7th EDITION
EMERGENCY MEDICINE CONCEPTS AND CLINICAL PRACTICE
Editor-in-Chief
John A. Marx, MD
Adjunct Professor of Emergency Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina Chair, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, North Carolina
Senior Editors
Robert S. Hockberger, MD
Professor of Clinical Medicine, David Geffen School of Medicine at UCLA, Los Angeles, California Chair, Department of Emergency Medicine, Harbor-UCLA Medical Center, Torrance, California
Ron M. Walls, MD
Professor of Medicine (Emergency Medicine), Harvard Medi cal School, Boston, Massachusetts Chairman, Department of Emergency Medicine, Brigham and Women’s Hospital, Boston, Massachusetts
Editors
James G. Adams, MD
Professor and Chair, Department of Emergency Medicine, Feinberg School of Medicine, Northwestern University, Chicago, Illinois Chair, Department of Emergency Medicine, Northwestern Memorial Hospital, Chicago, Illinois
William G. Barsan, MD
Professor and Chair, Department of Emergency Medicine, University of Michigan, Ann Arbor, Michigan
Volume 1
Michelle H. Biros, MD, MS
Professor, Emergency Medicine, University of Minnesota Medical School and Hennepin County Medical Center, Minneapolis, Minnesota Vice-Chair of Research, Emergency Medicine, University of Minnesota Medical School, Minneapolis, Minnesota Associate Research Director, Hennepin County Medical Center, Minneapolis, Minnesota
Daniel F. Danzl, MD
Professor and Chair, Department of Emergency Medicine, University of Louisville School of Medicine, Louisville, Kentucky
Marianne Gausche-Hill, MD
Professor of Clinical Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California Director of EMS and Pediatric Emergency Fellowships, Department of Emergency Medicine, Harbor-University of California at Los Angeles Medical Center, Los Angeles, California
Louis J. Ling, MD
Professor, Emergency Medicine and Pharmacy and Associate Dean for Graduate Medical Education, University of Minnesota Medical School, Minneapolis, Minnesota Associate Medical Director for Education, Hennepin County Medical Center; Senior Associate Medical Director, Hennepin Regional Poison Center, Minneapolis, Minnesota
Edward J. Newton, MD
Professor and Chairman of Emergency Medicine, Keck School of Medicine, Los Angeles, California Chair, Department of Emergency Medicine, LAC and USC Medical Center, Los Angeles, California
1600 John F. Kennedy Boulevard Suite 1800 Philadelphia, PA 19103-2899
ROSEN’S EMERGENCY MEDICINE: CONCEPTS AND CLINICAL PRACTICE, 7TH EDITION
ISBN-13: 978-0-323-05472-0 Volume 1 Part Number: 9996073467 Volume 2 Part Number: 9996073521
Copyright © 2010, 2006, 2002, 1998, 1992, 1988, 1983 by Mosby, an imprint of Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permissions may be sought directly from Elsevier’s Rights Department: phone: (+1) 215 239 3804 (US) or (+44) 1865 843830 (UK); fax: (+44) 1865 853333; e-mail:
[email protected]. You may also complete your request on-line via the Elsevier website at http://www.elsevier.com/permissions.
Notice Knowledge and best practice in this field are constantly changing. As new research and experience broaden our knowledge, changes in practice, treatment and drug therapy may become necessary or appropriate. Readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of the practitioner, relying on their own experience and knowledge of the patient, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the Editors and assumes any liability for any injury and/or damage to persons or property arising out or related to any use of the material contained in this book. The Publisher Library of Congress Cataloging-in-Publication Data Rosen’s emergency medicine : concepts and clinical practice / editor-in-chief, John A. Marx ; senior editors, Robert S. Hockberger, Ron M. Walls ; editors, James G. Adams ... [et al.].—7th ed. p. ; cm. Includes bibliographical references and index. ISBN 978-0-323-05472-0 1. Emergency medicine. I. Marx, John A. II. Hockberger, Robert S. III. Walls, Ron M. IV. Adams, James, 1962 May 8– V. Rosen, Peter, 1935– Emergency medicine. VI. Title: Emergency medicine. [DNLM: 1. Emergencies. 2. Emergency Medicine. WB 105 R815 2010] RC86.7.E5784 2010 616.02′5—dc22 2008037485
Acquisitions Editor: Stefanie Jewell-Thomas Developmental Editor: Dee Simpson Cover Design Direction: Steven Stave Interior Design: Steven Stave
Working together to grow libraries in developing countries Printed in China Last digit is the print number: 9 8 7 6 5 4 3 2
www.elsevier.com | www.bookaid.org | www.sabre.org
EMERITUS EDITORS
Peter Rosen Roger M. Barkin PAST EDITORS
Frank J. Baker II (Editions 1 and 2) G. Richard Braen (Editions 1, 2, and 3) Robert H. Dailey (Editions 1, 2, and 3) Jerris R. Hedges (Edition 3) Richard C. Levy (Editions 1, 2, and 3) Vincent Markovchick (Edition 4) Mark Smith (Edition 3) Glenn C. Hamilton (Editions 5 and 6)
We wish to dedicate this 7th edition to Peter Rosen, MD, the founder of this textbook and its Editor-in-Chief through the first four editions. Thirty-one years ago, in 1978, Peter came to the belief that an emergency medicine textbook should be written by emergency physicians, and 5 years later, in 1983, he was the first to carry out that mission when the inaugural edition of this textbook was published. While this eponymous work is significant, Peter is more widely acknowledged as one of the true fathers of academic emergency medicine. He continues to leave his imprimatur on numerous programs and innumerable medical students, residents, and faculty members throughout the country and indeed the world. Many of these individuals, like ourselves, have tried to emulate his unbending leadership, his passion for teaching, and his reverence for emergency medicine and the patients we have the great privilege to serve. This textbook represents just one of many exemplary accomplishments of one extraordinary man. We are indebted to Peter for his vision and his wisdom. We are unspeakably grateful for the honor of continuing this work, ever humbled by the responsibility of carrying his legacy forward, and beholden to him for helping craft our specialty into what it is today. John A. Marx Robert S. Hockberger Ron M. Walls
How this Medical Textbook Should Be Viewed by the Practicing Clinician and the Judicial System The editors and authors of this textbook strongly believe that the complex practice of medicine, the vagaries of human diseases, the unpredictability of pathologic conditions, and the functions, dysfunctions, and responses of the human body cannot be defined, explained, or rigidly categorized by any written document. Therefore, it is neither the purpose nor intent of our textbook to serve as an authoritative source on any medical condition, treatment plan, or clinical intervention, nor should our textbook be used to rigorously define a standard of care that should be practiced by all clinicians. Our written word provides the physician with a literature-referenced database, and a reasonable clinical guide, which is combined with practical suggestions from individual experienced practitioners. We offer a general reference source and clinical roadmap on a variety of conditions and procedures that may confront clinicians who are experienced in emergency medicine practice. This text cannot replace physician judgment; cannot describe every possible aberration, nuance, clinical scenario, or presentation; and cannot define rigid standards for clinical actions or procedures. Every medical encounter must be individualized, and every patient must be approached on a case-by-case basis. No complex medical interaction can possibly be reduced to the written word. The treatments, procedures, and medical conditions described in out textbook do not constitute the total expertise or knowledge base expected to be possessed by all clinicians. Finally, many of the described complications and adverse outcomes associated with implementing or withholding complex medical and surgical interventions may occur, even when every aspect of the intervention has been standard or performed correctly. The editors and authors of Rosen’s Emergency Medicine: Concepts and Clinical Practice, 7th Edition From Roberts JR, Marx JA: Position Statement: Use of Medical Textbooks in Malpractice Claims. Emerg Med News, Vol. XXXI, 2:3, February 2009.
Contributors Cynthia K. Aaron, MD
Professor, Emergency Medicine and Pediatrics, Wayne State University School of Medicine; Program Director, Medical Toxicology; Education Director, Regional Poison Center; Associate Medical Director, Regional Poison Center, Children’s Hospital of Michigan, Part of the Detroit Medical Center, Detroit, Michigan
Kumar Alagappan, MD
Associate Professor, Albert Einstein College of Medicine, Bronx, New York; Associate Chairman, Emergency Medicine, Long Island Jewish Medical Center, New Hyde Park, New York
James T. Amsterdam, DMD, MD, MMM
Professor Emeritus, Emergency Medicine; Faculty, Center for Bioethics and Humanities, University of Colorado School of Medicine; Attending Physician, Anschutz Medical Center, Aurora, Colorado
Adjunct Professor of Emergency Medicine, Drexel University College of Medicine, Philadelphia, Pennsylvania; Professor of Clinical Emergency Medicine, Pennsylvania State University College of Medicine, Hershey, Pennsylvania; Chair and Service Line Director, Department of Emergency Medicine, York Hospital/WellSpan Health, York, Pennsylvania
Riyad B. Abu-Laban, MD, MHSc, FRCPC
Christine Anderegg, MD
Jean T. Abbott, MD, MH
Associate Professor and Co-Research Director, Department of Emergency Medicine, University of British Columbia; Attending Physician, Department of Emergency Medicine, Vancouver General Hospital, Vancouver, British Columbia, Canada
Bruce D. Adams, MD, FACEP, Colonel, Medical Corps, U.S. Army
Clinical Professor of Emergency Medicine, Medical College of Georgia, Augusta, Georgia; Chief, Department of Clinical Investigations and Chief, Department of Emergency Medicine, William Beaumont Army Medical Center, El Paso, Texas
James G. Adams, MD
Professor and Chair, Department of Emergency Medicine, Feinberg School of Medicine, Northwestern University; Chair, Department of Emergency Medicine, Northwestern Memorial Hospital, Chicago, Illinois
Stephen L. Adams, MD
Professor and Chief, Division of Sports Medicine, Department of Medicine, Northwestern University, Feinberg School of Medicine; Medical Director, Emergency Preparedness/ Disaster Services, Northwestern Memorial Hospital; Team Physician, Chicago Cubs National League Baseball Club, Chicago, Illinois
Terry A. Adirim, MD, MPH
Associate Chief Medical Officer, Office of Health Affairs, U.S. Department of Homeland Security, Washington, DC; Attending Physician, Pediatric Emergency Department, Shady Grove Adventist Hospital, Rockville, Maryland
Attending Physician, Department of Emergency Medicine, Davis Hospital and Medical Center, Layton, Utah
Megan L. Anderson, MD
Department of Emergency Medicine, University of Michigan, Ann Arbor, Michigan
Deirdre Anglin, MD
Professor of Emergency Medicine, Keck School of Medicine, University of Southern California; Attending Physician, Los Angeles County and University of Southern California Medical Center, Los Angeles, California
Felix Ankel, MD
Associate Professor of Emergency Medicine, University of Minnesota, Minneapolis, Minnesota; Residency Director and Assistant Department Head, Emergency Medicine, Regions Hospital, St. Paul, Minnesota
Sanjay Arora, MD
Associate Professor of Clinical Emergency Medicine, University of Southern California, Keck School of Medicine, Los Angeles County Hospital, Los Angeles, California
Tom P. Aufderheide, MD, FACEP, FAHA
Professor of Emergency Medicine and Associate Chair of Research Affairs, Department of Emergency Medicine, Medical College of Wisconsin, Milwaukee, Wisconsin
Kevin M. Ban, MD
Assistant Clinical Professor, Harvard Medical School; Attending Physician, Beth Israel Deaconess Medical Center, Boston, Massachusetts vii
viii
Contributors
Emily Baran, MD
Assistant Professor, Department of Emergency Medicine, Feinberg School of Medicine, Northwestern University; Attending Physician, Northwestern Memorial Hospital, Chicago, Illinois
Christina E. Hantsch Bardsley, MD, FACEP, FACMT
Associate Professor, Department of Surgery, Division of Emergency Medicine, Stritch School of Medicine, Loyola University Chicago, Chicago, Illinois; Attending Physician, Emergency Medicine and Medical Toxicology, Loyola University Medical Center, Maywood, Illinois
Adam Z. Barkin, MD, MPH
Howard A. Bessen, MD
Professor of Medicine, David Geffen School of Medicine at University of California at Los Angeles; Senior Faculty Member, Department of Emergency Medicine, HarborUniversity of California at Los Angeles Medical Center, Torrance, California
Kriti Bhatia, MD
Clinical Instructor, Harvard Medical School; Attending Physician, Department of Emergency Medicine, Brigham and Women’s Hospital; Associate Residency Director, Harvard Affiliated Emergency Medicine Residency, Brigham and Women’s Hospital, Boston, Massachusetts
Clinical Instructor, Department of Surgery; Clinical Instructor, Department of Pediatrics, University of Colorado School of Medicine; Attending Physician, Rose Medical Center, Denver, Colorado
Elisabeth F. Bilden, MD
Andrew R. Barnosky, DO, MPH
Diane M. Birnbaumer, MD
Associate Professor, Emergency Medicine, University of Michigan Medical School; Associate Professor and Attending Physician, University of Michigan Health System, Department of Emergency Medicine, Ann Arbor, Michigan
William G. Barsan, MD
Professor and Chair, Department of Emergency Medicine, University of Michigan, Ann Arbor, Michigan
Bruce M. Becker, MD, MPH
Professor, Emergency Medicine and Community Health, Warren Alpert School of Medicine, Brown University; Attending Physician, Department of Emergency Medicine, Rhode Island Hospital and Hasbro Children’s Hospital, Providence, Rhode Island
Rimon N. Bengiamin, MD
Clinical Instructor, University of California San Francisco— Fresno, Fresno, California
Marc D. Berg, MD
Associate Professor of Clinical Pediatrics, University of Arizona, College of Medicine, Tucson, Arizona
Robert A. Berg, MD
Professor of Anesthesiology and Critical Care, University of Pennsylvania School of Medicine; Division Chief, Critical Care Medicine, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania
Carol D. Berkowitz, MD
Associate Medical Director, Hennepin County Medical Center, Minneapolis, Minnesota; Attending Physician, St. Mary’s Duluth Clinic, Duluth, Minnesota Professor of Clinical Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Associate Residency Program Director, Department of Emergency Medicine, Harbor-University of California at Los Angeles Medical Center, Torrance, California
Michelle H. Biros, MD, MS
Professor, Emergency Medicine, University of Minnesota Medical School and Hennepin County Medical Center; Vice Chair of Research-Emergency Medicine, University of Minnesota Medical School; Associate Research Director, Hennepin County Medical Center, Minneapolis, Minnesota
Robert A. Bitterman, MD, JD
President, Bitterman Health Law Consulting Group, Inc., Harbor Springs, Michigan; President, Emergency Physicians Insurance Company (EPIC), Auburn, California
Thomas H. Blackwell, MD
Clinical Associate Professor, School of Medicine, University of North Carolina—Chapel Hill, Chapel Hill, North Carolina; Medical Director, Center for Prehospital Medicine, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, North Carolina
Frederick C. Blum, MD, FACEP, FAAP, FIFEM
Associate Professor of Emergency Medicine and Pediatrics, West Virginia University School of Medicine, Morgantown, West Virginia
Professor of Clinical Pediatrics, David Geffen School of Medicine at the University of California at Los Angeles, Los Angeles, California; Executive Vice-Chair, Department of Pediatrics, Harbor-University of California at Los Angeles Medical Center, Torrance, California
Ira J. Blumen, MD
Edward Bernstein, MD
Jennifer M. Bocock, MD, FACEP
Professor and Vice-Chair for Academic Affairs, Department of Emergency Medicine, Boston University School of Medicine, Boston, Massachusetts
Judith Bernstein, PhD, RNC
Associate Professor, Department of Emergency Medicine, Boston University School of Medicine; Associate Professor, Department of Maternal and Child Health, Boston University School of Public Health, Boston, Massachusetts
Professor, Section of Emergency Medicine, Department of Medicine, University of Chicago; Program/Medical Director, University of Chicago Aeromedical Network (UCAN), University of Chicago Medical Center, Chicago, Illinois Attending Physician, Department of Emergency Medicine, Kettering Medical Center, Kettering, Ohio
Edward B. Bolgiano, MD, FACP, FACEP
Assistant Professor, Departments of Medicine and Surgery, University of Maryland School of Medicine; Chief, Department of Emergency Medicine, Bon Secours Hospital, Baltimore, Maryland
Laura J. Bontempo, MD
William J. Brady, MD
Professor of Emergency Medicine and Medicine, Vice-Chair, Department of Emergency Medicine, University of Virginia School of Medicine; Medical Director, Mondial Assistance USA and Canada, Charlottesville, Virginia
Assistant Professor of Clinical Medicine, Columbia College of Physicians and Surgeons, New York, New York; Adjunct Assistant Professor of Emergency Medicine, Warren Alpert School of Medicine, Brown University, Providence, Rhode Island; Visiting Senior Lecturer in International Health and Tropical Medicine, The Royal College of Surgeons, Dublin, Ireland; Senior Attending Physician in Emergency Medicine and Infectious Diseases, Global Health Fellowship Director, St. Luke’s Roosevelt Hospital Center, New York, New York
Sabina Braithwaite, MD
Kirsten K. Calder, MD, FACEP
Calvin A. Brown III, MD
Richard M. Cantor, MD, FAAP, FACEP
Associate Professor of Emergency Medicine, University of Virginia, Charlottesville, Virginia Instructor in Medicine (Emergency Medicine), Harvard Medical School; Attending Physician, Brigham and Women’s Hospital, Boston, Massachusetts
James E. Brown, MD
Program Director and Vice-Chair, Department of Emergency Medicine, Wright State University, Dayton, Ohio
Douglas D. Brunette, MD, MPH
Associate Professor, University of Minnesota Medical School, Department of Emergency Medicine; Assistant Chief, Hennepin County Medical Center, Department of Emergency Medicine, Minneapolis, Minnesota
Gavin R. Budhram, MD, RDMS
Assistant Professor of Emergency Medicine, Tufts University School of Medicine, Western Campus; Staff Physician, Director of Emergency Ultrasound, Baystate Medical Center, Springfield, Massachusetts
E. Bradshaw Bunney, MD
Associate Professor, Residency Director, University of Illinois at Chicago; Attending Physician, Department of Emergency Medicine, University of Illinois Hospital, Chicago, Illinois
David Burbulys, MD
Associate Professor of Clinical Medicine, David Geffen School of Medicine at University of California at Los Angeles; Director, Residency Program, Department of Emergency Medicine, Harbor-University of California at Los Angeles Medical Center, Torrance, California
Michael J. Burns, MD, FACEP, FACP
Clinical Professor, Departments of Emergency Medicine and Medicine, Division of Infectious Diseases, University of California, Irvine School of Medicine, Irvine, California; Attending Physician, Emergency Medicine and Infectious Diseases, University of California, Irvine Medical Center, Orange, California
Richard L. Byyny, MD, MSc
Assistant Professor, Division of Surgery, University of Colorado, School of Medicine, Aurora, Colorado; Associate Director of Research, Denver Health Medical Center Residency in Emergency Medicine, Denver Health Medical Center, Denver, Colorado
Staff Physician, Department of Emergency Medicine, Los Alamitos Medical Center, Los Alamitos, California Associate Professor and Director, Pediatric Emergency Medicine, Department of Emergency Medicine; Medical Director, Central New York Regional Poison Control Center, State University of New York, Upstate Medical University College of Medicine, Syracuse, New York
Stuart M. Caplen, MD
Lean Project Coordinator, Emergency Department, Montefiore North Division, Bronx, New York; Attending Physician, Emergency Department, Metropolitan Hospital Center, New York, New York
Andrea Carlson, MD
Attending Physician, Emergency Medicine, Director, Medical Toxicology, Advocate Christ Hospital, Oak Lawn, Illinois
Theodore C. Chan, MD
Professor of Clinical Medicine, University of California at San Diego; Medical Director, Emergency Department, University of California, San Diego Medical Center, San Diego, California
Lei Chen, MD
Assistant Professor, Section of Pediatric Emergency Medicine, Department of Pediatrics, Yale University School of Medicine; Attending Physician, Yale-New Haven Children’s Hospital, New Haven, Connecticut
Stephen B. Choi, MD, FRCPC
Associate Residency Director, Department of Emergency Medicine, University of Ottawa; Assistant Professor, University of Ottawa; Co-Editor-in Chief, Open Medicine, Ottawa, Ontario, Canada
Richard F. Clark, MD
Professor of Medicine, University of California at San Diego; Director, Division of Medical Toxicology, University of California at San Diego Medical Center, San Diego, California
Philip A. Clement, MD, FACEP
Clinical Assistant Professor, Department of Emergency Medicine, East Carolina University, Brody School of Medicine; Attending Physician, Pitt County Memorial Hospital, Greenville, North Carolina
ix
Contributors
Assistant Professor and Residency Program Director, Section of Emergency Medicine, Department of Surgery, Yale University School of Medicine, New Haven, Connecticut
John D. Cahill, MD
Contributors
x
Wendy C. Coates, MD
Professor of Medicine, Chair, Acute Care College, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, California; Director, Medical Education, Harbor-University of California at Los Angeles Medical Center, Department of Emergency Medicine, Huntington Beach, California
Robert E. Collier, MD
Assistant Professor of Emergency Medicine, University of Minnesota School of Medicine; Emergency Medicine Faculty, Hyperbaric Medicine Fellowship Director, Hennepin County Medical Center, Minneapolis, Minnesota
Jamie L. Collings, MD
Associate Professor, Department of Emergency Medicine, Northwestern University, Feinberg School of Medicine; Residency Director, Department of Emergency Medicine, Northwestern Memorial Hospital, Chicago, Illinois
Stephen A. Colucciello, MD, FACEP
Adjunct Professor of Emergency Medicine, University of North Carolina Medical School—Chapel Hill, Chapel Hill, North Carolina; Vice Chief Emergency Medicine, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, North Carolina
Christopher B. Colwell, MD
Associate Professor, Department of Surgery, Division of Emergency Medicine, University of Colorado at Denver, School of Medicine; Associate Director, Department of Emergency Medicine, Denver Health Medical Center, Denver, Colorado
Edward E. Conway, Jr., MD, MS
Professor of Clinical Pediatrics, Albert Einstein College of Medicine, Bronx, New York; Chairman, Milton and Bernice Stern Department of Pediatrics, Chief of Pediatric Critical Care Medicine, Beth Israel Medical Center, New York, New York
Jeremy L. Cooke, MD
Assistant Professor, Department of Emergency Medicine, University of California at Davis; Assistant Professor of Emergency Medicine, University of California at Davis Medical Center, Sacramento, California
Mary Ann Cooper, MD
Professor (Retired), Departments of Bioengineering and Emergency Medicine, University of Illinois at Chicago, Chicago, Illinois
Randolph J. Cordle, MD
Adjunct Assistant Professor, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina; Medical Director, Division of Pediatric Emergency Medicine; Program Director, Pediatric Emergency Medicine Fellowship, Levine Children’s Hospital, Department of Emergency Medicine, Charlotte, North Carolina
Sandy A. Craig, MD
Adjunct Associate Professor, Department of Emergency Medicine, University of North Carolina School of Medicine, Chapel Hill, North Carolina; Faculty, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, North Carolina
Hilarie Cranmer, MD, MPH
Assistant Professor, Harvard Medical School; Attending Emergency Medicine Director, Global Women’s Health Fellowship, Education Director, Harvard Humanitarian Initiative, Brigham and Women’s Hospital, Boston, Massachusetts
Todd J. Crocco, MD
Associate Professor and Chair, Department of Emergency Medicine, West Virginia University School of Medicine, Morgantown, West Virginia
Pat Croskerry, MD, PhD
Senior Research Scientist, Dalhousie University, Halifax, Nova Scotia, Canada; Attending Physician, Dartmouth General Hospital, Dartmouth, Nova Scotia, Canada
A. Adam Cwinn, MD, FRCPC
Professor, Department of Emergency Medicine, The University of Ottawa; Head, Department of Emergency Medicine and Medical Director of Critical Care and Emergency Medicine, The Ottawa Hospital, Ottawa, Ontario, Canada
Rita K. Cydulka, MD, MS
Professor, Emergency Medicine, Case Western Reserve University School of Medicine, Cleveland, Ohio; Vice Chair, MetroHealth Medical Center, Shaker Heights, Ohio
Daniel F. Danzl, MD
Professor and Chair, Department of Emergency Medicine, University of Louisville School of Medicine, Louisville, Kentucky
Ana M. Davitt, MD
Attending Physician, Pennsylvania Hospital, University of Pennsylvania Health System, Philadelphia, Pennsylvania
Mohamud Daya, MD
Associate Professor, Department of Emergency Medicine, Oregon Health and Science University, Portland, Oregon
Kathleen A. Delaney, MD, MS
Professor, Division of Emergency Medicine, University of Texas, Southwestern Medical School; Vice Chair of Emergency Medicine, Parkland Memorial Hospital, Dallas, Texas
Theodore R. Delbridge, MD, MPH
Professor of Emergency Medicine, Brody School of Medicine at East Carolina University; Chief of Emergency Services, Department of Emergency Medicine, Pitt County Memorial Hospital, Greenville, North Carolina
Robert A. De Lorenzo, MD, MSM
Professor of Military and Emergency Medicine, Uniformed Services University of the Health Sciences, Bethesda, Maryland; Colonel Medical Corps, U.S. Army, Brooke Army Medical Center, Fort Sam Houston, Houston, Texas
Robert W. Derlet, MD
Professor Emeritus, Department of Emergency Medicine, University of California Davis School of Medicine, Sacramento, California
Shoma Desai, MD, BA
Assistant Professor of Clinical Emergency Medicine, University of Southern California; Quality Assurance Director, Los Angeles County and University of Southern California Medical Center, Los Angeles, California
Bram A. Dolcourt, MD
Evelyn H. Duvivier, MD, MPH
Attending Physician, Pennsylvania Hospital, Philadelphia, Pennsylvania
Joshua S. Easter, MD
Clinical Fellow, Harvard Medical School; Clinical Pediatric Emergency Medicine Fellow, Department of Emergency Medicine, Children’s Hospital of Boston, Boston, Massachusetts
Marc Eckstein, MD, MPH
Associate Professor of Emergency Medicine, Keck School of Medicine of the University of Southern California; Medical Director, Los Angeles Fire Department; Director of Prehospital Care–Los Angeles County/University of Southern California Medical Center, Los Angeles, California
Mary Eisenhauer, MD, FRCPC
Associate Professor of Medicine, Schulich School of Medicine and Dentistry, University of Western Ontario; Consultant, London Health Sciences Centre, London, Ontario, Canada
Matt Emery, MD, FACEP
Assistant Professor of Emergency Medicine, Michigan State University-CHM, East Lansing, Michigan; Educational Assistant, MSU-MERC Program in Emergency Medicine, Spectrum Health, Butterworth Campus, Grand Rapids, Michigan
Jay L. Falk, MD, FACEP, FCCM
Professor of Medicine and Emergency Medicine, University of Central Florida, College of Medicine; Clinical Professor, Clinical Sciences, Florida State University, College of Medicine; Academic Chairman, Department of Emergency Medicine, Orlando Regional Medical Center; Vice President of Medical Education, Orlando Health, Orlando, Florida
Sing-Yi Feng, MD
Assistant Professor, Department of Pediatrics, Division of Emergency Medicine, University of Texas Southwestern Medical Center at Dallas; Medical Toxicologist, North Texas Poison Center, Parkland Memorial Hospital, Dallas, Texas
Madonna Fernández-Frackelton, MD
Associate Professor of Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Associate Residency Director, HarborUniversity of California at Los Angeles Medical Center, Torrance, California
James F. Fiechtl, MD
Assistant Professor, Department of Emergency Medicine, Vanderbilt University Medical Center, Nashville, Tennessee
John T. Finnell, II, MD, MSc
Associate Professor of Emergency Medicine, Indiana University; Research Scientist, Regenstrief Institute, Indianapolis, Indiana
Assistant Professor, Department of Emergency Medicine; Assistant Professor, Department of Orthopedics and Rehabilitation, Vanderbilt University Medical Center, Nashville, Tennessee
Mark Foran, MD
Clinical Fellow in Emergency Medicine, Harvard Medical School; Resident Physician, Harvard Affiliated Emergency Medicine Residency, Brigham and Women’s Hospital, Massachusetts General Hospital, Boston, Massachusetts
E. John Gallagher, MD
Professor and University Chair, Department of Emergency Medicine, Albert Einstein College of Medicine of Yeshiva University; Chief of Service, Emergency Medicine, Montefiore Medical Center, Bronx, New York
Boris Garber, DO
Assistant Professor, Case Western Reserve University School of Medicine; Attending Physician, MetroHealth Medical Center, Cleveland, Ohio
Marianne Gausche-Hill, MD, FACEP, FAAP
Professor of Clinical Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Director of EMS and Pediatric Emergency Fellowships, Department of Emergency Medicine, HarborUniversity of California at Los Angeles Medical Center, Torrance, California
Mark E. Gebhart, MD, FAAEM
Assistant Professor of Emergency Medicine, Wright State University School of Medicine; Staff Physician, Emergency and Trauma Center, Good Samaritan Hospital, Dayton, Ohio
Joel M. Geiderman, MD, FACEP
Professor of Emergency Medicine, Cedars-Sinai Medical Center; Professor of Medicine, David Geffen School of Medicine at University of California at Los Angeles; Co-Chairman, Department of Emergency Medicine, Cedars-Sinai Medical Center, Los Angeles, California
Michael A. Gibbs, MD, FACEP
Professor of Emergency Medicine, Tufts University School of Medicine, Boston, Massachusetts; Chief, Department of Emergency Medicine, Maine Medical Center, Portland, Maine
Casey M. Glass, MD
Assistant Professor, Department of Emergency Medicine, Wake Forest University Health Sciences; Director of Community Emergency Ultrasound Programs, Wake Forest University Health Sciences Department of Emergency Medicine; Assistant Medical Director, Emergency Medicine, Wilkes Regional Medical Center, North Wilkesboro, North Carolina; North Carolinas Baptist Medical Center, Winston-Salem, North Carolina
Richard Goldberg, MD
Clinical Professor of Emergency Medicine, Department of Emergency Medicine, Los Angeles County and University of Southern California Medical Center, Los Angeles, California; Staff Physician, Providence Saint Joseph Medical Center, Burbank, California
xi
Contributors
Assistant Professor, Wayne State University School of Medicine; Medical Toxicologist, Children’s Hospital of Michigan Regional Poison Control Center, Detroit, Michigan
Robert W. Fitch, MD
xii
Contributors
John E. Gough, MD
Professor, Department of Emergency Medicine, East Carolina University, Brody School of Medicine; Attending Physician, Pitt County Memorial Hospital, Greenville, North Carolina
Louis Graff IV, MD, FACP, FACEP
William G. Heegaard, MD, MPH
Associate Professor, University of Minnesota Medical School, Department of Emergency Medicine; Assistant Chief, Hennepin County Medical Center, Department of Emergency Medicine, Minneapolis, Minnesota
Professor of Emergency Medicine, Professor of Clinical Medicine, University of Connecticut School of Medicine, Farmington, Connecticut; Medical Director of Quality, Associate Director of Emergency Medicine, Hospital of Central Connecticut, New Britain, Connecticut
Jag S. Heer, MD, FAAEM
Richard O. Gray, MD
Katherine L. Heilpern, MD
Assistant Professor of Emergency Medicine, University of Minnesota Medical School; Department of Emergency Medicine, Hennepin County Medical Center, Minneapolis, Minnesota
Eric Gross, MD
Assistant Professor, Department of Emergency Medicine, University of Minnesota Medical School; Assistant Residency Director, Department of Emergency Medicine, Hennepin County Medical Center, Minneapolis, Minnesota
John A. Guisto, MD
Associate Professor, Department of Emergency Medicine, University of Arizona College of Medicine; Medical Director, Emergency Department, University Medical Center, Tucson, Arizona
Assistant Clinical Professor, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Attending Faculty Department of Emergency Medicine, Kern Medical Center, Bakersfield, California Professor and Chair, Department of Emergency Medicine, Emory University School of Medicine, Atlanta, Georgia
Robin R. Hemphill, MD, MPH
Associate Professor, Department of Emergency Medicine, Emory University School of Medicine; Director of Patient Safety and Quality, Emory University Hospital, Atlanta, Georgia
Sean O. Henderson, MD
Associate Professor of Emergency and Preventive Medicine, Keck School of Medicine of the University of California; Vice Chair, Department of Emergency Medicine LAC and USC Medical Center, Los Angeles, California
Robert G. Hendrickson, MD
Professor and Chair, University of California at San Diego, Department of Emergency Medicine, University of California San Diego School of Medicine, San Diego, California
Associate Professor, Department of Emergency Medicine, Oregon Health and Science University; Associate Medical Director, Medical Toxicologist, Oregon Poison Center; Associate Fellowship Director, Program in Medical Toxicology, Oregon Health and Science University, Portland, Oregon
Leon Gussow, MD
Philip L. Henneman, MD
Rania Habal, MD
H. Gene Hern, Jr., MD
Tenagne Haile-Mariam, MD
Kendall Ho, MD, FRCPC
David A. Guss, MD
Lecturer, Department of Emergency Medicine, University of Illinois; Instructor, Department of Emergency Medicine, Rush Medical College; Attending Physician, John H. Stroger, Jr. Hospital of Cook County, Chicago, Illinois Assistant Clinical Professor, Emergency Medicine, New York Medical College, Valhalla, New York; Attending Physician, Emergency Medicine, Metropolitan Hospital Center, New York, New York Assistant Professor, Department of Emergency Medicine, George Washington University Medical Center, Washington, DC
Glenn C. Hamilton, MD
Professor and Chair, Department of Emergency Medicine, Wright State University, Dayton, Ohio
Stephen W. Hargarten, MD, MPH
Professor, Department of Emergency Medicine, Medical College of Wisconsin; Director, Emergency Medicine, Froedtert Hospital, Milwaukee, Wisconsin
Richard A. Harrigan, MD
Professor of Emergency Medicine, Temple University School of Medicine, Temple University, Philadelphia, Pennsylvania
Professor of Emergency Medicine, Tufts University School of Medicine, Boston, Massachusetts; Attending Physician, Department of Emergency Medicine, Baystate Medical Center, Springfield, Massachusetts Assistant Clinical Professor of Emergency Medicine, University of California at San Francisco, San Francisco, California; Residency Director, Alameda County Medical Center, Oakland, California Associate Professor, Department of Emergency Medicine, Faculty of Medicine, University of British Columbia; Attending Staff, Department of Emergency Medicine, Vancouver General Hospital, Vancouver, British Columbia, Canada
Robert S. Hockberger, MD
Professor of Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Chair, Department of Emergency Medicine, Harbor-University of California at Los Angeles Medical Center, Torrance, California
Robert S. Hoffman, MD
Associate Professor of Emergency Medicine and Medicine (Clinical Pharmacology), New York University School of Medicine; Attending Physician, Bellevue Hospital Center, New York, New York
Benjamin Honigman, MD
Timothy Horeczko, MD
Clinical Instructor of Medicine, David Geffen School of Medicine at the University of California, Los Angeles, Los Angeles, California; Pediatric Emergency Medicine Fellow, Department of Emergency Medicine, Harbor-University of California at Los Angeles Medical Center, Los Angeles County Harbor-University of California at Los Angeles Medical Center, Torrance, California
Mark A. Hostetler, MD, MPH
Clinical Professor, Departments of Pediatrics and Emergency Medicine, The University of Arizona College of Medicine; Attending Physician, Phoenix Children’s Hospital, Phoenix, Arizona
Debra E. Houry, MD, MPH
Professor and Chair, Mt. Sinai School of Medicine; Medical Director, Mt. Sinai Medical Center, New York, New York
Thea L. James, MD
Assistant Professor of Emergency Medicine, Department of Emergency Medicine, Boston University School of Medicine, Boston Medical Center, Boston, Massachusetts
Timothy G. Janz, MD
Professor, Department of Emergency Medicine, Department of Internal Medicine, Boonshoft School of Medicine, Wright State University, Dayton, Ohio
Alan Jones, MD
Adjunct Assistant Professor of Emergency Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina; Director, Emergency Medicine Critical Care Services; Assistant Director, Emergency Medicine Research, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, North Carolina
Assistant Professor, Department of Emergency Medicine, Emory School of Medicine; Assistant Professor, Department of Environmental and Occupational Health and Department of Behavioral Sciences and Health Education, Rollins School of Public Health; Director, Center for Injury Control, Emory University, Atlanta; Attending Emergency Physician, Emory University Hospital, Atlanta, Georgia
James B. Jones, PharmD, MD
J. Stephen Huff, MD
Nicholas J. Jouriles, MD
Associate Professor of Emergency Medicine and Neurology, Department of Emergency Medicine, University of Virginia Health System, Charlottesville, Virginia
Oliver Hung, MD
Assistant Clinical Professor of Emergency Medicine, Mt. Sinai School of Medicine, New York, New York; Attending Physician, Department of Emergency Medicine, Morristown Memorial Hospital, Morristown, New Jersey
H. Range Hutson, MD
Assistant Professor, Department of Emergency Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts
Alson S. Inaba, MD, FAAP
Associate Professor of Pediatrics, University of Hawaii, John A. Burns School of Medicine; Director and Attending Physician, Pediatric Emergency Medicine Center, Kapi’olani Medical Center for Women and Children, Honolulu, Hawaii
Jennifer L. Isenhour, MD
Adjunct Assistant Professor, Department of Emergency Medicine, University of North Carolina—Chapel Hill, Chapel Hill, North Carolina; Associate Program Director, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, North Carolina
Kenneth V. Iserson, MD, MBA, FACEP
Professor Emeritus, Department of Emergency Medicine, University of Arizona College of Medicine, Tucson, Arizona
Kenneth Jackimczyk, MD, FACEP
Attending Physician, Maricopa Medical Center, Phoenix, Arizona
xiii
Staff Physician, Mercy Hospital, Scranton, Pennsylvania
Jonathan S. Jones, MD
Assistant Professor and Assistant Program Direcotr, Department of Emergency Medicine, University of Mississippi Medical Center, Jackson, Mississippi Professor, Emergency Medicine, Northeastern Ohio Universities College of Medicine and Pharmacy, Rootstown, Ohio; Emergency Medicine Resident Care Faculty, Akron General Medical Center, Akron, Ohio
Amy H. Kaji, MD, PhD
Assistant Clinical Professor of Emergency Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Assistant Clinical Professor of Emergency Medicine, Medical Director, Disaster Resource Center, Harbor-University of California at Los Angeles Medical Center, Torrance, California
Norman Kalbfleisch, MD
Associate Professor, Oregon Health and Science University, Portland, Oregon
Louise Kao, MD
Director, Medical Toxicology Fellowship Program; Assistant Professor of Clinical Emergency Medicine, Indiana University School of Medicine; Methodist Hospital/Clarian Health Partners, Indianapolis, Indiana
Dan Katz, MD, DTMH
Assistant Clinical Professor of Medicine, David Geffen School of Medicine at University of California at Los Angeles; Medical Director of Academic Affairs, Department of Emergency Medicine, Cedars-Sinai Medical Center, Los Angeles, California
Matthew T. Keadey, MD, FACEP
Assistant Professor, Emory University School of Medicine; Chief of Service, Department of Emergency Medicine, Emory University Hospital, Atlanta, Georgia
Contributors
Professor of Surgery, University of Colorado Denver, School of Medicine; Head, Division of Emergency Medicine, Department of Emergency Medicine, University of Colorado Hospital, Aurora, Colorado
Andy Jagoda, MD, FACEP
xiv
Contributors
Eugene E. Kercher, MD, FACEP, FAPA
Chief Medical Officer, Director of Graduate Medical Education, Kern Medical Center, Bakersfield, California; Associate Clinical Professor of Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California
Kianusch Kiai, MD, MS
Clinical Associate Professor of Anesthesiology, Department of Anesthesiology, David Geffen School of Medicine at University of California at Los Angeles; Attending Physician, University of California at Los Angeles Ronald Reagan Medical Center, Los Angeles, California
Kelly E. King, MD
Medical Director, Casualty Care Research Center, Assistant Professor of Military and Emergency Medicine, Uniformed Services University of the Health Sciences, Bethesda, Maryland
Susan Kirelik, MD
Medical Director, Pediatric Emergency Services; Chair, Department of Pediatrics, Sky Ridge Medical Center, Lone Tree, Colorado
Eileen J. Klein, MD, MPH
Associate Professor, Pediatrics, University of Washington; Attending Physician, Seattle Children’s Hospital, Seattle, Washington
Jeffrey A. Kline, MD
Adjunct Professor of Emergency Medicine, University of North Carolina at Chapel Hill, Charlotte, North Carolina; Professor of Emergency Medicine, University of North Carolina—Chapel Hill; Director of Research, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, North Carolina
Andrew L. Knaut, MD, PhD
Attending Physician, Emergency Physicians at Porter Hospitals, Denver, Colorado
Kristi L. Koenig, MD, FACEP
Professor of Emergency Medicine, Co-Director, EMS and Disaster Medical Sciences Fellowship, University of California at Irvine, School of Medicine; Director of Public Preparedness, University of California at Irvine, Orange, California
Amy V. Kontrick, MD
Assistant Professor of Emergency Medicine; Director, Undergraduate Medical Education, Northwestern University Feinberg School of Medicine, Chicago, Illinois
Dina Halpern Kornblau, MD, BA
Rashmi U. Kothari, MD
Associate Professor, Michigan State University/Kalamazoo Center for Medical Studies (MSU/KCMS); Director of Emergency Medicine Research, Borgess Research Institute, Borgess Hospital, Kalamazoo, Michigan
Baruch Krauss, MD, EdM
Associate Professor of Pediatrics, Department of Pediatrics, Harvard Medical School; Senior Associate Physician in Medicine, Division of Emergency Medicine, Children’s Hospital, Boston, Massachusetts
Ken Kulig, MD, FACMT, FAACT
Clinical Associate Professor, Emergency Medicine, University of Colorado; President Elect, Medical Staff, Porter Adventist Hospital, Denver, Colorado
Thomas Kwiatkowski, MD
Professor of Clinical Emergency Medicine, Albert Einstein College of Medicine, Bronx, New York; Medical Director, Patient Safety Institute; Faculty, Emergency Medicine, North Shore-Long Island Jewish Hospital Health System, Lake Success, New York
Frank W. Lavoie, MD
Vice President of Medical Affairs, Southern Maine Medical Center, Biddeford, Maine
Eric J. Lavonas, MD, FACEP, FACMT
Assistant Professor of Surgery (Emergency Medicine), University of Colorado, Denver, School of Medicine, Aurora, Colorado; Emergency Physician, Denver Health Medical Center; Associate Director, Rocky Mountain Poison and Drug Center, Denver Health Medical Center, Denver, Colorado
Christopher C. Lee, MD
Assistant Professor and Director of International Emergency Medicine Center, Stony Brook University, Stony Brook, New York
David C. Lee, MD
Clinical Associate Professor, New York University School of Medicine, New York, New York; Director of Research, Department of Emergency Medicine, North Shore University Hospital, Manhasset, New York
Jill F. Lehrmann, MD, MPH
Assistant Professor, Northwestern University Feinberg School of Medicine; Attending Physician, Northwestern Memorial Hospital, Chicago, Illinois
E. Brooke Lerner, PhD
Associate Professor, Medical College of Wisconsin, Milwaukee, Wisconsin
Assistant Professor, Albert Einstein College of Medicine; Attending Physician and Director, Division of Pediatric Neurology, St. Barnabas Hospital, Bronx, New York
Michael D. Levine, MD
Joshua M. Kosowsky, MD
Roger J. Lewis, MD, PhD
Assistant Professor, Harvard Medical School; Clinical Director, Department of Emergency Medicine, Brigham and Women’s Hospital, Boston, Massachusetts
Department of Medical Toxicology, Banner Good Samaritan Hospital Medical Center, Phoenix, Arizona Professor, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Vice Chair, Academic Affairs, Department of Emergency Medicine, Harbor-University of California at Los Angeles Medical Center, Torrance, California
Michelle Lin, MD
Louis J. Ling, MD
xv
Assistant Professor of Clinical Emergency Medicine, University of Southern California, Keck School of Medicine; Assistant Residency Director, Residency in Emergency Medicine, Los Angeles County and University of Southern California Medical Center, Los Angeles, California
Professor, Emergency Medicine and Pharmacy and Associate Dean for Graduate Medical Education, University of Minnesota Medical School; Associate Medical Director for Education, Hennepin County Medical Center; Senior Associate Medical Director, Hennepin Regional Poison Center, Minneapolis, Minnesota
William K. Mallon, MD, FACEP
Ari M. Lipsky, MD, MS
Assistant Professor, Department of Emergency Medicine, Case Western Reserve University; Attending Director of Medical Toxicology, Department of Emergency Medicine, MetroHealth Medical Center, Cleveland, Ohio
Assistant Professor, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Attending Physician, Department of Emergency Medicine, Harbor-University of California at Los Angeles Medical Center, Torrance, California
Eve D. Losman, MD
Assistant Professor, Associate Program Director, Department of Emergency Medicine, University of Michigan Medical School, University of Michigan Health System, Ann Arbor, Michigan
Mark J. Lowell, MD
Associate Professor of Emergency Medicine, University of Michigan Medical School; Medical Director, Survival Flight, University of Michigan Health System, Ann Arbor, Michigan
Douglas W. Lowery III, MD
Associate Professor of Clinical Emergency Medicine, Keck School of Medicine at the University of Southern California; Director, Division of International Emergency Medicine, LAC and USC Medical Center, Los Angeles, California
Gerald E. Maloney, Jr., DO
Diku P. Mandavia, MD, FACEP, FRCPC
Clinical Associate Professor of Emergency Medicine, Keck School of Medicine, University of California at Los Angeles; Attending Staff Physician, Department of Emergency Medicine, Cedars-Sinai Medical Center, Los Angeles, California
Mariann Manno, MD
Associate Professor, Clinical Pediatric and Emergency Medicine, University of Massachusetts Medical School; Division Director, Pediatric Emergency Services; Director, Pediatric Emergency Department and PediPlace, Children’s Medical Center, University of Massachusetts Memorial Hospital, Worcester, Massachusetts
Associate Professor of Emergency Medicine, Emory University School of Medicine; Vice Chair of Clinical Operations, Department of Emergency Medicine, Emory Healthcare, Atlanta, Georgia
Catherine A. Marco, MD, FACEP
Binh T. Ly, MD, FACEP, FACMT
Vincent Markovchick, MD
Associate Professor, University of California, San Diego; Director, Emergency Medicine Residency; Director, Medical Toxicology Fellowship, Division of Medical Toxicology, University of California at San Diego Medical Center, San Diego, California
Everett T. Lyn, MD, MSc
Assistant Professor, Harvard Medical School, Boston, Massachusetts; Chairman, Department of Emergency Medicine, North Shore Medical Center, Salem, Massachusetts
Malcolm Mahadevan, MD, MBBS (Singapore), MRCP (UK), FRCSEd (A&E), FAMS
Senior Clinical Lecturer, Yong Loo Lin School of Medicine, National University of Singapore; Clinical Director and Senior Consultant, Emergency Department, National University Hospital, Singapore
Brian D. Mahoney, MD
Associate Professor, Department of Emergency Medicine, University of Minnesota; Medical Director, Emergency Medical Services, Hennepin County Medical Center, Minneapolis, Minnesota
Professor, Department of Emergency Medicine; Director of Medical Ethics Curriculum, University of Toledo College of Medicine, Toledo, Ohio Professor of Surgery, Division of Emergency Medicine, University of Colorado at Denver School of Medicine; Director, Department of Emergency Medicine, Denver Health Medical Center, Denver, Colorado
Marc L. Martel, MD
Associate Professor, University of Minnesota; Program Director, Emergency Medicine; Co-Program Director, Emergency Medicine/Internal Medicine, Department of Emergency Medicine, Hennepin County Medical Center, Minneapolis, Minnesota
John A. Marx, MD
Adjunct Professor of Emergency Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina; Chair, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, North Carolina
Ryanne J. Mayersak, MD, MS
Assistant Professor of Emergency Medicine, The George Washington University, Washington, DC
Suzan S. Mazor, MD
Assistant Professor, Pediatrics, University of Washington; Attending Physician, Seattle Children’s Hospital, Seattle, Washington
Contributors
Associate Clinical Professor of Emergency Medicine, University of California at San Francisco; San Francisco General Hospital, Department of Emergency Medicine, San Francisco, California
Thomas Mailhot, MD
xvi
Contributors
Maureen McCollough, MD, FACEP, FAAEM
Associate Professor of Clinical Emergency Medicine and Pediatrics, Keck School of Medicine of USC; Medical Director, Emergency Department, Los Angeles County University of Southern California Medical Center, Los Angeles, California
Mary Pat McKay, MD, MPH
Connie Mitchell, MD, MPH
Assistant Clinical Professor, Department of Internal Medicine, School of Medicine, University of California at Davis, Davis, California; Policy Development, Maternal, Child, and Adolescent Health, California Department of Public Health, Sacramento, California
Associate Professor of Emergency Medicine and Public Health, The George Washington University; Director, Center for Injury Prevention and Control, The George Washington University, Washington, DC
Gregory P. Moore, MD, JD
L. Kendall McKenzie, MD
Professor of Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Department of Emergency Medicine and Division of Infectious Diseases, Olive View–University of California at Los Angeles Medical Center, Sylmar, California
Assistant Professor of Emergency Medicine, The University of Mississippi School of Medicine, Jackson, Mississippi
Nathanael J. McKeown, DO
Assistant Professor, Oregon Health and Science University; Attending Physician, Portland Veteran Affairs Medical Center; Oregon Health and Science University, Portland, Oregon
John McManus, MD, MCR, FACEP, FAAEM
Director, Center for Pre-Deployment Medicine, U.S. Army Medical Department Center and School, Fort Sam Houston; EMS Fellowship Program Director, San Antonio Uniformed Services Health Education Consortium; Medical Director, Fort Sam Houston and Camp Bullis Fire Department; Clinical Associate Professor, Emergency Medicine, University of Texas Health Science Center, San Antonio, Texas
Attending Physician, Emergency Medicine Residency, Madigan Army Medical Center, Tacoma, Washington
Gregory J. Moran, MD, FACEP, FIDSA
Laurie J. Morrison, MD, MSc, FRCPC
Professor of Emergency Medicine, Department of Medicine, University of Toronto; Director, Clinician Scientist, Keenan Research Centre, Li Ka Shing Knowledge Institute, St. Michael’s Hospital, Toronto, Ontario, Canada
Robert L. Muelleman, MD, FACEP
Chief of Emergency Medicine, University of Nebraska Medical Center, Omaha, Nebraska
Lindsay Murray, MD, MBBS, FACEM
Regional Medical Director, TEAM Health, Southeast, Emergency Services, The Medical Center, Columbus, Georgia
Clinical Associate Professor, University of Western Australia, Perth, Western Australia; Consultant Emergency Physician and Clinical Toxicologist, Sir Charles Gairdner Hospital, Perth, Western Australia
Kemedy K. McQuillen, MD
Michael F. Murphy, MD, FRCPC
David B. McMicken, MD, FACEP
Attending Physician, Central Maine Medical Center, Lewiston, Maine
Harvey W. Meislin, MD
Professor of Emergency Medicine, The University of Arizona College of Medicine; Department Head of Emergency Medicine, University Medical Center; Director, Arizona Emergency Medicine Research Center, Tucson, Arizona
Professor and Chair, Department of Anesthesia; Professor of Emergency Medicine, Dalhousie University; Chief, Department of Anesthesia, Capital Health District Health Authority, Halifax, Nova Scotia, Canada
Vinay M. Nadkarni, MD, MS
Assistant Clinical Professor, Department of Emergency Medicine, University of New Mexico, Albuquerque, New Mexico; President, Physician Practices, CHRISTUS–St. Vincent Regional Medical Center, Santa Fe, New Mexico
Endowed Chair, Pediatric Critical Care Medicine, University of Pennsylvania School of Medicine; Associate Professor, Anesthesia, Critical Care and Pediatrics, University of Pennsylvania School of Medicine, Director, Center for Stimulation, Advanced Education and Innovation, Endowed Chair, Pediatric Critical Care Medicine, Department of Anesthesia and Critical Care Medicine, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania
William J. Meurer, MD
Yoko Nakamura, MD
Frantz R. Melio, MD, FACEP
Clinical Lecturer, Departments of Emergency Medicine and Neurology, University of Michigan at Ann Arbor, Ann Arbor, Michigan
Nathan W. Mick, MD
Assistant Professor, University of Vermont College of Medicine, Burlington, Vermont; Director, Pediatric Emergency Medicine, Department of Emergency Medicine, Maine Medical Center, Portland, Maine
James R. Miner, MD
Associate Professor of Emergency Medicine, University of Minnesota Medical School; Research Director, Department of Emergency Medicine, Hennepin County Medical Center, Minneapolis, Minnesota
Emergency Medicine Resident, Oregon Health and Science University, Portland, Oregon
Lewis S. Nelson, MD
Associate Professor of Emergency Medicine; Director, Fellowship in Medical Toxicology, New York University School of Medicine; Associate Director, New York City Poison Control Center, New York, New York
Robert W. Neumar, MD, PhD
Associate Professor of Emergency Medicine, University of Pennsylvania School of Medicine; Associate Director, Center for Resuscitation Science, University of Pennsylvania School of Medicine, Department of Emergency Medicine, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania
Edward J. Newton, MD
Kim Newton, MD, FACEP
xvii
Associate Professor, Emergency Medicine, University of Virginia; Director, Division of Prehospital Care, University of Virginia Medical Center, Charlottesville, Virginia
Andrew D. Perron, MD
Assistant Professor of Emergency Medicine, Department of Emergency Medicine, Keck School of Medicine, University of Southern California, Los Angeles, California
Professor of Emergency Medicine, University of Vermont School of Medicine, Burlington, Vermont; Emergency Medicine Residency Program Director, Maine Medical Center, Portland, Maine
James T. Niemann, MD
Shawna J. Perry, MD
Professor of Medicine, The David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Senior Physician Specialist, Medicine/Emergency Medicine, Department of Emergency Medicine, HarborUniversity of California at Los Angeles Medical Center, Torrance, California
Richard M. Nowak, MD, MBA
Clinical Professor, Department of Emergency Medicine, Wayne State University School of Medicine, Detroit, Michigan; Clinical Associate Professor, Department of Emergency Medicine, University of Michigan School of Medicine, Ann Arbor, Michigan; Past Chair, Department of Emergency Medicine, Henry Ford Health System, Detroit, Michigan
John F. O’Brien, MD
Associate Professor of Emergency Medicine, University of Florida College of Medicine, Gainesville, Florida; Associate Professor of Emergency Medicine, University of South Florida College of Medicine, Tampa, Florida; Associate Residency Director, Department of Emergency Medicine, Orlando Regional Medical Center, Orlando, Florida
Jonathan S. Olshaker, MD
Professor and Chair, Department of Emergency Medicine, Boston University School of Medicine; Chief, Department of Emergency Medicine, Boston Medical Center, Boston, Massachusetts
Edward J. Otten, MD, FACMT, FAWM
Professor of Emergency Medicine and Pediatrics; Director, Division of Toxicology, University of Cincinnati College of Medicine, Cincinnati, Ohio
Leslie C. Oyama, MD
Assistant Clinical Professor, University of California at San Diego, Department of Emergency Medicine, University of California at San Diego School of Medicine, San Diego, California
Daniel J. Pallin, MD, MPH
Assistant Professor, Medicine (Emergency Medicine); Assistant Professor of Pediatrics, Harvard Medical School; Research Director, Department of Emergency Medicine, Brigham and Women’s Hospital; Attending Physician, Division of Emergency Medicine, Children’s Hospital Boston, Boston, Massachusetts
Paul M. Paris, MD, FACEP, LLD(Hon)
Professor and Chair, Department of Emergency Medicine, University of Pittsburgh School of Medicine; Chief Medical Officer, Center for Emergency Medicine of Western Pennsylvania, Pittsburgh, Pennsylvania
Associate Professor, Associate Chair, Department of Emergency Medicine, Virginia Commonwealth University, School of Medicine; Director for Patient Safety Systems Engineering, Virginia Commonwealth University Health Systems, Richmond, Virginia
Michael A. Peterson, MD
Associate Professor of Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Vice Chair, Clinical Affairs, Department of Emergency Medicine, Harbor-University of California at Los Angeles Medical Center, Torrance, California
James A. Pfaff, MD
Assistant Professor, Uniformed Services University of the Health Sciences, Bethesda, Maryland; Emergency Medicine Residency, Department of Emergency Medicine, San Antonio Uniformed Health Education (SAUSHEC), Brooke Army Medical Center, Fort Sam Houston, Houston, Texas
Sharon Pfeil, MD
Professor, Department of Emergency Medicine, The Ohio State University; Professor and Chair, Department of Emergency Medicine, The Ohio State University, Columbus, Ohio
William James Phillips, MD
Associate Professor, Departments of Anesthesiology and Emergency Medicine, University of Mississippi Medical Center, Jackson, Mississippi
Melissa Platt, MD
Assistant Professor, University of Louisville, Louisville, Kentucky
Michael Alan Polis, MD, MPH
Clinical Professor, Emergency Medicine, George Washington University Medical School, Washington, DC; Attending Physician, Division of Intramural Research, Warren Grant Magnuson Clinical Center, Bethesda, Maryland
Charles V. Pollack, Jr., MD, MA, FACEP, FAAEM, FAHA
Professor of Emergency Medicine, University of Pennsylvania School of Medicine; Chairman, Department of Emergency Medicine, Pennsylvania Hospital, Philadelphia, Pennsylvania
Timothy G. Price, MD
Associate Professor, Department of Emergency Medicine, University of Louisville, Louisville, Kentucky
Thomas B. Purcell, MD
Adjunct Assistant Clinical Professor, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Attending Faculty, Department of Emergency Medicine, Kern Medical Center, Bakersfield, California
Contributors
Professor, Emergency Medicine; Chair, Department of Emergency Medicine, Keck School of Medicine, Los Angeles; Chair, Department of Emergency Medicine, LAC and USC Medical Center, Los Angeles, California
Debra Perina, MD
Contributors
xviii
Prasanthi Ramanujam, MD, MAS, MBBS
Michael S. Runyon, MD
Rama B. Rao, MD
Christopher S. Russi, DO, FACEP
Neha P. Raukar, MD
Bisan A. Salhi, MD
Assistant Professor, University of California at San Francisco; Attending Physician, Department of Emergency Medicine, University of California at San Francisco Medical Center, San Francisco, California Assistant Professor, Emergency Medicine and Public Health, Weill Medical College at Cornell University; Faculty, Emergency Medicine, New York Presbyterian Hospital at the WeillCornell Medical Center, New York, New York Assistant Professor, Alpert Medical School of Brown University; Emergency Medicine Attending Physician, Primary Care Sports Medicine (University Orthopedics), Rhode Island Hospital and The Miriam Hospital, Providence, Rhode Island
James W. Rhee, MD
Assistant Professor of Medicine and Pediatrics, The University of Chicago; Director, Medical Toxicology; Attending Physician, Adult Emergency Department; Attending Physician, Pediatric Emergency Department, The University of Chicago Medical Center, Chicago, Illinois
David B. Richards, MD
Clinical Instructor, Department of Surgery, University of Colorado Denver, School of Medicine; Attending Physician, Denver Health Medical Center, Denver, Colorado
John R. Richards, MD
Professor, University of California, Davis Medical Center, Department of Emergency Medicine, Sacramento, California
David J. Roberts, MD
Adjunct Professor, University of Minnesota Medical School, Minneapolis, Minnesota; Consulting Toxicologist, Staff Emergency Physician, North Memorial Medical Center, Robbinsdale, Minnesota
Howard Rodenberg, MD, MPH
Director of the Division of Health and Environment and State Health Officer; Clinical Associate Professor, University of Kansas Medical School, Wichita, Kansas; Department of Health and Environment, Topeka, Kansas
Kevin G. Rodgers, MD
Clinical Professor of Emergency Medicine and Co-Program Director, Emergency Medicine Residency, Indiana University School of Medicine, Indianapolis, Indiana
Richard E. Rothman, MD, PhD, FACEP
Associate Professor, Department of Emergency Medicine, The Johns Hopkins University, The Johns Hopkins Hospital, Baltimore, Maryland
David H. Rubin, MD
Professor of Clinical Pediatrics, Albert Einstein College of Medicine; Chairman and Program Director, Department of Pediatrics, St. Barnabas Hospital, Bronx, New York
Douglas A. Rund, MD
Professor and Chair, Department of Emergency Medicine; Associate Dean, College of Medicine and Public Health, Ohio State University, Columbus, Ohio
Adjunct Assistant Professor of Emergency Medicine, University of North Carolina—Chapel Hill, Chapel Hill, North Carolina; Assistant Residency Director, Carolinas Medical Center, Charlotte, North Carolina Assistant Professor of Emergency Medicine, Mayo Clinic College of Medicine; Associate Director for EMS Research, Mayo Clinical Medical Transport, Department of Emergency Medicine, Rochester, Minnesota Professor, Department of Emergency Medicine, Emory School of Medicine; Attending Emergency Physician, Emory University Hospital, Atlanta, Georgia
Sally A. Santen, MD
Associate Professor, Department of Emergency Medicine, Office of Medical Education and Student Affairs, Emory University School of Medicine, Atlanta, Georgia
Radu V. Saveanu, MD
Chairman, Department of Psychiatry, Ohio State University; Executive Director, Ohio State University Harding Hospital, Columbus, Ohio
Richard J. Scarfone, MD
Associate Professor of Pediatrics, University of Pennsylvania School of Medicine; Attending Physician, Emergency Medicine; Medical Director, Emergency Preparedness, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania
Michael J. Schmidt, MD
Assistant Professor, Northwestern University, Feinberg School of Medicine; Medical Director, Northwestern Memorial Hospital, Chicago, Illinois
Diana C. Schneider, MD
Assistant Professor of Family and Internal Medicine, Keck School of Medicine, University of California at Los Angeles; Medical Director, Adult Protection Team, Los Angeles County and University of Southern California Medical Center, Los Angeles, California
Carl H. Schultz, MD
Professor of Emergency Medicine, Co-Director, EMS and Disaster Medical Sciences Fellowship, Department of Emergency Medicine, University of California at Irvine, School of Medicine, Irvine, California; Director, Disaster Medical Services, University of California at Irvine Medical Center, Orange, California
Richard B. Schwartz, MD
Chairman and Professor, Medical College of Georgia, Department of Emergency Medicine, Augusta, Georgia
Susan M. Scott, MD
Associate Professor, Department of Pediatrics, University of Texas, Southwestern Medical Center; Pediatric Emergency Medicine Fellowship Director, Emergency Services, Children’s Medical Center of Dallas, Dallas, Texas
Donna L. Seger, MD
Jeffrey A. Seiden, MD
Assistant Professor of Clinical Pediatrics, University of Pennsylvania School of Medicine; Attending Physician, Emergency Medicine, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania
Jennifer Seirafi, MD
Assistant Voluntary Professor of Medicine, Miller School of Medicine, University of Miami; Emergency Care Center Attending Physician, Jackson Memorial Hospital, Miami, Florida
Clare T. Sercombe, MD
Staff Physician, Emergency Department, North Memorial Medical Center, Robbinsdale, Minnesota
Joseph D. Sexton, MD, FACEP, AA
Clinical Assistant Professor, Penn State University Medical School, Hershey, Pennsylvania; Attending Physician, Department of Emergency Medicine, Lehigh Valley Health Network, Allentown, Pennsylvania
Assistant Professor of Clinical Emergency Medicine, Keck School of Medicine of the University of Southern California; Associate Residency Director, Department of Emergency Medicine, Los Angeles County and University of Southern California Medical Center, Los Angeles, California
Lee W. Shockley, MD, FACEP, FAAEM
Professor of Surgery, Division of Emergency Medicine, University of Colorado School of Medicine, Aurora, Colorado; Emergency Department Medical Director, Associate Residency Program Director, The Denver Health Medical Center Residency in Emergency Medicine, The Denver Health Medical Center, Denver, Colorado
Robert Silbergleit, MD
Associate Professor, University of Michigan, Ann Arbor, Michigan
Barry C. Simon, MD
University of California at San Francisco, San Francisco, California; Chairman, Department of Emergency Medicine, Alameda County Medical Center, Oakland, California
Adam J. Singer, MD
Professor and Vice Chairman for Research, Stony Brook University, Stony Brook, New York
Marc J. Shapiro, MD
Jonathan I. Singer, MD, FAAP, FACEP
Nathan I. Shapiro, MD, MPH
Amardeep Singh, MD, RDMS
Assistant Professor, Brown University; Attending Physician, Department of Emergency Medicine, Rhode Island Hospital, Providence, Rhode Island Assistant Professor, Harvard Medical School, Boston; Research Director, Beth Israel Deaconess Medical Center, Boston, Massachusetts
Ghazala Q. Sharieff, MD, FACEP, FAAEM, FAAP
Associate Clinical Professor and Division Director, Rady Children’s Hospital Emergency Care Center; Director of Pediatric Emergency Medicine, Palomar-Pomerado Health System, San Diego, California
Rahul Sharma, MD, MBA, FACEP
xix
Professor of Emergency Medicine and Pediatrics, Wright State University School of Medicine; Staff Physician, Children’s Medical Center, Dayton, Ohio Assistant Professor of Emergency Medicine, Chicago Medical School, North Chicago, Illinois; Emergency Room Physician, QI Director for Emergency Department, Ultrasound Director for Emergency Department, Mount Sinai Hospital, Chicago, Illinois
Laura Slaughter, MD, FACP
Consultant, Violence Intervention Program, University of Southern California Medical Center, Los Angeles, California; San Luis Obispo County SART, San Luis Obispo, California
Assistant Professor and Attending Physician, Co-Coordinator, Medical Student Sub-internship in Emergency Medicine, Weill-Cornell Medical College; Assistant Director, Emergency Department Operations, Department of Emergency Medicine, New York Presbyterian Weill-Cornell Medical Center, New York, New York
Jeffrey Paul Smith, MD, MPH
Peter Shearer, MD
Professor, Division of Protective Medicine, Department of Emergency Medicine; Director, Clinical Forensic Medicine Program, Department of Emergency Medicine, University of Louisville School of Medicine, University of Louisville Hospital, Louisville, Kentucky
Assistant Professor Emergency Medicine, Mount Sinai School of Medicine; Residency Program Director, Mount Sinai Medical Center, New York, New York
Richard D. Shih, MD
Associate Professor of Surgery, New Jersey Medical School, Newark, New Jersey; Emergency Medicine Residency Director, Morristown Memorial Hospital, Morristown, New Jersey
Associate Professor, Co-Director Ronald Reagan Institute of Emergency Medicine, George Washington University Medical Center; Director of Clinical Operations and Trauma Services, George Washington University Hospital, Washington, DC
William Spafford Smock, MD, MS
Peter E. Sokolove, MD
Professor, Vice Chair for Education, Program Director, Department of Emergency Medicine, University of California Davis Health System, Sacramento, California
Harry S. Soroff, MD
Professor Emeritus, Stony Brook University, Stony Brook, New York
Contributors
Associate Professor of Medicine and Emergency Medicine, Department of Medicine; Medical Director, Tennessee Poison Center, Vanderbilt University Medical Center, Nashville, Tennessee
Jan M. Shoenberger, MD
xx
Contributors
Benjamin Squire, MD
Clinical Instructor of Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; EMS/Research Fellow, HarborUniversity of California at Los Angeles Medical Center, Torrance, California
Stephen H. Thomas, MD, MPH
Kaiser Foundation Professor and Chair, Department of Emergency Medicine, University of Oklahoma School of Community Medicine, Tulsa, Oklahoma
Carrie D. Tibbles, MD
Assistant Professor, Emergency Medicine, University of Cincinnati Medical Center, Cincinnati, Ohio
Assistant Professor, Harvard Medical School; Associate Program Director, Beth Israel Deaconess Medical Center, Harvard Affiliated Emergency Medicine Residency, Boston, Massachusetts
Sara T. Stewart, MD, MPH
Joshua J. Tobias, MD
Brian A. Stettler, MD
Assistant Professor of Pediatrics, University of California at Los Angeles, Los Angeles, California; Medical Director, Child Crisis Center, Harbor-University of California at Los Angeles Medical Center, Torrance, California
David M. Stocker, MD
Assistant Clinical Professor, Department of Medicine, David Geffen School of Medicine at the University of California at Los Angeles, Los Angeles, California; Associate Program Director, Department of Emergency Medicine, Kern Medical Center, Bakersfield, California
Chairman of Pediatrics and Medical Director, Pediatric Emergency Department, Swedish Medical Center, Englewood, Colorado; Pediatric Emergency Physician, Carepoint P.C., Denver, Colorado
Glenn F. Tokarski, MD
Susan Stone, MD, MPH
Associate Professor of Emergency Medicine, University of Southern California at Los Angeles; Associate Professor of Clinical Emergency Medicine, Director of Palliative Care, University of Southern California, Los Angeles, California
Medical Director, HBO, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, North Carolina; Adjunct Associate Professor of Emergency Medicine, University of North Carolina School of Medicine, Chapel Hill, North Carolina
Jared Strote, MD, MS
Sam S. Torbati, MD, FAAEM
Assistant Professor, University of Washington Medical Center, Seattle, Washington
Stuart P. Swadron, MD, FACEP, FAAEM, FRCPC
Associate Professor of Clinical Emergency Medicine, Keck School of Medicine, University of Southern California; ViceChair of Education and Program Director, Department of Emergency Medicine, Los Angeles County/University of Southern California Medical Center, Los Angeles, California
Allison Tadros, MD
Assistant Professor, Health Science Center, West Virginia University; Assistant Residency Director, Health Science Center, West Virginia State University, Morgantown, West Virginia
Breena R. Taira, MD
Research Fellow, Stony Brook University, Stony Brook, New York
David A. Talan, MD, FACEP, FAAEM
Professor of Medicine in Residence, The David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Chairman, Department of Emergency Medicine and Faculty, Division of Infectious Diseases, Olive View-University of California at Los Angeles Medical Center, Sylmar, California
Vivek S. Tayal, MD
Director, Division of Emergency Ultrasound, Department of Emergency Medicine, Carolinas Medical Center; Clinical Associate Professor of Emergency Medicine, University of North Carolina, Charlotte, North Carolina
Senior Staff Physician, Emergency Medicine, Henry Ford Hospital, Henry Ford Healthcare System, Detroit, Michigan
Christian Tomaszewski, MD
Assistant Clinical Professor of Medicine, University of California at Los Angeles Medical Center; Associate Medical Director and Attending Physician, Cedars-Sinai Medical Center, Los Angeles, California
Susan P. Torrey, MD, FACEP
Assistant Professor of Emergency Medicine, Tufts University School of Medicine, Boston, Massachusetts; Associate Residency Director, Department of Emergency Medicine, Baystate Medical Center, Springfield, Massachusetts
T. Paul Tran, MD, MS, FACEP
Associate Professor and Research Director, Department of Emergency Medicine, University of Nebraska Medical Center, Omaha, Nebraska
Sandra Ugras-Rey, DO
Core Faculty, Newark Beth Israel Medical Center; Associate Medical Director, Department of Emergency Medicine, Newark Beth Israel Medical Center, Newark, New Jersey
Monira Vakil, DO
Assistant Professor of Emergency Medicine, Department of Emergency Medicine, University of Mississippi Medical Center, Jackson, Mississippi
Marshall G. Vary, MD
Assistant Clinical Professor of Psychiatry, Department of Psychiatry, Ohio State University; Active Medical Staff Member, Department of Psychiatry, Riverside Methodist Hospital, Columbus, Ohio
Larissa I. Velez, MD, FACEP
Associate Professor, Division of Emergency Medicine; Associate Residency Director, Emergency Medicine, University of Texas Southwestern Medical Center; Staff Toxicologist, North Texas Poison Center, Dallas, Texas
John M. Wightman, MD, MA
Robert J. Vissers, MD, FRCPC
Saralyn R. Williams, MD
Associate Professor of Emergency Medicine, Department of Emergency Medicine, University of Louis Hospital, University of Louisville School of Medicine, Louisville, Kentucky Adjunct Associate Professor, Department of Emergency Medicine, Oregon Health Sciences University; Chief, Emergency Medicine, Associate Chief Medical Officer, Legacy Emanuel Hospital, Portland, Oregon
Ron M. Walls, MD
Professor of Medicine (Emergency Medicine), Harvard Medical School; Chairman, Department of Emergency Medicine, Brigham and Women’s Hospital, Boston, Massachusetts
Mark Watson, MD
Professor and Education Director, Department of Emergency Medicine, Boonshoft School of Medicine, Wright State University, Dayton, Ohio Associate Professor of Clinical Medicine, Department of Medicine and Department of Emergency Medicine, Vanderbilt University, Nashville, Tennessee
Adria O. Winter, MD
Assistant Clinical Professor of Medicine, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Attending Faculty Physician, Department of Emergency Medicine, Kern Medical Center, Bakersfield, California
Vice President of Clinical Effectiveness/Emergency Medicine; Attending Physician, Hospital Administration/Department of Emergency Medicine, Newark Beth Israel Medical Center, Newark, New Jersey
Mary A. Wittler, MD
Paul M. Wax, MD
Jeannette M. Wolfe, MD, FACEP
Clinical Professor of Surgery, Department of Emergency Medicine, University of Texas Southwestern Medical School, Dallas, Texas; Executive Director, American College of Medical Toxicology, Phoenix, Arizona
Robert L. Wears, MD, MS
Professor, Department of Emergency Medicine, University of Florida Health Science Center; Attending Physician, Shands Medical Center, Jacksonville, Florida; Visiting Professor, Clinical Safety Research Unit, Imperial College and St. Mary’s Hospital, London, UK
Ellen J. Weber, MD
Professor of Clinical Emergency Medicine, University of California at San Francisco, San Francisco, California
Hugh H. West, MD
Assistant Professor of Emergency Medicine, Wake Forest University Baptist Medical Center, Winston-Salem, North Carolina Assistant Professor of Emergency Medicine, Tufts University School of Medicine, Baystate Campus, Springfield, Massachusetts
Allan B. Wolfson, MD
Professor of Emergency Medicine, University of Pittsburgh; Program Director, University of Pittsburgh Affiliated Residency in Emergency Medicine, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania
Karen G. H. Woolfrey, MD, FRCPC, FACEP
Assistant Professor, Department of Medicine; Deputy Director, Division of Emergency Medicine, McMaster University; Research Coordinator and Director of Residency Clinical Teaching Unit, Emergency Department, St. Joseph’s Healthcare Hamilton, Hamilton, Ontario, Canada
Assistant Professor, Department of Emergency Medicine, University of California San Francisco School of Medicine; Assistant Professor of Emergency Medicine, University of San Francisco School of Medicine, San Francisco, California
Michael Woolfrey, MD, BSc., BMedSc., FRCS(C)
Matthew A. Wheatley, MD
Joshua L. Wright, MD
Benjamin A. White, MD
Samuel Yang, MD
Suzanne R. White, MD
Michael Yaron, MD
Assistant Professor, Emory University; Attending Physician, Emory University Hospital, Atlanta, Grady Memorial Hospital, Atlanta, Georgia Clinical Fellow in Medicine, Harvard Medical School, Boston, Massachusetts Munuswamy Dayanandan Professor and Chair, Department of Emergency Medicine, Wayne State University School of Medicine; Emergency Physician-in-Chief, Detroit Medical Center; Medical Director, Children’s Hospital of Michigan, Regional Poison Control Center, Detroit, Michigan
Robert A. Wiebe, MD, FAAP, FACEP
Professor, Division of Pediatric Emergency Medicine, University of Texas Southwestern Medical Center, Dallas, Texas
xxi
Assistant Clinical Professor, McMaster University, Hamilton, Ontario, Canada; Chief, Department of Orthopaedic Surgery, Brantford General Hospital, Brantford, Ontario, Canada Associate Professor, Residency Director (Military Component), Department of Emergency Medicine, Wright State University, Dayton, Ohio Assistant Professor, Johns Hopkins University, Baltimore, Maryland Professor of Surgery, University of Colorado Denver, School of Medicine; Emergency Medicine Attending Physician, University of Colorado Hospital, Aurora, Colorado
Donald M. Yealy, MD
Professor and Vice-Chair of Emergency Medicine, University of Pittsburgh; Vice-Chair, University of Pittsburgh Physicians, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania
Contributors
Salvator Vicario, MD
xxii
Contributors
Amy Young, MD
Clinical Assistant Professor, University of Texas Southwestern, Dallas; Emergency Medicine and Toxicology Faculty, Parkland Memorial Hospital and Children’s Medical Center, University of Texas Southwestern, Dallas, Texas
Kelly D. Young, MD, MS
Associate Clinical Professor of Pediatrics, David Geffen School of Medicine at University of California at Los Angeles, Los Angeles, California; Director of Pediatric and Pain Management Education, Harbor-University of California at Los Angeles Medical Center, Torrance, California
John G. Younger, MD, MS
Associate Professor, Associate Chair for Research, Department of Emergency Medicine, University of Michigan, Ann Arbor, Michigan
Richard Zane, MD
Vice Chair, Department of Emergency Medicine, Brigham and Women’s Hospital, Boston, Massachusetts
David K. Zich, MD
Assistant Professor, Department of Emergency and Internal Medicine, Northwestern University, Feinberg School of Medicine; Attending Physician, Northwestern Memorial Hospital, Chicago, Illinois
Gary D. Zimmer, MD, FAAEM
Assistant Professor, Department of Emergency Medicine, Johns Hopkins University School of Medicine; Director, Department of Emergency Medicine, Harbor Hospital; Assistant Medical Director for Baltimore Operations, Aeromedical Transport Services Corporation, Baltimore, Maryland
Brian J. Zink, MD
Professor and Chair, Department of Emergency Medicine, Alpert Medical School of Brown University; Physician-inChief, Emergency Medicine, Rhode Island Hospital, The Miriam Hospital, and Hasbro Children’s Hospital, Providence, Rhode Island
David Zull, MD, FACEP, FACP
Associate Professor of Medicine and Emergency Medicine, Feinberg School of Medicine, Northwestern University; Director, Emergency Department Observation Unit, Northwestern Memorial Hospital, Chicago, Illinois
Leslie S. Zun, MD, MBA
Professor and Chairman, Department of Emergency Medicine, Rosalind Franklin University of Medicine and Science, The Chicago Medical School, North Chicago, Illinois; Chairman, Department of Emergency Medicine, Mount Sinai Hospital, Chicago, Illinois
Preface to the Seventh Edition We are pleased to provide this 7th edition of Rosen’s Emergency Medicine: Concepts and Clinical Practice, now in its 28th year of existence, and wish to recite several changes intended to enhance its content, readability, and purpose. The textbook has been converted into two volumes and reduced in size by more than 500 pages. This was accomplished with judicious editing but mostly through the transfer of the entire bibliography onto the book’s website at expertconsult.com, where the full text is available online, along with an image library, Q&A, and updates. The number of chapters is virtually unchanged while the number of annotations per chapter has been expanded. More importantly, we have strived to render the textbook as strongly evidenced based as possible through reliance on the vigilant selection of high-quality and recent references. We will continue to add online updates, wherein recent journal articles selected by the senior editors are abstracted and electronically plugged into the margins of the relevant area of the existing web-based version of the book. We also have the good fortune of adding numerous authors who are authoritative in the subjects about which they write as well as a new team that has composed a Question and Answer com-
pendium. Much of the artwork and the format have been reworked and many of the photographs, including radiographs, have been updated. We are grateful to many. The authors have committed their many hours and substantive expertise to lay the foundation. The editors have performed yeoman’s work to maintain consistency in templating each chapter and to help assure accuracy and clarity. We thank Judy Fletcher, our Publishing Director; Stefanie Jewell-Thomas, our Acquisitions Editor; and Dee Simpson, our Developmental Editor; for listening thoroughly and suggesting wisely and for all their work behind the scenes. We are most appreciative for the terrific support from our administrative assistants, Tricia Wyatt and Gail Franklin (JAM), Maria Figueroa (RSH), and Diane Pugh and Janice Bingham (RMW). We could not have committed the requisite time and energy to this labor of great love were we without the encouragement and buoying up from our families. Finally, we thank you, Peter, for making all this possible just because you chased a dream three decades ago. John A. Marx Robert S. Hockberger Ron M. Walls
xxiii
Preface to the First Edition From the vision and foresight of a few physicians who perceived the need for a unique, disciplined, sensitive approach to the identification and stabilization of patients threatened with loss of life or limb, emergency medicine has rapidly developed into an exciting, academically recognized medical specialty. This textbook is dedicated to those who have accepted its responsibilities, challenges, and excitements. We have attempted to define in depth the material on which our practice is based. There have been a number of efforts to write about emergencies, but we believe that this is the first to call solely on those people who themselves practice the specialty. In every chapter theory and knowledge pertinent to the practice of emergency medicine are presented. This book is not an easy one; it was written based on published literature, not anecdote or prejudice. In many instances where the data are not available, both sides are presented with a suggested practice. The book is intended for all with a serious interest in or a need to know emergency medicine, including those who do not practice full-time emergency medicine, as well as the dedicated specialists who do. The book is organized into two main sections—trauma and nontrauma. This division is artificial but does correspond to the first major decisions made in patient evaluations, because trauma usually affects individual anatomic structures, whereas nontrauma is more likely to affect systems. Despite this artificial separation, long and detailed discussion and instruction to authors concerning content and style
ensued. We realize that we could not tap all available talent for contributions to the book, but we have made an effort to represent different schools of thought and regions of the country. There are deliberate omissions; for example, we elected not to include any procedures. There was not enough room to create an atlas, but it was our desire to cover the chosen topics in detail. No effort has been made to address administration, management, disaster planning, or technical requirements of emergency medicine supplies or design. Prehospital care has been included only as it related to individual topics, not as suggested protocol or from the vantage point of technician training programs. It would be impossible to write a book this long and present nothing controversial. In fact we ourselves find sections we cannot totally accept, but in the process of working with multiple authors, we cannot with intellectual honesty put ideas into their material. We have, however, achieved our goal of presenting an in-depth vision of emergency medicine. We hope you will find the reading of this book as stimulating and enjoyable as we have found in its creation. Peter Rosen Frank J. Baker II G. Richard Braen Robert H. Dailey Richard C. Levy
xxv
Acknowledgments To all my editorial colleagues, especially Ron and Bob, as well as our authors, all of whose extraordinary and collective efforts created this edition; to my wife, Karin, for her immeasurable love and support; and to my children, Conner and Shelby, who have enriched my heart and soul beyond description. JAM To my mentors, my colleagues, and my students, who together comprised my professional family over the past 30 years, for their inspiration, friendship, and support; and to Patty, who brought beauty, love, and balance to my life. RSH With humble thanks to Peter, for his enduring friendship, for his great vision in the creation of an extraordinary specialty, and for his selfless mentorship of so many of its present and future leaders; to John and Bob, for being the engine of this amazing and rewarding collaboration; to the editors and authors, who give so freely of their precious time to help to bring so much to so many; to my amazingly resilient, loving, and supportive wife, Barb, and to my children, Andrew, Blake, and Alexa, whose love and intellectual energy inspire me. RMW To my father, James J. Adams, whose strength will forever inspire me; to my mother, Rita A. Adams, whose devotion to family will forever guide me; and to the many other members of my family: Cecelia, Joe, Jeff, Liz, Rob, David, Nicholas, Gregory, Leah, Katherine, Sydney, and Trent, whose support I rely on. JGA
To my family, friends, students, and teachers. I am continuously amazed, grateful, and humbled by your support and encouragement. Thank you. MHB To Joanna, Megan, and Julia and our dual-phase marriage; kind of “pre-Rosen” and now four editions together. May you never be tempted to join the literary “first-wives club.” DFD I dedicate this book to my husband, David, and our children, Sarah, Jeremiah, and Katie—for their patience, love, and support. MGH To emergency medicine residents and faculty everywhere in their constant pursuit of knowledge but especially those at Hennepin County Medical Center for continuing to teach me. I am grateful to my parents, Rose and Joseph, for their commitment to education. Special thanks to Eric, Ali, Amanda, and, most of all, Beth for their love, patience, and understanding. LJL I would like to thank my teachers—my parents and my children, professors and patients, colleagues and students—who have patiently taught me about medicine and life; and my steadfast companion and wife, Lynda, who has made the pursuit of wisdom possible. EJN
To all my former residents over the past 28 years from both Cincinnati and Michigan. I have been privileged to work with the best and I have learned a lot from all of you. I have to thank my children, David, Blake, and Anna, for keeping me thinking young and teaching me lots about life and family. Most of all, thanks to my wife and best friend, Mary, for her love and support over the past 33 years. She makes it all worthwhile. WGB
xxvii
PART I
Fundamental Clinical Concepts
•
Section One Critical Management Principles Chapter 1
Airway
Ron M. Walls
■ PERSPECTIVE Airway management is the cornerstone of resuscitation and is a defining skill for the specialty of emergency medicine. The emergency physician has primary responsibility for management of the airway. All techniques of airway management lie within the domain of emergency medicine. Rapid sequence intubation (RSI) with direct laryngoscopy is the most commonly used method for emergency intubation, but emergency airway management includes various intubation maneuvers, use of ancillary devices, approaches to the difficult airway, and rescue techniques when intubation fails. Since the first reported use of neuromuscular blocking agents (NMBAs) in the emergency department (ED) by emergency personnel in 1971, there has been progressive sophistication of emergency airway techniques, pharmacologic agents, and special devices used to facilitate intubation.1,2 In the 1990s, RSI was widely adopted as the method of choice for most emergency intubations in the ED, and increasing attention has focused on identification and management of anticipated difficult intubation.3,4
■ PATHOPHYSIOLOGY Decision to Intubate A decision to intubate should be based on careful assessment of the patient with respect to three essential criteria: (1) failure to maintain or protect the airway, (2) failure of ventilation or oxygenation, and (3) the patient’s anticipated clinical course and likelihood of deterioration.5
Failure to Maintain or Protect the Airway A patent airway is essential for adequate ventilation and oxygenation. If the patient is unable to maintain the airway, patency must be established by artificial means, such as repositioning, chin lift, jaw thrust, or insertion of an oral or nasal airway. Likewise, the patient must be able to protect against aspiration of gastric contents, which carries significant morbidity and mortality. Traditionally, the presence or absence of a gag reflex has been advocated as a reliable indicator of the patient’s ability to protect the airway, but the gag reflex is absent in 12 to 25% of normal adults, and there is no evidence that its presence or absence corresponds to airway protective
reflexes or the need for intubation.6 Testing the gag reflex in an obtunded, supine patient is unlikely to yield useful information with respect to the need to intubate and may precipitate vomiting. The patient’s ability to swallow or handle secretions is a more reliable indicator of airway protection.5 The recommended approach is to evaluate the patient’s ability to phonate in response to voice command or query (which provides information about level of consciousness and voice quality), level of consciousness, and ability to manage his or her own secretions (e.g., pooling of secretions in the oropharynx, absence of swallowing spontaneously or to command.) In general, a patient who requires a maneuver to establish a patent airway or who easily tolerates an oral airway probably requires intubation for protection of that airway, unless a temporary or readily reversible condition, such as opioid overdose, is present.
Failure of Ventilation or Oxygenation Ventilatory failure that is not reversible by clinical means or increasing hypoxemia that is not adequately responsive to supplemental oxygen is a primary indication for intubation. This assessment is clinical and includes evaluation of the patient’s general status, oxygenation by pulse oximetry, and changes in the ventilatory pattern. Continuous capnography also can be helpful, but is not essential if oximetry readings are reliable. Arterial blood gases (ABGs) generally are not required to determine the patient’s need for intubation. In most circumstances, clinical assessment, including pulse oximetry with or without capnography, and observation of improvement or deterioration lead to a correct decision. ABG results are rarely helpful, may cause delay in intubating a deteriorating patient, and may be misleading, so, if obtained, they must be interpreted carefully in the context of the patient’s clinical status. Patients who are clinically improving despite severe or worsening ABG alterations may not require intubation, whereas a rapidly tiring patient may require intubation when ABG values are only modestly disturbed or even improving. Regardless of the underlying cause, the need for mechanical ventilation generally mandates intubation. External mask devices increasingly have been used to provide assisted mechanical ventilation without intubation (see Chapter 2), but despite these advances, most patients who need assisted ventilation or positive pressure to improve oxygenation require intubation.7 3
4
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
Anticipated Clinical Course Certain conditions indicate the need for intubation even in the absence of frank airway, ventilatory, or oxygenation failure. These conditions are characterized by a moderate to high likelihood of predictable deterioration that would require airway intervention. Intubation may be indicated relatively early in the course of severe cyclic antidepressant overdose. Although the patient is awake, protecting the airway, and exchanging gas well, intubation is advisable to guard against the strong likelihood of clinical deterioration, which can occur relatively abruptly and includes coma, seizure, cardiac dysrhythmia or arrest, and possible aspiration of activated charcoal or gastric contents. Significant multiple trauma, with or without head injury, may be an indication for intubation.8,9 Many of these patients are ventilating normally through a patent airway, and oxygen levels frequently are normal or supernormal with supplemental oxygen. Despite this, anticipated deterioration, loss of the ability to protect the airway, the need for invasive and painful procedures, or the need for studies outside the ED (e.g., computed tomography, angiography) may mandate intubation.10 A patient with penetrating neck trauma may present with a patent airway and adequate gas exchange. Nevertheless, early intubation is advisable with any evidence of vascular or direct airway injury because these patients tend to deteriorate and because increasing hemorrhage or swelling in the neck tends to both compromise the airway and confound later attempts at intubation.11,12 Although these indications for intubation may seem quite different and individualized, the common thread is the anticipated clinical course over time. In each circumstance, it can be anticipated that future events will compromise either the patient’s ability to maintain and protect the airway or the patient’s ability to oxygenate and ventilate. A similar thought process is applied to any patient who will be leaving the ED for diagnostic studies (e.g., angiography) or who may be transported to another facility. If it seems clinically likely that the patient may deteriorate, then “preemptive” intubation is the prudent course.
■ CLINICAL FEATURES Identification of the Difficult Airway In most patients, even in the ED’s dynamic and unpredictable environment, intubation is technically easy and straightforward. In large ED studies, overall intubation failure rates are less than 1% for medical intubations and less than 3% in trauma patients.1,13,14 Intubation failure occurs in approximately 1 in 200 to 1 in 2000 elective general anesthesia cases.3,15,16 Bag-mask ventilation (BMV) is difficult in approximately 1 in 50 general anesthesia patients, and impossible in approximately 1 in 600.17,18 BMV is difficult, however, in up to one third of patients in whom intubation failure occurs, and difficult BMV-makes the likelihood of difficult intubation four times greater and the likelihood of impossible intubation 12 times greater.17,18 The combination of failure of intubation and failure of BMV in elective anesthesia practice is estimated to be exceedingly rare: 1 in 5000 to 1 in 20,000 elective anesthesia patients.16,18 These numbers cannot be applied directly to the ED situation, where patient selection cannot occur (as with a preanesthetic visit), but are reassuring in that they indicate a high degree of safety if a preintubation analysis of factors predicting difficult intubation is undertaken.19 The emergency nature of the patient’s presentation often precludes postponement of the intubation, even for a short
time, but knowledge of the difficulties presented by the patient’s airway permits thoughtful planning and preparation for possible intubation failure. Preintubation assessment should evaluate the patient for difficult intubation, difficult BMV, difficult ventilation using an extraglottic device (EGD, such as a laryngeal mask airway, see later discussion) and difficult cricothyrotomy. Knowledge of all four domains is crucial to successful planning.3,4 Neuromuscular paralysis should be avoided in patients for whom a high degree of intubation difficulty is predicted, unless the administration of the NMBA is part of a planned approach to the difficult airway. This approach may include use of a double setup, in which an alternative approach, such as cricothyrotomy, is simultaneously prepared. Preintubation evaluation should be as comprehensive as clinical circumstances permit. A systematic approach to the patient is required.
Difficult Direct Laryngoscopy: LEMON Most of the difficult airway markers discussed in the anesthesia and emergency medicine literature have not been scientifically validated.20 Nevertheless, a methodical approach can be used to evaluate the patient, based on the accepted markers of difficult intubation by direct laryngoscopy. One such approach uses the mnemonic LEMON (Box 1-1).3,21 L—Look Externally. The patient first should be examined for external markers of difficult intubation, which are determined based simply on the intubator’s clinical impression. For example, the severely bruised and bloodied face of a combative trauma patient, immobilized in a cervical collar on a spine board, might (correctly) invoke an immediate appreciation of anticipated difficult intubation. Subjective clinical judgment can be highly specific (>90%), but insensitive and so must be augmented by other evaluations.18 E—Evaluate 3-3-2. The second step in the evaluation of the difficult airway is to assess the patient’s anatomy to determine his or her suitability for direct laryngoscopy. Direct laryngoscopy requires the ability to visualize the glottis by direct vision through the mouth, using alignment of the oral, pharyngeal, and laryngeal axes. Visualization requires that the mouth open adequately, that the submandibular space be adequate to accommodate the tongue, and that the larynx be positioned low enough in the neck to be accessible. These relationships have been explored in various studies by external measurement of mouth opening, oropharyngeal size, neck movement, and thyromental distance.22 The “3-3-2 rule” is an effective summary of these geometric evaluations.3,21 The 3-3-2 rule requires that the patient be able to place 3 of his or her own fingers between the open incisors, 3 of his or her own fingers along the floor of the mandible beginning at the mentum, and 2 fingers from the laryngeal prominence to the floor of the
BOX 1-1
“LEMON” Approach for Evaluation of Difficult Direct Laryngoscopy
Look externally for signs of difficult intubation (by gestalt) Evaluate the “3-3-2 rule” Mallampati Obstruction/Obesity Neck mobility Adapted with permission from The Difficult Airway Course: Emergency and Walls RM and Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed, Philadelphia, Lippincott Williams & Wilkins, 2008.
5
Chapter 1 / Airway
Figure 1-1. Final two steps of the 3-3-2
rule. A, Three fingers are placed along the floor of the mouth beginning at the mentum. B, Two fingers are placed in the laryngeal prominence (Adam’s apple). (Adapted from Murphy MF, Walls RM: Identification of difficult and failed airways. In Walls RM and Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2008, pp. 81–91. The 3-3-2 Rule is copyrighted © 2004 by The Airway Course and Lippincott Williams & Wilkins, publishers of The Manual of Emergency Airway Management.)
1 2 3
A
Class I: soft palate, uvula, fauces, pillars visible
Class II: soft palate, uvula, fauces visible
No difficulty
No difficulty
Class III: soft palate, base of uvula visible
Class IV: only hard palate visible
Moderate difficulty
Severe difficulty
Figure 1-2. The Mallampati scale assesses oral access for intubation. (From Whitten CE: Anyone Can Intubate, 4th ed. San Diego, KW Publications, 2004, with permission.)
mandible (Fig. 1-1). A patient with a receding mandible and high-riding larynx can be impossible to intubate using direct laryngoscopy. Most patients are not sufficiently cooperative for such an evaluation, and the operator compares his or her fingers with the patient’s fingers to estimate the sizes for the three tests. M—Mallampati Scale. Oral access is assessed using the Mallampati scale (Fig. 1-2). Visibility of the oral pharynx ranges from complete visualization, including the tonsillar pillars (class I), to no visualization at all, with the tongue pressed against the hard palate (class IV). Class I and class II predict adequate oral
1 2
B
access, class III predicts moderate difficulty, and class IV predicts a high degree of difficulty.22,23 A recent meta-analysis confirmed that the four-class Mallampati score performs well as a predictor of difficult laryngoscopy (and, less so, difficult intubation), but that the Mallampati score, alone, is not a sufficient assessment tool.24 O—Obstruction or Obesity. Upper airway (supraglottic) obstruction may make visualization of the glottis, or intubation itself, mechanically impossible. Conditions such as epiglottitis, laryngeal tumor, Ludwig’s angina, neck hematoma, or glottic polyps can compromise laryngoscopy, passage of the endotracheal tube (ETT), BMV, or all three. Physical examination for airway obstruction is combined with assessment of the patient’s voice to satisfy this evaluation step. There is conflicting evidence regarding whether obesity is itself an independent marker of difficult intubation or whether patients with obesity simply are more likely to have other markers of difficult intubation.25,26 Regardless, obese patients generally are more difficult to intubate than their non-obese counterparts, and preparations must account both for this, and for the more rapid oxyhemoglobin desaturation and increased difficulty with ventilation using bag and mask or an EGD (see below) that will occur. N—Neck Mobility. Neck mobility is essential for the repositioning of the angled axes of the upper airway in order to permit direct visualization of the glottis. Neck mobility is assessed by having the patient flex and extend the head and neck through a full range of motion. Neck extension is the most important motion, and simple extension may be as effective as the “sniffing” position in achieving an optimal laryngeal view.27 A recent study also found that the “extension-extension” position, in which the neck is extended on the body (opposite of the sniffing position) with the head extended on the neck, provides superior laryngeal views to the sniffing position.28 Modest limitations of motion do not seriously impair laryngoscopy, but severe loss of motion, as can occur in ankylosing spondylitis or rheumatoid arthritis, for example, may render laryngoscopy impossible. Cervical spine immobilization in trauma artificially reduces cervical spine mobility and predicts a more difficult laryngoscopy, but direct laryngoscopy is still highly successful in this group of patients.1
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
6
Identification of a difficult intubation does not preclude use of an RSI technique (see Fig. 1-7). The crucial determination is whether the clinician judges that the patient has a reasonable likelihood of intubation success, despite the difficulties identified, and that ventilation with a bag and mask or an EGD will be successful in the event that intubation fails (hence the value of the BMV and EGD assessments; see Boxes 1-2 and 1-3).
Difficult Bag-mask Ventilation: MOANS Attributes of difficult BMV have largely been validated and can be summarized with the mnemonic MOANS (see Box 1-2).3,17,18 Difficulty with mask seal; obstruction (particularly supraglottic obstruction, but can be present anywhere in the airway) or obesity (because of redundant upper airway tissues, chest wall weight, and resistance of abdominal mass); advanced age (best judged by the physiologic appearance of the patient, but age older than 55 years increases risk); edentulousness (“no teeth”), which independently interferes with mask seal; and stiffness or resistance to ventilation (e.g., asthma, chronic obstructive pulmonary disease, pulmonary edema, restrictive lung disease, term pregnancy) all cause or contribute to increased difficulty with BMV. The difficulty with BMV of the edentulous patient is the basis of the adage: “Remove dentures to intubate, leave them in to bag-mask ventilate.” The wisdom of this approach recently was validated yet again.29
Difficult Extraglottic Device Placement: RODS Placement of an EGD, such as a laryngeal mask airway, a Combitube, or a similar upper airway device often can facilitate ventilation, and convert a “can’t intubate, can’t oxygenate” situation to a “can’t intubate, can oxygenate” situation, which allows time for more careful planning of the rescue of a failed airway (see following section.) Difficulty achieving placement or ventilation using an EGD is predicted by the mnemonic “RODS.” Fortunately, if the clinician has already performed the LEMON and MOANS assessments, only the “D” for distorted anatomy remains to be evaluated (see Box 1-3).
BOX 1-2
MOANS Mnemonic for Evaluation of Difficult Bag-mask Ventilation
Mask seal Obstruction or obesity Aged No teeth Stiffness (resistance to ventilation) Adapted with permission from The Difficult Airway Course: Emergency and Walls RM and Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed, Philadelphia, Lippincott, Williams & Wilkins, 2008.
BOX 1-3
RODS Mnemonic for Evaluation of Difficult Extraglottic Device Placement
Restricted mouth opening Obstruction or obesity Distorted anatomy Stiffness (resistance to ventilation) Adapted with permission from The Difficult Airway Course: Emergency and Walls RM and Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed, Philadelphia, Lippincott, Williams & Wilkins, 2008.
Difficult Cricothyrotomy Difficult cricothyrotomy can be anticipated whenever there is disturbance of the ability to locate and access the landmarks of the anterior airway via the neck. Prior surgery; the presence of hematoma, anatomic disruption, tumor, or abscess; scarring (as from radiation therapy or prior injury); or obesity, edema, or subcutaneous air each has the potential to make cricothyrotomy more difficult. The landmarks for cricothyrotomy are sought and identified as part of the preintubation assessment of the patient.
Measurement of Intubation Difficulty The actual degree to which an intubation is “difficult” is highly subjective, and quantification is challenging. Research has relied on laryngoscopic view to characterize the intubation difficulty, and the most widely used system is that of Cormack and Lehane, which grades laryngoscopy according to the extent to which laryngeal and glottic structures can be seen. In grade 1 laryngoscopy, the entire glottic aperture is seen. Grade 2 laryngoscopy visualizes only a portion of the glottis (arytenoid cartilages alone or arytenoid cartilages plus part of the vocal cords). Grade 3 laryngoscopy visualizes only the epiglottis. In grade 4 laryngoscopy, not even the epiglottis is visible. Research conducted on elective anesthesia patients suggests that true grade 4 laryngoscopy, which is associated with impossible intubation, occurs in less than 1% of patients. Grade 3 laryngoscopy, which represents extreme intubation difficulty, is found in less than 5% of patients. Grade 2 laryngoscopy, which occurs in 10 to 30% of patients, can be subdivided further into grade 2a, in which arytenoids and a portion of the vocal cords are seen, and grade 2b, in which only the arytenoids are seen. Intubation failure occurs in 67% of grade 2b cases but in only 4% of grade 2a cases.30 Approximately 80% of all grade 2 laryngoscopies are grade 2a; the rest are grade 2b. A grade 1 view is associated with virtually 100% intubation success. An alternative system, the POGO (percentage of glottic opening) also has been proposed and validated, but is not widely used or studied.31
Confirmation of Endotracheal Tube Placement The most serious complication of endotracheal intubation is unrecognized esophageal intubation with resultant hypoxic brain injury. Although direct visualization of the ETT passing through the vocal cords generally is a reliable indicator of tracheal intubation, such clinical anatomic observations are fallible, and additional means are required to ensure correct placement of the tube within the trachea. Traditional methods, such as chest auscultation, gastric auscultation, bag resistance, exhaled volume, visualization of condensation within the ETT, and chest radiography, all are prone to failure as means of confirming tracheal intubation.32 Other clinical techniques are readily available for detecting tracheal or esophageal intubation. Immediately after intubation, the intubator should apply an end-tidal carbon dioxide (ETco2) detection device to the ETT and assess it through six manual ventilations. Disposable, colorimetric ETco2 detectors are highly reliable, convenient, and easy to interpret, indicating adequate CO2 detection by color change (Figs. 1-3 and 1-4) (see Chapter 3). ETco2 detection is highly reliable in determining tracheal and esophageal intubation in patients with spontaneous circulation.33 These devices indicate the carbon dioxide content in exhaled
7
Chapter 1 / Airway
Figure 1-3. End-tidal CO2 detector before application. The indicator is
Figure 1-4. Positive detection of CO2 turns the indicator yellow, indicating tracheal placement of the endotracheal tube.
air either qualitatively or quantitatively. The persistence of detected CO2 after six manual breaths indicates that the tube is within the airway, although not necessarily within the trachea. Gas exchange is detected with the tube in the mainstem bronchus, the trachea, or the supraglottic space. Correlation of ETco2 detection with the depth markings on the endotracheal tube (particularly important in pediatric patients) confirms tracheal placement. Rarely, BMV before intubation or ingestion of carbonated beverages may lead to release of CO2 from the stomach after esophageal intubation, causing a transient false indication of tracheal intubation. Washout of this phenomenon occurs within six breaths, however, so persistence of CO2 detection after six breaths indicates tracheal intubation. Although colorimetric ETco2 measurement is highly sensitive and specific for detecting esophageal intubation, caution is required for patients with cardiopulmonary arrest. Insufficient gas exchange may hamper CO2 detection in the exhaled air, even when the tube is correctly placed within the trachea.33 In patients with cardiopulmonary arrest, a CO2 level greater than 2%, which is the threshold for color change on colorimetric capnometers, should be considered definitive evidence of correct ETT placement, but the absence of such CO2 cannot be used reliably as an indicator of esophageal intubation. This circumstance arises in approximately 25 to 40% of intubated cardiac arrest patients.33,34 In all other patients, absence of CO2 detection indicates failure to intubate the trachea, and rapid reintubation is indicated. When possible, continuous quantitative capnography is more accurate and yields more information than capnometry (including colorimetric devices; see Chapter 3). The other method of tube placement confirmation is the aspiration technique, which is based on the anatomic differences between the trachea and the esophagus. The esophagus is a muscular structure with no support within its walls. The trachea is held patent by cartilaginous rings. Vigorous aspiration of air through the ETT with the ETT cuff deflated results in occlusion of the ETT orifices by the soft walls of the esophagus, whereas aspiration after tracheal placement of the tube is easy and rapid. Bulb or syringe aspiration devices may be used in patients with cardiac arrest who have no detectable CO2, but although such devices are highly reliable at detecting esophageal intubation (sensitivity > 95%), false-positives, in which a correctly placed tracheal tube is incorrectly identified as esophageal, can occur in up to 25% of cardiac arrest patients.33 Aspiration
devices may be useful in the out-of-hospital setting when poor lighting hampers colorimetric ETco2 determination. They also are good backup devices when cardiac arrest confounds attempts to assess placement using ETco2. Detection of expired CO2 is more reliable and should be considered the standard for confirmation of tracheal placement of an ETT and for early detection of accidental esophageal intubation. Aspiration devices have a valuable, but secondary role. Repeat laryngoscopy generally is insufficient to “confirm” that the tube is through the glottis because error and misinterpretation can occur, especially if the clinician confirming the intubation is the same person who intubated in the first place. The objective instrument (ETco2) should be considered correct. Complete obstruction of the trachea or both main stem bronchi, which prevents ventilation of the patient with even small tidal volumes, can lead to failure to detect CO2 even when the tube is in the trachea. In the absence of known or suggested complete large airway obstruction, however, failure to detect CO2 should not be ascribed to other causes, such as severe asthma, in which the physician might postulate that adequate CO2 exchange is not occurring for physiologic reasons. Absent equipment failure, this generally does not occur, and detection failure should be equated with intubation failure. Accordingly, ETco2 detection, with aspiration as backup, should be considered the primary means of ETT placement confirmation. Secondary means include physical examination findings, oximetry, and radiography. The examiner should auscultate both lung fields and the epigastric area. Auscultation of typical hollow, gurgling, gastric sounds in the epigastrium is highly suggestive of esophageal intubation and should prompt consideration of immediate reintubation. Diminished or absent breath sounds on one side (usually the left side) indicate main stem bronchus intubation, in the absence of pneumothorax or an alternative cause of unilateral loss of breath sounds. Persistent, obvious leak despite positive ETco2 detection indicates cuff malfunction or supraglottic placement of the ETT, such that the tube is in the airway, detecting CO2, but above the vocal cords. In either case (main stem bronchus intubation or supraglottic intubation), tube malpositioning can be confirmed by inspection of the depth of insertion of the tube, supplemented by chest radiography when needed. If malpositioning is detected, repositioning is indicated. Pulse oximetry is indicated as a monitoring technique in all critically ill patients, not just those who require intubation. Oximetry is useful in detecting esophageal intubation, but
purple, which indicates failure to detect CO2. This also is the appearance when the esophagus is intubated.
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
8
may not show a decreasing oxygen saturation for several minutes after a failed intubation because of the oxygen reservoir (preoxygenation) created in the patient before intubation.35 Oximetry may be particularly misleading in a spontaneously breathing patient who has had an inadvertent nasoesophageal intubation and did not have the ETco2 measured. In this case, oxygen saturation may be preserved because of spontaneous respirations, but catastrophe can ensue if the patient is later paralyzed or heavily sedated in the mistaken belief that the tube is in the trachea. Although chest radiography is universally recommended after ETT placement, its primary purpose is to ensure that the tube is well positioned below the cords and above the carina. A single anteroposterior chest radiograph is not sufficient to detect esophageal intubation, although esophageal intubation may be detected if the ETT is clearly outside the air shadow of the trachea. In cases where doubt persists, a fiberoptic scope can be passed through the ETT to identify tracheal rings, a “gold standard” for confirmation of tracheal placement.
■ MANAGEMENT
both for planning intubation and for rescue in the event of intubation failure.4 The algorithm in Figure 1-5 assumes that a decision to intubate has been made and outlines such an approach. The approach is predicated on two key determinations that must be made before active airway management is begun (see Fig. 1-5). The first determination is whether the patient is in cardiopulmonary arrest or a state near to arrest and is predicted to be unresponsive to direct laryngoscopy. Such a patient (agonal, near death, circulatory collapse) is called a “crash airway” patient for the purposes of emergency airway management and is treated using the crash airway algorithm by immediate intubation without use of drugs, supplemented by a single dose of succinylcholine if the attempt to intubate fails and the patient is felt not to be sufficiently relaxed (Fig. 1-6). Next, it must be determined whether the patient represents a difficult intubation as determined by the LEMON, MOANS, and RODS evaluations. If so, the difficult airway algorithm is used (Fig. 1-7). For all other cases, that is, for all patients who require emergency intubation but who have neither a crash airway nor a difficult airway, RSI is recommended. RSI provides the safest and quickest method of achieving intubation in such
Approach to Intubation After it is determined that the patient requires intubation, an approach must be planned. Algorithms for emergency airway management have been developed and provide a useful guide,
Maintain oxygenation
Needs intubation
Unresponsive? Near death?
Yes
Crash airway
From difficult airway
Intubation attempt successful?
Yes
Difficult airway
Unable to bag ventilate?
RSI
Succinylcholine 2 mg/kg IVP
Attempt intubation
Attempt intubation Yes
Postintubation management
No Failure to maintain oxygenation?
Yes
Failed airway
Successful?
Yes
Postintubation management
No Yes
Failed airway
No
≥ 3 attempts at OTI by experienced operator?
Postintubation management
No
No
Successful?
Yes
No
No Predict difficult airway?
Crash airway
Failure to maintain oxygenation?
Yes
Failed airway
No Yes
No
Figure 1-5. Main emergency airway management algorithm. OTI, orotracheal intubation; RSI, rapid sequence intubation. (Adapted from Walls RM: The emergency airway algorithms. In Walls RM, Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, p. 11, 2008. Copyright © 2008 The Difficult Airway Course: Emergency and Lippincott Williams & Wilkins.)
≥ 3 attempts by experienced operator?
Yes
No
Figure 1-6. Crash airway algorithm. IVP, intravenous push. (Adapted from Walls RM: The emergency airway algorithms. In Walls RM, Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, p. 14, 2008. Copyright © 2008 The Difficult Airway Course: Emergency and Lippincott Williams & Wilkins.)
9 Difficult airway predicted
Yes
Call for assistance Extraglottic device may be attempted
Failed airway
Failure to maintain oxygenation?
No
No BMV or EGD predicted to be successful?
Yes
Yes
RSI*
No
No Awake DL, FO, or VL successful?
Intubation predicted to be successful?
Yes
Postintubation management or RSI
Cuffed ETT placed? Go to main algorithm
Figure 1-7. Difficult airway algorithm. BMV, bag-mask ventilation; BNTI,
blind nasotracheal intubation; DL, direct laryngoscopy; EGD, extraglottic device; FO, fiberoptic laryngoscopy; ILMA, intubating laryngeal mask airway; RSI, rapid sequence intubation; VL, video laryngoscopy. (Adapted from Walls RM: The emergency airway algorithms. In Walls RM, Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, p. 15, 2008. Copyright © 2008 The Difficult Airway Course: Emergency and Lippincott Williams & Wilkins.) *May require double set-up. † If not done earlier.
patients.1,20,36,37 After administration of the RSI drugs, intubation attempts are repeated until the patient is intubated or a failed intubation is identified. If more than one intubation attempt is required, oxygen saturation is monitored continuously, and if saturation falls to 90% or less, BMV is performed until saturation is recovered for another attempt. If the clinician cannot maintain oxygen saturation with BMV, despite optimal use of a two-person, two-handed technique with an oral airway in place, a failed airway exists. This is referred to as a “can’t intubate, can’t oxygenate” situation. In addition, if three attempts at direct laryngoscopy have been unsuccessful, a failed airway exists because subsequent attempts at laryngoscopy by the same clinician are unlikely to succeed. The three failed laryngoscopy attempts are defined as attempts by an experienced clinician, using best possible patient positioning and technique. A further attempt at direct laryngoscopy by the same clinician or one of equivalent experience is inadvisable, unless the clinician identifies a specific situation on the third laryngoscopy that is amenable to correction, justifying a fourth attempt. Also, if the clinician ascertains after even a single attempt that intubation will be impossible (e.g., grade IV laryngoscopic view despite optimal patient positioning), a failed airway is present. The failed airway is managed according to the failed airway algorithm (Fig. 1-8).
Difficult Airway When preintubation evaluation has identified a potentially difficult airway, a different approach is used (see Fig. 1-7).3,4 The approach is based on the fact that NMBAs should not be administered to a patient for intubation unless the clinician believes that (1) intubation is likely to be successful and (2)
Cricothyrotomy
If contraindicated
Choose one of: Fiberoptic method Video laryngoscopy Extraglottic device Lighted stylet Cricothyrotomy
No ILMA FO or VL† Cricothyrotomy BNTI, lighted stylet
Yes
Yes
Postintubation management
No Arrange for definitive airway management
Figure 1-8. Failed airway algorithm. ETT, endotracheal tube. (Adapted
from Walls RM: The emergency airway algorithms. In Walls RM, Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, p. 18, 2008. Copyright © 2008 The Difficult Airway Course: Emergency and Lippincott Williams & Wilkins.)
oxygenation via BMV or EGD is likely to be successful if a first intubation attempt does not succeed and oxygenation is required. The perception of a difficult airway is relative, and many emergency intubations could be considered “difficult.” The judgment regarding whether to treat the airway as a typical emergency airway or whether to use the difficult airway algorithm is based on the degree of perceived difficulty and the individual circumstances of the case.38 The LEMON, MOANS, and RODS assessments provide a systematic framework to assist in identifying the potentially difficult airway. When a difficult airway is identified, the first step is to ensure that oxygenation is sufficient to permit a planned, orderly approach (see Fig. 1-7). If oxygenation is inadequate and cannot be made adequate by supplementation with bag and mask, the airway should be considered a failed airway. The failed airway algorithm should be used because the predicted high degree of intubation difficulty combined with failure to maintain oxygen saturation is analogous to the “can’t intubate, can’t oxygenate” situation. When oxygenation is adequate, the next consideration is whether RSI is appropriate, based on the operator’s assessment of the likelihood of (1) successful ventilation using a bag and mask or an EGD in the event intubation is unsuccessful, and (2) the likelihood of successful intubation by direct laryngoscopy. In some cases, a double setup can be used in which RSI is performed, but all preparations are undertaken for rescue cricothyrotomy before the drugs are administered. If RSI is not advisable, an “awake” technique can be used. In this context, awake means that the patient continues to breathe and is able to respond to caregivers. Usually the technique involves sedation and topical anesthesia, often preceded by a drying agent, such as glycopyrrolate.
Chapter 1 / Airway
Failure to maintain oxygenation?
Failed airway criteria
Call for assistance
The awake technique often is direct laryngoscopy, assisted by topical anesthesia and sedation (comparable to that for a painful procedure), with the purpose of ascertaining whether intubation using direct laryngoscopy is possible. If the glottis is adequately visualized, the patient can be intubated at that time, or, in a stable difficult airway situation, the clinician may proceed with planned RSI, now assured of intubation success. Awake laryngoscopy can be performed using a direct laryngoscope, a flexible fiberoptic scope, a videolaryngoscope, or a rigid fiberoptic scope. If the awake laryngoscopy determines that oral intubation using a standard laryngoscope would likely be unsuccessful, the patient is intubated using any of numerous techniques shown in the last box in Figure 1-7. For each of these methods, the patient is kept breathing but variably sedated and anesthetized and each of the methods results in placement of a cuffed ETT in the trachea. The choice among these methods depends on clinician experience and preference, device availability, and patient attributes.
Failed Airway Management of the failed airway is dictated by an assessment of whether the patient can be oxygenated.3,4 If adequate oxygenation cannot be maintained, the rescue technique of first resort is cricothyrotomy (see Fig. 1-8). Multiple attempts at other methods in the context of failed oxygenation delay cricothyrotomy and place the patient at increased risk for hypoxic brain injury. If an alternative device (i.e., an EGD such as a laryngeal mask airway or Combitube) is readily at hand, however, an attempt can be made to use it simultaneously with preparations for immediate cricothyrotomy, as long as initiation of cricothyrotomy is not delayed. Only a single attempt with the EGD is recommended in this circumstance. If adequate oxygenation is possible, several options are available for the failed airway. In almost all cases, cricothyrotomy is the definitive rescue technique for the failed airway if time (i.e., preservation of oxygenation) does not allow for other approaches or if they fail. The fundamental difference in philosophy between the difficult airway and the failed airway is that the difficult airway is planned for, and the standard is to place a cuffed ETT in the trachea. The failed airway is not planned for, and the standard is to achieve an airway that provides adequate oxygenation to avert the immediate problem of hypoxic brain injury. Some of the devices used in the failed airway (e.g., EGDs) are temporary and do not provide airway protection.
■ THERAPEUTIC MODALITIES
authors correctly noted that virtually no studies have ever been designed to measure this precise endpoint.40 RSI is nevertheless the most widely used technique by far for emergency intubation of patients without identifiable difficult airway attributes.1,13,14 The central concept of RSI is to take the patient from the starting point (e.g., conscious, breathing spontaneously) to a state of unconsciousness with complete neuromuscular paralysis, then to achieve intubation without interposed assisted ventilation. The risk of aspiration of gastric contents is felt to be significantly higher for patients who have not fasted before induction. Application of positive-pressure ventilation can cause air to pass into the stomach, resulting in gastric distention and likely increasing the risk of regurgitation and aspiration.41 The purpose of RSI is to avoid positive-pressure ventilation until the ETT is placed correctly in the trachea with the cuff inflated. This requires a preoxygenation phase, during which the nitrogen reservoir in the functional residual capacity in the lungs is replaced with oxygen, permitting at least several minutes of apnea (see later discussion) in the normal adult before oxygen desaturation to 90% ensues (Fig. 1-9).35 Use of RSI also facilitates successful endotracheal intubation by causing complete relaxation of the patient’s musculature, allowing better access to the airway.19,36,37,42 Finally, RSI permits pharmacologic control of the physiologic responses to laryngoscopy and intubation, mitigating potential adverse effects. These effects include further intracranial pressure (ICP) increase in response to the procedure and to the sympathetic discharge resulting from laryngoscopy (Box 1-4).43 RSI is a series of discrete steps, and every step should be planned (see Box 1-5).5 Preparation. In the initial phase, the patient is assessed for intubation difficulty (unless this has already been done), and the intubation is planned, including determining dosages and
100
90 SaO2 (%)
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
10
80
70
Mean time to recovery of twitch height from 1 mg/kg succinylcholine IV
Methods of Intubation Although many techniques are available for intubation of the emergency patient, four methods are most common, with RSI being the most frequently used in nonarrested patients.1,13,14,39
Rapid Sequence Intubation RSI is the cornerstone of modern emergency airway management and is defined as the virtually simultaneous administration of a potent sedative (induction) agent and an NMBA, usually succinylcholine, for the purpose of endotracheal intubation. This approach provides optimal intubating conditions and has long been believed to minimize the risk of aspiration of gastric contents. A systematic review of the literature in 2007 failed to prove that rapid sequence intubation results in a lower incidence of aspiration than other techniques, but the
60 0
10% 0
1
2
3
4
5
6
7 6.8
50% 8
9 8.5
90% 10 10.2
⋅ Time of VE = 0 (min) Obese 127-kg adult Normal 10-kg child
Normal 70-kg adult Moderately ill 70-kg adult
Figure 1-9. Desaturation time for apneic, fully preoxygenated patients.
Children, patients with comorbidity, and obese patients desaturate much more rapidly than healthy, normal adults. The box on the lower righthand side of the graph depicts time to recovery from succinylcholine, which in almost all cases exceeds safe apnea time. Note also the precipitous decline of oxygen saturation from 90% to 0% for all groups. Modified from Benumof J, et al: Critical hemoglobin desaturation will occur before return to unparalyzed state following 1 mg/kg intravenous succinylcholine. Anesthesiology 87:979, 1997.
BOX 1-4
Reactive airways disease: Lidocaine: 1.5 mg/kg IV, to mitigate bronchospasm. Albuterol 2.5 mg by nebulizer (if time permits and not already given). Cardiovascular disease: Fentanyl: 3 µg/kg to mitigate sympathetic discharge. Elevated ICP: Lidocaine: 1.5 mg/kg IV to mitigate ICP increase in response to airway manipulation. Fentanyl 3 µg/kg to mitigate sympathetic discharge and attendant rise in ICP. *Given 3 minutes before induction and paralysis. ICP, intracranial pressure.
BOX 1-5
The Seven “Ps” of RSI
1. Preparation 2. Preoxygenation 3. Pretreatment 4. Paralysis with induction 5. Positioning 6. Placement of tube 7. Postintubation management sequence of drugs, tube size, and laryngoscope type, blade and size. Drugs are drawn up and labeled. All necessary equipment is assembled. All such patients require continuous cardiac monitoring and pulse oximetry. At least one and preferably two good-quality intravenous (IV) lines should be established. Redundancy is always desirable in case of equipment or IV access failure. Preoxygenation. Administration of 100% oxygen for 3 minutes of normal, tidal volume breathing in a normal, healthy adult establishes an adequate oxygen reservoir to permit 8 minutes of apnea before oxygen desaturation to less than 90% occurs (see Fig. 1-9).35 The time to desaturation to less than 90% in children, obese adults, late-term pregnant women, and patients with significant comorbidity is considerably less. Desaturation time also is reduced if the patient does not inspire 100% oxygen.44 Nevertheless, adequate preoxygenation usually can be obtained, even in ED patients, to permit several minutes of apnea before oxygen desaturation to less than 90% occurs. In children and adults, preoxygenation is essential to the “no bagging” approach of RSI. If time is insufficient for a full 3-minute preoxygenation phase, eight vital capacity breaths using high-flow oxygen can achieve oxygen saturations and apnea times that match or exceed those obtained with traditional preoxygenation.45 Preoxygenation of obese patients in the head up position results in significantly longer (approximately 45 seconds) apnea time before critical saturation.46 Preoxygenation should be done in parallel with the preparation phase and can be started in the field for high risk patients. Oxygen saturation monitors permit earlier detection of desaturation during laryngoscopy, but preoxygenation remains an essential step in RSI. Pretreatment. During this phase, drugs are administered 3 minutes before administration of the succinylcholine and induction agent to mitigate the effects of laryngoscopy and intubation on the patient’s presenting or comorbid conditions. Intubation is intensely stimulating and results in sympathetic discharge (the reflex sympathetic response to laryngoscopy), elevation of ICP in patients with ICP disturbance, and reactive bronchospasm. Bradycardia often occurs in children, particu-
11
Chapter 1 / Airway
Pretreatment Agents for Rapid Sequence Intubation*
larly young children, but appears multifactorial, likely involving both parasympathetic discharge in response to airway instrumentation and perhaps some contributory effect of succinylcholine. Pretreatment focuses on three main objectives, in certain at-risk patients. The three groups of patients at risk are those with reactive airways disease, elevated ICP, or a cardiovascular or neurovascular condition or acute event for which an acute elevation in blood pressure and heart rate might be hazardous. Patients with reactive airways disease often experience a worsening of their bronchospasm when intubated. Controversy exists regarding whether albuterol alone, lidocaine alone, or both drugs together are effective in reducing this intubationrelated bronchospasm.47-49 Asthmatic patients being intubated in the ED for status asthmaticus will have received albuterol before intubation, and, pending larger studies, it is reasonable also to administer lidocaine (1.5 mg/kg) as a pretreatment drug in these cases. When an asthmatic patient is being intubated for a condition (e.g., trauma) other than acute asthma, nebulized albuterol and IV lidocaine should be given before intubation, if possible. Patients with significant cardiovascular disease (e.g., ischemic coronary disease) who are being intubated in the ED may benefit from the administration of the synthetic opioid, fentanyl, in a dose of 3 µg/kg to mitigate the release of catecholamines in response to airway manipulation. Similarly, patients with intracranial hemorrhage, elevated ICP, or marked hypertension may benefit from pretreatment with fentanyl.49 Finally, there is some evidence that patients with elevated ICP may experience less exacerbation of the ICP during intubation if they are pretreated with lidocaine (1.5 mg/kg). These patients, unless hypotensive, should also receive fentanyl (3 µg/kg) to mitigate blood pressure surges that might translate to further increases in ICP. There is evidence supporting the physiologic effects of these agents, but outcome data are lacking. Individualization is necessary, and critical time should not be lost administering pretreatment drugs if the patient requires immediate intubation. Despite the lack of outcome studies, considerable inferential evidence supports this approach, and these agents probably provide protection for vulnerable patients against the adverse hemodynamic and intracranial effects of laryngoscopy and intubation.49 Although many variations are possible for pretreatment regimens in various conditions, pretreatment can be simplified to these three basic indications (see Box 1-4). When possible, 3 minutes should elapse between the administration of the pretreatment drug and the administration of the induction drug and NMBA. If time is insufficient to wait 3 minutes, even a reduced time may provide some benefit. Paralysis with Induction. In this phase, a potent sedative agent is administered by rapid IV push in a dose capable of rapidly producing unconsciousness. This is immediately followed by rapid administration of an intubating dose of an NMBA, usually succinylcholine. It is usual to wait 45 seconds from the time the succinylcholine is given to allow sufficient paralysis to occur. (See later discussion of drugs and doses.) Positioning. The patient should be positioned for intubation as consciousness is lost. Usually, positioning involves head extension, often with flexion of the neck on the body, but there is evidence that simple extension of the head alone, or extension of both the head and neck (the extension-extension position) are equivalent or superior.27,28 (See earlier discussion.) Sellick’s maneuver (application of firm backwarddirected pressure over the cricoid cartilage) has long been recommended to minimize the risk of passive regurgitation and, hence, aspiration, but two recent reviews have challenged this premise.50-52 In addition, there is evidence that Sellick’s maneuver may make laryngoscopy or intubation more difficult
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
12
in some patients.51 Accordingly, Sellick’s maneuver should be considered optional, applied selectively, and released or modified to improve laryngeal view or tube passage, as indicated. During this phase after administration of the induction agent and NMBA, although the patient becomes unconscious and apneic, BMV should not be initiated unless the oxygen saturation falls to 90%. Placement of Tube. Approximately 45 seconds after the administration of succinylcholine, the patient is relaxed sufficiently to permit laryngoscopy; this is assessed most easily by moving the mandible to test for absence of muscle tone. The ETT is placed under direct visualization of the glottis. If the first attempt is unsuccessful, but oxygen saturation remains high, it is not necessary to ventilate the patient with a bag and mask between intubation attempts. If the oxygen saturation is approaching 90%, the patient may be ventilated briefly with a bag and mask between attempts to reestablish the oxygen reservoir. When BMV is performed, Sellick’s maneuver is advisable to minimize passage of air into the stomach.41 Sellick’s maneuver may be continued or released during repeat laryngoscopy, according the judgment of the clinician and the glottic view obtained. As soon as the ETT is placed, the cuff should be inflated and its position confirmed as described earlier. Postintubation Management. A chest radiograph should be obtained to confirm that main stem intubation has not occurred and to assess the lungs. There is a trend away from the use of long-acting NMBAs (e.g., pancuronium, vecuronium) toward optimal management using opioid analgesics and sedative agents to facilitate mechanical ventilation.50 (See Chapter 3.) An adequate dose of a benzodiazepine (e.g., midazolam 0.1–0.2 mg/kg, IV) and an opioid analgesic (e.g., fentanyl, 3–5 µg/kg, IV, or morphine, 0.2–0.3 mg/kg, IV) is given to improve patient comfort and decrease sympathetic response to the ETT. Appropriate use of sedation and analgesia often obviates the need for an NMBA. Table 1-1 presents a sample RSI protocol using etomidate and succinylcholine. “Zero” refers to the time at which the induction agent and succinylcholine are pushed.
Table 1-1 Sample Rapid Sequence Intubation Using Etomidate and Succinylcholine
TIME
STEP
Zero minus 10 min Zero minus 5 min
Preparation Preoxygenation 100% oxygen for 3 min or eight vital capacity breaths Pretreatment as indicated Paralysis with induction Etomidate, 0.3 mg/kg Succinylcholine, 1.5 mg/kg Positioning Sellick’s maneuver optional Placement Laryngoscopy and intubation End-tidal carbon dioxide confirmation Postintubation management Sedation and analgesia as indicated Initiate mechanical ventilation NMBA only if needed after adequate sedation/analgesia
Zero minus 3 min Zero
Zero plus 30 sec Zero plus 45 sec Zero plus 2 min
Blind Nasotracheal Intubation Historically, blind nasotracheal intubation (BNTI) was used extensively in the ED and out-of-hospital setting, but has fallen out of favor largely because of the superiority of RSI.11,53 Success rates have been about 80 to 90%, and high complication rates are reported, most often epistaxis or delayed or incorrect tube placement.53 Long-term complications (e.g., sinusitis, turbinate destruction, laryngeal perforation) are uncommon and related to multiple attempts or prolonged intubation. Basilar skull fracture and facial trauma have been considered contraindications to nasotracheal intubation because of the risk of entering the cranial vault or increasing the incidence of intracranial infection. These contraindications are not based on scientific study, however, and two small studies failed to detect a difference in complications between orally and nasally intubated facial trauma patients.54,55 Two other studies compared the success rates of RSI and BNTI performed by physicians or paramedics on helicopter services. Results differed, with one study showing essentially equivalent success rates and the other showing a significant advantage for neuromuscular blockade over BNTI.56,57 ED studies have shown superiority of RSI over BNTI.11,39 Also, the incidence and severity of oxygen desaturation are greater in BNTI than with RSI.58 BNTI is a valid and useful method of intubation in the out-of-hospital setting and is still widely used by paramedics and other out-of-hospital first responders. In the ED, where NMBAs and RSI are available, BNTI should be considered a second-line approach and reserved for patients in whom the presence of a difficult airway makes RSI undesirable or contraindicated and alternatives (e.g., fiberoptics) are not available. Interestingly, the old recommendation that refrigeration of the tube before use increases success of nasotracheal intubation probably is not true. To the contrary, warming the tube to 40° before use appears to facilitate easy tube passage and reduce the incidence of epistaxis.59 Similarly, maintaining the head in a neutral position and inflating the ETT cuff to 15 mL in the oropharynx or hypopharynx before attempting to traverse the glottis also improves the success rate.60 Use of BNTI in the ED has declined sufficiently, and it is doubtful that emergency medicine residents will be adequately trained in the technique.53
Awake Oral Intubation Awake oral intubation is a technique in which sedative and topical anesthetic agents are administered to permit management of a difficult airway. Sedation and analgesia are achieved in a manner analogous to that for painful procedures in the ED. Topical anesthesia may be achieved by spray, nebulization, or local anesthetic nerve block. After the patient is sedated and topical anesthesia has been achieved, gentle direct, video, or fiberoptic laryngoscopy is performed to determine whether the glottis is visible and intubation possible. The patient may be intubated during the laryngoscopy, or the laryngoscopy may show that oral intubation is possible, permitting safe use of RSI (see earlier discussion). Awake oral intubation is distinct from the practice of oral intubation using a sedative or opioid agent to obtund the patient for intubation without neuromuscular blockade, which had been a typical ED practice. This latter technique can be referred to as “intubation with sedation alone” or, paradoxically, “nonparalytic RSI.” Proponents of intubation with sedation alone argue that administration of a benzodiazepine, opioid, or both provides improved access to the airway, decreases patient resistance, and avoids the risks inherent
Oral Intubation without Pharmacologic Agents The unconscious, unresponsive, near death patient may not require pharmacologic agents for intubation. If the patient is essentially dead, administration of any pharmacologic agent, including an NMBA, may needlessly delay intubation. Even an unconscious patient may retain sufficient muscle tone to render intubation difficult, however. If the glottis is not adequately visualized, administration of a single dose of succinylcholine alone may facilitate laryngoscopy. Success rates for intubating unconscious, unresponsive patients are comparable to those achieved with RSI, presumably because the patient is in a similar physiologic state (i.e., muscle relaxation, no ability to react to laryngoscopy or tube insertion).1
Pharmacologic Agents Neuromuscular Blocking Agents Muscle contraction is the result of membrane depolarization, which causes massive intracellular release of calcium ions from the sarcoplasmic reticulum, leading to active contraction of myofibrils. The inciting incident is the depolarization of portions of the myocyte membrane, called the motor endplates, which are adjacent to the innervating axons. Action potentials conducted down the innervating axons cause release of the neurotransmitter acetylcholine (ACh) from the terminal axon. The ACh traverses the synaptic cleft, binds reversibly to receptors on the motor endplate, and opens channels in the membrane to initiate depolarization. NMBAs are highly water-soluble, quaternary ammonium compounds that mimic the quaternary ammonium group on the ACh molecule. Their water solubility explains why these agents do not readily cross the blood-brain barrier or placenta. The NMBAs are divided into two main classes. The depolarizing agent, succinylcholine, exerts its effects by binding noncompetitively with ACh receptors on the motor endplate and causing sustained depolarization of the myocyte. The other major class of NMBA comprises the competitive, or nondepolarizing, agents, which bind competitively to ACh receptors, preventing access by ACh and preventing muscular activity. The competitive agents are of two pharmacologically distinct types, steroid-based agents (aminosteroid compounds) and benzylisoquinolines. Each of these basic chemical types has distinct properties, but only the aminosteroid compounds are used in the ED. Succinylcholine. Succinylcholine is a combination of two molecules of ACh. Succinylcholine is rapidly hydrolyzed by plasma pseudocholinesterase to succinylmonocholine, which is a weak NMBA, then to succinic acid and choline, which have no NMBA activity. Pseudocholinesterase is not present at the motor endplate and exerts its effects systemically before the succinylcholine reaches the ACh receptor.64 Only a small amount of the succinylcholine that is administered survives to
reach the motor endplate. When attached to the ACh receptor, succinylcholine is active until it diffuses away. Decreased plasma pseudocholinesterase activity can increase the amount of succinylcholine reaching the motor endplate, prolonging succinylcholine block, but this is of little significance in the emergency setting because the prolongation of action is rarely significant, reaching only 23 minutes at the extreme.64,65 Uses. Succinylcholine is rapidly active, typically producing intubating conditions within 60 seconds of administration by rapid IV bolus injection.37,66 The clinical duration of action before spontaneous respiration is 6 to 10 minutes (see Fig. 19).35 Full recovery of normal neuromuscular function occurs within 15 minutes. The combination of rapid onset, complete reliability, short duration of action, and absence of serious side effects maintains succinylcholine as the drug of choice for most ED intubations.1,13,50,62 The use of a competitive, or nondepolarizing, NMBA for RSI may be desirable when succinylcholine is contraindicated and in certain other settings. Cardiovascular Effects. As an ACh analogue, succinylcholine binds to ACh receptors throughout the body, not just at the motor endplate. It is difficult to separate the effects of succinylcholine on the heart that are caused by direct cardiac muscarinic stimulation from those caused by stimulation of autonomic ganglia by succinylcholine and from the effects induced by the autonomic responses to laryngoscopy and intubation. Succinylcholine can be a negative chronotrope, especially in children, and sinus bradycardia may ensue after succinylcholine administration. Sinus bradycardia is treated with atropine, if necessary, but is often self-limiting. Some pediatric practi tioners recommend pretreatment with atropine for children younger than 1 year old, but there is no evidence for benefit.67 Other cardiac dysrhythmias, including ventricular fibrillation and asystole, have been reported with succinylcholine, but it is impossible to distinguish the effects of the drug itself from those caused by the intense vagal stimulation and catecholamine release that accompany laryngoscopy and intubation. In addition, many of these catastrophic complications occur in critically ill patients, further confounding attempts to identify whether the illness or any particular drug or procedure is the cause. Fasciculations. The depolarizing action of succinylcholine results in fine, chaotic contractions of the muscles throughout the body for several seconds at the onset of paralysis in over 90% of patients. Muscle pain occurs in approximately 50% of patients who receive succinylcholine. Although it is widely believed that muscle pains are reduced or abolished by prior administration of a defasciculating dose of a competitive NMBA, the evidence is not conclusive.68 Use of 1.5 mg/kg of succinylcholine results in both less fasciculation and less myalgia than occur with 1 mg/kg.68 Hyperkalemia. Succinylcholine has been associated with severe, fatal hyperkalemia when administered in specific clinical circumstances (Table 1-2).69 Although the hyperkalemia occurs within minutes after administration of succinylcholine and may be severe or fatal, the patient’s vulnerability to succinylcholine-induced hyperkalemia does not become significant until at least 5 days after the inciting injury or burn. Succinylcholine remains the agent of choice for RSI in acute burn, trauma, stroke, spinal cord injury, and intra-abdominal sepsis if intubation occurs less than 5 days after onset of the condition. If doubt exists regarding the onset time, succinylcholine should be replaced with a competitive NMBA, usually rocuronium. Denervation syndromes (e.g., multiple sclerosis, amyotrophic lateral sclerosis) can be particularly troubling, however, because the risk begins with the onset of the disease and continues indefinitely, regardless of the apparent stability of the symptoms. Patients who have denervation caused by
13
Chapter 1 / Airway
in neuromuscular blockade. This technique actually is more hazardous than RSI, however. Intubating conditions achieved even with deep anesthesia are significantly inferior to the conditions achieved when neuromuscular blockade is used.36,37,61 The same superiority of neuromuscular blockade-assisted intubation over intubation with sedation alone has been observed in pediatric emergency medicine and in EMS care.62,63 In general, the technique of administering a potent sedative agent to obtund the patient’s responses and permit intubation in the absence of neuromuscular blockade is ill-advised and inappropriate for ETI in the ED, unless it is performed as part of an “awake” intubation as described earlier.
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
14
stroke or spinal cord injury are stabilized after 6 months, and thereafter can receive succinylcholine safely.65 Potassium release does not occur to any significant extent in the general population. Succinylcholine is not contraindicated in renal failure but probably should not be used in patients with known or presumed hyperkalemia sufficient to manifest on the electrocardiogram. The only published series of patients with hyperkalemia, many of whom had renal failure, failed to show a single adverse event related to succinylcholine administration.70 Increased Intraocular Pressure. Succinylcholine may cause a modest increase in intraocular pressure and historically has been considered relatively to absolutely contraindicated in penetrating globe injury. There is no published evidence to support this view, however, and several large series show safety when succinylcholine is used in patients with open globes. The admonition to avoid succinylcholine in open globe injuries is unjustified and should be abandoned.71 Masseter Spasm. Succinylcholine has been reported rarely to cause masseter spasm, primarily in children.64 The clinical significance of this phenomenon is unclear, but administration of a competitive NMBA terminates the spasm. Severe, persistent spasm should raise suspicion of malignant hyperthermia. Malignant Hyperthermia. Succinylcholine has been associated with malignant hyperthermia, a perplexing syndrome of rapid temperature rise and aggressive rhabdomyolysis. Malignant hyperthermia occurs in genetically predisposed individuals who receive certain volatile anesthetic agents or succinylcholine. The condition is extremely rare and has not been reported in the context of ED intubation. Treatment consists of cessation of any potential offending agents, administration of dantrolene (2 mg/kg IV every 5 min to a maximum dose of 10 mg/kg), and
Conditions Associated with Hyperkalemia after Table 1-2 Succinylcholine Administration CONDITION
PERIOD OF CONCERN
Burns >10% BSA Crush injury Denervation (stroke, spinal cord injury) Neuromuscular disease (ALS, MS) Intra-abdominal sepsis
>5 days until healed >5 days until healed >5 days until 6 months postinjury Indefinitely >5 days until resolution
ALS, amyotrophic lateral sclerosis; BSA, body surface area; MS, multiple sclerosis.
attempts to reduce body temperature by external means.72 A national malignant hyperthermia hotline is available for emergency consultation at 1-800-644-9737 (then dial zero). Refrigeration. The standard recommendation to keep succinylcholine refrigerated creates problems related to its storage, timely retrieval, and ready availability on intubation carts or kits in the ED. Succinylcholine undergoes degradation beginning at the time of manufacture, and the rate of this degradation is much lower when the drug is refrigerated. Succinylcholine retains more than 90% of its original activity when stored at room temperature for 3 months; it retains even more if protected from light.73 Succinylcholine may be kept at room temperature in the ED or EMS setting, provided that a proper inventory control system ensures that all supplies are replaced not more than 3 months after introduction. Competitive Agents. Competitive NMBAs are classified according to their chemical structure. The aminosteroid agents include pancuronium, vecuronium, and rocuronium. Vecuronium neither releases histamine nor exhibits cardiac muscarinic blockade and is an excellent agent for maintenance of neuromuscular blockade when this is desirable. Rocuronium is the best agent for use in RSI when succinylcholine is contraindicated. Rapid Sequence Intubation with a Competitive Agent. Competitive agents, especially vecuronium and rocuronium, have been studied extensively for RSI. Although vecuronium was the first competitive NMBA to establish a role in RSI, it works best when given as a split dose. First, 0.01 mg/kg is administered as a “priming” dose. Three minutes later, 0.15 mg/kg is given for paralysis, which is achieved in about 75 to 90 seconds. Rocuronium bromide (1 mg/kg IV) achieves intubating conditions closely approaching those of succinylcholine, lasts approximately 50 minutes, and has been used in the ED with success (Table 1-3).61,74 Paralysis after Intubation. After intubation, prolonged paralysis may be desired to optimize mechanical ventilation; however, current management trends are away from the use of prolonged paralysis in favor of deep sedation with analgesia. If neuromuscular blockade is desired, vecuronium (0.1 mg/kg IV) can be given. Longer term neuromuscular blockade must not be undertaken without attention to appropriate sedation and analgesia of the patient.50 An adequate dose of a benzodiazepine, such as midazolam (0.1–0.2 mg/kg IV), and an opioid analgesic, such as fentanyl (3–5 µg/kg IV) or morphine (0.2– 0.3 mg/kg IV), is required to improve patient comfort and decrease sympathetic response to the ETT. Appropriate use
Table 1-3 Sample Rapid Sequence Intubation Using Etomidate and Rocuronium TIME
STEP
Zero minus 10 min Zero minus 5 min
Preparation Preoxygenation 100% oxygen for 3 min or eight vital capacity breaths Pretreatment As indicated Paralysis with induction Etomidate, 0.3 mg/kg Rocuronium, 1.0 mg/kg
Zero minus 3 min Zero
Zero plus 30 sec Zero plus 60 sec Zero plus 2 min
Positioning Placement Laryngoscopy and intubation End-tidal carbon dioxide confirmation Postintubation management Sedation and analgesia mandatory because of prolonged (45 min) duration of paralysis with rocuronium
Induction Agents A patient who presents with any degree of clinical responsiveness, including reactivity to noxious stimuli, requires a sedative or induction agent at the time of administration of any NMBA. Patients who already are deeply unconscious and unresponsive may not require a full dose of an induction agent if the unconscious state is caused by drugs or alcohol (themselves general anesthetic agents.) Patients who are unconscious because of a central nervous system insult should receive an induction agent to attenuate adverse responses to airway manipulation. Induction agents also enhance the effect of the NMBA and improve intubation conditions because the intubation is done at the earliest phase of neuromuscular blockade, and the relaxation effects of the induction agent are additive to those of the NMBA.75 Etomidate. Etomidate is an imidazole derivative that has been in use since 1972. Its activity profile is similar to that for thiopental, with rapid onset, rapid peak activity, and brief duration, but it is remarkably hemodynamically stable.76,77 Etomidate has emerged as the agent of choice for ED RSI, and numerous reports attest to its effectiveness and safety.1,14 The induction dose is 0.3 mg/kg IV. Because etomidate is able to decrease ICP, cerebral blood flow, and cerebral metabolic rate without adversely affecting systemic mean arterial blood pressure and cerebral perfusion pressure, it is an excellent induction agent for patients with elevated ICP, even in cases of hemodynamic instability.77,78 Etomidate may cause brief myoclonus, but this is of no clinical significance. Etomidate by continuous infusion has been reported to cause suppression of endogenous cortisol production. Recently, controversy has emerged regarding the role of etomidate for intubation of patients with septic shock.79-81 Several retrospective studies have claimed to demonstrate that etomidate, used in a single dose for intubation, causes suppression of the adrenal response to exogenously administered adrenocorticotropic hormone, and have attempted to link this to increased mortality.82,83 Other retrospective studies have shown the opposite.84,85 Ironically, much of the original criticism of etomidate arose from the belief that adrenocortical response to exogenous corticotropin predicts outcome in patients with septic shock, a belief that has since been abandoned.86 Also, the most recent and comprehensive study of the role of corticosteroids in septic shock failed to show any benefit, casting further doubt about any possible mortality effect of a single dose of etomidate.87 Pending a properly constructed, prospective, randomized clinical trial, there is not sufficient evidence to support a recommendation that etomidate not be used in patients with septic shock.81,88 In fact, etomidate’s superior hemodynamic profile makes it an excellent choice in these generally unstable patients. Barbiturates. Although both the thiobarbiturate, sodium thiopental, and the methylated oxybarbiturate, methohexital, have been used as induction agents for RSI, thiopental has been used more widely. The use of these agents has declined significantly, however, with the adoption of newer agents, particularly etomidate and propofol. The rapidly acting barbiturates are highly lipid-soluble and readily cross the bloodbrain barrier, acting on the γ-aminobutyric acid receptor neuroinhibitory complex to rapidly depress central nervous system activity. A single dose of 3 mg/kg of thiopental pro-
duces loss of consciousness in less than 30 seconds, has a peak effect at 1 minute, and has a clinical duration of 5 to 8 minutes. Methohexital may have a slightly shorter duration of action but is more prone to cause central nervous system excitatory side effects, such as myoclonus. Thiopental is a negative inotrope and a potent venodilator and should be used with caution in patients whose cardiovascular reserve is diminished. For the same reason, thiopental should be avoided in a hypotensive patient who would not tolerate further compromise of circulation. Thiopental can release histamine and probably should not be used in asthmatic patients. Benzodiazepines. Of the benzodiazepines, only midazolam is suited to use as an induction agent, with a normal induction dose of 0.2 to 0.3 mg/kg IV.77 In a dose of 0.3 mg/kg IV, midazolam produces loss of consciousness in about 30 seconds and has a clinical duration of 15 to 20 minutes.89 Midazolam is a negative inotrope comparable to thiopental and should be used with caution in hemodynamically compromised and elderly patients, for whom the dose can be reduced to 0.1 mg/kg or 0.05 mg/kg. Onset is slower at these reduced doses. Much lower doses than indicated are often used in ED intubations, perhaps because practitioners are familiar with the sedation doses, but not the anesthetic induction doses, of midazolam.90 These inadequate doses reduce the effectiveness of laryngoscopy, do not provide optimal blunting of adverse physiologic effects of laryngoscopy and intubation, and may compromise the patient’s amnesia for the intubation. Midazolam may be cerebroprotective, but less so than etomidate or thiopental. Ketamine. Ketamine, a phencyclidine derivative, has been widely used as a general anesthetic agent since 1970. After an IV dose of 1 to 2 mg/kg, ketamine produces loss of awareness within 30 seconds, peaks in approximately 1 minute, and has a clinical duration of 10 to 15 minutes. As a dissociative anesthetic agent, ketamine induces a cataleptic state rather than a true unconscious state. The patient has profound analgesia but may have open eyes. Many protective reflexes, including airway reflexes, are preserved. The principal uses of ketamine in emergency airway management are for the induction of patients with acute, severe asthma and for hemodynamically unstable trauma patients. Ketamine is exceptionally hemodynamically stable, more so than etomidate, so although either drug is a good choice in the trauma patient, ketamine is probably superior in terms of preserving precarious cardiovascular stability.89 In patients with status asthmaticus, etomidate or any of most of the other induction agents is acceptable, with the notable exception of sodium thiopental, which releases histamine. Ketamine is a direct bronchodilator and releases catecholamines, so may be useful both for intubation and for intermittent administration as part of sedation for mechanical ventilation in patients with severe asthma, although no outcome studies clearly demonstrate its superiority. Controversy exists regarding the use of ketamine in patients with elevated ICP because ketamine has been believed to increase cerebral metabolic rate, ICP, and cerebral blood flow.91 The evidence that ketamine can produce harm in this way is conflicting, however, and its role as an induction agent in trauma is significant because of its superior hemodynamic stability.10 Because of its tendency to release catecholamines and increase blood pressure, ketamine should probably be avoided in head trauma patients with normal or elevated blood pressure. However, in the hypotensive head trauma patient, ketamine is a reasonable choice for induction.77 Ketamine tends to produce unpleasant emergence phenomena, especially disturbing or frightening dreams in the first 3 hours after awakening. These reactions, which are more prominent in
15
Chapter 1 / Airway
of sedation and analgesia often obviates the need for an NMBA. Additional medication may be required if the patient’s blood pressure and heart rate indicate excessive sympathetic tone.
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
16
adults than in children, in women than in men, in patients receiving larger doses, and in certain personality types, are mitigated by benzodiazepine administration. Patients (e.g., with asthma) who undergo RSI with ketamine should receive a sufficient dose of a benzodiazepine (e.g., 0.05 mg/kg of lorazepam) as part of postintubation management.
Rapid Sequence Intubation for Table 1-4 Status Asthmaticus TIME
STEP
Zero minus 10 min Zero minus 5 min
Preparation Preoxygenation (as possible) Continuous albuterol nebulizer 100% oxygen for 3 min or 8 vital capacity breaths Pretreatment Lidocaine, 1.5 mg/kg Paralysis with induction Ketamine, 1.5 mg/kg Succinylcholine, 1.5 mg/kg Positioning Placement Laryngoscopy with intubation End-tidal carbon dioxide confirmation Postintubation management Sedation and analgesia NMBA only if required after adequate sedation/analgesia In-line albuterol nebulization Additional ketamine as indicated
Special Clinical Circumstances Status Asthmaticus Status asthmaticus with supervening respiratory failure is a preterminal event. Respiratory failure in the asthmatic patient is not caused primarily by progressive worsening of the bronchospasm, but rather by eventual exhaustion and fatigue secondary to the effort of breathing against severe airway resistance. All patients who are intubated for status asthmaticus are heavily sedated and receive mechanical ventilation. RSI permits the most rapid attainment of intubation, protects against aspiration, and induces the unconsciousness and motor paralysis necessary for optimal initiation of mechanical ventilation; it is the recommended technique for intubation of a patient in status asthmaticus. Difficult airway considerations are complex in an asthmatic patient because of impending respiratory arrest and the patient’s inability to tolerate attempts at awake intubation. Even when a difficult airway is identified in an asthmatic patient, RSI is usually the intubation method of choice, with a double setup for rescue cricothyrotomy when indicated. The asthmatic patient has highly reactive airways, and steps should be taken to minimize any additional bronchospasm that may occur during intubation. Lidocaine has been shown to suppress the coughing that occurs in response to airway manipulation and may improve ETT tolerance and reduce reactive bronchospasm in asthmatic patients.48 The balance of evidence suggests that lidocaine (1.5 mg/kg) is indicated as a pretreatment drug before intubation in status asthmaticus and in asthmatic patients being intubated for reasons other than their asthma. High-dose, inhaled beta-agonists may provide maximal protection against reactive bronchospasm during intubation in asthmatics without active bronchospasm, and lidocaine may provide little additional benefit in this setting.47 This approach has not been tested in patients in status asthmaticus, however. Ketamine has been shown to produce bronchodilation in humans and animal models and may be the ideal induction agent in asthma. Although reports to date have been limited, there is a growing body of experience with ketamine as an induction agent for the emergency intubation of patients with status asthmaticus. Ketamine also has been reported to mitigate bronchospasm in patients who are not intubated and in patients who are already intubated and who are not improving with mechanical ventilation (Table 1-4).
Hemodynamic Consequences of Intubation Laryngoscopy and intubation are potent stimuli for the reflex release of catecholamines.92 This reflex sympathetic response to laryngoscopy (RSRL) produces only modest increases in blood pressure and heart rate and is of little consequence in otherwise healthy patients. The RSRL is of potential clinical significance in two settings: acute elevation of ICP and certain cardiovascular diseases (e.g., intracerebral hemorrhage, subarachnoid hemorrhage, aortic dissection or aneurysm, and is chemic heart disease). In these settings, the reflex release of catecholamines, increased myocardial oxygen demand, and attendant rise in mean arterial blood pressure and heart rate may produce deleterious effects. The synthetic opioids (e.g.,
Zero minus 3 min Zero
Zero plus 30 sec Zero plus 45 sec Zero plus 2 min
fentanyl) and beta-adrenergic blocking agents (e.g., esmolol) are capable of blunting the RSRL and stabilizing heart rate and blood pressure during intubation.92 Lidocaine also has been studied, but the results are contradictory and inconclusive.93 In patients at risk from acute blood pressure elevation, administration of fentanyl (3 µg/kg) during the pretreatment phase of RSI attenuates the heart rate and blood pressure increase. The full sympatholytic dose of fentanyl is 5 to 9 µg/ kg, but if this dose is administered as a single pretreatment bolus, hypoventilation or apnea can occur. The administration of 3 µg/kg is safer and can be supplemented with an additional 3 µg/kg immediately after intubation if full sympathetic blockade is desired or if hypertension and tachycardia ensue, providing evidence of excessive sympathetic activity. Fentanyl should be given as the last pretreatment drug over 60 seconds to prevent hypoventilation or apnea.
Elevated Intracranial Pressure When ICP is elevated as a result of head injury or acute intracranial catastrophe, maintenance of cerebral perfusion pressure and avoidance of further increases in ICP are desirable.43 Significant reductions in mean arterial blood pressure decrease cerebral perfusion pressure by reducing the driving gradient between arterial pressure and ICP, leading to increased cerebral ischemia. Maintenance of the systemic mean arterial blood pressure at 100 mm Hg or greater supports the cerebral perfusion pressure and reduces the likelihood of secondary injury. In addition, cerebral autoregulation may be lost, and increases in systemic blood pressure may lead to corresponding increases in cerebral blood flow and ICP. With elevated ICP, control of the reflex hemodynamic stimulation resulting from intubation is desirable to avoid further elevation of ICP. Fentanyl (3 µg/kg) given as a pretreatment drug is the best choice for this purpose in the emergency setting.43,94 Evidence suggests a separate reflex that increases ICP in response to laryngoscopy and intubation, although the precise mechanism is not understood. IV lidocaine reduces ICP and
TIME
STEP
Zero minus 10 min Zero minus 5 min
Preparation Preoxygenation (as possible) 100% oxygen for 3 min or 8 vital capacity breaths Pretreatment Lidocaine, 1.5 mg/kg Fentanyl, 3 µg/kg (slowly) Paralysis with induction Etomidate, 0.3 mg/kg Succinylcholine, 1.5 mg/kg* Positioning Placement Laryngoscopy with intubation End-tidal carbon dioxide confirmation Postintubation management Sedation and analgesia, consider propofol to permit frequent reexamination NMBA only if required after adequate sedation/analgesia
Zero minus 3 min Zero Zero plus 30 sec Zero plus 45 sec Zero plus 2 min
*May substitute rocuronium, 1 mg/kg, for succinylcholine.
blunts the ICP response to laryngoscopy and intubation. Lidocaine (1.5 mg/kg IV), administered during the pretreatment phase of RSI, is desirable to blunt the ICP response to laryngoscopy and intubation. Similarly, RSRL and ICP response to laryngoscopy and intubation relatively contraindicate BNTI, which should be undertaken only if RSI is impossible and fiberoptic intubation is not an option. The physician should choose an induction agent that balances a favorable effect on cerebral dynamics and ICP with a stable systemic hemodynamic profile. At present, etomidate (0.3 mg/kg) probably is the best choice for patients with elevated ICP, although thiopental also is an excellent choice when hypotension is not present (Table 1-5).
Potential Cervical Spine Injury Historically, it was believed that oral endotracheal intubation carried an unacceptably high risk of injury to the cervical spinal cord in patients with blunt cervical spine injury and was relatively contraindicated, but this assertion was never subjected to scientific scrutiny. Numerous studies and reports have asserted the safety and effectiveness of controlled, oral intubation with in-line cervical spine immobilization, whether done as an awake procedure or with neuromuscular blockade.95,96 The evidence favors RSI with in-line stabilization, which provides maximal control of the patient, the ability to mitigate adverse effects of the intubation, and the best conditions for laryngoscopy. In-line stabilization also seems to improve the laryngoscopic view of the larynx compared with conventional tape/collar/sandbag immobilization. The intubating laryngeal mask airway (ILMA) also has been compared with conventional laryngoscopy and may result in less movement of the cervical spine during intubation than that caused by direct laryngoscopy.97 A comparison of methods on a cadaver model of unstable injury of the third cervical vertebra reinforced the potential role for fiberoptic intubation and raised questions about the safety of the Combitube because of significant cervical spine movement during its placement.98 Newer devices have also shown promise for safe intubation of patients with cervical spine injury. A fluoroscopic study com-
Pediatric Intubation Although many considerations in pediatric intubation are the same as for adults, a few differences exist in regard to airway management. The larynx is higher in the child’s neck, causing a more acute angle between the oral pharynx and the larynx. Visualization is aided by gentle posterior pressure on the anterior aspect of the thyroid cartilage. The epiglottis is high and soft, making visualization of the cords more difficult. If the child is very small, the prominent occiput brings the mouth to a position far anterior to the larynx; an assistant can lift the chest gently by grasping both shoulders, immobilizing the head at the same time. The airway in the small child is short, and care must be taken not to intubate either bronchus.67 A straight laryngoscope blade is desirable to displace the floppy epiglottis, especially in young children, and positioning for intubation may be different. BNTI is relatively contraindicated in children younger than 12 years old. Although the product insert for succinylcholine now advises against its routine use in pediatric anesthesia because of the risk of hyperkalemia in children with undiagnosed congenital neuromuscular disorders (e.g., muscular dystrophy), it remains the drug of choice for emergency RSI of infants and children.62 Rocuronium has been used in children, but experience is too limited to recommend that it replace succinylcholine for pediatric RSI in the ED. RSI may be used in children in a similar manner to adults, with two important differences. Excessive bradycardia may be seen with succinylcholine in children younger than 1 year old, but it is not known whether administration of atropine (0.02 mg/kg) during the pretreatment phase prevents any possible adverse outcome. The dose of succinylcholine in infants is 2 mg/kg. Induction agents may be selected using similar criteria as for adults. The major difficulty in intubating children and infants is choosing the correct size of equipment and the correct drug doses for age or size. These obstacles can be overcome by use of a length-based system (Broselow-Luten Color Coding Kids; Vital Signs, Inc., Totowa, NJ), which provides dosing and equipment sizes based on the length of the child. Cricothyrotomy is impossible
17
Chapter 1 / Airway
Rapid Sequence Intubation for Elevated Table 1-5 Intracranial Pressure
paring intubation using the Shikani optical stylet (SOS) to that done with direct laryngoscopy showed significantly less cervical spine movement with the SOS, but a slightly longer time (28 sec vs. 17 sec) to achieve intubation.99 The Airtraq, a single-use intubation device, resulted in better glottic views and more rapid intubation of patients with cervical spine immobilization than direct laryngoscopy using a Macintosh blade.100 The Glidescope, a video laryngoscope, provides superior glottic views with reduced or comparable cervical spine movement when compared with conventional direct laryngoscopy using the Macintosh blade.101,102 Cervical spine immobilization of patients with penetrating head and neck trauma is poorly addressed in the literature. It is not proven whether patients with gunshot or shotgun injuries to the head or neck are at risk of exacerbation of cervical cord injury during intubation, but there is no report of such a patient, with or without clinical evidence of spinal cord injury, who was injured by intubation. If the path of the missile is felt not to involve the bony spinal column and there is no evidence of spinal cord injury, prudence would dictate immobilization of patients with gunshot wounds to the head or neck with a secondary injury mechanism (e.g., fall from height) or with neurologic deficit suggesting spinal involvement.103 Immobilization for intubation of patients with penetrating injury elsewhere in the body should be directed by the likelihood of secondary injury to the spine from a fall or other event distinct from the wounding.
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
18
in small children, and alternative rescue airway devices (e.g., percutaneous oxygenation via the cricothyroid membrane) are required.
Other Airway Devices and Techniques Regardless of the care taken by the intubator and the detailed assessment of the patient before intubation, some intubations are simply unsuccessful or impossible. In most circumstances when intubation is not possible, BMV or ventilation using an EGD provides adequate ventilation and oxygenation until a rescue airway can be established. This underscores the importance of evaluating the patient for ease of intubation, ventilation, and EGD use before deciding on the best approach and initiating the intubation sequence. Over the past 10 years, there has been a revolution in airway management, based primarily on the incorporation of video and fiberoptic technology into laryngoscopes and stylets. In addition, increasing experience with extraglottic devices and other approaches has proved useful both for routine and difficult or failed airways.
Extraglottic Devices Laryngeal Mask Airway. The laryngeal mask airway (LMA) is an irregular, ovoid, silicone mask with an inflatable rim, connected to a tube that allows ventilation (Fig. 1-10). It is available in both reusable and single-use configurations; single-use models are offered by several manufacturers and are probably equivalent.104 The mask is inserted blindly into the pharynx, then inflated, providing a seal that permits ventilation of the trachea with minimal gastric insufflation. In elective anesthesia, the LMA has an extremely high insertion success rate and low complication rate, including a low incidence of tracheal aspiration.105,106 In the emergency setting, studies to date have focused on use during resuscitation from cardiopulmonary arrest, although data are beginning to emerge for use of the LMA as a rescue device in the event of failed intubation and
Figure 1-10. The standard laryngeal mask airway (LMA Classic) is available in sizes from infant to large adult. (Courtesy LMA North America, Inc., San Diego.)
as an alternative to direct laryngoscopy for intubation or a bagvalve-mask for ventilation.107 Evaluations of LMA insertion by experienced and inexperienced personnel consistently have shown ease of insertion, high insertion success rates, and successful ventilation.108 Novice users appear to be able to both ventilate and intubate more easily and successfully with the intubating LMA (ILMA) than by bag-mask ventilation and direct laryngoscopy.108 The LMA may be a viable alternative to endotracheal intubation for in-hospital or out-of-hospital treatment of cardiac arrest, particularly when responders are inexperienced airway managers. At a minimum, the device may serve a temporizing role equal or superior to BMV until definitive airway management can be achieved. A new form of LMA, the iGel, has a viscous gel within the cuff, so does not require inflation. Initial experience with the device, even with minimally trained novice users, is promising, with high insertion success rates and short insertion times.109 The ILMA is designed to facilitate intubation through the mask after correct placement (Fig. 1-11). It differs from the LMA in two main ways: The mask is attached to a rigid, stainless steel ventilation tube that is bent almost to a right angle, and the mask incorporates an epiglottic elevator at its distal end. Placement of the ILMA results in successful ventilation in almost 100% of cases and successful subsequent intubation in 95%.97,110-112 The ILMA can also be used for both ventilation and intubation in obese patients with similarly high success rates.113 The ILMA has a special ETT and a stabilizer rod to remove the mask over the ETT after intubation is accomplished, but intubation can be comparably successful with a conventional polyvinylchloride (PVC) endotracheal tube.114 The ILMA is a better device than the standard LMA for use in the ED because it facilitates both rescue ventilation and intubation. Intubation through the ILMA has compared favorably in terms of success with direct laryngoscopy and is superior in the hands of relatively novice intubators.108,110 When the ILMA is placed, intubation can be performed blindly or guided by a lighted stylet or a fiberoptic scope. The ILMA comes only in sizes 3, 4, and 5 and so is not suitable for use in patients weighing less than about 30 kg. For smaller patients, the standard LMA, which has sizes down to size 1 (infant), should be used. Intubation can be achieved through the standard LMA,
Figure 1-11. The intubating laryngeal mask airway is modified to facilitate insertion of an endotracheal tube after placement and ventilation are achieved. The epiglottic elevator (triangle) lifts the epiglottis to allow passage of the special ETT (arrow).
19
Chapter 1 / Airway
but the success rate is significantly less than with the ILMA. As experience with both the LMA and ILMA grows, it is likely that there will be increasing adoption of the LMA as a primary airway management technique by nonhospital first-responders, and the ILMA is gaining attention as a primary rescue device in the ED. A new version of the ILMA, the CTrach incorporates fiberoptic bundles and a detachable viewing screen to provide a view of the glottis during intubation. The device performs better than the standard ILMA for first attempt intubation, where it achieves almost 93% success versus approximately 80% for the ILMA in one well-conducted study.115 Ultimately, though, the ILMA’s intubation success rate is so high (on three or fewer attempts) that it is not clear that the CTrach provides additional benefit overall. The view can uncommonly be obscured by secretions, but this is easily solved by removing and reinserting the device, or cleaning it with a swab through the airway lumen.116 The greatest issue is cost, with the CTrach priced almost five times higher than the corresponding set of standard ILMAs. Whether the additional cost provides additional benefit for application in the ED remains to be seen. In the ED, the primary use of the LMA or ILMA is as a rescue technique to provide a temporary airway when intubation has failed, bag ventilation is satisfactory, and the patient has been paralyzed or is otherwise in need of immediate airway management. In such cases, the LMA is one of numerous acceptable devices. In the “can’t intubate, can’t ventilate” situation, cricothyrotomy is indicated, but an ILMA may be placed rapidly in an attempt to achieve ventilation (converting the situation to “can’t intubate, can ventilate”) as long as this is done in parallel with preparations for cricothyrotomy and does not delay the initiation of a surgical airway.107 Availability of the LMA and adequate prior training of the clinician offer a legitimate option for the management of the failed airway, and the ILMA compares well with fiberoptic intubation in terms of successful intubation of difficult airways.112 The standard LMA may also offer advantages for providing ventilation in unconventional positions, such as when the patient is lying on his or her side.117 In the out-of-hospital setting, where concerns about esophageal placement of ETTs have focused interest on methods used for airway management, the LMA and Combitube offer excellent placement and ventilation characteristics and may be preferable to endotracheal intubation in this setting, especially when intubation is relatively infrequently performed.118 If the patient is in a difficult position in terms of intubation access, the LMA may facilitate more rapid ventilation.119 New LMA devices, from a number of manufacturers, are now available. Esophagotracheal Combitube. The Combitube is a plastic doublelumen tube with one lumen functioning as an airway after esophageal insertion and the other lumen functioning as a tracheal airway (Fig. 1-12). The tube is placed blindly into the esophagus, and proximal and distal balloons are inflated to prevent escape of ventilatory gases through the pharynx to the mouth or nose or down the esophagus. The tube is placed into the esophagus, as designed, almost 100% of the time, but both lumens are patent, so ventilation is still possible if the tube has been placed inadvertently into the trachea. The Combitube is primarily a substitute for endotracheal intubation for non-ETT-trained personnel, but it also has a role as a primary airway device in place of endotracheal intubation in the out-of-hospital setting.120 It has been used as a rescue device or as a primary intubating device in difficult airways that have precluded endotracheal intubation or successful LMA placement, both in patients with and those without cardiac arrest.121,122 Serious complications attributable
Figure 1-12. The Combitube is inserted through the mouth blindly,
although a laryngoscope can be used, if desired. It seats itself in the esophagus more than 95% of the time, and ventilation is performed through the proximal lumen and side ports (white arrows) after inflation of the two balloons. The lower balloon (triangle) occludes the esophagus. The upper (large) balloon (thick arrow) occludes the oropharynx. If the tube is in the trachea, the alternate lumen (clear arrow) is used for ventilation.
to Combitube use are uncommon.123 The tube may be difficult to insert blindly when the patient is in cervical spine precautions, raising concerns about first-responder use in trauma patients, but results have been conflicting.124,125 Standard methods for confirming tube placement, using ETco2, seem to be reliable in identifying whether the tube has been passed into the esophagus or trachea and in confirming the correct ventilation port. Although the Combitube has provided successful ventilation for several hours, it should be considered a temporizing measure only. Current use in the ED should be restricted to rescue placement after failed oral intubation with adequate BMV or a quick maneuver in the “can’t intubate, can’t oxygenate” patient simultaneous with preparation for a cricothyrotomy (analogous to the use of the ILMA in this situation). The Combitube has virtually no role in the ED as a primary airway management device except in cases of cardiopulmonary arrest when expertise for endotracheal intubation is not available.
Video Laryngoscopes New devices incorporate video imaging into modified laryngoscopes to allow superior visualization of the glottis without the need to create a straight-line visual axis through the mouth. The Glidescope uses an extended Macintosh blade with a sharply angulated tip to direct the video camera at the glottis, even in patients with difficult airways (Fig. 1-13). When compared with direct laryngoscopy, the Glidescope provides an equivalent or superior glottic view, and has a very high intubation success rate.126,127 The Glidescope appears to cause less cervical spine movement than conventional direct laryngoscopy with a Macintosh blade.101 The C-MAC video laryngoscope (Fig. 1-14) incorporates a complementarymetal-oxide-semiconductor (CMOS) video chip into otherwise conventional laryngoscope blades, to enhance glottic view. Other videolaryngoscopes are available or under development. Overall, videolaryngoscopy offers the promise of transforming laryngoscopy and has the potential to render conventional, direct laryngoscopy obsolete.128
Fiberoptic Intubating Stylets Several rigid fiberoptic intubating stylets have also been approved and adopted into clinical use.129 The Shikani Optical Stylet (SOS—Clarus Medical, Minneapolis, Minn.) is the most studied of these. The endotracheal tube is placed over the
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
20
Figure 1-15. The Shikani optical stylet (SOS, Clarus Optical) with Figure 1-13. The GlideScope (Verathon, Inc.) is a video laryngoscope that
endotracheal tube mounted. The eyepiece and battery pack are at the right.
uses a 50° deflection of the distal tip of the blade (which is otherwise similar to an extended MAC-3 blade) to direct the video camera and light source directly at the glottis without repositioning the head. The endotracheal tube insertion is done under direct vision via the video screen. (From Walls RM, Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, p. 171, 2008, with permission.)
Figure 1-16. The Bonfils intubating fiberscope (Karl Storz). The
endotracheal tube is mounted on the stylet, and intubation is facilitated through the eyepiece at right.
similarly.130 The Bonfils intubating fiberscope (Karl Storz Endoscopy of America, Culver City, Calif.) functions as a retromolar intubating stylet (Fig. 1-16). The ETT is loaded directly onto the nonmalleable fiberoptic stylet, then guided along the cheek and directed around posterior to the back molar, then through the glottic aperture by direct fiberoptic visualization.128,129
Flexible Fiberoptic Scopes
Figure 1-14. The C-MAC video laryngoscopy (Karl Storz Endoscopy) uses an integrated CMOS video chip to capture a video image from near the distal tip of an otherwise conventional laryngoscope blade. The image is conveyed to a video screen where it is viewed by the intubator. (From Walls RM, Murphy MF [eds]: Manual of Emergency Airway Management, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, p. 173, 2008, with permission.)
malleable stylet, then advanced through the mouth in the midline and into the trachea using built-in fiberoptic visualization (Fig. 1-15). The SOS appears to cause less movement of the cervical spine than conventional laryngoscopy during intubation with in-line stabilization.99 A simpler version, the Levitan scope, uses an LED-illuminated fiberoptic stylet to facilitate intubation by direct laryngoscopy.128 The device is recommended by the manufacturer to facilitate first-pass success when a limited glottic view is obtained by direct laryngoscopy. In the only study comparing the Levitan scope to the gum-elastic bougie, however, the two devices performed
Intubation using a flexible fiberoptic scope is increasingly applied to difficult airways in the ED, after many years of use for similar applications in the operating room. The intubating fiberoptic bronchoscope can be passed through the vocal cords under fiberoptic visualization, then can serve as an introducer over which the ETT is passed. Fiberoptic examination facilitates airway assessment for the need for intubation, without definitely committing the patient to intubation, as is the case when an NMBA is administered for RSI. For example, in a patient with smoke inhalation, examination with the fiberoptic scope might identify that intubation is not required, but will also facilitate intubation when it is indicated. Intubation of morbidly obese patients, those with distorted airway anatomy (e.g., penetrating or blunt anterior neck injury), or those with fixed cervical spine deformity, can be achieved using the fiberoptic scope, topical anesthesia, and moderate (procedural level) sedation, thus preserving the patient’s ability to breathe until intubation is achieved. The fiberoptic scope also has been used successfully in concert with the ILMA to achieve intubation in difficult cases, including when the cervical spine is immobilized, where it significantly outperforms conventional laryngoscopy.131 There is a significant learning curve for flexible fiberoptic intubation, and fiberoptic examination of the upper airway in
Other Intubation Techniques Retrograde Intubation. In retrograde intubation, a flexible wire is passed in retrograde fashion through a cricothyroid membrane puncture. The wire is retrieved through the mouth, then used to facilitate intubation by serving as a guide over which the ETT is passed. Purported advantages of retrograde intubation include ease of learning and application to the difficult airway. Although retrograde intubation theoretically may be useful when the upper airway is disrupted by trauma, rendering oral intubation difficult or impossible, it is unlikely to be used in the ED except in circumstances in which alternative devices, such as fiberoptic intubation, Trachlight, Combitube, and cricothyrotomy, are unavailable. Published reports of its use in emergency circumstances have been limited to case reports, very small series, and review articles. It is doubtful whether retrograde intubation would ever be the airway maneuver of first choice in the ED, but it may be a useful consideration in rare, unique difficult airway cases. Lighted Stylet. The lighted stylet is a device that incorporates a handle, a fitting for mounting an ETT, and an intubating stylet with a fiberoptic light mounted on the end (Fig. 1-17). The ETT is mounted as on a conventional intubating stylet, but transillumination of the soft tissues from within the neck permits identification of tracheal entry by the stylet and ETT. The lighted stylet has been used for oral and nasal intubation and has an excellent success rate.134 The lighted stylet is less stimulating to the heart rate and blood pressure than conventional laryngoscopy and may be useful when sympathetic stimulation is not desirable.135 Although overall success rates with the Trachlight lighted stylet have been high, it may be more difficult for novice intubators to learn than conventional laryngoscopy, if only minimal manikin training is used.136 The Trachlight can be used as a primary intubating device or as a rescue device in the “can’t intubate, can ventilate” failed airway. It is not appropriate for the “can’t intubate, can’t ventilate” failed airway, when cricothyrotomy is indicated. As a device for a difficult airway, the lighted stylet can be used as the intubating stylet for a standard oral intubation. The direct illumination by the stylet can aid in visualization during intubation. If direct laryngoscopy is unsuccessful, the first rescue procedure could be an immediate attempt at blind, oral intubation using the lighted stylet, as long as ventilation is possible. There is also some evidence that the Trachlight produces less cervical spine motion than does direct laryngoscopy.102
21
Chapter 1 / Airway
patients with pharyngitis or odynophagia, for example, is helpful, as it requires the same “navigation skills” as are required for intubation. Use of a video attachment for instruction, so that the instructor and learner can simultaneously see the same image appears to enhance learning.132 Models have been created to allow learners to navigate through a series of openings and around barriers, which also increases subsequent intubation performance.133 The role of flexible fiberoptic intubation in the ED is greatly expanding, as obesity increases in the population, and, increasingly, difficult airways are handled in the ED without backup. The transition from fiberoptic to CMOS video technology should make these flexible scopes more durable, less prone to fogging, and less expensive—all desirable attributes for emergency intubation. Emergency physicians should have immediate access to fiberoptic scopes and should endeavor to acquire training and practice in their use. Fiberoptic scopes are of great value in the patient with predicted difficulty in direct laryngoscopy, EGD use, and BMV. The expanding use of video laryngoscopy will redefine the role of flexible fiberoptic scopes, as video laryngoscopy solves many of the difficulties that occur with direct laryngoscopy.
Figure 1-17. The Trachlight lighted stylet facilitates placement of the endotracheal tube when the glottis cannot be visualized by direct laryngoscopy. It also is used as a primary intubation device. Here the endotracheal tube is correctly mounted on the stylet.
Surgical Airway Management Needle Cricothyrotomy with Transtracheal Jet Ventilation Needle cricothyrotomy involves the insertion of a large needle (ideally 10-gauge) through the cricothyroid membrane into the airway. When inserted, the needle is used to ventilate the patient with a standard wall oxygen source. Because of the high-velocity ventilation that ensues through the narrow catheter, this procedure is called transtracheal jet ventilation. Transtracheal jet ventilation has been used successfully in humans and has been subjected to various animal experiments to determine its uses and limitations. It rarely has been used in patients in EDs, however, where its role as a rescue device in the “can’t intubate, can’t ventilate” situation is vastly inferior to cricothyrotomy. The jet ventilator should include a regulator and gauge so that pressures can be monitored and reduced, especially in children (Fig. 1-18). Upper airway obstruction has been considered a contraindication to transtracheal jet ventilation, but ventilation still can be successful, although at the cost of higher intrapleural pressure and possibly pulmonary barotrauma. In general, when upper airway obstruction is present in adults, percutaneous or surgical cricothyrotomy is preferred. The primary indication for transtracheal ventilation in the ED is the initiation of emergency oxygenation for a pediatric patient who is apneic (either because of the presenting condition or because of administration of an NMBA) and in whom intubation and BMV are impossible. Cricothyrotomy is extremely difficult or impossible in children younger than 10 years old, and transtracheal ventilation should be considered the surgical rescue modality of choice in this age group. For children younger than 5 years old, bag ventilation is used with the percutaneous catheter, and pressurized devices are avoided.67
Cricothyrotomy Cricothyrotomy is the creation of an opening in the cricothyroid membrane through which a cannula, usually a cuffed tracheostomy tube, is inserted to permit ventilation. The techniques, and variations thereof, are well described elsewhere.137 When surgical airway management is required, cricothyrotomy is the procedure of choice in the emergency setting, where it is faster, more straightforward, and more likely to be successful than tracheotomy. Cricothyrotomy is indicated when oral or nasal intubation is impossible or fails and when BMV cannot maintain adequate oxygen saturation (the “can’t intubate, can’t ventilate” situation). Several large series have established that the incidence of cricothyrotomy is approximately 1% of all ED intubations.1,39
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
22
Figure 1-18. Transtracheal jet ventilation. High-pressure ventilation tubing (black triangle) attaches to standard wall oxygen outlet at 55 psi. Ventilation block (middle white triangle) is used to control oxygen flow through tubing (top left white triangle) to catheter (lower right triangle), which is inserted in the airway.
Figure 1-19. Melker universal cricothyrotomy kit. (Courtesy of Cook Cricothyrotomy is relatively contraindicated by distorted neck anatomy, preexisting infection, and coagulopathy; these contraindications are relative, however, and the establishment of the airway takes precedence over all other considerations. Successful cricothyrotomy after systemic fibrinolytic therapy has been reported.138 The procedure should be avoided in children younger than 10 years old, in whom anatomic considerations make it exceedingly difficult.67 Studies suggest that approximately five “practice” cricothyrotomies on a simulator or animal model are sufficient to achieve at least baseline capability with the procedure.139 Cricothyrotomes are devices used to perform percutaneous cricothyroidotomy. Percutaneous cricothyrotomy using the Seldinger technique appears comparable to formal open cricothyrotomy in terms of ease of learning and success rates.140 The safety and effectiveness of other cricothyrotomes are not clearly established. A recently released kit by Portex offers a small red flag indicator to warn when the posterior tracheal wall is touched, but a cadaver study showed that, although the device resulted in somewhat faster placement of an airway than did a Seldinger technique, the incidences of both failure and major complications (posterior airway wall laceration) were unacceptably high, so the device cannot be recommended.141 Only two percutaneous cricothyrotomy sets on the market currently have the ability to place a cuffed tracheostomy tube. One is a dedicated Seldinger cricothyrotomy set; the other is a combination set that has all necessary equipment for either a Seldinger percutaneous cricothyrotomy or a standard surgical cricothyrotomy (Melker universal cricothyrotomy kit; Cook Critical Care, Bloomington, Ind.) (Fig. 1-19).
■ OUTCOMES Few studies of emergency airway management have characterized complications and outcomes. The largest single-institution series reported a success rate for ED RSI of 99% and a complication rate of 9.3%; most complications were minor.14 Phase II of the large National Emergency Airway Registry Study (NEAR II) of almost 9000 ED intubations reported success rates of approximately 97% for RSI. The NEAR classification system characterizes potentially adverse occurrences during intubation as “adverse events.”1,39,142 In the NEAR study, the observed rate of adverse events was approximately
Critical Care.) (Disclosure: The author assisted in the design of this kit and receives a 10–35% royalty on its sales.)
9% in medical patients and 8% in trauma patients, and most of these were minor.1 No studies have evaluated the long-term outcome of intubated ED patients.
KEY CONCEPTS ■
Knowledge of the clinical course of the patient’s condition and anticipation of possible deterioration are crucial to the decision to intubate, especially if the patient is to leave the ED for a time (e.g., interfacility transfer, diagnostic testing). ■ Assessment of the patient for potential difficulty with intubation, bag-mask ventilation (BMV), ventilation using an extraglottic device (EGD), and cricothyrotomy is an essential step in planning airway management. The mnemonics LEMON, MOANS, and RODS can serve as useful aids. ■ In the absence of a “crash” patient (agonal, unresponsive to laryngoscopy) or a difficult airway, RSI is the airway management method of choice for ED patients. ■ Succinylcholine is the NMBA of choice for ED RSI, but it should be avoided in certain patient groups because of risk of significant hyperkalemia. ■ Pretreatment drugs given during RSI can mitigate adverse responses to intubation and improve the patient’s clinical condition. ■ Tube placement confirmation using end-tidal CO2 (ETco2) is essential after intubation, and failure to detect adequate quantities of exhaled CO2 is evidence of esophageal intubation until proven otherwise. ■ Videolaryngoscopy is transforming intubation by eliminating the traditional anatomic barriers to direct laryngoscopy. Emergency practitioners should evaluate video laryngoscopes for incorporation into their practice, both for difficult and routine intubations.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 2
Mechanical Ventilation and Noninvasive Ventilatory Support
Megan L. Anderson and John G. Younger
■ PERSPECTIVE Invasive and noninvasive ventilation are essential tools for treatment of critically ill patients. The indications for endotracheal intubation (ETT) and for assisted ventilation in the emergency department (ED) are not always the same. Some patients require support for respiratory failure or as part of comprehensive management of critical illness, while others’ cardiopulmonary function may be preserved and assistance is needed primarily for airway protection. The decision to intubate is discussed in Chapter 1 and in various other places throughout this textbook in the context of individual conditions. This chapter describes the modalities and techniques of mechanical ventilation.
■ FUNDAMENTALS Invasive Techniques Once the need for ventilatory support has been identified, three questions must be answered to initiate support: (1) What will constitute a breath—a delivered tidal volume or a delivered airway pressure? (2) To what extent will the patient be allowed to participate in breathing? (3) How will the support be delivered—by endotracheal tube or by some noninvasive means? These three are typically referred to, in order, as the cycle or limit, the mode, and the method of mechanical ventilation. Regardless of the specific method used, an oxygenated mixture of gases, usually warmed and humidified, is cyclically forced into the lungs under supra-atmospheric pressure during the inspiratory phase and allowed to exit passively during expiration. In pressure-limited, or cycled, ventilation, a breath is defined by the peak inspiratory pressure (PIP) that is achieved each cycle. When that pressure is reached, gas insufflation ends, and passive exhalation is permitted; the delivered tidal volume is thus a function of dynamic lung compliance. An advantage of pressure-cycled ventilation is that it serves as an explicit safeguard against iatrogenic, injurious high airway pressures. A significant disadvantage is the possibility of hyper- or hypoventilation in the event of an acute change in lung compliance during therapy. In volume-cycled ventilation, inhalation ends when a preset tidal volume has been delivered, and inspiratory pressure varies with the inverse of lung compliance. The benefit to this method is the ability to control tidal volume; the risk is poten-
tially high peak pressures when compliance is poor. There is no consensus regarding which approach to positive-pressure ventilation is better; pressure- and volume-cycling are opposite sides of the same coin, and both strategies are used clinically. The most commonly available modes of positive-pressure ventilation are (1) controlled mechanical ventilation (CMV), (2) assist/control (A/C) ventilation, and (3) synchronized intermittent mandatory ventilation (SIMV), all of which can be supplied through either pressure-cycled or volume-cycled ventilators. Two main factors differentiate these modes from one another: (1) how a breath is triggered (at a preset fixed rate or by a patient’s inspiratory effort sensed by the ventilator), and (2) the target capacity (pressure or volume) for each breath. During CMV, the ventilator delivers breaths at a preset rate, regardless of any ventilatory effort made by the patient. A person receiving CMV can neither trigger a breath nor inspire gas spontaneously through the ventilator circuit, so this mode is appropriate only for apneic, pharmacologically paralyzed, and deeply sedated patients. In contrast, a ventilator applying A/C mode continuously monitors the ventilator circuit for either negative pressure or air flow deflections (generated by spontaneous inspiratory effort) and responds with a full breath. In the absence of any such patient effort, the device automatically cycles at a preset minimum “backup” rate. For example, if an A/C ventilator is set to deliver 12 breaths per minute, a breath is provided every 5 seconds in the absence of spontaneous inspiratory effort. Should a patient attempt to inspire, an additional breath is provided and the ventilator’s timer resets for another 5 seconds. Thus a patient can breathe at a higher rate than the programmed A/C rate, with an attendant increase in work of breathing. A/C ventilation is a useful initial mode of mechanical ventilation in many ED patients. Commonly encountered disadvantages of this mode include poor tolerance in awake patients (often resulting in frequent elevated airway pressure alarms and underventilation) and worsening of intrathoracic air trapping in patients with chronic obstructive pulmonary disease (COPD).1 SIMV is a more complex mode developed to improve patient comfort by facilitating patient-device synchrony. The delivery of a mandatory breath is synchronized as much as possible with a patient’s spontaneous respirations, preventing breath stacking (i.e., the delivery of a mechanical breath before the previous breath has been completely exhaled). Stacking may result in hyperinflation and barotrauma.2 If spontaneous breathing occurs at a rate equal to or lower than the preset ventilator rate, 23
the patient’s inspiration, or an elapsed time, triggers the next breath delivery. If the patient’s spontaneous breathing is faster than the established SIMV rate, the patient breathes gas from the ventilator circuit and receives a volume consistent with his or her inspiratory effort, in addition to regular breaths at the set tidal volume and rate, which are triggered by the patient and delivered by the ventilator. Regardless of the ventilatory mode chosen, additional refinements are available and commonly used. The most important of them is positive end-expiratory pressure, or PEEP. PEEP and continuous positive airway pressure (CPAP), a closely related entity, refer to the maintenance of positive airway pressure after the completion of passive exhalation. By convention, PEEP refers to pressure applied during invasive mechanical ventilation, whereas CPAP is the application of positive pressure (invasively or noninvasively) during spontaneous breathing. The terms are occasionally used interchangeably. Acute lung injury and cardiogenic pulmonary edema are characterized by loss of surfactant function. The chief beneficial effect of PEEP or CPAP is to increase functional residual capacity (FRC) by maintaining patency of injured or flooded alveoli that would otherwise collapse at the end of exhalation. Increasing the FRC may improve both oxygenation and lung compliance. One of the potential adverse effects of PEEP is decreased cardiac output. Pressure support ventilation (PSV) is another adjunct in which breathing is controlled by the patient, and peak pressures are controlled by the ventilator. The primary goal of PSV is to support the patient’s spontaneous breathing effort while providing satisfactory oxygenation. PSV provides for the prompt attainment of a preset PIP each time the patient initiates inspiratory effort. Inspiratory time, inspiratory flow rate, and tidal volume (TV) are augmented, whereas inspiratory work of breathing is reduced. The machine likewise senses the end of inspiration or initiation of exhalation, at which time the pressure support ceases, and exhalation is allowed to proceed spontaneously. Increasing levels of PSV decrease the work of breathing.1 PEEP may be added to PSV, and a manda-
Airway pressure
Volume control
Set pressure
tory ventilation rate can be set in case spontaneous respirations deteriorate, typically using SIMV. PSV first was used as a weaning tool, but some authorities now recommend it as a primary means of ventilatory support. Careful monitoring is necessary during its use because TV is uncontrolled. Invasive ventilation with PSV is rarely used in the ED, but PSV may be applied noninvasively. Applied PEEP must be differentiated from intrinsic PEEP (iPEEP, or auto-PEEP), which may result from improper assisted ventilation when adequate time is not allowed between breaths for complete exhalation. This circumstance is discussed later.
Noninvasive Techniques Noninvasive positive-pressure ventilation (NPPV) includes CPAP or biphasic positive airway pressure (BiPAP) and is applied with a face or nose mask. CPAP provides constant pressure throughout the respiratory cycle, and its benefits were discussed earlier in relation to PEEP. BiPAP alternates between higher pressure during inspiration (IPAP) and lower pressure during expiration (EPAP). The machine senses and responds to the patient’s respiratory efforts. EPAP splints airways open and prevents alveolar collapse and atelectasis. IPAP decreases the work of breathing and improves TV (Fig. 2-1).
■ MANAGEMENT Invasive versus Noninvasive Approach Alert patients with a patent airway and an intact respiratory drive, even if that drive is insufficient, may be candidates for NPPV. Patients most likely to respond to NPPV in the ED are those with more readily reversible causes of their distress, such as COPD exacerbation or cardiogenic pulmonary edema, in which fatigue is a significant factor.3-6 Patient selection, comprehensive management of the underlying condition, and ongoing monitoring are essential for successful NPPV. NPPV
Pressure control
Pressure support
Figure 2-1. Pressure, flow, and
Set pressure
+ 0
– Set ti Set flow Set minimal flow
Flow
Inhaled 0 Exhaled
Volume delivered
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
24
Set volume
+ 0 Machine triggered
Machine triggered
Patient triggered
waveforms typically encountered during mechanical ventilation in the emergency department. The nature of fresh gas delivery during mechanical ventilation is in part a function of the definition of “breath” (i.e., a delivered volume or delivered pressure) and the means by which that breath is initiated (i.e., by the patient or by a timing decision made by the ventilator). In practice, only a handful of ventilator parameters commonly are managed in the emergency department (the mode, the magnitude of the delivered breath, the rate of delivery, and Fio2). However, as this figure shows, a number of additional features can be fine-tuned to optimize the effectiveness and comfort of mechanical ventilation in critically ill patients. (Modified from Mason RJ: Murray and Nadel’s Textbook of Respiratory Medicine, 4th ed. Philadelphia: Elsevier, 2005.)
Initial Settings and Ongoing Monitoring Recommended initial settings for BiPAP ventilators are an IPAP of 8 cm H2O and an EPAP of 3 cm H2O. Either a face mask or a nose mask can be used. The flow of supplemental oxygen bled into the circuit should be governed by pulse oximetry, as corroborated by arterial blood gas (ABG) results; it is appropriate to initiate therapy with 3 to 5 L/min of supplemental oxygen, but this should be adjusted with each titration of IPAP or EPAP. The ventilator should be in spontaneous mode to support the patient’s respiratory effort. As the patient’s response to NPPV and other therapy is monitored (using cardiac and blood pressure monitors, ABGs, and oximetry, and the patient’s own voiced assessment of tolerance and progress), support pressures are adjusted. Although this adjustment must be individualized, a reasonable approach for BiPAP support in hypoxemic patients is to increase EPAP in 2-cm H2O increments, with IPAP maintained at a fixed interval higher. Hypercapnia can be managed by increasing IPAP in 2-cm H2O increments, with EPAP being increased in approximately a 1 : 2.5 ratio to IPAP.6,13 For the intubated patients, initial ventilator settings depend on the goal of the ventilatory intervention (mechanical ventilation, assisted ventilation, or PSV) and on the underlying cause of respiratory insufficiency. The basic parameters to be set in volume-cycled ventilators in CMV, A/C, IMV, and SIMV modes are fraction of inspired oxygen (Fio2), TV, rate, and inspiratory/expiratory (I/E) ratio. (The I/E ratio reflects the duration of machine insufflation and the rest periods between them.) If atelectasis is a problem, PEEP should be added; the addition of PEEP may permit the use of more
physiologic Fio2 values as well. For an apneic or paralyzed patient, CMV, A/C, or IMV mode may be used. For a breathing patient with inadequate ventilatory effort, A/C is usually the best initial approach. Reasonable initial ventilator settings are a TV of 6 to 8 mL/kg body mass and a rate of 12 to 14 breaths/min. Initial Fio2 should be set at 1.0 but generally can be adjusted down quickly to maintain an oxygen saturation of 90% or greater. Ventilator settings are adjusted dynamically using pulse oximetry, end-tidal carbon dioxide monitoring, ventilation pressures, clinical status, and ABGs as a guide. PEEP, if indicated, should be initiated at 2.5 to 5 cm H2O. In pressure-cycled ventilators, the rate and Fio2 are set as described earlier. An inspiratory pressure should be chosen that results in a TV of 6 to 8 mL/kg, usually 25 to 40 cm H2O. There are additional specific considerations for many particular conditions (see section on Special Clinical Circumstances). Mechanical ventilation is a dynamic process that requires constant monitoring and regular adjustment of these parameters. Tachycardia and hypertension can indicate ventilator intolerance and a need for increased sedation or adjustment of the ventilator settings. Bradycardia and ventricular irritability represent hypoxemia until this is disproved. Unless capnometry and pulse oximetry are in use, an ABG should be measured approximately 20 minutes after initiating support. These results indicate the sufficiency of ventilation (using the pH and arterial partial pressure of carbon dioxide, Paco2) and oxygenation (using arterial oxygen partial pressure, Pao2). Adjustments in minute volume (the product of TV and rate) and Fio2 can be guided by baseline measurements supplemented by ongoing monitoring. To avoid oxygen toxicity, Fio2 should be reduced to the lowest level that provides acceptable (≥90%) oxygen saturation. In many instances PEEP will allow better oxygenation for a given Fio2. Important ventilator readouts include the PIP and expiratory volume. PIP is among the most frequently referenced measures of ventilatory function during mechanical ventilation. It reflects lung compliance and airway resistance; changes in the magnitude of PIP may reflect any of several potentially detrimental problems related to ventilation.14 In a practical sense, PIP can be considered an additional vital sign for patients on a ventilator. Acute decreases in PIP reflect inadequate volume delivery to the patient, which may be caused by insufficient gas supply to the ventilator, inadvertent change in settings, a leak in the breathing circuit, unintended extubation, or failure of the ventilator. Increases in PIP may indicate ETT occlusion by secretions in or kinking of the tube, acute bronchospasm, pneumothorax, or conditions causing decreased lung compliance such as the development of worsening pulmonary edema. PIP can serve as a useful measure of effectiveness of therapy in patients with asthma or COPD; as airway resistance lessens, the PIP decreases. High PIP may cause barotrauma and other acute lung injury.15,16 Measurement of expiratory volume and expiratory flow allows estimation of the effectiveness of spontaneous respiratory efforts and, by comparing expiratory volume with the set TV, assessment of the effectiveness of ventilation and the integrity of the breathing circuit. The expiratory volume measurement is particularly important in assessing mechanical ventilation in children, who often have air leaks around an uncuffed ETT.
Patient Treatment Even if a mechanically ventilated patient’s stay in the ED is brief, attention must be paid to ventilatory management. Routine concerns are sedation, neuromuscular paralysis if
25
Chapter 2 / Mechanical Ventilation and Noninvasive Ventilatory Support
has been shown to be a beneficial initial intervention for both congestive heart failure (CHF) and COPD.7,8 Although studies have significant flaws, they have consistently demonstrated reduced need for intubations in both conditions, as well as a mortality benefit in COPD patients. In COPD, NPPV decreases the work of breathing and splints airways open, improving ventilation-perfusion matching.7 Predictors for success in COPD include younger age, unimpaired consciousness, less severe acidemia, and prompt response (less than 2 hours), as measured by heart rate, respiratory rate, and gas exchange. Predictors for failure in COPD include a Glasgow Coma Score less than 11, an arterial pH less than 7.25, and tachypnea greater than 30 breaths per minute.9 In CHF, NPPV reduces work of breathing, improves cardiac output by decreasing preload and afterload, redistributes lung water, and improves ventilation-perfusion matching thereby reducing shunt. However, although symptoms during acute exacerbations of CHF are improved, a mortality benefit has not been detected.10 Some trials have used NPPV for hypoxemic respiratory failure such as in pneumonia; however, the results are not clearly beneficial.8,11 Contraindications to NPPV include severely impaired level of consciousness, cardiac arrest, acute MI, inability to protect airway, apnea, copious secretions, uncontrolled vomiting, upper airway obstruction and facial trauma.11 Patients in whom NPPV is initially chosen should be reassessed frequently for progress of therapy, tolerance of the mode of support, and any signs of clinical deterioration that indicate a need for intubation. NPPV is an attractive alternative to ETT because it reduces the risks of airway trauma and ventilator-acquired pneumonia and may be helpful in patients who decline intubation. NPPV also may be considered for a patient whose advanced directives proscribe intubation. An individualized approach is important, and discussion with the patient and family members may be helpful.12
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
26
necessary, analgesia, and suctioning. Sedation and analgesia should be titrated to provide the greatest patient comfort and ventilation performance. In addition, a rapid and systematic approach should be taken to manage the patient who becomes suddenly difficult to oxygenate or ventilate. An opioid (e.g., fentanyl or morphine) and a sedative agent (e.g., midazolam by intermittent bolus or infusion, or propofol by infusion) are commonly used for analgesia and sedation. Ketamine can provide both sedation and analgesia and is often used for children and patients with reactive airways disease. Prolonged neuromuscular paralysis can usually be avoided by the use of adequate sedation and analgesia (see Chapter 1). If neuromuscular blockade is required, a competitive, non depolarizing agent, such as pancuronium, vecuronium, or rocuronium, is often selected. Endotracheal suctioning should be performed regularly. The appropriate frequency is a balance between the need for clearing secretions (especially in pulmonary edema or asthma) and the disadvantage of interrupting ventilation, which can sacrifice gains in alveolar recruitment by allowing airway pressures to fall, even very briefly, to atmospheric levels. Orally intubated patients should have a bite-block placed to protect the endotracheal tube.
Complications PPV is a lifesaving therapy. However, its use is associated with complications that can become quickly life-threatening, and it is important for emergency physicians to be familiar with common problems associated with PPV. Most of these result from changes in thoracic physiology when positive pressure is present for part or all of the respiratory cycle and are outlined in Box 2-1. Many of these are discussed elsewhere in this text. Acute difficulty with oxygenation or ventilation, or the development of high airway pressures in a previously calm patient, may indicate undersedation or inadequate analgesia. Additional sedation is administered in concert with a systematic search for patient or device-associated abnormalities. The differential diagnosis, after initial acclimatization, includes ETT migration, ETT occlusion, pneumothorax, bronchospasm, pulmonary edema, acute pulmonary embolism, dynamic hyperinflation, abdominal distention, mechanical failure of the ventilator, and patient-ventilator asynchrony.17 ETT placement can be checked by capnometry, physical examination, and chest radiography. ETT patency should be assessed by passing a suction catheter. In patients with copious secretions,
BOX 2-1
Potential Adverse Effects of Positive-Pressure Ventilation
Increased intrathoracic pressure, leading to Decreased venous return to the heart and decreased cardiac output Increased ventilation/perfusion ratio Air trapping and intrinsic positive end-expiratory pressure (iPEEP) Barotrauma Nosocomial infections of the lungs and sinuses Respiratory alkalosis Agitation and increased respiratory distress Increased work of breathing due to asynchrony or improperly set triggers
the existence of a mucous plug acting as a “ball valve” in the endotracheal tube must be considered. This phenomenon presents as a sudden decrease in exhaled volume and elevated PIP, which responds immediately to suctioning but then quickly recurs. The diagnoses of pulmonary edema, pneumothorax, and bronchospasm can be made clinically, with chest radiography used as an adjunct. Pulmonary embolism in a ventilated patient may be an even more elusive diagnosis than in other ED patients. Abdominal distention should be apparent on physical examination and is relieved by passage of a nasogastric or orogastric tube. Dynamic hyper inflation and ventilator malfunction may be diagnosed by momentarily disconnecting the ventilator. In the former circumstance, allowing full exhalation results in improvement; in the latter, the patient can be ventilated satisfactorily with a bag and 100% oxygen. Patient-ventilator asynchrony may indicate incorrect ventilator mode selection, improper flow trigger sensitivity for A/C or SIMV modes, dynamic hyperinflation, or poor tolerance of mechanical ventilation despite sedation. In the last case, there is an indication for increased sedation and, if necessary, neuromuscular blockade. Patients treated in the ED with NPPV generally should not be given sedatives or major analgesics because preservation of respiratory drive is essential to the use of this technique. Small, incremental doses of benzodiazepines for patients who have difficulty tolerating the face mask or nose mask may be useful.18 Successful application of noninvasive methods is an acquired skill that takes advantage of not just drugs but of a calming bedside approach to frequently terrified patients. The authors’ experience suggests that allowing family members to stay at the bedside to offer reassurance is often very helpful during the use of NPPV. Intrinsic PEEP is an important issue, usually in patients with obstructive lung disease. In these cases, the expiratory flow rate is less than normal because of diminished elastic recoil from small airway obstruction (in emphysema) or because of dynamic airflow obstruction during exhalation (in reactive airway disease), or both. The time needed for intrapulmonary pressures to fall to ambient levels at the end of exhalation is prolonged. In spontaneously breathing intubated patients, iPEEP contributes to respiratory failure because this pressure must be matched by deep negative pressures generated by the respiratory bellows in order to initiate inhalation.19 In mechanically ventilated patients, failure to anticipate prolonged expiration in patients with chronic obstructive or severe reactive lung disease risks setting a respiratory rate too high to allow complete exhalation. Breath stacking results, and unexpectedly high PIPs, patient distress, and hypotension can occur. In patients with chronic lung disease or severe asthma exacerbations who suddenly develop hypotension or become difficult to mechanically ventilate, an appropriate measure is to tem porarily discontinue mechanical ventilation by switching to bag-valve breathing with deliberately prolonged exhalation. If this corrects the problem, then mechanical ventilation can be resumed by either using a slower respiratory rate or often with the aid of a respiratory therapist, customizing the inspiratory/expiratory duty cycle to allow the patient more time to exhale.
Special Clinical Circumstances In the following five common clinical indications for mechanical ventilation in the ED, special fine-tuning adjustments to the guidelines offered previously may be appropriate (Table 2-1).
27
Table 2-1 General Guidelines for Initial Invasive Ventilator Settings in Various Clinical Settings
COPD exacerbation, respiratory acidosis Cardiogenic pulmonary edema ARDS Hypovolemic shock
Fio2 (%)
TV (ML/KG)
RATE (BREATHS/MIN)
I/E RATIO
PEEP (CM H2O)
CMV, A/C, IMV, SIMV
100
8–10
10–12
1 : 2
0–5
CMV, A/C, IMV, SIMV PHC‡, CMV, IMV, SIMV CMV, A/C, IMV, SIMV PHC‡, CMV, IMV, SIMV CMV, A/C, IMV, SIMV
100 100 100 100 100
5–10 5–8 5–10 5–8 8–10
8–12 6–10 10–12 8–12 10–12
1 : 4 1 : 4 1 : 3–1 : 4 1 : 3–1 : 4 1 : 2
2.5–10† 2.5–10† 2.5–10† 2.5–10† 2.5–15
CMV, A/C, IMV, SIMV Inverse ratio CMV, A/C, IMV, SIMV
100 100 100
6–8 6–8 8–10
20–25 8–12
1 : 2 1 : 1–1 : 2.1 1 : 2
2.5–10 2.5–5 0–5
§
*CMV is the appropriate mode when the patient is apneic or paralyzed. † PEEP in air-trapping diseases should not exceed measured intrinsic PEEP. ‡ PHC is permissive hypercapnia, a ventilatory strategy that can be employed in multiple modes. § Rate should be set based on desired Paco2. A/C, assist/control (ventilation); ARDS, acute respiratory distress syndrome; CMV, controlled mechanical ventilation; Fio2, fraction of inspired oxygen; I/E, inspiratory/expiratory (ratio); IMV, intermittent mandatory ventilation; PEEP, positive end-expiratory pressure; SIMV, synchronized intermittent mandatory ventilation; TV, tidal volume. When using a pressure-cycled ventilator, set pressures to deliver desired TV. After 20 minutes on initial settings, arterial blood gases should be checked so that settings and down-titrations of Fio2 can be modified appropriately.
Acute Exacerbation of Chronic Obstructive Pulmonary Disease
Status Asthmaticus
In treating patients with COPD on the ventilator, respiratory acidosis should be corrected gradually over hours. Overcorrection or too-rapid correction of hypercapnia and acidosis may result in metabolic alkalosis, hypokalemia, and hypophosphatemia.20 Hypoxemia usually is easily correctable by increasing Fio2. Target values for Pao2, Paco2, and pH should reflect the patient’s predicted (or known) baseline function rather than usual “normal” values. The other major goal in the mechanical ventilation of patients with COPD is normalization of lung volume. Air trapping and resultant iPEEP in a patient with COPD increase the work of breathing and the likelihood of barotrauma with mechanical ventilation. Strategies used to address this problem center on reducing iPEEP. When inadequate expiratory time is allowed in the COPD patient, air trapping is exacerbated with each inspiration; this dynamic hyperinflation eventually results in a sufficiently high iPEEP that any additional breath necessarily overinflates the thorax. The immediate remedy for this problem is to disconnect the patient from the ventilator momentarily, allowing complete exhalation. The ongoing solution is to build adequate expiratory time into the ventilator settings. The rate should be kept as low as possible for patients with COPD, and the expiratory time should be maximized by increasing the I/E ratio to 1 : 3 or 1 : 4. The TV also should be minimized to reduce exhaled volumes. Often patients with COPD require higher flow rates (≥100 L/min) during inspiration to minimize inspiratory time. This approach allows more of the ventilatory cycle to be spent in exhalation. Each of these modifications in the settings reduces iPEEP. The iPEEP also may be reduced by the use of bronchodilators and corticosteroids. These agents increase inspiratory muscle strength and reduce the amount of secretions in the bronchial lumen, both of which decrease the work of breathing. Finally, iPEEP can be replaced in part by extrinsic PEEP. PEEP at a level of no more than the measured iPEEP (some authors suggest no more than 85% of iPEEP) unloads the work required to maintain iPEEP and allows the recruitment of the muscles providing the inspiratory effort. A consensus statement has suggested that BiPAP should be the initial ventilatory assistance modality of choice in COPD exacerbation.19
Interventions aimed at reducing hypercapnia in ventilated patients with status asthmaticus may result in dynamic hyperinflation and barotrauma. The best approach in these patients, similar to that used in COPD, is small TV and high inspiratory flow rates to reduce inspiratory time and peak airway pressures. Airway pressures also can be lowered by permissive hypercapnia, which uses a low TV (5–8 mL/kg) and relatively low rates (8–10 breaths/min) to prevent excessive alveolar distention. Paco2 is allowed to remain at supranormal values without ventilatory correction. The primary goal of permissive hypercapnia is the reduction of lung volume (and iPEEP) and the risk of barotrauma, while maintaining adequate oxygenation. This approach has not been studied thoroughly under controlled conditions, but permissive hypercapnia has potential applicability in status asthmaticus, acute respiratory distress syndrome, and severe COPD exacerbations.21 Occasional external chest compression also may be useful in assisting exhalation in asthma. This concept has shown promise in animal and uncontrolled human trials.22
Acute Lung Injury Acute lung injury (ALI) typically develops over several hours but may become evident in the ED. Pathophysiologically it is characterized by heterogeneous noncardiogenic pulmonary edema and surfactant failure that produces poor lung compliance and hypoxia. Pressure-limited special modes may be the optimal means of ventilating patients with ALI, but these techniques are often unavailable in the ED. On standard ventilators, settings should be adjusted to keep PEEP and Fio2 as low as possible. Small TV (6–8 mL/kg) and fast rates (20–25 breaths/min) are indicated.23 Although PEEP is considered primary therapy for ALI, these patients are highly susceptible to barotrauma. The risk of oxygen toxicity in ALI also is high and can be minimized by reducing the inspired oxygen concentration to the lowest level that maintains safe hemoglobin saturation levels. Sustained supraphysiologic oxygen tensions worsen inflammation. Iatrogenic barotrauma includes both mechanically induced tissue injury and the introduction of extrapulmonary air, such as pneumothorax or
Chapter 2 / Mechanical Ventilation and Noninvasive Ventilatory Support
Overdose in otherwise healthy patient Status asthmaticus
MODE*
28
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
BOX 2-2
Ventilator Recommendations for Acute Lung Injury (ARDSNET Reference)
A/C, volume-cycled ventilation Reduce TV to 6–8 mL/kg ideal body weight Plateau pressures < 30 cm H2O (2.9 kPa); may require reducing TV as low as 4 mL/kg Wean Fio2 to maintain a saturation of 88–95% Strategic use of PEEP to permit lower Fio2 and reduce risk of oxygen toxicity24 A/C, assist/control (ventilation); Fio2, fraction of inspired oxygen; PEEP, positive end-expiratory pressure; TV, tidal volume.
pneumomediastinum, due to airspace rupture. Repetitive opening and closing of alveoli lead to shear stress and worsening inflammation. Box 2-2 lists recommendations for minimizing the potential complications of PPV in the setting of ALI; these settings are often appropriate for other applications of mechanical ventilation as well. Collectively, this is often referred to as open lung approach.25
Cardiogenic Shock and Pulmonary Edema The ventilatory management of pulmonary edema with cardiogenic shock is complicated by the adverse effect of applied PEEP, which is ordinarily a primary mode of therapy for pulmonary edema, on cardiac output. A reasonable compromise is to use only sufficient PEEP to allow titration of the inspired oxygen concentration. Patients in cardiogenic shock require invasive hemodynamic monitoring. Hemodynamic and ventilatory parameters must be followed in tandem to achieve optimal benefit. Patients with less severe pulmonary edema secondary to CHF often may benefit from noninvasive ventilatory support while receiving appropriate pharmacologic therapy. CPAP and BiPAP have been studied in this setting, and one study indicated that there may be little difference between the two modes in either safety or efficacy.26
Hypovolemic Shock Appropriate volume resuscitation is the optimal means of managing respiratory compromise after trauma or with other causes of hypovolemic shock (e.g., massive gastrointestinal hemorrhage). PPV may exacerbate hypotension in hypovolemic patients. Patients in shock should be ventilated with 100% oxygen at a rate and TV predicted to produce near-physiologic Paco2. PEEP generally should be avoided until circulating volume is restored.
■ OUTCOMES When NPPV is successful (i.e., when ETT-mechanical ven tilation is avoided), several potential therapeutic, patient comfort, and fiscal benefits are derived. The advantages of NPPV over ETT-mechanical ventilation include (1) preservation of speech, swallowing, and physiologic airway defense mechanisms; (2) reduced risk of airway injury; (3) reduced risk of nosocomial infection; and (4) decreased length of stay in, and reduced need for admission to, the intensive care unit (ICU) because less weaning and less intensive monitoring are necessary.18,27-30 Patients treated with NPPV have an increased risk of pulmonary barotrauma, aerophagia, and pressure stress to the face compared with intubated and ventilated patients. (BiPAP is a
leak-tolerant system so that pressure sores are a much less common complication of extended BiPAP support than of CPAP.) In two series, patients successfully supported in the ED with NPPV usually could be admitted to a telemetry unit instead of an ICU, with significant cost savings.13,29 Uncontrolled studies without definitive inclusion criteria found NPPV successful in avoiding ETT-mechanical ventilation in 60 to 90% of patients.3,4 One study31 showed increased mortality rates from the use of NPPV, which was attributed to a delay in necessary intubation; the study suffered from selection bias, however, and its results have not been corroborated.6 In a prospective trial of NPPV in acute cardiogenic pulmonary edema, NPPV produced more improvement in subjective dyspnea, tachycardia, and acidosis at 1 hour compared to standard oxygen therapy, but did not produce a difference in mortality at 30 days.10 Treatment of mechanically ventilated patients usually extends beyond the ED. Adequate resuscitation and stabilization are the primary ED goals so that more focused therapy of the underlying problems can be pursued in the ICU. Occasionally, patients are extubated in the ED, most often those intubated solely for airway protection when the initial insult has been reversed or adequately tolerated. Adequate ventilatory drive and oxygen must be confirmed before ED extubation; before attempting to discontinue mechanical ventilation, patients should have a respiratory rate less than 30 breaths/ minute, PEEP 5 cm H2O or less, and Pao2 greater than 60 mm Hg with Fio2 less than 60%. NPPV may serve as a bridge between mechanical and spontaneous ventilation.
KEY CONCEPTS ■
Not all patients who require intensive ventilatory support in the ED require endotracheal intubationmechanical ventilation. Careful patient selection for noninvasive ventilatory support may spare some patients invasive therapy and its attendant risk of complications. ■ TV of 6 to 8 mL/kg, rate of 12 to 14 breaths/minute, Fio2 of 1.0 is a reasonable starting point for mechanical ventilator settings for patients whose primary pathology is not pulmonary. For patients with pulmonary pathology, settings specific for the cause of the patient’s respiratory failure should be used, then carefully adjusted based on clinical response. ■ Sudden difficulty in the treatment of patients receiving NPPV is usually the result of intolerance, inadequate ventilation or oxygenation, or air trapping. Intolerance should not be assumed until other causes are excluded. ■ Sudden difficulty in the treatment of patients receiving mechanical ventilation should prompt a quick and systematic evaluation for tube, ventilator, airway pressures, and physiologic problems; such difficulties must not be automatically assumed to result from undersedation. This frequently starts with disconnecting the ventilator from the tube and temporarily bagging with Fio2 of 1.0. ■ The goal of therapy in many patients is not prompt normalization of blood gas parameters but may be relief of significant work of breathing. Slow correction to a patient-specific baseline likely results in significantly better clinical outcomes. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 3
Monitoring the Emergency Patient
Michael F. Murphy and Baruch Krauss
■ PERSPECTIVE To monitor means to measure or observe a physiologic parameter either continuously or intermittently. Monitoring devices provide a “snapshot in time” and a window into the clinical status of the patient, detecting deterioration, tracking improvement, or measuring the effects of interventions. Monitoring parameters such as clinical observation, routine vital sign measurement, and electrocardiographic monitoring are basic tools in the practice of emergency medicine. This chapter focuses on the following monitoring modalities: oxygenation monitoring with pulse oximetry, ventilation monitoring with end-tidal carbon dioxide (ETco2) measurement and waveform analysis, and hemodynamic monitoring with noninvasive blood pressure (BP) measurement. Fetal monitoring immediately after maternal trauma is also briefly discussed.
■ NONINVASIVE BLOOD PRESSURE MEASUREMENT Automatic noninvasive BP measurement has become a popular and, if applied appropriately, an accurate method of determining BP. Advantages include more time for staff to attend to other tasks, timed repetition of BP measurements, continuous display of the systolic pressure, and a multiparameter display (e.g., systolic, diastolic, and mean BP; pulse rate). Two types of noninvasive BP measurement devices are currently available: 1. Cuff-type 2. Radial arterial noninvasive waveform analysis The noninvasive cuff-type devices use a detection system based on auscultatory, oscillometric, or Doppler principles.1,2 Automatic oscillometric devices determine BP by electronically determining the pulse amplitude. This method and Doppler are the most accurate of the indirect methods. The cuff is automatically inflated at predetermined intervals to a preset level. As the machine gradually deflates the cuff, it senses the amplitude of the oscillations (pulsations) transmitted to the cuff by movement of the arterial wall under the cuff. An abrupt increase in the magnitude of the oscillation signals an opening of the artery and an increase in volume under the cuff; this is the systolic pressure. The magnitude of the oscilla-
tion increases to a peak and then falls rapidly. The point where there is no longer an alteration in the magnitude of the oscillation is the diastolic pressure. Some devices calculate the mean arterial pressure (MAP); others identify it as the cuff pressure at the point of largest oscillation.1 Noninvasive cuff-type oscillometric devices can be cycled every 15 to 20 seconds in the “STAT” mode when necessary to provide rapid but intermittent BP readings.3 Accuracy during rapid cycling is the same as during less frequent sampling, but to prevent pressure injury from the high frequency of cycling, most cuff-type automatic BP devices revert to the intermittent mode after a brief period of rapid cycling. The shortcomings of cuff-based noninvasive BP monitoring are those of any cuff measurement technique; patients with obese arms, uncooperative moving patients, and patients with very high or very low BP. Even with these limitations, automatic devices are more accurate and reliable than manual auscultation in patients with very low or high BP because the sensing devices are more sensitive than the human ear.2 The cycle length of the inflation-deflation sequence of the older devices was exceedingly long and led to frequent failure. The newer devices have rectified this problem. A newer method of continuous, noninvasive BP monitoring measures radial artery BP and pulse rate every 12 to 15 beats. The Vasotrac (Medwave Inc., Arden Hills, St. Paul, MN) device measures BP and pulse rate and displays a radial arterial pressure waveform.4 It consists of a reusable circular sensor (diameter 1.20″; width 0.35″) which is strapped over the radial artery at the wrist. The wrist sensor module is designed to measure only the pulsatile energy perpendicular to the artery, using cyclical compression and decompression. The processor requires 12 to 15 consecutive beats without interference (movement artifacts) to obtain adequate energy information to generate the pulsatile calibrated beat.4 Although the device is expensive and requires the patient to remain relatively still, a limited number of studies have demonstrated that this noninvasive method of continuous BP measurement is comparable to that provided by an invasive arterial catheter.4-6 The most accurate method of measuring BP is with an intraarterial catheter transduced to an electronic display. The ability to identify beat-to-beat variability, respiratory variation, and longer trends is unsurpassed. In addition, arterial catheter placement enables frequent sampling of arterial blood without additional arterial punctures. Arterial pressure monitoring is 29
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
30
used increasingly in EDs, particularly as lack of available beds in the intensive care unit mandates longer stays in the emergency department (ED) for critically ill patients. The risk of arterial injury or thrombosis related to arterial line insertion is low, but real, and can result in vascular compromise. Situations when noninvasive BP measurement may prove inadequate and invasive monitoring via an arterial catheter should be considered include 1. Exceedingly high (>250 mm Hg systolic) or low (20 breaths/min or Paco2 1.0 mL/kg/hr), reduced (0.5–1.0 mL/kg/hr), or severely reduced (38° C or 90 beats/min 3. Respiratory rate >20 breaths/min or Paco2 12,000/mm3, 10% band neutrophilia Severe Sepsis SIRS with suspected or confirmed infection and associated with organ dysfunction or hypotension; organ dysfunction may include presence of lactic acidosis, oliguria, or altered mental status Septic Shock SIRS with suspected or confirmed infection with hypotension despite adequate fluid resuscitation; septic shock should still be diagnosed if vasopressor therapy has normalized blood pressure Hemorrhagic Shock Simple Hemorrhage Suspected bleeding with pulse 100 and pink Observational care
Persistent cyanosis
B
Effective ventilation, Provide positive-pressure HR > 100 and pink Postresuscitation care ventilation* HR < 60
30 sec
C
HR > 60
• Provide positive-pressure ventilation* • Administer chest compressions HR < 60
D
ventilation distends the stomach and worsens respiratory distress because of the presence of the stomach in the chest cavity. Infants with meningomyelocele should not be placed on their backs to avoid pressure on the defect, but on their stomachs or sides. The resuscitation should be conducted in this position if possible. The spinal defect should be covered with warm sterile gauze pads soaked in warm sterile saline and covered with a plastic covering. Infants with gastroschisis or omphalocele should be resuscitated as needed; in the same manner as for patients with meningomyelocele, the defect should be covered with an occlusive plastic covering to decrease water and heat loss. Because newborns are obligate nose breathers, bilateral choanal atresia causes upper airway obstruction and respiratory distress. It is diagnosed by the inability to pass a catheter through either nare into the oropharynx. An oral airway bypasses the obstruction. Patients with Pierre Robin sequence have small jaws and large tongues leading to upper airway obstruction. A nasal or oral airway should be able to bypass the
Administer epinephrine and/or volume*
obstruction; if not, intubation may be necessary. It is technically difficult to intubate a patient with Pierre Robin sequence, so a laryngeal mask airway (LMA) may be placed instead.27 Consultation with anesthesiology may be needed.
■ PREPARATION To maximize the effectiveness of resuscitation, the emergency department should have a prestocked drug pack, standardized equipment (Box 9-1), and staff familiar with newborn resuscitation.4,22,20,28 The pediatric Broselow Emergency Tape has a section that can be used to determine equipment size and drug dosages for newborn resuscitation for infants weighing greater than or equal to 3 kg.29,30 It is crucial to use universal precautions and to wear gown, gloves, and eye protection during neonatal resuscitation. For adequate preparation, the heat source must be turned on early, and the resuscitation table must be warm when the newborn is placed on it. Equipment of proper size is essential, especially respiratory equipment because it is most likely to be
Chapter 9 / Neonatal Resuscitation
• Term gestation? • Amniotic fluid clear? • Breathing or crying? • Good muscle tone?
80
PART I ■ Fundamental Clinical Concepts / Section One • Critical Management Principles
BOX 9-1
Equipment Needed for Neonatal Resuscitation
1. Gown, gloves, and eye protection (universal precautions) 2. Blankets (to warm and dry infant) 3. Radiant warmer 4. Bulb syringe 5. Suction and suction catheters (French #5, #8, and #10) 6. Self-inflating bags (450 and 750 mL) 7. Masks (premature, newborn, and infant sizes) 8. Laryngoscope with straight blades (No. 0 and 1) 9. Endotracheal tubes with stylets (2.5, 3, and 3.5 mm) 10. Scissors and tape to stabilize endotracheal tube 11. Meconium aspirator 12. Umbilical catheters (French #3.5 and #5) 13. Hemostats, sterile drapes and gloves, povidone-iodine solution, scalpel, umbilical tape, suture, and three-way stopcock for umbilical vessel catheterization
BOX 9-2 1. 2. 3. 4. 5.
Maternal History Questions
What is the estimated gestational age? Is this a multiple gestation? Is meconium present? Is there a history of vaginal bleeding? Were medications given or drugs taken?
used, and ventilation is the key to most resuscitative efforts. Appropriately sized self-inflating devices and ventilation masks decrease complications from overventilation and prevent injury or inability to ventilate because of improper mask fit. If available, additional information may be helpful in preparing for resuscitation (Box 9-2). The estimated gestational age provides information about possible prematurity and associated complications. Multiple births require more equipment and personnel, and the newborns are at greater risk for prematurity and its complications. Meconium present in the amniotic fluid may require suctioning of the neonate’s trachea after delivery. A history of maternal vaginal bleeding increases the likelihood of hypovolemic shock and respiratory distress in the newborn. A history of medication administration or drug use may provide clues to the cause of respiratory depression in the neonate.
■ MANAGEMENT An organized approach is key to successful resuscitation outside and inside the delivery room. The American Heart Association and the American Academy of Pediatrics have published a neonatal flow algorithm (see Fig. 9.1)—a stepwise approach to neonatal resuscitation. These steps are discussed in this section.1
Dry, Warm, Position, Suction, Stimulate, Assess Need for Further Intervention To prevent complications caused by hypothermia, all newborns should be dried off as soon as possible after delivery and placed under a radiant heat source. Wet blankets should be replaced after drying with dry, preferably warm blankets. Next, the neonate should be positioned to maximize air entry and avoid obstruction of airflow. Because of the infant’s relatively large occiput and anterior airway, an open airway is best
Table 9-2 Apgar Score SIGN
0
1
2
Heart rate (beats/min) Respirations Muscle tone
Absent
Slow (100
Absent Limp
Slow, irregular Some flexion
Reflex irritability Color
No response
Grimace
Blue, pale
Pink body, blue hands/feet
Good, crying Active, good flexion Cough, sneeze Pink
achieved with the neck in slight extension. The slightly extended position is best accomplished by placing a rolled diaper or small towel under the infant’s shoulders. Placement under the neck is not useful, and a towel that is too large and under the shoulders leads to hyperextension at the neck and possible airway occlusion. If meconium is present and the newborn has poor respirations, poor tone, or bradycardia (HR < 100 beats/min), the trachea should be suctioned with an ETT before any other intervention. If no meconium is present, the newborn may be suctioned with bulb or mechanical suction (2.5 kg) and less than 30 mg/dL in a premature infant (25%; normal pupillary responses; normal sensation; normal reflexes Improves with rest May have a coexisting thymoma (CXR, chest CT)
Supportive care, ICU admission Neurology consult Edrophonium/neostigmine test Bedside spirometry Measure serum acetylcholine receptor antibody levels Tx: Anticholinesterase drugs— neostigmine; pyridostigmine
Organophosphate/ Carbamante Poisoning Immediate–3 wk
Neurotransmission Cholinergic crisis from inhibition of acetylcholine Neuropathy (weeks after exposure)
History of insecticide exposure Gastrointestinal symptoms, agitation, miosis, paralysis, diaphoresis, muscle weakness, bradycardia Cramping muscle pain, distal numbness and paresthesias, progressive muscle weakness; decreased reflexes; can develop flaccid/ wasted leg muscles
Decontamination Supportive care, ICU admission Atropine Pralidoxime (2-Pam)
Tetanus Toxin 3 wk 10–60% fatality
Neurotransmission Toxin interferes with release of inhibitory transmitters including GABA; results in motor nerve hyperactivity
Immunization status History of cutaneous infection Trismus, laryngospasm, painful muscle spasms and rigidity (opisthotonos), autonomic instability
Supportive care, ICU admission Débridement of wounds Tetanus immunoglobulin Penicillin for the infection High-dose benzodiazapines Neuromuscular blockade
Tick Paralysis Toxin 2–7 days Rocky Mountain wood tick and the American dog tick are most common in the USA
Neurotransmission Toxin reduces motor neuron action potential and the action of acetylcholine
History of outdoor activities/tick bite Progressive, ascending, flaccid weakness over several hours may lead to respiratory failure; may present as acute ataxia without muscle weakness; decreased or absent reflexes; ophthalmoplegia and bulbar palsy can occur
Removal of the embedded tick (look at the hairline/in the scalp) Supportive care Full recovery if tick removed; 10% fatality if not recognized
Ciguatoxin Toxin 12–24 hr; neurologic symptoms can last months
Neuropathy Toxin causes cell membrane excitability and instability
History of ingestion of large, tropical fish Diarrhea, abdominal pain, nausea, and vomiting are followed by painful paresthesias, ataxia, altered hot/cold perception, myalgias, bradycardia, and hypotension Rarely, death occurs through respiratory failure
Supportive care, ICU admission Atropine for bradycardia Hydration IV mannitol can be helpful
Diphtheria Toxin 2 wk–3 mo after infection
Neuropathy Lower motor neuron
Immunization status History of throat infection with pseudomembrane; cutaneous infection Palatal weakness, impaired pupillary responses, generalized sensorimotor polyneuropathy; respiratory failure; motor weakness of the proximal muscle groups and extending distally
Supportive care, ICU admission Equine diphtheria antitoxin Erythromycin or penicillin G for 14 days to halt toxin production, treat localized infection and prevent transmission of organisms Immunization
89
Table 11-1 Neuromuscular Diseases: A Brief Description—cont’d MECHANISM
HISTORICAL FEATURES/EXAM FINDINGS
ED MANAGEMENT
Gullian-Barré Syndrome Idiopathic 1–4 wk 75% recover 5% fatality
Neuropathy Lower motor neuron Immune-mediated polyneuropathy Multiple variants
May have a history of infection; viral infection; Campylobacter jejuni in 15–40% Symmetrical ascending motor neuropathy; decreased/absence reflexes; mild sensory involvement; autonomic dysfunction; can progress to respiratory compromise
Lumbar puncture: CSF with elevated protein but normal WBC Bedside spirometry Plasmapheresis and IVIG Consider ICU admission Neurology consult
Transverse Myelitis Idiopathic, postinfectious, autoimmune Rapid onset (hours–days)
Neuropathy Upper motor neuron Axonal demyelination
Loss of spinal cord functions with symptoms depending on the level of the lesion; thoracic is most common Acute, focal back pain; distal muscle weakness; abnormal sensation; urinary retention/loss of bowel control; muscles may be flaccid; decreased or absent reflexes initially Differentiate from spinal cord compression, trauma or infarct; may be the first sign of multiple sclerosis
Supportive care, ICU admission if C-spine level for respiratory support Spine radiograph to evaluate for boney lesion Stat MRI/CT myelogram Decompress bladder
Electrolyte Imbalance
Myopathy
History of nausea/vomiting/diarrhea History of renal failure, alcohol dependence, new medication Ascending symmetric muscle weakness with normal to diminished reflexes
ECG Electrolyte panel: Na, K, Cl, PO4, Ca, and Mg Renal function Correct the abnormality; close hemodynamic monitoring
Polymyositis Autoimmune
Myopathy
History of connective tissues disorders or cancer Progressive at a variable rate; muscle weakness and wasting; ascending pattern with proximal limb and girdle muscle involvement; muscle pain; dysphagia; respiratory difficulty; can have an erythematous periorbital and eyelid rash (dermatomyositis)
Elevated CPK, rhabdomyolysis rare Normal ESR Supportive care Corticosteroids
Adapted and expanded from LoVecchio, et al: Approach to generalized weakness and peripheral neuromuscular disease. Emerg Med Clin North Am 15:605, 1997. CDC, Centers for Disease Control and Prevention; CPK, creatine phosphokinase; CSF, cerebrospinal fluid; CT, computed tomography; CXR, chest radiograph; ESR, erythrocyte sedimentation rate; GABA, γ-aminobutyric acid; GI, gastrointestinal; ICU, intensive care unit; IVIG, intravenous immunoglobulin; MRI, magnetic resonance imaging; WBC, white blood cell count.
electrolyte disturbances or generalized medication side effects. In adults older than age 50, particularly women, the complaint of generalized weakness should prompt consideration of cardiac ischemia. In adults older than age 65, a complaint of weakness may be the only symptom of a serious infection, electrolyte disturbance, or cardiovascular compromise. When this complaint is combined with a recent fall, altered mental status, or urinary incontinence, urosepsis should be considered. When this complaint is accompanied by the report of poor sleep, dyspnea, or decreased exercise tolerance, acute coronary syndrome or heart failure should be considered. Consideration should be given to situational orthostasis resulting in a sensation of weakness accompanied by a presyncopal feeling; examples include postparandial hypotension and post-tussive or micturition near-syncope.
The neurologic exam should focus on clarifying if the patient is experiencing true loss of strength along with the distribution of the deficits (Table 11-3). A complete examination, including cranial nerves, and gait, where possible, is helpful. The motor exam should be systematic and thorough. Muscle bulk, strength, tone, and the presence or absence of abnormal movements should be noted. Sarcopenia (age-associated loss of muscle mass and function) is normal in the older adult. In this situation, the loss of power is uniform in all limbs. Walking on heels, toes, and in tandem is a good test of strength as well as coordination and proprioception. Gait apraxia has a wide differential and should prompt investigation for cerebellar abnormality; normal pressure hydrocephalus should be considered in the patient who has simultaneous incontinence and decreased cognitive function. Fine muscle fasciculations typically point to an LMN disorder, whereas spasticity, greater in the flexors than extensors, is seen in UMN lesions.
Examination
Ancillary Testing
Fever, hypotension, tachycardia, or tachypnea may provide clues regarding the source of the patient’s complaint (Table 11-2). If severe weakness is present, an assessment of the patient’s ability to maintain the airway and the adequacy of respiration is indicated (Fig. 11-1).
Patients presenting with weakness can have myriad underlying abnormalities. Although testing will be guided by the history and exam, virtually all patients require a complete blood count to evaluate for anemia or blood loss and serum electrolytes, glucose, and creatinine. An electrocardiogram
Chapter 11 / Weakness
DISEASE
90
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
Patient with a complaint of WEAKNESS
Vital signs Assess airway, breathing, circulation IV access, cardiopulmonary monitoring Is the patient protecting the airway and ventilating adequately?
If NO, proceed to immediate intubation.
Is the patient maintaining the blood pressure and pulse?
If NO, proceed to appropriate stabilization.
Complete the history and physical exam
Non-neuromuscular weakness • BROAD differential • Assessment guided by history and exam
• ECG • Labs: CBC, chemistries, LFTs, cardiac enzymes, blood cultures, U/A, drug levels • Chest radiograph • Consider brain CT
• Supportive care; close monitoring • Correct abnormalities • Consider admission
Neuromuscular weakness • Detailed neurologic exam • Deep tendon reflexes - Muscle tone - Strength - Ascending vs. descending weakness - Sensation - Plantar reflexes
• Bedside spirometry • Supportive care; close monitoring • Neurology consult • Consider imaging: CT/MRI • Consider LP • Consider antidotes • Consider ICU admission
Indications for intubation: • Severe fatigue • Inability to protect the airway and/or handle secretions • Rapidly rising PaCO2 • Hypoxemia • Forced vital capacity < 12 mL/Kg • Negative inspiratory force < 20 cm H2O
Any patient who appears to have respiratory compromise or for whom Guillian-Barré syndrome or myasthenia gravis is possible should undergo bedside spirometry. A forced vital capacity (FVC) of less than 10 to 12 mL/kg or a negative inspiratory force (NIF) of less than 20 cm H2O is an indication for respiratory support. An arterial blood gas test for carbon dioxide tension or capnography may also be helpful.
■ DIFFERENTIAL DIAGNOSIS AND INITIAL MANAGEMENT The differential diagnosis of weakness is very broad, and at times, a definitive diagnosis is impossible in the span of an ED visit. Ensuring appropriate disposition and follow-up is particularly important in these patients. In a patient with neuromuscular weakness, the respiratory drive is preserved, but the ability to ventilate adequately can be impaired and the patient may complain of dyspnea. Patients with rapid progression of weakness may require early airway intervention and mechanical ventilation. Warning signs of worsening respiratory status include the inability to lift the head, ineffective cough, alteration of the voice, and difficulty controlling secretions. As a crude measure of vital capacity, the patient’s inability to count to 20 in a single exhalation suggests that the FVC is compromised; when the patient can only count to 10, the FVC can be estimated at 1 L and preparations to intubate should be made. About 30% of patients with GuillianBarré syndrome require mechanical ventilation, and several studies have demonstrated that an elective, controlled intubation leads to better outcomes in terms of ventilation-associated pneumonia and total time on the ventilator3 (Fig. 11-2). Patients with neuromuscular weakness increase their respiratory rate to compensate for low tidal volumes, and the Paco2 is maintained within the normal range. If the patient develops muscle fatigue or increased muscle weakness, the Paco2 will rise and respiratory failure can occur quickly. In this situation, intubation, usually using a rapid sequence technique, is indicated. Succinylcholine should be avoided when a progressive denervation syndrome is suggested.4 The up-regulation and redistribution of acetylcholine receptors on denervated myocytes that occurs with succinylcholine can lead to significant hyperkalemia with administration of this drug (see Chapter 1). In addition, autonomic instability in these patients can make intubation challenging; the physician should anticipate the possibility of labile blood pressures and bradycardia.5 Bradycardia responds to atropine administration, and the blood pressure should be closely monitored but not necessarily treated as it can fluctuate rapidly.
Figure 11-1. Diagnostic management algorithm. CBC, complete blood
count; CT, computed tomography; ECG, electrocardiogram; ICU, intensive care unit; LFTs, liver function tests; LP, lumbar puncture; MRI, magnetic resonance imaging; Paco2, arterial partial pressure of carbon dioxide; U/A, urinalysis.
may indicate cardiac ischemia or hypo- or hyperkalemia. In patients with infectious symptoms and in elders with nonlocalized symptoms, a chest radiograph and urinalysis are helpful. Computed tomography or magnetic resonance imaging of the brain or spine are not generally indicated in the absence of focal findings, altered mental status, history of cancer, or anticoagulation with (even minor) trauma. Lumbar puncture can help confirm central nervous system infection when suggested, or Guillian-Barré syndrome when protein levels are elevated and a white blood cell count of 10 or fewer per milliliter strongly supports the diagnosis.
Special Situations Myasthenia Gravis: Myasthenic Crisis versus Cholinergic Crisis Myasthenia gravis is discussed in detail in Chapter 106. Myasthenic crisis refers to a rapid worsening of neuromuscular function with respiratory compromise. It occurs in approximately 15% of patients with this disease and can be triggered by infection (≈ 30% of cases), change in medications, metabolic derangement, or physical stress; a third of the time no cause is found.6 A crisis can be precipitated by a recent change in the dose of the patient’s anticholinesterase inhibitor, recent initiation or tapering of corticosteroids, and recent initiation of several commonly used medications (aminoglycoside and quinolone antibiotics, beta-blockers, and antiarrhythmic agents). Myasthenic crisis is associated with prolonged hospitalization and intubation periods of 1 to 3 weeks.
91
Table 11-2 Vital Signs: Weakness ELEVATED
DECREASED
POTENTIAL INTERVENTIONS / ANCILLARY TEST
Heart rate
Arrhythmia Blood loss Dehydration Hyperthyroidism Pain Serious infection
Electrolyte imbalance Medication effect (BB, CCB)
ECG Fluid bolus and reevaluate Orthostatic blood pressure/pulse Rate control based on ECG findings Antibiotics if infection suspected
Blood pressure
Hyperthyroidism Medication noncompliance Pain
Arrhythmia Blood loss Dehydration Medication effect (BB, CCB) Serious infection
ECG Fluid bolus and reevaluate Orthostatic blood pressure/pulse Pressors
Respiratory rate
Serious infection COPD/Asthma
Impending respiratory failure
Bronchodilators CXR Respiratory support: oxygen, BiPAP, intubation
Temperature
Serious infection Medication effect
Serious infection Environmental exposure
Antipyretics/cooling measures Passive rewarming ECG Infectious workup
Oxygen saturation
N/A
Serious infection COPD/asthma Impending respiratory failure
Bronchodilators CXR Respiratory support: oxygen, BiPAP, intubation
BB, beta-blocker; BiPAP, Bi-level positive airway pressure; CCB, calcium channel blocker; COPD, chronic obstructive pulmonary disease; CXR, chest radiograph, ECG, electrocardiogram.
Table 11-3 Physical Examination: Localizing Neuromuscular Lesions LOCATION OF LESION
DEEP TENDON REFLEXES
MUSCLE TONE
PLANTAR REFLEXES
STRENGTH
Upper motor neuron
Increased
Upgoing
Weak/paralysis
Lower motor neuron
Decreased or absent
Normal or absent
Weak/paralysis
Neuromuscular junction Muscle
Normal or decreased Normal or decreased
Normal (Increased/spastic as disease progresses) Decreased/flaccid (may see fasciculations) Decreased/flaccid Decreased/flaccid
Normal or absent Normal or decreased
Variable weakness pattern Constant/progressive Proximal > distal
Figure 11-2. Differential diagnosis algorithm: Weakness.
Non-neuromuscular
Neuromuscular
Critical: Hemodynamic instability Myocardial infarction Arrhythmia Severe infection/sepsis Respiratory failure Hyperkalemia
Critical: Potential for respiratory compromise Rabies Botulism Tetanus Organophosphate poisoning Myasthenia gravis crisis
Emergent
Emergent
Acute anemia Dehydration Metabolic disorder Hypothyroidism Diabetes Electrolyte imbalance Other Fatigue Psychiatric (anxiety, depression) Rheumatologic (fibromyalgia; SLE) Malignancy Renal or hepatic disease Metabolic disease Alcoholism and other toxin-related disease Malingering
Guillian-Barré syndrome Transverse myelitis Impingement syndromes Spinal cord infarction Electrolyte imbalance
Other Lambert-Eaton syndrome ALS Paraneoplastic syndrome Diphtheria Porphyria Drugs and toxins Tick paralysis Poliomyelitis
Chapter 11 / Weakness
VITAL SIGNS
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
92
Cholinergic crisis results from an excess of cholinesterase inhibitor medication that produces a flaccid muscle paralysis and generalized weakness. Respiratory failure may be present with or without other cholinergic symptoms. These two forms of crisis can be difficult to distinguish but are managed similarly with a focus on airway protection and ensuring adequate ventilation. Potential triggers for the crisis should be sought, and because many of these patients are on immunosuppressive medications, infection should be strongly considered. An urgent neurology consultation may be helpful, and an edrophonium challenge test may be performed if a myasthenic crisis is suggested (see Chapter 106).
Older Adults and Frailty In the older adult with a complaint of weakness, an accurate history can be difficult to obtain. In addition, comorbidities and polypharmacy can often complicate the presentation in these patients. Frailty is a biologic syndrome defined by decreased reserve and resistance to stressors and is an independent predictor of future functional decline, falls, hospitalization, and mortality.7,8 Hallmarks of this syndrome include generalized muscle weakness, poor endurance, weight loss, low physical activity, and slow gait speed. The prevalence of frailty in communitydwelling persons older than age 65 is approximately 7% and it increases with age and female gender.
Older adults with disability, frailty (as measured by their functional status, grip strength, and ability to ambulate), and comorbid chronic illness are at high risk for poor outcome after the ED visit.9 A thorough evaluation with attention to potential life-threatening conditions and a low threshold for admission and further evaluation is warranted in this population. One year post-ED visit with hospitalization, mortality in this population is in the range of 25%.
■ DISPOSITION Most patients presenting to the ED with a complaint of weakness have a non-neuromuscular cause for their complaint. The history, exam, and results of ancillary testing dictate treatment and disposition of these patients. Those patients who present with acute symmetrical neuromuscular weakness require a thorough evaluation with par ticular attention to their airway and ventilation. Disposition decisions should be made in conjunction with neurology consultation, and admission to the intensive care unit for close respiratory monitoring should be considered.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 12
Dizziness and Vertigo
Jonathan S. Olshaker
■ PERSPECTIVE An estimated 7.5 million patients with dizziness are seen each year in ambulatory care settings. It is one of the most common principal complaints in the emergency department.1 Benign paroxysmal positional vertigo, felt to be caused by loose particles in the semicircular canals, is the most common cause of vertigo with an incidence estimated to be 107 cases per 100,000 population per year.2 Dizziness in older persons is associated with a variety of cardiovascular, neurosensory, and psychiatric conditions and with the use of multiple medications.3 Among patients older than 60 years, 20% have experienced dizziness severe enough to affect their daily activity.4 In a study of 1000 outpatients, dizziness was the third most common complaint. Vertigo is defined more clearly as a sensation of disorientation in space combined with a sensation of motion. In 1921, Bárány published the first detailed description of benign paroxysmal positional vertigo. A complaint of “dizziness” is an imprecise term. The emergency department physician may believe that the patient will be difficult to interview and that the condition will be problematic to diagnose and treat. But in reality, most of these patients have an organic basis for symptoms that can be successfully identified and treated. The diagnostic process is consistently based on two basic concepts: deciding whether the patient has true vertigo and, if vertigo exists, deciding whether the cause is central or peripheral.5
Pathophysiology The maintenance of equilibrium and awareness of the body in relationship to its surroundings depend on the interaction of three systems: visual, proprioceptive, and vestibular. The eyes, muscles, joints, and otic labyrinths continuously supply information about the position of the body. Visual impulses, mediated through the higher brain centers, provide information about body position in space. Impulses from proprioceptors of the joints and muscles supply data about the relative positions of the parts of the body. Impulses from the neck are of special importance in relating the position of the head to the rest of the body. The sense organs of the visual, vestibular, and proprioceptive systems are connected with the cerebellum by way of the vestibular nuclei in the brainstem. Any disease that interrupts the integration of these three systems may give rise to symptoms of vertigo and disequilibrium. The vestibular apparatus helps maintain head position and stabilize head movement. It is housed in the inner ear, or laby-
rinth, which lies embedded in the petrous portion of the temporal bone, where it is vulnerable to trauma; blood-borne toxins; and infections in nearby structures, including the middle ear and meninges. The vestibular apparatus consists of three semicircular canals with their cristae and two otolithic structures: the utricle and saccule. The semicircular canals provide information about movement and angular momentum; the otoliths provide information about the orientation of the body with respect to gravity. The semicircular canals are paired structures that normally respond to motion in a symmetrical manner. With inner ear disease, the resting discharge or the discharge stimulated by motion can be altered in one ear. This alteration produces asymmetrical responses and results in the perception of vertigo. Freely moving debris within the semicircular canals can produce positional vertigo as the debris moves under the influence of gravity. Impulses leave the vestibular apparatus by the vestibular part of the acoustic nerve (cranial nerve VIII), enter the brainstem just below the pons and anterior to the cerebellum, and proceed to the four vestibular nuclei of the brainstem and to the cerebellum. From there, impulses travel along two pathways that contribute to the clinical manifestations of vertigo: the medial longitudinal fasciculus and the vestibulospinal tract. In individuals with healthy vestibular systems, these connections allow the eyes to compensate for body movement in different directions and to maintain a visual axis that is stable with respect to the environment. Nystagmus occurs when the synchronized vestibular information becomes unbalanced. Typically, it results from unilateral vestibular disease, which causes asymmetrical stimulation of the medial and lateral rectus muscles. This unopposed activity causes a slow movement of the eyes toward the side of the stimulus, regardless of the direction of deviation of the eyes. The cerebral cortex then corrects for these eye movements and rapidly brings the eyes back to the midline, only to have the process repeated. By convention, the direction of nystagmus is denoted by the direction of the fast “cortical” component. Nystagmus caused by vestibular disease tends to be unidirectional and horizontorotary. If the nystagmus is vertical, a central lesion (either brainstem or cerebral) is usually the cause. The vestibular nuclei send information to the lateral vestibulospinal tract, where they connect with motor neurons that supply the muscles of the extremities. This phenomenon explains the false steps or other body movements made by people with a defective vestibular apparatus who are 93
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
94
attempting to correct for an imagined change in position. Connections between the vestibular nuclei and the autonomic system account for the perspiration, nausea, and vomiting that commonly accompany an attack of vertigo. Connections between the vestibular nuclei and the cerebellum account for the modulating influence of this organ on motor activity.
■ DIAGNOSTIC APPROACH Differential Considerations Patients use the term dizzy to describe a variety of experiences, including sensations of motion, weakness, fainting, lightheadedness, unsteadiness, and depression. To clarify the picture, it is often helpful to have patients describe the sensation without using the word dizzy. True vertigo may be defined as a sensation of disorientation in space combined with a sensation of motion. There is a hallucination of movement either of the self (subjective vertigo) or the external environment (objective vertigo). Descriptions of light-headedness or feeling faint are more consistent with presyncope. The differential diagnosis for these patients should include dysrhythmias, myocardial infarction, sepsis, hypovolemia, drug side effects, and pulmonary embolism. For some patients, dizziness is simply a metaphor for malaise, representing a variety of other causes, such as anemia, viral illness, or depression. The primary focus of this chapter is to review the evaluation of the vertiginous patient. If the patient has true vertigo, the clinician must determine whether the cause is a peripheral lesion (e.g., of the inner ear) or a central process, such as cerebrovascular disease or a neoplasm. In most cases, peripheral disorders are benign, and central processes have more serious consequences. Occasionally, as in the case of a cerebellar hemorrhage, immediate therapeutic intervention is indicated. Acute suppurative labyrinthitis is the only cause of peripheral vertigo that requires urgent intervention. Box 12-1 lists causes of vertigo and identifies the peripheral, central, and systemic diagnoses. Table 12-1 summarizes the different characteristics of peripheral and central vertigo.
BOX 12-1 Causes of Vertigo Peripheral Causes Foreign body in ear canal Cerumen or hair against tympanic membrane Acute otitis media Labyrinthitis (suppurative, serous, toxic, chronic) Benign positional vertigo Ménière’s disease Vestibular neuronitis Perilymphatic fistula Trauma (labyrinth concussion) Motion sickness Acoustic neuroma* Central Causes Infection (encephalitis, meningitis, brain abscess) Vertebral basilar artery insufficiency Subclavian steal syndrome Cerebellar hemorrhage or infarction Vertebral basilar migraine Post-traumatic injury (temporal bone fracture) Postconcussive syndrome Temporal lobe epilepsy Tumor Multiple sclerosis Cervical spine muscle and ligamentous injury Systemic Causes Diabetes mellitus Hypothyroidism *Cause of peripheral vertigo that proceeds centrally.
of Peripheral and Table 12-1 Characteristics Central Vertigo CHARACTERISTIC
PERIPHERAL
CENTRAL
Onset Intensity Duration
Sudden Severe Usually seconds or minutes; occasionally hours, days (intermittent)
Direction of nystagmus
One direction (usually horizontorotary), never vertical
Effect of head position
Worsened by position, often single critical position None
Gradual or sudden Mild Usually weeks, months (continuous) but can be seconds or minutes with vascular causes Horizontal, rotary, or vertical (different directions in different positions) Little change, associated with more than one position Usually present
Pivotal Findings History The medical history is the most important source of information. A first key question is, “Does true vertigo exist?” Does the patient have a sensation of disorientation in space or a sensation of motion? The sensation of spinning usually indicates a vestibular disorder. Some nausea, vomiting, pallor, and perspiration accompany almost all but the mildest forms of vertigo. The presence of these symptoms without vertigo should suggest a different cause. The labyrinth has no effect on the level of consciousness. The patient should not have an associated change in mentation or syncope. A sensation of imbalance often accompanies vertigo, but true instability, disequilibrium or ataxia makes a higher likelihood of a central process.6 Because nystagmus accompanies acute vertigo, it is often helpful to ask members of the patient’s family if they have noted any unusual eye movements during the dizzy spells. This question is especially important in children unable to offer a concise history.7 Occasionally, the patient may be able to describe a flickering or oscillating visual field immediately after a change in position, such as rolling over in bed. In addition, interviewing family and other witnesses can often uncover
Associated neurologic findings Associated auditory findings
May be present, including tinnitus
None
evidence suggesting seizures, syncope, or imbalance unrelated to feelings of vertigo. The time of onset and the duration of vertigo are important clues to the cause. Episodic vertigo that is severe, lasts several hours, and has symptom-free intervals between episodes sug-
Physical Examination Vital Signs. In some cases, pulses and blood pressure should be checked in both arms. Most patients with subclavian steal syndrome, which also can cause vertebrobasilar artery insufficiency, have pulse or systolic blood pressure differences between the two arms. Head and Neck. Carotid or vertebral artery bruits suggest atherosclerosis. The neck is auscultated along the course of the
carotid artery from the supraclavicular area to the base of the skull. Vertigo can be caused by impacted cerumen or a foreign object in the ear canal. Accumulation of fluid behind the eardrum as a result of a middle ear infection may cause mild vertigo, as can occlusion of the eustachian tubes associated with an upper respiratory tract infection. A perforated or scarred eardrum may indicate a perilymphatic fistula, especially if the history includes previous trauma. Examination of the eyes is key in assessing a patient with vertigo or disequilibrium. The focus is on any pupillary abnormalities indicating third cranial nerve or descending sympathetic tract involvement or optic disk signs of early increased intracranial pressure. Extraocular movements should be assessed carefully. Relatively subtle ocular movement abnormalities can be the only clue to a cerebellar hemorrhage. A sixth cranial nerve palsy ipsilateral to the hemorrhage may result from early brainstem compression by the expanding hematoma. Internuclear ophthalmoplegia is recognized when the eyes are in a normal position on straight-ahead gaze, but on eye movement the adducting eye (cranial nerve III) is weak or shows no movement while the abducting eye (cranial nerve VI) moves normally, although often displaying a coarse nystagmus. This finding indicates an interruption of the medial longitudinal fasciculus on the side of the third cranial nerve weakness. It indicates brainstem pathology and is virtually pathognomonic of multiple sclerosis. Abnormal nystagmus is the cardinal sign of inner ear disease and the principal objective evidence of abnormal vestibular function. In nystagmus, the patient has difficulty maintaining the conjugate deviation of the eyes or has a postural control imbalance of eye movements. The abnormal jerk nystagmus of inner ear disease consists of slow and quick components. The eyes slowly “drift” in the direction of the diseased, hypoactive ear, then quickly jerk back to the intended direction of gaze. Positional nystagmus, induced by rapidly changing the position of the head, strongly suggests an organic vestibular disorder. The characteristics of nystagmus are one of the most valuable tools for distinguishing peripheral from central causes of vertigo (Table 12-2). Positional Testing. If nystagmus is not present at rest, positional testing can be helpful in determining its existence and characteristics. In the Hallpike maneuver, the patient is moved quickly from an upright seated position to a supine position, and the head is turned to one side and extended (to a headdown posture) approximately 30° from the horizontal plane off the end of the stretcher. The eyes should be observed for nystagmus and the patient queried for the occurrence of symp-
Characteristics of Nystagmus Table 12-2 Distinguishing with Central and Peripheral Vertigo CHARACTERISTIC
CENTRAL
PERIPHERAL
Direction
Any direction
Laterality Position testing effects: Latency Duration Intensity Fatigability Effect of visual fixation
Unilateral or bilateral
Horizontal or horizontorotary Bilateral
Short Sustained Mild Nonfatigable Not suppressed, may be enhanced
Long Transient Mild to severe Fatigable Suppressed
95
Chapter 12 / Dizziness and Vertigo
gests a peripheral labyrinth disorder. Vertigo produced primarily by a change in position also suggests a peripheral disorder. Vestibular neuronitis and benign positional vertigo fit this pattern. The presence of auditory symptoms suggests a peripheral cause of the vertigo, as in middle and inner ear problems, or a peripheral cause that progresses centrally, such as an acoustic neuroma. The abnormally hearing ear is usually the side of end-organ disturbance. Progressive unilateral hearing loss of several months’ duration may be the earliest symptom of an acoustic neuroma. Tinnitus occurs in most patients with acoustic neuroma and, along with vertigo, is what often prompts patients to seek medical attention. Hearing loss, vertigo, and tinnitus form the characteristic triad of Ménière’s disease. Are there associated neurologic symptoms? The patient or family members should be questioned about the time of onset of ataxia or gait disturbances. Ataxia of recent and relatively sudden onset suggests cerebellar hemorrhage or infarction in the distribution of the posterior inferior cerebellar artery or the superior cerebellar artery. The salient feature of chronic cerebellar disorders is a slowly progressive ataxia. True ataxia may be difficult to discern from the unsteadiness that occurs when a patient with significant vertigo attempts to walk. Vertiginous symptoms are common after head injury. The presence of recent head or neck trauma should be explored because vertiginous symptoms are common after both.8,9 Head injury can cause vertigo occasionally from intercerebral injury and more commonly from labyrinth concussion. Neck injury can cause vertigo from strain of muscle proprioceptors. In addition, vertebral artery injury has been seen resulting from activities such as chiropractic manipulation and even hair shampooing with marked hyperextension in a salon.10 It has clearly been shown that isolated vertigo can be the only initial symptom of cerebellar and other posterior circulation bleeds, transient ischemic attacks (TIAs), and infarction.11-13 One study showed that emergency physicians often did not make the correct diagnosis in patients with validated strokes or TIAs that presented with only vertigo.6 Risk factor assessment and symptom patterns can be extremely helpful in deciding which patients warrant imaging and admission. Older age, male sex, hypertension, coronary artery disease, diabetes mellitis, and atrial fibrillation put patients at higher risk. In addition, frequent episodes lasting only minutes or prolonged episodes of a day or more are more often associated with central processes.6,11,12 A recent retrospective study showed emergency physicians often failed to chart triggers and duration of dizziness, information that could potentially lead to increased likelihood of a more serious cause of symptoms.14 Past Medical History. Many medications have direct vestibulotoxicity. The most commonly encountered are the aminoglycosides, anticonvulsants, alcohols, quinine, quinidine, and minocycline. In addition, caffeine and nicotine can have wideranging autonomic effects that may exacerbate vestibular symptoms. The history of past and present illnesses should be explored, with specific questioning about the existence of diabetes, drug or alcohol use, and the risk factors mentioned earlier.
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
96
or stopping suddenly while walking. Patients with gait ataxia cannot perform heel-to-toe walking.
Ancillary Testing
Figure 12-1. Testing for positional vertigo and nystagmus.
toms. This test should be repeated with the head turned to the other side. Positive elicitation of symptoms and signs to one side or the other generally indicates vestibular pathology on that same side. This test should be performed with caution if vertebrobasilar insufficiency (VBI) is suggested because sudden twisting movements theoretically might dislodge atheromatous plaques (Fig. 12-1). Neurologic Examination. The presence of cranial nerve deficits suggests a space-occupying lesion in the brainstem or cerebellopontine angle. The corneal reflex is a sensory cranial nerve V and motor cranial nerve VII circuit. Its diminution or absence can be one of the early signs of an acoustic neuroma. Vertigo caused by eighth cranial nerve involvement is likely to be accompanied by a unilateral hearing loss. Patients cannot hear a tuning fork when it is held alongside the affected ear, but they can hear it when it is held against the mastoid process. Involvement of the eighth cranial nerve suggests an acoustic tumor. Seventh cranial nerve involvement causes facial palsy that affects the entire side of the face. In supranuclear facial paralysis, the forehead is spared because these muscles receive bilateral cortical innervation. The patient should be evaluated specifically for evidence of cerebellar dysfunction. This examination must be performed in bed and standing because truncal ataxia may be occult on testing of limbs in bed and may become obvious only when the patient has to sit, stand, or walk unaided. Dysmetria is the inability to arrest a muscular movement at the desired point. Dysmetria should be assessed using finger-to-finger/finger-tonose pointing, and dysdiadochokinesia (an inability to perform coordinated muscular movement smoothly) is assessed with rapid alternating movements. The gait must be evaluated when the patient gives a history suggesting ataxia, although examination may be impossible during an attack of vertigo. Any marked abnormality (e.g., consistent falling or a grossly abnormal gait) should suggest a central lesion, especially in a patient whose vertiginous symptoms have subsided. The main features of a cerebellar gait are a wide base (separation of legs), unsteadiness, irregularity of steps, tremor of the trunk, and lurching from side to side. The unsteadiness is most prominent on arising quickly from a sitting position, turning quickly,
Most routine laboratory testing is not helpful in the evaluation of a vertiginous patient. A finger-stick blood glucose test should be performed in most cases because hypoglycemia can present as vertigo.15 Blood counts and blood chemistries are sometimes helpful when it is difficult to distinguish whether “dizziness” is vertigo or near-syncope. An electrocardiogram should be obtained if there is any possibility of myocardial ischemia. Radiologic Imaging. If cerebellar hemorrhage, cerebellar infarction, or other central lesions are suggested, emergent computed tomography (CT) or magnetic resonance imaging (MRI) of the brain is indicated. MRI, when available, has become the diagnostic modality of choice when cerebellar processes other than acute hemorrhage are possible. MRI is particularly useful for the diagnosis of acoustic neuromas and for sclerotic and demyelinating lesions of the white matter, as seen in multiple sclerosis. Acute vertigo by itself does not usually warrant urgent CT or MRI in all patients, particularly patients in whom a clear picture of peripheral vertigo emerges. But as mentioned earlier, many studies strongly support the use of imaging in patients of advanced age or at risk for cerebrovascular disease.6,11,12,16 Conventional angiography or magnetic resonance angio graphy can be used in cases of suspected VBI to document the presence of vascular disease. It is used most often in patients with changing neurologic signs and symptoms, suggesting impending posterior circulation occlusion. Audiology and electronystagmography are helpful in the follow-up evaluation of a vertiginous patient. Audiology can locate the anatomic site of a lesion causing vertigo. Electronystagmography is a collection of examinations that, when abnormal, suggest vestibular dysfunction but do not yield the specific diagnosis.
■ DIFFERENTIAL DIAGNOSIS The differential diagnosis for other peripheral, central, and systemic causes of vertigo is large (see Box 12-1). More detailed information is given on selected causes in Table 12-3, including the most common peripheral causes of true vertigo: benign positional vertigo, labyrinthitis, Ménière’s disease, and vestibular neuronitis.
Diagnostic Algorithm Most cases of vertigo are of peripheral origin and are not usually life-threatening. The diagnostic approach must focus on identifying entities that either immediately or in the near future can lead to death or significant morbidity (Fig. 12-2).
■ MANAGEMENT Management is based on an accurate diagnosis that distinguishes the serious central causes of vertigo from the less serious, albeit more debilitating, peripheral causes (Fig. 12-3). Any suggestion of cerebellar hemorrhage should warrant immediate imaging with CT or MRI and neurosurgery consultations. VBI should be considered in any patient of advanced age or at high risk of cerebrovascular disease with isolated, new-onset vertigo without an obvious cause.6,11,17,18 Because of the possibility of progression of new-onset VBI in the first 24 to 72 hours, hospital or observation unit admission and consid-
97 Dizziness
Spinning or sensation of motion
Ménière’s Tinnitus Hearing loss Attacks in clusters Long symptom-free intervals
Anemia Infection Depression
Vertigo
Central Attacks: gradual, mild, usually continuous for weeks or months but can be sudden, severe and seconds or minutes with vascular causes Nystagmus: horizontal, rotary, or vertical Little change with head position Neurologic findings usually present No auditory findings
Peripheral Attacks: sudden, severe, usually seconds or minutes Nystagmus: horizontorotary, worsened by head position No neurologic findings Auditory findings may be present
BPPV Short-lived, positional episodes probably caused by stray otoconial particles
Malaise
Vestibular neuronitis Severe vertigo for days Mild persistent positional vertigo No auditory symptoms
Acoustic neuroma Peripheral cause that can become central Vertigo, hearing loss, tinnitus
Cerebellar hemorrhage Severe vertigo, headache, vomiting, ataxia Hypoglycemia Head/neck trauma
Labyrinthitis
Multiple sclerosis Acute suppurative Signs of toxicity Toxic patient Severe vertigo Hearing loss
Serous No signs of toxicity Milder symptoms Inflammatory response to nearby infections
Toxic Hearing loss Tinnitus Medication exposure
Chronic Chronic symptoms Secondary to fistula
Vertebrobasilar migraine
Vertebrobasilar insufficiency Usually associated neurologic abnormalities More likely in the elderly and those with history of cardiac or cerebrovascular disease
Figure 12-2. Diagnostic algorithm for dizziness and vertigo. BPPV, benign paroxysmal positional vertigo.
eration of early magnetic resonance angiography probably are warranted, even in a stable patient. Changing or rapidly progressive symptoms should raise awareness of impending posterior circulation occlusion. If CT or MRI excludes hemorrhage as the source of the patient’s symptoms, an immediate neurologic consultation, emergency angiography, and possibly anticoagulation are indicated. Acute bacterial labyrinthitis requires admission, intravenous antibiotics, and, occasionally, surgical drainage and débridement. In cases of toxic labyrinthitis, the offending medication should be discontinued immediately. Some cases of Ménière’s disease have been treated successfully by vasodilation and diuretic therapy. Diets low in sodium and caffeine and cessation of smoking also have been helpful. Chemical ablation of vestibular function with gentamicin and streptomycin is an option in severe Ménière’s disease. In one small controlled study, corticosteroids were more effective than placebo in treating the acute symptoms of vestibular neuritis. The treatment of acute attacks of vertigo caused by peripheral disorders is symptomatic. Intravenous diazepam in 2- to
5-mg doses is extremely effective in stopping vertigo. It has a sedative effect that acts on the limbic system, the thalamus, and the hypothalamus. Outpatient treatment with diazepam can be continued at doses of 5 to 10 mg three times daily. The neurons involved in vestibular reactions are mediated by acetylcholine. Anticholinergic drugs or antihistamines with anticholinergic activity are extremely useful in treating vertigo. Meclizine hydrochloride (Antivert) is usually prescribed as 25 mg every 8 hours, but has a wide therapeutic margin and can be taken much more frequently to control symptoms. Diphenhydramine hydrochloride (Benadryl), 25 to 50 mg every 6 to 8 hours, and dimenhydrinate (Dramamine, Gravol) are also effective, but are more sedating than meclizine. Either drug also can be given intravenously. Transdermal scopolamine has shown disappointing results for treatment of peripheral vertigo but may be considered a third-line or fourth-line option. Promethazine hydrochloride (Phenergan), 25 mg orally or rectally every 6 to 8 hours, is effective because of its strong antiemetic and mild anticholinergic properties; it also can be used intravenously in doses of 12.5 to 25 mg. Buccal
Chapter 12 / Dizziness and Vertigo
Dysrhythmias Myocardial infarction Hypovolemia Vasovagal Sepsis Panic disorder Drug side effect
Near-syncope/ light-headedness
98
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
Table 12-3 Differential Diagnosis of Patients with True Vertigo CAUSE
HISTORY
ASSOCIATED SYMPTOMS
PHYSICAL
Peripheral 1. Benign paroxysmal positional vertigo
Short-lived, positional, fatigable episodes
Nausea, vomiting
Single position can precipitate vertigo. Horizontorotary nystagmus often can be induced at bedside.
Mild to severe positional symptoms. Usually coexisting or antecedent infection of ear, nose, throat, or meninges Coexisting acute exudative infection of the inner ear. Severe symptoms Gradually progressive symptoms: Patients on medication causing toxicity Recurrent episodes of severe rotational vertigo usually lasting hours. Onset usually abrupt. Attacks may occur in clusters. Long symptom-free remissions Sudden onset of severe vertigo, increasing in intensity for hours, then gradually subsiding over several days. Mild positional vertigo often lasts weeks to months. Sometimes history of infection or toxic exposure that precedes initial attack. Highest incidence is found in third and fifth decades Gradual onset and increase in symptoms. Neurologic signs in later stages. Most occur in women between 30 and 60
Mild to severe hearing loss can occur
Usually nontoxic patient with minimal fever elevation
Usually severe hearing loss, nausea, vomiting
Febrile patient showing signs of toxicity. Acute otitis media Hearing loss. Ataxia common feature in chronic phase
2. Labyrinthitis A. Serous
B. Acute suppurative C. Toxic 3. Ménière’s disease
4. Vestibular neuronitis
Hearing loss that may become rapid and severe, nausea and vomiting Nausea, vomiting, tinnitus, hearing loss
Positional nystagmus not present
Nausea, vomiting. Auditory symptoms do not occur
Spontaneous nystagmus toward the involved ear may be present.
Hearing loss, tinnitus. True ataxia and neurologic signs as tumor enlarges
Unilateral decreased hearing. True truncal ataxia and other neurologic signs when tumor enlarges. May have diminution or absence of corneal reflex. Eighth cranial nerve deficit may be present.
Should be considered in any patient of advanced age with isolated new-onset vertigo without an obvious cause. More likely with history of atherosclerosis. Initial episode usually seconds to minutes Sudden onset of severe symptoms
Often headache. Usually neurologic symptoms including dysarthria, ataxia, weakness, numbness, double vision. Tinnitus and deafness uncommon
Neurologic deficits usually present, but initially neurologic examination can be normal.
Headache, vomiting, ataxia
C. Occlusion of posterior inferior cerebellar artery (Wallenberg’s syndrome)
Vertigo associated with significant neurologic complaints
Nausea, vomiting, loss of pain and temperature sensation, ataxia, hoarseness
Signs of toxicity. Dysmetria, true ataxia. Ipsilateral sixth cranial nerve palsy may be present. Loss of pain and temperature sensation on the side of the face ipsilateral to the lesion and on the opposite side of the body, paralysis of the palate, pharynx, and larynx. Horner’s syndrome (ipsilateral ptosis, miosis, and decreased facial sweating)
D. Subclavian steal syndrome
Classic picture is syncopal attacks during exercise, but most cases present with more subtle symptoms.
Arm fatigue, cramps, mild light-headedness may be only other symptoms than vertigo
5. Acoustic neuroma
Central 1. Vascular disorders A. Vertebrobasilar insufficiency
B. Cerebellar hemorrhage
Diminished or absent radial pulses in affected side or systolic blood pressure differentials between the two areas occur in most patients.
99
Table 12-3 Differential Diagnosis of Patients with True Vertigo—cont’d HISTORY
ASSOCIATED SYMPTOMS
PHYSICAL
2. Head trauma
Symptoms begin with or shortly after head trauma. Positional symptoms most common type after trauma. Self-limited symptoms that can persist weeks to months Usual onset 7–10 days after whiplash injury. Symptoms may last weeks to months. Episodes seconds to minutes when turning head Vertigo almost always followed by headache. Patient has usually had similar episodes in past. Most patients have a family history of migraine. Syndrome usually begins in adolescence Vertigo presenting symptoms in 7– 10% and appears in the course of the disease in a third. Onset may be severe and suggest labyrinth disease. Disease onset usually between ages 20 and 40. Often history of other attacks with varying neurologic signs or symptoms Can be initial or prominent symptom in some patients with the disorder Should be considered in diabetics and any other patient with unexplained symptoms
Usually mild nausea
Occasionally, basilar skull fracture
Neck pain
Neck tenderness, pain on movement, and positional nystagmus and vertigo when head is turned to side of the whiplash No residual neurologic or otologic signs are present after attack.
3. Neck trauma
4. Vertebrobasilar migraine
5. Multiple sclerosis
6. Temporal lobe epilepsy 7. Hypoglycemia
Dysarthria, ataxia, visual disturbances, or paresthesias usually precede headache
Nausea and vomiting, which may be severe
May have horizontal, rotary, or vertical nystagmus. Nystagmus may persist after the vertiginous symptoms have subsided. Bilateral internuclear ophthalmoplegia and ataxic eye movements suggest multiple sclerosis.
Memory impairment, hallucinations, trancelike states, seizures Sweating, anxiety
May have aphasia or convulsions Tachycardia, mental status change may be present.
Vertigo ABCs
Finger-stick blood glucose
ECG
Ischemic changes
Hypoglycemia
Possible AMI algorithm
Central causes
Figure 12-3. Management algorithm for vertigo. AMI, acute myocardial infarction; BPPV, benign paroxysmal positional vertigo; CT, computed tomography; ECG, electrocardiogram; ENT, ear, nose, and throat; MRI, magnetic resonance imaging.
VBI CT or MRI Neurology consult Admit Head/neck trauma Trauma evaluation Symptomatic treatment
Vertebrobasilar migraine Migraine treatment Cerebellar hemorrhage Stat CT or MRI Stat neurosurgery consult Unknown or other diagnosis Neurology consult
Dextrose
Peripheral causes
Toxic labyrinthitis Stop offending medication
BPPV Drug treatment Consider repositioning maneuvers
Acute suppurative labyrinthitis IV antibiotics ENT consult Admission
Other causes Drug treatment: Diazepam Meclizine Benadryl Promethazine
Unable to walk: admission Significant improvement: discharge Primary care or ENT follow-up for all ENT follow-up for suspected acoustic neuroma or perilymphatic fistula
Chapter 12 / Dizziness and Vertigo
CAUSE
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
100
A B E
have been shown to be extremely effective in treating benign paroxysmal positional vertigo.19-21 The Epley maneuver involves sequential movement of the head into four positions, staying in each position for approximately 30 seconds, as demonstrated in Figure 12-4. One study has even shown these maneuvers have resulted in long-term efficacy in successful symptom treatment.22 One of the most useful tools the physician has is patient reassurance. Most patients with vertigo have self-limited disease processes that have a specific organic cause. By combining patient education and reassurance with judicious use of medications, the treatment of a dizzy patient can be rewarding for the patient and the physician.
■ DISPOSITION
D
C (c) 2001 Northwestern University
Figure 12-4. The Epley maneuver for benign paroxysmal peripheral
vertigo, also known as the particle repositioning or canalith repositioning procedure. Image used with permission of 〈http://www.dizziness-andbalance.com/disorders/bppv/bppv.html〉 Timothy C. Hain, Professor of Neurology, Feinberg School of Medicine, Northwestern University, http://www.dizziness-and-balance.com/disorders/bppv/bppv.html.
prochlorperazine (Compazine, Stemetil) also has been shown to be a safe and effective treatment for vertigo. Avoidance of stimulants (e.g., caffeine, pseudoephedrine, nicotine) may ease symptoms in some cases. In addition, canalith repositioning procedures, such as the Epley and Semont maneuvers,
Documented or suggested cerebellar hemorrhage or infarction, VBI, and acute bacterial labyrinthitis require workup and hospitalization. In patients older than age 55 years, particularly patients with vascular disease, admission for observation and imaging of cerebral vasculature is often warranted. Most younger patients with peripheral causes of vertigo can be discharged from the emergency department after symptoms are controlled. Some patients may have such severe symptoms (e.g., vomiting, inability to walk) despite a trial of medication that they require admission for intravenous hydration and observation. All discharged patients should receive primary care; neurology; or a follow-up consult with an ear, nose, and throat specialist. This follow-up is especially important for possible cases of acoustic neuroma and toxic labyrinthitis.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 13
Confusion
J. Stephen Huff
■ PERSPECTIVE The term confusion connotes an alteration in higher cerebral functions, such as memory, attention, or awareness. Confusion is a symptom, not a diagnosis. Clinical jargon includes “altered mental status,” “delta MS” (change in mental status), “altered mentation,” and “change from baseline.” Additionally, the ability to sustain and focus attention is impaired. Symptoms of confusion may fluctuate, as may the level of consciousness. Implicit in the definition is a recent change in behavior. Chronic mental status changes such as dementia typically have a different clinical chronology. Other forms of altered mentation include states of diminished alertness on the coma spectrum; these presentations may result from some of the same pathophysiologic processes causing confusion and are discussed in Chapter 15. Confusion may range in severity from a mild disturbance of short-term memory to a global inability to relate to the environment and process sensory input. This extreme state is termed delirium. Delirium has two subtypes: hyperactive and hypoactive.1 Hyperactive delirium is characterized as an acute confusional state associated with increased alertness, increased psychomotor activity, and disorientation and is often accompanied by hallucinations. In hypoactive delirium (sometimes referred to as quiet delirium), the confusional state is present but the patient has a reduction in alertness and behavior. Confusion has many causes, and an orderly approach is necessary to discover the causative diagnosis.
Epidemiology Physicians underestimate the incidence of confusion in patients.2,3 Often, confusion is accepted as an incidental or secondary component of another condition. A patient with injuries from a motor vehicle crash or with dyspnea may be confused, but the primary condition overshadows the underlying abnormal mental status. When confusion exists as an isolated or unexplained finding, it is more likely to receive full and immediate consideration by the clinician. Confusion is estimated to occur in 2% of emergency department (ED) patients, 10% of all hospitalized patients, and 50% of elderly hospitalized patients.2,4
Pathophysiology Conceptually, consciousness may be divided into elements of alertness or arousal and elements constituting content of consciousness. Confusion is largely a problem of the content
portion of consciousness. Many different clinical processes may disrupt optimal cortical functioning and result in confusion. The pathophysiology is not straightforward. Widespread cortical dysfunction is thought to result from substrate deficit (hypoglycemia or hypoxemia), neurotransmitter dysfunction, or circulatory dysfunction. Compounding this problem is the idea that the reserve of central nervous system (CNS) function varies from individual to individual; individuals with a pre existing impairment may become confused after even minor changes in their normal state.
■ DIAGNOSTIC APPROACH Differential Considerations The observation of acute confusion prompts a search for an underlying cause. Four groups of disorders encompass most causes of diffuse cortical dysfunction: (1) systemic diseases secondarily affecting the CNS, (2) primary intracranial disease, (3) exogenous toxins, and (4) drug withdrawal states (Box 131).1 Focal cortical dysfunction, such as from tumor or stroke, typically does not cause confusion, although exceptions are encountered. Likewise, subcortical or brainstem dysfunction most frequently results in a diminished level of alertness and consciousness, not confusion.
Rapid Assessment and Stabilization Most patients with acute confusion do not require immediate interventions. Three crucial exceptions are hypoglycemia, hypoxemia, and shock. A complete set of vital signs, including temperature and oxyhemoglobin saturation, and a bedside blood glucose level should be determined promptly for all confused patients. Oral or intravenous glucose therapy is indicated if low blood glucose is discovered. Supplemental oxygen and intravenous fluid are administered as necessary. Patients should be protected from harming themselves or others. Close observation may need to be supplemented by medications or physical restraint. Family members may offer valuable assistance in observing and comforting the patient. In a patient with abnormal or unstable vital signs, initial diagnostic and management efforts are directed toward treatment of the systemic condition. A confused patient with acute pulmonary edema, hypoxia, and confusion obviously requires evaluation and treatment of the pulmonary edema, not a screening test for cognitive functioning. 101
102
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
BOX 13-1 Major Categories: Differential Considerations Primary intracranial disease Systemic diseases secondarily affecting the central nervous system Exogenous toxins Drug withdrawal
Findings That May Help Differentiate between
BOX 13-2 Organic and Functional Causes of Confusion Organic History Acute onset Any age Mental status examination Fluctuating level of consciousness Disoriented Attention disturbances Poor recent memory Hallucinations: visual, tactile, auditory Cognitive changes Physical examination Abnormal vital signs Nystagmus Focal neurologic signs Signs of trauma
Functional (psychiatric) History Onset over weeks to months Onset ages 12 to 40 years Mental status examination Alert
Oriented Agitated, anxious Poor immediate memory Hallucinations: most commonly auditory Delusions, illusions Physical examination Normal vital signs No nystagmus Purposeful movement No signs of trauma
Generally, in patients with schizophrenia and other psychiatric disorders, tests of cognition, orientation, and attention are normal unless the condition is severe. The term psychosis implies a disorder of reality testing and thought organization severe enough to interfere with normal daily functioning. Psychosis is a nonspecific syndrome, and careful evaluation is required to differentiate between psychiatric and organic origins (e.g., drug intoxication or other systemic process) (Box 13-2).
Pivotal Findings A patient with an altered state of consciousness including confusion is evaluated by taking a focused history and conducting a pertinent examination, performing rapid bedside screening investigations, and observing the response to certain therapies (e.g., dextrose or naloxone). Additional evaluation may include laboratory testing and diagnostic imaging with various modalities. Useful information that provides the diagnosis or strongly suggests the etiology is found roughly in descending order from the patient’s history, the examination including results of rapid bedside testing, and the response to ED therapies; the results of laboratory testing and diagnostic imaging are less often useful.5
History Confusion is often reported by family members or caregivers; frequently the patient is not aware of the confusion and seem-
ORIENTATION TO TIME “What is the date?” REGISTRATION “Listen carefully. I am going to say three words. You say them back after I stop. Ready? Here they are . . . HOUSE (pause), CAR (pause), LAKE (pause). Now repeat those words back to me.” [Repeat up to 5 times, but score only the first trial.] NAMING “What is this?” [Point to a pencil or pen.] READING “Please read this and do what it says.” [Show examinee the words on the stimulus form.] CLOSE YOUR EYES
Figure 13-1. Mini-mental state examination sample items. (Reproduced by special permission of the Publisher, Psychological Assessment Resources, Inc., 16204 North Florida Avenue, Lutz, Florida 33549, from the Mini Mental State Examination, by Marshal Folstein and Susan Folstein. Copyright © 1975, 1998, 2001 by Mini Mental LLC, Inc. Published 2001 by Psychological Assessment Resources, Inc. Further reproduction is prohibited without permission of PAR, Inc. The MMSE can be purchased from PAR, Inc., by calling (800) 331-8378 or (813) 968-3003.)
ingly glosses over problems. Families may articulate the complaint as confusion but also may describe rambling, disorientation, speaking to persons not there, the patient’s inability to find his or her way around familiar surroundings, or simply “not being right.” An essential goal of the history is to determine when the patient last exhibited “normal” thinking and behavior. Attention deficit is the common denominator in confusional states. The initial task in evaluating the patient is to define the symptoms and severity of confusion. The specific behaviors that are of concern to the patient or caregivers should be defined. Often, the family is the most valuable source for information; a physician or other caregiver with an established relationship with the patient also may be helpful. The duration of the confusion, any recent changes in medications, and recent illnesses are important points in the clinical history. Hallucinations are not unique to psychiatric illness and can commonly occur in confusion states, especially delirium. Hallucinations in delirium tend to be visual (with or without auditory components), powerful, fleeting, and poorly organized. A history of medication or substance abuse and any recent changes, especially cessation of benzodiazepines or ethanol, should be sought.
Physical Examination The patient’s confusion may be obvious at the bedside. In other cases, confusion may be subtle, and informal assessment of mental status and cognitive abilities may fail to detect it. The mini-mental state examination (MMSE) (Fig. 13-1) commonly is recommended as a screening instrument but is used infrequently in the ED because of the time required to administer it.6,7 A more rapidly performed screening tool, the Quick Confusion Scale (QCS; Fig. 13-2), has been developed and tested in ED patients.8-10 This tool objectively measures elements of the patient’s mental status in 2 to 3 minutes and correlates well with the MMSE.9,11 The tasks measured by either the MMSE or the QCS require adequate attention on
103 SCORE (highest number in category indicates correct response; decreased scoring indicates increased number of errors)
What year is it now?
0 or 1 (score 1 if correct; 0 if incorrect)
x2
What month is it?
0 or 1 (score 1 if correct; 0 if incorrect)
x2
About what time is it? (answer correct if within the hour)
0 or 1 (score 1 if correct; 0 if incorrect)
x2
Count backwards from 20 to 1
0, 1, or 2 (score 2 if correct; 1 if 1 error; score 0 if more than 2 errors)
x1
Say the months in reverse
0, 1, or 2 (score 2 if correct; 1 if 1 error; score 0 if more than 2 errors)
x1
Repeat the memory phrase (each underlined portion is worth 1 point)
0, 1, 2, 3, 4, 5 (score 5 if correctly performed; each error drops score by one)
x1
WEIGHT
SCORE
Repeat phrase and remember it: “John Brown, 42 Market Street, New York”
Figure 13-2. Quick Confusion Scale.
TOTAL Final score is sum of the totals; score less than 15 suggests the presence of altered cognition and need for further assessment.
the part of the patient. If the patient’s attention span is greatly impaired, detailed testing may be impossible. Digit repetition forward (five or six digits) and backward (four digits) is a brief screen for attention function. Alternatively, spelling a commonly used word backward (“world” is frequently used) measures a patient’s ability to concentrate. Screening tests may detect confusion not obvious in casual conversation, identifying the need for further investigations.12,13 The physical examination may suggest a cause for confusion such as congestive heart failure or pneumonia. A fever suggests an infection as the cause of altered mental status and should prompt a search for the source, particularly urinary tract infection in the elder patient. Any new focal neurologic findings suggest a possible mass lesion or stroke and should trigger neuroimaging. In this regard, testing of gait and tandem gait, if possible, may be invaluable. Aphasia, fluent or nonfluent, is a focal sign suggesting a lesion in the dominant cerebral hemisphere. In confusional states, speech may be abnormal and is often incoherent, and the rate of speech may be either rapid or slowed. Involuntary movements, such as asterixis or tremor, may be present. The various toxidromes may assist in the identification of an intoxication or drug effect as the cause of confusion.
Laboratory Tests The results of the history and physical examination frequently guide the clinician in the choice of laboratory tests most likely to yield valuable diagnostic information. Pulse oximetry may reveal hypoxia, or bedside glucose testing may reveal hypoglycemia or hyperglycemia. In the presence of a fever, chest radiography and urinalysis often reveal the source of the infection causing the altered mentation. In elder patients, urinalysis should be performed whether or not fever is present. Other
tests commonly available in the ED and useful in the evaluation of a confused patient are serum electrolyte testing (especially sodium) and electrocardiography. Electrocardiography is indicated in elderly patients because myocardial infarction may present as confusion. The complete blood count, although commonly performed, is unlikely to provide useful diagnostic clues. Arterial blood gas testing is rarely indicated or useful, unless pulse oximetry is not reliable. If common and simple tests do not suggest a solution, more complex testing should be initiated in the ED, observation unit, or inpatient service. The clinical situation and overall condition of the patient determine the speed and direction of evaluation. Additional laboratory work is often of decreasing yield but may reveal the cause of confusion. Serum ammonia, calcium, thyroid function, and selected drug and toxicologic testing may be ordered in this second tier of evaluation. Blood and urine cultures should be obtained in the febrile patient when hospital admission is anticipated and a clear infectious source is not evident. Paracentesis or thoracentesis may be appropriate if ascites or pleural effusion is present. Cranial computed tomography (CT) scanning is usually done to screen for CNS lesions in the absence of another identified source for the confusion. Focal findings on CT increase the yield of this test, but unanticipated abnormalities are often found on neuroimaging. Lumbar puncture may discover or exclude CNS infection if no other source has been identified. Cerebrospinal fluid examination may clarify a diagnosis of bacterial meningitis, encephalitis, aseptic meningitis, or subarachnoid hemorrhage. If the cause of confusion remains unclear or if the patient is unable to function safely in his or her current environment, admission may be necessary for additional ongoing assessment, including diagnostic testing not usually available in the ED, such as magnetic resonance imaging or electroencephalography.5
Chapter 13 / Confusion
ITEM
104
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
BOX 13-3 Critical and Emergent Diagnoses Critical Hypoxia/diffuse cerebral ischemia Respiratory failure Congestive heart failure Myocardial infarction Shock Systemic processes Hypoglycemia CNS infections Hypertensive encephalopathy Elevated intracranial pressure—medical and surgical origin Emergent Hypoxia/diffuse cerebral ischemia Severe anemia Systemic diseases Electrolyte and fluid disturbance Endocrine disease Thyroid Adrenal Hepatic failure Nutrition/Wernicke’s encephalopathy Sepsis, infection Intoxications and withdrawal CNS sedatives Ethanol Other medication side effects, particularly anticholinergics CNS disease Trauma Infections Stroke Subarachnoid hemorrhage Epilepsy/seizures Postictal state Nonconvulsive status epilepticus Complex partial status epilepticus Neoplasm Note: These represent a partial diagnosis; causes are myriad. “Critical” in this case means conditions that need immediate assessment and correction within moments, such as oxygenation and ventilation problems or hypoglycemia. Because confusion represents CNS failure, other problems may be considered critical as well and may require intensive care unit admission, depending on severity. CNS, central nervous system.
■ DIFFERENTIAL DIAGNOSIS Certain critical and emergent diagnoses require prompt recognition to prevent morbidity or mortality (Box 13-3). The diagnosis of confusion implies the exclusion of other states of altered mental status, such as coma and decompensated psychiatric syndromes. A new focal neurologic deficit points to a focal defect of the CNS, which is less likely to cause the global cortical dysfunction necessary for confusion. Stroke rarely causes confusion, but resulting disturbances in speech or understanding may mimic a confusional state. The diagnosis of stroke is relatively straightforward if a new motor deficit is present. Occasionally, other focal neurologic abnormalities may mimic a confusional state. A person with a new visual field deficit and visual neglect may have difficulty ambulating in familiar surroundings and be labeled as confused, but this
Altered mental status
Diminished level of consciousness?
Yes
Coma, stupor
No Acute focal neurologic deficit? Hemiparesis, aphasia, visual field cut?
Yes
Stroke, mass
No Abnormal attention span, mental status testing
Yes
Confusion, delirium ( agitation)
No Thought disorder Possible psychiatric disorder
Figure 13-3. Diagnostic algorithm for confusion.
reflects focal neurologic injury and not a confusional state from global CNS dysfunction. Careful assessment of mental status assists in resolving the diagnostic dilemma. Frontal lobe dysfunction from stroke, subdural hematoma, or tumor may result in personality changes and the report of “confusion” by family or friends. Altered mental status may be divided into three different categories depending on the findings of diminished level of consciousness, acute focal neurologic deficit, or abnormal attention span. Placement into one of these categories may guide the differential assessment and therapy (Fig. 13-3).
■ EMPIRICAL MANAGEMENT Ideally, treatment is directed at the underlying cause of the confusion. Investigations continue until a likely diagnosis is discovered or consultation and admission are deemed necessary (Fig. 13-4). Many febrile patients are found to have a systemic infectious cause of the confusion. Urinary tract infections and pneumonia are the more common sources, but soft tissue infections also warrant consideration. CNS infections are encountered less frequently but have potentially devastating consequences if not recognized promptly. Antibiotic treatment for coverage of common causes of meningitis may be considered in ill febrile patients while definitive evaluation is in progress. Postictal confusion is common in patients with seizures but should improve within 20 to 30 minutes. If the patient remains unconscious or confused after a seizure, the possibility of ongoing or intermittent seizure activity (i.e., nonconvulsive seizures) should be considered. Nonconvulsive status epilepticus, an epileptic twilight state, is unusual but does occur, and may be particularly difficult to recognize in the elderly14 (see also Chapter 15). Sometimes it may be necessary to treat confusion or agitation for patient safety. Environmental manipulations, such as dim lighting or psychosocial support, may be helpful. Confinement or physical restraint may be necessary at times for patient safety; institutional guidelines should be followed. Benzodiazepines or butyrophenones may be used if necessary to decrease agitation. These medications may alter mental status further, making evaluation more difficult.
105
■ DISPOSITION
Confusion
Correct as needed
Normal Fever?
Yes
Search for infection source
Yes
Pursue likely cause
No History, physical examination Suggest likely cause No Diagnosis uncertain
ED basic testing (CBC, electrolytes, UA, CXR, ECG)
ED advanced testing or consultation (may include CT, ABG, CSF, thyroid, EEG, Ca2+, toxicology analysis. Other specific tests as necessary, e.g., cultures, MRI, ECG)
Admission, advanced diagnostic testing
Figure 13-4. Management algorithm for confusion. ABG, arterial
blood gas; CBC, complete blood count; CT, computed tomography; CSF, cerebrospinal fluid; CXR, chest x-ray; ECG, electrocardiogram; MRI, magnetic resonance imaging; UA, urinalysis.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 13 / Confusion
Abnormal Substrate verification Glucose, oxygenation checks
Most patients presenting with confusion are admitted to the hospital or ED observation unit for additional diagnostic procedures, extended observation, and treatment. Exceptions include patients with rapidly resolved confusional states after treatment for insulin-induced hypoglycemia, after generalized seizures of known origin, or after recovering from self-limiting intoxicants or withdrawal states, such as those related to ethanol or recreational drugs. These patients may be observed and then discharged after successful identification and resolution of acute confusional state. Unresolved confusion or unexplained findings on repeat mental status screen should prompt admission or careful reevaluation before considering discharge.
Chapter 14
Depressed Consciousness and Coma
Jeremy L. Cooke
■ PERSPECTIVE Epidemiology Depressed consciousness is a common presenting complaint in the emergency department (ED). It represents a spectrum of disease that ranges from sleepiness or decreased alertness to frank coma. The majority of cases of depressed consciousness and coma are caused by metabolic or systemic derangements, and the remainder are caused by structural lesions.1 The differential diagnosis for depressed level of consciousness often overlaps that for confusion (see Chapter 13).
Pathophysiology Consciousness includes the properties of arousal, which is defined as the awareness of one’s self or surroundings, and cognition. Cognition is the combination of orientation, the accurate perception of what is experienced, judgment, the ability to process input data to generate more meaningful information, and memory, the ability to store and retrieve information. The ascending reticular activating system (ARAS) is the neuroanatomic structure primarily responsible for arousal. It is located in the paramedian tegmental zone in the dorsal part of the brainstem (Fig. 14-1). The input of somatic and sensory stimuli to the cerebral cortex is controlled by the ARAS and functions to initiate arousal from sleep. The brain’s cognition centers are located primarily in the cerebral cortex. Insults to the cerebral cortex or brainstem can each independently cause depressed consciousness or coma. These structures are vulnerable to metabolic derangements, toxins, or mechanical injury. Typically, both cerebral hemispheres need to be affected to induce coma and this also depends on the size and speed of progression of the insult. Localized, unilateral lesions in the cerebral cortex usually do not induce depressed consciousness or coma even if other cognitive functions are impaired. In contrast, a completely intact brainstem is necessary for arousal. Small focal lesions in the brainstem can affect the ARAS. If the ARAS is impaired, the cerebral cortex cannot be aroused and depressed consciousness or coma occurs. Potential causes of depressed consciousness can be broken down into a few general categories. Metabolic or systemic causes of coma can include hypoxia, hypoperfusion, infection, toxic drug effects, or electrolyte disturbances. Hypoxia can be the result of congestive heart failure (CHF), pulmonary embolism, carbon monoxide poisoning, or severe pulmonary com106
promise such as occurs in chronic obstructive pulmonary disease (COPD), cystic fibrosis, and asthma. The various causes of shock can result in hypoperfusion states leading to depression of consciousness. These include anaphylactic, septic, hypovolemic, cardiogenic, and neurogenic origins of shock. Each type of shock has its own special characteristics, which are detailed in other chapters. Infection, both systemic (sepsis) and focal, can be another general cause of depressed consciousness. This is particularly true if central nervous system (CNS) structures are involved as in meningitis, encephalitis, or CNS abscess. Toxic drug effects ranging from recreational drug use and intentional overdoses to therapeutic doses with adverse side effects are common general causes of depressed consciousness seen in the ED. In the elderly, adverse side effects from prescription medications are common. In addition, electrolyte and glucose abnormalities can be caused by multiple conditions, including diabetes, renal dysfunction, malignancy, and medication interactions or dosing errors. Structural causes of coma and depressed consciousness are those most commonly arising from head trauma, stroke, tumor, or infection. Traumatic causes can include subdural and epidural hematomas, intraparenchymal or subarachnoid hemorrhage, or simply contusion or concussion. Strokes occur with embolic, thrombotic, or hemorrhagic mechanisms, but it is extremely unusual for ischemic (i.e., nonhemorrhagic) stroke to depress consciousness unless a massive insult to both hemispheres has occurred (e.g., diffuse, severe cerebral edema after a massive infarct). Depression of consciousness with CNS infections may be caused by mass effect and is common with severe bacterial meningitis, cerebral abscess or empyema, or parasitic mass. Malignancies, whether primary or metastatic, can cause depressed consciousness if their mass effect is sufficient or if surrounding edema develops rapidly. Special consideration should be given to specific populations of patients. The elderly, in particular, are susceptible to alterations in therapeutic medication dosage and drug-drug interactions. Even seemingly minor infections, such as urinary tract infections, upper respiratory infections, or viral gastroenteritis, may cause altered mental status (see Chapter 13), depressed consciousness, or coma. In addition, immunocompromised patients such as those with AIDS or those undergoing chemotherapy treatment for transplants, malignancy, or immunologic disease are vulnerable to a multitude of opportunistic infections not commonly seen in the general patient population.
107 Medial geniculate body Inferior brachium Pineal body Optic commissure
Superior colliculi Inferior colliculi
Figure 14-1. Brainstem anatomy. (Adapted
from Adams J: Emergency Medicine. Fig. 89-1. Philadelphia, Elsevier. Copyright © 2008.)
Peduncle of cerebrum
Frenulum veli Trochlear nerve Lateral lemniscus Superior peduncle
Oculomotor nerve
Pons Trigeminal nerves
Middle peduncle Rhomboid fossa
Acoustic nerve Facial nerve Abducent nerve
Clava
Hypoglossal nerve
Glossopharyngeal and vagus nerves
Clinical Evaluation The clinical evaluation and stabilization of patients with depressed consciousness occur simultaneously with the diagnosis in the ED. The differential diagnosis of depression of consciousness is extensive but can be greatly simplified by focusing attention on the distinguishing characteristics of the available patient history and physical examination (Boxes 14-1 and 14-2).2 Approaching the patient’s presentation systematically, beginning with a broad differential diagnosis, usually allows development of a short list of likely diagnoses early in the encounter.
History Chief complaints relating to depressed consciousness vary widely. Family members may report the patient as being more difficult to arouse from sleep or less interactive. Often, family members or friends have alerted emergency medical services after the patient is “found down” and unarousable even with vigorous stimulation. Family members, caregivers, or friends often can provide information that is unobtainable or unreliable from the patient who presents with depressed consciousness. They usually have some knowledge regarding the patient’s past medical history, which may include diabetes, liver or renal disease, vascular disease such as hypertension, stroke or transient ische mic attacks, malignancy, seizures, immunocompromised states such as HIV, sickle cell disease or a history of organ transplant, or psychiatric illness. Symptoms in the hours to days preceding the occurrence of depressed consciousness are important. Specifically, the patient may have complained of headache, focal weakness or numbness, incoordination, or vision disturbances. The patient may have experienced nausea, vomiting, or fever. There may be a history of a traumatic fall or exposure to drugs or toxins. Family members may also be
BOX 14-1
Structural Etiology of Altered Mental Status and Coma
Trauma Subdural hematoma Epidural hematoma Cerebral concussion/contusion Stroke Syndromes Embolism Cardiac (atrial fibrillation, endocarditis) Paradoxical (fat embolus) Thrombosis Cerebral venous sinus thrombosis Hemorrhage Subarachnoid hemorrhage (SAH) Pontine hemorrhage Cerebellar hemorrhage Intracerebral hemorrhage Tumor Brainstem tumors Metastatic disease Angiomas Pituitary apoplexy Acute hydrocephalus Infection Subdural empyema/abscess
able to relay additional diagnostic clues such as rate of onset or waxing-waning characteristics of the patient’s symptoms. Causes of depression of consciousness vary with patient age (Box 14-3). The elderly are particularly vulnerable to infec-
Chapter 14 / Depressed Consciousness and Coma
Optic tract
Pulvinar
108
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
BOX 14-2 Metabolic/Systemic Etiology of Altered Mental Status and Coma Hypoxia Severe pulmonary disease (hypoventilation) Severe anemia Environmental/toxin Methemoglobinemia Cyanide Carbon monoxide Decreased atmospheric oxygen (high altitude) Near-drowning Disorders of Glucose Hypoglycemia Chronic alcohol abuse and liver disease Excessive use of insulin or other hypoglycemic agents Insulinoma Hyperglycemia Diabetic ketoacidosis Nonketotic hyperosmolar coma Decreased Cerebral Blood Flow Hypovolemic shock Cardiac Vasovagal syncope Arrhythmias Myocardial infarction Valvular disorders Congestive heart failure Pericardial effusion/tamponade Myocarditis Infectious Septic shock Bacterial meningitis Vascular/hematologic Hypertensive encephalopathy Pseudotumor cerebri Hyperviscosity (sickle cell, polycythemia) Hyperventilation Cerebral lupus vasculitis Thrombotic thrombocytopenic purpura Disseminated intravascular coagulation Metabolic Cofactor Deficiency Thiamine (Wernicke-Korsakoff syndrome) Pyridoxine (isoniazid overdose) Folic acid (chonic alcohol abuse) Cyanocobalamine Niacin
Electrolyte/pH Disturbances Acidosis/alkalosis Hypernatremia/hyponatremia* Hypercalcemia/hypocalcemia Hypophosphatemia Hypermagnesemia/hypomagnesemia Endocrine Disorders Myxedema coma, thyrotoxicosis Hypopituitarism Addison’s disease (primary or secondary) Cushing’s disease Pheochromocytoma Hyperparathyroidism/hypoparathyroidism Endogenous Toxins Hyperammonemia (liver failure) Uremia (renal disease) Carbon dioxide narcosis (pulmonary disease) Porphyria Exogenous Toxins Alcohols Ethanol, isopropyl alcohol, methanol, ethylene glycol Acid poisons Salicylates Paraldehyde Ammonium chloride Antidepressant medications Lithium Tricyclic antidepressants (TCAs) Selective serotonin reuptake inhibitors (SSRIs) Monamine oxidase inhibitors (MAOIs) Stimulants Amphetamines/methamphetamines Cocaine Over-the-counter sympathomimetics Narcotics/opiates Morphine Heroin Codeine, oxycodone, meperidine, hydrocodone Methadone Fentanyl Propoxyphene Sedative-hypnotics Benzodiazepines Barbiturates Rohypnol Bromide
Hallucinogens Lysergic acid diethylamide (LSD) Marijuana Mescaline, peyote Mushrooms Phencyclidine (PCP) Herbs/plants Aconite Jimson weed Morning glory Volatile substances Hydrocarbons (gasoline, butane, toluene, benzene, choroform) Nitrites Anesthetic agents (nitrous oxide, ether) Other γ-Hydroxybutyrate (GHB) Ketamine Penicillin Cardiac glycosides Anticonvulsants Steroids Heavy metals Cimetidine Organophosphates Disorders of Temperature Regulation/Environmental Hypothermia Heat stroke Malignant hyperthermia Neuroleptic malignant syndrome High-altitude cerebral edema (HACE) Dysbarism Primary Glial or Neuronal Disorders Adrenoleukodystrophy Creutzfeldt-Jakob disease Progressive multifocal leukoencephalopathy Marchiava-Bignami disease Gliomatosis cerebri Central pontine myelinolysis Other Disorders of Unknown Etiology Seizures Postictal states Reye’s syndrome† Intussuception†
*Can be associated with dilution of formula in infant feeding. † Prominent in the pediatric population.
tious etiologies, medication changes, and alterations in their living environments. Young adults and adolescents are more likely to present after recreational drug use or trauma. Accidental toxic ingestions are often seen in younger children. In infants, infectious causes of depressed consciousness are most common; however, trauma secondary to physical abuse and metabolic derangements from inborn errors of metabolism can be seen.3
Physical Examination The severity of presenting symptoms dictates the speed needed for stabilization and diagnosis. After necessary stabilization measures have been instituted (e.g., intubation of the frankly comatose patient), a systematic examination is conducted. Level of consciousness is determined by the patient’s ability to speak in full, coherent sentences and to respond
109
Common Age-Related Etiology of Altered
BOX 14-4 Glascow Coma Scale Score
Infant Infection Trauma/abuse Metabolic
Eye Opening
Child Toxic ingestion
Verbal Response Adult
Adolescent/Young Adult Toxic ingestion Recreational drug use Trauma Elderly Medication changes Over-the-counter medications Infection Alterations in living environment Stroke
appropriately to the examiner. A rapid, directed neurologic screening examination can determine whether the patient has a significant focal motor deficit. The presence of a distinctive odor on the breath, although uncommon, can cue the examiner to the presence of alcohol, ketones (diabetic/ alcoholic ketoacidosis), or bitter almonds (cyanide toxicity). Undressing the patient promptly and completely permits evaluation for signs of trauma or skin lesions suggesting overwhelming infection. Vital signs are paramount in the initial assessment of all patients. Significant hypotension with depressed consciousness suggests shock, and both causes and therapy should be addressed immediately. Late-stage, severe elevation in intracranial pressure (ICP) can cause bradycardia and hypertension. Tachycardia and hypotension can be the result of primary cardiac, infectious, or toxic/metabolic causes. Both hypothermia and hyperthermia can result in altered mental status whether from infectious, structural, or toxic/metabolic causes. Hyperventilation, Kussmaul’s or Cheyne-Stokes breathing, agonal breathing, apnea, or other alterations in respiratory patterns can suggest primary CNS abnormalities or toxic/ metabolic derangements. Immediately after an assessment of the patient’s vital signs, a head-to-toe physical examination is performed.4 A methodical and complete head and neck examination is conducted, with particular emphasis on examination of the papillary reflexes and eye movements (see later discussion) and any indications of head trauma (hemotympanum, scalp hematoma). The mucous membranes may suggest specific toxidromes. Examination of the neck should focus on evidence of infection, including nuchal rigidity, lymphadenopathy, or fluctuance. The cervical spine should be immobilized if there are signs of neck trauma, such as cervical spine tenderness, or evidence of blunt external trauma. Stridor indicates respiratory distress typically from infection, edema, or foreign body aspiration. Chest examination focuses on pulmonary function, infection, cardiac output, and the presence of injury. Potentially helpful abdominal findings include ascites, hepatosplenomegaly, ecchymosis, or striae. Gross blood, purulent drainage, or retained foreign bodies should be sought on genitourinary and rectal examination. In the absence or presence of signs of
Pediatric
Motor Response
Spontaneous To voice To pain None
4 3 2 1
Oriented Confused Inappropriate words Incomprehensible words None Appropriate Cries, consolable Persistently irritable Restless, agitated None Obeys commands Localizes pain Withdraws to pain Flexion to pain Extension to pain None
5 4 3 2 1 5 4 3 2 1 6 5 4 3 2 1
trauma, lesions on the skin such as rashes, signs of drug use (needle “tracks” or medication patches), or embolic phenomena can be differential clues. A systematic neurologic examination, with particular attention paid to the eyes, is the most useful tool in differentiating a structural from a systemic or metabolic etiology of depressed consciousness or coma. A head-to-toe approach is a proven strategy. This should include evaluation of the patient’s Glasgow Coma Scale (GCS) (Box 14-4), level of alertness, cranial nerves, strength, reflexes, and cerebellar functions with emphasis on gait, pronator drift, finger-to-nose, heel-to-shin, rapid alternating movements, and Romberg testing. A change of two or more points in serial GCS testing represents a significant change in the patient’s level of consciousness. Discovery of a focal neurologic deficit is suggestive of a structural etiology. Particular attention should be paid to a focused eye examination during which a helpful amount of information can be obtained. Unilateral dilatation of a pupil (“blown pupil”) and loss of reactivity in a comatose patient are ominous signs of uncal herniation requiring immediate neurosurgical consultation and intervention. Papilledema in the setting of increased ICP or retinal hemorrhage associated with trauma can be found on funduscopic examination. The eye examination should also include testing of eye movements, which are coordinated by the medial longitudinal fasciculus located in the brainstem and ocular centers located in the cerebral cortex. Cranial nerves III, IV, and VI are responsible for control of the extraocular muscles. Cranial nerve III paralysis results in a persistently abducted eye, whereas a persistently adducted eye is caused by paralysis of cranial nerve VI. In the setting of trauma, a unilateral third cranial nerve palsy suggests an ipsilateral compressive lesion such as seen with epidural hematoma. Cranial nerve VI palsies are often nonlocalizing as the nerve has a long intracranial course and compressive forces from intracranial mass effects (tumor, traumatic hematoma, increased ICP, etc.) may compromise cranial nerve function anywhere in its course. Horizontal disconjugate gaze is an important finding and is commonly seen in patients who are sedated, drowsy, or intoxicated. Disconjugate gaze found
Chapter 14 / Depressed Consciousness and Coma
BOX 14-3 Mental Status
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
110
COWS, which stands for “cold-opposite, warm-same.” If there is no response to the irrigation, brainstem dysfunction is possible. Ice water irrigation right ear
Alert wakefulness (nystagmus with rapid movement opposite side of cold stimulation)
Bilateral cerebral hemisphere dysfunction
Left MLF dysfunction
Brainstem dysfunction
Left oculomotor nerve dysfunction
Figure 14-2. Oculocephalogyric (caloric) responses to various central nervous system pathologic conditions. MLF, medial longitudinal fasciculus.
in the vertical plane is usually more serious and suggests cerebellar or pontine dysfunction. Oculocephalic (doll’s eyes) and oculovestibular reflex testing are useful in looking at the functional integrity of the brainstem. These tests, if negative, make structural lesions in the brainstem very unlikely as the source of the patient’s altered mental status. If there are no contraindications, such as suspected cervical spine injury, oculocephalic testing is accomplished by observing the patient’s eye movements while the head is turned from side to side. Patients who exhibit a maintained forward gaze despite head turning (“doll’s eyes reflex”) are unlikely to have a brainstem-mediated cause of coma. If the eyes remain in a fixed position within the orbits, turning in unison with the head, brainstem dysfunction is suggested. Oculovestibular or “cold water caloric” testing is a more sensitive test for brainstem involvement and cannot voluntarily be resisted (Fig. 142). After elevation of the patient’s head to 30° (this can be done in patients whose cervical spine is not cleared by placing the bed in the reverse Trendelenburg position), 10 to 30 mL of ice water is used to irrigate the external auditory canal. Tympanic membrane perforation and cerumen impaction should be ruled out prior to performing this test. In patients who have an intact brainstem, the response is a slow conjugate deviation of gaze toward the side of the cold water stimulus lasting 30 to 120 seconds. The reflex is short-lived and followed by corrective fast beats of nystagmus toward the midline. This corrective nystagmus is described by the mnemonic
Diagnostic Algorithm Information gathered from the history and physical examination of the patient with depressed consciousness must be used to direct the approach to diagnostic testing (Fig. 14-3). Most often, this information points toward a systemic or metabolic rather than a structural etiology. Neuroimaging studies are performed early in patients with suggested structural causes, but should not precede treatment of quickly reversible conditions such as possible opioid overdose or hypoglycemia. Systemic or metabolic causes of depressed consciousness and coma are most often found on analysis of laboratory studies. Bedside glucose testing definitively confirms or excludes hypoglycemia. Serum electrolytes identify derangements in sodium, CO2, or the anion gap. Changes in serum calcium can be a marker for metastatic disease. A urine dip test is a quick way to identify infection, ketones, or spilling of glucose. Urinalysis itself provides valuable information regarding volume status (specific gravity), infection, and the possible presence of calcium oxalate crystals in the setting of ethylene glycol ingestion. Urine drug testing may be helpful if another cause is not forthcoming. Although an elevated white blood cell count can be a marker for infection, it is nonspecific and rarely helpful. An abnormally low white blood cell count, however, suggests an immunocompromised state and should urgently direct clinical investigation toward an infectious etiology. Thrombocytopenia can be a marker for sepsis or intracranial hemorrhage and may sound a cautionary note against an invasive procedure such as obtaining central venous access or performing a lumbar puncture. Elevated results from serum coagulation studies can be a marker for bleeding tendencies or liver disease. Serum ammonia levels are controversial and have not been shown to be a reliable marker in the setting of depressed consciousness. Thyroid function studies can reveal myxedema coma from hypothyroidism. When CNS pathology such as infection or hemorrhage is suggested but not seen on neuroimaging studies, cerebrospinal fluid analysis is undertaken. Noncontrast computed tomography (CT) of the brain is the preliminary imaging modality of choice in the setting of depressed consciousness and coma. In the majority of ED settings, it is quickly available, making it more suitable for the patient with borderline stability. It is sufficiently sensitive to detect most intracranial hemorrhages that are large enough to cause coma. Contrast-enhanced CT may be used if a tumor or infection is possible. Linear artifacts created by the thick skull base can limit the view of the posterior fossa on CT. For this reason, magnetic resonance imaging of the brain is generally more useful for identifying structural lesions in this region; however, this modality is less feasible in most ED settings due to its cost and limited availability. Angiography may be available in larger tertiary care centers for use in the diagnosis or treatment of intracerebral aneurysms or arteriovenous malformations after initial identification of an intracranial hemorrhage on noncontrast CT. Plain radiography may identify severe pneumonia or acute respiratory distress syndrome, or it may rarely reveal specific types of heavy-metal ingestions such as mercury, iron, or lead in the pediatric population. Electrocardiograms can point to certain ingestions (tricyclic antidepressants, etc.), electrolyte abnormalities (potassium, calcium, etc.), or hypothermia. If nonconvulsive status epilepticus is suggested, or if a patient with status epilepticus has required neuromuscular blockade,
111 DIAGNOSTIC APPROACH TO ALTERED MENTAL STATUS AND COMA
No gag reflex, GCS < 8
Chapter 14 / Depressed Consciousness and Coma
Initial assessment: Vital signs Airway, breathing, circulation IV, oxygen, monitor Bedside glucose Narcan
RSI
History/suggestion of trauma? Yes
No History/suspicion of infection?
C-spine immobilization Yes
Signs of impending herniation? • Cushing’s response (↑ BP, ↓ HR) • Unilateral blown pupil Yes
No
IV antibiotics Consider steroids
Toxic/metabolic workup ECG CT
CT
Toxic/metabolic etiology?
No
Elevate head Neurosurgery consult ↑ RR (PCO2 ~ 35 mm Hg) Mannitol Craniotomy/ventriculostomy CT/OR
CT
Yes
Lesion on CT? Lesion on CT?
Yes
Yes No
Neurosurgery consult OR
Neurosurgery consult OR
Toxic/ metabolic etiology?
Yes Treat appropriately
No
Treat appropriately
Lumbar puncture
Yes Continue antibiotics Admit ICU Neurology consult
Lesion on CT?
Yes
Lumbar puncture
No
Neurosurgery consult OR
Infection? No
No
Hemorrhage?
No Toxic/ metabolic workup
Lumbar puncture
Yes Neurosurgery consult Angiography OR
No Anoxia CVA
MRI Neurology consult
Figure 14-3. Algorithm for diagnostic approach to altered mental status and coma. BP, blood pressure; C-spine, cervical spine; CT, computed tomography; CVA, cerebrovascular accident, HR, heart rate; ICU, intensive care unit; IV, intravenous; OR, operating room; RR, respiration rate; RSI, rapid sequence intubation. (Adapted from Adams J: Emergency Medicine. Fig. 89-2. Philadelphia, Elsevier. Copyright ©2008.)
continuous electroencephalographic monitoring, if available, can provide key information about the patient’s status and guide therapy.
Empirical Management Initial establishment of airway, breathing, and circulation (ABCs) is of primary importance in stabilizing the patient with
altered mental status. Initiation of IV access combined with the administration of oxygen and continuous telemetry monitoring should happen concomitantly within the first few minutes of the patient’s arrival.5 In patients with a GCS score lower than 8, definitive airway control should be obtained unless the coma is readily reversible (e.g., hypoglycemia, opioid overdose). Patients with possible increased ICP should receive lidocaine prior to rapid sequence intubation (see
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
112
Chapter 1). Trauma patients require spinal immobilization in addition to indicated fluid resuscitation. Reversible causes of the patient’s condition should be sought concomitantly with initial stabilization. Administration of the components of the “GI cocktail,” which include dextrose, nalaxone, and thiamine, can quickly reverse the alterations in mental status caused by hypoglycemia, narcotic overdose, and thiamine deficiency, respectively. Further therapy and workup will be dictated by the patient’s history and physical examination. Specific attention should be given to identifying the focal neurologic abnormalities, including pupillary reflexes and pathologic eye movements that suggest mass effect or depressed brainstem function, prompting neuroimaging and evaluation by a neurosurgeon. Empirical administration of mannitol is indicated when there is evidence of transtentorial herniation in this setting. Ventriculostomy and ICP monitoring are commonly performed by the neurosurgeon in the ED. In trauma patients with suspected epidural hematoma who have evidence of brain herniation, the use of burr holes in the skull on the side of the dilated pupil may be a last resort. In patients with compromised brainstem function who lack evidence of herniation, a workup to investigate possible exposure to toxins or metabolic imbalances should proceed while supportive care is provided. Brain tissue is considered to be unsalvageable in patients who have not received sedative medications, are normothermic, and demonstrate lack of brainstem reflex activity.
In patients who demonstrate normal brainstem function, the workup proceeds as supportive care is provided. When an infectious cause is suggested, empirical administration of a broad-spectrum antibiotic should not be delayed for lumbar puncture or other diagnostic tools. Lesions or masses found on brain imaging should prompt evaluation by a neurosurgeon and, if indicated, early operative intervention. In patients in whom a toxic ingestion is possible, activated charcoal is of no proven benefit in most cases and gastric emptying is rarely indicated (see Chapter 145). Specific toxin antidotes, if indicated, can be given, with consultation with a local or regional poison center when required. Early hemodialysis after consultation with a nephrologist should also be considered in patients who have toxic or metabolic abnormalities amenable to this therapy. The vast majority of patients who present with depressed consciousness or coma require admission to the hospital for further treatment and workup. Some patients who have returned to their baseline mental status after reversal of hypoglycemia or opioid overdose may be suitable for discharge directly from the ED or ED observation unit after a period of observation. Patients with alcohol or recreational drug intoxication and no other discernible cause of altered mental status can be discharged when they are clinically sober. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 15
Seizures
Ana M. Davitt and Charles V. Pollack, Jr.
■ PERSPECTIVE Seizure is defined as abnormal neurologic functioning caused by abnormally excessive activation of neurons, either in the cerebral cortex or in the deep limbic system. Epilepsy is defined as recurrent unprovoked seizures due to a genetically determined or acquired brain disorder1; it is not an appropriate term for seizures that occur intermittently and predictably after a known insult, such as alcohol intoxication and withdrawal. Presentation to the emergency department (ED) with a generalized convulsive seizure prompts immediate concern for airway protection and stabilization, followed by a focused search for the cause. Nonconvulsive seizures, which are much less common, may be relatively obscure in their presentation, more diverse in their etiology, and are sometimes more difficult to recognize and control acutely.
Epidemiology and Classification It is estimated that 6% of the U.S. population experience at least one nonfebrile seizure during their lifetime; the annual incidence among adults is 84 per 100,000 population, and more than half of these individuals develop epilepsy.2 In one study, approximately 1% of ED visits were for seizure-related complaints.3 Nearly half of these patients had alcohol or low antiepileptic drug levels implicated as contributing factors. Seizures can be classified as primary or secondary (the latter also termed reactive), as generalized or focal (partial), or as convulsive or nonconvulsive. Table 15-1 shows the distribution of seizures in a typical population of patients. A generalized seizure is defined as abnormal neuronal activity in both cerebral hemispheres. Seizues may be divided into tonicclonic, absence, and myoclonic. Partial seizures or focal seizures usually involve one hemisphere. They are divided into simple partial (in which consciousness is maintained), complex partial (in which consciousness is lost), and those that become secondarily generalized. Some seizures are impossible to classify because of inadequate or inaccurate description of the ictal activity.2,3 Status epilepticus is defined as at least 30 minutes of persistent seizures or a series of recurrent seizures without intervening return to full consciousness,4 although several authors have proposed shortening the time criterion from 30 minutes to 5 minutes.5 Secondary seizures may occur as a result of a vast array of injuries and of illnesses such as intoxication or poisoning,
encephalitis, encephalopathy, organ failure, other metabolic disturbances, infections of the central nervous system, cerebral tumors, pregnancy, and, paradoxically, supratherapeutic levels of anticonvulsants. Seizures in children follow a different distribution, primarily because of the relatively high incidence of febrile seizures and the frequently uncertain observational history of possible ictal activity. Febrile seizure is the most common pediatric seizure, occurring in 2 to 5% of children between 6 months and 5 years of age; 20 to 30% of those children have at least one recurrence. It is important to differentiate between febrile seizure and seizure with fever.6 First-time seizures in infants younger than 6 months may indicate significant underlying pathology and warrant a full assessment.7
Pathophysiology Seizures occur when the abnormal increased electrical activity of the initiating neurons activates adjacent neurons and propagates until the thalamus and other subcortical structures are similarly stimulated. At a cellular level, the pathophysiology is not well understood, although recent research in specific epilepsy syndromes is elucidating possible mechanisms. Investigation of rare inherited epilepsy syndromes has identified mutations in neuronal ion channel proteins, limiting intracellular passage of potassium. Given that the potassium current is the primary force behind repolarization of membranes, depolarization is prolonged in these patients, leading to an increase in neuronal hyperexcitability.1 Other studies have found that malformations of cortical development and glial cells may play a role in epileptogenesis.1 Clinical seizure activity typically, but not always, reflects the initiating focus. When the ictal discharge extends below the cortex to deeper structures, the reticular activating system in the brainstem may be affected, altering consciousness. In generalized seizures, the focus is often deep and midline, which explains the prompt loss of consciousness and bilateral involvement. Seizures are typically self-limited; at some point the hyperpolarization subsides and the bursts of electrical discharges from the focus terminate. This cessation may be related to reflex inhibition, neuronal exhaustion, or alteration of the local balance of neurotransmitters. Partial seizures may represent a similar pathophysiologic process in which less recruitment occurs and the ictal activity does not cross the midline. Because of the more limited focus of abnormal activity, convulsive motor activity may not be the predominant clinical manifestation.3 113
114
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
■ DIAGNOSTIC APPROACH Differential Considerations Because an incorrect diagnosis is expensive and involves loss of driving privileges and exposure to potentially toxic medicines, the first diagnostic task is to determine whether the patient is having a “true” seizure.8 Ictal activity can be irrefutably verified only by electroencephalography (EEG). Other abnormal movements and states of consciousness, including pseudoseizures, can be confused with ictal activity. Other disorders mimicking seizures are listed in Table 15-2.7 Syncope, whether vasodepressive (vagal syncope), orthostatic, or dysrhythmia related, can be confused with seizures by observers. A sudden loss of consciousness followed by abnormal movements can be ictal or syncopal in origin, hence
Table 15-1 Classification of Seizures in a General Adult Population
SEIZURE TYPE
PERCENTAGE
Generalized Tonic-clonic Absence Myoclonic Others
35 1 2 years: 2 mL/kg IV of D50W 6 g over 15–20 min followed by 2 g/hr 0.2–0.5 mg/kg IV/IO/ET or 0.5–1.0 mg/kg PR up to 20 mg 0.05–0.1 mg/kg IV 0.15 mg/kg IV, then 2–10 mcg/kg/min
Phenytoin
20 mg/kg IV at ≤40 mg/min
20 mg/kg IV at 1 mg/kg/min
Fosphenytoin
15–20 mg/kg IV at 100– 150 mg/min or 20 mg/kg IM
20–25 mg/kg IV, then up to 3 mg/kg/min IV up to 159 mg/min IV
Propofol
3–5 mg/kg initial dose, then 1–15 mg/kg/hr infusion 20–30 mg/kg IV at 60– 100 mg/min or as single IM dose 20 mg/kg PR or 10–15 mg/kg IV (initial dose) 5 mg/kg IV at 25 mg/min, then titrate to EEG Via general endotracheal anesthesia
Phenobarbital Valproate Pentobarbital Isoflurane
First-line therapy for eclamptic seizures Monitor airway protection and respiratory drive Monitor airway protection and respiratory drive Monitor airway protection and respiratory drive
During infusion patient should have continuous cardiac and blood pressure monitoring Level of monitoring directed by patient’s status, not drug use Used for status epilepticus; intubation required Intubation may be required Maximum dosage 60 mg/kg/day Dilute 1 : 1 with water; onset is slow Intubation, ventilation, and pressor support are required Monitor with EEG
*Although alternative routes of administration (e.g., IO, PR) have not all been studied in adults, appropriate weight or length based on dosing by pediatric guidelines can be used when the clinical situation dictates. EEG, electroencephalogram; ET, endotracheal; IM, intramuscular; IO, intraosseous; IV, intravenous; PR, rectal administration.
Chapter 15 / Seizures
A pulse oximeter should be applied and oxygen administered as necessary. Optimally, the patient is turned on his or her side to protect the airway from aspiration. If the patient is immobilized on a spine board after trauma, the entire board is tipped up to one side. Preparation should be made for endotracheal intubation in case anticonvulsant drugs fail to terminate the seizure. While these procedures are accomplished, an assistant should be establishing IV access.11 Hypoglycemia is the most common metabolic cause of seizure activity. The only treatment required for the patient may be administration of IV glucose. Prolonged seizure activity may also cause hypoglycemia, so that the cause-and-effect relationship may sometimes be reversed and further therapy is required. Benzodiazepines are the optimal first-line agents for stopping seizure activity in patients of all ages. Available agents include lorazepam (Ativan), diazepam (Valium), and midazolam (Versed). All three are efficacious in terminating seizure activity (see Table 15-3 for doses), but if IV access cannot be achieved, diazepam may be given rectally, endotracheally, or intraosseously; rectal diazepam stops seizures in 70% of patients, compared with 60 to 80% for IV dosing.12 Midazolam can be given intramuscularly, and recent research shows that buccal midazolam works in children.13 If IV access is obtained, however, lorazepam is the agent of choice for
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
116
is the only fully effective pharmacologic treatment for toxic isoniazid seizures, although benzodiazepines have been shown to suppress seizure activity in some cases.26 In seizing females of childbearing age, eclampsia should be considered; in this case, intravenous magnesium (6 g) is the drug of choice. (See Chapter 177.) Approximately 10% of patients will have a second seizure despite magnesium; these patients should get a second 2-g bolus of magnesium.27 If the eclamptic patient continues seizing, magnesium dosing should be repeated; refractory eclamptic seizures can also respond to benzodiazepines or barbiturates with or without phenytoin. Children and psychiatric patients at risk for water intoxication should be considered potential candidates for hypertonic saline therapy, after laboratory confirmation of hyponatremia. Patients who remain unresponsive to the third-level choice of pharmacologic intervention are by definition in refractory status epilepticus. Further choices for therapy at that juncture are general anesthetic doses of midazolam or propofol, barbiturate coma and isoflurane anesthesia; all of which mandate endotracheal intubation.22,23,29 A neuromuscular blocking agent is administered concomitantly to reduce the metabolic burden and potential hyperthermia that can ensue from prolonged status seizures. Anesthetic dosing of midazolam is 0.2 to 0.3 mg/kg bolus, then 0.05 to 2.0 mg/kg/hr, and for propofol it is 2 to 4 mg/kg, then 1 to 15 mg/kg/hr. Both drugs are usually well-tolerated and can be titrated to effect, although propofol is preferable because of its rapid onset and offset of action, which allows the patient to be “awakened” intermittently for examination in the event that continuous EEG monitoring is not available.16
Pivotal Findings When the patient is stabilized with a secure airway and ictal activity is controlled, attention is turned to gathering more complete data.
History History taking in the patient with seizure is directed by two main questions. First, “Was the incident truly a seizure?” This is important because of the broad differential diagnosis for seizures (see Table 15-2) and the notoriously inaccurate descriptions of seizure-like activity from laypersons.10 In general, however, ictal events have six properties: 1. Abrupt onset: Generalized seizures typically occur without an aura. 2. Brief duration: Seizures rarely last longer than 90 to 120 seconds, although bystanders typically overestimate the duration. 3. Altered mental status: Present by definition, except for simple partial seizures. 4. Purposeless activity: For example, automatisms and undirected tonic-clonic movements. 5. Unprovoked: Especially with regard to emotional stimuli; fever in children and substance withdrawal in adults are notable exceptions. 6. Postictal state: An acute confusional state that typically occurs with all seizures except simple partial and absence. Information regarding focality of onset, loss of bowel or bladder control, or tongue biting should also be elicited. The second question to direct the history is, “Does this patient have a history of seizures?” If he or she does have a documented history of seizures, ED evaluation may be limited to a thorough history and consideration of measurement of
anticonvulsant drug levels. History should focus on intercurrent illness or trauma, drug or alcohol use, potential adverse drug-drug interactions with anticonvulsants, medication compliance, a recent change in anticonvulsant dosing regimens, or a change in ictal pattern or characteristics.2,11 Supratherapeutic and toxic levels of some anticonvulsants such as phenytoin and carbamazepine, whether attained chronically or after acute overdose, may cause seizures. If empiric anticonvulsant therapy is indicated before the serum level is available, only 50% of a full loading dose should be given unless the patient is known reliably not to be taking anticonvulsant medication. If the patient does not have a history of seizures and the description of the event is truly consistent with a seizure, the history should focus on potential underlying medical, toxicologic, or neurologic causes. A personal history from the patient, close friend, relative, or medical record may reveal potential ictogenic factors such as recent or remote head trauma, developmental abnormalities, metabolic diseases, drug or alcohol abuse, sleep deprivation, pregnancy, recent travel, previous seizures, or use of herbal supplements. When no witness or family member is available, extensive questioning must await clearance of the postictal confusional state.
Physical Examination The physical manifestations of convulsive ictal activity include hypertension, tachycardia, and tachypnea from sympathetic stimulation. These signs typically resolve quickly after the seizure activity ceases. With more prolonged convulsions, skeletal muscle damage, lactic acidosis, and, rarely, frank rhabdomyolysis may ensue. Autonomic discharges and bulbar muscle involvement may result in urinary or fecal incontinence, vomiting (with significant aspiration risk), tongue biting, and airway impairment. All of these signs are helpful discriminators in the differential evaluation of seizure-like spells. After the seizure activity has ceased, resting vital signs should be evaluated. Fever and underlying infection can cause seizures, although there may be a low-grade temperature elevation immediately after a convulsive generalized seizure. Tachypnea, tachycardia, or an abnormal blood pressure that persists beyond the immediate postictal period may indicate toxic exposure, hypoxia, or a central nervous system lesion. Pertinent physical findings may include nuchal rigidity, stigmata of substance abuse, lymphadenopathy suggestive of HIV disease or malignancy, dysmorphic features, or skin lesions. The examination should also focus on potential adverse sequelae of convulsive seizures, such as head trauma, tongue injury, posterior shoulder dislocation, or back pain. Finally, a complete neurologic examination must be performed. A persistent focal deficit after a seizure (e.g., Todd’s paralysis) often indicates the focal origin of the event but also can be evidence of an underlying stroke. The patient should be carefully examined for papilledema; elevated intracranial pressure can both cause and result from ictal activity. Failure to note steady improvement of postictal depression of consciousness suggests the possibility of an underlying encephalopathy or nonconvulsive status epilepticus.
Ancillary Testing Laboratory. Routine screening studies such as a complete blood count and chemistry profile have little use in the neurologically normal, otherwise healthy, postictal patient with a known seizure disorder for whom a reliable history can be obtained.
117
Acute intracranial process is suspected History of acute head trauma History of malignancy Immunocompromise Fever Persistent headache History of anticoagulation New focal neurologic examination Age older than 40 years Focal onset before generalization Persistently altered mental status CT, computed tomography.
Bedside blood glucose is measured early. Anticonvulsant levels are appropriate in patients known or thought to be taking anticonvulsant medication. Febrile patients are evaluated for the source of the fever. For medically ill adults (e.g., diabetic patients, cancer patients, patients with liver disease, patients taking medications that can affect serum electrolyte levels) and in those presenting with a first-time seizure, appropriate chemistry studies are ordered, including electrolytes and liver function tests.8 Directed toxicologic screens should be obtained if substance abuse is possible. Serum sodium should be evaluated, particularly if mental status remains altered after apparent recovery from the postictal state. Pregnancy testing is useful if eclampsia is possible. If there is any suggestion of meningitis or subarachnoid hemorrhage, lumbar puncture should be performed, with a preceding cranial computed tomography (CT) scan.30 Imaging. In the fully recovered patient without headache and with fully normal mental status and neurologic examination who has had a single, brief seizure, a cranial CT scan can be obtained in the ED or at a follow-up visit at the discretion of the treating physician.4,29 Table 15-4 lists the circumstances under which a head CT is recommended in the ED due to a higher likelihood of discovering an acute abnormality.4 The literature on this issue for first-time nonfebrile seizures in children is also inconclusive.31 Cranial CT is indicated in any age group when there is a possibility of head trauma, elevated intracranial pressure, intracranial mass, persistently abnormal mental status or focal neurologic abnormality, or HIV disease. Electroencephalography. EEG is not consistently available in the ED. It may be particularly useful in specific cases, such as the diagnosis of nonconvulsive status epilepticus, to monitor seizure activity after intubation and neuromuscular blockade,
■ MANAGEMENT Usually, an acute seizure self-terminates or can be pharmacologically terminated before a need arises for active airway management. Rapidly reversible ictal insults (e.g., hypoglycemia, hypoxemia, isoniazid ingestion) should be considered and, if found, treated. Primary abortive therapy in the ED is accomplished as described earlier. Although a number of newer antiepileptic medications have become available, their therapeutic purpose is directed toward chronic rather than acute seizures.32 Identifying a new-onset seizure in the ED generates consideration for further management. The choice to initiate anticonvulsant therapy depends on the risk of seizure recurrence and any underlying predisposing disease, and the risk of initiating anticonvulsant therapy is typically not made by the emergency physician. The initiation of anticonvulsant therapy after a single seizure is an issue of considerable controversy and should be undertaken in consultation with the neurologist who will be following the patient after discharge from the ED.33,34 Prompt treatment of any apparent ictal source discovered in the ED, however, is always appropriate.
■ DISPOSITION Disposition plans must be individualized according to the findings of the ED evaluation and the presence or absence of underlying disease. One quarter of adult patients presenting with seizure-related complaints have new-onset seizures.3 Almost half of them require admission, most because of abnormal CT scans or persistent focal abnormalities; 95% of those who retrospectively required admission were correctly identified by using an ED evaluation consistent with that recommended previously. Patients may be discharged home with early referral to a neurologist if they have a normal neurologic exam, no comorbidities, no known structural brain disease, do not require the use of an antiepileptic drug in the ED, and are felt to be sufficiently resourceful and reliable to comply with follow-up instructions.4 Patients discharged home from the ED should receive appropriate state-specific guidance regarding driver’s license privileges and information for prompt follow-up with a neurologist. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 15 / Seizures
for Emergent Head CT for Table 15-4 Indications New-Onset Seizure Patients
and to help differentiate seizures from other similar presentations. In general, EEGs are most appropriate for the follow-up evaluation of first-time seizures without clear cause after a complete ED evaluation.19
Chapter 16
Headache
Christopher S. Russi
■ PERSPECTIVE Epidemiology Up to 85% of the U.S. adult population complains of significant headaches at least occasionally, and 15% does so on a regular basis. Headache as a primary complaint represents between 3 and 5% of all emergency department (ED) visits. The vast majority of patients who have the primary complaint of headache do not have a serious medical cause for the problem. Tension headache accounts for approximately 50% of patients presenting to the ED, another 30% have headache of unidentified origin, 10% have migraine-type pain, and 8% have headache from other potentially serious causes (e.g., tumor, glaucoma). It is estimated that less than 1% of patients who present to the ED with headache have a life-threatening organic disease.1 The percentages can create a false sense of security, and headache is disproportionately represented in emergency medicine malpractice claims. Although still rare, the most commonly encountered life-threatening cause of severe sudden head pain is subarachnoid hemorrhage (SAH); approximately 20,000 potentially salvageable cases of SAH present to EDs each year. It is estimated that between 25 and 50% of these are missed on the first presentation to a physician.2 The other significant, potentially life-threatening causes of headache occur even less frequently. Meningitis, carbon monoxide poisoning, temporal arteritis, acute angle-closure glaucoma, intracranial hemorrhage (ICH), cerebral venous sinus thrombosis, and increased intracranial pressure can often be linked with specific historical elements and physical findings that facilitate their diagnosis.
Pathophysiology The brain parenchyma is insensitive to pain. The painsensitive areas of the head include the coverings of the brain— the meninges—and the blood vessels, both arteries and veins supplying the brain, and the various tissues lining the cavities within the skull. The ability of the patient to specifically localize head pain is often poor. Much of the pain associated with headache, particularly with vascular headache and migraines, is mediated through the fifth cranial nerve. Such pain may proceed back to the nucleus and then be radiated through various branches of the fifth cranial nerve to areas not directly involved. A specific inflammation in a specific structure (e.g., periapical abscess, sinusitis, or tic douloureux) is much easier to localize than the relatively diffuse pain that may be generated by tension or traction headaches. Pains in the head and 118
neck may easily overlap. They should be thought of as a unit when considering complaints of headache.
■ DIAGNOSTIC APPROACH Differential Considerations The differential diagnosis of headache is complex because of the large number of potential disease entities and the diffuse nature of many types of pain in the head and neck region (Table 16-1). However, in evaluating the patient with a headache complaint, the top priority is to exclude intracranial hemorrhage (SAH and ICH), meningitis, encephalitis, and mass lesions. Carbon monoxide is an exogenous toxin, the effects of which may be reversible by removing the patient from the source and administering oxygen. Carbon monoxide poisoning is a rare example of a headache in which a simple intervention may quickly improve a critical situation. On the contrary, returning the patient to the poisoned environment without a diagnosis could be lethal.
Rapid Assessment and Stabilization If the patient presents in a critical or comatose state, initial stabilization, including airway management, is undertaken as indicated, preceded by a neurologic examination if at all possible. For purposes of the initial assessment, headache can be divided into two categories: accompanied by altered mental status and without altered mental status. Whenever a patient’s mental status is impaired, brain tissue is initially assumed to be compromised. The principles of care centered on cerebral resuscitation address the seven major causes of evolving brain injury: lack of substrate (glucose, oxygen), cerebral edema, intracranial mass lesion, endogenous or exogenous toxins, metabolic alterations (fever, seizure), ischemia, or elevated intracranial pressure.
Pivotal Findings History The history is the pivotal part of the workup for the patient with headache (Table 16-2). 1. The patient should be asked to describe the pattern and onset of the pain. Patients often relate frequent and recurrent headaches similar to the one they have
119
Table 16-1 Differential Diagnosis CRITICAL DIAGNOSES
EMERGENT DIAGNOSES
NONEMERGENT DIAGNOSES
Neurologic, CNS, vessels
Subarachnoid hemorrhage
Shunt failure Traction headaches Tumor/other masses Subdural hematomas
Migraine, various types Vascular, various types Trigeminal neuralgia Post-traumatic Postlumbar puncture Headaches
Toxic/metabolic Environmental Collagen vascular disease Eye/ENT
Carbon monoxide poisoning
Mountain sickness
Temporal arteritis Glaucoma/sinusitis
Musculoskeletal Allergy Infectious disease
Bacterial meningitis/encephalitis
Brain abscess Anoxic headache Anemia Hypertensive crisis
Pulmonary/O2 Cardiovascular Unspecified
Dental problems/temporomandibular joint disease Tension headaches Cervical strain Cluster/histamine headaches Febrile headaches/nonneurologic source of infection Hypertension (rare) Effort-dependent/coital headaches
CNS, central nervous system; ENT, ear, nose, and throat.
Table 16-2 Significant Symptoms SYMPTOM
FINDING
POSSIBLE DIAGNOSES
Sudden onset of pain
Lightning strike or thunder clap with any decreased mentation, any positive focal finding or intractable pain Associated with sudden onset Associated with sudden onset Clicking or snapping. Pain with jaw movement Fulminant pain of the forehead and area of maxillary sinus. Nasal congestion Tender temporal arteries Sudden onset with tearing
Subarachnoid hemorrhage
“Worst headache of their life” Near syncope or syncope Increase with jaw movement Facial pain Forehead or temporal area pain (or both) Periorbital or retro-orbital pain
on this ED visit. A marked variation in headache pattern can signal a new or serious problem. The rate of onset of pain may have significance. Pain with rapid onset of a few seconds to minutes is more likely to be vascular in origin than pain that developed over several hours or days. Almost all studies dealing with subarachnoid bleeding report that patients moved from the pain-free state to severe pain within seconds to minutes. The “thunder clap” or “lightning strike” headache is a real phenomenon, and this response to questioning may lead to the correct diagnosis of subarachnoid hemorrhage, even if the pain is improving at the time of evaluation.3 2. The patient’s activity at the onset of the pain may be helpful. Certainly, headaches that come on during severe exertion have a relationship to vascular bleeding, but again, there is enough variation to make assignment to any specific cause highly variable. The syndrome of coital or postcoital headache is well known, but coitus is also a common time of onset for SAH. These headaches require the same evaluation on initial presentation as any other exertion-related head pain. If the patient can recall the precise activity in
Subarachnoid hemorrhage Subarachnoid hemorrhage Temporomandibular joint disease Sinus pressure or dental infection Temporal arteritis Temporal arteritis or acute angleclosure glaucoma
which he or she was engaging at the time of the onset of the headache (e.g., “I was just getting up out of the chair to answer the doorbell”), sudden onset is extremely likely and evaluation for SAH is warranted. 3. If the patient or nonhospital medical personnel can relate a history of head trauma, the differential diagnosis and emergent causes have narrowed significantly. The considerations now focus on epidural and subdural hematoma, traumatic SAH, skull fracture, and closed-head injury (i.e., concussion and diffuse axonal injury). 4. Toxoplasmosis, cryptococcal meningitis, and abscess are considered higher in the differential in patients with a history of HIV or immunocompromised state. Although such entities are rare, it is important to remember that this subset of patients may have serious disease without typical signs or symptoms of systemic illness (e.g., fever and meningismus). 5. The intensity of head pain is difficult to quantify objectively. Almost all patients who present to the ED consider their headache to be “severe.” Use of a pain scale of 1 to 10 may help differentiate patients initially but has more value in monitoring their response to therapy.
Chapter 16 / Headache
ORGAN SYSTEM
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
120
6. The character of the pain (i.e., throbbing, steady), although sometimes helpful, may not be adequate to differentiate one type of headache from another. 7. The location of head pain is helpful when the patient can identify a specific area. It is useful to have the patient point or try to indicate the area of pain and the emergency physician to then properly examine that area. Unilateral pain is more suggestive of migraine or a localized inflammatory process in the skull (e.g., sinus) or soft tissue.4 Occipital headaches are classically associated with hypertension. Certainly, temporal arteritis, temporomandibular joint disease, dental infections, and sinus infections frequently have a highly localized area of discomfort. Meningitis, encephalitis, SAH, and even severe migraine, although intense in nature, are usually more diffuse in their localization. 8. Exacerbating or alleviating factors may be important. Patients whose headaches rapidly improve when they are removed from their environment may have carbon monoxide poisoning. Most other severe causes of head pain are not rapidly relieved or improved when patients get to the ED. Headaches on awakening are typically described with brain tumors. Intracranial infections, dental infections, and other regional causes of head pain tend not to be improved or alleviated before therapy is given. 9. Associated symptoms and risk factors may relate to the severity of headache but rarely point to the specific causes (Box 16-1). Nausea and vomiting are completely nonspecific. Migraine headaches, increased intracranial pressure, temporal arteritis, and glaucoma can all manifest with severe nausea and vomiting, as can some systemic viral infections with headache. Such factors may point toward the inten-
sity of the discomfort but are not specific in establishing the diagnosis. 10. A prior history of headache, although helpful, does not rule out current serious problems. It is extremely helpful, however, to know that the patient has had a workup for severe disease. Previous ED visits, computed tomography (CT) magnetic resonance imaging, and other forms of testing should be inquired about. Patients with both migraine and tension headaches tend to have a stereotypical recurrent pattern. Adherence to these patterns is also helpful in deciding the degree to which a patient’s symptoms are pursued.
Physical Examination Physical findings associated with various forms of headache are listed in Table 16-3.
Ancillary Testing The vast majority of headache patients do not require additional testing (Table 16-4). The single largest consistent mistake made by emergency physicians in the workup of the headache patient is believing a single CT scan clears the patient of the possibility of SAH or other serious intracranial disease. The CT scan can miss 6 to 8% of patients with SAH, especially in patients with minor (grade I) SAH, who are most treatable.5 The sensitivity of CT for identifying SAH is reduced by nearly 10% for symptom onset greater than 12 hours and by almost 20% at 3 to 5 days. The basic approach to integrating CTs and lumbar puncture in the assessment of headache is outlined in Figure 16-1.6,8,9
BOX 16-1 Risk Factors Associated with Potentially Catastrophic Illness 1. Carbon monoxide poisoning a. Breathing in enclosed or confined spaces with engine exhaust or ventilation of heating equipment b. Multiple family members with the same symptoms c. Pattern of recurrence in one setting (where the exposure is occurring), relief when not in that setting d. Wintertime and working around machinery or equipment producing carbon monoxide (furnaces, etc.) 2. Meningitis/encephalitis/abscess a. History of sinus or ear infection or recent surgical procedure b. Immunocompromised state c. General debilitation with decreased immunologic system function d. Acute febrile illness—any type e. Extremes of age f. Impacted living conditions (e.g., military barracks, college dormitories) g. Lack of primary immunizations 3. Temporal arteritis a. Age > 50 b. Females > males 4 : 1 c. History of other collagen vascular diseases (e.g., systemic lupus) d. Previous chronic meningitis e. Previous chronic illness such as tuberculosis, parasitic infection, fungi 4. Glaucoma—sudden angle-closure a. Not associated with any usual or customary headache pattern
b. History of previous glaucoma c. Age >30 d. History of pain increasing in a dark environment 5. Increased intracranial pressure a. History of previous benign intracranial hypertension b. Presence of a cerebrospinal fluid shunt c. History of congenital brain or skull abnormalities 6. Cerebral venous sinus thrombosis 7. Intracranial hemorrhage (ICH) a. Subarachnoid hemorrhage (SAH) i. Sudden severe pain. “Worst headache of life.” ii. Acute severe pain following sexual intercourse or straining (i.e., heavy lifting) iii. History of SAH or cerebral aneurysm iv. History of polycystic kidney disease v. Family history of subarachnoid hemorrhage vi. Hypertension—severe vii. Previous vascular lesions in other areas of the body viii. Young and middle-aged b. Subdural hematoma (SDH) i. History of alcohol dependency with or without trauma ii. Current use of anticoagulants c. Epidural hematoma (EDH) i. Traumatic injury ii. Lucid mentation followed by acute altered mentation or somnolence iii. Anisocoria on physical examination
121
Table 16-3 Pivotal Findings on Physical Examination FINDING
POSSIBLE DIAGNOSES
General appearance
Alteration of mental status—nonfocal
Meningitis/encephalitis Subarachnoid hemorrhage Anoxia Increased CSF pressure Intraparenchymal bleed Tentorial herniation Stroke Increased CSF pressure Acute angle-closure glaucoma Subarachnoid hemorrhage Increased CSF pressure Subarachnoid hemorrhage Tentorial herniation Intraparenchymal bleed Anoxia/anemia Febrile headache Exertional/coital headaches Febrile headaches Meningitis/encephalitis Temporal arteritis
Alterations of mental status with focal findings Severe nausea/vomiting Vital signs
Hypertension with normal heart rate or bradycardia
Tachycardia Fever HEENT Fundi—loss of spontaneous venous pulsations or presence of papilledema
Neurologic
Tender temporal arteries Increased CSF pressure Subhyaloid hemorrhage Acute red eye (severe ciliary flushing) and poorly reactive pupils Enlarged pupil with third nerve palsy Lateralized motor or sensory deficit Acute cerebellar ataxia
Mass lesions Subarachnoid hemorrhage Acute angle-closure glaucoma Tentorial pressure cone Mass effect (i.e., subdural, epidural, tumor, intraparenchymal hemorrhage) Stroke (rare) Subdural hematoma, epidural hematoma, hemiplegic or anesthetic migraine (rare) Acute cerebellar hemorrhage Acute cerebellitis (mostly children) Chemical intoxication—various types
CSF, cerebrospinal fluid; HEENT, head, eyes, ears, nose, and throat.
Table 16-4 Diagnostic Adjuncts in Headache Assessment TEST
FINDING
DIAGNOSIS
Erythrocyte sedimentation rate (ESR) ECG
Significant elevation Nonspecific ST-T wave changes
CBC CT—head
Severe anemia Increased ventricular size Blood in subarachnoid space Blood in epidural or subdural space Bleeding into parenchyma of brain Areas of poor vascular flow Structural/mass lesion Increased pressure
Temporal arteritis Subarachnoid hemorrhage Increased CSF pressure Anoxia Increased CSF pressure Subarachnoid hemorrhage Epidural/subdural hematoma Intraparenchymal hemorrhage Pale infarct Traction headache secondary to mass effect Pseudotumor cerebri Mass lesions Shunt failure Tumor/other structural lesions Subarachnoid hemorrhage Infection Infection Infection
Lumbar puncture/CSF analysis
Increased protein Increased RBCs Increased WBCs Positive Gram’s stain Decreased glucose
CBC, complete blood count; CSF, cerebrospinal fluid; CT, computed tomography; ECG, electrocardiogram; RBC, red blood cell; WBC, white blood cell.
Chapter 16 / Headache
SIGN
Usually gradual, subtle, dull, nonfocal throbbing pain
Sudden onset, “thunder clap” or “lightning strike,” severe throbbing
Gradual—as general symptoms increase, headache increases—nonfocal
Often pain developing over a few hours from mild to severe. Virtually always focal in nature Sudden in onset
Gradual, dull, nonfocal
Carbon monoxide poisoning
Subarachnoid hemorrhage
Meningitis/ encephalitis/ abscess
Temporal arteritis
Increased intracranial pressure syndromes Vomiting, decreased mentation
Nausea, vomiting, decreased vision
Decreased mentation prominent, irritability prominent. With abscess, focal neurologic findings may be present Decreased vision, nausea, vomiting intense—may confuse diagnosis
May wax and wane as they leave and enter the involved area of carbon monoxide. Throbbing may vary considerably Whenever altered mental status is present, the outcome is decidedly worse
ASSOCIATED SYMPTOMS
CSF, cerebrospinal fluid; CT, computed tomography; ENT, ear, nose, and throat.
Acute angleclosure glaucoma
PAIN HISTORY
DISEASE ENTITIES
History of glaucoma. History of pain going into dark area History of CSF shunt or other congenital brain or skull abnormality
Age over 50. Other collagen vascular diseases or inflammatory diseases
Recent infection Recent facial or dental surgery or other ENT surgery
History of polycystic kidney disease. History of chronic hypertension
Exposure to engine exhaust, old or defective heating systems, most common in winter months
SUPPORT HISTORY
Uncommon
Rare
Uncommon
Uncommon
Uncommon
Rare
PREVALENCE
CT. Shunt function study. If OK, lumbar puncture
Measurement of intraocular pressure
“Steamy” cornea. Midposition pupil poorly reactive. Acute red eye Papilledema. Loss of spontaneous venous pulsations
Sedimentation rate
CT. Lumbar puncture
CT. Lumbar puncture
Carbon monoxide level, cognitive testing
USEFUL TESTS
Tender temporal arteries
Frequently decreased mentation— meningismus, increased blood pressure, decreased pulse, decreased spontaneous venous pulsations, rarely subhyaloid hemorrhage Fever—late in course, decreased spontaneous venous pulsations
No focal neurologic findings. May need cognitive testing
PHYSICAL EXAMINATION
Table 16-5 Causes and Differentiation of Potentially Catastrophic Illness Presenting with Nontraumatic Headache
Rapid intervention with medications required—if no relief, immediate surgery may be required Shunt failure or other cause of significant increased CSF pressure requires involvement of neurosurgery
Usually unrelated and rapidly progressive
When such infection suspected, treat. Do not delay antibiotics and steroids awaiting laboratory results
If CT positive, immediate involvement of neurosurgery. If CT negative, lumbar puncture
May improve on the way to the hospital. Occurs in groups, may involve entire families or groups of people exposed to the carbon monoxide
ATYPICAL OR IMPORTANT ASPECTS
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
122
123
Initial assessment: history and physical
Decreased mentation Focal neurologic finding or meningismis
If meningitis suggested, there should be no unreasonable delay. Do not wait for test results to start empirical antibiotic therapy.
CT
CT +: treat underlying condition
Blood/abscess/ tumor: notify neurosurgery
LP may be performed immediately if the patient has no focal neurologic findings and normal funduscopic examination
CT –: LP
LP + blood: treat subarachnoid hemorrhage
■ MANAGEMENT Empirical
LP –: treat as a basic pain problem, anticipate post-LP headache
LP + infection: treat infection appropriately
Figure 16-1. Initial assessment and treatment of headache. CO, carbon
Patients with headache represent a spectrum of disease. Patients with headache need to receive triage for evaluation according to their symptoms. Clearly, patients with abnormal vital signs or altered mental status require evaluation before patients with less severe symptoms. If history and physical examination point toward potentially lethal causes, however, effort should be made to establish the diagnosis rapidly with ancillary testing. Pain treatment should be started early. The pain medication of choice depends on the particular patient, underlying vital signs, allergies, and general condition; but relief of pain is still an essential part of the physician’s job and should have little effect on the diagnostic workup.
monoxide; CT, computed tomography; LP, lumbar puncture.
Specific Obtaining cerebrospinal fluid should not delay antimicrobial treatment if intracranial infection is suggested. Intravenous antibiotics should precede lumbar puncture. Abnormal mental status, signs of increased intracranial pressure, papilledema focal findings on the neurologic examination, or any other indication suggestive of focal intracranial lesion requires CT before lumbar puncture.
■ DIFFERENTIAL CONSIDERATIONS Certain historical and physical findings can help the emergency physician decide whether the patient falls into an “all clear” or a “warning signal” group. In the warning group, further investigation and testing should be performed on all patients who present with any of the following: (1) sudden onset of headache, (2) “the worst headache ever,” (3) decreased
Specific management for headache is described in Chapter 101. The challenge in emergency medicine, however, is to eliminate life-threatening causes of headache and to treat the patient’s pain.
■ DISPOSITION Most patients presenting with headache are discharged from the ED with appropriate analgesia and follow-up. These represent patients in the all-clear category or those found to have no serious disease after a careful evaluation and testing. Any patients in whom warning findings are noted require more extensive assessment. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 16 / Headache
Suspicion of CO poisoning: check CO level and administer O2
or altered mental status, (4) true meningismus, (5) unexplained fever or bradycardia, (6) focal neurologic deficits on examination, (7) symptoms refractory to treatment or worsening under observation, (8) new onset of headache with exertion, or (9) history of HIV. These patients have the highest risk for significant disease. In addition, a group of reliable “all clear signals” indicates patients who do not require further investigation when all are present: (1) previous identical headaches, (2) normal alertness and cognition by both examination and history of the event, (3) normal examination of the neck showing no meningismus, (4) normal vital signs, (5) normal or nonfocal neurologic examination, and (6) improvement under observation or with treatment. Sequential evaluation and assessment of data are ongoing processes. Patients should be reevaluated while in the ED, and inconsistent findings may require a rapid review of the situation and rethinking of the diagnosis (Table 16-5).7
Chapter 17
Dyspnea
Sabina Braithwaite and Debra Perina
■ PERSPECTIVE
Pathophysiology
Dyspnea is the term applied to the sensation of breathlessness and the patient’s reaction to that sensation. It is an uncomfortable awareness of breathing difficulties that in the extreme manifests as “air hunger.” Dyspnea is often ill defined by patients, who may describe the feeling as shortness of breath, chest tightness, or difficulty breathing. Dyspnea results from a variety of conditions, ranging from nonurgent to lifethreatening. Neither the clinical severity nor the patient’s perception correlates well with the seriousness of underlying pathology and may be affected by emotions, behavioral and cultural influences, and external stimuli.1,2 The following terms may be used in the assessment of the dyspneic patient:
The actual mechanisms responsible for dyspnea are unknown. Normal breathing is controlled both centrally by the respiratory control center in the medulla oblongata, as well as peripherally by chemoreceptors located near the carotid bodies, and mechanoreceptors in the diaphragm and skeletal muscles.3 Any imbalance between these sites is perceived as dyspnea. This imbalance generally results from ventilatory demand being greater than capacity.4 The perception and sensation of dyspnea are believed to occur by one or more of the following mechanisms: increased work of breathing, such as the increased lung resistance or decreased compliance that occurs with asthma or chronic obstructive pulmonary disease (COPD), or increased respiratory drive, such as results from severe hypoxemia, acidosis, or centrally acting stimuli (toxins, central nervous system events). Pulmonary stretch receptors also are thought to play a role.
Tachypnea: A respiratory rate greater than normal. Normal rates range from 44 cycles/min in a newborn to 14 to 18 cycles/ min in adults. Hyperpnea: Greater than normal minute ventilation to meet metabolic requirements. Hyperventilation: A minute ventilation (determined by respiratory rate and tidal volume) that exceeds metabolic demand. Arterial blood gases (ABG) characteristically show a normal partial pressure of oxygen (Po2) with an uncompensated respiratory alkalosis (low partial pressure of carbon dioxide [Pco2] and elevated pH). Dyspnea on exertion: Dyspnea provoked by physical effort or exertion. It often is quantified in simple terms, such as the number of stairs or number of blocks a patient can manage before the onset of dyspnea. Orthopnea: Dyspnea in a recumbent position. It usually is measured in number of pillows the patient must use to lie in bed (e.g., two-pillow orthopnea). Paroxysmal nocturnal dyspnea: Sudden onset of dyspnea occurring while reclining at night, usually related to the presence of congestive heart failure.
Epidemiology Dyspnea is a common presenting complaint among emergency department patients of all ages. Causes vary widely and may be due to a benign, self-limited condition or significant pathology that can produce long-term morbidity and premature mortality. 124
■ DIAGNOSTIC APPROACH Differential Considerations Dyspnea is subjective and has many different potential causes.5 The differential diagnosis list can be divided into acute and chronic causes, of which many are pulmonary. Other etiologies include cardiac, metabolic, infectious, neuromuscular, traumatic, and hematologic (Table 17-1).
Pivotal Findings History Duration of Dyspnea. Chronic or progressive dyspnea usually denotes primary cardiac or pulmonary disease.6 Acute dyspneic spells may result from asthma exacerbation; infection; pulmonary embolus; intermittent cardiac dysfunction; psychogenic causes; or inhalation of irritants, allergens, or foreign bodies. Onset of Dyspnea. Sudden onset of dyspnea should lead to consideration of pulmonary embolism (PE) or spontaneous pneumothorax. Dyspnea that builds slowly over hours or days may represent a flare of asthma or COPD; pneumonia; recurrent, small pulmonary emboli; congestive heart failure; or malignancy.
125
Table 17-1 Differential Diagnoses for Acute Dyspnea CRITICAL DIAGNOSES
EMERGENT DIAGNOSES
NONEMERGENT DIAGNOSES
Pulmonary
Airway obstruction Pulmonary embolus Noncardiogenic edema Anaphylaxis Ventilatory failure Pulmonary edema Myocardial infarction Cardiac tamponade
Spontaneous pneumothorax Asthma Cor pulmonale Aspiration Pneumonia Pericarditis
Pleural effusion Neoplasm Pneumonia (CAP score < = 70) COPD
Mechanical interference Hypotension, sepsis from ruptured viscus, bowel obstruction, inflammatory/infectious process
Pregnancy Ascites Obesity Hyperventilation syndrome Somatization disorder Panic attack Fever Thyroid disease
Cardiac
Congenital heart disease Valvular heart disease Cardiomyopathy
Primarily Associated with Normal or Increased Respiratory Effort Abdominal Psychogenic
Metabolic/endocrine
Toxic ingestion DKA
Infectious Traumatic
Epiglottitis Tension pneumothorax Cardiac tamponade Flail chest Carbon monoxide poisoning Acute chest syndrome
Hematologic
Renal failure Electrolyte abnormalities Metabolic acidosis Pneumonia (CAP score < = 70) Simple pneumothorax, hemothorax Diaphragmatic rupture
Pneumonia (CAP score < = 70) Rib fractures
Anemia
Primarily Associated with Decreased Respiratory Effort Neuromuscular
CVA, intracranial insult Organophosphate poisoning
Multiple sclerosis Guillain-Barré syndrome Tick paralysis
ALS Polymyositis Porphyria
ALS, amyotrophic lateral sclerosis; CAP, community-acquired pneumonia; COPD, chronic obstructive pulmonary disease; CVA, cerebrovascular accident; DKA, diabetic ketoacidosis.
Positional Changes. Orthopnea can result from left-sided heart failure, COPD, or neuromuscular disorders. One of the earliest symptoms seen in patients with diaphragmatic weakness from neuromuscular disease is orthopnea.7 Paroxysmal nocturnal dyspnea is most common in patients with left-sided heart failure,6 but also can be found in COPD. Exertional dyspnea commonly is associated with COPD, but also can be seen with poor cardiac reserve and abdominal loading. Abdominal loading, caused by ascites, obesity, or pregnancy, leads to elevation of the diaphragm, resulting in less effective ventilation and dyspnea. Trauma. Dyspnea can result from trauma, causing fractured ribs, flail chest, hemothorax, pneumothorax, diaphragmatic rupture, pericardial effusion, cardiac tamponade, or neurologic injury.
Symptoms Patient descriptions of dyspnea vary significantly and generally correlate poorly with severity. Fever suggests an infectious cause. Anxiety may point to panic attack or psychogenic dyspnea, if no organic cause can be isolated. PE or myocardial infarction may present with isolated dyspnea or with associated chest pain, particularly if the pain is constant, dull, or visceral.8,9 If the pain is sharp and worsened by deep breathing but not by movement, pleural effusion and pleurisy or pleural irritation from pneumonia or PE are possible. Spontaneous pneumothorax also may produce sharp pain with deep breathing that is not worsened by movement.
Signs Physical signs in dyspneic patients may be consistent with specific illnesses (Table 17-2). Physical findings found in specific diseases also can be grouped as presenting patterns (Table 17-3).
Ancillary Studies Specific findings obtained from the history and physical examination should be used to determine which ancillary studies are needed (Table 17-4). Bedside oxygen saturation determinations, or selective use of ABGs when oximetry is not reliable, are useful in determining the degree of hypoxia and the need for supplemental oxygen or assisted ventilation. An additional resource for quickly assessing ventilatory status is noninvasive waveform capnography. Using both the end-tidal CO2 value and the shape of the waveform itself can be helpful in assessing the adequacy of ventilations as well as potential causes of the dyspnea (See Chapter 3). An electrocardiogram may be useful if the etiology is cardiac or suggests acute pulmonary hypertension. Serum electrolytes may suggest less common possible causes, such as hypokalemia, hypophosphatemia, diabetic ketoacidosis, or hypocalcemia. A complete blood count may identify severe anemia or thrombocytopenia associated with sepsis. The white blood cell count is not sufficiently sensitive or specific to be of discriminatory value. Cardiac markers and D-dimer assay may be useful in pursuing etiologies such as
Chapter 17 / Dyspnea
ORGAN SYSTEM
126
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
Table 17-2 Pivotal Findings in Physical Examination SIGN
PHYSICAL FINDING
DIAGNOSES TO CONSIDER
Vital signs
Tachypnea Hypopnea Tachycardia Hypotension Fever Cachexia, weight loss Obesity Pregnancy Barrel chest “Sniffing” position “Tripoding” position Traumatic injury
Pneumonia, pneumothorax Intracranial insult, drug/toxin ingestion PE, traumatic chest injury Tension pneumothorax Pneumonia, PE Malignancy, acquired immune disorder, mycobacterial infection Hypoventilation, sleep apnea, PE PE COPD Epiglottitis COPD/asthma with severe distress Pneumothorax (simple, tension), rib fractures, flail chest, hemothorax, pulmonary contusion COPD, malignancy, infection Chronic hypoxia, intracardiac shunts or pulmonary vascular anomalies Anemia Neuromuscular disease Chest wall: rib fractures, pneumothorax Diffuse: thrombocytopenia, chronic steroid use, anticoagulation Rib fractures, pneumothorax, tracheobronchial disruption Allergic reaction, infection, tick-borne illness Upper airway edema/infection, foreign body, traumatic injury, anaphylaxis Tension pneumothorax, COPD or asthma exacerbation, fluid overload/CHF, PE CHF, anaphylaxis Bronchospasm CHF, pneumonia, PE Pneumothorax, pleural effusion, consolidation, rib fractures/ contusion, pulmonary contusion Malignancy, infection, bleeding disorder, CHF Infection (viral, bacterial) Pleurisy Intracranial insult
General appearance
Skin/nails
Tobacco stains/odor Clubbing Pallid skin/conjunctivae Muscle wasting Bruising
Neck
Subcutaneous emphysema Hives, rash Stridor JVD
Lung examination
Wheezes Rales Unilateral decrease
Chest examination
Cardiac examination
Extremities Neurologic examination
Hemoptysis Sputum production Friction rub Abnormal respiratory pattern (e.g., Cheyne-Stokes) Crepitance or pain on palpation Subcutaneous emphysema Thoracoabdominal desynchrony Flail segment Murmur S3 or S4 gallop S2 accentuation Muffled heart sounds Calf tenderness, Homans’ sign Edema Focal deficits (motor, sensory, cognitive) Symmetrical deficits Diffuse weakness Hyporeflexia Ascending weakness
Rib or sternal fractures Pneumothorax, tracheobronchial rupture Diaphragmatic injury with herniation; cervical spinal cord trauma Flail chest, pulmonary contusion PE PE PE Cardiac tamponade PE CHF Stroke, intracranial hemorrhage causing central abnormal respiratory drive; if long-standing, risk of aspiration pneumonia Neuromuscular disease Metabolic or electrolyte abnormality (hypocalcemia, hypomagnesemia, hypophosphatemia), anemia Hypermagnesemia Guillain-Barré syndrome
CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; JVD, jugular venous distention; PE, pulmonary embolism.
cardiac ischemia or PE. Amino-terminal pro-B-type natriuretic peptide (NT-proBNP) analysis adds both diagnostic and prognostic value for several causes of dyspnea, including heart failure, PE, and ischemic cardiac disease.9-11 Combinations of specific serum markers can also help define pathology.12-14 Specialized tests, such as ventilation-perfusion scans, chest computed tomography, pulmonary angiography, or, rarely, conventional pulmonary angiography, may confirm the diagnosis of PE.15 If dyspnea is believed to be upper airway in origin, direct or fiberoptic laryngoscopy or a soft tissue lateral radiograph of the neck may be useful.
■ DIFFERENTIAL DIAGNOSIS The range and diversity of pathophysiologic states that produce dyspnea make a simple algorithmic approach difficult.16 After initial stabilization and assessment, findings from the history, physical examination, and ancillary testing are collated to match patterns of disease that produce dyspnea. This process is updated periodically as new information becomes available. Table 17-3 presents recognizable patterns of disease for common dyspnea-producing conditions, along with specific associated symptoms.
Fever, productive cough, chest pain
SH: tobacco use
Exposure (e.g., influenza, varicella) Immune disorder, chemotherapy Exposure (e.g., birds), indolent onset Abrupt onset ± trauma, chest pain, thin males more likely to have spontaneous pneumothorax
Pneumonia
Bacterial
Viral Opportunistic Fungal/parasitic Pneumothorax
Tobacco use, medication noncompliance, URI symptoms, sudden weather change PMH: environmental allergies FH: asthma Weight loss, tobacco or other occupational exposure Gradual onset, dietary indiscretion or medication noncompliance, chest pain PMH: recent MI, diabetes, CHF Abrupt onset, exposure to allergen Dysphagia
Worsening orthopnea, PND
Dysphagia
Air hunger, diaphoresis
Diaphoresis
Episodic fever, nonproductive cough Localized chest pain
Anorexia, chills, nausea, vomiting, exertional dyspnea, cough
Diaphoresis, exertional dyspnea
ASSOCIATED SYMPTOMS
JVD, peripheral edema, S3 or S4 gallop, new cardiac dysrhythmia, hepatojugular reflux Oral swelling, stridor, wheezing, hives
Hemoptysis
Above JVD, tracheal deviation, muffled heart sounds, cardiovascular collapse Retractions, accessory muscle use, tripoding, cyanosis
Decreased breath sounds, subcutaneous emphysema, chest wall wounds or instability
Fever, tachycardia, tachypnea, rales or decreased breath sounds
Tachycardia, tachypnea, low-grade fever
SIGNS AND PHYSICAL FINDINGS
CXR, chest CT: mass, hilar adenopathy, focal atelectasis CXR: pleural effusion, interstitial edema, Kerley B lines, cardiomegaly ECG: ischemia, dysrhythmia NT-proBNP
Waveform capnography
Ultrasound positive for pneumothorax Clinical diagnosis: requires immediate decompression. May verify using bedside ultrasound CXR: rule out infiltrate, pneumothorax, atelectasis (mucus plug)
CXR: pneumothorax, rib fractures, hemothorax
ABG if hypoxia suspected Waveform capnography if altered mental status
ABG (A-a gradient), D-dimer ECG (dysrhythmia, right heart strain) CXR (Westermark sign, Hampton’s hump) , spiral CT, MRV Q V Pulmonary angiogram Ultrasound positive for DVT CXR, CBC, sputum and blood cultures
TESTS
Chapter 17 / Dyspnea
ABG, arterial blood gas; CBC, complete blood count; CHF, congestive heart failure; CT, computed tomography; CXR, chest x-ray; DVT, deep vein thrombosis; ECG, electrocardiogram; FH, family history; HPI, history of present illness; JVD, jugular venous distention; MI, myocardial infarction; MRV, magnetic resonance venography; NT-proBNP, amino-terminal pro-brain natriuretic peptide; PE, pulmonary embolism; PMH, past medical history; PND, paroxysmal nocturnal dyspnea; SH, social history; URI, upper respiratory infection.
Anaphylaxis
Fluid overload
Malignancy
COPD/asthma
Decompensation of simple pneumothorax
HPI: abrupt onset, pleuritic pain, immobility (travel, recent surgery) PMH: malignancy, DVT, PE, hypercoagulability, oral contraception, obesity
Pulmonary embolism
Simple Tension
HISTORY: (DYSPNEA)
DISEASE
Table 17-3 Diagnostic Table: Patterns of Diseases Often Resulting in Dyspnea
127
128
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
Table 17-4 Ancillary Testing in the Dyspneic Patient CATEGORY
TEST
FINDINGS/POTENTIAL DIAGNOSES
Laboratory
Pulse oximetry, selective ABG use Waveform capnography
Hypoxia, hyperventilation (muscular weakness, intracranial event) CO2 retention (COPD, sleep apnea), obstructive or restrictive pulmonary pattern Metabolic versus respiratory acidosis (DKA, ingestions) A-a gradient (PE) Elevated carboxyhemoglobin (inhalation injury or CO poisoning) WBC Increase: infection, stress demargination, hematologic malignancy Decrease: neutropenia, sepsis Hgb/Hct: anemia, polycythemia Smear: abnormal Hgb (i.e., sickling), inclusions Platelets: thrombocytopenia (marrow toxicity) BUN/Cr: acute/chronic renal failure K/Mg/Phos: low levels resulting in muscular weakness Glucose: DKA D-dimer: abnormal clotting activity NT-proBNP: heart failure, PE Troponin: cardiac ischemia or infarct Ischemia, dysrhythmia, S1Q3T3 (PE), right heart strain Pulmonary hypertension, valvular disorders Wall motion abnormalities related to ischemia, intracardiac shunts Bony structures: fractures, lytic lesions, pectus, kyphoscoliosis Mass: malignancy, cavitary lesion, infiltrate, foreign body Diaphragm: eventration, elevation of hemidiaphragm, bowel herniation Mediastinum: adenopathy (infection, sarcoid), air Cardiac silhouette: enlarged (cardiomyopathy, fluid overload) Soft tissue: subcutaneous air Lung parenchyma: blebs, pneumothorax, effusions (blood, infectious), interstitial edema, local consolidation, air bronchograms, Hampton’s hump, Westermark’s sign PE
Complete blood count
Chemistry
Cardiac
ECG Echocardiogram
Radiologic
Chest radiograph
Fiberoptic
scan Q V Pulmonary angiogram CT MRI Soft tissue neck radiograph Ultrasound Bronchoscopy
Laryngoscopy
PE, intervention (thrombolysis) Mass lesion, adenopathy, trauma, PE PE, bony and soft tissue lesions, vascular abnormality Epiglottitis, foreign body Pneumothorax, pleural effusion, impaired cardiac function or pericardial effusion Mass lesion, foreign body Intervention (stenting, biopsy) Mass lesion, edema, epiglottitis, foreign body
A-a, alveolar-arterial; ABG, arterial blood gas; BUN, blood urea nitrogen; CHF, congestive heart failure; CO, carbon monoxide; COPD, chronic obstructive pulmonary disease; Cr, creatinine; CT, computed tomography; DKA,diabetic ketoacidosis; ECG, electrocardiogram; MRI, magnetic resonance imaging; NT-proBNP, amino-terminal pro-brain ventilation-perfusion; WBC, white blood cell. natriuretic peptide; PE, pulmonary embolism; V/Q,
Critical Diagnoses Several critical diagnoses should be promptly considered to determine the best treatment options to stabilize the patient. Tension pneumothorax is such a critical diagnosis. If a dyspneic patient has diminished breath sounds on one side, ipsilateral hyper-resonance, severe respiratory distress, hypotension, and oxygen desaturation, prompt decompression of presumptive tension pneumothorax is necessary. Bedside ultrasonography may assist in confirming pneumothorax. If obstruction of the upper airway is evidenced by dyspnea and stridor, early, definitive assessment and intervention must occur in the emergency department or operating room. Complete obstruction by a foreign body warrants the Heimlich maneuver until the obstruction is relieved or the patient is unconscious, followed rapidly by direct laryngoscopy. Congestive heart failure and pulmonary edema can produce dyspnea and respiratory failure and should be treated as soon as possible if severe.17 Significant dyspnea and wheezing can be seen in anaphylaxis and must be treated promptly to prevent further deterioration. Severe bronchospastic exacerbations of asthma at any age may lead rapidly to respiratory failure and arrest and
should receive vigorous attention, including continuous or frequent administration of a beta-agonist aerosol.18 As mentioned earlier, waveform capnography is a valuable tool for assessing the severity and determining the cause of respiratory distress.
Emergent Diagnoses Asthma and COPD exacerbations can result in marked dyspnea with bronchospasm and decreased ventilatory volumes.19 Sudden onset of dyspnea with a decreased oxygen saturation on room air accompanied by sharp chest pain may represent PE.15 Dyspnea accompanied by decreased breath sounds and tympany to percussion on one side is seen with spontaneous pneumothorax. Dyspnea associated with decreased respiratory effort may represent a neuromuscular process, such as multiple sclerosis, Guillain-Barré syndrome, or myasthenia gravis.14 Unilateral rales, cough, fever, and dyspnea usually indicate pneumonia. Figure 17-1 provides an algorithm for assessment and stabilization of a dyspneic patient. The initial division is based on the degree of breathing effort associated with the symptoms.
129 Pulse oximetry Oxygen supplementation History and physical Directed evaluation
No
Yes Supplement oxygen, pulse oximetry, IV access, cardiac monitor, continuous waveform capnography
Able to maintain own airway or evidence of respiratory failure?
No
Assist ventilation
Assess breath sounds
Unequal, chest trauma
Hypotension, shock, tracheal deviation, JVD?
Yes Bedside U/S with cardiac collapse, ptx
CXR
No Yes
Assess breath sounds, obtain history and physical
+BNP, interstitial pattern on CXR: consider pulmonary edema
Wheezes, clear CXR: consider anaphylaxis, bronchospasm, sharkfin EtCO2 waveform
ECG
CXR
Rhonchi, diminished breath sounds, infiltrate on CXR: consider pneumonia, pleural effusion
Endotracheal intubation or cricothyrotomy and mechanical ventilation
Clear, ± chest pain: consider pulmonary embolism, anginal equivalent, neuromuscular disorder or central neurogenic cause
Bronchodilators, epinephrine, steroids, antihistamines
Antibiotics blood cultures
Unequal, trauma: large pneumothorax or hemothorax on U/S
Thoracostomy ECG
Diuretics, nitrates, ACE inhibitors, morphine, consider CPAP or BiPAP
Immediate chest decompression
Troponin
D-dimer
ABG
Ischemic: cardiac evaluation and treatment
S1Q3T3, right-sided heart strain, sinus tachycardia NSSTWC: consider PE
Abnormal A-a gradient
Anticoagulate, V/Q scan, pulmonary angiogram, spiral CT
Figure 17-1. Rapid assessment and stabilization of a dyspneic patient. ABG, arterial blood gas; ACE, angiotensin-converting enzyme; BiPAP, biphasic
positive airway pressure; BNP, B-type natriuretic peptide; CO, carbon monoxide; CPAP, continuous positive airway pressure; CT, computed tomography; CXR, chest x-ray; ECG, electrocardiogram; EtCO2, end-tidal carbon dioxide; IV, intravenous; JVD, jugular venous distention; NSSTWC, nonspecific ST wave , ventilation-perfusion ratio; U/S, ultrasound. changes (on ECG); PE, pulmonary embolism; RR, respiratory rate; V Q
The most critical diagnoses must be considered first and appropriate intervention taken as necessary. All patients experiencing dyspnea, regardless of possible cause, should be promptly transported to the treatment area. Bedside pulse oximetry should be obtained, and the patient should be placed on a cardiac monitor. If the pulse oximetry is less than 98% saturated on room air, the patient should be placed on supplemental oxygen either by nasal cannula or
mask depending on the degree of desaturation detected. If necessary, the patient should be intubated, and breathing should be assisted with manual or mechanical ventilation. When the airway has been secured, rapid assessment of the patient’s appearance and vital signs can help determine the need for further stabilization. Decreased mental alertness, inability to speak in more than one-word syllables, or certain types of body positioning, signal the presence of significant
Chapter 17 / Dyspnea
Respiratory distress? (RR >24 or 33 years Women >40 years Diabetes mellitus Hypertension Cigarette use/possible passive exposure Elevated cholesterol (LDL)/triglycerides Sedentary lifestyle Obesity Postmenopausal Left ventricular hypertrophy Cocaine abuse Pulmonary embolism Prolonged immobilization Surgery >30 minutes in last 3 mo Prior deep vein thrombosis or pulmonary embolus Pregnancy or recent pregnancy Pelvic or lower extremity trauma Oral contraceptives with cigarette smoking Congestive heart failure Chronic obstructive pulmonary disease Obesity Past medical or family history of hypercoagulability
Table 18-4 Ancillary Testing of Patients with Chest Pain TEST
FINDING
DIAGNOSIS
ECG
New injury
Acute MI Aortic dissection Coronary ischemia Coronary spasm PE Pericarditis
New ischemia
CXR
RV strain Diffuse ST segment elevation Pneumothorax with mediastinal shift Wide mediastinum Pneumothorax Effusion Increased cardiac silhouette Pneumomediastinum
ABG scan or Q V
spiral CT
Hypoxemia, A-a gradient High probability or any positive in patient with high clinical suspicion
Tension pneumothorax Aortic dissection Esophageal rupture Pneumothorax Esophageal rupture Pericarditis Esophageal rupture Mediastinitis PE PE
ABG, arterial blood gas; CT, computed tomography; ECG, electrocardiogram; MI, myocardial infarction; RV, right ventricular.
level may be seen in esophageal rupture. Increased cardiac silhouette may indicate pericarditis or cardiomyopathy. Pneumomediastinum is seen with esophageal rupture and mediastinitis. A serum D-dimer assay may help discriminate patients with PE from those with a possible gastrointestinal cause. A low serum D-dimer in a patient without a high pretest
Aortic dissection Hypertension Congenital disease of the aorta or aortic valve Inflammatory aortic disease Connective tissue disease Pregnancy Arteriosclerosis Cigarette use Pericarditis or myocarditis Infection Autoimmune disease (e.g., systemic lupus erythematosus) Acute rheumatic fever Recent myocardial infarction or cardiac surgery Malignancy Radiation therapy to mediastinum Uremia Drugs Prior pericarditis Pneumothorax Prior pneumothorax Valsalva’s maneuver Chronic lung disease Cigarette use
Findings in Ischemic Table 18-5 Electrocardiogram Chest Pain Classic myocardial infarction Subendocardial infarction Unstable angina Pericarditis
ST segment elevation (>1 mm) in contiguous leads; new LBBB; Q waves ≥0.04 sec duration T wave inversion or ST segment depression in concordant leads Most often normal or nonspecific changes; may see T wave inversion Diffuse ST segment elevation; PR segment depression
LBBB, left bundle-branch block.
probability of PE effectively excludes the diagnosis.13,17,18 (see Chapter 87.) Patients at high pretest probability for PE should undergo diagnostic imaging (multidetector computed tomography [CT], or, less commonly, pulmonary angiography or a ventilation-perfusion lung scan).19 High pretest probability warrants initiation of anticoagulation (heparin or lowmolecular-weight heparin) therapy in the ED before the imaging study, in the absence of a contraindication. Patients with suspected thoracic aortic dissection may be evaluated by CT angiography, transesophageal echocardiog raphy, or magnetic resonance imaging. Selection of imaging modality depends on patient status and availability of the testing equipment.20 CT with a 64 or higher detector scanner has the potential to rule out all of the life-threatening causes of chest pain. Although the “triple rule out” of ACS, PE, and thoracic dissection are the causes most commonly discussed, pneumothorax, mediastinitis, and pericardial effusions are also diagnosed with CT.21,22
137 Completed initial evaluation Chapter 18 / Chest Pain
Initiate emergency care Cardiac monitor Oxygen therapy Aspirin Nitroglycerin IV access Laboratory tests
No
Stable angina — resolved
Possible ACS? Yes
Discharge
Acute STEMI: ST ↑ > 1 mm or New LBBB
Yes
High Risk Elevated troponin New ST ↓ 0.5 mm Recurrent ischemia Heart failure with ischemia Depressed LV function Hemodynamic instability PCI in last 6 months Prior CABG
No ED Chest Pain Center Provocative testing
Low Risk No intermediate or high risk features Nondiagnostic ECG and cardiac markers Age < 70
Discharge
Heparin or LMWH IV nitroglycerin Beta-blocker Revascularization: Fibrinolysis or GP llb-llla inhibitor + PCl
Risk stratification
Intermediate Risk >10 minutes rest pain - resolved T-wave inversion >2 mm Intermediate troponin elevation (TnT >0.01 ng/mL, < 0.1 ng/mL)
Early Conservative Treatment Observation bed Heparin or LMWH IV nitroglycerin Beta-blocker Continuous ECG monitoring Repeat ECGs at regular intervals Cardiac markers
Evidence of ongoing ischemia
Early Invasive Treatment Heparin or LMWH IV nitroglycerin Beta-blocker GP llb-llla inhibitor Diagnostic catheterization in 12–48 hr
Yes
No Provocative testing
Discharge
Figure 18-2. Clinical guidelines for emergency department management of chest pain of myocardial ischemic origin. ACS, acute coronary syndrome;
CABG, coronary artery bypass graft; ECG, electrocardiogram; GP, glycoprotein; IV, intravenous; LBBB, left bundle-branch block; LMWH, low-molecularweight heparin; LV, left ventricular; MI, myocardial infarction; PCI, percutaneous coronary intervention; ST, echocardiographic peak; STEMI, ST segment evaluation myocardial infarction; TnT, troponin T. (Adapted from Gibler WB, Cannon CP, Blonikalns AL, et al: Practical implementation of the guidelines for unstable angina/non-ST-segment elevation myocardial infarction in the emergency department: A scientific statement from the American Heart Association Council on Clinical Cardiology (Subcommittee on Acute Cardiac Care), Council on Cardiovascular Nursing, and Quality of Care and Outcomes Research Interdisciplinary Working Group, in Collaboration with the Society of Chest Pain Centers. Circulation 111:2699, 2005.)
Unstable Angina
Myocardial Infarction
Discomfort is usually moderately severe to severe and rapid in onset. May be more “pressure” than pain. Usually retrosternal, may radiate to neck, jaw, both arms, upper back, epigastrium, and sides of chest (left more than right). Lasts more than 15–30 min and is unrelieved by NTG Changes in pattern of preexisting angina with more severe, prolonged, or frequent pain (crescendo angina). Pain usually lasts >10 min. Angina at rest lasting 15–20 min or new-onset angina (duration 40 years, positive risk factors, and male sex increase possibility
Not clearly related to precipitating factors. May be a decrease in amount of physical activity that initiates pain. Previous history of MI or angina. Over 40 years old, presence of risk factors, and male sex increase probability
Often minimal. May have mild diaphoresis, nausea, dyspnea with pain. Increasing pattern of dyspnea on exertion
SUPPORTING HISTORY
Diaphoresis, nausea, vomiting, dyspnea
ASSOCIATED SYMPTOMS
Common
Common
PREVALENCE IN EMERGENCY DEPARTMENT
Nonspecific findings of a transient nature, may have similar cardiac findings as in MI, especially intermittent diaphoresis
Patients are anxious and uncomfortable. Blood pressure usually is elevated, but normotension and hypotension are seen. The heart rate is usually mildly increased, but bradycardia can be seen. Patients may be diaphoretic and show peripheral poor perfusion. There are no diagnostic examination findings for MI, although S3 and S4 heart sounds and new murmur are supportive
PHYSICAL EXAMINATION
Often no ECG or enzyme changes. Variant angina (Prinzmetal’s) has episodic pain, at rest, often severe, with prominent ST segment elevation
ECG changes (new Q waves or ST segment–T wave changes) occur in 80% of patients. CK-MB and troponins are helpful if elevated, but may be normal
USEFUL TESTS
Table 18-6 Causes and Differentiation of Potentially Catastrophic Illness Presenting with Central Chest Pain or Discomfort
May be pain-free at presentation. Full history is essential. Fewer than 15% of patients hospitalized for unstable angina go on to acute MI. May respond to NTG. May manifest similarly to non–Q wave infarction
Pain may present as “indigestion” or “unable to describe.” Other atypical presentations include altered mental status, stroke, angina pattern without extended pain, severe fatigue, syncope. Elderly may present with weakness, congestive heart failure, or chest tightness. 25% of nonfatal MIs are unrecognized by patient. The pain may have resolved by the time of evaluation
ATYPICAL OR ADDITIONAL ASPECTS
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
138
90% of patients have rapid-onset severe chest pain that is maximal at beginning. Radiates anteriorly in chest to the back interscapular area or into abdomen. Pain often has a “tearing” sensation, and may migrate
Pain is more often lateral-pleuritic. Central pain is more consistent with massive embolus. Abrupt in onset and maximal at beginning. May be episodic or intermittent
Aortic Dissection
Pulmonary Embolism
PAIN HISTORY
Hemoptysis occurs in 16/min. Tachycardia, inspiratory rales, and an increased pulmonic second sound are common. Fever, phlebitis, and diaphoresis are seen in 30–40% of patients. Wheezes and peripheral cyanosis are less common
PHYSICAL EXAMINATION
Arterial blood gases show Po2 < 80 mm Hg in 90%. Widened A-a gradient is helpful. Chest film is usually normal, although 40% show some volume loss, oligemia, or signs of consolidation due to pulmonary infarction. Lung perfusion scan rules out, if truly negative
ECG usually shows left ventricular hypertrophy, nonspecific changes. Chest film shows abnormal aortic silhouette (90%). Aortic angiography has diagnostic accuracy of 95–99%. Transesophageal echocardiogram, CT, MRI most useful in screening
USEFUL TESTS
Patients may present with dyspnea with or without bronchospasm. Acute mortality rate is 10%. Emboli usually from lower extremities above knee, prostate/pelvis venous plexus, right heart. May be subtle cause of COPD exacerbation
Rare for patient to present pain-free. May present with neurologic complications. Physical examination findings may be minimal. Dissection into coronary arteries can mimic MI Ascending aortic aneurysms are more often approached surgically. Descending are generally managed medically
ATYPICAL OR ADDITIONAL ASPECTS
Chapter 18 / Chest Pain
ASSOCIATED SYMPTOMS
139
Pain usually is preceded by vomiting and is abrupt in onset. Pain is persistent and unrelieved, localized along the esophagus, and increased by swallowing and neck flexion
Dull, aching recurrent pain unrelated to exercises or meals. Or it may be a sharp, stabbing, pleuritic-type pain that does not change with chest wall motion. May be severe. Not relieved by NTG
Esophageal Rupture
Pericarditis Dyspnea, diaphoresis
Diaphoresis, dyspnea (late), shock
Dyspnea has a prominent role. Hypotension and altered mental states occur in tension pneumothorax
Pain is often worse when supine, but improves sitting up. Often preceded by viral illness or underlying disease (SLE or uremia)
Older individual with known gastrointestinal problems. History of violent emesis, foreign body, caustic ingestion, blunt trauma, alcoholism, esophageal disease
Chest trauma, previous episode, or asthenic body type
SUPPORTING HISTORY
Rare Tamponade even more rare complication
Rare
Infrequent
PREVALENCE IN EMERGENCY DEPARTMENT
Friction rub may be heard, often fleeting, position-dependent (50% of patients).
Signs of lung consolidation, subcutaneous emphysema may be present
Decreased breath sounds, increased resonance on percussion. Elevated pressure in neck veins occurs in tension pneumothorax
PHYSICAL EXAMINATION
Chest film definitive. Inspiratory and expiratory films may enhance contrast between air and lung parenchyma. Tension pneumothorax should be diagnosed on physical examination Chest film usually has mediastinal air, a left-sided pleural effusion, pneumothorax, or a widened mediastinum. pH of pleural effusion is 16,000/mm3, glucose >200 dL, base deficit >4, LDH >350 IU/L, AST >250 F Units Within 48 hours—Hct drop of 10%, BUN >2 mg/dL, Po2 < 60 mm Hg, calcium 4 L Early appendicitis can be a difficult diagnosis to make. It is still frequently missed on the first physician encounter.
DKA may be the first manifestation of diabetes in some patients. These patients often do not recognize the importance of polydipsia and polyuria. They often present complaining only of nausea, vomiting, and epigastric pain.
Not all patients present with chest pain. A subset of patients, particularly diabetics and the elderly, may present with only nausea, vomiting, and epigastric discomfort.
Normal temperature, WBC, and spontaneous resolution of symptoms suggest biliary colic. Fever, Murphy’s sign, elevated WBC, and suggestive ultrasound indicate cholecystitis.
COMMENTS
Chapter 20 / Nausea and Vomiting
DISORDER
155
Classically, abdominal pain consists of intermittent cramps occurring at regular intervals. The frequency of the cramps varies with the level of the obstruction; the higher the level, the more frequent the cramps. The location of the pain also varies with the level of the obstruction; high obstruction causes epigastric pain, mid-level obstruction causes periumbilical pain, colonic obstruction causes hypogastric pain. Headache is usually present. CO poisoning often occurs during winter months when furnaces are turned on. Family members may have similar symptoms if they also have been exposed. Patients may have neck, chest, or epigastric pain. Forceful, protracted vomiting usually causes the tear. Most cases follow a bout of heavy eating and drinking. Other reported causes include childbirth, defecation, seizures, and heavy lifting.
Bowel obstruction
Uncommon
Uncommon
Common
PREVALENCE
Tachypnea, tachycardia, and hypotension may be present. Escaped air from the esophagus may produce subcutaneous emphysema. Air in the mediastinum produces a “crunching” sound as the heart beats (Hamman’s sign).
No reliable signs of early CO poisoning
Abdominal distention, mild diffuse tenderness, and highpitched “tinkling” bowel sounds may be present. Thorough search for hernias should be performed.
PHYSICAL EXAMINATION
CXR may show pleural effusion, widened mediastinum, pneumothorax, or pneumomediastinum. Esophagogram using water-soluble contrast is definitive.
CO level
Supine and upright plain abdominal films Abdominal CT
USEFUL TESTS
The classic presentation includes forceful vomiting, severe chest pain, subcutaneous emphysema, and multiple CXR findings. There is a growing body of evidence that most cases do not have this “classic” picture. In more subtle presentations, the diagnosis can be difficult to make.
Because CO is a tasteless, odorless gas, patients may not realize they have been exposed. It is important to keep a high index of suspicion during the winter months.
Adhesions, hernias, and tumors account for 90% of bowel obstructions. Other causes include intussusception, volvulus, foreign bodies, gallstone ileus, inflammatory bowel disease, stricture, cystic fibrosis, and hematoma.
COMMENTS
ABGs, arterial blood gases; AST, aspartate aminotransferase; β-hCG, β-human chorionic gonadotropin; BUN, blood urea nitrogen; CK, creatine kinase; CT, computed tomography; CXR, chest radiography; DKA, diabetic ketoacidosis; ECG, electrocardiogram; ERCP, endoscopic retrograde cholangiopancreatography; ETOH, ethyl alcohol; LDH, lactate dehydrogenase; MI, myocardial infarction; NSAID, nonsteroidal anti-inflammatory drug; PUD, peptic ulcer disease; WBC, white blood cell.
Boerhaave’s syndrome
Carbon monoxide (CO) poisoning
HISTORY
DISORDER
Table 20-3 Disorders Commonly Associated with Vomiting—cont’d
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
156
157
Table 20-4 Etiology of Nausea and Vomiting in Pediatric Age Groups NEWBORN
INFANT
CHILD
ADOLESCENT
Infectious
Sepsis, meningitis, UTI, thrush Atresia and webs, malrotation, stenosis, meconium ileus, Hirschsprung’s disease Reflux, overfeeding, gastric outlet obstruction, volvulus
Pneumonia, otitis media, thrush Pyloric stenosis, intussusception, Hirschsprung’s disease Reflux, gastritis, milk intolerance
Gastroenteritis
Gastroenteritis, URI
Bezoars, chronic granulomatous disease
PUD, superior mesenteric syndrome Achalasia, hepatitis
Neurologic
Subdural hematoma, hydrocephalus
Subdural hematoma
Metabolic
Organic or amino acidemias, urea cycle defects, galactosemia, hypercalcemia, phenylketonuria, kernicterus Idiopathic, cardiac failure
Hereditary fructose intolerance, disorders of fatty acid metabolism, uremia, adrenal hyperplasia, kernicterus
Anatomic Gastrointestinal
Other
Rumination, cardiac failure
Appendicitis, pancreatic, hepatitis, other food intolerance Neoplasia, migraine, Reye’s syndrome, motion sickness, hypertension Diabetes, vitamin A excess
Cyclic vomiting syndrome, toxins, food poisoning, Munchausen syndrome by proxy
Neoplasia, migraine, motion sickness, hypertension Diabetes, pregnancy, acute intermittent porphyria
Psychogenic, anorexia
PUD, peptic ulcer disease; URI, upper respiratory infection; UTI, urinary tract infection. Adapted from Li HK, Sunku BK: Vomiting and nausea. In Wyllie R, Hyams JS (eds): Pediatric Gastrointestinal and Liver Disease: Pathophysiology, Diagnosis, Management. Philadelphia, Saunders, 2005, pp 127–149.
Table 20-5 Commonly Used Medications for the Treatment of Nausea and Vomiting MEDICATION
DOSE
COMMENTS
Promethazine (Phenergan)
Adult: 12.5–25 mg IV, IM, PO, or by rectum Pediatric: 0.25–1 mg/kg/dose q4–6 h prn IV, IM, PO, or by rectum; max 25 mg/dose Adult: 5–10 mg IM, or PO; 2.5–10 mg IV; 25 mg by rectum Pediatric: 0.4 mg/kg/24 hr tid-qid PO or by rectum; 0.1–0.15 mg/ kg/dose tid-qid IM; max 40 mg/24 hr
May be repeated every 4-6 hr, until cessation of vomiting. Dry mouth, dizziness, blurred vision. Boxed warning for use under 2 yrs old May be repeated every 4 hr by IV or IM or every 12 hr by rectum, until cessation of vomiting. Lethargy, hypotension, extrapyramidal effects Dystonic reactions, tardive dyskinesia, neuroleptic malignant syndrome Headache, dizziness, and musculoskeletal pain
Prochlorperazine (Compazine) Metoclopramide (Reglan) Ondansetron (Zofran)
Adult: 10 mg IM or IV, may repeat q6 h Pediatric: 1–2 mg/kg/dose q2–6 h IV q2–3 hr Adult: 4 mg IV single dose Pediatric: up to 40 kg: 0.1 mg/kg; >40 kg: 4 mg/dose IV single dose
IM, intramuscular; IV, intravenous.
generated much interest because of their beneficial effect in chemotherapy-induced emesis. Their principal site of action is the area postrema, although they also affect receptors in the GI tract. Several studies in small series of patients have looked at their effect in overdose of theophylline and acetaminophen. With both of these agents, overdose causes vomiting, and oral intake is required as part of therapy (multiple-dose charcoal and N-acetylcysteine). It is well documented that the vomiting often prevents effective oral therapy in patients with overdose of these agents. These studies showed that ondansetron stopped the vomiting and allowed oral therapy to proceed. The dose was 8 mg given intravenously over 20 minutes. The side effects of the serotonin receptor antagonists are mild and include headache and constipation.7,8 The prokinetic agents are useful in patients with gastropa resis, gastroesophageal reflux disease, and other putative dys motility syndromes. Metoclopramide (Reglan) has the most applicability in the ED. It has dopamine antagonist activity at the CTZ and exerts anticholinergic and antiserotonin effects. The primary effect is increased gastric emptying; the exact mechanism for this is not understood. Metoclopramide has
multiple antiemetic actions and may be used as a generalpurpose agent. Other prokinetic agents, such as cisapride (Pro pulsid), do not cross the blood-brain barrier. They are not useful as general-purpose antiemetics. Prokinetic agents are used in patients with isolated gastric motility disorders. The most common side effects of metoclopramide are restlessness, drowsiness, and diarrhea. These effects are usually mild and transient. Antihistamines are useful in nausea and vomiting associated with motion sickness and vertigo. Agents such as dimenhydri nate (Gravol, Dramamine) and meclizine (Antivert) directly inhibit vestibular stimulation and vestibular-cerebellar path ways. Their anticholinergic effect also may contribute to their effectiveness in vertigo and motion sickness. Antihista mines have some role as general antiemetics but are better used in the prevention of motion sickness; for nausea and vomiting, they are less effective than the phenothiazines. The most common side effects of antihistamines are drowsi ness, blurred vision, dry mouth, and hypotension. The newer, less-sedating antihistamines are thought to be less effective as antiemetics.
Chapter 20 / Nausea and Vomiting
ETIOLOGIC CATEGORY
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
158
The anticholinergic agent scopolamine in a transdermal patch (Transderm Scō p) or hyoscine (Buscopan) in an oral form may be used for prophylaxis and treatment of motion sickness. These agents also have mild efficacy in preventing cytotoxic chemotherapy-related nausea and vomiting but are not useful in the emergency department. Benzodiazepine medications have been used for nausea and vomiting, with variable results. Limited studies have evalu ated the efficacy of benzodiazepines in the treatment of hyperemesis gravidarum, in prophylaxis for emetogenic che motherapy, and preoperatively for minor gynecologic surgery. Although this aspect was not directly measured in these studies, the studies inferred that part of the response may be related to the anxiolytic component. No studies have addressed the use of benzodiazepines to treat nonspecific nausea and vomiting in an ED population. Many of the new medications for nausea and vomiting are first tested as agents for prevention and treatment related to chemotherapy and postoperative nausea and vomiting (PONV). A new oral neurokinin-1 antagonist, aprepitant (Emend), has been found to be an effective adjunctive agent for use in patients receiving cancer chemotherapy.9 Aprepitant blocks the effects of substance P in the brain. Currently, it is not indicated for use in patients with established nausea and vomiting. The medication choice is directed at the underlying cause of the nausea and vomiting, if known, such as motion sickness, PONV, or nausea and vomiting related to cancer chemother apy. For all other patients, the choice of antiemetic agent has not been well studied in emergency medicine. One study found droperidol to be more effective than prochlorperazine or metoclopramide as compared with placebo for moderate to severe nausea of any cause.10 The same limitation is true of preferred agents used in the field. One study found that ondansetron was moderately effective in the treatment of nausea and vomiting in this setting.11 As with adults, the underlying cause of nausea and vomiting is first addressed in determining treatment choices.6 Most of the same agents used in adults are recommended for children in a weight-based dosing regimen. Ondansetron and metoclo pramide have value for antiemetic treatment to reduce nausea and vomiting in pediatric patients. These agents are particu larly effective in improving gastroenteritis patients’ ability to maintain oral hydration.12
Special Situations Medications such as antihistamines are frequently used to reduce the incidence of nausea and vomiting when opioid anal gesics are administered in the ED for pain control. Studies have demonstrated that the incidence of nausea and vomiting related to opioid administration in the ED is low and that these medica tions have little efficacy in reducing nausea and vomiting.13,14 Many agents have been advocated for the treatment of nausea and vomiting in pregnancy (NVP). The treatments include nonpharmacologic—avoiding triggers, dietary changes, acupunc ture acupressure, ginger, and behavioral therapy—and pharmacologic—pyridoxine, antihistamines, metoclopramide, ondansetron, or prochlorperazine. Hyperemesis gravidarum is treated essentially as for NVP. For mild symptoms, pyridoxine (vitamin B6), acupressure, ginger, and administration of anti emetics including antihistamines, metoclopramide, ondanse tron, prochorperazine, and phenothiazines may be used. Pyridoxine, acupressure, and ginger are thought to be of benefit but are not commonly used in the ED. For severe symptoms, hospitalization, fluids, corticosteroids, and electrolyte replace ment may be needed. No specific medication has been shown to be superior in the treatment of hyperemesis gravidarum.15
Treatment of PONV is well known. Approximately one third of the patients undergoing surgery may experience nausea and vomiting unless they receive appropriate prophy lactic treatment, and the incidence is surgical procedure– dependent. Droperidol, metoclopramide, ondansetron, and dexamethasone have been used to reduce PONV.16 However, the need for antiemetic therapy during procedural sedation in the ED is not well studied. Many of the same drugs associated with PONV used by anesthesiologists in the operating room are used in the ED for procedural sedation. Nitrous oxide and propofol have been associated with a higher incidence of nausea and vomiting. Although the optimal medication for PONV has yet to be determined, ondansetron is considered a first-line agent in some studies; this recommendation could be extended to postprocedural sedation–related nausea and vom iting in the ED. Chemotherapy-related nausea and vomiting may be seen in the ED. The chemotherapy-induced nausea and vomiting may be acute (up to 24 hours) or delayed (after 24 hours).17 The incidence of nausea and vomiting is correlated with the emetic potential of the chemotherapeutic agents, the patient’s risk factors and other comorbid disorders, and antiemetic treat ment. Patients commonly are given the serotonin antagonists dexamethasone and aprepitant for both immediate and delayed prophylaxis. Cannabinoids also have been used to control che motherapy-induced nausea and vomiting. Choice of agents for treatment in the ED has not been studied, but for this indication the serotonin antagonists and aprepitants are used.
■ DISPOSITION Hospital admission is appropriate when the patient has a sig nificant underlying disease, has an unclear diagnosis and responds poorly to fluid and antiemetic therapy, continues to experience uncontrolled emesis refractory to medication, or is at the extremes of age with poor response to treatment. A cat egory subject to broad interpretation is patients in whom the diagnosis is unclear and prospects for timely follow-up are poor (e.g., the patient has no family physician, lacks transportation, is indigent, habitually abuses drugs or alcohol, or has a lan guage barrier). Discharge may be considered if no serious underlying illness is present, the response to fluid and anti emetic therapy is good, the patient is able to take clear liquids before discharge, and the prospects for follow-up and observa tion at home are favorable. Close follow-up is arranged for most discharged patients, preferably with their primary care physician, in 24 to 48 hours. At discharge, the patient is prescribed medications as needed and is advised to restart oral intake with small feedings of a liquid diet with gradual return to a normal diet. Some experts have recommended the nausea and vomiting diet, which requires the least amount of gastric neuromuscular work. It is a three-step diet: Sports drinks and bouillon are recommended in step 1; soups are recommended in step 2; and foods that require the least amount of gastric “work” are recommended in step 3, such as meals high in protein and low in lipids.16 Clear instruc tions are given to return to the emergency department if there is a recurrence, change, or deterioration in symptoms. Causes for nausea and vomiting frequently remain undiag nosed. Some cases declare themselves or resolve over time; reevaluation and close follow-up are imperative for patients with continuing symptoms. In patients with persistent or recurring symptoms, psychogenic causes or cyclic vomiting syndrome should be considered. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 21
Abdominal Pain
Rimon N. Bengiamin, Gavin R. Budhram, Kelly E. King, and John M. Wightman
■ PERSPECTIVE Abdominal pain is a common emergency department (ED) complaint but, for many reasons, is often diagnostically challenging. The nature and quality of abdominal pain may be difficult for the patient to convey. Physical examination findings with this complaint are variable and can be misleading. The location and severity of the pain may change over time. Benign-appearing symptoms and presentations may evolve into life-threatening conditions. Conversely, patients presenting with severe symptoms may carry a relatively benign diagnosis. All of these factors make evaluation of patients with acute abdominal pain challenging in the ED setting.
Epidemiology Abdominal pain is a common presenting complaint, accounting for up to 10% of all ED visits. Some of the most common causes of acute abdominal pain are listed in Table 21-1. Many patients present with pain and other symptoms that are not typical of any specific disease process. A specific diagnosis may not be possible in about one in every four individuals presenting with this chief complaint.1 In addition, several adult groups deserve special consideration: the elderly (older than 65 years of age), the immunocompromised, and women of reproductive age. Elderly patients with acute abdominal pain are more likely to have a life-threatening process as the cause of their pain. Conditions such as diverticulitis, ruptured abdominal aneurysm, or mesenteric ischemia may manifest atypically and be rapidly progressive. Decreased diagnostic accuracy, coupled with increased probability of severe disease, results in increased mortality in elderly patients with abdominal pain.2 Increasingly, emergency physicians are seeing patients in immunocompromised states secondary to HIV/AIDS, chemotherapy, and immunosuppressive drugs. For many reasons, these patients also prove challenging. Their clinical presentation can be misleading owing to atypical physical and laboratory findings, such as lack of fever or elevated white count. In regard to infection, the scope of the differential diagnosis also should be broader than usual.3–7 Presentations in the immunocompromised patient may be highly variable and subtle and are discussed in Chapter 181. The evaluation of abdominal pain in women involves a differential diagnosis of considerable extent and often requires a more in-depth physical exam and further diagnostic testing. Pelvic organs may be the source of significant pathology in
both the pregnant and the nonpregnant patient. The possibility of ectopic pregnancy in women of reproductive age greatly increases the risk of serious disease with a high potential for misdiagnosis. During pregnancy the uterus becomes an abdominal rather than a pelvic organ and may displace the normal intraperitoneal contents, adding complexity to the evaluation of these patients.8 Nonpregnant patients require evaluation for various ovarian and uterine pathology states.
Pathophysiology Pathology in the gastrointestinal and genitourinary tracts remains the most common source of pain perceived in the abdomen. Also, pain can arise from a multitude of other intraabdominal and extra-abdominal locations (Box 21-1). Abdominal pain is derived from one or more of three distinct pain pathways: visceral, somatic, and referred. Visceral pain results from stimulating autonomic nerves invested in the visceral peritoneum surrounding internal organs. It is often the earliest manifestation of a particular disease process. Distention of hollow organs by fluid or gas and capsular stretching of solid organs from edema, blood, cysts, or abscesses are the most common stimuli. This discomfort is poorly characterized and difficult to localize. If the involved organ is affected by peristalsis, the pain often is described as intermittent, crampy, or colicky. In general, visceral pain is perceived from the abdominal region that correlates with the embryonic somatic segment: ■
Foregut structures (stomach, duodenum, liver, and pancreas) are associated with upper abdominal pain. ■ Midgut derivatives (small bowel, proximal colon, and appendix) are associated with periumbilical pain. ■ Hindgut structures (distal colon and genitourinary tract) are associated with lower abdominal pain. Visceral pain can be perceived in a location remote from the actual disease process. Localization occurs with the extension of the disease process beyond the viscera. A classic example is that of the early periumbilical pain of appendicitis (midgut). When the parietal peritoneum becomes involved, the pain localizes to the right lower quadrant of the abdomen, the usual location of the appendix. Somatic pain occurs with irritation of the parietal peritoneum. This is usually caused by infection, chemical irritation, or another inflammatory process. Sensations are conducted by 159
Peak age in adolescence and young adulthood; less common in children and elders. Higher perforation rate in women, children, and elders and in pregancy. Mortality rate is 0.1% but increases to 2–6% with perforation. Peak age 35–60 yr; rare in patients younger than 20. Female-to-male ratio of 3:1. Risk factors include multiparity, obesity, alcohol intake, and use of birth control pills. Average age 30–40 yr, primarily in men. Prior history or family history of stones is common.
Acute appendicitis
Ureteral colic
Biliary tract disease
Occurs in all age groups.
EPIDEMIOLOGY
Gastric, esophageal, or duodenal inflammation
CAUSATIVE DISORDER/CONDITION
Crampy RUQ pain radiates to right subscapular area. Prior history of pain is common. May have nausea or postprandial pain. Longer duration of pain favors diagnosis of cholecystitis or cholangitis. Acute onset of flank pain radiating to groin. Nausea, vomiting, and pallor are common. Patient usually writhing in pain.
Passage of gallstones causes biliary colic. Impaction of a stone in cystic duct or common duct causes choecystitis or cholangitis. Family history, gout, Proteus infection. Renal tubular acidosis and cystinuria lead to stone formation.
Epigastric or periumbilical pain migrates to RLQ over 8–12 hr (50–60%). Later presentations associated with higher perforation rates. Pain, lowgrade fever (15%), and anorexia (80%) common; vomiting less common (50–70%).
Epigastric radiating or localized, associated with certain foods. Pain may be burning. In some cases, exacerbation in supine position.
PRESENTATION
Appendiceal lumen obstruction leads to swelling, ischemia, infection, and perforation.
Caused by gastric hypersecretion, breakdown of mucoprotective barriers, infection, or exogenous sources.
ETIOLOGY
Table 21-1 Common Causes of Abdominal Pain
Temperature normal in biliary colic, elevated in cholecystitis and cholangitis. RUQ tenderness, rebound, and jaundice (less common) may be present. Vital signs usually normal. Tenderness on CVA percussion with benign abdominal examination.
Mean temperature 38° C (100.5° F). Higher temperature associated with perforation. RLQ tenderness (90–95%) with rebound (40–70%) in majority of cases. Rectal tenderness in 30%.
Epigastric tenderness without rebound or guarding. Perforation or bleeding leads to more severe clinical findings.
PHYSICAL EXAMINATION
WBC count elevated in cholecystitis and cholangitis. Lipase and liver function tests may help differentiate this from gastritis or ulcer disease. Ultrasound shows wall thickening, pericholecystic fluid, stones, or duct dilatation. Hepatobiliary scintigraphy diagnoses gallbladder function. Urinalysis usually shows hematuria. Noncontrast CT is sensitive and specific. US with fluid bolus useful diagnostically.
Uncomplicated cases are treated with antacids or histamine H2 blockers before invasive studies are contemplated. Gastroduodenoscopy is valuable in diagnosis and biopsy. Testing for H. pylori with blood or biopsy specimens. If perforation is suspected, an upright chest radiograph is obtained early to rule out free air. CT may be beneficial. Leukocyte count usually elevated or may show left shift. Urinalysis may show sterile pyuria. CT is sensitive and specific. US may have use in women, pregnancy, and children with RLQ pain.
USEFUL TEST(S)
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
160
More common in persons of young and middle age, women of childbearing age or persons of low socioeconomic status, and patients with psychiatric disorders. Up to 10% of patients older than 50 years of age will have intraabdominal cancer.
Incidence increases with advancing age, affects males more often than females. Recurrences are common. Often called “left-sided” appendicitis. Common diagnosis. Seasonal. Most common misdiagnosis of appendicitis. May be seen in multiple family members. History of travel or immune compromise. More common in females, the elderly, the very young, and patients on narcotics.
EPIDEMIOLOGY
Idiopathic or hypokinesis secondary to disease states (low motility) or exogenous sources (diet, medications). Unknown. Early or undiagnosed presentation of pathologic conditions.
Colonic diverticula may become infected or perforated or cause local colitis. Obstruction, peritonitis, abscesses, fistulas result from infection or swelling. Usually viral. Consider invasive bacterial or parasitic in prolonged cases, in travelers, or immune-compromised patients.
ETIOLOGY
Variable but tends to be chronic or recurrent.
Abdominal pain; change in bowel habits.
Pain usually poorly localized, intermittent, crampy, and diffuse. Diarrhea is key element in diagnosis; usually large-volume, watery. Nausea and vomiting usually begin before pain.
Change in stool frequency or consistency commonly reported. LLQ pain is common. Associated with fever, nausea/vomiting; rectal bleeding may be seen.
PRESENTATION
Abdominal examination usually nonspecific without peritoneal signs. Watery diarrhea or no stool noted on rectal examination. Fever is usually present. Variable, nonspecific without peritoneal signs. Rectal exam may reveal hard stool or impaction. Variable but no peritoneal signs. Rectal exam should be done to evaluate for subtle signs of pathology, including heme-positive stool, fistulas, and fissures.
Fever usually of low grade. LLQ pain without rebound is common. Stool may be heme-positive.
PHYSICAL EXAMINATION
Variable and often can be done on an outpatient basis.
Radiographs may show large amounts of stool. This is a diagnosis of exclusion.
Usually symptomatic care with antiemetics and volume repletion. Heme-positive stools may be a clue to invasive pathogens. Key is not using this as a “default” diagnosis and missing more serious disease.
Results on most tests usually normal. Plain radiographs may show obstruction or mass effect. CT is often diagnostic.
USEFUL TEST(S)
Chapter 21 / Abdominal Pain
CT, computed tomography; CVA, costovertebral angle; LLQ, left lower quadrant; LUQ, left upper quadrant; RLQ, right lower quadrant; RUQ, right upper quadrant; US, ultrasonography; WBC, white blood cell.
Nonspecific abdominal pain
Constipation and obstipation
Acute gastroenteritis
Diverticulitis
CAUSATIVE DISORDER/CONDITION
161
162
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
Important Extra-abdominopelvic Causes of
BOX 21-1 Abdominal Pain
Thoracic Myocardial infarction/unstable angina Pneumonia Pulmonary embolism Herniated thoracic disk (neuralgia) Pericarditis/myocarditis Genitourinary Testicular torsion Abdominal Wall Muscle spasm Muscle hematoma Herpes zoster Infectious Streptococcal pharyngitis (more often in children) Rocky Mountain spotted fever Mononucleosis Systemic Diabetic ketoacidosis Alcoholic ketoacidosis Uremia Sickle cell disease Porphyria Systemic lupus erythematosus Vasculitis Glaucoma Hyperthyroidism Toxic Methanol poisoning Heavy metal toxicity Scorpion bite Snake bite Black widow spider bite Adapted from Purcell TB: Nonsurgical and extraperitoneal causes of abdominal pain. Emerg Med Clin North Am 7:721, 1989.
the peripheral nerves and are better localized than the visceral pain component. Figure 21-1 illustrates some more typical pain locations corresponding to specific disease entities. Somatic pain is often described as intense and constant. As disease processes evolve to peritoneal irritation with inflammation, better localization of the pain to the area of pathology generally occurs. Referred pain is defined as pain felt at a distance from its source because peripheral afferent nerve fibers from many internal organs enter the spinal cord through nerve roots that also carry nociceptive fibers from other locations, as illustrated in Figure 21-2. This makes interpretation of the location of noxious stimuli difficult for the brain. Both visceral pain and somatic pain can manifest as referred pain. Two examples of referred pain are the epigastric pain associated with an inferior myocardial infarction and the shoulder pain associated with blood in the peritoneal cavity irritating the diaphragm. Gynecologic and obstetric presentations are discussed in other chapters. Notably, any abdominal pain in a female may represent referred pain from pelvic structures or an extension of a pelvic process, as in the case of perihepatic inflammation with pelvic inflammatory disease.
■ DIAGNOSTIC APPROACH The clinical approach should focus on early stabilization, history, physical examination, and any ancillary tests collectively facilitating appropriate management and disposition plans.
Differential Considerations Classically, potential diagnoses are divided into intraabdominopelvic (intraperitoneal, retroperitoneal, and pelvic) causes (e.g., appendicitis, cholecystitis, pancreatitis) and extraabdominopelvic processes (e.g., pneumonia, myocardial infarction, ketoacidosis). Although significant morbidity and mortality can result from many disorders causing abdominal pain, a few processes warrant careful consideration in the ED. Table 21-2 lists important potentially life-threatening nontraumatic causes of abdominal pain. This group represents the major etiologic disorders likely to be associated with hemodynamic com promise and for which early therapeutic intervention is critical.
Rapid Assessment and Stabilization As with any complaint, triage is the first critical step in management. Most patients presenting with abdominal pain do not have hemodynamic instability, but up to 7% of these patients may have a life-threatening process. This percentage is higher in elders and immunocompromised patients.1 Physiologically compromised patients should be brought to a treatment area immediately and resuscitation initiated. Profound shock or protracted emesis can lead to airway com promise necessitating intubation. These patients are often severely volume depleted and require rapid intravenous access and volume resuscitation with an isotonic crystalloid solution, titrated to a physiologic endpoint. Extreme conditions such as ruptured abdominal aortic aneurysm, massive gastrointestinal hemorrhage, ruptured spleen, and hemorrhagic pancreatitis may require blood or blood product replacement. Bedside ultrasonography can be used to quickly evaluate patients for free intraperitoneal fluid, volume status, and presence of aortic pathology. Ultrasound assessment should be part of the initial physical examination and can be invaluable in guiding treatment and disposition. Because any of the immediately life-threatening entities may necessitate surgical intervention or management, early surgical consultation is indicated.
Pivotal Findings History A careful and focused history is central to unlocking the puzzle of abdominal pain. Box 21-2 lists some historical questions with high yields for serious pathology. Language and cultural differences may influence accurate communication and mutual understanding. Abrupt onset often is indicative of a more serious cause; however, delayed presentations also may represent a surgical condition. Surgical causes of abdominal pain are more likely to manifest with pain first, followed by nausea and vomiting, rather than with nausea and vomiting followed by pain. Localization and pain migration also are helpful components of the pain history. Diffuse pain generally is nonsurgical, but it may represent the early visceral component of a surgical process. Colicky pain is indicative of hollow viscus distention, and
163
Figure 21-1. Differential diagnosis of
acute abdominal pain. CHF, congestive heart failure; GERD, gastroesophageal reflux disease; LLL, left lower lobe; RLL, right lower lobe.
RIGHT UPPER QUADRANT PAIN Biliary colic Cholecystitis Gastritis GERD Hepatic abscess Acute hepatitis Hepatomegaly due to CHF Perforated ulcer Pancreatitis Retrocecal appendicitis Myocardial ischemia Appendicitis in pregnancy RLL pneumonia RIGHT LOWER QUADRANT PAIN Appendicitis Meckel’s diverticulitis Cecal diverticulitis Aortic aneurysm Ectopic pregnancy Ovarian cyst Pelvic inflammatory disease Endometriosis Ureteral calculi Psoas abcess Mesenteric adenitis Incarcerated/strangulated hernia Ovarian torsion Tubo-ovarian abscess Urinary tract infection
Perforated duodenal ulcer or ruptured spleen
LEFT UPPER QUADRANT PAIN Gastritis Pancreatitis GERD Splenic pathology Myocardial ischemia Pericarditis Myocarditis LLL pneumonia Pleural effusion
LEFT LOWER QUADRANT PAIN Aortic aneurysm Sigmoid diverticulitis Incarcerated/strangulated hernia Ectopic pregnancy Ovarian torsion Mittelschmerz Ovarian cyst Pelvic inflammatory disease Endometriosis Tubo-ovarian abscess Ureteral calculi Psoas abscess Urinary tract infection
duration and time of colic may give clues to the identity of the culprit organ, as displayed in Figure 21-3. The severity and descriptive nature of the pain are the most subjective aspects of the pain history, but a few classical descriptions are recognized, such as the following: ■ ■
Acute pancreatitis or renal colic
Biliary colic Uterine or rectal pain
The diffuse, severe, colicky pain of bowel obstruction The “pain out of proportion to examination” observed in patients with mesenteric ischemia ■ The radiation of pain from the epigastrium straight through to the midback associated with pancreatitis, either related to primary organ inflammation or secondary to a penetrating ulcer ■ The radiation of pain to the left shoulder or independent pain in the left shoulder associated with splenic pathology, diaphragmatic irritation, or free intraperitoneal fluid ■ The onset of pain associated with syncope seen in perforation of gastric or duodenal ulcer, ruptured aortic aneurysm, or ruptured ectopic pregnancy
Physical Examination Figure 21-2. Common locations of referred pain from abdominal etiology.
The objective evaluation begins with measurement of the vital signs. Significant tachycardia and hypotension are indicators that hypovolemia or sepsis may be present. Tachypnea
Chapter 21 / Abdominal Pain
DIFFUSE PAIN Peritonitis Pancreatitis Sickle cell crisis Early appendicitis Mesenteric thrombosis Gastroenteritis Dissecting or ruptured aneurysm Intestinal obstruction Diabetes mellitis Inflammatory bowel disease Irritable bowel
EPIDEMIOLOGY
Occurs in females of childbearing age. No method of contraception prevents ectopic pregnancy. Approximately 1 in every 100 pregnancies.
Incidence increases with advancing age. More frequent in men. Risk factors include HTN, DM, smoking, COPD, and CAD.
Occurs most commonly in elders with CV disease, CHF, cardiac dysrhythmias, DM, sepsis, and dehydration. Responsible for 1 of 1000 hospital admissions. Mortality 70%. Mesenteric venous thrombosis associated with hypercoagulable states, hematologic inflammation, and trauma.
CAUSE
Ruptured ectopic pregnancy
Ruptured or leaking abdominal aneurysm
Mesenteric ischemia 20–30% of lesions are nonocclusive. The causes of ischemia are multifactorial, including transient hypotension in the presence of preexisting atherosclerotic lesion. The arterial occlusive causes (65%) are secondary to emboli (75%) or acute arterial thrombosis (25%).
Exact etiology is undetermined. Contributing factors include atherosclerosis, genetic predisposition, HTN, connective tissue disease, trauma, and infection.
Risk factors include nonwhite race, older age, history of STD or PID, infertility treatment, intrauterine contraceptive device placed within the past year, tubal sterilization, and previous ectopic pregnancy.
ETIOLOGY
Table 21-2 Potentially Life-threatening Causes of Abdominal Pain
Severe pain, colicky, that starts in periumbilical region and then becomes diffuse. Often associated with vomiting and diarrhea. Sometimes postprandial. “Mesenteric or abdominal angina.”
Severe, sharp constant pain localized to the affected side. More diffuse abdominal pain with intraperitoneal hemorrhage. Signs of shock may be present. Midline pain tends not to be ectopic pregnancy. Patient often asymptomatic until rupture. Acute epigastric and back pain often associated with or followed by syncope or signs of shock. Pain may radiate to back, groin, or testes.
PRESENTATION
Shock or evidence of peritonitis may be present. Lateralized abdominal tenderness. Localized adnexal tenderness or cervical motion tenderness increase the likelihood of ectopic pregnancy. Vaginal bleeding does not have to be present. Vital signs may be normal (in 70%) to severely hypotensive. Palpation of a pulsatile mass is usually possible in aneurysms 5 cm or greater. The physical examination may be nonspecific. Bruits or inequality of femoral pulses may be evident. Early examination results can be remarkably benign in the presence of severe ischemia. Bowel sounds often still present. Rectal examination important because mild bleeding with positive guaiac stools can be present.
PHYSICAL EXAMINATION
Often a pronounced leukocytosis is present. Elevations of amylase and creatine kinase levels are seen. Metabolic acidosis due to lactic acidemia is often seen with infarction. Plain radiographs of limited benefit. CT, MRI, and angiography are accurate to varying degrees.
β-hCG testing necessary in all females of childbearing age (10–55 yr); combined with ultrasonography, preferably transvaginal in early pregnancy, usually is diagnostic. FAST exam is useful in evaluating for free fluid in patients with shock or peritonitis. Abdominal plain films abnormal in 80% of cases. Ultrasound can define diameter and length but can be limited by obesity and bowel gas. FAST exam can be helpful in evaluating for leak by looking for free fluid. Spiral CT test of choice in stable patients.
USEFUL TOOL(S)
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
164
Peaks in infancy and older age. More common with history of previous abdominal surgery.
Incidence increases with advancing age. History of peptic ulcer disease or diverticular disease common.
Peak age in adulthood; rare in children and elders. Male preponderance. Alcohol abuse and biliary tract disease are risk factors.
Intestinal obstruction
Perforated viscus
Acute pancreatitis Alcohol, gallstones, hyperlipidemia, hypercalcemia, or endoscopic retrograde pancreatography causing pancreatic damage, saponification, and necrosis. ARDS, sepsis, hemorrhage, and renal failure are secondary.
More often a duodenal ulcer that erodes through the serosa. Colonic diverticula, large bowel, and gallbladder perforations are rare. Spillage of bowel contents causes peritonitis.
Adhesions, carcinoma, hernias, abscesses, volvulus, and infarction. Obstruction leads to vomiting, “third spacing” of fluid, or strangulation and necrosis of bowel.
ETIOLOGY
Acute onset of epigastric pain radiating to the back. Nausea and vomiting are common. Pain disproportionate to physical findings. Adequate volume repletion is important in the initial therapy.
Acute onset of epigastric pain is common. Vomiting in 50%. Fever may develop later. Pain may localize with omental walling off of peritonitis. Shock may be present with bleeding or sepsis.
Crampy diffuse abdominal pain associated with vomiting.
PRESENTATION
Vital signs usually normal unless dehydration or bowel strangulation has occurred. Abdominal distention, hyperactive bowel sounds, and diffuse tenderness. Local peritoneal signs indicate strangulation. Fever, usually of low grade, is common; worsens over time. Tachycardia is common. Abdominal examination reveals diffuse guarding and rebound. “Boardlike” abdomen in later stages. Bowel sounds are decreased. Low-grade fever common. Patient may be hypotensive or tachypneic. Some epigastric tenderness usually present. Because pancreas is retroperitoneal organ, guarding or rebound not present unless condition is severe. Flank ecchymosis or periumbilical ecchymosis may be seen if process is hemorrhagic.
PHYSICAL EXAMINATION
Lipase determination is test of choice. Ultrasound exam may show edema, pseudocyst, or biliary tract disease. CT scan may show abscesses, necrosis, hemorrhage, or pseudocysts. CT is ordered if severe acture pancreatitis is suspected. Rule out gallstones with ultrasound exam.
WBC count usually elevated due to peritonitis. Amylase may be elevated; LFT results are variable. Upright radiographic view reveals free air in 70–80% of cases with perforated ulcers.
Elevated WBC count suggests strangulation. Electrolytes may be abnormal if associated with vomiting or prolonged symptoms. Abdominal radiographs and CT are useful in diagnosis.
USEFUL TOOL(S)
Chapter 21 / Abdominal Pain
ARDS, acute respiratory distress syndrome; β-hCG, β human chorionic gonadotropin; CAD, coronary artery disease; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; CT, computed tomography; CV, cerebrovascular; DM, diabetes mellitus; FAST, focused assessment with sonography in trauma; HTN, hypertension; LFT, liver function test; MRI, magnetic resonance imaging; PID, pelvic inflammatory disease; STD, sexually transmitted disease; WBC, white blood cell.
EPIDEMIOLOGY
CAUSE
165
166
1. How old are you? Advanced age means increased risk. 2. Which came first—pain or vomiting? Pain first is worse (i.e., more likely to be caused by surgical disease). 3. How long have you had the pain? Pain for less than 48 hours is worse. 4. Have you ever had abdominal surgery? Consider obstruction in patients who report previous abdominal surgery. 5. Is the pain constant or intermittent? Constant pain is worse. 6. Have you ever had this before? A report of no prior episodes is worse. 7. Do you have a history of cancer, diverticulosis, pancreatitis, kidney failure, gallstones, or inflammatory bowel disease? All are suggestive of more serious disease. 8. Do you have human immunodeficiency virus (HIV)? Consider occult infection or drug-related pancreatitis. 9. How much alcohol do you drink per day? Consider pancreatitis, hepatitis, or cirrhosis in patients with history or signs of significant intake. 10. Are you pregnant? Test for pregnancy—consider ectopic pregnancy. 11. Are you taking antibiotics or steroids? Effects of these drugs may mask infection. 12. Did the pain start centrally and migrate to the right lower quadrant? High specificity for appendicitis. 13. Do you have a history of vascular or heart disease, hypertension, or atrial fibrillation? Consider mesenteric ischemia and abdominal aneurysm. From Colucciello SA, Lukens TW, Morgan DL: Abdominal pain: An evidence-based approach. Emerg Med Pract 1:2, 1999.
Pain scale
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
BOX 21-2 High-Yield Historical Questions
Time Ureteral colic Biliary colic Small intestinal colic Large intestinal colic
Figure 21-3. The characteristics of colicky abdominal pain. may be an indication of metabolic acidosis from gangrenous viscera or sepsis, hypoxemia from pneumonia, or simply a catecholamine-induced reaction to pain. Elevated temperature often is associated with intra-abdominal infections. Although important, vital signs may be misleading and should be inter-
preted in the context of the entire presentation. Tachycardia may develop late for various reasons in hypovolema. Temperature often is normal in elderly patients with laparotomy-proven intraperitoneal infections.9 Septic elderly patients also may present with hypothermia. A thorough abdominal examination is an essential part of evaluation of the patient with abdominal pain. This requires properly positioning the patient supine and exposing the abdomen. The examination should begin with inspection for any signs of trauma, bruising, or skin lesions. The patient should be asked to localize the area of maximal tenderness by pointing with one finger. The abdomen can be mentally divided into four quadrants: right upper, right lower, left upper, and left lower; each area is then examined individually. Tenderness in one quadrant often corresponds with the location of the diseased organ, which will direct the workup (see Fig. 21-1). Some disease processes may manifest with pain that is not exclusively within one specific quadrant, such as the suprapubic pain of a urinary tract infection or the midepigastic pain of a gastric ulcer. Although 80% of patients with suspected appendicitis present with right lower quadrant abdominal tenderness, 20% of patients with proven appendicitis do not.10 Rectal examination may have limited use in the evaluation of abdominal pain, except that associated with intraluminal gastrointestinal hemorrhage, prostatitis, or perirectal disease. The main utility of the rectal examination is in the detection of heme-positive stool, anal fissures or fistulas, or stool impaction. Rectal examination has not been shown to increase diagnostic accuracy for appendicitis when added to external physical examination of the abdomen.11 The abdominal evaluation should include a pelvic examination in female patients with lower abdominal pain or an otherwise uncertain diagnosis. The pelvic exam should be done early in the evaluation of the female patient with abdominal pain to help differentiate an abdominal from a pelvic source. This information is helpful in choosing an imaging modality. Pelvic ultrasound exam is helpful in evaluating uterine and ovarian pathology, whereas computed tomography (CT) is more beneficial in evaluation of suspected intra-abdominal pathology. Although the pelvic exam may guide the initial choice of imaging modality, overlap in exam findings is common. For example a patient with right lower quadrant tenderness may have both right adnexal tenderness and tenderness over McBurney’s point—necessitating exclusion of both appendicitis and ovarian torsion. The diagnosis highest on the differential list should be ruled out first using the corresponding imaging modality. In the male patient with abdominal pain, the urogenital system should be examined. Diseases such as prostatitis, orchitis, and epididymitis commonly cause abdominal pain in males. Furthermore, inguinal hernias are more common in males, with the possibility of strangulation or incarceration in the inguinal canal making a thorough genitourinary examination mandatory. In view of the evolving nature of abdominal pain, repetitive examinations may be useful. This is common practice with respect to suspected appendicitis and has improved the diagnostic accuracy in patients whose presentations were atypical.2
Ancillary Testing Urinalysis and testing for pregnancy are perhaps the most time- and cost-effective adjunctive laboratory tests available. Results often can be obtained quickly, so the former can lead to an early diagnosis and the latter may significantly affect
167
of CT studies will reduce morbidity, mortality, and medical expenses.18,19 CT has increased diagnostic utility in elderly patients for several reasons. Older people with abdominal pain may have twice the rate of surgery20-22 and a six- to eight-fold increase in mortality compared with younger adults.20,21 Furthermore, evaluation of abdominal pain in the elderly often is more challenging owing to unreliable findings on physical examination including vital signs, difficulties in history taking, physiologic age-related changes, and comorbid conditions. In the elderly population, CT results change management or disposition decisions in a significant proportion of patients.23 Table 21-3 lists the most common findings on CT scans in elderly patients with abdominal pain. Some controversy surrounds the use of oral contrast in abdominal CT in the critically ill ED patient. Technologic advances have improved image acquisition and resolution, and preliminary studies have shown that intravenous contrast alone may now be adequate in the evaluation of certain suspected pathologic processes, such as solid organ or bowel wall disease.24 CT with intravenous contrast alone also has been shown to be sensitive and specific for the confirmation or exclusion of acute appendicitis.25 The exclusion of oral contrast in these patients significantly decreases ED time to disposition and improves patient satisfaction. Bedside transabdominal and transvaginal ultrasonography have emerged as extremely useful adjuncts, decreasing time to diagnosis of life-threatening abdominopelvic conditions. Useful indications include the following: ■
Identification of an intrauterine pregnancy, effectively lowering the chances of an ectopic pregnancy to less than 1 in 20,000 (In women using fertility aids, however, identification of intrauterine pregnancy does not exclude ectopic pregnancy, in keeping with an increased incidence of heterotopic pregnancy.) ■ Measurement of the cross-sectional diameter of the abdominal aorta to determine whether an abdominal aortic aneurysm exists ■ Detection of free intraperitoneal fluid indicating hemorrhage, pus, or extrusion of gut contents ■ Use as a diagnostic aid for detection of the following non–life-threatening conditions: ■ Gallstones or a dilated common bile duct, which may be a clue to the presence of choledocholithiasis ■ Pericholecystic fluid or gallbladder wall thickening, which may be indicative of cholecystitis ■ Free intraperitoneal fluid indicating ascites ■ Hydronephrosis indicating possible obstructive uro pathy ■ Inferior vena cava distention or collapse as an indicator of volume status
Common Diagnostic Computed Tomography (CT) Findings in Older Patients Presenting to Table 21-3 Most the Emergency Department with Acute Abdominal Pain FINDING
PERCENT OF ABDOMINAL CT SCANS
Small bowel obstruction or ileus Diverticulitis Urolithiasis Cholelithiasis Abdominal mass/neoplasm Pyelonephritis Pancreatitis
18% 18% 10% 10% 8% 7% 6%
From Hustey FM, Meldon SW, et al: The use of abdominal computed tomography in older ED patients with acute abdominal pain. Am J Emerg Med 23:259–265, 2005.
Chapter 21 / Abdominal Pain
further evaluation and management approaches. It is necessary to interpret urinalysis results within the context of the patient’s clinical picture. Pyuria, with or without bacteriuria, often is present in a variety of conditions besides a simple urinary tract infection. For example, appendicitis may feature sterile pyuria.12 Similarly, hematuria usually is present with the relatively benign condition of nephrolithiasis but also may indicate an abdominal aortic aneurysm. Complete blood counts frequently are ordered for patients with abdominal pain, but findings seldom are contributory to a diagnosis. Despite the association of elevated white blood cell (WBC) counts with many infectious and inflammatory processes, the WBC count is neither sufficiently sensitive nor specific to be considered a discriminatory test to help establish or rule out a serious cause for the pain. Even serial WBC counts have failed to differentiate surgical from nonsurgical conditions. The WBC count is therefore not helpful for diagnosis. Serum electrolytes, even in the presence of protracted emesis or diarrhea, are abnormal in less than 1% of patients. These studies are not indicated for most patients in the absence of another indication. Blood urea nitrogen concentrations can be elevated in gastrointestinal hemorrhage and dehydration, but such conditions are better detected and quantified by history and physical examination. Increased serum creatinine usually is indicative of renal dysfunction. Blood glucose, anion gap, and serum ketone determinations are useful in diabetic ketoacidosis, one cause of acute abdominal pain and tachypnea. Liver enzymes and coagulation studies are helpful only in a small subset of patients with suspected liver disease.13 If pancreatitis is suspected, the most useful diagnostic result is serum lipase elevated to at least double the normal value, because it is more specific and more sensitive than serum amylase for this process. Measurement of serum amylase is of no value if a serum lipase level is available.14 Serum phosphate and serum lactate levels are elevated late in bowel ischemia, and such determinations may be useful if this entity is suspected but cannot be considered either sufficiently sensitive or specific to establish or exclude the diagnosis on their own. Plain radiography of the abdomen has limited usefulness in the evaluation of acute abdominal pain. Suspected bowel obstruction, foreign body, and perforated viscus are the main indications. CT of the abdomen has become the imaging modality of choice with nonobstetric abdominal pain. It allows visualization of both intraperitoneal and extraperitoneal structures and has a high degree of accuracy, establishing a diagnosis in more than 95% of cases in one study15 and increasing the confidence in diagnosis in another.16 Incidental findings are common on CT scans and may lead to a diagnosis. Patients who undergo CT have a change in diagnosis more often than those who do not.17 The proper execution and interpretation
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
168
The results of sonographic examinations are operatordependent, and misdiagnosis can occur because of failure to detect or identify pathology, incorrect identification of normal anatomy as pathologic, or overinterpretation of correctly identified findings (e.g., the mere presence of gallstones does not indicate that cholelithiasis is the cause of the pain). The emergency physician must be properly trained in image acquisition and interpretation, and ultrasound evaluation in the radiology department should be sought if there is ambiguity or uncertainty in findings.
■ DIFFERENTIAL DIAGNOSIS The differential considerations with abdominal pain include a significant number of potentially life- or organ-threatening entities, particularly in the setting of a hemodynamically unstable or toxic-appearing patient. Severely ill patients require timely resuscitation and expeditious evaluation for potentially life-threatening conditions. A focused history and exam should be performed, and the patient should be placed in a monitored acute care area well equipped for airway control, quick intravenous access, and fluid administration. Only then should appropriate diagnostics be initiated (bedside focused assessment with sonography in trauma [FAST] and aorta ultrasound assessment and radiographic, electrocardiographic, and laboratory studies). This approach is particularly important in dealing with elderly or potentially pregnant patients (see Tables 21-1 and 21-2). Women of reproductive age who present with abdominal pain should undergo pregnancy testing early, and a known pregnancy or a positive result on urine or serum pregnancy testing associated with abdominal pain in the ED should be considered to represent an ectopic pregnancy until proved otherwise. If evidence of blood loss is present, early obstetric consultation and diagnostic ultrasonography should be promptly sought. Bedside transabdominal sonography may identify free intraperitoneal fluid during the evaluation of shock, which may be sufficient evidence to justify operative intervention in the context of a positive pregnancy test and appropriate history and physical findings. Despite the limitations already described, the approach to the differential diagnosis of abdominal pain generally is based on the location of maximum tenderness. Figure 21-1 shows locations of subjective pain and maximal tenderness on palpation related to various underlying causes. In women of childbearing age, a positive result on pregnancy testing may indicate ectopic pregnancy, but the entire spectrum of intra-abdominal conditions remains in the differential diagnosis, as for the nonpregnant patient. When the very broad differential list is compartmentalized by both history and physical examination, ancillary testing should proceed to either confirm or support the clinical suspicion. Despite the significant variety of tests available, close to one half of the patients presenting to the ED with acute abdominal pain will have no conclusive diagnosis. It is incumbent on the clinician to reconsider the extra-abdominal causes of abdominal pain (see Box 21-1), with special consideration in elderly and immunocompromised patients, before arriving at the diagnosis of “nonspecific abdominal pain.”
■ EMPIRICAL MANAGEMENT The main therapeutic goals in managing acute abdominal pain are physiologic stabilization, mitigation of symptoms (e.g., emesis control, pain relief), and expeditious diagnosis, with consultation, if required.
There is no evidence to support withholding analgesics from patients with acute abdominal pain to preserve the accuracy of subsequent abdominal exams; in fact, the preponderance of evidence supports the opposite. Pain relief may facilitate the diagnosis in patients ultimately requiring surgery.26-28 In the acute setting, analgesia usually is accomplished with intravenously titrated opioids. Meperidine (Demerol) has an unfavorable side effect profile and should be avoided. Intravenous ketorolac, the only parenteral nonsteroidal anti-inflammatory drug available in North America, is useful for both ureteral and biliary colic,29,30 as well as some gynecologic conditions, but is not indicated for general treatment of undifferentiated abdominal pain. Among patients with gastrointestinal hemorrhage and potential surgical candidates, ketorolac has been shown to increase bleeding times in healthy volunteers.31 Aside from analgesics, a variety of other medications may be helpful to patients with abdominal pain. The burning pain caused by gastric acid may be relieved by antacids.32 Intestinal cramping may be diminished with oral anticholinergics such as the combination agent atropine-scopolamine-hyoscyaminephenobarbital (Donnatal), although evidence for this is scant and highly variable. Antiemetics such as promethazine, prochlorperazine, ondansetron, granisetron, or inapsine can be useful for nausea and vomiting. Gastric emptying by nasogastric tube with suction is appropriate for suspected small bowel obstruction and intractable pain or vomiting. If intra-abdominal infection is suspected, broad-spectrum antibiotic therapy should be initiated promptly. Abdominal infections are often polymicrobial and coverage for enteric gram-negative, gram-positive, and anaerobic bacteria must be included. In the choice of antibiotic or combination, the following should be considered: ■
Unless local antibiotic resistance surveillance indicates otherwise, second-generation cephalosporins (e.g., cefamandole, cefotetan, cefoxitin) or quinolone (ciprofloxacin, levofloxacin) may be combined with metronidazole for the initial dose of antibiotics in the ED. Other noncephalosporin, β-lactam agents with β-lactamase antagonists (e.g., ampicillin-sulbactam, piperacillin-tazobactam, ticarcillinclavulanate) are alternatives. ■ Many enteric gram-negative bacilli mutate rapidly to produce β-lactamases that are poorly antagonized by specific drug combinations containing clavulanate, sulbactam, or tazobactam. A carbapenem (e.g., imipenem, meropenem) or cefepime is an alternative for patients who may have recently received other antibiotics.10,33 Whether to provide coverage for Enterococcus species is still a subject of debate, and the decision to treat for these bacteria specifically can be made after consultation. Immunocompromised patients may require antifungal agents.
■ DISPOSITION Because up to 40% of patients presenting with acute abdominal pain receive the diagnosis of nonspecific abdominal pain, the disposition can be as difficult as the diagnosis in these patients. Categories for disposition may include surgical versus nonsurgical consultation and management, admission for observation, and discharge to home with follow-up evaluation.34 The decision to admit a patient to an observation unit or a hospital bed must factor in the following:
■
Clinically stable patients may be discharged from the ED with appropriate follow-up care, possibly to include repeated physical exam or additional diagnostic imaging if indicated. In the case of nonspecific abdominal pain that is considered potentially worrisome, it is prudent to have the patient reevaluated after 8 to 12 hours. This can be done through a return visit to the ED, an appointment with a primary care physician, or an observation unit protocol. Before discharge of a patient with an undiagnosed cause of nonspecific abdominal pain, several conditions should be met: The abdominal examination findings should be benign overall, with normal vital signs. Pain and nausea should be controlled,
169
and the patient should be able to eat and drink. If a patient is to be discharged home without a specific diagnosis, clear instructions to the patient must include the following information: ■
What the patient has to do for relief of symptoms or to maximize chances of resolution of the condition (e.g., avoiding exacerbating food or activities, taking medications as prescribed) ■ Under what circumstances, with whom, and in what time frame to seek follow-up evaluation, if all goes as desired on the basis of what is known when the patient is in the ED ■ Under what conditions the patient should seek more urgent care because of unexpected changes in his or her condition (such as with natural progression of the process before improvement, incorrect diagnosis made in the ED, or untoward reactions to medications) The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 21 / Abdominal Pain
Information gained from the history, physical examination, and test results ■ The likelihood of any suspected disease ■ Any potential ramifications of progression of a known disease, or of incorrect diagnosis or management ■ The likelihood of appropriate (or any) and timely follow-up after hospital discharge
Chapter 22
Gastrointestinal Bleeding
Philip L. Henneman
■ PERSPECTIVE Epidemiology Gastrointestinal (GI) bleeding is a relatively common problem encountered in emergency medicine that often requires early consultation and hospital admission. The overall mortality rate for GI bleeding is approximately 10% and has not changed significantly since the 1960s. Diagnostic modalities have improved much more than therapeutic techniques. GI bleeding is often easy to identify when there is clear evidence of vomiting blood or passing blood in the stool, but the clinical presentation may be subtle, with signs and symptoms of hypovolemia, such as dizziness, weakness, or syncope. The approach to GI bleeding depends on whether the hemorrhage is located in the proximal or the distal segment of the GI tract (i.e., upper or lower GI bleeding). These segments are anatomically defined by the ligament of Treitz in the fourth section of the duodenum. In the United States, upper GI bleeding (UGIB) affects 50 to 150 people per 100,000 population each year and results in more than 300,000 admissions and about 30,000 deaths per year at an estimated annual cost of almost $1 billion. Lower GI bleeding (LGIB) affects a smaller portion of patients and results in proportionally fewer hospital admissions than UGIB.1 GI bleeding can occur in persons of any age but most commonly affects people in their 40s through 70s (mean age, 59 years). Most deaths caused by GI bleeding occur in patients older than 60 years. UGIB is more common in men than in women (in a 2 : 1 ratio), whereas LGIB is more common in women. Significant UGIB requiring admission is more common in adults, whereas LGIB requiring admission is more common in children.2
■ DIAGNOSTIC APPROACH
LGIB. Despite improved diagnostic techniques, no source of bleeding is identified in approximately 10% of patients with GI bleeding. In patients with abdominal aortic grafts who present to the emergency department (ED) with GI bleeding, the possibility of aortoenteric fistula should be considered. Prompt surgical consultation in the ED should be obtained if this is suspected, because bleeding can be massive and fatal.
Rapid Assessment and Stabilization Most patients with GI bleeding are easy to diagnose because they present to the ED complaining of vomiting blood or passing black or bloody stool. The diagnosis is confirmed quickly by examination of the stool for the presence of blood. Patients with suspected GI bleeding who are hemodynamically unstable should undergo rapid evaluation and resuscitation. They should be undressed quickly to permit placement of cardiac and oxygen saturation monitors, and supplemental oxygen should be given as needed. At least two large-bore peripheral intravenous lines should be placed (minimum 18gauge); blood should be drawn for hemoglobin or hematocrit, platelet count, prothrombin time (PT), and type and screen or type and crossmatch studies; and crystalloid resuscitation should be initiated. Intravenous crystalloid fluid should be given as a 2-L bolus in adults or 20 mL/kg in children until the patient’s vital signs have stabilized or the patient has received 40 mL/kg of crystalloid in an adult or 60 mL/kg as a child. Patients who remain unstable after 40 to 60 mL/kg of crystalloid should be given type O, type-specific, or crossmatched blood, depending on availability. Persistently unstable patients should receive immediate consultation with a gastroenterologist for UGIB and with a surgeon for LGIB.3
Differential Considerations
Pivotal Findings
Peptic ulcer disease, gastric erosions, and varices account for approximately three fourths of adult patients with UGIB (Box 22-1). Diverticulosis and angiodysplasia account for approximately 80% of adults with LGIB. In children, esophagitis, gastritis, and peptic ulcer disease are the most common causes of UGIB, and infectious colitis and inflammatory bowel disease are the most common causes of LGIB (Box 22-2). In children younger than 2 years of age, massive LGIB is most often a result of Meckel’s diverticulum or intussusception. At all ages, anorectal abnormalities are the most common cause of minor
History, physical examination, testing a stool sample for blood, and measuring hemoglobin or hematocrit are the keys to diagnosing GI bleeding in most patients.
170
History Patients typically complain of vomiting red blood or coffee grounds–like material, or passing black or bloody stool. Hematemesis (vomiting blood) occurs with bleeding of the esophagus, stomach, or proximal small bowel. Approximately
171
Etiology of Significant Gastrointestinal (GI)
Upper Peptic ulcer disease Gastric erosions Varices Mallory-Weiss tear Esophagitis Duodenitis
Lower Diverticulosis Angiodysplasia Upper GI bleeding Cancer/polyps Rectal disease Inflammatory bowel disease
*Potential causes listed in decreasing frequency.
BOX 22-2
Etiology of Gastrointestinal Bleeding in Children*
Upper Esophagitis Gastritis Ulcer Esophageal varices Mallory-Weiss tear
Lower Anal fissure Infectious colitis Inflammatory bowel Polyps Intussusception
*Potential causes listed in decreasing frequency.
50% of patients with UGIB present with this complaint. Hematemesis may be bright red or darker (i.e., coffee grounds– like) as a result of the conversion of hemoglobin to hematin or other pigments by hydrochloric acid in the stomach. The color of vomited or aspirated blood from the stomach does not differentiate between arterial and venous bleeding. Melena, or black tarry stool, will result from the presence of approximately 150 to 200 mL of blood in the GI tract for a prolonged period. Melena is seen in approximately 70% of patients with UGIB and in one third of patients with LGIB. Black stool that is not tarlike may result from presence of 60 mL of blood from the upper GI tract. Blood from the duodenum or jejunum must remain in the GI tract for approximately 8 hours before turning black. Occasionally, black stool may follow bleeding into the lower portion of the small bowel and ascending colon. Stool may remain black and tarry for several days, even though bleeding has stopped. Black stool also may be seen after ingestion of bismuth (e.g., PeptoBismol), which can confuse the situation because such preparations often are taken for UGI distress. In contrast with melena, stool rendered black by bismuth is not positive on Hemoccult testing. Hematochezia, or bloody stool (bright red or maroon), most often signifies LGIB but may be due to a brisk UGIB with rapid transit time through the bowel in 10 to 15% of patients. Because UGIB is much more common than LGIB, a more proximal source of significant bleeding must be excluded before assuming the bleeding is from the lower GI tract. Approximately two thirds of patients with LGIB present with red blood from bleeding per rectum. Small amounts of red blood (e.g., 5 mL) from rectal bleeding, such as bleeding due to hemorrhoids, may cause the water in the toilet bowl to appear bright red. Bright red stools also can be seen after ingestion of a large quantity of beets; in this case, Hemoccult testing would be negative and the patient also will report pinkcolored water in the toilet bowl. In taking the history, specific questions should address the duration and quantity of bleeding, associated symptoms, previous history of bleeding, current medications, alcohol, nonsteroidal anti-inflammatory drug use and long-term aspirin
Physical Examination Vital Signs Vital signs and postural changes in heart rate and blood pressure have been used to assess the amount of blood loss in patients with GI bleeding but are insensitive and nonspecific, with the exception of significant, sustained heart rate increase and hypotension. All patients with a history suggesting GI bleeding who are hypotensive, are tachycardic, or experience sustained posture-induced changes in heart rate of greater than 20 beats per minute should be assumed to have a significant hemorrhage. Normal vital signs do not exclude a significant hemorrhage, and postural changes in heart rate and blood pressure may occur in individuals who are not bleeding (e.g., elderly patients, many normal individuals, individuals on certain medications such as beta-blockers, individuals with hypovolemia from other causes). General Examination The physical examination is valuable in establishing a specific diagnosis and assessing the severity of blood loss and the physiologic response to that loss. Careful attention is given to the patient’s general appearance, vital signs, mental status (including restlessness), skin signs (e.g., color, warmth, and moisture to assess for shock, or presence of lesions such as telangiectasia, bruises, or petechiae to assess for vascular diseases or hypocoagulable states), pulmonary and cardiac findings, abdominal examination, and rectal and stool examination. Frequent reassessment is important because a patient’s status may change quickly. Rectal Examination Rectal and stool examinations are often key to making or confirming the diagnosis of GI bleeding. The finding of red, black, or melenic stool early in the assessment is helpful in prompting early recognition and management of patients with GI bleeding. The absence of black or bloody stool, however, does not exclude the diagnosis of GI bleeding. Regardless of the apparent character and color of the stool, occult blood testing is indicated.
Ancillary Testing Tests for Occult Blood The presence of hemoglobin in occult amounts in stool is confirmed by tests such as guaiac assays (e.g., Hemoccult, HemaPrompt). Stool tests for occult blood may have positive results 14 days after a single, major episode of UGIB. False-positive results have been associated with the ingestion of certain fruits (e.g., cantaloupe, grapefruit, figs), uncooked vegetables (e.g., radish, cauliflower, broccoli) and red meat, methylene blue, chlorophyll, iodide, cupric sulfate, and bromide preparations. False-negative results are uncommon but can be caused by bile or ingestion of magnesiumcontaining antacids or ascorbic acid. Tests to evaluate gastric contents for occult blood (e.g., Gastroccult) can be unreliable
Chapter 22 / Gastrointestinal Bleeding
BOX 22-1 Bleeding in Adults*
ingestion, allergies, associated medical illnesses, previous surgery, treatment by nonhospital personnel, and the response to that treatment.4,5 Patients with GI bleeding may report symptoms of hypovolemia, such as dizziness, weakness, or loss of consciousness, most often after standing up. Other nonspecific complaints include dyspnea, confusion, and abdominal pain. Rarely an elderly patient may present with ischemic chest pain precipitated by significant anemia due to a GI bleed. One in five patients with GI bleeding may have only nonspecific complaints. The history is of limited help in predicting the site or quantity of bleeding. Patients with a previously documented GI lesion bleed from the same site in only 60% of cases. Gross estimates of blood loss based on the volume and color of the vomitus or stool (e.g., brown or black, pink or red) or the number of episodes of hemorrhage are notoriously inaccurate.
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
172
and should not be used for this purpose. In newborns, maternal blood that is swallowed may cause bloody stools; the Apt test may show that it is maternal in origin. Clinical Laboratory Tests Blood should be drawn for evaluation of baseline hematocrit or hemoglobin, coagulation studies (PT and platelet count), and type and crossmatch studies (or type and screen studies if the patient is stable). The initial hematocrit may be misleading in patients with preexisting anemia or polycythemia. Changes in the hematocrit may lag significantly behind actual blood loss. Infusion of normal saline speeds equilibration of the hematocrit; however, rapid infusion of crystalloid in nonbleeding patients also may cause a decrease in hematocrit by hemodilution. The optimal hematocrit with respect to oxygen-carrying capacity and viscosity in critically ill patients has been reported to be 33%. In general, patients with a hemoglobin concentration of 8 g/dL or less (hematocrit 60 yr Initial SBP < 100 mm Hg Mild ongoing tachycardia for 1 hr Transfusions required, ≤4 units Stable major comorbid diseases Mild liver disease—PT normal or near-normal No high-risk clinical features
Persistent SBP < 100 mm Hg Persistent moderate/severe tachycardia Transfusion required, >4 units Unstable major comorbid diseases Decompensated liver disease—e.g., coagulopathy, ascites, encephalopathy
No moderate-risk or high-risk clinical features PT, prothrombin time; SBP, systolic blood pressure.
Data from Terdiman JP, Lindenauer GF: Acute gastrointestinal bleeding. In Wachter RM, Goldman L, Hollander H (eds): Hospital Medicine. Philadelphia, Lippincott Williams & Wilkins, 2005, pp. 767–779.
Table 22-2 Management by Risk Category for Patients with Upper Gastrointestinal Bleeding after Endoscopy Recommended Management RISK STRATIFICATION
LOW RISK
MODERATE RISK
HIGH RISK
Low risk
Immediate discharge*
23-hr observation (floor)†
Moderate risk
48-hr inpatient stay†
48- to 72-hr inpatient stay (floor)†
High risk
ICU monitoring for 48 hr (48- to 72-hr hospitalization)
ICU monitoring for 24 to 48 hr (72-hr hospitalization)
ICU monitoring for 24 hr‡ (48- to 72-hr hospitalization) ICU monitoring for 24 hr (48- to 72-hr hospitalization) ICU monitoring for 72 hr (≥72-hr hospitalization)
*Patients with low-risk clinical and endoscopic findings can be discharged home with appropriate treatment based on diagnosis, scheduled follow-up evaluation within 24 hours, and proper patient education to ensure immediate return if signs of rebleeding appear. † Patients may be discharged after 24 to 48 hours of in-hospital observation if there is no evidence of rebleeding, vital signs are normal, there is no need for further transfusion, and the hemoglobin or hematocrit has remained stable. They should be provided with appropriate treatment based on diagnosis, scheduled follow-up evaluation within 24 hours, and proper patient education to ensure immediate return if signs of rebleeding appear. ‡ Patients with high-risk clinical or endoscopic findings should be hospitalized and monitored closely for evidence of rebleeding. Data from Terdiman JP, Lindenauer GF: Acute gastrointestinal bleeding. In Wachter RM, Goldman L, Hollander H (eds): Hospital Medicine. Philadelphia, Lippincott Williams & Wilkins, 2005, pp. 767–779.
low-risk patients with GI bleeding. Studies have shown that combining clinical and endoscopic criteria provides an accurate estimation of the risk of rebleeding and mortality in patients with UGIB. These combined criteria have been used to identify patients with UGIB at low risk, who can be discharged home, and patients at moderate or high risk, who need to be admitted to an appropriate care site in the hospital. Risk stratification for patients with LGIB is less well studied, so nearly all patients with significant LGIB are admitted. Risk stratification can be used for patients with LGIB, however, to decide an appropriate inpatient care site. Table 22-1 presents an initial risk stratification tool for patients with upper and lower GI bleeding. Combining clinical and endoscopic findings allows for final risk stratification, as shown in Table 22-2, to decide disposition, inpatient care site, and treatment.1,2,6,20,23-25 Patients with clinical evidence of GI bleeding should undergo endoscopy as soon as it is available for final risk stratification, inpatient triage, and determination of appropriate treatment (see Table 22-2). If endoscopy is not immediately available, patients with low clinical risk may be admitted to an ED observation unit or short-stay hospital bed until endoscopy can be performed. Patients with moderate clinical risk criteria may be admitted to an inpatient floor, intermediate care unit, or ICU, as indicated by specific patient management needs and depending on the capabilities of the institution. Patients with high clinical risk should be admitted to a closely monitored step-down unit or an ICU. The timing of endoscopy depends on availability, the acuity of the patient, the need for
emergent therapy, the need to determine final care site, and the need to minimize length of stay.3,26 Patients with LGIB that is not clearly due to hemorrhoids, fissure, or proctitis should be admitted to an inpatient bed. Patients with low risk may be admitted to an inpatient floor bed and prepared for a nuclear medicine imaging study (e.g., red blood cell–labeled study) or colonoscopy. Patients with high-risk criteria should be admitted to a step-down unit or ICU and considered for angiography to identify the site of LGIB. Patients with moderate-risk criteria require individualized determination of the most appropriate inpatient care site (floor, intermediate care bed, or ICU) and the most useful diagnostic studies (nuclear imaging or angiography). Consultation with a surgeon should be obtained if it appears that more than 5 units of blood is required to achieve hemodynamic stability or if there is reasonable suspicion that operative intervention may be needed. This is especially true of patients older than 65 years of age. In general, the older the patient, the more aggressive the surgical management ought to be. Patients with a history of varices, persistent postural changes in heart rate, or significant bleeding of bright red blood per rectum are more likely to require surgery than are patients without these findings. Emergent vascular surgical consultation is needed for patients who have abdominal aortic grafts who present to the ED with GI bleeding, because of the possibility of an aortoenteric fistula. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 22 / Gastrointestinal Bleeding
LOW RISK
Chapter 23
Diarrhea
Robert E. Collier, John E. Gough, and Philip A. Clement
■ PERSPECTIVE Diarrhea is a common presenting complaint in the emergency department (ED) and can account for up to 5% of all visits. Most cases of diarrhea are self-limited and require only supportive care. Conversely, patients with more serious infection and associated comorbidity may present with life-threatening dehydration and shock. There may be an associated sepsis and septic shock component. Numerous but relatively rare, noninfectious causes of diarrhea should also be considered.
Incidence Worldwide, diarrhea remains a major health problem, accounting for approximately 4% of all deaths each year, which is estimated by the World Health Organization (WHO) to be 2.2 million victims (World Health Report, 2000).1 A large proportion of these deaths occur in small children in developing countries. Rotavirus causes 25% to 65% of childhoodassociated diarrheal illnesses (3.5 million cases per year in the United States), whereas adults experience 74 million episodes of diarrhea annually. In the United States, 90% of diarrheal illnesses are caused by noroviruses (caliciviruses), of which more than 100 different strains are recognized.2 Patients at the extremes of age, those with significant comorbidity, those who are immunologically compromised, and those with iatrogenic illness are most vulnerable to significant morbidity and mortality. An estimated 60% of patients infected with human immunodeficiency virus (HIV) experience significant diarrhea during the course of their illness.3 Most adults experience diarrhea many times during their lifetime. Diarrhea illnesses are the primary cause of many hospitalizations and hours of lost work.
Definition and Categorization The term diarrhea is derived from the Greek words dia (“through”) and rhein (“to flow”). The two main categories of diarrhea-associated illness are infectious and noninfectious. Infectious causes represent about 85% of cases, whereas noninfectious causes account for only 15% of the total. Infectious diarrhea may be divided into viral, bacterial, and parasitic causes (Box 23-1), with estimates of their relative contributions being 70% for viral, 24% for bacterial, and 6% for parasitic infections.4 Definitions for diarrhea have been proposed to standardize nomenclature, help the clinician determine a probable etiology, and direct empirical therapy if indicated5: 176
■ ■ ■
Acute diarrhea is defined as lasting for 14 days or less. Persistent diarrhea lasts for longer than 14 days. Chronic diarrhea lasts 30 days or longer.
Acute diarrhea presentations usually will be infectious. A majority of these cases are self-limited and caused by viral and bacterial pathogens. Persistent diarrhea suggests an enteric pathogen other than viral, such as bacterial or protozoan. Chronic diarrhea usually is associated with noninfectious causes and requires further testing to determine the etiology. Normally, the small and large bowel absorb 99% of gastrointestinal tract secretions produced and liquids ingested each day. Any pathologic state that reduces water absorption by 1% can cause diarrhea.6 Diarrhea results from one or more of four different pathologic processes that are characteristic of the primary cause and that contribute to the decreased absorption of the gut. Secretory diarrhea is caused by pathogens that produce cytotoxins that increase cellular permeability and cause the oversecretion of water and electrolytes. Most cases of diarrhea encountered in the ED are secretory. Noninfectious causes of secretory diarrhea include medications, toxic substances, endocrine disorders, and neoplasias (Box 23-2). Inflammatory diarrhea, also described as invasive or severe diarrhea, or dysentery, results from cellular damage to the intestinal mucosa, leading to the hypersecretion of water, electrolytes, blood, mucus, and plasma proteins. This diarrhea most commonly is caused by invasive bacterial and parasitic pathogens that produce dysenteric illnesses (see Box 23-1). Some noninfectious causes of inflammatory diarrhea include chemotherapy, radiation therapy, hypersensitivity reactions, autoimmune disorders, ischemic colitis, and inflammatory bowel disease. With inflammatory diarrhea, fecal leukocytes and erythrocytes typically are present, as are systemic symptoms, and the diarrhea continues despite fasting. Osmotic diarrhea occurs with the ingestion or malabsorption of osmotically active solutes. These solutes cause the osmotic movement of water into the intestinal lumen, which then overwhelms the gut’s ability to reabsorb it. Examples include the effects of osmotic laxatives and carbohydrate malabsorption. Steatorrhea results from osmotic effects of lipids not absorbed in malabsorption and maldigestion syndromes. Abnormal motility generally is seen in patients with chronic diarrhea but is always a component of acute diarrhea. Hypermotility decreases contact time between luminal contents and the absorbing mucosa, limiting water and electrolyte absorption.
177
BOX 23-1 Etiologic Agents of Infectious Diarrhea
Bacterial (20% of cases) Invasive* Aeromonas spp. Campylobacter spp. Clostridium difficile Enteroinvasive E. coli Mycobacterium spp. Plesiomonas shigelloides Salmonella spp. Shigella spp. Vibrio fluvialis Vibrio parahaemolyticus Vibrio vulnificus Yersinia enterocolitica Yersinia pseudotuberculosis Toxigenic food poisoning with preformed toxins
Bacillus cereus Clostridium botulinum Staphylococcus aureus
Enterohemorrhagic E. coli* O157:H7 Enterotoxigenic E. coli Klebsiella pneumoniae Shigella spp. Vibrio cholerae Other bacteria Parasitic (5% of cases) Protozoa Balantidium coli* Blastocystis hominis Cryptosporidium Cyclospora Dientamoeba fragilis Entamoeba histolytica* Entamoeba polecki Enteromonas hominis Giardia lamblia Isospora belli Microsporidia Sarcocystis hominis Helminths Angiostrongylus costaricense Anisakiasis Ascaris lumbricoides Diphyllobothrium latum Enterobius vermicularis Hookworms Schistosoma spp. Strongyloides stercoralis Taenia spp. Trichinella spiralis Trichuris trichiura
toxin formation after colonization
Aeromonas hydrophila Clostridium perfringens *Associated with fever, abdominal pain, and fecal red blood cells or white blood cells. % indicates the estimated contribution to total cases.
■ CLINICAL APPROACH
Secondary Survey
Emergency Assessment and Stabilization
The physical examination should assess the patient’s overall health, toxicity, fever, volume status, signs of a surgical abdomen, and determine the presence of blood in the stool. Young healthy adults may maintain a normal blood pressure and heart rate even with significant dehydration. In patients who are taking antiarrhythmic or beta-blocker medications or have conduction disease or fixed-pace rhythms, heart rate may not be a reliable indicator of volume status. Signs of volume depletion and impending shock include dry mucosa, poor skin turgor, decreased urine output, and mental status changes. Children will present with sunken eyes, depression of the fontanel, decrease in urine output (number of wet diapers), and decrease in alertness and activity.7 Particular attention should be given to the abdominal examination. Focal abdominal pain with peritoneal findings may be due to an acute surgical abdomen with symptoms mimicking those of severe gastroenteritis. A rectal examination should be performed to detect fecal impaction, melena, or hematochezia.
An immediate assessment should be made of the patient’s stability, including maintenance of the airway, adequacy of oxygenation and ventilation, and circulation, with particular attention to volume status. Tachycardia, orthostatic hypotension, poor skin turgor and color, diaphoresis, and mental status changes all are characteristic of hypovolemia and hypoperfusion. Associated septic shock may contribute to the hypotension and general organ hypoperfusion, and diarrhea may be a manifestation of toxic shock syndrome. A diarrhea-associated acid-base disorder should be suspected in patients with Kussmaul respirations, a significant anion gap on basic metabolic panel reflecting a lactic acidosis from significant volume loss, or a non–anion gap metabolic acidosis associated with massive bicarbonate loss. After stabilization, a secondary survey may elucidate the potential cause of the diarrhea and direct further evaluation and treatment.
Chapter 23 / Diarrhea
Viral (60% of cases) Astrovirus Calicivirus Coronavirus Cytomegalovirus* Enteric adenovirus Hepatitis A through G Herpes simplex virus HIV enteropathy Norwalk-like agents Norwalk virus Pararotavirus Picornavirus Rotavirus Small round viruses
178
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
BOX 23-2 Causes of Noninfectious Diarrhea Toxins Drugs ACE inhibitors Alprazolam (Xanax) Antacids (Mg) Antibiotics Antidepressants Antiepileptic drugs Antihypertensives Antiparkinson drugs Beta-blockers Caffeine Cardiac antiarrhythmics Chemotherapy agents Cholesterol-lowering drugs Cholinergic agents Cholinesterase inhibitors Colchicine Digitalis Diuretics Fluorouracil Fluoxetine (Prozac) Histamine H2-receptor antagonists Hydralazine Lactulose Laxatives/cathartics Levodopa Lithium NSAIDs Neomycin Podophyllin Procainamide Prostaglandins Quinidine Ricinoleic acid Theophylline Thyroid hormone Valproic acid Dietetic Foods Mannitol Sorbitol Xylitol Fish-Associated Toxins Amnestic shellfish poisoning Ciguatera Echinoderms
Neurotoxic shellfish poisoning Paralytic shellfish poisoning Scombroid Tetroton Plant-Associated Toxins Herbal preparations Horse chestnut Mushrooms—Amanita spp. Nicotine Other plant toxins Pesticides—organophosphates Pokeweed Rhubarb Miscellaneous Allergic reactions Carbon monoxide poisoning Ethanol Heavy metals Monosodium glutamate (MSG) Opiate withdrawal Gastrointestinal Pathology Appendicitis Autonomic dysfunction Bile acid malabsorption Blind loop Bowel obstruction Celiac disease Cirrhosis Defects in amino acid transport Diverticular disease Familial dysautonomia Fecal impaction Fecal incontinence GI bleed GI cancer Hirschsprung’s disease Inflammatory bowel disease (ulcerative colitis, Crohn’s disease) Intussusception Irritable bowel syndrome Ischemic bowel Lactose/fructose intolerance Malabsorption syndromes Malrotation Postsurgical Postvagotomy
Radiation therapy Short gut syndrome Small bowel resection Strictures Toxic megacolon Tropical sprue Volvulus Whipple’s disease Endocrine-Related Carcinoid syndrome (serotonin) Hormonal hypersecretion Hyperthyroidism (thyroid hormone) Medullary carcinoma of the thyroid (calcitonin) Pancreatic cholera (VIP) Somatostatinoma (somatostatin) Systemic mastocytosis (histamine) Zollinger-Ellison syndrome (gastrin) Endocrine Pathology Adrenal insufficiency Diabetes enteropathy Hypoparathyroidism Pancreatic insufficiency Systemic Illness/Other Alcoholism Amyloidosis Connective tissue disease Cystic fibrosis Ectopic pregnancy Hemolytic-uremic syndrome Henoch-Schönlein purpura Lymphoma Otitis media—infants Pelvic inflammatory disease Pneumonia/sepsis Pyelonephritis Scleroderma/SLE Severe malnutrition Stevens-Johnson syndrome Toxic shock syndrome Wilson’s disease Miscellaneous Factitious diarrhea Runner’s diarrhea
ACE, angiotensin-converting enzyme; GI, gastrointestinal; NSAIDs, nonsteroidal anti-inflammatory drugs; SLE, systemic lupus erythematosus; VIP, vasoactive intestinal polypeptide.
Gross blood may be consistent with invasive, infectious diarrhea but may be the harbinger of many other pathologic states that manifest with gastrointestinal bleeding. Histamineinduced skin changes may be indicative of an intestinal parasitic infection. The patient should be assessed for specific toxidromes, such as cholinergic or sympathomimetic states that may be clues to a noninfectious cause.
Characterization of the Diarrheal Syndrome Acute Infectious Diarrhea Most viral and many bacterial agents cause a self-limited, secretory diarrhea that lasts less than 14 days and causes only mild dehydration and minimal systemic symp-
toms. These infections do not require extensive testing and are treated symptomatically. In the United States, monitoring of pathogens causing this type of acute gastroenteritis demonstrates that 90% of the infections are caused by norovirus species.8 All other potential causes of diarrhea are highly improbable unless certain historical and clinical findings are present. Bacterial and protozoan agents less commonly cause diarrhea syndromes indistinguishable from norovirus infection with a nontoxic, self-limited course. A Bayesian approach to diagnosing and treating acute diarrhea has been proposed.8 The clinical evaluation should screen for all factors (Table 23-1) that may change the probability of “not norovirus” from 10% to 50% or greater.8 With one or more of these findings
179
Table 23-1 Factors Increasing Probability of Nonbenign Diarrhea SPECIFIC PATHOGEN(S)/OTHER CONSIDERATIONS
Presentation to a health care facility
Degree of illness overall greater in patients presenting for evaluation; increased probability of “not norovirus” etiology to 50% Especially foreign travel and to endemic areas of dysenteric disease C. difficile from antibiotic exposure Rotavirus, Shigella, Giardia C. difficile, medication side effects, tube feedings, ischemic colitis, fecal impaction, and overflow diarrhea Giardia or Cryptosporidium C. difficile, antibiotic side effects Salmonella spp., E. coli O157:H7 and non-O157 Shiga toxin–producing E. coli, Vibrio spp. Norovirus; less commonly, Campylobacter jejuni, Salmonella spp., Cryptosporidium Bacillus cereus, Clostridium botulinum, Staphylococcus aureus
Travel history Recent hospitalization Day care attendance Nursing home residence Wilderness exposure Antibiotic therapy Raw shellfish, farm animals and fair livestock, pet reptiles or amphibians, petting zoos Epidemic of multiple patients with a short time of onset Acute vomiting and diarrhea after suspected contaminated food Epidemic of severe gastroenteritis traced to eggs, poultry, meat, or dairy products Homosexual lifestyle (males) Abdominal pain Nausea, vomiting Bloody stool Fever Rectal pain Tenesmus Diarrhea >7–14 days’ duration Hemolytic uremic syndrome Stool WBC count Colonic ulcerations Proctitis Pseudomembranes Chronic disease (e.g., cirrhosis, DM) Organ transplantation
HIV infection, other immunodeficiency disorders
Campylobacter jejuni, Salmonella spp. Giardia lamblia, Entamoeba hystolytica Severe bacterial infections: Salmonella, Campylobacter, Shigella, EPEC, Yersinia or Vibrio spp. Also consider surgical abdomen, GI bleeding Inflammatory bowel disease Protozoa and microsporidia, Clostridium difficile, Campylobacter, Shiga toxin– producing E. coli E. coli O157:H7 or other species Not reliable for diagnosis of bacterial etiology Inflammatory bowel disease Bacterial etiology highly probable Toxic megacolon, Clostridium difficile Complicated course expected with any form of diarrheal illness Abnormally severe illness from rotavirus and adenovirus Increased frequency of cytomegalovirus Severe illness from dysenteric diarrhea Spore-forming protozoa and microsporidia Severe illness from common bacteria/spore-forming protozoa and microsporidia Increased frequency of cytomegalovirus and Mycobacterium avium complex
DM, diabetes mellitus; EPEC, enteropathogenic E. coli; WBC, white blood cell.
present, empirical antibiotic or other specific therapy may be indicated, as well as clinical testing to determine the exact etiologic disorder. Chronic Infectious Diarrhea Persistent diarrhea is defined as that lasting for more than 14 days, and chronic diarrhea, more than 30 days.5 Infectious agents of persistent and chronic diarrhea include bacteria, parasites, and rarely viruses.9 Common bacterial pathogens include Aeromonas, Plesiomonas, Campylobacter, Clostridium difficile, Salmonella, and Mycobacterium tuberculosis. Parasites causing chronic diarrhea are colonic forms such as Amoeba, Trichuris, Yersinia, and Schistosoma species or small intestinal pathogens such as Giardia, Cryptosporidium, Cyclospora, Isospora, and Strongyloides.9 In developing countries, chronic diarrhea is more likely to have a bacterial cause. In developed countries, chronic diarrhea is caused by noninfectious disorders such as irritable bowel syndrome, malabsorption syndromes, laxative abuse, and inflammatory bowel disease.6 Categorization of the stool type as watery, inflammatory, or fatty may assist in proper classification of the chronic
diarrhea syndrome as infectious and noninfectious. Testing for HIV or immune deficiency is important because these patients commonly present with chronic diarrhea. Evaluation should include testing for cryptosporidia, microsporidia, mycobacteria (i.e., Mycobacterium avium complex), herpes simplex virus (HSV), Isospora, Cyclospora, and cytomegalovirus. In addition, parasitic and helminthic infections should be ruled out. Noninfectious Diarrhea Noninfectious causes of diarrhea (see Box 23-2) are responsible for approximately 15% of all cases of diarrheal illness. The distinction between infectious and noninfectious causes may not be clinically apparent. A complete evaluation should consider possible surgical pathology of the abdomen, including gastrointestinal bleeding, ischemic bowel, acute appendicitis, intussusception, ectopic pregnancy, and partial bowel obstruction.7 The differential diagnosis also includes possible toxic exposures or ingestions, such as heavy metal poisoning, or ingestion of plant-borne or fish-borne toxins. Endocrine pathology, such as adrenal insufficiency, hyperthyroidism, diabetic enteropathy, and
Chapter 23 / Diarrhea
FACTOR
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
180
hormone-secreting tumors, and other systemic illnesses should be considered, and special attention should be directed at underlying medical conditions, medication use, and past surgical history.
Ancillary Testing Most cases of acute diarrhea are self-limited, and laboratory and diagnostic tests should be kept to a minimum unless required for epidemiologic studies. Testing is indicated in patients who have a high probability of a “non-norovirus” clinical picture and have worrisome historical data, signs, and symptoms associated with an increased probability of those causes. Ancillary testing should never compromise empirical treatment when indicated (as discussed later on). Fever with a toxic appearance and volume depletion, blood- or mucuscontaining stools, frequent voluminous stools, and other risk factors (Table 23-1) should prompt a diligent search for a specific causative disorder in order to guide appropriate therapy.8,10 A white blood cell count is rarely helpful and not sensitive or specific enough to aid in diagnostic decision-making, although hemoglobin determination is useful to screen for anemia from blood loss, and abnormalities in platelet and coagulation parameters may contribute to identification of a cause for gastrointestinal bleeding. A comprehensive chemistry panel including renal function tests can be important when significant volume loss is suspected, or when significant diarrhea has been present for 48 to 72 hours. Liver function studies, thyroid tests, serum lipase assay, and a pregnancy test may be helpful in selected cases. Hemoccult and fecal cell count: The presence of fecal leukocytes is not specific or sensitive enough to use as the sole criterion to decide which patients with presumed bacterial gastroenteritis should be treated empirically with antibiotics. With inflammatory diarrhea of various causes, red and white blood cells are seen on stool examination. Included are bacterial, parasitic, and noninfectious causes, such as chemotherapy, radiation therapy, hypersensitivity reactions, autoimmune disorders, and inflammatory bowel disease. The presence of fecal leukocytes does not delineate which patients would benefit from empirical antimicrobial therapy. The presence of blood does not always correlate with the presence of fecal leukocytes, so reliance on positive stool guaiac test result alone as a rationale for antibiotic therapy is not recommended. The presence of blood without fecal leukocytes may indicate amebiasis, malignancies, heavy metal poisoning, fissures, hemorrhoids, bowel ischemia, or primary gastrointestinal bleeding. Assays for calprotectin and lactoferrin, produced by leukocytes, are sensitive and specific and may be more useful than microscopic examination of the stool, but these tests are rarely, if ever, of use in the ED.6 Clostridium difficile toxin assay: This test is indicated if the patient reports recent antibiotic use. C. difficile–associated diarrhea most commonly occurs during or shortly after the antibiotic course. In 25% to 40% of cases, however, onset of the diarrhea may be delayed as long as 12 weeks after antibiotic therapy. The most commonly implicated antibiotics are cephalosporins, penicillins, and clindamycin. Although C. difficile accounts for only 10% to 20% of antibiotic-associated diarrhea, an assay for C. difficile toxin gives a positive result in nearly all cases of antibiotic-associated pseudomembranous colitis.11 Approximately 3% of adult patients and 65% of newborns may be colonized with C. difficile. E. coli O157:H7 toxin assay: This test is considered in endemic areas and in patients with suspected hemolyticuremic syndrome.12
Stool culture for bacteria: Stool cultures may be warranted in patients who are febrile, toxic-appearing, immunocompromised, at the extremes of age, experiencing a prolonged course, or not responding to conventional treatment. Studies have shown a 2% positive rate, thus proving that routine cultures are of limited value.12 Stool examination for ova and parasites: The assessment of stool for ova and parasites is not routinely recommended. This study is used in patients with chronic diarrhea (E. histolytica, Cryptosporidium); patients with a history of travel to developing countries, particularly to Nepal or areas of Russia (Cryptosporidium, Giardia, Cyclospora)13; patients with exposure to infants in day care centers (Cryptosporidium, Giardia); and patients with HIV infection (E. histolytica, Giardia).14 Giardia antigen assay and serologic testing for amebiasis may be considered in patients exposed to poor sanitation, HIVinfected patients, patients with a history of travel to developing countries, patients with a history of backpacking, and patients with day care exposures. Urinalysis: A urinalysis and a urine pregnancy test should be obtained only when urinary tract infection is a possibility, a gastrointestinal origin for the symptoms is not clear, or pregnancy is suspected. Radiographic studies: Plain radiographs and contrast computed tomography (CT) may be indicated for patients thought to have a surgical abdomen and to identify pathologic abnormalities, such as tumor, obstruction, free air, fistulas, blind loops, and those associated with Crohn’s disease. Gastrointestinal referral: Referral may be indicated in the evaluation of chronic diarrhea and for workup beyond the scope of the ED (e.g., endoscopy, further stool studies, biopsy).
■ EMPIRICAL MANAGEMENT Oral rehydration is the treatment of choice for mild to moderate fluid losses (Fig. 23-1). Oral rehydration can be accomplished using sports beverages, commercial rehydration solutions, or a balanced clear liquid diet in the home (e.g.,
Acute diarrhea syndrome
Primary and secondary survey: • Toxic, inflammatory • Hypovolemia or shock • Acidosis, bloody stool? No
• Oral rehydration • Symptomatic outpatient treatment • Probiotic agents
Yes
• Volume resuscitation • Laboratory studies • Imaging (US, CT) • Empirical antibiotic therapy
Consider admission with: • Comorbid conditions • Poor response to resuscitation • Suspicion of other than infectious etiology
Figure 23-1. Approach to the patient with acute diarrhea. CT, computed tomography; US, ultrasonography.
3.5 g of sodium chloride 2.9 g of trisodium citrate or 2.5 g of sodium bicarbonate 1.5 g of potassium chloride 20 g of glucose or 40 g of sucrose Replacement of micronutrients, particularly copper and zinc, has been recommended, especially in developing countries.15 The concept of bowel rest has been abandoned because it may worsen diarrhea and lead to more severe dehydration. The choice of oral rehydration fluids depends on the extent of dehydration and the underlying health of the patient. In otherwise healthy patients with mild to moderate dehydration, fluids such as sports drinks, diluted fruit juices, and soft drinks supplemented with soups, broths, or crackers may be sufficient to replace the fluid and sodium losses associated with acute diarrhea. Such frequently used “clear liquids” may contain excess sugars and insufficient sodium content, however, leading to an osmotic diarrhea. Beverages containing caffeine should be avoided because caffeine increases cyclic adenosine monophosphate levels and may cause a secretory diarrhea. Milk and other products containing lactose also should be avoided because viral and bacterial pathogens, responsible for many cases of diarrhea, may cause a transient lactase deficiency, leading to malabsorption and osmotic diarrhea. Food intake is encouraged, but foods high in simple sugars should be avoided because the osmotic effect is counterproductive. Foods with a high fat content may delay gastric emptying and should be avoided. The BRAT (bananas, rice, apples, and toast) diet has long been recommended, particularly with pediatric patients. Although no controlled studies have examined the efficacy of the BRAT diet, it remains a commonly recommended strategy. The pectin in the peel of apples is constipating (pectin, found in fruit peel, is the “pectate” in Kaopectate), and bananas provide potassium. If this diet is used for extended periods, adequate provision of protein and energy needs of the patient becomes a concern. In patients with evidence of more severe dehydration, intravenous fluid resuscitation with normal saline or lactated Ringer’s solution is the preferred treatment. Pediatric patients should receive a bolus of 20 mL/kg of normal saline, which may be repeated as indicated. Specific treatment for diarrhea should be directed toward the suspected cause. In patients with suspected surgical pathology, further diagnostic testing and surgical consultation may be required. With toxic exposures, treatment consists of early decontamination, supportive care, and, if appropriate, administration of specific antidotes. Other noninfectious causes of diarrhea are treated as indicated. Because the specific pathogen causing infectious diarrhea is rarely identified in the ED, and the results of cultures are usually unavailable, any antimicrobial treatment must be empirical and guided by knowledge of the common causes of infectious diarrhea (see Box 23-1). Viral and noninvasive bacterial gastroenteritis tend to be self-limiting and require only supportive therapy. Empirical antibiotic treatment is directed against invasive bacterial and parasitic organisms that cause the greatest harm. Antibiotic treatment is initiated in patients with a suspected invasive process and severe diarrhea, sys-
181
temic symptoms, fever, or abdominal pain and in patients who appear toxic. The current recommendation for empirical treatment of a systemically ill–appearing adult is ciprofloxacin, 500 mg orally twice a day, or levofloxacin, 500 mg orally every 24 hours for 3 to 5 days.16 Fluoroquinolones are efficacious against most organisms that cause dysenteric illnesses and have been shown to be more effective than trimethoprimsulfamethoxazole.13,17 Fluoroquinolones should not be administered to pregnant patients or children younger than 18 years of age. The antibiotic treatment of severe gastroenteritis in children has been associated with the development of hemolytic-uremic syndrome and thrombotic thrombocytopenic purpura if the bacterial cause is enterohemorrhagic E. coli O157:H7, although Salmonella, Shigella, and Campylobacter also have been implicated. If possible, treatment for pediatric patients should be based on culture results with supportive care initially. If amebic dysentery is of concern in high-risk patients (see Table 23-1), treatment with metronidazole after stool analysis for ova and parasites is recommended. In patients with a history of recent antibiotic use suspected of having C. difficile colitis, a C. difficile toxin assay followed by vancomycin or metronidazole is appropriate.18 The use of antimotility agents in the treatment of acute enteritis has been controversial, with the literature divided over this issue. Patients with simple, acute viral gastroenteritis benefit from antimotility agents and often obtain significant relief of symptoms, with less fluid loss and without significant complications. Loperamide is the safest and most effective medication. Relief of symptoms is achieved much more rapidly than with bismuth subsalicylate (Pepto-Bismol) in patients with inflammatory diarrhea or antibiotic-associated colitis. In pediatric age groups, the use of opioids, loperamide, or diphenoxylate with atropine rarely has been associated with the precipitation of toxic megacolon and hemolytic-uremic syndrome. Because the beneficial effects of these medications are modest, they should be avoided or used with extreme caution in these high-risk patients. Probiotics have been used as an alternative to traditional antibiotic therapy for diarrhea. Lactobacillus and other bacteria have proved to be effective in restoring the normal gastrointestinal flora that is disrupted during diarrhea illness. This approach has been most effective with traveler’s diarrhea and nonspecific diarrhea in children.
■ DISPOSITION Most patients with uncomplicated, acute diarrhea can be discharged home after assessment and symptomatic relief. Hospitalization rarely is required for diarrhea secondary to viral and many forms of bacterial gastroenteritis, which tend to be self-limiting. Often the exact etiologic agent of diarrhea is not identified in the ED. An understanding of common causes and their treatment and recognition of patients at risk for a more severe clinical course are essential to make the appropriate disposition. In patients with severe dehydration, hemodynamic instability, or a toxic appearance and in high-risk groups, hospital admission is warranted for continuous monitoring, further treatment, and definitive management when initial evaluation and stabilization are complete. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 23 / Diarrhea
consisting of water, salt-containing liquids such as canned soups, and potassium from oranges or bananas). The WHO has defined an oral rehydration solution (WHO-ORS) that can be made by dissolving the following in 1 liter of clean water:
Chapter 24
Constipation
Jan M. Shoenberger
■ PERSPECTIVE Constipation is a symptom, not a disease. Patients and doctors often define constipation differently. Patients often use the term constipation to describe a broad set of complaints including straining, hard or infrequent stools, feeling of incomplete evacuation, and abdominal discomfort. Constipation may be acute (new for the patient) or chronic. Chronic constipation is defined as the presence of symptoms for at least 3 months. The Rome III criteria constitute a consensus definition of functional chronic constipation often used in research (Table 24-1).1 Attempting to identify the cause of this symptom will often result in the best chance of effective treatment and will help determine disposition. A definitive diagnosis often is not possible in the emergency department, and appropriate followup evaluation should be arranged in those cases. When constipation becomes severe with constant pain, some clinicians use the term obstipation. Obstipation represents the progression of the symptom of constipation toward bowel obstruction. In the emergency department (ED), the complaint of constipation should be of concern when it represents a significant change from a patient’s own normal pattern that is creating discomfort for the patient. This change may manifest as a decrease in frequency of defecation, sudden and persistent change in the character or amount of stools (especially decrease in stool caliber), blood in the stool, or problems expelling the stool.2
Epidemiology The prevalence of constipation in North America is approximately 15%. Constipation is more common in women than in men, in nonwhites than in whites, and in the elderly. A consistent trend of increasing prevalence of constipation is observed with age, with significant increases after the age of 70 years. The high prevalence among elderly patients is multifactorial and related to a diet low in fiber, sedentary habits, multiple medications, and various disease processes that impair neurologic and motor control. Constipation also is common with patients who are institutionalized, debilitated, or neurologically impaired.3
Pathophysiology Normally the gastrointestinal tract is presented with 9 to 10 L/ day of secretions and ingested fluids. The small intestine 182
usually absorbs all of this except for approximately 500 mL. The colon mixes the ileal effluent, ferments and salvages the unabsorbed carbohydrate residues, and desiccates the contents to form stool. The process of stool transport and evacuation is complex and is regulated by neurotransmitters, intrinsic colonic reflexes, and a multitude of learned and reflex mechanisms that are not fully understood. Constipation may result from structural, metabolic, mechanical, neurologic, or behavioral disorders that affect the colon or anorectum either directly or indirectly.4-6
■ DIAGNOSTIC APPROACH Differential Considerations The causes of constipation are numerous. Causes of constipation can be divided into primary (no apparent external cause) and secondary causes (summarized in Table 24-2). These two groupings have some overlap. In the ED, patients commonly present with acute constipation due to side effects of medications or avoidance of defecation secondary to presence of painful perianal lesions such as fissures, hemorrhoids, or perirectal abscesses.6
Pivotal Findings History A thorough, detailed history usually identifies the most likely cause of the patient’s constipation. Defining what the patient means by “constipation” is a good starting point. Essential information includes the presence or absence of alarming signs or symptoms. These include fevers, anorexia, nausea, vomiting, new onset or worsening of constipation, blood in the stool, weight loss, and a family history of inflammatory bowel disease or colon cancer. Additional elements of the history are directed toward elucidating a possible cause. Questions about the character of the stools may reveal a decrease in caliber of the stool, suggesting possible mass lesion, or diarrhea alternating with constipation, which may indicate irritable bowel syndrome. Frequency of stools and what the patient considers “normal” should be assessed. The review of systems may need to include questions regarding associated symptoms if no obvious cause is elicited in the cursory history. Questions directed at associated neuro-
183
Table 24-1 Rome III Criteria for Functional Constipation
logic symptoms, activity level, and status of comorbid diseases may provide clues to contributing factors.7 A medication history is essential and should include any recent changes in dosing of any prescription medications, herbal agents, and over-the-counter (OTC) medications. Many patients experience constipation as a side effect of medication. Drugs of abuse also may cause changes in bowel patterns. Opiates are the most common cause of constipation among medications and drugs of abuse.
Physical Examination The physical examination should initially focus on two major aspects: the abdominal and rectal portions of the physical examination. The abdominal examination usually yields normal findings but may reveal tenderness, a mass, distention, or possibly evidence of obstruction. Bowel sounds should be auscultated. The anorectal examination and an evaluation of the stool are the most important parts of the physical assessment. Anorectal inspection may reveal fissures, skin excoriations, hemorrhoids, or rectal prolapse. The digital rectal examination should include careful palpation for masses, and the presence or absence of pain should be noted. Other possible findings include strictures, high sphincter tone, and the presence of blood. Having the patient bear down may be helpful in assessing sphincter function and may reveal milder forms of prolapse. The quantity and the characteristics of the stool should be recorded. Testing the stool for occult blood may yield additional information, although straining at stool can produce local anal lesions and bleeding. If results of occult blood testing are positive, diverticular disease, carcinoma, and simply trauma from repeated attempts at straining all are possibilities. Patients with acute constipation who present to the ED most commonly have large amounts of hard stool in the rectum. Results of rectal examination have not been shown, however, to correlate with complaints of constipation or with evidence of colonic loading on abdominal radiographs. The rectal examination alone should not be used to confirm or exclude the presence of constipation.8
Ancillary Testing A majority of patients who present to the ED with a chief complaint of constipation do not need any testing. Plain radiographs may provide information about extent of stool retention but also may suggest emergent diagnoses such as megacolon or volvulus. Although constipation can cause cramping and abdominal pain, plain radiographs documenting an increased stool load in the constipated patient cannot be used to rule out more serious underlying etiologic disorders, especially if the
Primary Causes Functional Disorders Idiopathic Irritable bowel syndrome Pelvic dyssynergia (anismus) Slow-transit constipation Neuropathic Congenital anal sphincter myopathy Hirschsprung’s disease Spinal cord injury Obstructive Anal stenosis Crohn’s disease Colon cancer Stricture Rectal prolapse Gynecologic Large rectocele Pelvic relaxation Secondary Causes Lifestyle/General Condition Dehydration Inadequate dietary fiber Sedentary Voluntary suppression of defecation Medications Antacids Anticholinergics Anticonvulsants Antidepressants Antihistamines Antiparkinsonian drugs Antipsychotics Calcium channel blockers Calcium supplements Diuretics Iron supplements Laxatives (chronic abuse) Nonsteroidal anti-inflammatory drugs Opiates Metabolic/Endocrine Diabetes mellitus Hypercalcemia Hypokalemia Hypothyroidism Hypomagnesemia Porphyria Uremia Myopathic Scleroderma Amyloidosis Neurologic Cerebrovascular accident Autonomic neuropathy Mutiple sclerosis Paraneoplastic neuropathy Parkinson’s disease Amyotrophic lateral sclerosis Psychological Anxiety Depression Eating disorders Situational stress Sexual abuse Adapted from Swegle JM, Logemann C: Management of common opioidinduced adverse effects. Am Fam Physician 74:1347, 2006.
Chapter 24 / Constipation
(1) At least 2 of the following for a minimum of 3 months, with symptom onset at least 6 months before diagnosis: (a) Straining during ≥25% of bowel movements (b) Lumpy or hard stools for ≥25% of bowel movements (c) Sensation of incomplete evacuation for ≥25% of bowel movements (d) Manual maneuvers to facilitate ≥25% of bowel movements (e.g., digital evacuation, support of the pelvic floor) (e) >indirect
Indirect>>direct
Normal/mild ↑ transaminases ↑↑↑ Alkaline phosphatase ± ↑ PT/PTT + / ↑ Amylase
↑↑↑ Transaminases Normal or ↑ alkaline phosphatase Normal or ↑ PT/PTT Normal amylase
Normal transaminases Normal alkaline phosphatase Normal PT/PTT
Suggests obstructive process
Suggests hepatocellular/cholestatic process (including fulminant hepatic failure)
Suggests hematologic process
• Choledocholithiasis • Intrinsic bile duct disease • Cholangitis • AIDS cholangiopathy • Strictures • Neoplasms • Extrinsic biliary compression • Neoplasms (pancreatic/liver)
• Viral hepatitis • Fulminant hepatic failure • Alcoholic hepatitis • AST >> ALT • Ischemia • Toxins • Autoimmune hepatic disease • HELLP syndrome
Figure 25-1. Laboratory approach to differential diagnosis of jaundice. AIDS, acquired immunodeficiency syndrome; ALT, alanine aminotransferase; AMS, altered mental status; AST, aspartate aminotransferase; CBC, complete blood count; HELLP, hemolysis, elevated liver enzymes, and low platelets; PT, prothrombin time; PTT, partial thromboplastin time.
• Hemolytic disorder • Hematoma resorption • Ineffective erythropoiesis • Gilbert’s syndrome*
*A benign hereditary condition characterized by hyperbilirubinemia and jaundice due to inadequate hepatic conjugation of bilirubin.
Fever with right upper quadrant tenderness suggests cholangitis.3 In this clinical scenario, the liver should not be engorged. A large tender liver may represent an exacerbation of acute or chronic hepatitis or malignant infiltration. A palpable gallbladder, a rare finding, suggests chronic cholestasis or malignancy. The presence of splenomegaly suggests hemolysis, malignancy, or portal hypertension. Ascites may be associated with acute or chronic liver disease. Ascites associated with abdominal tenderness raises suspicion for spontaneous bacterial peritonitis. Rapid onset of hepatomegaly and ascites may indicate portal vein thrombosis (Budd-Chiari syndrome).4 Jaundice associated with a large pulsatile abdominal mass may indicate a rapidly enlarging or ruptured abdominal aortic aneurysm. The patient’s mental status should be assessed for evidence of hepatic encephalopathy. Physical examination findings associated with chronic liver disease and cirrhosis include spider angiomas, gynecomastia, testicular atrophy, and caput medusae. Excoriations from scratching in attempts to relieve pruritus suggest chronic liver
disease. Asterixis, a sign of hepatic encephalopathy, usually is found only in patients with chronic liver disease. Table 25-1 summarizes the clinical stages of hepatic encephalopathy.
Ancillary Testing Figure 25-1 lists the laboratory tests that should be considered in the evaluation of the patient with jaundice. Alkaline phosphatase (AP) also can be elevated in diseases affecting bone or the placenta in the first trimester. In the setting of isolated elevated AP, increased serum gamma-glutamyl transpeptidase (GGT) or 5′-nucleotidase points to a hepatic source. A reticulocyte count and evaluation of the peripheral blood smear may identify hemolysis. In the setting of toxic ingestion or unexplained hepatocellular injury, serum acetaminophen concentration level is indicated. Rapid stool guaiac testing should be performed to assess for the presence of gastrointestinal bleeding. Patients with altered mental status should have a rapid bedside glucose assessment in addition to determination of
189
Table 25-1 Clinical Stages of Hepatic Encephalopathy CLINICAL STAGE
Subclinical
Stage 1
Stage 2 Physical Exam • Assess mental status • Abdominal tenderness/liver size • Hepatomegaly • Skin findings: petechiae/purpura, caput medusae, spider angiomata • Ascites • Pulsatile mass
Lab Tests • CBC with platelets • PT/PTT • Hepatic panel: transaminases, alkaline phosphatase, bilirubin with fractionation, amylase • ABG • Alcohol level/tox screen • Pregnancy test
Figure 25-2. Pivotal points in the assessment of the jaundiced patient. ABG, arterial blood gas (analysis); Alk phos, alkaline phosphatase; CBC, complete blood count; IVDU, intravenous drug use; PT, prothrombin time; PTT, partial thromboplastin time.
serum ammonia concentration. Although elevated serum ammonia may aid in diagnosis, the degree of hyperammonemia has not been shown to correlate with the degree of encephalopathy.5 In the presence of abdominal tenderness and ascites, ascitic fluid should be tested for cell count, Gram staining, culture, and protein.6 Two sets of blood cultures should be obtained in patients with fever and jaundice. If the patient appears ill or there is evidence of gastrointestinal bleeding, type and screen or type and crossmatch studies should be performed.
Imaging The best radiologic study for the emergent evaluation of obstructive biliary disease remains somewhat controversial. Both ultrasonography (US) and computed tomography (CT) are available in the emergency department (ED), and each has its advantages. The choice of imaging procedure depends on the pretest probability of biliary obstruction and on the index of suspicion of malignancy. In cases in which the probability of malignant obstruction is high, CT is the preferred imaging methodology. It is more sensitive than US in locating the site of the obstruction. Additionally, it is 70% accurate in staging disease and determining resectable versus unresectable disease.7 Patients with a high likelihood of benign obstruction
Stage 3
Stage 4
INTELLECTUAL FUNCTION
NEUROMUSCULAR FUNCTION
Normal examination findings, but work or driving may be impaired Impaired attention, irritability, depression, or personality changes Drowsiness, behavioral changes, poor memory, disturbed sleep Confusion, disorientation, somnolence, amnesia
Subtle changes in psychometric testing
Stupor and coma
Tremor, incoordination, apraxia Asterixis, slowed or slurred speech, ataxia Hypoactive reflexes, nystagmus, clonus, muscular rigidity Dilated pupils and decerebrate posturing, oculocephalic reflex
From Fitz G: Hepatic encephalopathy, hepatopulmonary syndromes, hepatorenal syndrome and other complications of liver disease. In Feldman M, Friedman L (eds.): Gastroenterology and Liver Disease, 8th ed., Philadelphia, WB Saunders, 2006, p 1966.
are best screened with US. It is safe, rapid, and less expensive and less invasive than CT. Some common duct stones may be missed with US, but it is as sensitive as CT in determining the presence of obstruction. US with Doppler flow can detect obstruction of the hepatic, portal, and splenic veins. Sonographic features that suggest acute cholecystitis include presence of pericholecystic fluid and gallbladder wall thickening.8 If gallstones are present on ultrasound images, a sonographic Murphy sign has a positive predictive value of 90% for acute cholecystitis. In patients with low or intermediate clinical likelihood of mechanical obstruction, US is the preferred modality to evaluate whether or not biliary obstruction is present. CT is preferred if the entire abdomen needs to be evaluated.
■ DIFFERENTIAL DIAGNOSIS Using a systems approach, jaundice can be classified into critical, emergent, and nonemergent categories (Table 25-2). Patients are considered to be critically ill if they present with jaundice and any of the following: altered level of consciousness, hypotension, fever with abdominal pain, or active bleeding. Any patient with a new triad of jaundice, encephalopathy, and coagulopathy is considered to have fulminant hepatic failure.9 In general, these patients have no previous history of liver disease and experience sudden onset of illness or toxic exposure that leads to hepatic necrosis. The time course from insult to fulminant hepatic failure ranges from 1 to 8 weeks. Patients with fulminant hepatic failure require aggressive stabilization, consideration for toxic exposures, and admission to an intensive care unit or possible transfer to a center with liver transplantation capabilities.
Chapter 25 / Jaundice
History • Viral prodrome • Liver disease • Alcohol/IVDU • Biliary tract surgery • Fever/abdominal pain • Pregnancy • Toxic or therapeutic ingestion • Malignancy • Recent or remote blood products • Occupational exposure • Cardiovascular disease • Recent trauma • Travel history
190
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
Table 25-2 Causes of Jaundice Grouped by Level of Urgency ETIOLOGIC CATEGORY
CRITICAL
EMERGENT
NONEMERGENT
Hepatic
Fulminant hepatic failure Toxin Virus Alcohol Ischemic insult Reye’s syndrome
Hepatitis with normal mental status, normal vital signs, and no active bleeding
Biliary
Cholangitis
Systemic
Sepsis Heatstroke
Cardiovascular
Obstructing AAA Budd-Chiari syndrome Severe congestive heart failure Transfusion reaction
Hepatitis of any etiology with confusion, bleeding, or coagulopathy Wilson’s disease* Primary biliary cirrhosis Autoimmune hepatitis Liver transplant rejection Infiltrative liver disease Drug-induced (isoniazid, phenytoin, acetaminophen, ritonavir, halothane, sulfonamides) Toxin ingestion or exposure Bile duct obstruction (stone, inflammation, stricture, neoplasm) Sarcoidosis Amyloidosis Graft-versus-host disease Right-sided congestive heart failure Veno-occlusive disease Hemolytic anemia Massive malignant infiltration Inborn error of metabolism Pancreatic head tumor Metastatic disease Hyperemesis gravidarum
Gilbert’s syndrome Physiologic neonatal jaundice
Hematologic-oncologic
Reproductive
Preclampsia/HELLP syndrome Acute fatty liver of pregnancy
Post-traumatic hematoma resorption Total parenteral nutrition
Cholestasis of pregnancy
*In Wilson’s disease, hereditary deficiency of ceruloplasmin causes copper to accumulate in the liver, leading to fulminant hepatic failure. AAA, abdominal aortic aneurysm; HELLP, hemolysis, elevated liver enzymes, low platelets.
■ EMPIRICAL MANAGEMENT AND DISPOSITION Specific therapies depend on the likely clinical entity causing the jaundice (Fig. 25-3). The patient with depressed mental status should have bedside glucose testing. If mental status remains significantly depressed, endotracheal intubation for maintaining airway patency or protection may be required. Intravenous access should be obtained immediately, and crystalloid infusion may be indicated in the hypotensive patient. A quick assessment of volume status is required because hepatic congestion with jaundice can occur in the setting of congestive heart failure. Owing to the risk of coagulopathy, compressible sites should be used for central venous access. Significant bleeding from any source requires aggressive management. Crystalloid infusion is initiated and continued until blood products become available. Coagulopathy should be corrected with fresh frozen plasma and blood volume repleted with packed red blood cells. If ascites is present, diagnostic paracentesis should be considered to rule out spontaneous bacterial peritonitis (SBP). This disease can have a subtle presentation and may be missed without a diagnostic paracentesis. The presence of more than 250 polymorphonuclear cells per cm3 of ascitic fluid is diagnostic for SBP. The empiric antibiotic of choice is a third-generation cephalosporin.6 Patients with jaundice and elevated transaminase levels out of proportion to the elevation of alkaline phosphatase have a hepatocellular injury pattern. Liver failure with hepatic encephalopathy, if present, can be treated with lactulose, either 60 mg orally or 300 mg by retention enema. Patients with fulminant hepatic failure should be admitted to the intensive care unit or possibly transferred to a liver transplantation center.
Even in the absence of acute liver failure, patients with encephalopathy or unstable vital signs should be hospitalized. On the basis of laboratory data alone, patients with new-onset jaundice should be hospitalized if transaminases are greater than 1000 IU/L, the bilirubin exceeds 10 mg/dL, or there is evidence of coagulopathy. Any of these laboratory abnormalities suggests significant hepatic dysfunction. Patients with hepatitis or cholestatic jaundice may be managed as outpatients if they have a normal mental status and stable vital signs, are able to tolerate oral fluids, have no evidence of acute bleeding, and have no complicating infectious process. Intravenous fluids and antiemetics may be required in the ED. Medications with potential hepatotoxicity, particularly acetaminophen, should be avoided. If the laboratory evaluation and diagnostic imaging point to an obstructive picture, ascending cholangitis must be ruled out. If it is suspected, blood cultures should be obtained, followed by prompt administration of broad-spectrum antibiotics with coverage for gram-negative aerobes and anaerobes. Patients with this disorder usually require emergent decompression by means of endoscopic retrograde cholangiopancreatography (ERCP) or cholecystostomy, which dramatically improves survival.10 Some stable patients can undergo a trial of antibiotics and have drainage performed subacutely.11 However, conservative treatment is more likely to fail in patients older than 75 years or chronic smokers.12 Patients with extrahepatic obstructive jaundice without cholangitis should be admitted for drainage. ERCP is therapeutic for benign obstructions such as gallstones or strictures. Patients with obstructive jaundice due to malignancy also benefit from biliary decompression, whether operative, endoscopic, or palliative. Once jaundice develops, malignancy is associated with more advanced disease and increased morbid-
191 Patient with jaundice
History • Abdominal pain, fever, chills • Prior abdominal surgery • Older age Physical • High fever • RUQ abdominal tenderness • Palpable mass • Evidence of prior abd surgery
Figure 25-3. Management of the patient with
jaundice. Alk phos, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; bili, bilirubin; CT, computed tomography; ERCP, endoscopic retrograde cholangiopancreatography; GI, gastrointestinal; H/O, history of; IVDU, intravenous drug use; LFT, liver function test; PT, prothrombin time; PTT, partial thromboplastin time; RUQ, right upper quadrant.
History • Viral prodrome • Alcohol/IVDU • H/O transfusion • Hepatotoxin exposure • Known hepatitis exposure • Pregnancy • Malignancy
History • Trauma • Recent transfusion • Hematopoietic disorder Physical • Hematoma • Evidence of trauma • Paucity of exam findings
Physical • Hepatomegaly • Ascites • Asterixis • Encephalopathy • Spider angiomata • Caput medusae • Gynecomastia • Testicular atrophy • Excoriations Laboratory evaluation
Direct bili > indirect bili
Indirect bili > direct bili
• ± ↑ AST/ALT • ↑↑ Alk phos • ± ↑ Amylase
Suggests obstructive process
Direct bili > indirect bili • ↑↑ AST/ALT • Mild ↑ Alk phos • Normal amylase: normal / ↑ PT/PTT
• Normal LFT results • Abnormal hemogram
Suggests hepatocellular/cholestatic process (including fulminant hepatic failure)
Suggests hematologic process
Reassess and treat signs and symptoms Radiographic evaluation • Ultrasound or CT • Direct bile duct visualization • ERCP/surgical • GI and surgical consultations
ity and mortality.13,14. Biliary drainage has been correlated with improvements in cardiac function15 and, not insignificantly, food intake.16 In patients with obstructive jaundice secondary to malignancy, preoperative drainage is not beneficial in those undergoing surgery.17 Palliative biliary drainage is recommended for patients who are not surgical candidates. Endoscopic drainage with biliary stenting has been found to result in fewer complications, although the rate of recurrent obstruction is higher18 than with percutaneous drainage. In general, patients with uncomplicated cholecystitis should receive intravenous fluids in the ED, parenteral analgesics, and antiemetics as needed and should be hospitalized. For uncomplicated cholecystitis, antibiotic therapy usually is not indicated. Patients with temperature greater than 38.8°C
• Observation • GI consultation • Remove toxins • Viral markers
• Type and crossmatch blood • Hematologic consultation
(102°F), a toxic appearance, or frank sepsis should receive broad-spectrum antibiotic therapy with coverage for enteric pathogens, streptococcal species, and anaerobes. These patients should undergo emergent imaging and consultation with a surgeon or gastroenterologist.8 Choledocholithiasis, presence of a stone in the common bile duct, may not be as easily visualized by sonography but is suggested by significant obstructive signs and symptoms and dilation of the common bile duct beyond 6 mm. Affected patients require hospitalization for possible ERCP and cholecystectomy.19-21 In immune-mediated hemolytic anemia, appropriate crossmatching may be difficult and fatal if not done properly. The decision to transfuse should be based on the achievable level
Chapter 25 / Jaundice
Stabilize serious signs and symptoms
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
192
of oxygenation and the feasibility of instituting alternative treatments. An urgent hematology consultation is recommended. In the case of drug-induced hemolytic anemia, the mainstay of treatment is removal of the offending agent. For patients with glucose-6-phosphate deficiency, blood transfusions are rarely indicated, and the focus of management should be on maintaining urine output to prevent renal failure. Patients with hemoglobinopathies rarely require transfusion therapy unless they present with severe anemia without evidence of reticulocytosis. Fluids, oxygen, and analgesics can be given for an acute crisis.
■ SPECIAL POPULATIONS One specific presentation that warrants discussion is the pregnant patient who presents with jaundice. Normal physiologic changes in pregnancy have little effect on the liver, so jaundice always indicates serious pathology. Jaundice can occur in pregnancy as a result of any of the conditions discussed earlier, as well as conditions specific to pregnancy, such as hyperemesis gravidarum, acute fatty liver of pregnancy, and intrahepatic cholestasis of pregnancy. Hyperemesis gravidarum usually manifests in the first trimester and, in severe cases, can be associated with elevated serum bilirubin. The exact mechanism for jaundice is unknown but is likely to be related to malnutrition and impaired excretion of bilirubin. ED treatment is unchanged in these cases: hydration and antiemetics. Patients with hyperemesis and jaundice should be admitted for intravenous hydration.
Intrahepatic cholestasis of pregnancy is an idiopathic cause of jaundice that occurs early in the third trimester. It manifests with pruritus mainly on the trunk, extremities, palms, and soles, followed by jaundice after 1 to 4 weeks. Other features of obstructive jaundice such as acholic stools and dark urine may be present. Laboratory analysis reveals a cholestatic picture. Affected patients are at increased risk for preterm delivery and intrauterine fetal demise and should therefore be managed in conjunction with the obstetric team or transferred to a center capable of caring for premature neonates. Specific treatments include cholestyramine for pruritus and vitamin K. Acute fatty liver of pregnancy (AFLP) occurs in the third trimester and is characterized by accumulation of microvesicular fat within hepatocytes. It is rare, occurring in 1 in 13,000 deliveries. There is a slight predilection toward primiparous and multiple gestation pregnancies. Clinical manifestations include nausea, vomiting, right upper quadrant or epigastric pain, malaise, anorexia, and jaundice progressing to fulminant hepatic failure and encephalopathy. Treatment consists of prompt delivery. Jaundice and liver dysfunction may progress after delivery but generally resolve. AFLP generally does not recur in subsequent pregnancies. Liver transplantation has been successful for this condition.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 26
Acute Pelvic Pain in Women
Ari Lipsky
■ PERSPECTIVE
Pathophysiology
Women of childbearing age who present with low abdominal pain often have pathologic conditions related to the female reproductive system or bladder, although additional causes must also be considered. Potential etiologic disorders range from the very benign to the immediately life-threatening. Pregnancy presents its series of considerations, and pregnancy status should be determined in all patients.
The female pelvis contains the vagina, uterus, fallopian tubes and ovaries, ureters and urinary bladder, and sigmoid colon and rectum, as well as components of the musculoskeletal system. Although pelvic pain often originates from the reproductive organs, it may arise from any structures that lie adjacent to or course through the pelvis. Visceral pain afferents supplying the pelvic organs have common innervation with the appendix, ureters, and colon. Their significant overlap makes accurate localization difficult for both patient and clinician. Pain may be initiated by inflammation, distention, or ischemia of an organ, or by spillage of blood, pus, or other material into the pelvis. Parietal pain develops when the afferent nerves in the parietal peritoneum adjacent to an affected organ are stimulated.
Epidemiology Acute pain due to pelvic pathology is common, although the presenting complaint is often abdominal pain or lower abdominal pain; a complaint of low back pain may also signal pelvic pathology. A flare of chronic pelvic pain may manifest as an acute process. In a survey of reproductive-age adult women, 39% reported that they experience nonmenstrual pelvic pain at least sometimes.1 Among women who present to an emergency department (ED) and receive a gynecologic diagnosis, 24% of those diagnoses are for pelvic inflammatory disease (PID), 23% for lower genital tract infections (e.g., cervicitis, candidiasis, Bartholin’s abscess), 12% for menstrual disorders, 12% for noninflammatory ovarian and tubal pathology (including cysts and torsion), and 4.3% for ectopic pregnancy.2 In the general population, annually 5.8 of every 1000 women present to an ED and receive a diagnosis of PID, and 1.1 of every 1000 women are diagnosed with an ectopic pregnancy.2 Younger patients and those with multiple sexual partners are more likely to have PID, and a previous episode increases the likelihood of a subsequent episode.3 The risk of ectopic pregnancy is higher in women who have had PID, pelvic surgery, or an intrauterine device. Heterotopic pregnancy is of special concern in women undergoing fertility treatment.4 Common nongynecologic diseases, such as appendicitis, diverticulitis, urinary tract infection, and urolithiasis, remain important considerations in the woman with acute pelvic pain. Box 26-1 lists conditions accounting for pelvic pain in most women.5,6 Some causes of pelvic pain may lead to serious sequelae. PID carries the short-term risk of tubo-ovarian abscess, and the long-term risks of impaired fertility, chronic pelvic pain, and increased predisposition to ectopic pregnancy.3 Rupture of an ectopic pregnancy or a hemorrhagic ovarian cyst may be acutely life-threatening. Unrecognized abuse may have serious or lethal consequences as well.
■ DIAGNOSTIC APPROACH Differential Considerations The differential diagnosis of pelvic pain is broad in scope (see Box 26-1). Most causes of pelvic pain fit into three categories, however: (1) those that originate in the reproductive tract, (2) those that originate in the urinary tract, and (3) those that originate in the intestinal tract. Within the reproductive tract, a subset of causes of pelvic pain is only found in pregnancy; the pregnancy test is therefore a key branch point in the diagnostic process. Potential pregnancy-related disorders can be divided into complications of early pregnancy and complications that occur further along in pregnancy. Although the specific cause of pelvic pain is not always determined at the initial ED visit, an organized approach usually leads to the confirmation or exclusion of disorders most likely to result in significant morbidity.
Pivotal Findings It is rare that any particular finding on history or physical examination (summarized in Table 26-1) is reliable enough to conclusively make or exclude a particular diagnosis, so ancillary testing (beyond a simple pregnancy test) is commonly required in the evaluation of patients with acute pelvic pain. The bimanual examination may at times provide important and convincing information. Unfortunately, however, findings on pelvic examination are somewhat subjective and unreliable,7,8 and the test may be more helpful to localize the process to one side or the other, or to help focus the workup of the 193
194
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
BOX 26-1 Potential Causes of Pelvic Pain in Women Reproductive Tract Ovarian torsion Ovarian cyst Salpingitis/tubo-ovarian abscess Septic pelvic thrombophlebitis Endometritis Endometriosis Uterine perforation Uterine fibroids Dysmenorrhea Pregnancy-Related First Trimester Ectopic pregnancy Threatened abortion Nonviable pregnancy Ovarian hyperstimulation syndrome Second and Third Trimesters Placenta previa Placental abruption Round ligament pain Intestinal Tract Appendicitis Diverticulitis Ischemic bowel Perforated viscus Bowel obstruction Incarcerated/strangulated hernia Inflammatory bowel disease Gastroenteritis Urinary Tract Pyelonephritis Cystitis Ureteral stone pathologic process to the reproductive organs. For instance, tenderness on examination that seems to arise from the right ovary may be appropriately used to guide the subsequent workup, perhaps the ordering of a pelvic ultrasound study. The lack of certainty of the findings on the bimanual examination, however, do not allow the examiner to completely exclude appendicitis, especially if the pelvic ultrasound study fails to identify a clear explanation for the pain. A sequential approach, as outlined next, allows the clinician to progressively limit the diagnostic possibilities until a sound provisional diagnosis is reached.
Symptoms The location of pain and the radiation pattern often are helpful in focusing the differential diagnosis toward a specific cause or group of causes. Lateral pelvic pain usually is related to a process in the tube or ovary. In right-sided pain, appendicitis is considered, and in left-sided pain (especially in patients older than 40 years of age), the differential diagnosis includes diverticulitis and colitis. Urolithiasis may also manifest as lateral pelvic pain, especially when the stone is impacted at the ureterovesicular junction. Central pelvic pain usually is due to processes involving the uterus or bladder, or involving both adnexae. Pain radiating to the rectum may be secondary to pooling of fluid or blood in the cul-de-sac. Diffuse pain may occur with a bilateral process, such as PID, or with diffuse peritonitis secondary to infection or intra-abdominal hemorrhage.
Information regarding the onset and duration of pain may also be useful. Patients with uncomplicated appendicitis (without perforation or abscess) typically present within 48 hours of symptom onset. Sudden-onset pain suggests acute intrapelvic hemorrhage, cystic rupture, or ovarian torsion. Gradual-onset pain is more consistent with inflammation (such as in PID) or obstruction. Chronic or recurrent pain is consistent with endometriosis, recurrent ovarian cysts, or a persistent ovarian mass. The quality of pain may differentiate the crampy, intermittent pattern of muscular contractions along a hollow viscus (arising from, e.g., uterine, ureteral, or bowel pathology) from the steady, progressive pain associated with inflammatory or neoplastic causes, but this finding is highly variable. Pain associated with PID often manifests at the end of menses. Ovarian cyst pain may fluctuate through several menstrual cycles, finally manifesting as rupture, which often occurs in the middle of the menstrual cycle. A complaint of fever and chills is more common with an infectious process. Nausea and vomiting occur more frequently when the process originates within the gastrointestinal tract but also may accompany ovarian torsion, ureteral colic, other causes of severe pain, and pregnancy. Dysuria and frequency occur in many local vulvar and vaginal processes, such as herpesvirus infection, candidiasis, and other types of vulvovaginitis, but urgency typically signals an irritated bladder or urethra, focusing attention on the urinary tract. Information about the patient’s last menstrual period, pattern of menses, and sexual activity pattern is useful, although such data cannot be used to rule out pregnancy. Accordingly, a pregnancy test is always indicated except in women who have had a hysterectomy or are clearly postmenopausal. In a pregnant patient, the obstetric history may provide some helpful diagnostic clues. Recurrent spontaneous abortion or previous ectopic pregnancy increases the likelihood of these conditions, respectively. Patients who are actively undergoing infertility treatment are at increased risk for ectopic pregnancy, heterotopic pregnancy, ovarian torsion, and ovarian hyperstimulation syndrome. Round ligament pain usually is noted in the second trimester. Postpartum patients are at increased risk for endometritis. The presence, quantity, and duration of associated vaginal bleeding should be ascertained. (See also Chapters 27 and 176.) In a nonpregnant patient, bleeding may be associated with PID, trauma, dysfunctional uterine bleeding, or cervical or uterine cancer. In a pregnant patient, bleeding may be associated with a subchorionic hemorrhage in an otherwise viable pregnancy or with an ectopic pregnancy or a nonviable intrauterine pregnancy (which may continue to cause bleeding after expulsion of the uterine contents, especially if any products of conception are retained), or later in pregnancy with placenta previa or abruption. In some cases, the amount of bleeding may be substantial enough to necessitate blood transfusion and surgical intervention. As part of the past medical history, any recent procedures should be ascertained. All women are interviewed in private to permit disclosure of sensitive information, such as a known pregnancy or recent abortion. The onset of pelvic pain shortly after uterine instrumentation increases the possibility of uterine perforation or infection. Sexual history is important, with an emphasis on recent sexual contact and previous history of sexually transmitted diseases.
Signs The physical examination is directed toward the abdomen and pelvis. Pelvic examination is performed in virtually all patients, including pregnant patients at less than 20 weeks of gestation.
Abrupt moderate to severe lateral pain
Ruptured corpus luteum cyst (emergent-critical with significant hemorrhage; otherwise, urgent) Ovarian torsion (emergent)
Dysmenorrhea, dyspareunia
Often minimal
Urinary urgency and frequency; fever and vomiting if patient has associated pyelonephritis Nausea and vomiting
Fever, vaginal discharge
Low-grade fever, nausea, anorexia
Light-headedness if bleeding is severe; rectal pain arises from fluid in cul-de-sac. Nausea and vomiting
Vaginal bleeding
ASSOCIATED SYMPTOMS
Prior history of same type of pain in association with menstrual cycle
Prior history of similar pain
Prior history of stones
Migration of pain to RLQ from center, abdominal pain before vomiting Vaginal discharge, history of PID, history of unprotected intercourse/ multiple partners Recent urologic procedure, prior history of UTI
History of ovarian mass
Missed period; history of previous ectopic pregnancy, infertility, tubal ligation, PID, or IUD use
SUPPORTING HISTORY
Common
Common
Common
Common
PID: common TOA: uncommon
Common
Uncommon
Uncommon
Common
PREVALENCE IN ED
Lateral pelvic tenderness, with or without a mass Unilateral or bilateral adnexal tenderness, occasionally pelvic mass present, peritoneal findings uncommon
Patient often appears uncomfortable, but physical examination can be otherwise unremarkable
Suprapubic tenderness, flank tenderness, and fever with pyelonephritis
Pus from cervical os, (+) CMT, adnexal tenderness. Peritonitis suggests severe PID or TOA.
Adnexal mass and tenderness, possible peritonitis RLQ tenderness, possible peritonitis
Hypotension and tachycardia if blood loss is significant; possible peritonitis
Classically unilateral adnexal tenderness, adnexal mass, and CMT
PHYSICAL EXAMINATION
Pelvic US, laparoscopy
Pelvic US, CBC
Urinalysis: hematuria present in ∼80% of cases; abdominal CT
CBC, ESR, CRP, pelvic US, laparoscopy, cervical cultures, cervical smear for WBCs Urinalysis, urine culture
US with Doppler flow studies, laparoscopy US or CT in unclear cases
Pelvic US, CBC, T&C
Pelvic US, quantitative β hCG, T&C progesterone?, laparoscopy
USEFUL TESTS
Symptoms can mimic other types of pelvic pathology; laparoscopy often is needed for confirmation.
If stone is at junction of ureter and bladder, can have localized pain that can mimic appendicitis or other acute pelvic pathology
WBC can be present in urine with PID and appendicitis.
Early in course, tenderness may be minimal or poorly localized. History and physical may be inaccurate for diagnosis, particularly in patients presenting subacutely.
Torsion can be intermittent.
Cannot reliably exclude diagnosis based on history and physical; severe pain, hypotension, or peritonitis suggests rupture. Physical examination findings often do not correlate with volume of blood in pelvis at US.
ATYPICAL OR ADDITIONAL ASPECTS
Chapter 26 / Acute Pelvic Pain in Women
CBC, complete blood count; CMT, cervical motion tenderness; CRP, C-reactive protein; CT, computed tomography; ED, emergency department; ESR, erythrocyte sedimentation rate; β hCG, β human chorionic gonadotropin; IUD, intrauterine device; PID, pelvic inflammatory disease; RLQ, right lower quadrant; T&C, type and crossmatch; TOA, tubo-ovarian abscess; US, ultrasonography; UTI, urinary tract infection.
Endometriosis
Nonruptured ovarian cyst/tumor
Ureteral colic (urgent)
Unilateral or bilateral pelvic pain, often recurrent
Without TOA, pain usually bilateral. May present acutely within 48 hr, or subacutely with up to 3 wk of pain. Pain with urination usually is not severe unless patient has flank pain from associated pyelonephritis. Acute onset, presents within hours. Pain is lateral, usually moderate to severe. Often radiates into the groin. Lateral ache, gradual onset
PID/TOA (TOA: emergent; PID: urgentemergent)
UTI (urgent)
Duration often 1st trimester
1st trimester
■ DIAGNOSTIC ALGORITHM
Unilateral symptoms/ signs?
UTI Ureteral stone
No
No Abdominal tenderness or rebound?
No
Yes
Yes Placental abruption Placenta previa SAB Round ligament pain Labor
Definite IUP on ultrasound?
Yes
Torsion Salpingitis/TOA Ruptured ovarian cyst Mittelschmerz
PID Endometritis Dysmenorrhea Fibroids
Appendicitis Diverticulitis IBD IBS Other
Musculoskeletal Abuse Depression Psychogenic
No Threatened abortion Corpus luteum cyst
Ectopic pregnancy Spontaneous abortion Early pregnancy
Figure 26-1. Diagnostic algorithm for acute pelvic pain. H&P, history and physical; IBD, inflammatory bowel disease; IBS, irritable bowel syndrome; IUP, intrauterine pregnancy; PID, pelvic inflammatory disease; SAB, spontaneous abortion; TOA, tubo-ovarian abscess; UTI, urinary tract infection.
197
Critical? No
Undifferentiated right lower quadrant pain
Analgesia indicated Use medications safe in pregnancy until ruled out
Urine dipstickpositive?
Consider UTI and/or ureterolithiasis If not convincing, continue with algorithm If pregnant at 20 weeks
38° C (>100° F) Intravenous drug use Immunocompromised status
Emergent Aortic dissection Cauda equina syndrome Epidural abscess or hematoma Meningitis Ruptured/expanding aortic aneurysm Spinal fracture or subluxation with cord or root impingement
Physical Examination Abnormal vital signs—hypotension, tachycardia, fever Unequal blood pressure readings in the upper extremities Pulse deficit or circulatory compromise of the lower extremities Pulsatile abdominal mass Loss of rectal sphincter tone, urinary retention, or focal lower extremity weakness Focal back pain with fever abscess or compressive mass, spinal column injury with cord or root compression, and cauda equina syndrome. An accurate history and physical examination guides the investigation of possible more serious underlying pathologic process (Boxes 28-1 and 28-2).4,12 Laboratory and imaging are needed in some cases, but it is usually possible to rule out significant pathology without recourse to extensive testing.
Rapid Assessment and Stabilization If the initial history and physical examination identify any suggestion of serious disease, rapid stabilization measures should ensue consistent with the cause of concern (Fig. 28-1). Management of aortic dissection, ruptured abdominal aortic aneurysm, and spinal cord and column injuries are covered in other chapters. If epidural abscess or cauda equina syndrome is suggested, emergent MRI and neurosurgical consultation then should be obtained based on the results of the scan. For epidural abscess, blood cultures are obtained followed by intravenous (IV) administration of antibiotics against Staphylococcus aureus. For cauda equina syndrome, an urgent neurosurgical consultation is required. Although the evidence supporting steroid use is conflicting, dexamethasone is commonly used with the hope of decreasing compression from inflammation or to shrink tumor mass. For all patients with significant pain, including patients with “benign” causes for back pain, effective analgesia should be provided early in the evaluation.
Pivotal Findings History History of Present Illness. The history helps to localize pain to the most likely structure and mechanism. The following questions are useful in differentiating between mechanical and nonmechanical causes and will help guide appropriate management.
Urgent Back pain with neurologic deficits Disk herniation causing neurologic compromise Malignancy Sciatica with motor nerve root compression Spinal fractures without cord impingement Spinal stenosis Transverse myelitis Vertebral osteomyelitis Common or Stable Acute ligamentous injury Acute muscle strain Ankylosing spondylitis Degenerative joint disease Intervertebral disk disease without impingement Pathologic fracture without impingement Seropositive arthritis Spondylolisthesis Referred or Visceral Cholecystitis Esophageal disease Nephrolithiasis Ovarian torsion, mass, or tumor Pancreatitis Peptic ulcer disease Pleural effusion Pneumonia Pulmonary embolism Pyelonephritis Retroperitoneal hemorrhage or mass
Where is the pain? The patient is asked to point with one finger to the one spot where it hurts the most. Does the pain radiate to the legs and, if so, specifically where in the legs? Does the pain conform to a specific dermatomal area? Radicular pain, particularly extending below the knee in a dermatomal distribution, implies nerve root involvement. Pain mainly in the paralumbar musculature without dermatomal radiculopathy implies nonspecific low back pain. Any associated chest or abdominal pain may indicate a possible visceral cause. Flank location implies a renal origin, and a higher location can be from the chest or pleura. When did the pain start? The patient should describe in detail what he or she was doing when the pain started. Has there been a recent change in type or intensity of physical activity? Is there any past history of back pain, and what therapeutic modalities were used to treat it? If there is a history of back pain, is there any difference between present and past pain? Acute onset associated with a specific task suggests a mechanical cause. Sudden-onset, severe back pain suggests aortic dissection. Slow onset or onset unrelated to activity suggests a nonmechanical cause (e.g., tumor). Non-
Chapter 28 / Back Pain
BOX 28-1 “Red Flags”
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
206 Shock/syncope: Abdominal aortic aneurysm Aortic dissection (severe AI or tamponade) Sepsis (from infectious causes) Pulmonary embolus Are there abnormal vital signs?
Yes
Febrile: Epidural abscess (local tenderness) Osteomyelitis Pyelonephritis Pneumonia Cholecystitis Pericarditis
No
Evidence of neurologic deficits or focal tenderness
Hypertension/unequal UE blood pressures: Aortic dissection
Yes
No Pain without evidence of neurologic deficits
Visceral: Renal colic Biliary colic Pulmonary embolus Peptic ulcer disease Pancreatitis
Afebrile: Epidural hematoma Tumor (metastatic or primary) Fracture (pathologic or traumatic) Osteomyelitis Herniated nucleus pulposus Spinal stenosis Fractures Febrile: Epidural abscess Osteomyelitis Meningitis Musculoskeletal: Lumbosacral strain Paraspinal muscle strain Vertebral fracture (traumatic or pathologic) Herniated nucleus pulposus Tumor (metastatic or primary)
Figure 28-1. Rapid assessment of acute low back pain. AI, aortic insufficiency; UE, upper extremity.
mechanical pain may improve then recur, but the trend is progressive worsening. Are there any aggravating or alleviating factors? Cough or Valsalva maneuver that aggravates the pain in general favors a mechanical cause and may point specifically to a herniated disk. Patients with back pain associated with tumors and infectious causes often present with nighttime pain and persistent pain unrelieved by rest and analgesics. Spinal stenosis presents with diffuse back pain, numbness, and tingling in one or both legs (pseudoclaudication). Symptoms are aggravated by ambulation (especially “downhill”) and relieved with spinal flexion, which increases spinal canal diameter, temporarily relieving the stenosis. Direct trauma may suggest contusion, strain or fracture, while deceleration may suggest aortic dissection. Is there motor or sensory loss, bowel or bladder dysfunction? Back pain associated with progressive or severe neurologic symptoms, motor loss, or urinary retention or bowel incontinence requires MRI or CT, which may indicate the need for emergent decompression. Is there other pertinent history? Other pertinent history should include work history, past and present (a history of repeated
loading would suggest mechanical cause); fever (suggesting infectious cause); medications (anticoagulants associated with epidural hematomas, steroids associated with infection and compression fractures); hematuria (suggest nephrolithiasis or pyelonephritis); and pending litigation or worker’s compensation status (possible secondary gains). Past Medical History. In addition to any history of back disorders, a thorough inquiry about any systemic disease is important. Ask if there is a history of (1) cancer (metastatic disease), (2) inflammatory disease, (3) IV drug abuse (diskitis), (4) arthropathies, (5) endocrinopathies (hyperparathyroidism), (6) bleeding disorders, (7) osteoporosis, or (8) sickle cell disease. Previous atherosclerotic or vascular disease suggests aortic disease; previous kidney stones or alcohol-related disease may suggest related disease. Knowledge of medications or other modalities used to treat present and past symptoms informs direct treatment decisions. Knowledge of current medications used by the patient gives clues about the presence of other systemic disease. The family history also is assessed. Diseases such as spondyloarthropathies (e.g., ankylosing spondylitis) have a familial component.
Physical Examination Vital Signs. Vital signs are important because alterations may suggest a life-threatening process (e.g., hypotension and tachycardia with ruptured abdominal aortic aneurysm, hypertension with aortic dissection, fever with abscess, osteomyelitis, or diskitis). Lower Back Inspection. 1. Observe the patient’s gait and movement in the examining room. Does the patient move cautiously, protecting himself or herself, or freely and appear to be in little pain? 2. Examine the patient while standing, searching for scoliosis (may be structural or secondary to muscle spasm), increase or decrease of lumbar lordosis or thoracic kyphosis (may predispose to mechanical pain), or pelvic obliquity (may indicate muscle spasm, leg-length discrepancy, or uncompensated scoliosis). 3. Assess the range of motion for the low back. Patients with significant mechanical pain usually flex without reversing the normal lumbar lordosis, and extension may aggravate facet causes or nerve root impingement. 4. Perform the palpation in an orderly fashion with the fingertips to localize the area of greatest tenderness (e.g., specific spinous process, paravertebral musculature). Other Examinations, Including Neurologic Examination. 1. The neurologic assessment evaluates the asymmetry of reflexes (clinically, reflexes diminish with age, and uncovering asymmetry is key), dermatomal sensory loss, and focal muscle weakness (suggests nerve root impingement). If possible, motor testing of the legs is best done with the patient standing. Heel-walking and toe-walking indicate normal plantar and dorsiflexion strength, and a partial knee bend while bearing weight on one leg, then the other, indicate normal hip, buttock, and thigh muscle strength. A patient with a long history of back pain should be asked about previous motor, sensory, or reflex abnormality. The presence of clonus, hyper-reflexia, or upgoing toes (Babinski’s sign) indicate an upper motor neuron lesion. 2. A rectal examination can assess sphincter tone and anal wink. Testing for perianal sensation is necessary if there is any history of bowel or bladder dysfunction.
Straight Leg Raise. The straight leg raise is the classic test for sciatic nerve root irritation. It is sensitive but not specific for disk disease.4 This test is often negative in patients with spinal stenosis. With the knee extended, the leg is elevated until pain is elicited. A positive result is pain radiating down the leg below the knee in a dermatomal distribution when the leg is elevated to less than 90° (not back, buttocks, or thigh pain). Pain referred to an affected leg (“crossover pain”) with straight leg raise of the unaffected leg is insensitive but highly specific for nerve root irritation. In a patient who may be malingering, the straight leg raise can be done with the patient sitting with the knees flexed at the side of the bed and then passively straightening the legs. If there is true nerve root irritation, results should be similar in the sitting and the supine positions.
Ancillary Testing Laboratory Tests. For mechanical causes of back pain, laboratory studies are of little use. For nonmechanical causes, erythrocyte sedimentation rate and complete blood count may be useful if inflammatory disease is suggested, but are rarely of use in the ED. Urinalysis is helpful in possible cases of renal disease
207
with referred back pain (nephrolithiasis, pyelonephritis, urinary tract infection). Imaging. Although patient satisfaction is reportedly improved when imaging is performed,13 plain radiographs are not useful in uncomplicated mechanical low back pain of less than 6 weeks duration.12 If the patient has a history of trauma with bony tenderness or focal signs of trauma, neurologic deficit, cancer, unexplained weight loss, pain that persists at rest or at night, advanced age, osteoporosis, prolonged glucocorticoid use, or fever, plain radiographs may be helpful.13 Plain radiographs should not be obtained, however, if advanced imaging (e.g., CT, MRI) is planned. Most patients do not require radiographic evaluation while in the ED.14 Emergency MRI, CT, or myelogram (in order of preference) is indicated if an acute, significant neurologic deficit such as motor loss or cauda equina syndrome is present. For patients with acute back and radicular pain but no motor weakness, and for those with chronic low back pain without neurologic deficit, obtaining MRI and CT does not improve outcome.15-17 Eightyfour percent of patients with sciatica will recover without surgery.18 For patients in whom infection or tumor is suggested, MRI (or bone scan followed by MRI) is the diagnostic test of choice.19 The degree of neurologic impairment and patient stability dictates whether these tests are obtained on an emergent or urgent basis.
■ DIFFERENTIAL DIAGNOSIS After stabilization and assessment, the clinical findings aid in narrowing the differential diagnosis (Table 28-1). An algorithm
Table 28-1 Classic Findings in Selected Serious Causes of Acute Back Pain
Critical Vascular
Infectious
DIAGNOSES
HISTORY
Aortic dissection
Often suddenonset, “tearing” severe pain. Associated nausea, vomiting, acute anxiety are common. Syncope can occur
Abdominal aortic aneurysm (ruptured/ expanding)
Pain may radiate to back, flank, or testicle. Patient may present with syncope
Spinal epidural abscess
At-risk population with diabetes, chronic renal failure, intravenous drug use, alcoholism, cancer, or recent spinal surgery or trauma. Sepsislinked history is common
IMPORTANT PHYSICAL EXAMINATION FINDINGS
ANCILLARY TESTING
Associated diaphoresis, unstable vital signs. Hypertension is common. Unequal upper extremity blood pressure. Newonset aortic insufficiency murmur. Central and peripheral neurologic deficits secondary to ischemia Pulsatile abdominal mass (especially if right of midline), abdominal bruits. Diminished lower extremity pulses or hypoperfusion or both
Choice of CT, MRI, aortogram depends on patient stability and availability of equipment
More common as a chest pain cause, but low back pain may be only complaint
Bedside US. If “stable,” abdominal CT with contrast. Plain films may show a calcified enlarged aortic contour CBC, blood cultures useful but nonspecific. MRI modality of choice. CT or myelography can be used. Search for source of infection. Staphylococcus aureus common cause (70%)
Can also mimic renal colic, GI bleeding, diverticulitis, and myocardial infarction. 30% of signs are misdiagnosed Presents as massoccupying lesion compressing spinal cord; may be hematoma, malignancy, disk. Often begins as focal pyogenic infection in disk. Biopsy may be necessary
Fever, reproducible radicular pain, other signs of sepsis. Localized body tenderness along spine Focal neurologic deficits are late findings (4–5 g/dL in whole blood) is elevated. It is not a percent of desaturated total hemoglobin mass or a decreased amount of oxyhemoglobin. For this reason patients with a relatively low hemoglobin exhibit cyanosis at a much lower partial pressure of oxygen (Pao2) and arterial oxygen saturation (Sao2) than those with normal hemoglobin levels. Cyanosis is an insensitive indicator of tissue oxygenation.2 Its presence suggests hypoxia, buts its absence does not exclude it. Abnormal hemoglobin forms contribute significantly to cyanotic disease. Under normal conditions, red blood cells (RBCs) contain hemoglobin with iron in the reduced ferrous state (Fe2+). The iron molecule may be oxidized to the ferric state (Fe3+) to produce methemoglobin. This reaction impairs the ability of hemoglobin to transport oxygen to and carbon dioxide from the tissues. The oxygen dissociation curve is shifted to the left, resulting in tissue hypoxia and lactic acid production (Fig. 29-1). Methemoglobin normally accounts for less than 1% of total hemoglobin.3 Cyanosis results when greater than 10 to 15% of the total hemoglobin is methemoglobin (≥1.5 g/dL) that has a dark purple-brown color, even when exposed to room air. Methemoglobin is reduced to ferrous hemoglobin primarily by nicotinamide adenine dinucleotide (NADH) cytochrome-b5 reductase, an enzyme system present within RBCs. A secondary NADPH-dependant system uses glutathi-
one production and glucose-6-phosphate dehydrogenase (G6PD) to reduce methemoglobin to hemoglobin. This secondary pathway normally plays a minor role, but is accelerated by methylene blue.3 Primary methemoglobinemia is a congenital error of enzyme metabolism, with either diminished levels of NADH reductase or an abnormally functioning enzyme. Patients may present with cyanosis in a stable compensated state. Acquired methemoglobinemia occurs when methemoglobin production (hemoglobin oxidation) is accelerated beyond the capacity of NADH reductase activity. This usually occurs as a drug reaction. (See Box 29-1 for common causes.) Newborns are at risk for methemoglobinemia because their NADH reductase activity is relatively low.3
■ DIAGNOSTIC APPROACH Differential considerations for patients presenting with cyanosis are listed in Box 29-2.
Pivotal Findings Symptoms The onset, duration, and time of day of symptoms, and any previous episodes should be noted. Precipitating factors may include exposure to cold air or water, high altitude, or exercise in patients with a history of cardiopulmonary disease. Additional history should include known congenital heart disease or cardiopulmonary disease, hypercoagulable states, and any family history of cyanotic disease or hematologic illness. A history of home or occupational exposures to fumes or chemicals should be obtained, including aniline, azo dyes (pyridium), phenacetin, and nitrates.4 A drug history should be reviewed, including use of prescription and over-the-counter medications, health food supplements, and herbal or alternative preparations.5 The potential of pseudocyanosis resulting from exposure to dyes, heavy metals, or topically absorbed pigments should be explored.2 In infants, congenital heart disease is suggested by difficulty feeding, sweating, lethargy, poor weight gain, or respiratory distress. Episodic cyanotic events, or “Tet spells,” may be seen in children with tetralogy of Fallot (ventricular septal defect, overriding aorta, pulmonary stenosis or atresia, and right ventricular hypertrophy with outlet obstruction). These patients present with cyanosis, tachypnea, and anxiety due to 211
Less ↑ pH ↓ 2,3-BPG O2 delivered ↓ T°
100 Percent saturation of hemoglobin
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
212 Oxyhemoglobin
Hereditary Hemoglobin M NADH methemoglobin reductase deficiency (homozygote and heterozygote)
75
↓ pH More ↑ 2,3-BPG O2 ↑ T° delivered
50
P50 25
Deoxyhemoglobin 0 0
25
50
BOX 29-1 Common Causes of Methemoglobinemia
75
100
Tissue PO2 (mm Hg)
Figure 29-1. Hemoglobin-oxygen dissociation curve. Deoxyhemoglobin does not bind oxygen efficiently. Methemoglobin has a high affinity for oxygen molecules and does not readily release oxygen to the peripheral tissues. This shifts the normal oxygen dissociation curve to the left, resulting in hypoxia and lactic acid production. Typically, when acid is produced in the tissues, the dissociation curve shifts back to the right, facilitating oxygen release; however, the high affinity of methemoglobin prevents this normal process. (Redrawn from Benz EJ Jr: Hemoglobinopathies. In Harrison’s online.) 2,3-BPG, 2,3-bisphosphoglycerate; Po2, partial pressure of oxygen; P-50, oxygen half saturation pressure of hemoglobin; T°, temperature.
decreased pulmonary blood flow with shunting of unoxygenated blood into the peripheral circulation.6,7
Signs There is significant interobserver variability in detecting cyanosis on physical examination. Room lighting and temperature may affect examination of the skin and mucous membranes. A patient’s natural skin tone, thickness, and pigmentation also may alter findings. Central cyanosis is often secondary to the shunting of venous unsaturated hemoglobin into the arterial circulation or the presence of abnormal hemoglobin. A bluish discoloration of the skin and mucous membranes is best seen on perioral skin, oral mucosa, or conjunctivae. Peripheral cyanosis is secondary to vasoconstriction and slow flow of normally oxygenated hemoglobin in arterial blood, allowing for greater oxygen extraction by the tissues. Peripheral cyanosis affects capillary beds and typically is seen in the extremities and nail beds. Differential cyanosis may occur in either the upper or lower (or the right or the left) half of the body, with the remainder appearing well oxygenated. This form of cyanosis usually is seen in cases of cyanotic heart disease with multiple anomalies. Vital signs should be obtained on all patients. Temperature is typically normal. Blood pressure and heart rate may be high, normal, or low depending on the underlying cause. Upper airway obstruction and other signs of respiratory insufficiency should be sought. Intermittent apnea in infants suggests central nervous system immaturity or a central lesion. Infants with cyanosis, increased respiratory depth, periodic apnea epi-
Acquired Medications Amyl nitrite Antineoplastics (cyclophosphamide, ifosfamide, flutamide) Celecoxib Dapsone Local anesthetics (benzocaine, lidocaine, prilocaine) Nitroglycerin Nitroprusside Phenacetin Phenazopyridine (pyridium) Quinones (chloroquine, primaquine) Sulfonamides (sulfanilamide, sulfathiazide, sulfapyridine, sulfamethoxazole) Chemical Agents Aniline dye derivatives (shoe dyes, marking inks) Butyl nitrite Chlorobenzene Fires (heat-induced denaturation) Food adulterated with nitrites Food high in nitrates Isobutyl nitrite Naphthalene (mothballs) Nitrophenol Nitrous gases (seen in arc welders) Paraquat Silver nitrate Trinitrotoluene Well water (nitrates) Pediatric Reduced NADH methemoglobin reductase activity in infants (4 mo) Seen in association with low birth weight, prematurity, dehydration, acidosis, diarrhea, and hyperchloremia NADH, reduced nicotinamide adenine dinucleotide. Modified from Goldfrank LR: Toxicologic Emergencies, 6th ed. Stamford, Conn, Appleton and Lange, 1998.
sodes, or diaphoresis with feeding may have congenital heart disease.6 Tachypnea (>60 breaths/min) in a newborn may indicate a pulmonary disorder, congenital heart disease, infection, a metabolic disorder, or central nervous system pathology.8 General appearance and mental status should be evaluated. The head, eyes, ears, nose, and throat examination may reveal central cyanosis. Funduscopic examination may detect dilated tortuous veins and papilledema in patients with cyanotic congenital heart disease.9 Jugular venous distention may be seen on the neck examination in patients with pulmonary edema. The chest examination may reveal crackles, wheezing, or inadequate ventilation. Heart sounds should be assessed for tachycardia, abnormal rhythm, or gallops, and the presence and quality of murmurs, especially in newborns. Central pulse strength should be noted. The abdomen should be examined for the presence of hepatosplenomegaly, a pulsatile mass, or abdominal bruit. Extremity examination includes evaluation of nail beds for peripheral cyanosis, strength and symmetry of distal pulses,
213
BOX 29-2 Differential Diagnosis of Cyanosis
and capillary refill. Evidence of chronic vascular disease, such as hair loss and temperature difference, should be noted. Clubbing of the nails may occur due to increased soft tissue and expansion of the capillary beds (Fig. 29-2). Clubbing may be idiopathic or hereditary, but is usually the result of chronic hypoxemic states, such as cyanotic heart disease, infective endocarditis, pulmonary disease (chronic obstructive pul
Chapter 29 / Cyanosis
I. Peripheral cyanosis A. Low cardiac output states 1. Shock 2. Left ventricular failure 3. Hypovolemia B. Environmental exposure (cold) 1. Air or water C. Arterial occlusion 1. Thrombosis 2. Embolism 3. Vasospasm (Raynaud’s phenomenon) 4. Peripheral vascular disease D. Venous obstruction E. Redistribution of blood flow from extremities II. Central cyanosis A. Decreased arterial oxygen saturation 1. High altitude (>8000 ft) 2. Impaired pulmonary function a. Hypoventilation b. Impaired oxygen diffusion c. Ventilation-perfusion mismatching (1) Pulmonary embolism (2) Acture respiratory distress syndrome (3) Pulmonary hypertension d. Respiratory compromise (1) Upper airway obstruction (2) Pneumonia (3) Diaphragmatic hernia (4) Tension pneumothorax (5) Polycythemia B. Anatomic shunts 1. Pulmonary arteriovenous fistulae and intrapulmonary shunts 2. Cerebral, hepatic, peripheral arteriovenous fistulae 3. Cyanotic congenital heart disease a. Endocardial cushion defects b. Ventricular septal defects c. Coarctation of aorta d. Tetralogy of Fallot e. Total anomalous pulmonary venous drainage f. Hypoplastic left ventricle g. Pulmonary vein stenosis h. Tricuspid atresia and anomalies i. Premature closure of foramen ovale j. Dextrocardia k. Pulmonary stenosis of atrial septal defect l. Patent ductus arteriosis with reversed shunt C. Abnormal hemoglobin 1. Methemoglobinemia a. Hereditary b. Acquired 2. Sulfhemoglobinemia 3. Mutant hemoglobin with low oxygen affinity (e.g., hemoglobin Kansas)
Figure 29-2. Symmetrical cyanosis. Equal cyanosis and clubbing of hands and feet resulting from transposition of great vessels and a ventricular septal defect without patent ductus arteriosus.
monary disease, cystic fibrosis), and some gastrointestinal disorders (cirrhosis, Crohn’s disease, and regional enteritis). Thrombotic events should also be considered with findings of skin and nail bed hemorrhages or end-organ damage (eye, kidney). A neurologic examination should be performed focusing on mental status, symmetry of motor and sensory function, and any gross deficit.
Laboratory and Ancillary Testing The complete blood count should be checked to assess for polycythemia or anemia.10 Peripheral smear assesses RBC morphology and fragments, as well as white blood cell differential count. Interpretation of pulse oximetry is problematic in the setting of cyanosis (see Chapter 3). Assessment of distal perfusion usually determines if poor circulation is a cause of low pulse oximetry. Pulse oximetry measures light absorbance of tissue at 660 nm (red reduced hemoglobin) and 940 nm (infrared oxyhemoglobin). The ratio of these two readings is the basis of the pulse oximetry calculation. Methemoglobin absorbs well at both wavelengths, resulting in a saturation approximation of 85%, regardless of the actual Pao2 and Sao2.11 Arterial blood gas testing assesses Sao2, often sampled when the patient is breathing room air (see Fig. 29-1). Co-oximetry measurements should be specifically ordered if carbon monoxide exposure or methemoglobinemia is suspected. Sulf hemoglobin is reported as methemoglobin on co-oximetry, so if sulfhemoglobinemia is possible, measured oxygen saturation should be specifically requested.12
Imaging A chest radiograph should be ordered to evaluate lung fields for consolidation, infiltrates, and increased vasculature or pulmonary edema. The cardiac silhouette and mediastinum may suggest congenital heart disease. In patients thought to have pulmonary embolism, imaging may include lower extremity venous Doppler ultrasound (if deep venous thrombosis
214
An electrocardiogram should be performed on all patients with cyanosis to assess for arrhythmias and acute ischemic changes. Right-axis deviation or right ventricular hypertrophy may be seen with significant cardiopulmonary disease (e.g., cor pulmonale, acute pulmonary hypertension). An echocardiogram may be helpful in detecting septal defects in infants or valvular disease in infants and adults.
a clue to the presence of congenital cardiac disease. If heart size is normal, impaired pulmonary function, pulmonary embolus, or other noncardiac causes should be considered. If no improvement occurs with 100% oxygen therapy, the patient’s respiratory status should be reassessed, and tension pneumothorax or upper airway obstruction considered. Pul monary embolus should be considered and a ventilationperfusion scan or spiral computed tomography pulmonary angiogram performed. If a patient exhibits no respiratory distress and remains resistant to oxygen therapy, cardiac shunting or abnormal hemoglobin forms should be considered and treated accordingly.
■ DIFFERENTIAL ALGORITHMS
Critical Diagnoses
Figures 29-3 and 29-4 outline the differential diagnosis and treatment for peripheral and central cyanosis, respectively. After the initial assessment is completed, and the distribution of cyanosis is noted, the clinician should begin 100% oxygen therapy and follow steps to determine the cause of cyanosis. Clinical improvement with oxygen suggests diffusion impairment. Patients who do not respond to oxygen are more likely to have ventilation-perfusion ratio abnormalities, such as shunting from a consolidated pulmonary lobule, or congenital heart disease with right-to-left shunting. Cardiac size and silhouette on chest radiograph may provide
Acute cardiovascular and respiratory compromise must be considered in a patient presenting with cyanosis and symptoms or signs of shock. The differential diagnosis for these critical presentations includes acute congestive heart failure, acute coronary syndromes, hypovolemic or cardiogenic shock, acute respiratory insufficiency or failure, massive pulmonary embolism, an exacerbation or decompensation in a patient with known congenital heart disease, or the first presentation of pediatric congenital heart disease. These patients require emergent treatment, critical therapeutic intervention, and admission to the intensive care unit.
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
symptoms are present), ventilation-perfusion scanning or computerized tomography pulmonary angiogram.
Electrocardiogram and Echocardiogram
ABCs Check O2 saturation Administer oxygen
Improvement with O2
No improvement with O2
Consider low cardiac output states
Consider vascular occlusion
Hypovolemia
Sepsis
Cardiogenic shock
Administer IV fluids
Administer IV fluids and antibiotics; vasopressors as indicated
Inotropes, chronotropes, and vasopressors as needed
Consider causes of hypovolemia and treat accordingly
Look for source of infection
Consider causes of cardiogenic shock and echocardiogram to direct treatment
Warm extremity
Improvement
Consider vasospasm from environmental exposure
Consult rheumatology and consider calcium channel blockers or beta blockers
Figure 29-3. An algorithmic approach to peripheral cyanosis. ABCs, airway, breathing, circulation; ABI, ankle brachial index; IV, intravenous. No improvement
Peripheral vascular disease likely
Arterial embolism or thrombosis suspected
Measure ABIs and consult vascular surgery
215
No improvement with O2 or PaO2 ≤100 mm Hg or SaO2 ≤70
Improvement with O2 or PaO2 ≥100 mm Hg
CXR
CXR and consider co-oximetry, MetHg, CO, CN levels, Abnormal cardiac silhouette
Normal cardiac silhouette: Consider pulmonary causes
Infiltrate
No infiltrate
CHF
Inotropes, chronotropes, and vasopessors as needed
Antibiotics and respiratory support as needed
Consider PE, hypoventilation, polycythemia, AV fistulas, decreased pulmonary function
ECG and Echo
Consider causes of cardiogenic shock
V/Q or CTPA
Negative for PE
Digoxin Diuresis
Consult cardiology and admit to ICU1
Positive for PE
Hct >65
Hct 30% or >15% with symptoms: Admit • Treat with methylene blue and O2 • No improvement with methylene blue → SulfHg CO elevation: Admit O2 or hyperbaric O2 and consider other causes.2 CN elevation: Admit O2 Lilly Kit Consider other causes2
• No infiltrate or pneumothorax on CXR • Effusion may be present • Bronchospasm may be present
LMWH or heparin as indicated and admit to ICU
Figure 29-4. An algorithmic approach to central cyanosis. ABCs, airway, breathing, circulation; ABG, arterial blood gas; AV, arteriovenous;
CHF, congestive heart failure; CN, cyanide; CO, carbon monoxide; CTPA, computed tomography pulmonary angiography; CXR, chest radiograph; ECG, electrocardiogram; Echo, echocardiogram; G6PD, glucose-6-phosphate dehydrogenase; Hct, hematocrit; ICU, intensive care unit; IV, intravenous; LMWH, low-molecular-weight heparin; MetHgb, methemoglobin; Pao2, partial pressure of arterial oxygen; PE, pulmonary embolus; prn, as needed; RA, ventilation-perfusion scan. room air; Sao2, arterial oxygen saturation; SulfHg, sulfhemoglobin; V/Q, 1 Patients with chronic cyanotic heart disease may not require ICU care or even admission. Disposition should be discussed with patient’s cardiologist. 2 Cyanide and carbon monoxide toxicity do not present with cyanosis. If either of these is present, consider coexisting diagnosis.
Chapter 29 / Cyanosis
ABCs Check O2 saturation and ABG Administer oxygen
216
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
Emergent Diagnoses Methemoglobinemia is an infrequent cause of cyanosis, but should be considered in patients presenting without a history or physical findings suggestive of cardiovascular or pulmonary disease. Sulfhemoglobinemia is a rare cause of cyanosis, most commonly occurring after exposure to hydrogen sulfide from organic sources, medications that are sulfonamide derivatives, or gastrointestinal sources (bacterial overgrowth). Strong consideration should be given to sulfhemoglobin toxicity in patients with cyanotic findings and methemoglobin on cooximetry, but who do not improve with methylene blue treatment. Polycythemia is defined as an elevated RBC mass due to one of three causes. Polycythemia vera is a disorder of bone marrow stem cells with increased RBC mass, cyanosis, and splenomegaly. Patients may present with hyperviscosity syndrome. Secondary polycythemia occurs with either an appropriate or inappropriate increase of erythropoietin, a physiologic response to chronic hypoxemia (≤92% oxygen saturation), cyanotic congenital heart disease, cigarette smoking, or high altitude exposures. Relative polycythemia is an increased RBC mass, often due to dehydration or reduced plasma volumes. Finally, vascular disease, such as Raynaud’s phenomenon, may present with a cyanotic appearance. Raynaud’s phenomenon occurs in 15% of the population and has a female predominance. Patients have an abnormal response to excessive cold or emotional stress and report vasoconstriction, profound cold sensitivity, and recurrent events of sharply demarcated pallor or cyanosis of the digits. Most commonly, the cutaneous arterial capillary beds of the fingers and toes are affected, but tongue, ear, and other distal areas are sometimes also affected.13
■ EMPIRICAL MANAGEMENT Administration of high-flow oxygen is the first therapy for patients with cyanosis. Any clinical improvement, or lack thereof, should be noted. At this point, consideration of abnormal hemoglobin and toxin-induced cyanosis is crucial because the administration of appropriate antidotes and systemic therapies may decrease morbidity and improve outcome. Intravenous fluid resuscitation should be initiated in patients with hypovolemia. Treatment of congestive heart failure, arrhythmia, or poor cardiac output should occur as clinical conditions indicate. Cardiology consultation is recommended in patients thought to have congenital or ischemic heart disease. Although several specific treatments are discussed here, the cause of the cyanosis may be elusive, and hospitalization is required to determine it.
Specific Strategies
equipment. Urgent treatment with oxygen and methylene blue (1–2 mg/kg IV over 5 minutes)14 is indicated for patients with symptomatic hypoxia (dysrhythmias, angina, respiratory distress, seizures, or coma) and methemoglobin levels greater than 30%. Sulfhemoglobinemia is suggested when the laboratory reports an elevated methemoglobin level and the patient does not respond to methylene blue. Treatment of sulfhemoglobinemia is supportive in addition to removing the causative agent.
Other Causes of Cyanosis Acute therapy for patients with symptomatic hyperviscosity syndrome and secondary polycythemia includes phlebotomy and volume expansion with isotonic crystalloid. The goal of therapy is a normal hematocrit (45% in men and 42% in women). Long-term therapy is focused on the underlying cause, and patients may require referral to a hematologist.10,15 Raynaud’s phenomena is treated with warming the affected digits and extremities. Systemic vasodilating agents (e.g., calcium channel blockers [nifedipine] or nitrates) may be useful in the acute setting.16 If there is no improvement of peripheral cyanosis with warming and administration of 100% oxygen, arterial insufficiency or occlusion may be present. In cases of critical limb ischemia, intravenous heparin should be considered in consultation with a vascular surgeon. Embolic sources, such as endocarditis and abdominal aortic aneurysms should be considered. Vascular bypass, intra-arterial thrombolysis, or stenting may be indicated. Carbon monoxide and cyanide poisoning do not typically present with cyanosis and are covered elsewhere.
■ PATIENT DISPOSITION Admission All patients with a first episode of cyanosis or an uncertain cause require admission. Cardiology consultation and referral is recommended for children with a first episode of congestive heart failure and newly diagnosed or suggested congenital heart disease. Surgical consultation and intervention are indicated for acute arterial occlusion from embolic or thrombotic sources.
Discharge Patients with peripheral cyanosis from vasospasm, asymptomatic methemoglobinemia less than 15%, and stable patients with primary pulmonary disease may be treated as outpatients, after several hours of monitoring in the ED. Unless the patient has a previous diagnosis of chronic cyanosis, follow-up must be arranged within 24 hours. Instructions should clearly state that if the cyanosis worsens, or if dyspnea, altered mentation, or chest pain occur after discharge, the patient must return immediately to the ED.
Methemoglobinemia and Sulfhemoglobinemia If cutaneous exposure with an inciting agent occurred (i.e., aniline dyes), complete decontamination with soap and water is recommended. The staff should use appropriate protective
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 30
Sore Throat
Amy V. Kontrick, Jonathan I. Singer, and Mark E. Gebhart
■ PERSPECTIVE Epidemiology Sore throat is a frequent complaint among patients presenting to the emergency department (ED). The National Health Care Survey in 2001–2002 reported more than 2.4 million ED visits with complaints relating to the throat and acute pharyngitis diagnosed in more than 1.9 million patients.1 The chief complaint of sore throat is seen in every age group and has no sex predilection. Sore throat and other upper respiratory tract infections are some of the most common diseases for which care is sought and for which antibiotics are prescribed.1,2
Pathophysiology Sore throat results from irritation or inflammation of any anatomic surface within the oropharynx. The oropharynx is defined by the following borders: (1) posteriorly by the prevertebral fascia, (2) laterally by the buccinator muscle groups, (3) superiorly by the base of the skull, and (4) inferiorly by the vocal cords (Fig. 30-1). Pain may originate within the buccal mucosa, tongue, palatine tonsils, lingual tonsils, adenoids, soft palate, and posterior pharyngeal wall. In addition, pain may result from infection, inflammation, or invasive diseases of the potential spaces within and surrounding the oropharynx—the peritonsillar, retropharyngeal, sublingual, submental, lateral pharyngeal, parotid, buccal, and pretracheal spaces. Sore throat also occurs with inflammatory changes of the epiglottis, aryepiglottic folds, vocal cords, and subglottic region. Infectious diseases of dental structures and cervical nodes and the presence of middle ear fluid may cause sore throat through referred pain. The 9th and 10th cranial nerves provide sensory input from the oropharynx, larynx, middle ear, and external auditory canal.3 Many systemic diseases, including hepatitis, infectious mononucleosis, retroviral disease, and neutropenia, may also have sore throat as part of their symptom complex or initial presentation. Sore throat commonly results from infections within the oropharynx, and the majority of these illnesses are self-limited. Table 30-1 lists common infectious and noninfectious causes of sore throat. Although the majority of infections are mild and not associated with serious complications, several infections may result in airway compromise, systemic disease, or sepsis. Viruses cause the majority of cases of sore throat—up to 80% by some reports.2,4 Enterovirus infection accounts for the majority of sore throats in all age groups from late spring through
autumn. Adenovirus, rhinovirus, parainfluenza virus, influenza virus, and respiratory syncytial virus predominate during winter months. Epstein-Barr virus (EBV), herpes simplex, and varicella-zoster virus have less seasonal predilection. Acute pharygitis due to bacterial infection is much less common than viral infection, and the cause can usually be discerned by a combination of clinical evaluation and rapid strep testing, because group A β-hemolytic streptococcus (GABHS) is the most common bacterial pathogen.2,4,5 Aerobes such as GABHS with anaerobes or anaerobes alone cause infection in the deeper planes of the pharynx and neck. GABHS is isolated as the offending pathogen in 10 to 15% of all patients with sore throat. The incidence of GABHS in school-age children with sore throat may reach 15 to 30%, and some studies have reported the incidence as high as 50%.2,5-8 GABHS is implicated in as few as 5% of adults with sore throat, but 47 to 73% of adults with pharyngitis are prescribed antibiotics.2,9 GABHS is most often isolated from patients between late winter and spring. GABHS infection may cause coinfection with other viral agents, but distinguishing acute infection from carrier state is difficult. Fungal colonization and systemic infection with Candida albicans may occur throughout the oral cavity. Immunocompromised patients may present with severe infections or repeated infections. Recent antibiotic therapy, chemotherapy, and radiation therapy increase the risk for fungal colonization with Candida species. Sore throat may be a manifestation of noninfectious systemic disease, trauma, tumor, or congenital anomaly. Additional systemic complaints or physical findings will often accompany these diseases.
■ DIAGNOSTIC APPROACH Differential Considerations The stable patient should receive a directed history and physical examination followed by judicious use of ancillary testing. Table 30-1 lists the possible causes of acute sore throat.
Pivotal Findings History Characteristic of Pain. Rapidly progressing symptoms, high fever, or severe pain suggest the possibility of invasive disease.10,11 A duration of several days accompanied by fever suggests deeper plane infection or systemic disease. Inflam217
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
218
Nasopharynx
Uvula
Oropharynx
Palatine tonsil
Hypopharynx
Root of tongue Epiglottis
Median view of pharynx
Posterior view of pharynx (opened)
Figure 30-1. Anatomy of the nasopharynx, oropharynx, and hypopharynx.
Table 30-1 Differential Diagnosis for Sore Throat INFECTIOUS CAUSES AEROBES VIRAL
COMMON
UNCOMMON
ANAEROBES
OTHER
Rhinovirus Adenovirus
Streptococcus pyogenes (GABHS) GABHS Peptostreptococcus sp. Non–group A streptococcus Neisseria gonorrhoeae Neisseria meningitides Mycoplasma pneumoniae Arcanobacterium hemolyticum Chlamydia trachomatis Staphylococcus aureus
Haemophilus influenzae Haemophilus parainfluenzae Coccidioides sp. Corynebacterium diphtheriae Streptococcus pneumoniae Yersinia enterocolitica Treponema pallidum Francisella tularensis Legionella pneumophila Mycobacterium sp.
Bacteroides sp.
Candida sp.
Coronavirus Herpes simplex 1, 2 Influenza A, B Parainfluenza Cytomegalovirus Epstein-Barr Varicella-zoster Hepatitis virus
Peptococcus sp. Clostridium sp. Fusobacterium sp. Prevotella sp.
NONINFECTIOUS CAUSES SYSTEMIC
TRAUMA, MISCELLANEOUS
TUMOR
Kawasaki disease
Penetrating injury Angioneurotic edema Retained foreign body Anomalous aortic arch Laryngeal fracture Calcific retropharyngeal tendinitis Retropharyngeal hematoma Caustic exposure
Tongue
Stevens-Johnson syndrome Cyclic neutropenia Thyroiditis Connective tissue disease
Larynx Thyroid Leukemia
GABHS, group A beta-hemolytic streptococcus.
mation or infection within Waldeyer’s ring is accompanied by pain localized to the oropharynx. Pain that radiates to the back of the neck or between the shoulder blades suggests prevertebral or retropharyngeal pathology (abscess or calcific tendinitis). Sore throat with radiation to the jaw or ear may be seen with dental abscess or deeper tissue plane infection.12 Associated Complaints. Odynophagia is almost universal, and many viral infections can cause a raspy dysphonia (laryngitis). The presence of severe pain or significant dysphagia, drooling, voice muffling (“hot potato” voice), or difficulty
breathing, however, may indicate serious infection and potential for airway compromise. In the febrile patient, these symptoms suggest glossal abscess, severe infection of the lingual tonsils or palatine tonsils (peritonsillar cellulitis or abscess), epiglottitis, or Ludwig’s angina (submental or sublingual space infection).3,12 Systemic Symptoms. Prolonged fever (more than 5 to 7 days) may be seen in Kawasaki disease. Cough, myalgia, and arthralgia are seen with influenza A and B, parainfluenza, Neisseria meningitidis, and Mycoplasma pneumoniae infection. Hepatitis,
Physical Examination Assessing for airway compromise or potential airway compromise is the critical first step in the approach to the patient with sore throat (Fig. 30-2). Rapid assessment of the patient can be accomplished by observing the patient’s posture, color, level of consciousness, and phonation. Observation alone is especially important in the pediatric patient with potential airway compromise because attempts at a more thorough physical examination may result in agitation and progression to complete airway obstruction. The presence of air hunger, stridor, drooling, or toxic appearance may indicate pending airway obstruction.3,4 A complete ear, nose, and throat (ENT) and general examination will help narrow the differential diagnosis (Table 30-2). A reduced functional caliber of the airway may occur acutely, subacutely, or insidiously, depending on the cause of the disease process.10,11,15,16 Pending airway loss leads to airpreserving posturing especially in children. Infants unable to sit without support choose the lateral decubitus position with the neck hyperextended when partial obstruction occurs. Children capable of sitting may support their heads with their hands. Airway obstruction in an older child is typically associated with fixed upright posturing. The patient has forced flexion at the waist and maintains the neck flexed and the head extended with an open mouth. Alternatively, patients may assume tripod posturing, in which additional support is gained by hands held on a surface.
Ancillary Testing Laboratory procedures, other than rapid group A strep testing or throat culture, generally are not necessary to develop a working diagnosis of viral pharyngitis or GABHS pharyngitis (Table 30-3). Use of the Centor criteria, with or without rapid antigen detection test or culture, is a rational but not universally accepted approach.8,17-19 Patients with a score of 0 or 1 do not require treatment or additional testing. The goal is to
219 Sore throat Chapter 30 / Sore Throat
infectious mononucleosis, cytomegalovirus (CMV), and human herpesvirus 6 are associated with fatigue, malaise, and loss of appetite.13,14 Retroviral disease may present with similar symptoms and will often be accompanied by a rash.13 Epidemiology. For children, sick contacts within a daycare or school setting may provide important clues to infectious causes of sore throat. Secondary spread of disease is common for persons exposed to M. pneumoniae, GABHS, Haemophilus influenzae, N. meningitides, and many viruses. Among adults, intrafamilial spread is common with viruses, Mycoplasma sp., and GABHS. A history of recent orogenital contact may point to gonococcal or herpetic infection. Trauma. Blunt or penetrating trauma to the oropharynx can result in deep space infections. Recent medical or dental procedures may increase the likelihood of certain infections. Potential exposures to caustics or foreign body ingestions should be sought, especially in young children. Immunizations/Specific GABHS History. The patient’s immunization status is assessed for diphtheria, pertussis, H. influenza, and tetanus vaccines. The presence of previous GABHS infections should be ascertained. Immune Status. The immune status of the patient should be assessed, including (1) the presence of diabetes, (2) known immune disorders, or (3) recent chemotherapeutic or radiation therapy. Underlying alcoholism or malnutrition may place the patient at risk for more serious infections. Recent antibiotic use may indicate the presence of resistant organisms or atypical pathogens.
Anxiety, fixed posturing, or stridor
Yes
No
Airway compromising disease
Airway stable
Fever
Altered phonation or drooling
Yes
No
Yes
No
Epiglottitis RPA Tracheitis Ludwig’s angina Lingual abscess
Foreign body Caustic Tumor Congenital anomaly Trauma
Uvulitis Peritonsillar cellulitis Peritonsillar abscess Parapharyngeal abscess
Stomatitis Tonsillitis Pharyngitis Esophagitis Mass lesion
Figure 30-2. Diagnostic alogorithm for the patient with a sore throat. RPA, retropharyngeal abscess.
decrease the cost of additional testing and decrease the inappropriate use of antibiotics while still treating those with GABHS to prevent suppurative and nonsuppurative complications. A complete blood count is rarely helpful but may be used, along with serologic test for EBV, for the patient with a compatible presentation—severe sore throat, fever, and lymphadenopathy. Hematologic findings of leukocytosis, relative and absolute lymphocyte predominance, and the presence of atypical lymphocytes constituting more than 10% of the total leukocyte count suggest EBV. A serologic test such as the heterophile antibody screen (Monospot) may provide evidence of primary EBV infection.13 Patients with a negative serologic test but with compatible symptom complex should be retested a week later because heterophile antibodies may not be present in the first week in 10% of patients.13,14 In addition, CMV, acute retroviral illness, herpes simplex virus, and human herpes 6 viral infections should be considered. A lateral portable upright neck radiograph may be used in the pediatric patient to narrow the differential diagnosis of infectious conditions associated with potential airway obstruction. The lateral neck film may demonstrate swelling in the prevertebral soft tissue in a patient with a retropharyngeal abscess (RPA).4 Plain radiographic imaging is rarely warranted in the adult patient with an acute sore throat. The adult with severe symptoms should be considered for direct nasopharyngoscopy to search for epiglottitis. Use of the H. influenza b vaccine in children has resulted in a dramatic decrease in the incidence of epliglottitis, but the incidence in adults has not changed.10,11,16,20 Ultrasonography may be a useful tool in the diagnosis and treatment of some deep space infections. The advantages of ultrasound are as follows: (1) It can be used at the bedside, (2) it can be used to guide incision and drainage
220
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
Table 30-2 Pivotal Findings in Physical Examination SIGN
FINDING
DIAGNOSES
Appearance
Toxic
Posturing
Fixed, upright, leaning forward
Phonation
Torticollis Absent Muffled
Stridor, drool
Either present
Noninvasive ENT
Conjunctivitis
Epiglottitis RPA Bacterial tracheitis Kawasaki disease Epiglottitis RPA Tracheitis Laryngotracheobronchitis Parapharyngeal abscess Epiglottitis RPA Peritonsillar cellulitis Peritonsillar abscess Epiglottitis RPA Tracheitis Peritonsillar abscess Kawasaki disease Stevens-Johnson syndrome Adenovirus Stevens-Johnson syndrome Behçet disease Enterovirus Herpes simplex Ludwig’s angina Adenovirus EBV Mycobacterium sp. HIV Epiglottitis Thyroiditis Thyroglossal duct cyst infection Parapharyngeal abscess Peritonsillar abscess Kawasaki disease GABHS GABHS Infectious tonsillopharyngitis Caustic Trauma GABHS Corynebacterium diphtheriae Fusobacterium sp. EBV Adenovirus RPA Uvulitis Peritonsillar abscess Parapharyngeal abscess Epiglottitis EBV, hepatitis Lemierre’s syndrome GABHS Arcanobacterium sp. EBV Kawasaki disease
Mucous membrane sore
Submental, sublingual mass Adenopathy
Tender hyoid Tender thyroid Augmented ENT findings
Trismus Tongue coating Palatal petechiae Pharyngeal hyperemia Exudative tonsillitis
Bulged retropharynx Uvular erythema Displaced uvula
Abdomen Joint examination Rash
Inflamed epiglottis Hepatosplenomegaly Arthritis Scarlatiniform
EBV, Epstein-Barr virus; ENT, ear, nose, throat; GABHS, group A beta-hemolytic streptococcus; HIV, human immunodeficiency virus; RPA, retropharyngeal abscess.
221
of peritonsillar abscesses (PTAs), and (3) it decreases the exposure to ionizing radiation in computed tomography (CT) scans.4 CT scanning defines the extent of infection and is superior to ultrasonography for this purpose. It can also help distinguish cellulitis from abscess.4,21 Magnetic resonance imaging provides superior resolution of deep tissue planes and may supplant the use of CT scanning in the future.
■ DIFFERENTIAL DIAGNOSIS
■ EMPIRICAL MANAGEMENT The management of the patient presenting with a sore throat begins with a rapid assessment for potential airway compromise (Figs. 30-2 and 30-3). If the patient is in extremis, immediate airway control is obviously necessary. If the airway is patent and ventilation is adequate, diagnostic and therapeutic efforts may simultaneously commence. Infections within the parotid, buccal, parapharyngeal, submental, and sublingual spaces create masses that are readily apparent.3,12,21,22 The purulent material rapidly expands the tissues but rarely occludes the airway. A thorough head and neck examination accompanied by fiberoptic nasopharyngoscopy, ultrasonography, or CT scan may be necessary to identify the severity and extent of the process. Needle aspiration of a PTA can be both diagnostic and therapeutic. Intravenous antibiotics are administered to treat mixed infection with aerobic and anaerobic organisms. ENT consultation is often necessary for definitive management of PTA, RPA, and other infectious or mass lesions, and early consultation is often warranted. The patient who is febrile and appears toxic, is in respiratory distress, has an abnormal voice or prefers not to speak, or is drooling through a persistently open mouth may require emergent airway management before any other diagnostic
Centers for Disease Control and Prevention:
Table 30-3 Practice Guidelines for Acute Pharyngitis in Adults
Population: Adults (patients older than 15 years) Patients with viral symptoms: Do not test or treat Patients with symptoms of GABHS: Use Centor criteria* Centor score = 4: Perform RADT or treat presumptively Centor score = 3: Perform RADT or treat presumptively Centor score = 2: Perform RADT or do not test or treat Centor score = 1 or 0: Do not test or treat In all cases in which an RADT is performed, only those with positive results are treated. Culture after negative RADT: No Recommended antibiotic: Penicillin (erythromycin if penicillin allergic) *Centor criteria history of fever; absence of cough; swollen, tender anterior cervical lymph nodes; and tonsillar exudate.
GABHS, group A beta-hemolytic streptococcus; RADT, rapid antigen detection test.
Sore throat
Airway compromising disease
Figure 30-3. Management algorithm for the patient with a sore throat. CT, computed tomography, ENT, ear, nose, and throat; GABHS, group A beta-hemolytic streptococcus.
Yes
No
Superficial mass
Trismus, muffled tone, or drooling
Yes
No
Yes
No
Noninvasive exam Consult ENT Ultrasound or CT Optional aspiration Antibiotics IV Admission ICU
No further exam Optional portable x-ray Secure airway in OR Acquire culture Antibiotics IV Admission ICU
Invasive exam Consult ENT Optional imaging Optional culture Antibiotics IV Admission floor
Complete exam
Consistent GABHS?
Yes
No
Supportive care Antibiotics IM, PO Home going
Supportive care Optional culture Home going
Chapter 30 / Sore Throat
Table 30-1 lists the infectious and noninfectious causes of sore throat.
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
222
maneuvers are attempted due to impending airway compromise. If time permits, immediate transfer to the operating room may be warranted, with ENT and anesthesiology consultation. This requires that the operating room be ready to receive the patient and the patient be accompanied by a physician or surgeon capable of surgical cricothyrotomy. If the patient cannot be transported, then fiberoptic intubation (nasal or oral) is the preferred route, with light sedation and topical anesthesia (see Chapter 1). Equipment for cricothyrotomy should be readily available because instrumentation can lead to airway obstruction or laryngospasm. After the airway is secured, the infected surface and secretions can be swabbed for culture; tissue aspiration and blood culture specimens can be submitted for culturing.15 Broad-spectrum parenteral antibiotics are begun for mixed aerobic and anaerobic infection, and the patient should be admitted to the intensive care unit. If the febrile patient does not have evidence of airway compromise but has vocal changes (e.g., muffled or “hot potato” voice), epiglottitis, peritonsillar cellulitis, or abscess may be present. If examination of the oropharynx does not identify
the offending condition, then fiberoptic examination of the upper airway for epiglottitis is indicated. ENT need not be consulted for peritonsillar cellulitis or uvulitis. ENT may be consulted for peritonsillar abscess, even after needle aspiration, for incision and drainage. Intravenous antibiotics are provided to cover Streptococcus pyogenes, non-group A streptococci, and Staphylococcus aureus. The patient may require admission for further care if he or she has severe symptoms or is unable to tolerate liquids by mouth. In the patient with a sore throat who has no evidence of airway compromise, the pain may be a problem within the oropharynx, referred from another location, or part of a systemic illness. Further workup may continue in the ED or on an outpatient basis. The patient with presumed or proven GABHS should be treated with antibiotics. Penicillin remains the drug of choice.7-9,23 Details of treatment are provided in Chapter 73. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 31
Hemoptysis
Calvin A. Brown III
■ PERSPECTIVE Epidemiology Hemoptysis is defined as the expectoration of blood from the respiratory tract that originates below the vocal cords. Most cases are mild and consist of blood-tinged sputum or minute amounts of frank blood. The most common cause of smallvolume hemoptysis is bronchitis. Rarely, hemoptysis is accompanied by massive blood loss, generally accepted as 100 to 600 mL of blood loss in any 24-hour period. In addition to manifesting as hemoptysis, endobronchial bleeding may impair alveolar oxygen exchange and cause significant morbidity and mortality. Rapid blood loss can also result in hemodynamic instability and shock. Although hemoptysis is a common complaint in emergency populations, only 1 to 5% of hemoptysis patients have massive or life-threatening hemorrhage, with mortality rates approaching 80%.1 Large, contemporary series of patients with massive hemoptysis are lacking. Most etiologic data originate from small, often rural studies where tuberculosis (TB) and bronchiectasis are responsible for the vast majority of cases.2,3 In developed nations, cancer, cystic fibrosis, arteriovenous malformations, and postprocedural complications play a more prominent role. Pediatric hemoptysis is rare but can be caused by infection, congenital heart disease, cystic fibrosis, or bleeding from a preexisting tracheostomy.4
caliber airways, or lung parenchyma. Modes of vessel injury include acute and chronic inflammation (from bronchitis and arteritis), local infection (especially lung abscesses, TB, and aspergillosis), trauma, invasion from a growing tumor, infarction following a pulmonary embolus and fistula formation (specifically aortobronchial fistulae). Bronchiectasis, a chronic necrotizing infection resulting in bronchial wall inflammation and dilation, is one of the most common causes of massive hemoptysis. As tissue destruction and remodeling continue, rupture of nearby bronchial vessels results in bleeding. Brochiectasis can complicate chronic airway obstruction, necrotizing pneumonia, TB, or cystic fibrosis. Broncholithiasis, the formation of calcified endobronchial lesions following a wide array of granulomatous infections, is an uncommon problem with a propensity to erode nearby vessels. Hemorrhage control often requires surgical inter vention.5-7 Iatrogenic hemoptysis may complicate 2 to 10% of all endobronchial procedures, especially percutaneous lung biopsies.8,9 Additionally, bleeding can be exacerbated by coagulopathy and thrombocytopenia. An uncommon cause includes ectopic endometrial tissue within the lung that can result in monthly catamenial episodes of hemoptysis. Diffuse alveolar hemorrhage can be seen with autoimmune vasculitides such as Wegener’s granulomatosis, systemic lupus erythematosus, and Goodpasture’s syndrome. Still others include pulmonary hereditary telangectasias and hydatiform infections.
Pathophysiology Trace hemoptysis typically originates from tracheobronchial capillaries that become disrupted with vigorous coughing or minor bronchial infections. Massive hemoptysis almost exclusively involves one of the two sets of vessels that constitute the lung’s dual blood supply. Bronchial arteries, direct branches from the thoracic aorta, are responsible for supplying oxygenated blood to lung parenchyma. Disruption of these vessels from arteritis, trauma, or bronchiectasis or erosion from an adjacent malignancy can result in sudden and profound hemorrhage. Although small in caliber, the bronchial circulation is a high-pressure system and the culprit in nearly 90% of the cases of massive hemoptysis requiring embolization. Pulmonary arteries, although transmitting large volumes of blood, are at much lower pressure and, unless affected at a very central location, are less likely to cause massive hemoptysis. Nearly all causes of hemoptysis have a common mechanism—vascular disruption within the trachea, bronchi, small-
■ DIAGNOSTIC APPROACH Differential Considerations When a patient presents with apparent hemoptysis, two other potential sources of bleeding should be investigated. Nasal, oral, or hypopharyngeal bleeding sometimes inadvertently contaminates the tracheobronchial tree and can mimic true hemoptysis. The clinician should closely inspect the nasopharynx and oral cavity to exclude this possibility. Differentiating hemoptysis from a gastric or proximal duodenal source of bleeding is the principle diagnostic dilemma, since further evaluation and management follow divergent pathways. Usually, this can be done by the patient and physician discriminating coughing from vomiting. In unclear cases, inspection and pH testing may help to distinguish gastrointestinal from tracheobronchial hemorrhage. Unless an active, brisk upper GI hemorrhage is present, the acidification of blood in the stomach results in fragmentation and 223
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
224
darkening of its color. This produces specks of brown or black material often referred to as “coffee-grounds” emesis. Pulmonary blood appears bright red or as slightly darker clots and is alkaline.
endocarditis causing septic pulmonary emboli. Symptoms and signs of deep venous thrombosis should suggest pulmonary embolism. Ecchymoses and petechiae can indicate coagulop athy and thrombocytopenia, respectively.
Rapid Assessment and Stabilization
Ancillary Testing
Although hemodynamic instability can occur as a result of hemorrhage, the most lethal sequela of massive hemoptysis is hypoxia resulting from the ventilation-perfusion mismatch that occurs as small airways and alveoli are submerged with blood. The clinician should consider the standard indications for emergency airway management in such cases. As a mitigating maneuver in patients with a known lateralizing source of bleeding, the “lung down” position can be employed in which the patient is positioned so the bleeding lung is more dependent. This position can promote continued protection and ventilation of the unaffected lung and improve oxygenation.10,11 Large-bore (8.0) endotracheal tubes should be used to facilitate emergent fiberoptic bronchoscopy. In selected cases of confirmed left-sided bleeding, a single-lumen right-mainstem intubation can be successfully performed by advancing the tube in either the neutral position or by using a 90° rotational technique.12 Left-mainstem intubations are more difficult and should be attempted with caution. The use of double-lumen endotracheal tubes for lung isolation should be reserved for dire circumstances and usually requires an experienced anesthetist. The correct positioning of blindly placed double-lumen tubes is difficult and requires confirmation by auscultation and fiberoptic bronchoscopy, both of which have severely impaired accuracy in massive hemoptysis. Complications of double-lumen tubes include unilateral and bilateral pneumothorax, pneumomediastinum, carinal rupture, lobar collapse, and tube malposition.13
Initial laboratory studies include a complete blood count, coagulation tests, and a type and screen or crossmatch. Renal function tests should be obtained if vasculitis is suggested or contrast computed tomography (CT) is planned. Plain chest radiography should be ordered, although its sensitivity is marginal. A prospective study of 184 consecutive patients with varying degrees of hemoptysis reveals that more than 40% of patients with a normal chest radiograph have a positive chest CT.15 In patients with massive hemoptysis, plain films may localize the site of hemorrhage in as many as 80% of patients.6 High-resolution multidetector CT of the chest is the principle diagnostic test for investigating both bronchial and non bronchial arterial causes of massive hemoptysis. CT is diagnostically comparable, yet less invasive, than conventional angiography, which is now done primarily as a combined diagnostic-therapeutic modality.16-18 A chest CT should be obtained in the high risk patient (smokers, oncology patients) or in any patient with moderate to severe bleeding even if the initial chest radiography is normal. CT localization of hemorrhage can expedite bronchoscopic evaluation or guide subsequent interventional procedures.
Pivotal Findings History Although patient reports of bleeding severity are historically inaccurate, a rough estimate of the rate, volume, and appearance of expectorated blood should be obtained. Any history of parenchymal pulmonary disorders should be obtained, including the presence of bronchiectasis, recurrent pneumonia, chronic obstructive pulmonary disease, bronchitis, TB, and fungal infection. Inflammatory disorders that secondarily involve the lungs or pulmonary vasculature include Wegener’s granulomatosis, Goodpasture’s syndrome, and systemic lupus erythematosus. Risk factors for platelet dysfunction, thrombocytopenia, and coagulopathy may be present. Hypercoagulable states can contribute to deep venous thrombi and pulmonary embolism. The presence of primary or metastatic cancer can cause hemoptysis by erosion into pulmonary and bronchial vessels. Recent percutaneous or transbrochial procedures can cause immediate or delayed postprocedural bleeding, and any recent history of trauma should also be noted. A pertinent travel history to areas endemic with TB or pulmonary paragonimiasis is crucial.
Physical Examination After a primary survey and stabilization, a targeted examination may suggest the location and etiology of bleeding, but does so in less than 50% of cases.14 Focal adventitious breath sounds may indicate pneumonia or pulmonary abscess. A new heart murmur, especially in a febrile patient, might reflect
■ DIFFERENTIAL DIAGNOSIS Potential causes of hemoptysis vary and include systemic illnesses as well as pulmonary parenchymal disease. Box 31-1 includes the most common causes.
■ MANAGEMENT Since the advent of high-resolution CT, radiologic evaluation has had an integral role in the evaluation and treatment of patients with hemoptysis. The challenge to the emergency physician is to rapidly assess the need for airway control prior to hemodynamic stabilization. Unless the initial chest radiograph is diagnostic or the patient is hemodynamically unstable, a chest CT should be obtained in most cases. Further management strategy should be developed in conjunction with pulmonary and thoracic surgery consultants, guided by the CT results (Fig. 31-1).
Bronchoscopy Early bronchoscopy facilitates both localization of bleeding and therapeutic intervention. Balloon and topical hemostatic tamponade, thermocoagulation, and injection of vasoactive agents can all control arterial bleeding. Optimal timing for bronchoscopy remains conjectural. Stable patients with mild to moderate bleeding may benefit from early bronchoscopy. In unstable patients or those with brisk hemorrhaging, bronchoscopy sometimes can facilitate airway management, but is less likely to control bleeding. Chest CT is as diagnostically accurate as bronchoscopy in locating peripheral vessels not accessible by a flexible bronchoscope.19 Chest CT is used to identify the bleeding site and to determine whether angiography is indicated. There may be little benefit to bronchoscopy prior to interventional angiography if a CT scan has accurately identified a bleeding source.20
225
BOX 31-1 Differential Diagnosis: Hemoptysis
Parenchymal Disease Tuberculosis Pneumonia/lung abscess Fungal infection Neoplasm Vascular Disease Pulmonary embolism Arteriovenous malformation Aortic aneurysm Pulmonary hypertension Vasculitis (Wegener’s granulomatosis, SLE, Goodpasture’s syndrome) Hematologic Disease Coagulopathy (cirrhosis or warfarin therapy) Disseminated intravascular coagulation Platelet dysfunction Thrombocytopenia Cardiac Disease Congenital heart disease (especially in children) Valvular heart disease Endocarditis Miscellaneous Cocaine Post-procedural injury Tracheal-arterial fistula
Massive hemoptysis
Hemodynamic instability? No Chest radiograph Consider: CBC, ECG, electrolytes, coags, UA, chest CT, V/Q scan
Yes
Pulmonary consult
Hemodynamic instability?
No
Yes Consider early intubation
Yes Supplemental oxygen IV access Cardiac monitor Pulse oximetry
Stat chest radiograph Angled head down, position with affected side down (if known)
Consider intubation with large-bore (8.0) ETT
SLE, systemic lupus erythematosus.
CBC, electrolytes, Abnormal coags, UA, type and cross
Interventional Angiography
Pulmonary, CT surgical consult
Bronchial arterial embolization is an effective first-line therapy and is the procedure of choice for patients unable to tolerate surgery, or those in whom bronchoscopy was unsuccessful. Hemostatic rates range from 91 to 98%, but as many as 20 to 50% of patients have early episodes of repeat bleeding. The risk of delayed bleeding may exist for up to 36 months.21-24 In order to guide therapy, initial localization of bleeding by bronchoscopy or CT is preferred. Rare complications include arterial perforation and dissection.
Chapter 31 / Hemoptysis
Airway Disease Bronchitis (acute or chronic) Bronchiectasis Neoplasm (primary and metastatic) Trauma Foreign body
No
Correct coagulopathy, electrolyte disturbances, blood products as needed
Figure 31-1. Emergency department management of hemoptysis.
CBC, complete blood count; coags, coagulation studies; CT, computed tomography; ECG, electrocardiogram; ETT, endotracheal tube; ventilation-perfusion. IV, intravenous; UA, urinalysis; V Q,
■ DISPOSITION
and can be discharged with follow-up. High risk patients with minor hemoptysis and all patients with moderate or large amounts of hemoptysis should undergo plain chest radiography followed by emergent chest CT. Brief hospitalization or admission to an observation unit for bronchoscopy should be considered. All patients with massive hemoptysis require admission to an intensive care unit and expedited multidisciplinary treatment involving the emergency physician, pulmonologist, and thoracic surgeon.
Healthy patients with blood-streaked sputum and normal vital signs do not require imaging beyond plain chest radiography
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Surgery Emergency thoracotomy is reserved for life-threatening hemoptysis or for persistent, rapid bleeding that is uncontrolled by bronchoscopy and percutaneous embolization. Pulmonary arterial hemorrhage from tumor necrosis represents a surgical emergency.25
Chapter 32
Red and Painful Eye
Joshua L. Wright and John M. Wightman
■ PERSPECTIVE
Pivotal Findings
Epidemiology and Pathophysiology
Measurement of the patient’s best corrected visual acuity (i.e., with glasses on, if available) with each eye individually and with both eyes provides vital information when evaluating eye complaints. Only a few situations preclude early and accurate visual acuity testing. Eyes exposed to caustic materials should be decontaminated as soon as possible. Patients with sudden and complete visual loss in one eye require prompt funduscopic examination to determine the possibility of acute central retinal artery occlusion. This condition is readily apparent as a diffusely pale retina with indistinct or unseen retinal arteries (Fig. 32-2). Other pivotal findings, which are more likely to be associated with a serious diagnosis, in patients with a red or painful eye are listed in Box 32-1.
Most eye complaints are not immediately sight-threatening and can be managed by an emergency physician. Nontraumatic diseases, such as glaucoma and peripheral vascular disease leading to retinal ischemia, are more common with advancing age. Ocular injuries are the leading cause of visual impairment and blindness in the United States.1 More patients with postoperative complications can be expected to present to the emergency department as more vision correction sur geries are performed. The external and internal anatomy of the eye is depicted in Figure 32-1A and B. The globe has a complex layer of blood vessels in the conjunctiva, sclera, and retina. Redness reflects vascular dilation and may occur with processes that produce inflammation of the eye or surrounding tissues. Eye pain may originate from the cornea, conjunctiva, iris, or vasculature. Each is sensitive to processes causing irritation or inflammation.
■ DIAGNOSTIC APPROACH Rapid and accurate triage is the most critical consideration in the approach to the red and painful eye. The first question should be, “Did anything get in your eye?” If so, the second question should be, “What do you think it is?” This helps separate trauma from nontrauma, but, more importantly, seeks to identify quickly eyes that may have been exposed to a caustic substance. Patients exposed to caustic substances require rapid decontamination to prevent permanent loss of visual acuity.
Differential Considerations Diagnoses are classically divided into traumatic and nontraumatic. Traumatic pain and redness can be caused by caustic fluids and solid materials, low-velocity contact with a host of materials that can fall or be rubbed into the eye, higher velocity blunt-force impacts to the orbit or globe, or potentially penetrating injuries. Causes of nontraumatic pain and redness require a more detailed history, including contact lens use and questions directed toward determining the likelihood of systemic illnesses. 226
History Chief complaints of pain can be manifestations of a variety of sensations. When carefully questioned, some patients may differentiate between itching, burning, dull pain, sharp pain, and perception of a foreign body. Itching tends to be more often due to blepharitis, conjunctivitis, or dry eye syndrome. Burning is associated with these conditions and with other mostly extraocular problems such as irritation of a pterygium or pinguecula, episcleritis, or limbic keratoconjunctivitis. Dull pain may be a manifestation of increased intraocular pressure (IOP) or referred from an extraorbital process such as sinusitis, migraine headache, or temporal arteritis. Sharp pain generally results from abnormalities of the anterior eye, such as keratitis, uveitis, and acute angle-closure glaucoma. A foreign body sensation is more typical of corneal irritation or inflammation. A chief complaint of redness commonly results from palpebral or limbal injection of the conjunctiva. However, free blood can be noted behind the bulbar conjunctiva (i.e., subconjunctival hemorrhage) or in the anterior chamber (i.e., hyphema). Both of these can be spontaneous or post-traumatic. Spontaneous subconjunctival hemorrhages may follow coughing or straining or may be due to systemic hypertension. Often, it occurs without any identifiable precipitating incident and is simply noticed by the patient when looking in the mirror. Spontaneous subconjunctival hemorrhage is painless, and the presence of pain raises concern for a more serious cause of the hemorrhage, such as direct globe injury. Hyphema of
227 Pupillary margin Cornea overlying iris
Inner canthus
Sclera
Caruncle
Outer canthus Inferior punctum Iris sphincter Limbus (corneoscleral junction)
Collarette Iris crypt
Cilia (eyelashes)
A
External appearance of the eye Ciliary body Canal of Schlemm Posterior chamber
Fovea
Iris Optic nerve
Anterior chamber
Figure 32-1. External (A) and internal (B)
Lens
anatomy. (From Ragge NK, Easty DL: Immediate Eye Care. St. Louis, Mosby-Year Book, 1990.)
Cornea Limbus Pars plana Retina
B
Cross section of the eye
BOX 32-1
Pivotal Findings More Likely Associated with a Serious Diagnosis in Patients with a Red or Painful Eye
Severe ocular pain Persistently blurred vision Proptosis Reduced ocular light reflection Corneal epithelial defect or opacity Limbal injection (i.e., ciliary flush) Pupil unreactive to a direct light stimulus Wearer of soft contact lenses Neonate Immunocompromised host Worsening signs after 3 days of pharmacologic treatment Adapted and reprinted, with permission, from Trobe JD: The Physician’s Guide to Eye Care. San Francisco, Foundation of the American Academy of Ophthalmology, 2001. Retinal edema
Cherry-red spot
Figure 32-2. Key funduscopic findings in acute central retinal artery
occlusion include general pallor of the retina (except for a characteristic cherry-red spot where the perfused choroid shows through the thinner fovea) and attenuation of retinal arteries (possibly with retinal veins preserved as in the photograph). (From Kaiser PK, Friedman NJ, Pineda R, II: The Massachusetts Eye and Ear Infirmary Illustrated Manual of Ophthalmology, 2nd ed. Philadelphia, WB Saunders, 2004, p 297.)
Chapter 32 / Red and Painful Eye
Superior punctum
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
228
sufficient size to be noted by the patient or bystander usually arises with pain and blurred vision. Other subjective findings may be transient and detected only by history. The patient may relate lid swelling, tearing, discharge, crusting, or sensitivity to light. Lid swelling can be caused by inflammatory and noninflammatory processes. Concurrent erythema of the lid favors the former. In the absence of trauma or other external irritant (e.g., contact dermatitis), inflammatory processes include primary lid problems such as hordeolum (i.e., stye) or blepharitis as well as extension from concomitant conjunctivitis or cellulitis in orbital or periorbital structures. When pain is present, tearing is usually secondary. Discharge and crusting are most commonly associated with conjunctivitis, whether allergic, viral, or bacterial. Blepharitis, dacryocystitis, and canaliculitis are other inflammatory processes that may create a discharge and subsequent crusting. Other eye status review questions include the following: ■
■ ■ ■ ■
Are contact lenses used? If so, what type, how are they cleaned, and how old are the lenses? Has there been a change in the pattern of use (especially increased use)? Were the lenses worn for a particularly long period recently? Are there problems with the lenses drying out? Does insertion of the lenses worsen or relieve the symptoms? Are glasses worn? If so, when was the last assessment for adequate refraction? Has previous eye surgery or injury occurred? What is the patient’s usual state of health? What medications are being taken? Are there any allergies, including environmental allergies?
Physical Examination A complete eye examination usually includes eight components, although many patients require only a limited or directed eye examination, depending on the presentation.2 The mnemonic VVEEPP (pronounced “veep”) plus slit-lamp and funduscopic examinations represent these components (Box 32-2).3 Slit-lamp examination is recommended for any complaint involving trauma and for any medical presentation involving foreign body sensation or alteration of vision.
Funduscopic examination is usually pursued if there is visual loss, visual alteration, or suggestion of serious pathology in the history and initial physical examination. A thorough physical examination can be conducted in the following order.
Visual Acuity The initial determination of a patient’s visual acuity provides a baseline from which deterioration or improvement may be followed. It is also predictive of functional outcome after ocular trauma. Visual acuity is quantitatively assessed by use of a Snellen chart test at a distance of 20 feet (6 m) or a Rosenbaum chart at a distance of 14 inches. Young patients who cannot yet read letters and numbers should be tested with an Allen chart that depicts easily recognizable shapes. Each eye is tested separately with the opposite eye carefully covered. Patients who present without their prescribed corrective lenses may be evaluated by having them view the chart through a pinhole eye cover, which negates most refractive errors in vision. If the patient cannot distinguish letters or shapes on a chart, visual acuity must be determined qualitatively. Any printed material suffices. The result may be recorded as, for example, “patient able to read newsprint at 3 feet.” If this is not possible, visual acuity is recorded as: ■ ■ ■
Unable/able to count fingers (CF) Unable/able to perceive hand motion (HM) Unable/able to perceive light (LP)
Visual Field Testing Confrontation is the most common method of testing visual fields in the emergency department.4 Detection of a scotoma usually represents a retinal problem. However, glaucoma may cause scotomata that can be crescent-shaped, involve just the binasal visual fields, or affect all peripheral vision. Hemi- or quadrantanopia is more commonly a problem of the neural pathways to the brain.
External Examination
BOX 32-2 Complete Eye Examination Visual acuity (best possible using correction) Visual fields (tested by confrontation) External examination Globe position in orbit Conjugate gaze Periorbital soft tissues, bones, and sensation Extraocular muscle movement Pupillary evaluation (absolute and relative) Pressure determination (tonometry) Slit-lamp examination Lids and lashes Conjunctiva and sclera Cornea (with fluorescein in some cases) Anterior chamber Iris Lens Funduscopic examination Adapted from Wightman JM, Hurley LD: Emergency department management of eye injuries. Crit Decis Emerg Med 12:1, 1998.
Gross abnormalities are assessed by a visual inspection of both eyes simultaneously. Findings may be more apparent if compared with the opposite side. Fractures of facial bones are associated with ocular injuries, some of which require immediate intervention by an ophthalmologist.5 Globe position is part of the external examination. Subtle exophthalmos and enophthalmos are rare, and are best detected by looking inferiorly, tangentially across the forehead, from over the patient’s scalp.6 Exophthalmos may have traumatic or nontraumatic causes, but is due to increased pressure or a space-occupying lesion within the orbit, which may manifest as pain. Medical causes include cellulitis or intraorbital or lacrimal tumors. Hyperthyroidism may cause enlargement of extraocular muscles. The most important cause of exophthalmos in the emergency department is retrobulbar hematoma, a condition characterized by hemorrhage within the bony orbit, behind the globe. Orbital compartment syndrome pushes the globe forward, stretching the optic nerve and retinal artery and increasing IOP. The resulting microvascular ischemia is sight-threatening if sufficiently severe and persistent. Orbital emphysema and inflammation caused by a retained foreign body behind the eye are other causes of exophthalmos. The discovery of exophthalmos should prompt ocular tonometry measurements to determine the urgency of intervention. Trauma, particularly penetrating globe injury with extrusion
Extraocular Muscle Function Limitation of ocular movement in one eye may be detected by having the patient follow the examiner’s finger or a bright light through the cardinal movements of gaze. The eyes may move in a disconjugate fashion, or the patient may admit to diplopia if asked. Diplopia on extreme gaze in one direction may indicate entrapment of one of the extraocular muscles within a fracture site, but more often is caused simply by edema or hemorrhage related to the injury and is functional rather than actual entrapment. In the absence of trauma, diplopia is rarely associated with redness or pain.
Pupillary Evaluation The pupils are inspected for abnormalities of shape, size, and reactivity. These examinations are conducted with light specifically directed into the pupil and by means of the swinging flashlight test. Previous surgery (e.g., iridotomy for cataract extraction) and synechiae from prior iritis or other inflammatory condition are the most common causes of irregularly shaped pupils. Asymmetrically sized pupils may represent normal or pathologic conditions. Physiologic anisocoria is a slight difference in pupil size that occurs in up to 10% of the population. Topical or systemic medications, drugs, and toxins may cause abnormal pupillary constriction or dilation. Pathologic reasons for failure of one pupil to constrict with a direct light stimulus include globe injury, abnormalities of afferent or efferent nerves, and paralysis of the ciliaris or sphincter pupillae muscles in the iris. Potentially serious problems, which also cause pain and redness, include uveitis and acute angle-closure glaucoma. The swinging flashlight test is used to determine whether a relative afferent pupillary defect (RAPD) exists.4 The patient fixes the gaze on a distant object and the examination room is darkened. The size of the pupils in lowered light is noted, and unless there is physiologic anisocoria, the pupils should be equal in size. The direct and consensual light responses of the eyes are compared as a light source, angled into the pupil from in front of the cheeks, is swung back and forth between the two. When the light source shines into an eye with an RAPD, the pupil dilates because the consensual response from withdrawal of light from the opposite eye with normal afferent activity is stronger than the direct constrictive response to light in the affected eye with inhibited afferent activity. It is termed “relative” because the response is compared with that of the opposite side as the light source is alternated between eyes. An RAPD may be partial or complete and due to inhibition of light transmission to the retina because of vitreous hemorrhage, loss of some or all of the retinal surface for light contact because of ischemia or detachment, or the presence of lesions affecting the prechiasmal optic nerve (e.g., optic neuritis).
229
Pressure Determination Ocular tonometry is usually the last examination performed in the emergency department. Common methods of determining the IOP in the emergency department include use of electronic, manual (e.g., Schiøtz), or applanation tonometers. IOPs in the 10- to 20-mm Hg range are considered normal. Causes of intraocular hypertension include glaucoma in its many forms, suprachoroidal hemorrhage, and space-occupying retrobulbar pathology. Patients presenting with IOPs exceeding 20 mm Hg should have ophthalmologic consultation. Rapid treatment is usually not necessary until the pressure exceeds 30 mm Hg.
Slit-Lamp Examination The slit lamp permits a magnified, binocular view of the conjunctivae and anterior globe for diagnostic purposes and to facilitate delicate procedures. It allows depth perception in otherwise clear structures, such as the cornea, aqueous humor, and lens. The slit-lamp examination can include the following: ■
Lids and lashes may be inspected for blepharitis and pointing of a lid abscess (i.e., hordeolum). The inner canthus and lacrimal punctum may be better viewed for evidence of dacryocystitis. ■ Punctures, lacerations, and inflammatory patterns of the conjunctiva or sclera may be discovered with magnification. ■ Corneal abrasions, ulcers, foreign bodies, and other abnormalities may be seen. The depth of these lesions may be accurately assessed with an angled beam. Edema, which appears as a white haze or cloudiness within clear structures, can be differentiated within the epithelium or deeper stroma. ■ The anterior chamber may be examined for cells (e.g., red and white blood cells) and “flare.” Cells are seen as small floating objects caught in the beam of a highly angulated slit-lamp light, as dust floating in the movie theater glows from the reflected light of the projector beam. Flare is a diffuse haziness, related to cells and proteins suspended in the aqueous humor, and is often visible only when illuminated directly (Fig. 32-3). It usually represents deep
Reflection from cornea
Space in which to look for particulate matter, “flare”
Reflection from lens
Figure 32-3. Technique of slit-lamp examination with a short, narrow light
beam projected from an extreme temporal angle across the contrasting black pupil to better find cells or “flare” indicative of acute anterior uveitis. (From Ragge NK, Easty DL: Immediate Eye Care. St. Louis, Mosby-Year Book, 1990.)
Chapter 32 / Red and Painful Eye
of vitreous, can cause the globe to recede into the orbit, but the most common cause of enophthalmos is actually pseudoenophthalmos when the contralateral globe is proptotic. Inspection also involves examination of the upper and lower palpebral sulci for foreign bodies or other abnormalities. The lower sulcus is easily viewed after manual retraction of the lower lid toward the cheek and having the patient gaze upward. The upper sulcus is inspected by pulling its lashes directly forward and looking under the lid with white light. The lid can then be everted by pressing a cotton-tipped applicator in the external lid crease and folding the lid margin over the applicator.
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
230
inflammation of the eye and is often seen in iritis. Collections of layered blood or pus in the dependent portions of the anterior chamber are called hyphema or hypopyon, respectively, and are graded by the percentage of the vertical diameter of the visible iris when the head is upright. Foreign bodies that have penetrated the cornea may be found floating in the anterior chamber. ■ The trabeculated pattern of the iris can be seen in detail. Spiraling muscle fibers may be seen in acute angle-closure glaucoma. If the beam is shown almost coaxially with the examiner’s line of sight such that the red reflex is elicited, tears in the iris may be seen by light returning through the iris itself instead of just through the pupil. ■ The lens should be examined for position, general clarity, and the presence of opacities or foreign bodies. The type and position of any lens implants can also be better assessed during a slit-lamp examination.
Direct Funduscopic Examination Emergency physicians most commonly perform a nondilated funduscopic examination because there are several eye conditions in which dilation may be harmful (e.g., glaucoma). Iridodialysis, lens dislocation, and conditions requiring early intervention are usually identifiable along the visual axis. Inability to obtain a red reflex or visualize the fundus of the eye can be due to: ■ ■ ■ ■ ■ ■
Opacification of the cornea, most commonly by edema secondary to injury or infection Hyphema or hypopyon within the anterior chamber Extremely miotic pupil Cataract of the lens Blood in the vitreous or posterior eye wall Retinal detachment
In the absence of trauma, few posterior findings are associated with chief complaints of external redness. Findings associated with visual loss include pallor of the retina indicating ischemia, “cupping” of the optic disk indicating glaucoma, indistinctness of disk margins indicating papilledema or optic neuritis or neuropathy, air or plaque emboli in retinal arteries, and a host of other signs indicating more chronic ocular or systemic pathology not normally amenable to management in the emergency department.
Bedside Testing Fluorescein solution and the cobalt blue lamp are the best means for identifying damage to the corneal epithelium, including that which cannot be seen with conventional slitlamp examination. Fluorescein highlights defects, making them easy to identify, because the fluorescing liquid is thicker in defects than it is across the normally smooth corneal surface. Use of fluorescein may reveal corneal abrasions and ulcers as well as damage from keratitides related to chemicals, ultraviolet light, or infections (e.g., herpes). Relief of discomfort after instillation of a topical anesthetic can be used as a diagnostic test for an external source of pain. In general, abolition of pain by local anesthetic drops indicates pain of corneal origin. Modest but incomplete relief suggests a conjunctival process. Intraocular pain is not diminished by local anesthetic solution.7 When ocular penetration is suggested, Seidel’s test can be used. This test involves placing a fluorescein strip directly over an area of possible corneal disruption. The high localized concentration of fluorescein may facilitate identification of the corneal defect with a slit lamp by allowing visualization of leaking aqueous fluid diluting the fluorescein. This test does not work on the conjunctiva overly-
ing the sclera, and a negative test result does not rule out a full-thickness corneal injury.
Ancillary Testing An erythrocyte sedimentation rate may be used to evaluate for temporal arteritis, which may arise with eye pain and decreased visual acuity. Infections are usually evident by examination, and laboratory tests such as a complete blood count are not necessary. Microbiologic cultures are rarely ordered in the emergency department. Plain radiography is used to identify facial fractures associated with facial or ocular trauma or indirectly by detecting an air-fluid level in the orbit or fluid in the paranasal sinuses. Computed tomography (CT), using 1.5-mm axial and coronal cuts, provides superior imaging, but is not necessary in many cases. CT also reliably localizes metal and many nonradiopaque foreign bodies in the globe and orbit. It can also detect small amounts of intraocular air following penetrating trauma. Magnetic resonance imaging (MRI) clearly delineates the orbital and retro-orbital structures, but cannot be employed with metallic (magnetic) foreign bodies, which can migrate to cause additional damage.8 It is less often used in emergency eye assessment, for which, in general, CT is the initial imaging modality of choice.9 Ultrasonography is more sensitive for detecting intraocular foreign bodies, but CT is better at delineating the damage caused by them, so they are complementary tests.10
■ DIFFERENTIAL DIAGNOSIS Clinical findings most indicative of serious eye disorder are listed in Box 32-1.
Critical Diagnoses Caustic injury to the eye can rapidly lead to a destructive keratoconjunctivitis (Fig. 32-4A and B) if the agent is not removed immediately. The diagnosis is made on history alone, before any other examination is performed. Early and copious irrigation is indicated. Many patients have already undergone extensive irrigation at the job site, but when the exposure has occurred in the home, irrigation prior to arrival in the emergency department is uncommon. Alkaline caustic agents cause a liquefactive necrosis of the cornea by progressively reacting with the corneal layers, and destruction is severe and relentless. Continuous irrigation is the only effective method to terminate the reaction and should be continued for at least 30 minutes. Acid injury is much less severe and requires less irrigation than alkaline exposures, but irrigation should continue until the pH of the tears is neutral or the patient is essentially asymptomatic. Acute angle-closure glaucoma is a relatively rare but important critical diagnosis to make in the emergency department. Patients present with pain, the onset of which is often sudden in low-light conditions requiring pupillary dilation through contraction and thickening of the iris peripherally. The iris becomes immobile and often irregular, and the pupil is commonly fixed at 5 to 6 mm in diameter. Inability of the pupil to constrict may result in photophobia, and accommodation may be affected. These reactions and the increased IOP can lead to frontal headache, nausea, and vomiting. As inflammation progresses, limbal injection of the conjunctiva is almost universally seen. Figure 32-5 demonstrates many of these findings. Immediate medical intervention in the emergency
231
A
B Corneal alkali burn
Posterior cornea
Slit-beam
Iris surface
Figure 32-5. Primary angle-closure glaucoma with very shallow anterior
chamber and iridocorneal touch (no space between slit-beam view of cornea and iris). (From Kaiser PK, Friedman NJ, Pineda R, II: The Massachusetts Eye and Ear Infirmary Illustrated Manual of Ophthalmology, 2nd ed. Philadelphia, WB Saunders, 2004.)
department and urgent ophthalmologic consultation are warranted. Retrobulbar hematoma (blood) is usually caused by orbital trauma, but it can also occur spontaneously in patients with coagulopathy. Retrobulbar abscess (pus) or emphysema (air) can also occur.11,12 Elevated IOP in any of these conditions constitutes orbital compartment syndrome and a surgical emergency. Emergency intervention is to decompress the orbit by performing lateral canthotomy and cantholysis.13,14
Emergent Diagnoses Most emergent diagnoses involve some kind of inflammation secondary to trauma, infection, or systemic disease. These include keratitis, anterior uveitis, scleritis, and endophthalmitis. Any of these may be complications of surgical procedures, and an appropriate ophthalmologic history must be obtained. Keratitis, or inflammation of the cornea, is most commonly viral in origin but can also be caused by exposure to intense ultraviolet light (e.g., snow blindness, arc welder’s blindness), various chemicals, or ischemia related to contact lens use. Patients present with an intense foreign-body sensation, ciliary spasm causes photophobia that is often severe, and the affected eyes are often clenched shut. Topical anesthesia provides immediate (but temporary) relief of pain, thus reinforcing the
Corneal alkali burn
Corneal abrasion
Figure 32-6. Corneal abrasion demonstrating fluorescein pooling of a small inferior epithelial defect. (From Kaiser PK, Friedman NJ, Pineda R, II: The Massachusetts Eye and Ear Infirmary Illustrated Manual of Ophthalmology, 2nd ed. Philadelphia, WB Saunders, 2004.)
corneal origin of the process and facilitating examination and definitive diagnosis.6 Corneal abrasions are very common and may be identified by white light or fluorescein-facilitated blue light using a slit lamp or any other magnification (Fig. 32-6). Following thorough irrigation, thermal and chemical burns must receive a careful slit-lamp examination for potential fullthickness injury. If this is not found, the corneal injury may be treated similarly to an abrasion. In immunocompetent hosts, corneal ulcerations are most commonly due to overuse of contact lenses. They are seen as a denuding of epithelium with surrounding edema, the increased interstitial water of which is seen as whitish clouding of the normally clear tissue (Fig. 32-7). Almost all ulcerations require same-day evaluation by an ophthalmologist. Infections of the cornea with herpes simplex virus can rapidly lead to opacification and significant visual loss. It is most commonly recognized by a characteristic dendritic pattern of fluorescein pooling under blue light (Fig. 32-8). Anterior uveitis, which includes iritis and iridocyclitis, often occurs secondary to a traumatic injury or infectious process or can be associated with serious systemic immune diseases, such as adult and juvenile rheumatoid arthritis, sarcoidosis, and ankylosing spondylitis. Scleritis is rare and may be difficult to differentiate from episcleritis, which is somewhat more common and a more
Chapter 32 / Red and Painful Eye
Figure 32-4. A, Alkali burn demonstrating corneal burns and conjunctival injection on the day of the accident. B, Complete corneal tissue destruction 7 days after alkali burn. (From Kaiser PK, Friedman NJ, Pineda R, II: The Massachusetts Eye and Ear Infirmary Illustrated Manual of Ophthalmology, 2nd ed. Philadelphia, WB Saunders, 2004.)
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
232
Neovascularization
Corneal ulcer
Figure 32-7. Bacterial keratitis demonstrating large, central Streptococcus
pneumoniae corneal ulcer. Note the dense, white corneal infiltrate and the extreme conjunctival injection. (From Kaiser PK, Friedman NJ, Pineda R, II: The Massachusetts Eye and Ear Infirmary Illustrated Manual of Ophthalmology, 2nd ed. Philadelphia, WB Saunders, 2004.)
Necrotizing scleritis
Figure 32-9. Diffuse scleritis with slight bluish region in addition to injection of scleral, episcleral, and conjunctival vessels. (From Kaiser PK, Friedman NJ, Pineda R, II: The Massachusetts Eye and Ear Infirmary Illustrated Manual of Ophthalmology, 2nd ed. Philadelphia, WB Saunders, 2004.) translucent, sclera (Fig. 32-9). Scleritis may be associated with anterior uveitis, cataract, and secondary glaucoma. Endophthalmitis usually results from an infection of structures inside the globe. It is most common following penetrating trauma but may begin after hematogenous seeding from a remote or systemic infection, particularly in immunocompromised hosts. Unless it is detected early, and is responsive to aggressive antimicrobial therapy, endophthalmitis is a devastating process that frequently requires enucleation.
Urgent Diagnoses
Herpes simplex virus dendrite
Figure 32-8. Patient demonstrating fluorescein pooling of herpes simplex
virus dendrite. (From Kaiser PK, Friedman NJ, Pineda R, II: The Massachusetts Eye and Ear Infirmary Illustrated Manual of Ophthalmology, 2nd ed. Philadelphia, WB Saunders, 2004.)
benign inflammation. The former is commonly idiopathic but may be associated with a systemic inflammatory process, such as a connective tissue disease, gout, or infection (e.g., Lyme disease, syphilis, tuberculosis). Eye redness in episcleritis results from dilation of the episcleral blood vessels just underneath the conjunctiva, usually in a small sector of the visible portion of the globe. If the location of the involved layer is in doubt, a topical anesthetic allows the examiner to move the conjunctiva and its contained vessels with a cotton-tipped applicator, thus differentiating between vessels of scleral or conjunctival origin. The pain of scleritis is typically slower in onset but is often described as a severe “boring” pain that radiates to the ipsilateral forehead, cheek, or jaw. Engorgement of scleral vessels is usually more prominent and more diffuse than that of the episcleral vessels in episcleritis. A bluish hue may be seen as the underlying pigmented epithelium shows through the edematous, and hence more
Penetrating ocular trauma is evaluated by history (e.g., working with high-speed grinding equipment), examination (extrusion of aqueous humor or other globe content; direct visualization of a foreign body in the anterior chamber, vitreous, or retina), or identification of the offending object by biplanar plain radiography, thin-cut CT, or ultrasonography. MRI should not be used if there is any possibility that the foreign object may be metallic. Indirect indicators of globe penetration are hyphema, an irregularly shaped pupil from traction on or injury to the iris’ attachments, or lack of a red reflex. If penetrating ocular injury is confirmed or if the possibility persists after evaluation, an ophthalmologic consultation is indicated.4 Spontaneous or traumatic hyphema is often managed conservatively. Blood in the anterior chamber is usually the result of direct ocular trauma and may be associated with traumatic mydriasis or an obvious tear of the iris. If penetration and rupture can be reasonably excluded, the hyphema should be graded and IOP determined. Intraocular hypertension (or hypotension in the case of occult globe rupture) following trauma must also be evaluated by an ophthalmologist urgently. Inability to view posterior structures through the anterior blood may necessitate radiologic or ultrasonographic imaging.
Diagnostic Algorithm A recommended algorithmic approach to the patient with an acute red eye is provided in Figure 32-10.
233
Yes
No
Proptosis or external swelling?
Yes
No
Severe pain, FB sensation, or limbal injection?
Yes
No
Figure 32-10. Diagnostic algorithm for red eyes. *Indicates
Focal redness of bulbar conjunctiva?
potentially serious diagnoses if not identified on initial emergency department evaluation. †Purulent implies true pus, as opposed to the mucoid discharge more commonly associated with nonbacterial causes of conjunctivitis. a.k.a., also known as; FB, foreign body; incl, including. (Modified from Trobe JB: The Physician’s Guide to Eye Care. San Francisco, Foundation of the American Academy of Ophthalmology, 2001.)
Yes
1. Caustic keratoconjunctivitis*
2. Blepharitis 3. Chalazion 4. Dacrocystitis and dacroadenitis 5. Hordeolum (a.k.a. stye) 6. Inflammatory pseudotumor* 7. Orbital celluitis* 8. Orbital tumor* 9. Periorbital cellulitis or erysipelas 10. Retrobulbar abscess* 11. Retrobulbar emphysema* 12. Retrobulbar hematoma* 13. Keratitis* (incl abrasion and ulcer) 14. Keratoconjunctivitis 15. Episcleritis 16. Scleritis* 17. Anterior uveitis and hypopyon* 18. Acute angle-closure glaucoma* 19. Hyphema* 20. Endophthalmitis* 21. Inflamed pingueculum 22. Inflamed pterygium 23. Scleral penetration* 24. Subconjunctival hemorrhage
No
Purulent discharge?†
Yes
25. Bacterial conjunctivitis*
No
Itching sensation with or without other symptoms?
Yes
26. Allergic conjunctivitis
No
Airborne allergen, topical med, or cosmetics?
Yes
27. Contact dermatoconjunctivitis 28. Toxic conjunctivitis
No 29. Chlamydia conjunctivitis 30. Viral conjunctivitis
■ EMPIRICAL MANAGEMENT Irrigation Any clean water is appropriate for irrigation, and prompt initiation takes precedence over procurement of a particular irrigating solution. The most important principles are rapid and copious dilution and removal of the offending material. An eyewash station or faucet with tap water may be employed.
Normal saline may be instilled through the end of macrodrip intravenous administration tubing. If there is no gross eye injury, a Morgan lens may be attached to this tubing, but emergency department staff do not have to help the patient hold the eye open. Quickly administering two drops of topical anesthetic and allowing 30 seconds or so for the anesthetic to become effective greatly facilitates patients’ tolerance of the prolonged irrigation required. It is recommended that the first 500 to 1000 mL of irrigation fluid be
Chapter 32 / Red and Painful Eye
Caustic contamination?
234
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
administered while examining the eye; then the Morgan lens may be placed.
Pain Relief Pain often interferes with obtaining an adequate assessment. A topical anesthetic, such as proparacaine 0.5%, may facilitate cooperation in patients with possible injury or inflammation of the anterior eye by reducing pain and blepharospasm long enough to obtain a targeted history and focused examination. Topical anesthetic agents should not be given to patients to use at home. Parenteral or oral analgesics can be used for severe deep pain not amenable to topical relief in the emergency department, or for outpatient management of discomfort after discharge.
Mydriatic and Cycloplegic Agents Dilation of the pupil is not usually necessary in the emergency department for funduscopic examination, but may relieve pain associated with ciliary spasm in anterior uveitis. Mydriatic agents (e.g., phenylephrine, tropicamide) merely prevent constriction of the pupil by paralyzing the sphincter pupillae muscle of the iris. Cycloplegic agents (e.g., cyclopentolate, homatropine) paralyze the ciliaris muscle, with an accompanying mydriatic effect. The agent chosen should be guided by the desired length of time of mydriasis for the particular condition being treated (Table 32-1). Mydriatic agents are contraindicated in patients with narrow-angle glaucoma.
Antimicrobial Agents Most conjunctivitis is viral in origin, but it is often difficult to distinguish bacterial from viral types of conjunctivitis based solely on clinical grounds.15,16 Although no definitive empirical evidence dictates the use of antibiotic solutions or ointments for surface infections, the use of broad-spectrum topical antibiotics in cases of proven bacterial conjunctivitis is associated with benefit showing significantly higher clinical remission rates.17 Antimicrobial prophylaxis should be used for penetrating wounds to prevent bacterial keratitis or endophthalmitis. Antibiotics are typically used to treat identified or suggested bacterial infections, even if the exact bacterial agent has not been determined. The most common causes of bacterial conjunctivitis are nontypable Haemophylis influenzae, Streptococcus pneumoniae, and Staphylococcus aureus.18,19 Trachoma, a chronic keratoconjunctivitis caused by Chlamydia trachomatis, is the
most common infectious cause of blindness.20 Bacterial keratitis is usually seen in contact lens wearers, particularly those who wear them overnight.21 In descending order of frequency of cultured organisms, microbial keratitis is caused by Pseudomonas aeruginosa, streptococcal or staphylococcal species, filamentous fungi, nonpseudomonal gram-negative rods, Acanthamoeba, other bacteria, and yeast.22 The most common organisms cultured from deeper eye structures, particularly following open-globe injuries, are Bacillus cereus, Propionibacterium acnes, and various species of Bacillus, Streptococcus, and Staphylococcus.23,24 While awaiting emergent ophthalmologic consultation for possible vitrectomy, empirical parenteral antibiotic combinations include cefazolin plus gentamicin or vancomycin plus cefotaxime, ceftazidime, or ceftriaxone. Possible cases of mycotic endophthalmitis have historically been treated with amphotericin B,23 though voriconazole has been shown to have good intraocular penetration, broad-spectrum activity, and relatively low systemic toxicity.25 Open wounds also require tetanus prophylaxis, if the patient’s immunization status is not up to date. There is no current evidence supporting the practice of administering tetanus immunization to patients with superficial corneal abrasions.
Other Protective Interventions Significantly increased IOP must be reversed as rapidly as possible, often before the specific cause is known. After placing the patient in at least a 30° head-up position, two drops of timolol 0.5%, a topical beta-adrenergic blocking medication, should be administered as a first-line agent to decrease the production of aqueous humor. This may be followed by two drops of dorzolamide 2%, a topical carbonic anhydrase inhibitor, to reduce aqueous humor production further. If not available, 500 mg of acetazolamide may be given orally or intravenously. If the patient has sickle cell disease or trait, oral methazolamide 50 mg must be used instead. Patients with suggested intraocular hypertension who also have nausea or vomiting should receive a parenteral antiemetic so that they do not gag or vomit, which may further increase IOP.
Specific Management Management of the specific entities listed in the diagnostic algorithm presented in Figure 32-10 is presented in Table 32-2. Specific management of ophthalmologic conditions is also discussed in Chapter 69.
■ SPECIAL CONSIDERATIONS of Action for Common Mydriatic and Table 32-1 Duration Cycloplegic Medications NAME
CONCENTRATION (%)
COMMON DURATION
MAXIMUM DURATION
Ephedrine* Phenylephrine* Tropicamide Cyclopentolate Homatropine Scopolamine Atropine
5.0 2.5 0.5 0.5 1.0 0.5 0.5
0.5–1 hr 0.5–1 hr 3–4 hr 12–18 hr 1–2 days 2–5 days 5–10 days
3 hr 3 hr 6 hr 24 hr 3 days 7 days 14 days
*Mydriatic action only, no cycloplegic effect. Combination products such as Cyclomydril, which is cyclopentolate 0.2% and phenylephrine 10%, are also available.
Pediatrics A red eye in a neonate or infant is always abnormal. It is usually caused by corneal abrasion or infection. Corneal abrasions can also be a cause of inconsolable crying in an infant. Fluorescein examination helps to identify abrasions and herpes keratitis, acquired from the birth canal. Chlamydia infections may also be acquired during vaginal deliveries but may not arise for weeks. These infections should be treated with oral azithromycin as well as parenteral ceftriaxone to cover Neisseria gonorrhoeae. Conjunctivitis associated with respiratory symptoms or infiltrates on a chest radiograph in an infant younger than 3 months should be treated with an oral macrolide. Oral antibiotics are also indicated for conjunctivitis associated with otitis media. Mycoplasma is a common infectious agent in these cases, and a macrolide is indicated.14
5. Hordeolum (a.k.a. stye) Abscess in eyelash follicle or modified sebaceous gland at lid margin: external or internal based on side of lid margin that abscess is pointing. 6. Inflammatory pseudotumor* Nonspecific idiopathic retrobulbar inflammation with eyelid swelling, palbebral injection of conjunctiva, chemosis, proptosis, blurred vision, painful or limited ocular mobility, binocular diplopia, edema of optic disk, or venous engorgement of retina. 7. Orbital cellulitis* Eyelid swelling, redness and warmth of skin overlying orbit, tenderness of skin overlying bone palbebral injection of conjunctiva, and chemosis. Differentiated from periorbital cellulitis by presence of any finding of fever, ill appearance, blurred vision, proptosis, painful or limited ocular mobility, binocular diplopia, edema of optic disk, or venous engorgement of retina. 8. Orbital tumor* Blurred vision, proptosis or other displacement of globe, painful or limited ocular mobility, or binocular diplopia (but can be asymptomatic).
3. Chalazion Inflammation of meibomian gland causing subcutaneous nodule within the eyelid. 4. Dacrocystitis and dacroadenitis Eye tearing and inflammation of lower eyelid inferior to lacrimal punctum finding redness and tenderness over nasal aspect of lower lid and adjacent periorbital skin.
2. Blepharitis Inflammation of eyelid margins often a/w crusts on awakening, FB sensation, and tearing.
1. Caustic keratoconjunctivitis
Admit all cases of orbital cellulitis.
IOP > 20 mm Hg may be surgical emergency, Rx to decrease IOP in ED.
IOP > 20 mm Hg may be surgical emergency, Rx to decrease IOP in ED. Obtain blood cultures and start antibiotics. Axial and coronal CT of orbits and sinuses to r/o FB, retrobulbar abscess, orbital gas, subperiosteal abscess, osteomyelitis, and changes in cavernous sinus. Consider LP. IOP > 20 mm Hg may be surgical emergency, Rx to decrease IOP in ED. Ophthalmologist may want MRI, MRA, or orbital US.
Measure IOP. Evaluate for infection, diabetes mellitus, and vasculitis with CBC, BMP, UA, and ESR. Obtain axial CT of brain and axial and coronal CT of orbits and sinuses. Measure IOP. Start IV Rx with secondgeneration cephalosporin (e.g., cefuroxime, cefoxitin, or cefotetan) or with ampicillin/sulbactam to cover sinus and skin flora. Alternative Rx is ticarcillin/ clavulanate, piperacillin/tazobactam, vancomycin, or clindamycin + thirdgeneration cephalosporin (e.g., cefotaxime or ceftriaxone). Measure IOP. Evaluate for extraocular signs of malignancy. Obtain axial CT of brain and axial and coronal CT of orbits and sinuses.
Continued
Based on findings and discussion with consultant.
May discharge if no systemic problems, no findings of particular concern on CT, and IOP > 20 mm Hg. Start high-dose PO steroids after discussion with ophthalmologist and ensure reevaluation in 2–3 days.
Ophthalmologist may admit if systemically ill, case is moderate or severe, or no social support for patient. Ask about culturing before Rx if admitting, then Rx same as for periorbital cellultitis (#9). Outpatient referral only for treatment failure after 2 wk.
First r/o periorbital cellulitis (#9) and orbital cellulitis (#7). Inspect for obstruction of punctum by SLE, may express pus by pressing on sac, PO Rx for nasal and skin flora if not admitting. External: Warm compresses often all that is needed, may Rx anti-Staph ointment bid. Internal: PO Rx for β-lactamase Staph.
Outpatient referral only for treatment failure after 2 wk.
Discharge with instructions to apply warm compresses to eyelids for 15 min qid and scrub lid margins and lashes with mild shampoo on washcloth bid. Discharge with instructions to apply warm compresses to eyelids for 15 min and gently massage nodule qid. May discharge mild cases with PO analgesics and antibiotics (e.g., amoxicillin/clavulanate), and instructions to apply warm compresses to eyelids for 15 min and gently massage inner canthal area qid. Discharge with instructions to apply warm compresses to eyelids for 15 min and gently massage abscess qid.
May discharge only if tear film pH = 7 and no findings on examination except conjunctival injection and ophthalmologist can reevaluate next day.
DISPOSITION
Outpatient referral only for treatment failure after 2 wk.
Ophthalmologist must come to ED if there is any abnormal visual acuity or objective finding on examination after sufficient irrigation, with exception of expected injection of conjunctiva secondary to treatment.
CONSULTATION
None.
Immediate and copious irrigation with tap water or sterile normal saline until tearfilm pH = 7. Solids: lift particles out with dry swab before irrigation Acids: minimum of 2 L and 20 min Alkalis: minimum of 4 L and 40 min None except artificial tears for dry eye.
MANAGEMENT
Chapter 32 / Red and Painful Eye
DIAGNOSIS FROM FIGURE 32-10
Table 32-2 Management Algorithm for Red Eyes Extended from Diagnostic Algorithm in Figure 32-10
235
Keratitis (ulceration)* Symptoms and signs as above. Ulceration from complications of contact wear or neglected corneal abrasion has “scooped out” epithelium with surrounding edema appearing as white “cloudiness” in clear tissue. Keratitis (herpetic infection)* Symptoms and signs as above. Look for other signs of herpes, varicella, zoster (or CMV infection in immunocompromised patient). Look for “dendritic” defects of cornea with fluorescein under blue light.
12. Retrobulbar hematoma* Findings of pseudotumor (#6) but occurs due to trauma, coagulopathy, or thrombocytopenia and a/w diffuse subconjunctival hemorrhage anteriorly and extending posteriorly as well as increased IOP. 13. Keratitis (abrasion or UV injury) Pain, FB sensation, blepharospasm, tearing, photophobia, epithelial disruption on inspection under white light or fluorescein pooling under blue light. SPK appears as stippling of corneal surface [often lower 2/3 of cornea if due to light exposure]. First r/o corneal penetration either grossly or employing Seidel’s test. Relieve pain and blepharospasm with topical anesthetic. Inspect all conjunctival recesses and superficial cornea for any foreign material that can be removed by irrigation or manually lifted from surface. Tetanus prophylaxis is standard of care even if cornea not penetrated. Relieve pain and blepharospasm with topical anesthetic. Staph. and Strep. species still most common organisms, but Pseudomonas greater percentage in existing infections (especially contact lens wearer), so Rx with topical fluoroquinolone is preferred. Relieve pain and blepharospasm with topical anesthetic. Rx with trifluridine 1% solution, vidarabine ointment, or acyclovir ointment. Varicella-zoster and CMV not normally given antivirals if immunocompetent.
Abscess: Antibiotics as in orbital cellulitis (#7). Emphysema: Prophylax with antibiotics to cover sinus flora. Hematoma: Correct any coagulopathy or thrombocytopenia.
Discuss with ophthalmologist any potential need to débride or culture before starting antiviral.
Discuss any potential need to débride or culture before starting antibiotic.
Ophthalmologist must come to ED if there is any concern for globe penetration. Otherwise consult for follow-up examination in 1–2 days. One-time administration of cycloplegic agent may limit photophobia until follow-up examination.
Based on findings and discussion with consultant. Typical ciprofloxacin dosing is 1 gt. q 15 min for 1 hr, then 1 gt. q hr for 8 hr, then 1 gt. q 4 hr until seen by consultant next day. PO NSAIDs or narcotics for analgesia. No patch. Based on findings and discussion with consultant. Typical trifluridine dosing is 1 gt. q 2 hr for 7 days, then taper over 2 more wk. Typical vidarabine or acyclovir dosing is five times a day for 7 days, then taper over 2 more wk. PO NSAIDs or narcotics for analgesia. No patch.
May discharge cases not infected or ulcerated on topical antibiotic prophylaxis using polymyxin B combinations with bacitracin (ointment) or trimethoprim (solution). Gentamicin and sulfacetamide are less desirable single-agent alternatives. PO NSAIDs or narcotics for analgesia. Patching not necessary.
Admit all cases of retrobulbar pathology causing increased IOP. Others might be candidates for discharge depending on cause of problem.
IOP > 20 mm Hg may be surgical emergency, Rx to decrease IOP in ED. Obtain axial CT of brain and axial and coronal CT of orbits and sinuses.
Measure IOP unless possibility of ruptured globe. IOP > 30 mm Hg may require emergent needle aspiration or lateral canthotomy and cantholysis in ED.
11. Retrobulbar emphysema* Findings of pseudotumor (#6) but a/w increased IOP.
May discharge mild cases with PO antibiotics. Ophthalmologist must reevaluate next day to ensure no orbital extension.
Ophthalmologist may admit if systemically ill, case is moderate or severe, or no social support for patient.
First r/o orbital cellulitis (#7). PO Rx for sinus and skin flora if not admitting.
DISPOSITION
9. Periorbital cellulitis or erysipelas Eyelid swelling, redness and warmth of skin overlying orbit, tenderness of skin overlying bone, palbebral injection of conjunctiva, and chemosis. Differentiated from orbital cellulitis by absence of any other finding listed in #7. 10. Retrobulbar abscess* Findings of orbital cellulitis (#7) but a/w increased IOP.
CONSULTATION
MANAGEMENT
DIAGNOSIS FROM FIGURE 32-10
Table 32-2 Management Algorithm for Red Eyes Extended from Diagnostic Algorithm in Figure 32-10—cont’d
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
236
Based on findings and discussion with consultant, which primarily depends on speed of onset and response to treatment.
Discuss findings and use of prednisolone acetate 1% (frequency determined by ophthalmologist but range is q 1–6 hr). Discuss any IOP > 20 mm Hg with ophthalmologist.
First r/o glaucoma with IOP measurement. Rx in ED if IOP > 20 mm Hg. Otherwise OK to dilate pupil with 2 gtt. of cyclopentolate 1%. Decrease production of aqueous humor. Timolol 0.5% 1 gt., then repeat in 30 min. Apraclonidine 1% 1 gt. once. Dorzolamide 2% 2 gtt. or if sickle cell disease or trait then methazolamide 50 mg PO. Decrease inflammation. Prednisolone 1% 1 gt. every 15 min four times. Constrict pupil. Pilocarpine 4% 1 gt., then repeat in 15 min Consider establishing osmotic gradient Mannitol 2 g/kg IV. First r/o globe rupture. May require ultrasound if cannot visualize posterior structures. Measure IOP unless possibility of ruptured globe. IOP > 30 mm Hg may require acute treatment as in glaucoma (#18). If IOP > 20 mm Hg and no iridodialysis, may use cycloplegic to prevent iris motion.
19. Hyphema* Pain, decreased visual acuity, gross or microscopic blood in anterior chamber, may be a/w dilated and fixed pupil following blunt trauma. Graded by amount of blood Percentage of vertical diameter of anterior chamber when blood layers with patient in upright position. Microhyphema shows no layering and only suspended red blood cells.
Discuss findings and use of εaminocaproic acid and steroids, other medical therapy, best disposition, and follow-up examination by ophthalmologist within 2 days. Some patients may be admitted for observation, bedrest, head elevation, and frequent medication administration.
May discharge patient with medications recommended by ophthalmologist and ensure reevaluation in 2–3 days. Patients with hypopyon are generally admitted.
Discuss findings and use of topical or PO steroids.
Decrease inflammation with PO NSAIDs.
Rx in ED if IOP > 30 mm Hg.
May discharge patient with PO NSAIDs alone or in combination with topical ketarolac gtt. May discharge patient with medications recommended by ophthalmologist and ensure reevaluation in 2–3 days.
Outpatient referral only for treatment failure after 2 wk.
Relieve irritation with artificial tears and decrease inflammation with ketorolac gtt.
Continued
Most patients can be discharged with careful instructions to return for any increased pain or change in vision. Patients should decrease physical activity and sleep with an eye shield in place. Eyes should be left open while awake, so any change in vision can be immediately recognized. PO NSAIDs or narcotics for analgesia.
May discharge patient with medications recommended by ophthalmologist and ensure reevaluation in 2–3 days.
Discuss findings and use of prednisolone acetate 1% (frequency determined by ophthalmologist).
Treat for conjunctivitis by likely etiologic category (#25–30).
14. Keratoconjunctivitis Conjunctivitis with subepithelial infiltrates in cornea causing pain and decreased vision, possibly with halos reported. 15. Episcleritis Rapid onset of localized pain, injection of episcleral vessels, and localized tenderness. 16. Scleritis* Progressively increasing eye pain with radiation to ipsilateral face and decreasing vision, photophobia, tearing, and possible pain with eye motion. 17. Anterior uveitis and hypopyon* Eye pain, photophobia, tearing, limbal injection of conjunctiva, and cells or flare in anterior chamber. Hypopyon is layering of white cells (pus) in anterior chamber. 18. Acute angle-closure glaucoma Sudden-onset eye pain and blurred vision that may be a/w frontal headache, nausea, and vomiting. Anterior eye may manifest shallow or closed angle between iris and cornea, pupil fixed in mid-dilation, or limbal injection of conjunctiva.
DISPOSITION
Chapter 32 / Red and Painful Eye
CONSULTATION
MANAGEMENT
DIAGNOSIS FROM FIGURE 32-10
237
Ophthalmologist must come to ED if there is any concern for globe penetration. None required if no complications.
Protect eye from further pressure, provide pain relief, and prevent vomiting. Tetanus prophylaxis. Exclude coagulopathy or thrombocytopenia, if indicated by history. Topical polymyxin B trimethoprim in infants and children, because more Staph. spp. Topical sulfacetamide or gentamicin clinically effective in 90% of uncomplicated adult cases. Use topical fluoroquinolone if Pseudomonas possible.
Outpatient referral only for severe cases or treatment failure after 2 wk. Same as #27 Culture drainage and consult ophthalmology in all neonates and those at risk for vision loss or systemic sepsis. Culture drainage and consult ophthalmology in all neonates and those at risk for vision loss or systemic sepsis.
Irrigation with tap water or sterile normal saline. Decrease irritation with naphazoline gtt. Same as #27 Rx PO azithromycin for Chlamydia. Consider parenteral ceftriaxone for concurrent Neisseria gonorrhoeae. Decrease irritation with artifical tears, naphazoline, or ketorolac gtt.
Outpatient referral only for treatment failure after 2 wk.
Decrease irritation with naphazoline gtt.
Ask about pregnant mothers, infants, and immunocompromised individuals in close contact. Discharge uncomplicated cases with instructions on respiratory and direct-contact contagion for 2 wk.
Discharge uncomplicated cases on 5 days of PO azithromycin.
Identify offending agent and avoid subsequent exposure. Discharge uncomplicated cases on continued naphazoline. Same as #27
Identify antigen if possible. Consider treating other allergic symptoms with PO antihistamines.
Reassure patient that red should resolve over 2–3 wk. Discharge uncomplicated cases with 10 days of topical antibiotics in both eyes, regardless of laterality of apparent infection. Use ointments in infants and gtt. in others.
a.k.a., also known as; a/w, associated with; bid; twice daily; BMP, basic metabolic profile (includes electrolytes, glucose, and renal function tests); CBC, complete blood count; CMV, cytomegalovirus; CT, computed tomography; ED, emergency department; ESR, erythrocyte sedimentation rate; FB, foreign body; gt., drop; gtt. drops; IOP, intraocular pressure; LP, lumbar puncture; MRA, magnetic resonance angiography; MRI, magnetic resonance imaging; NSAIDs, nonsteroidal anti-inflammatory drugs; PO, oral; q, every; qid, 4 times a day; r/o, rule out; Rx, prescribe; SLE, slit-lamp examination; SPK, superficial punctuate keratitis; spp, species; Staph., Staphylococcus; Strep., Streptococcus; UA; urinalysis; US, ultrasonography; UV, ultraviolet.
*Potentially serious diagnoses if not identified on initial emergency department evaluation. Antibiotic choices should be based on current practice.
28. Toxic conjunctivitis Diffuse conjunctival injection, chemosis, and lid edema. 29. Chlamydia conjunctivitis Often bilateral palpebral injection of conjunctiva in neonate or other individual at risk for sexually transmitted disease. 30. Viral conjunctivitis Often bilateral palpebral injection of conjunctiva and follicular cobblestoning of inner surface of lids. Inflammation of eyelid margins often a/w crusts on awakening, FB sensation, and tearing.
Same as #21
Same as #21
Culture drainage and ophthalmology consult in all neonates and those at risk for vision loss or systemic sepsis. Neisseria gonorrhoeae can be rapidly sight-threatening.
Discharge to follow-up with ophthalmologist for possible steroid therapy or surgical removal. Same as #21
Outpatient referral only for treatment failure after 2 wk.
Admit for IV antibiotics and possible procedural intervention.
Admit all cases of endophthalmitis.
DISPOSITION
Ophthalmologist must admit for parenteral and possibly intraocular antibiotics.
20. Endophthalmitis* Progressively increasing eye pain and decreasing vision, diminished red reflex, cells and flare (and possibly hypopyon) in anterior chamber, chemosis, and eyelid edema. 21. Inflamed pingueculum Inflammation of soft yellow patches in temporal and nasal edges of limbal margin. 22. Inflamed pterygium Inflammation of firmer white nodules extending from limbal conjunctiva onto cornea. 23. Scleral penetration* Localized redness at site of entry, teardrop pupil, blood in anterior chamber or loss of red reflex. 24. Subconjunctival hemorrhage Red blood beneath clear conjunctival membrane. 25. Bacterial conjunctivitis* Hyperpurulent discharge not typical of common “pink eye” and more commonly unilateral in adults. Inflammation of eyelid margins a/w lid edema, chemosis, and possibly subconjunctival hemorrhage, but usually no follicular “cobblestoning.” 26. Allergic conjunctivitis Often bilateral palpebral injection of conjunctiva and follicular cobblestoning of inner surface of lids that may be seasonal and a/w other allergic symptoms such as rhinitis. 27. Contact dermatoconjunctivitis Localized lid and conjunctival redness and edema.
CONSULTATION
Empirical parenteral antibiotic administration with cefazolin + gentamicin or vancomycin + cefotaxime, ceftazidime, or ceftriaxone to cover Bacillus, enterococcus, and Staphylococcus spp. Decrease inflammation with naphazoline or ketorolac gtt.
MANAGEMENT
DIAGNOSIS FROM FIGURE 32-10
Table 32-2 Management Algorithm for Red Eyes Extended from Diagnostic Algorithm in Figure 32-10—cont’d
PART I ■ Fundamental Clinical Concepts / Section Two • Cardinal Presentations
238
239
Trauma
KEY CONCEPTS ■
Prompt and prolonged irrigation is advised for patients who experience caustic injury to the eye. ■ Headache and nausea may be prominent symptoms in patients with acute angle-closure glaucoma. ■ Keratitis, inflammation of the cornea, is most commonly caused by a viral infection, but may also be caused by recent ultraviolet light exposure, chemical injury, or hypoxic injury from contact lens use. ■ A localized corneal defect with edematous, inflammatory changes may signal corneal ulceration. ■ A corneal dendritic pattern may signal a herpetic infection, which can progress to corneal opacification and visual loss. ■ Pain, consensual photophobia, perilimbic conjunctivial infection, and a miotic pupil that is caused by ciliary spasm could signal iritis, which is inflammation of the iris and ciliary body, or uveitis, inflammation of the iris, ciliary body, and also choroids. The cause may be trauma or underlying autoimmune disease. The presence of cells and flare in the anterior chamber can help signal these conditions.
■ DISPOSITION Most emergency department patients with eye complaints are candidates for discharge and, if indicated, follow-up in the
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 32 / Red and Painful Eye
Blunt trauma is a common cause of a red and painful eye. Large hyphemas and those with clots are likely to require hospitalization for bedrest with 30° of head elevation. Systemic analgesia and, if required, antiemetics are indicated. Medications affecting platelet function should be avoided. Treatment may be indicated when the IOP exceeds 30 mm Hg, as it is in acute angle-closure glaucoma. If the iris is not injured, a long-acting cycloplegic agent (e.g., topical homatropine) may be recommended to prevent repetitive motion of the iris. Some reliable adult patients may be discharged with daily follow-up by a specialist. Strong analgesia and patching are not indicated, so that the patient may immediately identify increases in pain or decreases in visual acuity. Corneal abrasions are common problems in the emergency department. When the emergency physician is convinced that the cornea has not received a full-thickness laceration or penetration by a foreign body, management is relatively simple. Foreign bodies (on or in the epithelium only) should be removed when possible. These may frequently adhere to a saline-moistened cotton-tipped applicator. Ones that do not may sometimes be lifted off with a blunt-tipped tool (“spud”) under the binocular magnification of a slit lamp. The common use of hypodermic needle removal may damage surrounding cornea and is not recommended. Whether or not the object can be successfully removed, management is the same as for corneal abrasions. Rust staining of the corneal epithelium does not require removal in the emergency department, but patients are referred to a specialist for examination within 3 days. Prophylactic topical antibiotics are indicated for all epithelial defects of the cornea. Patching is not necessary and may be harmful. Systemic analgesia appropriate to the patient’s level of pain should be provided. Larger lesions may require a prophylactic mydriatic or cycloplegic agent anticipating a secondary iritis. Topical anesthetics should not be given to the patient for home use.4
emergency department or with an ophthalmologist in 1 to 2 days. Others may require referral only if there is lack of resolution or treatment fails. A few patients require admission for procedural intervention, parenteral antibiotic regimens, management of intractable pain, or further diagnostic evaluation. General consultation and disposition considerations for the most important entities are outlined in Table 32-2.
PART II
Trauma
•
Section One General Concepts Chapter 33
Multiple Trauma
Eric Gross and Marc Martel
■ INTRODUCTION The care of the injured patient remains one of the mainstays of emergency medicine practice. Emergency physicians play a vital role in the stabilization and diagnostic phases of trauma care. Management of these patients involves complex, timedependent decision-making, leadership capability, and technical skill. Proper resuscitation can lead to functional outcomes, even in severely injured patients.1
■ EPIDEMIOLOGY In 2004, there were 167,184 injury-related deaths, of which 73% were motor vehicle–related, firearms-related, or caused by poisonings or falls. Unintentional accidents were the leading cause of death in people ages 1 to 44 years. Motor vehicle collisions (MVCs) made up the largest percentage of those accidents, accounting for 26% of all injury-related deaths.2 The number of motor vehicle deaths has remained relatively stable for the past decade. The number of people injured in motor vehicle crashes, however, has declined 22% during the same time period, to 2.7 million in 2005.3 Homicide is one of the top five leading causes of death in people ages 1 to 44 years, with firearm injuries accounting for 17.7% of all injury deaths in 2004.2 The economic cost of traumatic injuries is staggering. It is estimated that the total cost of injuries that occurred in 2000 is $406 billion; this includes medical costs and lost productivity. Motor vehicle and fall injuries account for 22% ($89 billion) and 20% ($81 billion) of this total, respectively.4 Many of these injuries are avoidable. Proper use of lap/ shoulder belts can reduce the risk of fatal injury by 45%.5 Yet, in 2005, it is estimated that 55% of MVC occupant fatalities were unrestrained; this percentage increases to 65% for the 21- to 24-year-old age group.5 Child safety seats reduce the risk of death in passenger cars by 71% for infants and by 54% for toddlers ages 1 to 4 years.6 Educational and law enforcement initiatives addressing seat belts, proper child restraint, drinking and driving, gun safety, and fall prevention can assist in raising public awareness. The National Highway Traffic Safety Administration’s (NHTSA) “Click-it or Ticket” campaign increased belt use in 41 of 50 states and the District of Columbia during a 2-month time period.7 NHTSA has a similar program aimed at drunk driving called “Over the limit. Under arrest.” The effect of firearm laws on decreasing firearm vio-
lence is less concrete (even though a decrease has been seen).8 Further study is needed to determine the impact of legislation, public education, and prevention programs on firearm violence.
■ TRAUMA SYSTEMS The first document to set criteria for categorizing hospitals as trauma centers was promulgated by the American College of Surgeons (ACS) Committee on Trauma in 1976.9 As other groups recognized the importance of structured trauma care, legislation and funding to promote the development of trauma systems grew. In the early 1990s, the Health Resources and Services Administration developed the Model Trauma Care System Plan, a well-designed framework for progress measurement in trauma systems. Unfortunately, this program lost funding in 2006. As of 1998, 38 states and the District of Columbia had at least one critical element in place for a formal trauma system.10 More up-to-date data are not readily available because the program was the main source for this information. The benefit of regionalized trauma systems has been shown in multiple studies. A meta-analysis of 14 studies demonstrated an overall 15% decline in mortality due to the presence of a trauma system.11 However, this apparent decline may be confounded by other factors. A nationwide study suggests that mortality reduction could not be solely attributed to the presence of a trauma system because its impact was small and statistically not significant. Rather, the presence of a primary seat belt law and mean per capita income were associated with a reduction in occupant mortality rates, whereas rural population and speed limits faster than 65 mph were associated with an increase in mortality rates.12 As new trauma systems mature, more research will be needed to guide implementation of new system strategies to further reduce morbidity and mortality from traumatic injury. One goal of the out-of-hospital trauma system is to transport the patient to the closest appropriate facility in a timely manner. Problems with over- and undertriage occur. Most efforts are aimed at reducing undertriage (transport of severely injured patients to lower level trauma centers), which may result in preventable morbidity and mortality from delay in definitive care. Overtriage (transport of minimally injured patients to higher level trauma centers) has no deleterious effects on patient care; however, it may contribute to unneces243
PART II ■ Trauma / Section One • General Concepts
244
sary resource utilization and potential overcrowding in tertiary care, level 1 trauma centers. The ACS published a field triage decision scheme to assist in appropriate transport decisions (Fig. 33-1).10 Limitations to effective use of regionalized trauma systems remain. Hospital crowding, ambulance diversion, lack of specialist on-call availability, and reimbursement issues all contribute. In addition, from 1995 to 2005, the number of hospital emergency departments declined from 4176 to 3795.13 Trends such as this increase the burden of emergency care on those remaining hospitals, potentially compromising care. Future funding and legislative decisions at the state and national levels will have a significant impact on the future of quality trauma care in the United States.
■ PRINCIPLES OF DISEASE The emergency physician faces significant clinical uncertainty when a multiple trauma patient presents to the emergency department. Much of the diagnostic dilemma and subsequent evaluation can be directed based on knowledge of the mechanism of injury. Although mechanisms of injury alone are not good predictors of major trauma,14 common patterns of injuries can be anticipated and specifically assessed in emergency department patients. Table 33-1 outlines several blunt trauma mechanisms of injury with potential associated clinical findings. Basic anatomic principles are useful in the assessment of patients with penetrating trauma. In contrast to penetrating trauma from knife wounds, in which injuries can be expected along the track of the weapon, gunshot wounds depend on several factors. The amount of tissue damage is related to the kinetic energy of the bullet imparted to the patient. The bullet weight (caliber) and velocity (determined by the weapon) play a role in anticipating injuries. Gunshot wounds result in trauma to the surrounding tissue by direct laceration, crush injury, shock waves and cavitation—the displacement of tissue forward and radially. Because of these dynamic forces, the emergency physician should anticipate more widespread injuries from high-velocity weapons, such as rifles, than from lowvelocity weapons, such as handguns. Similar to knife wounds, handguns generally cause injury based on direct laceration and crush generated by the missile along its track. Shotgun wounds from close range are characterized by massive tissue injury. Injury patterns can differ significantly between adults and children subjected to similar mechanisms of trauma. The major anatomic distinctions relate to the smaller size and surface area, larger head-to-body ratio, and less protected abdominal cavity of the child. As a result, children are more vulnerable to multisystem injury in blunt trauma, more frequently sustain significant head and intra-abdominal injuries, and are more at risk for hypothermia.15-18 Trauma is the seventh leading cause of death in patients older than the age of 65 years.19 Elder patients commonly sustain extremity, craniofacial, and closed head injuries. The majority of these occur as the result of a fall or an MCV. Elder trauma patients typically have normal, age-related changes in organ system function related to decreased cardiopulmonary functional reserve, decreased renal function, decreased bone density, and cerebral atrophy. These changes can increase susceptibility to sheer forces and other aspects of trauma.20 Comorbidities and preexisting medication use further complicate the management of elder trauma patients. Lower extremity weakness, gait disturbances, decreased visual acuity, and the use of psychotropics, antihypertensives, and sedatives have been associated with falls in elders, resulting in major injury.21 The use of these medications, particularly antihyper-
tensives, should not be considered causative in trauma patients with hypotension until acute hemorrhage is assessed and managed. In addition, anticoagulants, antiplatelet drugs, and aspirin are commonly prescribed, and their effects should be suspected and reversed if possible in elder trauma patients.
■ MANAGEMENT Out-of-Hospital Management of the trauma patient frequently begins prior to arrival in the emergency department by first responders. The goals of out-of-hospital care include intervening in immediately life-threatening injuries, preventing additional injury, and rapid transport to trauma centers for definitive care. Although accepted as tenets of out-of-hospital care, controversy exists regarding each of these goals. The majority of life-threatening injuries that require intervention by out-of-hospital providers are related to airway, breathing, and circulation (the ABCs). Preventing aspiration of gastric contents and providing adequate tissue oxygenation are the primary goals of endotracheal intubation. Although controversy exists regarding the use of out-of-hospital rapid sequence induction,22-27 securing an unprotected airway is essential in this phase of trauma management. Tension pneumothorax is the fundamental threat to adequate ventilation and requires immediate needle thoracostomy. Systemic hypotension with impaired end-organ perfusion mandates treatment in the trauma patient, despite the debate surrounding controlled hypotension versus aggressive fluid resuscitation. Preventing additional injury requires an awareness of not only clinically evident abnormalities but also potentially more serious injuries. Coordinated extrication and transport with rigid cervical immobilization, complete spinal precautions, intensive hemodynamic monitoring, and stabilization of fractures to prevent neurovascular compromise are examples of assuming the most serious injuries exist in multiple trauma patients. In the United States, rapid transport to the nearest appropriate facility is one of the fundamental concepts in trauma management. Much of the controversy regarding various out-of-hospital approaches to the ABCs is rooted in attempts to limit transport times and avoid further infringement on the “golden hour” of trauma care. In contrast, physician-operated emergency medical service (EMS) systems more aggressively manage airway and ventilatory issues and are more likely to commit out-of-hospital time resources to establishing hemodynamic stability prior to transport.28,29 Rural EMS systems in the United States, where transport times may be prolonged because of the distance to a receiving facility, may benefit from more advanced interventions such as rapid sequence induction/intubation and more aggressive fluid resuscitative measures.
Emergency Department General Principles Care of the multiple trauma patient is complex and involves the coordination of multiple providers, including EMS personnel, emergency physicians, nurses, technicians, trauma surgeons, and subspecialists. A systematic and comprehensive approach to these patients is necessary, incorporating providers from each discipline. Advanced Trauma Life Support (ATLS) guidelines delineate the use of defined trauma response teams, with providers performing assessments, diagnostics, and interventions simultaneously. With this approach,
245 Measure vital signs and level of consciousness.
Glasgow Coma Scale Systolic blood pressure, mm Hg Respiratory rate, /min
20 mph) impact • Motorcycle crash >20 mph
Yes
No
Transport to closest appropriate trauma center, which, depending on the trauma system, need not be the highest level trauma center.
Assess special patient or system considerations.
• Age
Step 4
• Older adults: risk of injury/death increases after age 55 • Children: Should be triaged preferentially to pediatric-capable trauma centers • Anticoagulation and bleeding disorders • Burns • Without other trauma mechanisms: triage to burn facility • With trauma mechanism: triage to trauma center • Time-sensitive extremity injury • End-stage renal disease requiring dialysis • Pregnancy >20 weeks • EMS provider judgment
Yes
No
Contact medical control and consider transport to trauma center or a specific resource hospital.
Transport according to protocol.
When in doubt, transport to a trauma center.
Figure 33-1. Triage decision scheme. Redrawn from American College of Surgeons, Committee on Trauma: Resources for the Optimal Care of the Injured Patient. Chicago, American College of Surgeons, 2006.
Chapter 33 / Multiple Trauma
Step 1
246
PART II ■ Trauma / Section One • General Concepts
Table 33-1 Blunt Trauma Mechanisms and Associated Injuries MECHANISM OF INJURY
ADDITIONAL CONSIDERATIONS
Motor vehicle collisions Head-on collision Rear-end collision Lateral (T-bone) collision
Rollover
Windshield damage
Greater chance of ejection Significant mechanism of injury Likely unrestrained Significant mortality Likely unrestrained
Steering wheel damage
Likely unrestrained
Ejected from vehicle
Dashboard involvement/damage Restraint/seat belt use Proper three-point restraint Lap belt only Shoulder belt only Airbag deployment
Decreased morbidity
Front-end collisions Less severe head/upper torso injuries Not effective for lateral impacts More severe injuries in children (improper front seat placement)
Pedestrian versus automobile Low speed (braking automobile) High speed
Closed head injuries, coup and countercoup injuries Facial fractures Skull fractures Cervical spine fractures Thoracic injuries Sternal and rib fractures, flail chest Cardiac contusion Aortic injuries Hemo/pneumothoraces Pelvic and acetabular injuries Dislocated hip Sternal and rib fractures, pulmonary contusions Chance fractures, abdominal injuries, head and facial injuries/fractures Cervical spine injuries/fractures, “submarine” out of restraint devices (possible ejection) Upper extremity soft tissue injuries/fractures Lower extremity injuries/fractures
Closed head injuries “Handlebar” injuries Spleen/liver lacerations Additional intra-abdominal injuries Consider penetrating injuries Extremity injuries “Handlebar” injuries
Nonautomobile related
Horizontal impact
Facial injuries Lower extremity injuries Aortic injuries Hyperextension injuries of cervical spine Cervical spine fractures Central cord syndrome Thoracic injuries Abdominal injuries—spleen, liver Pelvic injuries Clavicle, humerus, rib fractures Crush injuries Compression fractures of spine Spinal injuries
Tibia and fibula fractures, knee injuries Waddle’s triad—tibia/fibula or femur fractures, truncal injuries, craniofacial injuries “Thrown” pedestrians at risk for multisystem injuries
Bicycle Automobile related
Falls Vertical impact
POTENTIAL ASSOCIATED INJURIES
LD50 36–60 ft
Calcaneal and lower extremity fractures Pelvic fractures Closed head injuries Cervical spine fractures Renal and renal vascular injuries Craniofacial fractures Hand and wrist fractures Abdominal and thoracic visceral injuries Aortic injuries
247
American College of Surgeons Requirements Major Resuscitations
A surgeon should be present in the emergency department on trauma patient arrival or within 15 minutes if any of the following major criteria are found: Confirmed hypotension (systolic blood pressure < 90 mm Hg) Respiratory compromise requiring intubation Penetrating gunshot wound to the neck, chest, abdomen, or pelvis Glasgow Coma Scale score of 92%) and limiting the plateau airway pressures to 35 cm H2O or less to reduce the incidence of barotrauma.47 This may require allowing the partial pressure of carbon dioxide (Pco2) levels to rise to 35 to 55 mm Hg (permissive hypercapnia) as long as the pH remains greater than 7.25. A recent randomized trial found that low tidal volumes (6 mL/kg of predicted weight) reduced mortality.48 Therefore, starting tidal volumes should be set at 6 to 8 mL/kg of predicted body weight. Larger tidal volumes may be required if oxygenation is not adequate. Positive end expiratory pressure may also help support oxygenation. In order to reduce hyperoxic lung injury, oxygen concentrations should be titrated to the lowest concen trations that maintain adequate oxygenation. A sample of initial ventilator settings is presented in Table 60-4. Highfrequency percussive ventilation may be used to obtain ade quate oxygenation while minimizing airway pressures. This mode of ventilation administers high-frequency, time-cycled, pressure-limited subtidal volumes, which may reduce the incidence of barotrauma.49,50 Patients with inhalation injury may develop bronchospasm, and therefore bronchodilators should be used in patients with wheezing. Bronchodilators may improve mucociliary function. Frequent airway suctioning and chest physiotherapy are also helpful at removing secretions. Aerosolized N-acetylcysteine with or without aerosolized heparin has also been shown to help break down thick mucous secretions.51,52 An example of a treatment protocol for patients with inhalation injury may include 5000 to 10,000 units of heparin and 3 mL of normal saline nebulized every 4 hours, alternating with 3 to 5 mL of 20% N-acetylcysteine.45
Circulation and Fluid Resuscitation Fluid resuscitation is one of the most important elements in the treatment of burn patients. Prior to World War II, many burn patients died of hypovolemic shock and renal failure. The first formal fluid resuscitation protocol was developed treating the burns that occurred during the Cocoanut Grove nightclub fire in Boston in 1942.53 Since then, many fluid resuscitation regi mens have been described; however, the Parkland formula described by Baxter remains the most popular.54 Burn injury results in the activation of the compliment system and the release of a large number of inflammatory mediators such as histamine, prostaglandins, and leukotrienes. These mediators increase the permeability of the local and systemic vasculature, resulting in the extravasation of intravas cular fluids and proteins into the interstitial space, contributing to fluid depletion and soft tissue edema.55 The leakage of plasma proteins during the first 3 to 5 hours reduces the intra vascular oncotic pressure and increases the interstitial oncotic pressure, which leads to an increase in edema formation.56 With large burns, half of the fluid requirements are due to extravasation into unburned areas.57
Table 60-4 Recommended Initial Ventilator Settings Tidal volume Respirator rate Plateau pressures I/E ratio Flow rates PEEP
6–8 mL/kg 8–12 in adults 12–45 in children 1 cranial bone Occipital
Figure 63-5. Metaphyseal fracture, “bucket handle type,” associated with child physical abuse. (Courtesy of Sara T. Stewart, MD, MPH.)
Figure 63-4. Complex skull fracture associated with child maltreatment. (Courtesy of Sara T. Stewart, MD, MPH.)
Skeletal injuries may involve any of the bones in the body. Although approximately 42% of boys and 27% of girls have sustained a fracture before the age of 16 years, certain fractures are highly suspicious of an inflicted injury (Table 63-3).14 In particular, metaphyseal fractures, rib fractures, and certain types of skull fractures should raise concern about inflicted trauma (Fig. 63-4). Metaphyseal fractures (i.e., classic metaphyseal lesions) occur because of yanking or pulling on an extremity. On radiographs, they may appear as chips or what is referred to as a bucket-handle injury of the long bone (Fig. 63-5). They usually are noted in children younger than 2 years. Rib fractures, particularly posterior rib fractures in a small infant, are virtually pathognomonic for inflicted injury15 (Fig. 63-6). The rib cage, because of its archlike structure, must be subjected to a good deal of force to cause disruption. Rib fractures are extremely uncommon as a consequence of cardiopulmonary resuscitation.16 The presence of multiple skull fractures and skull fractures in the occipital region are unusual in cases of accidental trauma and should raise concern for inflicted trauma. Fractures that are diastatic (>3 mm in width), grow in size, or involve more than one cranial bone are less specific as markers for nonaccidental trauma; however, they are usually associated with mechanisms of injury that involve greater traumatic forces. Head injuries, including those associated with the classically described “shaken baby syndrome,” account for most child abuse–related fatalities. The syndrome includes evidence of head trauma in association with retinal hemorrhages and skeletal injuries and occurs generally in infants younger than 1 year but may be seen in children up to 3 years old. Often,
Figure 63-6. Posterior rib fractures caused by compression of the ribs from child physical abuse. (Courtesy of Sara T. Stewart, MD, MPH.)
there is no evidence of an impact injury to the head, such as a scalp hematoma or skull fracture.17 The impact may be against a soft or compressible surface, such as the mattress of the crib. Such an impact results in a rapid deceleration of the head, and the brain experiences a coup-contrecoup injury by moving back and forth within the confines of the skull. Shaking of an infant or young child subjects the brain to rotational acceleration, which is capable of generating greater force and speed than linear acceleration, the type of acceleration that occurs with a fall. It has been postulated that severe repeated shaking of an infant or young child can lead to disruption of
Child Sexual Abuse Physical findings present when a child or adolescent has been sexually abused depend on the nature of the abuse, the time since the abuse, and whether the abuse was repetitive or isolated. Acute injuries include disruptions (tears) of the hymen, petechiae, hematomas, or rarely vaginal tears. The prepubescent hymen is significantly more fragile and more easily traumatized than the postpubertal hymen, which thickens and becomes redundant under the influence of estrogen. Physical changes in the anogenital area also may be noted when there has been prior or recurrent sexual abuse. Changes include loss of hymenal tissue and the appearance of U-shaped disruptions in the hymenal contour. Anal findings include the presence of scars and changes in anal tone and the anal contour.23 Evidence of recent trauma also may be visible in the anal area. Acutely, there may be lacerations that appear as perianal fissures, which are characteristically wider distal to the anus. There may be post-traumatic dilatation, or alternatively anal spasm may occur in response to submucosal injury. In abused boys, the penis
795
rarely has a noticeable injury. More recent studies have evaluated genital healing that occurs after traumatic injury.24
■ CLINICAL FEATURES Signs and Symptoms Child Physical Abuse Children who have been physically abused may have complaints related to the abuse, or injuries may be noted during the course of an evaluation for an unrelated medical condition. Infants with head injuries may present with nonspecific symptoms that go unrecognized as being related to inflicted head trauma. These symptoms may include apnea, altered mental status, an apparent life-threatening event, vomiting, or a seizure. Clinicians should remain alert that these symptoms may be related to intracranial bleeding or elevated intracranial pressure. A careful physical examination should include an attempt to visualize the eye grounds to determine if there is any evidence of retinal hemorrhages. Bruises on a young infant’s face are suspicious for head injury. Refusal to use an extremity or to bear weight may be an indication of a fracture. Similarly, swelling of an extremity may be a sign of a skeletal injury. Children with abdominal injuries may present with abdominal pain, vomiting, abdominal distention, or shock. In cases of inflicted trauma, the history may be unrevealing or may be inconsistent with the medical findings. It is common for a parent to state that he or she is uncertain how a child sustained the injury, or that the child was well at bedtime and awoke in the morning refusing to walk or with severe abdominal pain. Caregivers should be queried about how they think the injury was sustained. In cases of inflicted head trauma, the scenario often includes a young infant left in the care of a male friend or partner of the mother who has gone to work or to run errands. The male companion asserts that the infant was fine, sleeping in the crib when he went to take a shower or make some coffee, and when he returned to check on the infant, the infant was not breathing or was seizing. Often the mother is called before calling 911, or the infant is scooped up by the individual and carried to a hospital. The part of the history that is omitted is that the infant was crying, and the companion shook the infant and thrust the infant back into the crib. The infant sustained acute traumatic axonal injury and stopped crying. The infant is then left unattended, seemingly asleep. As bleeding and cerebral edema develop, other symptoms intervene. Sometimes a family member fallaciously relates a history of a fall, usually from a height (8 years of age: azithromycin 1 g PO Chlamydia (1 dose), or doxycycline, 100 mg PO bid × 7 days 45 kg: azithromycin 1 g PO (1 dose) Metronidazole 500 mg PO bid × 7 days Bacterial vaginosis or metronidazole gel 0.75% 5 g intravaginally daily × 5 days or clindamycin 300 mg PO bid × 7 days Metronidazole 2 g PO (or 500 mg PO bid Trichomonas × 7 days) Syphilis Benzathine penicillin 2.4 million units IM (1 dose) Acyclovir 400 mg PO tid × 7–10 days, or HSV (first clinical valacyclovir 1 g PO bid × 7–10 days episode) HBIG 0.06 mg/kg IM vaccine series Hepatitis B* HIV Contact local infectious disease specialist bid, twice daily; HBIG, hepatitis B immunoglobulin; HIV, human immunodeficiency virus; HSV, herpes simplex virus; tid, three times daily. *Unimmunized child and perpetrator with acute hepatitis B infection. Adapted from Workowski KA, Berman SM: Sexually transmitted diseases treatment guidelines, 2006. MMWR 55:1, 2006.
times a child or adolescent needs to be questioned about the alleged abusive events.
■ DIFFERENTIAL CONSIDERATIONS Child Physical Abuse The major differential diagnosis when considering child abuse is unintentional injury. Differentiating between inflicted and noninflicted injuries requires the consideration of multiple factors, including the developmental stage of the child, the extent of the injuries, whether the injuries appear to have occurred over a period of time, whether there were witnesses to the alleged event, whether medical care was sought in a timely manner, and whether the injuries could have been sustained in the stated manner. It is important for the evaluating clinician to have an understanding of normal child development, particularly the acquisition of motor skills. Children who are ambulatory, particularly toddlers and young schoolage children, are prone to bruises over bony prominences, such as the shins and forehead. Noninflicted bruises are usually unilateral, occurring on the side where a fall or collision with a solid object has occurred. Bruises on the buttocks and flank must be explained by the mechanism of injury and are unusual in noninflicted trauma.
Chapter 63 / Child Maltreatment
rib injuries, may not be readily visible. Repeat radiography in 1 to 2 weeks shows evidence of callus formation and makes the fracture more readily appreciated. Alternatively a radionucleotide bone scan can detect subtle injuries and should be obtained if there is a skeletal injury but a negative skeletal survey. Urinalysis, liver function studies, and serum amylase and lipase levels should be considered in children with symptoms of abdominal injury, such as vomiting, abdominal pain, or guarding.28 Plain x-rays are rarely helpful but should be considered if clinical findings suggest perforation or obstruction. Computed tomography scan of the abdomen is a more precise way of delineating any abdominal injury. Computed tomography scan of the head is indicated when symptoms are consistent with head trauma or in an infant with facial bruising. Computed tomography scans may reveal extra-axial hemorrhage or findings consistent with cerebral edema. If the child is sufficiently stable, magnetic resonance imaging is helpful in determining the age of the hemorrhages and in assessing whether there has been prior intracranial hemorrhage. In recent years, certain metabolic disorders have been described that also may be associated with intracranial hemorrhage. In particular, glutaric aciduria type I may be mistaken for inflicted head trauma.29 Urinalysis for organic and amino acids would detect this condition. A careful ophthalmologic examination is crucial. The examination is best done by a pediatric ophthalmologist. Photographic recording of the retinal findings is an important part of the diagnostic workup. Photographing cutaneous injuries is also important and frequently can be carried out by a police criminalist, whose equipment includes color bars for accurate documentation of the coloration of the bruising.
PART II ■ Trauma / Section Five • Violence and Abuse
798
Figure 63-7. Mongolian spots in an infant. (Courtesy of the EMSC Slide Set, National EMSC Resource Alliance.)
Mongolian spots are bluish discolorations that are seen normally over the buttocks and lower spine in children with darker complexions (Fig. 63-7). Mongolian spots can appear on other parts of the body, such as the face and upper arm. They are usually present from birth but may not appear until the infant is several weeks old. When seen in a typical location, they are readily recognized as Mongolian spots. When located elsewhere on the body, they may be mistaken for bruises. Bruises resolve over time, Mongolian spots remain unchanged (do not go through purple-green-yellow-brown transformation) because they are undistributed melanocytes. Phytophotodermatitis also may be mistaken for bruises. This is a condition that develops on sun-exposed areas of the body that have been in contact with certain fruits or juices, such as lime or lemon juice. The lesion appears as a brown discoloration, which may take the shape of the dripped juice or the object with which the juice came in contact. For instance, if a mother is making lemonade, has the lemon juice on her hand, and holds her child, a brown discoloration in the form of a handprint may appear if the child is in the sun. These lesions fade over time, and with a careful history and physician familiarity with the condition, the correct diagnosis can be made. Burns also may occur unintentionally. Unintentional burns are usually secondary to spills and may take the form of drip marks down a child’s chest. Bullous impetigo can be mistaken for second-degree burns because of its blister-like appearance. Culturing the lesion reveals the presence of S. aureus in the case of bullous impetigo. Certain dermatologic conditions, such as epidermolysis bullosa, also may cause bullous lesions that may resemble second-degree burns. The history and generalized appearance of these lesions help establish the correct diagnosis. Fractures may occur unintentionally. In young infants, fractures may be a result of birth-related injuries. The most common fractures sustained during birth are clavicular and humeral fractures. These fractures may not be appreciated immediately after birth but become apparent when callus formation is noted. Ambulatory children may sustain fractures related to falls. A toddler’s fracture, also referred to as a CAST fracture (childhood accidental spiral tibial fracture), occurs when there is a twisting injury to the tibia as the child falls on it.30 In general, the fracture is a nondisplaced distal fracture of the tibia that is detected when a child presents with a limp. Sometimes the fracture may not be apparent on the initial radiograph, but delayed radiographs (1–2 weeks after injury) may show callous formation or a bone scan may show the presence of increased bony uptake.
Fractures occur with increased frequency and with lower amounts of force in certain conditions. Premature infants may experience osteopenia of prematurity (sometimes referred to as rickets of prematurity), the radiographic appearance of which can be mistaken for a metaphyseal fracture.31 In addition, osteopenic bones may fracture more easily. More advanced cases of osteopenia can be noted on a plain film of the bones. Osteogenesis imperfecta is a condition in which the bone is more brittle and easily disrupted.32 There are four types of osteogenesis imperfecta, each with a different gene frequency. The overall incidence of osteogenesis imperfecta is 1 in 20,000. Generally, osteogenesis imperfecta is associated with other clinical findings, such as blue sclerae and brown discoloration of the teeth (dentogenesis imperfecta). Rarely, bone fragility may be present in isolation. Scurvy, congenital syphilis, and congenital rubella are associated with bony changes that may be misinterpreted as evidence of prior bony injury. Cerebral edema may occur with infection, such as encephalitis and meningitis, or after a hypoxic event. The history and the presence of associated medical findings help with the differentiation of these conditions.
Child Sexual Abuse Numerous medical conditions may be misdiagnosed as child sexual abuse. Accidental trauma, most commonly straddle injuries, may occur after a fall onto the perineum. Such falls occur with climbing on monkey bars, riding on boys’ bicycles, or exiting from a swimming pool. Straddle injuries usually involve the labia minora, labia majora, or periurethral area. The hymen remains uninjured. Lichen sclerosis et atrophicus is a dermatologic condition of unclear etiology affecting prepubertal girls and boys and postmenopausal women. The hymen is unaffected, but the adjacent skin becomes atrophic and may sustain blood blisters or petechiae. Characteristically the skin in the perianal and perihymenal areas becomes hypopigmented and surrounds these orifices with a pale figureof-eight configuration. Urethral prolapse characteristically affects African American girls between the ages of 5 and 8 years. The mucosal lining of the urethra slides forward and protrudes from the urethral orifice, appearing as an erythematous, edematous mass. Symptoms include pain and bleeding. Management may involve sitz baths, antibiotic ointment, or referral to a urologist for ligation. Vaginal discharge may occur secondary to conditions other than sexually transmitted infections. Shigella, group A betahemolytic streptococcis, Candida, pinworm infestation, and vaginal foreign bodies can cause a vaginal discharge. Penile swelling may occur with priapism (often secondary to sickle cell disease), paraphimosis, or an infestation with chiggers. Fissures and tags in the perianal area may result from trauma, but also may be associated with constipation and inflammatory bowel disease. Group A beta-hemolytic streptococcis can cause painful inflammation with erythema in the perianal area. Affected children may be febrile and experience pain with defecation. Hemorrhoids are rare in children and are associated with conditions that lead to an elevation of intraabdominal venous pressure as occurs with cirrhosis of the liver.
■ MANAGEMENT The focus of management is to attend to serious or lifethreatening injuries, such as significant head or abdominal trauma, and stabilize the patient. Physical problems requiring medical intervention, such as fractures, lacerations, burns, or sexually transmitted infections, should be managed
799
chairs or be blind or otherwise disabled. For other children, intervention and therapy for themselves and their offending parents may help reverse the adverse psychological effects of the abuse. The emergency physician has a key role in the early detection of the problem. The entire medical team, along with social services, law enforcement, and the judicial system, is responsible for implementing a treatment plan that prevents recidivism.
Disposition Admission to the hospital may be warranted because of the patient’s injuries, to complete the medical evaluation, and to protect the child while the evaluation is occurring. Many hospitals have suspected child abuse and neglect (SCAN) teams. These teams can offer expert consultation either in the emergency department or after the child has been admitted. SCAN teams usually have unique expertise in assessing the genital findings in prepubertal girls. Issues may arise with the family regarding “allowing” the child to be admitted to the hospital once the concerns of the emergency physician are related to the family. A strong security presence is advised during both disclosure of the intent to file a report of possible child abuse and when the decision is made to admit the child to the hospital. If security is not available to provide support to the emergency physician, and there is concern for personal safety, the addition of other hospital personnel should be solicited. The outcome for an abused child varies with the nature, extent, and duration of the abuse. Some children die as a result of their inflicted injuries. Others have irreversible brain damage and may spend the remainder of their lives confined to wheel-
KEY CONCEPTS ■
The emergency physician is mandated to report suspicion of child physical or sexual abuse. ■ In infants and young children presenting with any injury, the emergency physician should consider child physical abuse as a diagnosis and especially with certain injuries, such as unusual facial bruising; skull, rib, or metaphyseal fractures; patterned bruises or burns; and injuries in preambulatory infants. ■ Children who have been sexually abused may present with vague complaints, such as poor sleep pattern or abdominal pain, or may present with vaginal bleeding or discharge. ■ Numerous conditions mimic child abuse and should be considered in the differential diagnosis, such as Mongolian spots, lichen sclerosis, impetigo, or urethral prolapse. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 63 / Child Maltreatment
appropriately. Key to the ultimate management of the abused child is the precise recording of the pertinent history, particularly any disclosure made by the child, and the physical findings. Most states require the completion of a specific child abuse reporting form as a means of notifying the authorities about the suspected case of child abuse. In addition, many jurisdictions require immediate telephonic notification to initiate an investigation of the circumstances surrounding the abuse.
Chapter 64
Sexual Assault
Laura Slaughter
■ PERSPECTIVE Background Sexual violence is a significant problem in the United States. The Centers for Disease Control and Prevention have defined it as sexual activity where consent is not obtained or freely given. It pertains to a wide variety of sexual conduct and may entail, but does not require, penetration, completion, or, in certain cases, physical contact (e.g., voyeurism). Referred to in statute as sexual assault, it is explicated more precisely by state and local governments. Emergency physicians must be aware of their local laws because they are mandated reporters and may be involved in the evaluation, treatment, evidence collection, and documentation of sexual assault.1 Major advances in the evaluation and management of sexual assault victims (SAVs) have occurred over the past 30 years. The formation of community-based multidisciplinary teams that first took hold in California is probably the most important. The development of a sexual assault response team (SART) with representatives from the District Attorney’s office, law enforcement, crime laboratory, medical personnel, including both physicians and nurses, social service agencies, and victim advocates came together as a group. These major players have sought to solve the logistical, medical, psychological, legal, and social problems incurred by SAVs. The commitment and mutual cooperation of the SART led to the development of standardized protocols for the care and treatment of SAVs. These protocols specify the procedures for interviewing and examining the SAV, collection, preservation, and storage of evidentiary materials, and include the evidence kit and forms for documentation. Many jurisdictions have adopted this standardized system, and in 2004 the first national protocol was released by the U.S. Department of Justice.2 With a protocol in place, the necessity of having trained forensic examiners (FEs), irrespective of academic credentials, is evident.3 Currently, this role subsumes the sexual assault nurse examiner (SANE); these ubiquitous, mostly hospitalbased SANE programs have contributed substantially to our knowledge and understanding of SA and improved the care of these victims.4,5 Additionally, because of the use of special technologies, including colposcopy, digital photography/ videography, and the alternative light source, the establishment of designated examination centers for sexual violence is increasingly popular.6 Much information about the characteristics, physical examination findings, and correlates of injury in SAVs is now available. This information facilitates the 800
management of the SAV, improves the experience of the victim, and, ultimately, will assist in the identification of the perpetrator.
Epidemiology Sexual assault does not seem to be declining at the rate of other violent crimes; nearly every category of crime was significantly lower in 2001–2002 than in the preceding 2 years except for SA.7,8 Importantly, the cost of sexual violence to the health care system is high because it may include not only the initial visit for the crime but also subsequent visits for other health issues.6 Women remain the predominant victims (94%). Historically, SA has been a largely unreported crime, with only about a third of SAVs coming forward. The major reasons given for not reporting include that the matter was personal, fear of reprisal, or fear of police bias. The closer the relationship between the victim and offender, the less likely the SAV has been to report the crime; the relationship between SAV and perpetrator is often absent in the medical record, yet it is critical for safety planning.9 A substantial proportion of SAVs report at or after 72 hours, and they are typically adolescents. There is a high positive correlation between reporting to the police and receiving medical treatment.10 SA is an extremely common crime, with estimates of one in three females and one in seven males being assaulted during a lifetime.11 The mean age of the victim is approximately 20 years. She is most often single. Adolescents account for less than half of all victims seen, yet the incidence of SA peaks in the age group of 16 to 19 years.12 For nearly 40% of victims, SA is the first sexual experience.13 A person known to the victim commonly perpetrates the assault. Former and current boyfriends are equally common as perpetrators, lending support to the position that leaving a violent partner does not always end the terror.9 The younger the victim, the more likely the perpetrator is to be a relative. The location of the SA varies with the victim and the type of perpetrator. In general, adults are usually assaulted in their own home, whereas adolescents are more likely to be assaulted in the assailant’s residence.14 Stranger assaults are less common; they are more likely to involve adults, occur outdoors, and include the use of a weapon and a greater likelihood of injury.15–18 Alcohol and drug use in both the assailant and the victim are common accompaniments to SA.19,20
■ DISTINGUISHING PRINCIPLES OF DISEASE Emergency Department Preparation A standardized approach to the management of the SAV is important. This should include, if at all possible, the development of a multidisciplinary team that works together under a protocol. That protocol should address every detail, from the handling of the SAV’s first call to dispatch to the referral for psychological support. The SAV should be taken out of the medical triage system that typifies the emergency department not only to provide privacy and security for the SAV but also to prevent the deterioration of evidence. This approach assures a consistent process of evaluation, treatment, and collection of evidence. The medical team should receive forensic training on interviewing and examination techniques; collecting, preserving, and storing evidence; and chain of custody issues. The team should be trained to use the examination form and be thoroughly familiar with the SA kit (rape kit) provided by the state or local crime laboratory. Advocates, whether provided through law enforcement or by a separate entity, need to receive training from the SART about examination procedures and staff roles so they can best advise and counsel the SAV. A SART program cannot be successful unless advocates feel that the SAV will be treated with respect and receive the best treatment available. Because the SART examiner is the first person on the medical team to greet the victim, a kind, calm, knowledgeable, and professional demeanor is of the utmost importance. The emergency department should also be prepared for the unusual victim with significant or life-threatening injuries. In this instance, the emergency physician needs to delegate the forensic responsibility to another forensically trained staff member whose sole purpose is to collect the evidence and, if required, follow the SAV to surgery. In cases in which there is substantial injury, following the patient from intake allows the examiner to understand better and document the nature and extent of the trauma and continue the forensic process with little further distress to the patient.35
Obtaining the History and Consent Numerous consents should be obtained from the SAV, depending upon state and local laws (Box 64-1). The FE should be knowledgeable about all statutes governing consent, including
801
BOX 64-1 Consent Issues in Sexual Assault Cases Consents Should Specify that the SAV Signature Acknowledges: That hospitals and health care professionals are mandated reporters Receipt of information about victim compensation funds Specific understanding of the examination and evidence collection procedures Specific understanding of the use of photography in documenting physical and genital injuries That information collected will be sent to law enforcement and is obtainable by defense counsel That data without patient identity can be collected for valid educational and scientific interest That consent may be withdrawn at any time
BOX 64-2
Preliminary Issues and Strategies in Preparing for Taking the History from an SAV
Provide quiet, confidential, safe environment. Briefly review the interview and examination process in private with the SAV. Explain sensitive/personal/embarrassing nature of questions and right to be interviewed without family or friends. Show concern for immediate comfort (e.g., if thirsty, take oral swabs first so SAV may drink). Provide advocacy. Always conduct interview the same way. Leave difficult questions until the end. Explain why you are asking the question and the possible responses. Explain that all questions must be asked.
whether minors need parental consent. In some states, even if parental consent is not required, the FE may still be required to contact the parents and document the success or failure of this attempt. Although not part of the official consent process, most SAVs are concerned about access to the SART record and photographs, particularly when the SART examination facility is within the hospital setting. Keeping these records separate from the primary hospital chart system helps protect the privacy of the victim and has a precedent in the similar handling of psychiatric records. Surveys of SAVs have identified that they desire information about sexually transmitted diseases (STDs), pregnancy, emergency contraception, followup care, and physical and psychological health effects of SA.36 Providing written information on these topics to the patient with his/her signature confirmation documents the patient’s receipt of follow-up recommendations. Before getting started with the history, the SAV’s immediate privacy and personal needs should be addressed. Box 64-2 outlines some of the issues that may make the interview more comfortable for the patient and ensure reliable historical information from the patient. The group taking the history may include a law enforcement officer, patient advocate, medical assistant, and FE. A detailed history is important. California was the first state to mandate a uniform examination protocol and specific training for examiners (Fig. 64-1).37 text continued on p. 806
Chapter 64 / Sexual Assault
Most assaults involve penile-vaginal penetration,11,21 and penile penetration is significantly associated with genital injury in females.22,23 Typically, digital-vaginal penetration is the second most common sexual act reported. Oral-genital contact occurs in less than 30%, with anal assault slightly less common11,13,24,25 The use of a foreign object is unusual (10%).11,21 Anal assault is associated with increased violence,13,15,22 offender preference for anal sex,26 and offender problems with sexual dysfunction.27,28 Injury is not an inevitable consequence of SA.18,29–31 Adolescent SAVs have been shown to sustain more anogenital injuries than their adult counterparts.17,21,24,32 Nongenital trauma occurs in 40 to 81% of SAVs, and its presence is associated with anogenital injury.13,22,31 The extremities are most commonly injured, followed by the head and neck.33 Serious injury involving hospitalization occurs in about 5%,6,10 and death associated with SA is estimated at 1% or less, although this latter figure is probably a gross underestimate. Psychological distress and interpersonal difficulties are the major sequelae after an SA. These problems are exacerbated in SAVs with known attackers who delay reporting.34
ID Number
2. Responding Officer
Time
ID number
Agency
Telephone
Date
Time
I have been informed that victims of crime are eligible to submit crime victim compensation claims to the State Victims of Crime (VOC) Restitution Fund for out-of-pocket medical expenses, psychological counseling, loss of wages, and job retraining and rehabilitation.
CalEMA 2-923 (Rev 7/02)
Original - Law Enforcement
Copy within evidence kit - Crime Lab
Copy - Child Protective Services (if patient is a minor) 1
DISTRIBUTION OF CalEMA 2-923
Guardian
Copy - Medical Facility Records
Parent
________ (Initial)
I understand that data without patient identity may be collected from this report for health and forensic purposes and provided to health authorities and other qualified persons with a valid educational or scientific interest for demographic and/or epidemiological studies.
Patient
________ (Initial)
________ (Initial)
________ (Initial)
I hereby consent to a forensic medical examination for evidence of sexual assault.
I understand that collection of evidence may include photographing injuries and that these photographs may include the genital area.
I understand that a forensic medical examination for evidence of sexual assault at public expense can, with my consent, be conducted by a health care professional to discover and preserve evidence of the assault. If conducted, the report of the examination and any evidence obtained will be released to law enforcement authorities. I understand that the examination may include the collection of reference specimens at the time of the examination or at a later date. I understand that I may withdraw consent at any time for any portion of the examination.
Minors: Family Code Section 6927 permits minors (12 to 17 years of age) to consent to medical examination, treatment, and evidence collection for sexual assault without parental consent. See instructions for parental notification requirements for minors.
________ (Initial)
_________ (Initial)
Case Number
___________________________________________________________________________________
Law enforcement officer
Signature _____________________________________________________
• •
•
•
Date
Discharge Time
___________________________________________________________________________________
Telephone
Reported by: Name
other):
Telephone (W) (H) Discharge Date
Patient Identification
I understand that hospitals and health care professionals are required by Penal Code Sections 11160-11161 to report to law enforcement authorities cases in which medical care is sought when injuries have been inflicted upon any person in violation of any state penal law. The report must state the name of the injured person, current whereabouts, and the type and extent of injuries.
D. PATIENT CONSENT
•
•
C. PATIENT INFORMATION
Telephone Authorization Agency: Authorizing party: ID number: Date/time:
3. I request a forensic medical examination for suspected sexual assault at public expense.
Agency
ID Number
county
Arrival Time
State
Telephone
Jurisdiction ( city
Arrival Date
1. Telephone report made to law enforcement agency Name of Officer Agency
Ethnicity
County
Patient ID number
Name of Medical Facility:
B. REPORTING AND AUTHORIZATION
F
Gender M
3. Age
DOB
City
2. Address
1. Name of patient
A. GENERAL INFORMATION (print or type)
Confidential Document
CalEMA 2-923
STATE OF CALIFORNIA California Emergency Management Agency
FORENSIC MEDICAL REPORT: ACUTE (2 MDI short-acting beta2-agonist canisters per month Current use of or recent withdrawal from systemic corticosteroids Difficulty perceiving asthma symptoms or severity of exacerbations Social History Low socioeconomic status or inner-city residence Serious psychosocial problems Illicit drug use, especially inhaled cocaine and heroin Comorbidities Cardiovascular disease Other chronic lung disease Chronic psychiatric disease ED, emergency department; ICU, intensive care unit; MDI, metered-dose inhaler.
Chapter 71 / Asthma
Necropsies of patients with status asthmaticus reveal grossly inflated lungs that may fail to collapse on opening of the pleural cavities. Histologic examination reveals luminal plugs consisting of inflammatory cells, desquamated epithelial cells, and mucus. Marked thickening of the airway basement membrane, submucosal inflammatory cells, increased deposition of connective tissue, mucous gland hyperplasia, and hypertrophy of airway smooth muscle are also observed. Bronchiectasis is described in 15 to 20% of asthmatics.44 Reports of patients experiencing sudden-onset fatal asthma demonstrate less mucus in the airway lumens, suggesting that terminal events in this group may be dominated by bronchoconstriction without excessive luminal plugging; however, the putative roles of bronchoconstriction versus mucous plugging in sudden-onset fatal asthma are unclear.49,50
PART III ■ Medicine and Surgery / Section Two • Pulmonary System
894
The brief history pertinent to the current exacerbation should include onset and possible triggers, severity of symptoms especially as compared with previous exacerbations, and other comorbidities (especially those that may be worsened by systemic corticosteroids such as diabetes, peptic ulcer, hypertension, and psychosis). In addition, all current asthma medications should be noted, including times and amounts recently used, and any potential asthma exacerbators such as aspirin or NSAIDs, beta-blockers (including topical agents used for glaucoma), and angiotensin-converting enzyme inhibitors.
Physical Assessment Those who suffer from mild acute asthma usually speak in sentences, moderate in phrases, and severe in words. Although alterations in mentation or consciousness indicate severe asthma, restlessness and agitation do not reliably indicate hypoxia or hypercapnia. Patients who sit upright have severe airway obstruction; cyanosis is uncommon because of the left shift of the oxyhemoglobin dissociation curve produced by respiratory alkalosis. Diaphoresis can be seen secondary to the work of breathing, but if profound it is usually accompanied by a decreasing level of agitation and interaction with caregivers and may be preterminal. Tachypnea and tachycardia greater than 120 beats/min are associated with severe obstruction, but a lower rate does not rule out severe asthma. The respiratory rate correlates poorly with PFT and indicates severe obstruction only if it is more than 40 breaths/min.64 A pulsus paradoxus or inspiratory fall in systolic blood pressure greater than 10 mm Hg usually signifies severe disease, but its absence does not exclude it, and it rarely is contributory when taken in the context of the patient’s overall evaluation. Its use predated common availability of bedside spirometry, which is a more accurate and reliable test. When pulsus paradoxus is present, it may disappear with minimal improvement in airflow through larger airways.64 Similarly, use of accessory muscles of respiration (sternocleidomastoid and scalenus muscles) is not prognostic. Wheezing does not designate the presence, severity, or duration of asthma. It correlates poorly with the degree of functional derangement and may be absent when maximal effort produces minimal airflow. Physical examination may help to identify such complications of asthma as pneumonia, pneumothorax, or pneumomediastinum that may arise atypically as subcutaneus emphysema or simulate upper airway obstruction.
■ DIAGNOSTIC STRATEGIES Pulmonary Function Studies Because the severity of airflow obstruction cannot be accurately assessed from symptoms and physical exam alone,65 physicians tend to underestimate the degree of airway obstruction in acute asthma, particularly on initial assessment. Therefore, routine PFTs should be part of ED assessment and monitoring. The forced expiratory volume in 1 second from maximal inspiration (FEV1) or the peak expiratory flow rate (PEFR) in liters per second starting with fully inflated lungs and sustained for at least 10 msec may be used. Both measurements require the patient’s cooperation for maximal effort and are effort-dependent. Whenever possible, the best of three consecutive values should be recorded. Any patient not able to perform a pulmonary function study should be considered to have severe airway obstruction.
Most asthmatic assessments in the ED use single-patientuse portable peak flow meters because PEFR is easier to measure. There is wide limit of agreement between different devices, so a single device should be used to assess then reassess an individual patient, and different portable meters should not be used interchangeably.66 Lastly, although generally similar, the FEV1 and PEFR measurements do not appear to be interchangeable in assessing acute airway obstruction, which is not addressed in all management guidelines.67 Although absolute PFT measurements can be used,68 percentage of predicted performance (% predicted) values are preferable because they account for the individual’s age (now to age 85), sex, and height. Ideally, the percentage of the patient’s personal best effort individualizes the assessment and treatment.
Arterial Blood Gas Analysis Using pulse oximetry, changes in equilibration of oxygen saturation with supplemental oxygen occur in 3 to 4 minutes during an acute asthma attack. With initial onset of an asthma attack, stimulated hyperventilation leads to a modest fall in the partial pressure of carbon dioxide in arterial blood (Paco2). As airway obstruction increases, the Paco2 normalizes (PFT 15–25% predicted) and then increases (PFT < 15% predicted) with worsening hypoxemia. Because neither pretreatment nor post-treatment arterial blood gases (ABGs) correlate with PFTs or predict clinical outcome, ABG determination is rarely clinically useful in acute asthma exacerbations unless oxygen saturation cannot be obtained reliably using pulse oximetry. ABG determination is of no value in determining the need for tracheal intubation. ABG sampling, if used, should be limited to a subset of patients with predicted PFTs of less than 30%, whose clinical course is perplexing and for whom capnography is not available. Occasionally, despite PFTs improving with bronchodilator therapy, some patients have a transient fall in the Pao2 secondary to pulmonary vasodilatation and worsening ventilation-perfusion mismatch.69 The assessment of ventilation may be simplified since there is a high concordance between endtidal partial pressure of CO2 (Pco2) measured by capnography and the Paco2 obtained with ABG measurements.70
Other Blood Testing Leukocytosis is common in patients with acute asthma exacerbation, but is not of discriminatory value in determining whether patients with fever or purulent sputum have acute superimposed pulmonary infection. Of note, corticosteroid and epinephrine therapy demarginate polymorphonuclear leukocytes after 1 to 2 hours, and patients on chronic steroid therapy may have normal or significantly elevated WBC counts. Serum electrolytes are not primarily altered unless the patient is on corticosteroids or diuretics or has cardiovascular disease and is receiving aggressive beta2-agonist therapy. Frequent albuterol treatments can cause transient hypokalemia, hypomagnesemia, and hypophosphatemia, but this is rarely of clinical significance. The few remaining patients who are receiving chronic theophylline therapy should always have serum levels measured to assess for possible toxicity and for appropriate further dosing if deemed necessary. In the older asthmatic with cardiovascular comorbidities who presents with wheezing, measure the B-type natriuretic peptide (BNP) level to determine the contribution of unrecognized congestive heart failure to the clinical picture.
895
Radiology Studies
Electrocardiogram and Cardiac Monitoring The electrocardiogram (ECG) need not be routinely obtained, except in patients older than 40, those with a separate complaint (e.g., chest pain) that would prompt an ECG, or those with a history of significant cardiovascular disease, in whom the asthma attack may be a form of physiologic stress test. In patients with severe asthma, the ECG may show a right ventricular strain pattern that reverses with improvement in airflow. Older patients, especially those with coexistent heart disease or with severe exacerbation, may require continuous cardiac monitoring to detect dysrhythmias. All patients with severe hypoxemia, and those for whom intubation is contemplated, should have continuous cardiac monitoring.
Future Monitoring Strategies Noninvasive monitoring of bronchial inflammation may customize the ED assessment of acute asthma. This may include measurement of biologic biomarkers such as cytokine profiles in the blood, evaluation of LTE4 in the urine, and the monitoring of exhaled pentane, hydrogen peroxide, nitric oxide, or carbon monoxide levels. Of these measurements, exhaled nitric oxide shows the most promise in chronic asthma management72 but provides little aid in assessing the severity of acute exacerbations.73
Assessment Summary The severity of airflow obstruction cannot be accurately judged by patients’ symptoms, physical examination, and laboratory tests. Serial measurements of airflow obstruction (FEV1 or PEFR) are key components of disease assessment and response to therapy. A more detailed analysis between commonly measured variables and patients with severe asthma (FEV1 < 1 L) is shown in Table 71-1.
■ DIFFERENTIAL CONSIDERATIONS See Box 71-2.
■ MANAGEMENT OF ACUTE EXACERBATIONS
FACTOR
SEVERE ASTHMA (FEV1 < 1.0 L)
Pulse rate (beats/min)
≥120, but may be less with equally severe asthma ≥40, but most are >20, therefore nondiscriminating ≥10, but may be absent with equally severe asthma in 50% of cases If all three abnormal, 90% with severe asthma, but only 40% with FEV1 < 1.0 L have all three abnormal If present, may indicate severe asthma; if absent, may have equally severe asthma in 50% of cases Pao2 ≤ 60 or Paco2 ≥ 42 indicates severe asthma; all other values difficult to interpret unless PEFR or FEV1 known PEFR and FEV1 measure directly the degree of airflow obstruction; most useful in assessing severity and guiding treatment decisions
Respiratory rate (breaths/min) Pulsus paradoxus (mm Hg) Pulse rate ≥120, respiratory rate ≥20, pulsus paradoxus ≥10 Use of accessory muscles of respiration ABG analysis (mm Hg)
Pulmonary function studies
ABG, arterial blood gas; FEV1, forced expiratory volume in 1 second; Paco2, partial pressure of CO2 in arterial blood; PEFR, peak expiratory flow rate.
BOX 71-2 The Differential Diagnosis of Asthma Cardiac Conditions Valvular heart disease Congestive heart failure COPD Exacerbation Pulmonary Infection Pneumonia Allergic bronchopulmonary aspergillosis Löffler’s syndrome Chronic eosinophilic pneumonia Upper Airway Obstruction Laryngeal edema Laryngeal neoplasm Foreign body Vocal cord dysfunction Endobronchial Disease Neoplasm Foreign body Bronchial stenosis Pulmonary Embolus
Home and First-Responder Strategies
Carcinoid Tumor
Patients should be educated to monitor their symptoms, signs, and PEFR to recognize early deterioration and should be provided with a written action plan in the event of an exacerbation. Early therapy can prevent progression to severe attacks. Home management includes increased use of inhaled beta2agonists, early administration of systemic corticosteroids (not simply doubling the dose of current inhaled corticosteroids), and specific instructions on when and how to seek emergency care.1 Ideally, emergency medical service providers should
Allergic/Anaphylactic Reaction Miscellaneous Conditions GERD Noncardiogenic pulmonary edema Addison’s disease Invasive worm infection COPD, chronic obstructive pulmonary disease; GERD, gastroesophageal reflux disease.
Chapter 71 / Asthma
A chest radiograph is of little value in most acute asthma exacerbations, and its use should be restricted to patients thought to have a complicating cardiopulmonary process such as pneumonia, pneumothorax, pneumomediastinum, or congestive heart failure. Also, patients who do not respond to optimal therapy and require hospital admission have a higher likelihood of radiographically identifiable, unsuspected, clinically significant pulmonary complications of asthma (15% of cases).71
Table 71-1 Objective Findings in Asthma Assessment
896
PART III ■ Medicine and Surgery / Section Two • Pulmonary System
Table 71-2 Initial Severity Assessments and Therapies in the Emergency Department MILD TO MODERATE
SEVERE
FEV1 or PEFR (% predicted/personal best) Oxygen therapy Nebulized albuterol solution Levalbuterol (optimal)
≥40%
Unable or 95% in pregnant women and those with coexistent heart disease) rather than using predetermined concentrations or flow rates. Serial oxygen saturation monitoring is essential during the acute phase. Humidification of the inspired air-oxygen mixture is not essential, although studies suggest that active airway rehydration should be revisited.78
Adrenergic Medication Controversies in Use. Epidemiologic studies report an association between death and near death from asthma and the use of inhaled beta2-agonists, with use of more than one canister per month increasing this risk that doubles for each additional monthly canister used.79 This relationship does not imply causality but may be a marker for more severe disease, particularly if anti-inflammatory treatment is underused. Guidelines for chronic use of inhaled beta2-agonists, however, recommend limited daily use in a rescue-only mode.1
One form of albuterol is a racemic mixture of equal amounts of R and S isomers. Data from animal and human studies suggest that the S isomer, which contributes no bronchodilator activity, is proinflammatory, spasmogenic, and induces bronchial hyper-reactivity. This possibly explains the adverse effects of increased morbidity and mortality rates associated with regular or excessive use of this drug.80 Some investigations of the beta-adrenergic receptor polymorphisms show differential responsiveness to inhaled albuterol, a possible explanation for the widely varying responses seen clinically when treating patients with acute disease.81 Short-Acting Inhaled Beta2-Agonist Choice and Dosing Schedule. Racemic albuterol has been the main beta2-agonist used in the ED for over 30 years. It is more beta2-selective, longer acting, and has fewer side effects than other previously available drugs such as metaproterenol or isoetharine. Levalbuterol, the R isomer of racemic albuterol, is commercially available as a preservative-free nebulizer solution (unit doses of 0.31, 0.63, or 1.25 mg) for prevention and treatment of bronchospasm. In chronic asthma, levalbuterol provides a better therapeutic index than the standard dose of racemic albuterol, further fueling the debate on the potential adverse effects of the S isomer of beta-agonists.82 Clinical studies in acute disease report that levalbuterol on a milligram for milligram basis is a better bronchodilator than similar amounts of R-albuterol delivered with the S isomer in the racemic mixture.83-85 This reinforces the notion that the S isomer has a negative rather than neutral effect. The amount and frequency of delivery of levalbuterol and racemic albuterol depend on the initial severity and response to therapy, as shown in Table 71-2 and Table 71-3. Patients with more severe obstruction with a poor response to initial therapy should receive higher dosing schedules and possible continuous administration.86 When patients are stable but require nonintensive care unit (ICU) admission, it may be possible to dose nebulized levalbuterol at 1.25 mg every 8
897
Table 71-3 Response after 1 Hour of Initial Treatment
Albuterol therapy Levalbuterol (optimal) Racemic albuterol Ipratropium therapy Corticosteroids
SEVERE EXACERBATION
40–69%
2 3 × (upper limit of normal for serum LDH level) From Light RW, et al: Pleural effusions: The diagnostic separation of transudates and exudates. Ann Intern Med 77:507, 1972.
setting. The primary goal of pleural fluid analysis is to distinguish between transudative and exudative effusions. The presence of a transudate suggests an underlying process (e.g., congestive heart failure, nephrotic syndrome), whereas the presence of an exudate mandates a more extensive diagnostic evaluation. Although numerous alternative measurements are proposed, Light’s criteria remain a widely accepted means of differentiating transudates and exudates (Box 75-3).34 In the presence of an exudative effusion, additional pleural fluid analyses further classify the effusion. A pleural fluid pH of less than 7.3 is associated with parapneumonic effusions, malignancies, rheumatoid effusions, tuberculosis, and systemic acidosis. A pH of less than 7.0 strongly suggests empyema (or esophageal rupture). A pleural fluid pH of less than 7.0 and glucose less than 50 mg/dL are reasonable indications for tube thoracostomy.35 Normal pleural fluid contains less than 1000 white blood cells/mm3; exudative pleural fluid may contain over 10,000 white blood cells/mm3. Although the absolute cell count has limited diagnostic value, a predominance of neutrophils suggests an acute process, such as pneumonia, pulmonary embolus, or acute tuberculous pleuritis. A predominance of monocytes or lymphocytes suggests a more chronic process, such as malignancy or established tuberculosis. Pleural fluid from any patient with an undiagnosed exudative pleural effusion should undergo Gram’s staining and culture for bacteria (aerobic and anaerobic), mycobacteria, and fungi.
PART III ■ Medicine and Surgery / Section Two • Pulmonary System
946
In the absence of a traumatic tap, bloody fluid suggests trauma, neoplasm, or pulmonary infarction.35 If the hematocrit of the pleural fluid is more than 50% that of the peripheral blood, the effusion is, by definition, a hemothorax. Atraumatic hemothorax is relatively rare but can occur with spontaneous rupture of a tumor or blood vessel (e.g., ruptured aortic aneurysm). If the diagnosis of a malignant pleural effusion is being considered, pleural fluid should be submitted for cytologic examination. Contrary to popular perception, the sensitivity for diagnosis of pleural malignancy does not depend on the volume of pleural fluid extracted during thoracentesis.36 Cytologic analysis provides the diagnosis of cancer in 40 to 87% of malignant effusions.35
Management In patients with large effusions, urgent therapeutic thoracentesis may stabilize respiratory or circulatory status. The presence of empyema mandates insertion of a chest tube to drain the pleural space adequately and prevent the development of loculations. If an effusion is already loculated, streptokinase or urokinase can be injected by a thoracic surgeon, pulmonologist, or interventional radiologist into the pleural space in an attempt to dissolve adhesions and allow fluid to drain freely. Hemothorax requires tube thoracostomy to evacuate the pleural space, quantify bleeding, and allow apposition of the two pleural surfaces to tamponade hemorrhage. If bleeding exceeds 200 mL/hr, thoracotomy should be considered. In most other cases, the decision to proceed with therapeutic thoracentesis in the ED can be individualized. For example, therapeutic thoracentesis may be considered in patients with known, recurrent malignant effusion, in whom symptomatic relief may permit discharge. Pain relief is an important consideration in the management of patients with pleuritis, which may have a significant inflammatory component. Nonsteroidal anti-inflammatory drugs are relatively successful in treating pleural pain.37 Opioid analgesia is safe and effective, but care should be exercised in debilitated patients or patients with severe lung disease because of potential respiratory depression. Relative contraindications to thoracentesis include coagulopathy and other bleeding disorders. Thoracentesis may be safe with a prolonged prothrombin time in the absence of active bleeding.38 Pleural adhesions, suggested by a prior history of empyema, represent a relative contraindication to thoracentesis because of the high risk of pneumothorax associated with blind needle insertion. After thoracentesis is completed, a chest radiograph should be obtained to rule out iatrogenic pneumothorax. Other potential complications of thoracentesis include hemothorax, lung laceration, shearing of the catheter tip, and infection. Transient hypoxia caused by ventilation-perfusion mismatch often occurs, whereas unilateral, postexpansion pulmonary edema is rare except when large volumes (>1500 mL) are drained in one session. Hypotension also can occur after removal of a large volume of fluid, particularly in patients who are already intravascularly volume depleted.
Outcome Some pleural effusions reflect little clinical significance. For example, small pleural effusions are common after abdominal surgery and in the postpartum state and resolve spontaneously within a few days. Effusions associated with viral pleuritis are generally self-limited and resolve without specific treatment.
For patients with congestive heart failure, pleural effusions generally respond well to diuretic therapy. If an effusion persists despite several days of aggressive diuresis, a diagnostic thoracentesis should be considered. Pleural effusions associated with malignancy are a significant cause of morbidity in patients with advanced cancer. The presence of a malignant effusion indicates disseminated disease, and most of the malignancies that cause pleural effusions—mainly lung or breast carcinoma and lymphoma—are not curable by this stage. Therapeutic thoracentesis can relieve dyspnea in the short term, but malignant effusions tend to be recurrent, often rapidly so. Management strategies include chemical or mechanical pleurodesis to obliterate the pleural space or placement of a pleuroperitoneal shunt to provide continual drainage. Control of pleural effusions can improve quality of life in these patients.39 Parapneumonic effusions contribute significantly to the morbidity and mortality of pneumonia.40 Therefore, the presence of a parapneumonic effusion may influence the decision to hospitalize a patient with community-acquired pneumonia.41 Empyema can develop in 5 to 10% of patients with a parapneumonic effusion, but in most cases it responds well to parenteral antibiotics and pleural drainage. Early surgical drainage results in shorter hospital stays and may be more cost-effective than conservative management.42 In nearly 20% of pleural effusions, no definitive diagnosis can be established even after extensive investigation. A sizable percentage of these effusions may be due to viral infections, and most resolve spontaneously without sequelae.
KEY CONCEPTS ■
For healthy, young patients with a small (90–120 minutes).17
Emergency Department Evaluation The History The character of the chest discomfort as well as the onset, location, radiation, duration, prior presence, and any exacerbating or alleviating factors should be sought. Associated symptoms, especially of a cardiac, pulmonary, gastrointestinal, and neurologic nature, should be elicited. Results from any prior cardiac testing should be obtained. Traditionally, a history of risk factors for CAD is sought; these include male gender, age, tobacco smoking, hypertension, diabetes mellitus, hyperlipidemia, family history, artificial or early menopause, and cocaine abuse. Approximately 80% of a population of more than 122,000 patients with known CAD had at least one of the four conventional risk factors (diabetes mellitus, cigarette smoking, hypertension, or hyperlipidemia).18 Cardiac risk factor burden has little impact on the ED diagnosis of ACS; however, in patients greater than 40 years of age, ACS is 22 times more likely if four of the five major risk factors (diabetes mellitus, smoking, hypertension, hyperlipidemia, and family history) are present (compared to none).19 Nevertheless, Bayesian analysis indicates that risk factors are a populational phenomenon and do not increase or decrease the likelihood of any condition in any one patient. Thus, the presence of an individual risk factor, or a collection of risk factors, is far less important in diagnosing acute cardiac ischemia in the ED than the history of presenting illness, the presence of ST segment or T wave changes, or cardiac marker abnormalities.20 Risk assessment tools, such as the PURSUIT (The Platelet Glycoprotein IIb-IIIa in Unstable Angina: Receptor Suppres sion Using Integrilin Therapy) risk model, the GRACE (Global Registry of Acute Coronary Events) risk model, and the TIMI (Thrombosis in Myocardial Infarction) risk score, can be used to determine risk of death and ischemia in NSTEMI and STEMI. These may aid in decision making and in risk stratifying patients so as to properly disposition them (telemetry bed vs. intensive care unit). The TIMI risk score assigns a point each for seven factors based on history, cardiac markers, and the ECG. It can be accessed at www.timi.org.6 There are several nontraditional risk factors for coronary disease. Antiphospholipid syndrome, rheumatoid arthritis, and particularly systemic lupus erythematosus (SLE) are each associated with a higher risk of cardiovascular disease.21 Women
Chapter 76 / Acute Coronary Syndrome
vidual plaques vary greatly in composition. Fibrous plaques are considered stable but can produce anginal symptoms with exercise and increased myocardial oxygen consumption because of the reduction in coronary artery blood flow through the fixed, stenotic lesions. Vulnerable or unstable fibrolipid plaques consist of a lipid-rich core separated from the arterial lumen by a fibromuscular cap. These lesions are likely to rupture, resulting in a cascade of inflammatory events, thrombus formation, and platelet aggregation that can cause acute obstruction of the arterial lumen and myocardial necrosis.14 Thrombus formation is considered an integral factor in ACS, including UA, NSTEMI, and STEMI. All of these syndromes are initiated by endothelial damage and atherosclerotic plaque disruption, which leads to platelet activation and thrombus formation. Platelets play a major role in the thrombotic response to rupture of coronary artery plaque and subsequent ACS. Platelet-rich thrombi are also more resistant to fibrinolysis than fibrin- and erythrocyte-rich thrombi. The resulting thrombus can occlude the vessel lumen, leading to myocardial ischemia, hypoxia, acidosis, and eventually infarction. The consequences of the occlusion depend on the extent of the thrombotic process, the characteristics of the preexisting plaque, the extent of the vessel obstruction, and the availability of collateral circulation. In the setting of UA, acute stenosis of the vessel is noted; complete obstruction, however, is encountered in only 20% of cases. In these cases, it is likely that extensive collateral vessel circulation prevents total cessation of blood flow, averting frank infarction.13,15 With AMI, the occlusive fibrin-rich thrombus is fixed and persistent, resulting in myonecrosis of the cardiac tissue supplied by the affected artery. Angiographic studies demonstrate that the preceding coronary plaque lesion is often less than 50% stenotic, indicating that the most important factors in the infarction are the acute events of plaque rupture, platelet activation, and thrombus formation rather than the severity of the underlying coronary artery stenosis. Another important aspect of ACS is vasospasm. After significant coronary vessel occlusion, local mediators and vasoactive substances are released, inducing vasospasm, which further compromises blood flow. Central and sympathetic nervous system input increases within minutes of the occlusion, resulting in vasomotor hyperreactivity and coronary vasospasm. Sympathetic stimulation by endogenous hormones such as epinephrine and serotonin may also result in increased platelet aggregation and neutrophil-mediated vasoconstriction. Approximately 10% of MIs occur as a result of coronary artery spasm and subsequent thrombus formation without significant underlying CAD. This mechanism may be more prevalent during UA and other coronary syndromes that do not result in infarction. Further myocardial injury occurs at the cellular level as inflammatory, thrombotic, and other debris from the occlusive plaque lesion are released and embolize into the distal vessel. Such embolization can result in obstruction at the microvasculature, leading to hypoperfusion and ischemia of the distal myocardial tissue, even after reopening of the more proximal, initial, obstructing lesion. In particular, the introduction of calcium, oxygen, and cellular elements into ischemic myocardium can lead to irreversible myocardial damage that causes reperfusion injury, prolonged ventricular dysfunction (known as myocardial stunning), or reperfusion dysrhythmias. Neutrophils probably play an important role in reperfusion injury, occluding capillary lumens, decreasing blood flow, accelerating the inflammatory response, and resulting in the production of chemoattractants, proteolytic enzymes, and reactive oxygen species.
PART III ■ Medicine and Surgery / Section Three • Cardiac System
950
with SLE 35 to 44 years of age are over 50 times more likely to have an MI than a similar age- and gender-matched Framingham population.22
The Classic History The term angina refers to “tightening,” not pain. Classic angina pectoris may not be pain at all but rather a “discomfort,” with a “squeezing,” “pressure,” “tightness,” “fullness,” “heaviness,” or “burning” sensation. Classically, it is substernal or precordial in location and may radiate to the neck, jaw, shoulders, or arms. If the discomfort does extend down the arm, it classically involves the ulnar aspect. Discomfort in the left chest and radiation to left-sided structures is typical, but location and radiation to both sides or to only the right side may be consistent with angina. Radiation of the discomfort to the right arm or shoulder, or to both arms or shoulders, exceeds radiation to the left arm or shoulder in terms of likelihood of the chest pain being due to ACS, although all exceeded a positive likelihood ratio of 2.23 Further, classic features of angina pectoris include exacerbation with exertion, a heavy meal, stress, or cold and alleviation by rest. The onset of pain at rest in no way excludes the diagnosis of angina. Anginal discomfort characteristically lasts from 2 to 5 minutes up to 20 minutes, and it is rare for it to last only a few seconds or to endure for hours or incessantly, “all day” (Table 76-1). Symptoms characteristically associated with angina pectoris, or other entities of ACS, include the following: dyspnea, nausea, vomiting, diaphoresis, weakness, dizziness, excessive fatigue, or anxiety (Table 76-2). If these symptoms arise, either alone or in combination, as a presenting pattern of known ischemic coronary disease, they are termed anginal equivalent symptoms. Recognition that coronary ischemia may arise with an anginal equivalent, rather than a classic symptom is the key to understanding the atypical presentation of ACS. Complaints of “gas,” “indigestion,” or “heartburn” in the absence of a known history of gastroesophageal reflux disease or reproducible pain upon abdominal palpation should raise suspicion of ACS, and similarly if the heartburn is different from the patient’s usual gastroesophageal reflux.
The Atypical History A description of typical symptoms (crushing, retrosternal chest pain or pressure) is often lacking in ACS; this may be due to Characteristics of Classic Anginal Table 76-1 Clinical Chest Discomfort CHARACTERISTIC
Type of pain Duration Onset Location Reproducible Associated symptoms Palpation of chest wall
MORE LIKELY TO BE ANGINA
LESS LIKELY TO BE ANGINA
Dull, pressure 2–5 min, always 0.14 sec
Fully compensatory pause (unless interpolated) No preceding P waves (although retrograde atrial conduction can cause inverted P wave after QRS) Left bundle branch block, right bundle branch block, or hybrid pattern Frequently bizarre QRS axis QRS often >1.14 sec
1220 msec
1240 msec
1360 msec
Figure 77-20. Premature atrial contractions (PACs) with noncompensatory pauses and one aberrantly conducted impulse (upper strip). Note that both conducted and nonconducted PACs reset the sinus node, with the latter creating a pause.
Chapter 77 / Dysrhythmias
pause may result. Fully compensatory pauses are more commonly seen with premature ventricular contractions (PVCs). Table 77-5 lists ECG features to help distinguish PACs from PVCs. If it is conducted to the ventricles, a PAC results in a QRS complex that occurs earlier than the expected sinus QRS complex. The QRS complex from a PAC is narrow and identical to the sinus rhythm complex unless aberrant conduction occurs (Fig. 77-20). Aberrancy is likely to occur if a PAC arrives early within the cardiac cycle, with an RBBB pattern commonly seen on the ECG. Since bundle conductivity depends on the previous cycle length, a PAC that follows a long cardiac cycle (reflected as a preceding long R-R interval) may also be aberrantly conducted because the bundles require more time to repolarize. In the latter setting, aberrant conduction occurs because of the relatively early arrival of the PAC for the given cycle length. This “long/short” aberrant conduction is called the Ashman phenomenon and can occur with any irregular atrial rhythm, including PACs and atrial fibrillation. A PAC is the most common cause of a pause on the ECG. Although the source of this type of pause is obvious when a PAC is conducted, nonconducted PACs are frequently responsible for pauses. In this situation, the sinus node is depolarized by the PAC, causing an interruption and resetting of the regular rate. If the same extrasystolic impulse reaches the AV node or infranodal conducting system during the refractory period, no ventricular depolarization is possible. This combined sinus
PART III ■ Medicine and Surgery / Section Three • Cardiac System
1004
A
B
Figure 77-21. Premature ventricular contractions with compensatory pause. Note that a sinus P wave can be seen in the T wave of the extrasystolic beat. Also note the secondary T wave changes in beats 1 and 4 (T wave is opposite the main deflection of the QRS complex).
BOX 77-6
Causes of Premature Ventricular Contractions and Ventricular Tachycardia
Acute or previous myocardial infarction/ischemia Hypokalemia Hypoxemia Ischemic heart disease Valvular disease Catecholamine excess* Other drug intoxications (especially cyclic antidepressants) Idiopathic causes† Digitalis toxicity Hypomagnesemia Hypercapnia Class I/antidysrhythmic agents Ethanol Myocardial contusion Cardiomyopathy Acidosis Alkalosis Methylxanthine toxicity *Relative increase in sympathetic tone from drugs (direct or indirect) or conditions that augment catecholamine release or decrease parasympathetic tone. † Isolated premature ventricular contractions (PVCs) can occur in up to 50% of young subjects without obvious cardiac or noncardiac disease; however, multiform and repetitive PVCs and ventricular tachycardia are rarely seen in this population.
Fully compensatory pauses and aberrant conduction occur more often with PJCs than with PACs. The causes and treatment of PJCs are the same as those of PACs.
Premature Ventricular Contractions PVCs can occur in a variety of pathologic and nonpathologic states. Their major importance is related to the clinical scenario accompanying their presence and the risk of more serious ventricular dysrhythmias, such as ventricular tachycardia and fibrillation. Extrasystoles that occur during ventricular repolarization (the R-on-T phenomenon) are believed to carry a higher risk of precipitating ventricular tachycardia, although the magnitude of this effect is debated. Other data suggest that PVCs occurring during the next atrial depolarization (the R-on-P phenomenon) carry as high or a higher risk of precipitating serious ventricular dysrhythmias than R-on-T PVCs.33 PVCs can be caused by varying mechanisms (Box 77-6), including reentry, abnormal automaticity, and triggered afterdepolarizations. Classically, PVCs appear on the ECG as wide
II
Figure 77-22. Interpolated premature ventricular contraction. QRS complex extrasystoles (>0.12 second) unassociated with a preceding P wave (Fig. 77-21). In a single lead, a PVC may appear as a narrow QRS complex. This narrow complex occurs if the wave of depolarization is traveling directly perpendicular to the ECG lead and underscores the need to examine multiple leads to identify PVCs accurately. Although P waves from nonconducted sinus impulses may be seen on the ECG, these should have no consistent relationship with the QRS complexes from the PVCs. Rarely, retrograde conduction of PVCs can produce an inverted P′ wave after each QRS complex. PVCs usually cause a fully compensatory pause, with the resulting RR interval encompassing the PVC equal to twice the intrinsic RR interval length (see Fig. 77-21). Rarely, noncompensatory or subcompensatory pauses can be seen with PVCs and are associated with retrograde conduction and sinus node depolarization. Interpolated PVCs refer to another rare instance when the underlying sinus rhythm is unaffected by a PVC (Fig. 77-22). The structure of the QRS complexes depends on the origin of the impulse. PVCs with a left bundle branch appearance result from a wave of depolarization beginning in a right ventricular source and vice versa. Multiform (or “multifocal”) PVCs refer to ventricular extrasystoles from more than one source and appear as varying QRS complex structures. When a PVC depolarizes the ventricles at a similar time as a conducted atrial beat, a fusion QRS complex is seen (Fig. 77-23). Identification of fusion QRS beats indicates the presence of PVCs. PVCs produce abnormal repolarization as a direct result of the abnormal depolarization of the ventricles. Secondary T wave abnormalities refer to the repolarization changes seen as a result of pathologic depolarization and are seen with PVCs along with bundle branch blocks and left ventricular hypertrophy. These secondary T wave changes consist of widening and deflection opposite of the main QRS deflection (see Fig. 7721). Primary T wave abnormalities refer to changes in ventricular repolarization caused by underlying cardiac disease (such as ischemia) and are not solely the result of depolarization abnor-
1005 F
F Chapter 77 / Dysrhythmias
V4
V4
Figure 77-23. Sinus rhythm with run of accelerated idioventricular rhythm. Note fusion beats (F) displaying hybrid appearance of both morphologies.
Classification of Premature Table 77-6 Lown Ventricular Contractions CLASS
DESCRIPTION
0 1 2 3 4A 4B 5
None 200 beats/min in an adult) are a clue to the possibility of an accessory pathway because normal AV nodal tissues rarely allow a ventricular response rate of more than 150 to 165 beats per minute. The use of any predominantly AV nodal blocking agent in the presence of an accessory pathway and atrial flutter or fibrillation may allow unbridled rapid ventricular response rates and precipitate ventricular fibrillation. Type IA agents (especially procainamide) can convert atrial flutter if the previously mentioned drugs fail, or they may be used to prevent recurrence in an outpatient setting. The type III agents amiodarone and ibutilide are alternative primary converting agents. Finally, synchronized electrical cardioversion with sedation, beginning at 25 to 50 J, is effective in terminating atrial flutter in refractory or unstable patients. If electrical therapy is successful but atrial flutter recurs, a type IA or IC agent should be used before repeated electrical cardioversion to help prevent this from occurring again.
1010
PART III ■ Medicine and Surgery / Section Three • Cardiac System
Lead III
Lead II
C Figure 77-28, cont’d C, Atrial flutter with 1 : 1 conduction. This is rare and can be mistaken for ventricular tachycardia (lead II).
Figure 77-29. Atrial fibrillation with rapid ventricular response.
Atrial Fibrillation Atrial fibrillation is the result of chaotic depolarization of atrial tissues. This chaotic activity can lead to reduced cardiac output from a loss of coordinated atrial contractions and a rapid ventricular rate, both of which may limit the diastolic filling and stroke volume of the ventricles. Atrial fibrillation may be paroxysmal or chronic; the paroxysms may last for minutes to days. On the ECG, fibrillatory waves are seen and accompanied by an completely irregular QRS pattern, the hallmark of atrial fibrillation (Fig. 77-29). These fibrillatory waves are best seen in the inferior leads or lead V1 and are described as fine to coarse on the basis of their amplitude. Atrial fibrillation is the result of multiple microreentry circuits, creating 300 to 600 atrial impulses per minute.
The QRS complexes are usually narrow unless an underlying bundle branch block is present. The Ashman phenomenon can cause isolated or repeated aberrant ventricular conduction, usually in an RBBB pattern (Fig. 77-30). These Ashman beats can be mistaken for PVCs if the long-short cycle sequence is not recognized. The ventricular response rate depends on the conduction path and ratio, with the normal AV node maximal response rate being no greater than 150 to 170 beats per minute. As noted with atrial flutter, the presence of a chaotic rhythm (irrespective of the QRS duration) at a ventricular rate of more than 200 beats per minute strongly suggests atrial fibrillation coupled with conduction down an accessory pathway. This rhythm may deteriorate to ventricular fibrillation,34 especially if the rate reaches 250 beats per minute or greater or an AV nodal blocking agent is administered.
1011 MCL1 Chapter 77 / Dysrhythmias
Figure 77-30. Atrial fibrillation with classical Ashman phenomenon series of beats. Note long-short cycle before aberrantly conducted impulses. (From Marriott HJL, Conover MB: Advanced Concepts in Arrhythmias, 2nd ed. St. Louis, Mosby, 1989.)
BOX 77-9 Causes of Atrial Fibrillation Ischemic heart disease* Valvular disease (especially mitral)* Pericarditis Hyperthyroidism Sick sinus syndrome Myocardial contusion Acute ethanol intoxication (holiday heart syndrome) Idiopathic Hypertensive heart disease* Cardiomyopathy* Cardiac surgery Catecholamine excess Pulmonary embolism Congestive heart failure*† Accessory pathway (Wolff-Parkinson-White) syndrome‡ *Related to increased left atrial size. † Can also be a result of atrial fibrillation. ‡ Especially in patients with a ventricular response rate greater than 200 beats per minute.
Atrial fibrillation often occurs in an otherwise normal heart (lone atrial fibrillation) but is also associated with a variety of underlying diseases (Box 77-9). One etiology of new-onset atrial fibrillation is the “holiday heart” syndrome. This can occur after an ethanol binge, producing atrial fibrillation, atrial flutter, or atrial tachycardia. These rhythms usually revert spontaneously to a sinus rhythm after 24 to 48 hours. Atrial fibrillation may also result from the degeneration of atrial flutter, irrespective of the cause. In this case, an intermediary condition called atrial fibrillation-flutter shows characteristics of both rhythms on ECG. It may be manifested as fine fibrillatory waves with irregular QRS complexes intermixed with flutter waves and a stretch of regular QRS complexes. Rapid atrial fibrillation followed by sinus bradycardia in an elderly patient suggests the aforementioned bradycardiatachycardia syndrome. Finally, irregular atrial fibrillatory waves coupled with regular narrow or wide QRS complexes may represent atrial fibrillation coupled with complete heart block and an accelerated junctional or ventricular rhythm; this syndrome strongly suggests digitalis toxicity. The treatment of atrial fibrillation is based on distinguishing it from other chaotic rhythms (primarily MAT or other atrial rhythms with varying conduction) and the recognition of any underlying causes and symptoms. Asymptomatic atrial fibrillation at a ventricular rate of 100 beats per minute or less
requires no specific emergency therapy. Patients who are unstable from acute rapid atrial fibrillation should receive sedation and synchronized cardioversion starting at 50 to 100 J. Electrical cardioversion is not associated with an increased risk of malignant ventricular dysrhythmias in patients receiving digitalis unless clinical or laboratory evidence of toxicity coexists. In the ED, the therapeutic course chosen depends on the aforementioned principles plus the duration of the dysrhythmia. Both chronic and paroxysmal atrial fibrillation are associated with atrial thrombus formation and embolic events. Chronic but rapid atrial fibrillation is best managed by treating any trigger (notably volume deficits, infection, or decompensated heart failure) and rate control if needed. Embolic risk increases with the duration of atrial fibrillation and when there is underlying valvular disease or chamber enlargement. Prolonged outpatient attempts to establish or maintain a sinus rhythm in patients with recurrent atrial fibrillation do not offer a clear benefit compared with rate control. Patients with new-onset atrial fibrillation of more than 72 hours’ duration should receive ED rate control if needed. Before considering cardioversion in this group through any means, a search for atrial clot or empiric anticoagulation is needed. This usually occurs in an observation or formal admission setting. Clot and embolism can develop during fibrillation and for days after conversion, the latter related to stunned myocardium after restoration of a sinus rhythm. In stable patients with new-onset atrial fibrillation for 72 hours or less, ventricular rate control first is recommended.4,5,35 After carefully assessing the functional implications (seeking evidence of myocardial ischemia, heart failure, syncope, or other symptoms directly attributed to the dysrhythmia), cardioversion may be undertaken, often in consultation with a cardiologist. Intravenous calcium channel blockers (diltiazem or verapamil) or beta-adrenergic blockers are first-line rate-controlling agents in atrial fibrillation. As in atrial flutter, the relative effectiveness and complication rates in clinical practice between these agents are similar absent overt heart failure; titration is the most important guiding principle. Beta-adrenergic blockers are particularly effective in atrial fibrillation secondary to hyperthyroidism or catecholamine excess. Digitalis is a second-line agent for ventricular rate control because of its relatively slow onset of action. Calcium channel and beta-adrenergic blockers, adenosine, and digitalis are not indicated for patients in atrial fibrillation with an accessory pathway because of the risk of precipitating ventricular fibrillation.
PART III ■ Medicine and Surgery / Section Three • Cardiac System
1012
Intravenous magnesium sulfate (2–4 g over 2 minutes) is an adjunctive or third-line therapy to decrease the ventricular response rate. As with its use in atrial flutter, adenosine is not indicated as a primary therapy because of its short duration of effect. For outpatient rate control, calcium channel and betaadrenergic blockers are preferred over digitalis in patients without contraindications. Again, despite intuition, there appears to be no long-term benefit of converting atrial fibrillation to sinus rhythm, as opposed to rate control alone. This is especially true for older patients. However, it may be appropriate to attempt pharmacologic cardioversion for younger patients who present with a first known episode of atrial fibrillation of less than 72 hours’ duration. Pharmacologic cardioversion of atrial fibrillation is best accomplished with procainamide, amiodarone, flecainide, propafenone, or ibutilide.4,5,15,36 Each has conversion rates around 50 to 70%, although there may be a modestly better success rate with ibutilide. Flecainide and propafenone are reserved for patients without significant structural heart disease, hypertension, ischemia, or heart failure. In other words, these agents may be used when the patient’s only indication of heart disease is atrial fibrillation (“lone” atrial fibrillation). In all other cases, procainamide, amiodarone, or ibutilide is an appropriate choice. A calcium channel or betaadrenergic blocker should be given before a type IA agent to control the ventricular rate, with a target of 100 to 120 beats per minute (Box 77-10). This approach may limit any increase in the ventricular rate from the type IA agent. Irrespective of the route and specific agents chosen, pharmacologic cardioversion requires close monitoring and observation. If a class IA or III agent fails, one should not switch to agents of another class for several hours to avoid complications. Type IC agents (flecainide and propafenone) are not available for intravenous use in the United States (but are used in Europe for this indication) and therefore have a limited role for emergency management of atrial fibrillation. Vernakalant
Pharmacologic Approach to Atrial
BOX 77-10 Fibrillation Conversion
Intravenous procainamide, 50 mg/min, up to a total dose of 18 to 20 mg/kg (12 mg/kg in patients with congestive heart failure) or until conversion or side effects occur or Amiodarone, 3 to 5 mg/kg IV, over 15 to 20 minutes or Ibutilide, 0.015 to 0.02 mg/kg IV, over 10 to 15 minutes (conversion usually occurs within 20 minutes if successful) or Oral propafenone 600 mg (contraindicated in setting of structural heart disease or ischemia) or Oral flecainide 300 mg (contraindicated in setting of structural heart disease or ischemia) If needed: A calcium channel blocker (verapamil, 40–80 mg PO or 5-10 mg IV, or diltiazem, 60–120 PO or 15–25 mg IV) can be given before the type IA agent (if no contraindications are present) to lower the ventricular response rate to 250 SF units
Within 48 Hours of Admission Hematocrit fall >10% BUN rise >5 mg/dL Calcium 6 L
Substitute if Gallstone Induced Admission: Age >70 years WBC >18,000/mm3 Glucose >220 mg/dL LDH >400 IU/L AST >250 SF units
Within 48 Hours of Admission Hematocrit fall >10% BUN rise >2 mg/dL Calcium 5 mEq/L Fluid sequestration >4 L
Add the Total Number of Signs at 48 Hours: Number Associated Mortality Rate 0-3 1% 3-4 15% 5-6 40% 100% >7 AST, aspartate transaminase; BUN, blood urea nitrogen; LDH, lactate dehydrogenase; SF, sigma-Frankel; WBC, white blood cells.
two major forms, with trypsinogen-2 present in high serum concentrations in acute pancreatitis. In a recent study, the sensitivity and specificity for urinary trypsinogen-2 dipstick testing were 93% and 92%, respectively. At present, however, neither this test nor any others have proved useful enough for inclusion in standard practice.16,35,40
Additional Laboratory Evaluation In evaluating a patient with abdominal pain, amylase or lipase assays, along with other blood tests, are necessary to narrow the differential diagnosis, detect complications, and determine prognosis. Ranson developed a two-step list of primarily laboratory parameters, performed at hospital admission and after 48 hours, to determine the risk of death from pancreatitis41,42 (Box 89-3). With this in mind, additional testing should consist of a complete blood count (CBC), lactate dehydrogenase (LDH) determination, and a comprehensive metabolic panel (including measurement of liver enzymes, calcium, renal function, and glucose). In patients with liver disease, coagulation studies should be performed to determine the degree of liver dysfunction. Arterial blood gas analysis should be done selectively in patients who are acidotic or hypoxic. This information can be used for treatment decisions and to determine prognosis based on Ranson’s criteria. Magnesium should be checked in alcoholic patients and in those patients with electrolyte abnormalities. Both hypocalcemia and hyperglycemia are common in pancreatitis, with the hyperglycemia resulting from glucagon and insulin abnormalities. Calcium is best determined using the ionized calcium level. Serum calcium is falsely low in the presence of low albumin levels, as may be seen in patients with pancreatitis. Elevations of creatinine and blood urea nitrogen (BUN) may indicate the presence of hypovolemia or renal involvement, or both. Elevation in liver enzymes may result from biliary-induced pancreatitis or from other diseases of the liver or biliary tract. In addition, liver enzyme levels may increase from the pres-
and Specificity of Liver Enzymes for Table 89-1 Sensitivity the Etiology of Pancreatitis35 ENZYME/LEVEL
SENSITIVITY (%)
SPECIFICITY (%)
PPV (%)
ALT >150 mmol/L AST >150 mmol/L Alkaline phosphatase >300 units/L Bilirubin 2.8 mg/dL
48
96
95
44
95
87
24
95
87
38
93
89
ALT, alanine aminotransferase; AST, aspartate transaminase; PPV, positive predictive value.
Data from Yadav D, Agarwal N, Pitchumoni CS: A critical evaluation of laboratory tests in acute pancreatitis. Am J Gastroenterol 97:1309, 2002.
sure on the common bile duct that results from the surrounding pancreatic inflammation. Mild elevations of bilirubin are common in all types of pancreatitis, as well as in many other liver disorders. For the patient diagnosed with pancreatitis, higher elevations of aspartate transaminase (AST) and LDH are related to worse prognosis according to Ranson’s criteria. When liver enzymes are elevated, the pattern of elevation may help determine the underlying cause of the pancreatitis (Table 89-1). Alanine aminotransferase (ALT) is the best single marker for a biliary etiology; levels greater than 3 times baseline support the diagnosis of biliary pancreatitis.5,6,43 The higher the elevation of ALT, the greater the specificity and predictive value for gallstones. ALT levels more than 150 IU/L have 96% specificity, positive predictive value (PPV) of 95%, and 48% sensitivity for gallstone pancreatitis. Significant rises in AST, alkaline phosphatase, and bilirubin also are more likely to be related to biliary pancreatitis but are not as sensitive as ALT.43 The CBC may be notable for an elevated white blood cell count; the hematocrit may be either high or low. Early in the course, the hematocrit may be elevated because of third space volume loss. A decrease in hematocrit is a poor predictor of prognosis because it indicates intra-abdominal hemorrhage and severe pancreatitis. An electrocardiogram also should be done early to determine whether the patient’s abdominal pain may be cardiac in origin.
Prognosis At present, a variety of markers are available to aid in the detection of severe pancreatitis, including specific laboratory values that measure the systemic inflammatory response, scoring systems that tabulate the extent of inflammation or organ failure, and findings on imaging studies. The most commonly used scoring system is Ranson’s criteria (see Box 89-3). In this system, the five criteria evaluated on hospital admission indicate the degree of local inflammation, whereas the six criteria assessed at 48 hours reflect the development of systemic complications. Ranson recognized that the model did not work well for patients with gallstone pancreatitis, so he revised the criteria to reflect the improved mortality. Although Ranson’s criteria have an 89% negative predictive value, the obvious drawback to use of this system in the ED is that the scoring cannot be completed until 48 hours after diagnosis.6,44 Furthermore, in patients with AIDS, Ranson’s criteria may not be as accurate because of HIV-induced changes in the laboratory values such as those of calcium and LDH.25,26
Chapter 89 / Disorders of the Pancreas
BOX 89-3 Pancreatitis-Associated Mortality
1178
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Atlanta Criteria for Predicting Severe
BOX 89-4 Acute Pancreatitis
Criteria for severe acute pancreatitis—one or more of the following: 1. Ranson score 3 or higher on admission (or during the first 48 hours) 2. APACHE II score 8 or higher at any time during course 3. Presence of organ failure ■ Shock (systolic blood pressure less than 90 mm Hg) or ■ Pulmonary insufficiency (Pao2 60 mm Hg or less on room air) or ■ Renal failure (serum creatinine >2 mg/dL after fluid resuscitation) 4. Systemic complications ■ DIC (thrombocytopenia and hypofibrinogenemia and fibrin split products) or ■ Metabolic complications (serum calcium 7.5 mg/dL or less) or 5. Presence of one or more local complications (pancreatic necrosis, pancreatic abscess, pancreatic pseudocyst) APACHE II, Acute Physiology and Chronic Health Evaluation II; DIC, disseminated intravascular coagulation. From Bradley EL III: A clinically based classification system for acute pancreatitis. Summary of the International Symposium on Acute Pancreatitis, Atlanta, GA. Sept 11-13, 1992. Arch Surg 128:586, 1993.
Although Ranson’s criteria constitute a simple and wellknown scoring system, reliance on these criteria alone may result in delayed recognition of illness severity23; therefore, the Acute Physiology and Chronic Health Evaluation II (APACHE II) system also may be used to judge severity.44-46 This system uses 12 physiologic variables, age, and chronic health status to generate a total point score. Scoring can be performed on admission and throughout the hospital stay. An APACHE score of 8 or higher indicates severe disease.37,47 A score greater than 13 is associated with high likelihood of death.5 An APACHE III scoring system also has been developed and includes additional physiologic variables, but this system has not proved to be better at predicting outcomes in patients with acute pancreatitis.48 C-reactive protein levels above 150 mg/L at 48 hours have proved to be a poor prognostic predictor 48 hours after symptom onset.21,47 To account for the various factors that indicate the severity of disease in patients with acute pancreatitis, the Atlanta criteria for severity are widely accepted49 (Box 89-4). Finally, a four-variable predictive model using BUN, age, LDH, and interleukin-6 has been proposed; initial analysis indicates that it is as accurate as other scoring systems.50 This scoring system will need further validation, however.
Radiographic Studies In pancreatitis, abdominal radiographs may show an ileus with a sentinel jejunal loop or spasm of the transverse colon and dilatation of the ascending colon. Pancreatic calcifications of chronic pancreatitis or gallstones may rarely be seen. The portion of the chest seen as part of the abdominal series may show left-sided or bilateral pleural effusions, atelectasis, or ARDS. Up to 80% of radiographs obtained in patients with pancreatitis will demonstrate some abnormality.23 Unfortunately, many of these findings are nonspecific, and plain radiography has largely been replaced by more advanced imaging. CT and ultrasound imaging are complementary studies in the evaluation of pancreatitis. Magnetic resonance imaging
(MRI) offers imaging similar to that achieved with CT without significant advantages. Ultrasonography images the biliary tract with better accuracy than CT; however, the pancreas itself as well as local complications is less well visualized by this modality. It is recommended that an ultrasound study be performed within the first 24 hours of admission, particularly if a biliary etiology is suspected, to determine whether gallstones or dilation of the common bile duct is present.7,38 In one study that compared the results of CT and ultrasonography among patients with pancreatitis, ultrasound findings resulted in a change in treatment in 55% of patients, compared with no changes after CT. CT was 39% sensitive for biliary disease, whereas ultrasonography was 83% sensitive.51 In another study, ultrasound imaging was 94% sensitive for gallstones but only 19% for common duct stones and 38% for common duct dilatation.52 Because of these limitations, when gallstone pancreatitis is highly suspected, an endoscopic ultrasound study may be more accurate and can help guide the emergency use of ERCP.47,53 MRI cholangiopancreatography (MRCP) is a noninvasive test that can image the pancreas and may be used to help determine the cause of acute pancreatitis.54 Although ultrasound imaging is more sensitive for investigating biliary causes of pancreatitis, there are several reasons to perform CT in pancreatitis. The first is to rule out other causes of abdominal pain; the second is to evaluate for the presence of peripancreatic complications such as hemorrhage, pseudocyst, abscess, or vascular abnormalities; and the third is to help determine the extent of any pancreatic necrosis.33,55 The Atlanta International Symposium recommended CT in patients with (1) an uncertain diagnosis; (2) severe clinical pancreatitis, abdominal distention, tenderness, fever with temperatures higher than 102° C, and leukocytosis; (3) a Ranson score greater than 3 or APACHE score greater than 8; (4) no improvement within 72 hours; and (5) acute deterioration.49 If the diagnosis is clear and evidence of obstruction is lacking, CT or ultrasound imaging can be delayed until after the patient has been admitted to the hospital. The main indication for obtaining a CT in the ED is to exclude other diagnoses; however, if the patient is significantly ill and can tolerate the procedure, early CT may help determine if complications are already present. If a CT scan is obtained, a dynamic helical CT study with oral and intravenous contrast is recommended. This study will help differentiate unopacified bowel from a pancreatic abscess or pseudocyst. Recent studies have shown that contrast does not aggravate pancreatitis in humans; however, if the patient cannot tolerate contrast, a noncontrast study will still be helpful.33,56 CT also may be used to stage the severity and prognosis of acute pancreatitis.5 Grades A (no abnormality) and B (focal or diffuse pancreatic enlargement) indicate lower levels of inflammation. Grade C shows mild peripancreatic inflammation and is associated with an increased risk of complication. Grade D (enlarged pancreas with fluid in the anterior pararenal space) and grade E (enlarged pancreas with two or more fluid collections) are associated with significant risk of infection, with mortality rates of up to 15%. The CT severity index is an additional grading system that uses the CT to evaluate how the pancreas looks, as well as evidence of fluid collections or gas adjacent to the pancreas.57
Differential Considerations Pancreatitis must be differentiated from other abdominal processes, cardiopulmonary disorders, and systemic diseases (Box 89-5). An important point is that a number of acute surgical conditions may mimic pancreatitis in presentation and
1179
BOX 89-5 Differential Diagnosis for Pancreatitis
Cardiopulmonary Disorders Myocardial infarction Pericarditis Pneumonia ARDS Pleural effusion Systemic Disease Sickle cell crisis ARDS, acute respiratory distress syndrome.
also may cause elevated amylase. Examples are bowel perforation, peritonitis, ischemic bowel, small and large bowel obstruction, and ruptured ectopic pregnancy.
Management The management of pancreatitis is primarily supportive; rehydration, pain and nausea control, nutritional supplementation, and monitoring for complications constitute the foundations of care. The supportive care given to the patient with acute pancreatitis has multiple objectives. The first is volume replacement. Because of vomiting and fluid sequestration, most patients with pancreatitis are dehydrated. Fluids should be replaced with normal saline; several liters may be required. Vital signs and urine output should be used to judge the adequacy of volume replacement. Electrolytes should be monitored and replenished. A second objective is pain control. Abdominal pain associated with pancreatitis is severe and will generally require narcotic analgesia. Meperidine historically has been used in pancreatitis and biliary disease. Although morphine may increase the tone of the sphincter of Oddi, evidence that it worsens the disease process in pancreatitis is lacking.58 Patientcontrolled analgesia may be the most effective method of pain control.38 Antiemetics are indicated to control nausea or vomiting. A third objective is to ensure adequate nutrition. In the past, patients were allowed nothing by mouth and nasogastric suctioning was initiated out of concern that oral intake would stimulate the release of pancreatic enzymes. However, randomized clinical trials in patients with mild to moderate pancreatitis have shown no benefit from either fasting or use of nasogastric suctioning.59,60 At this time, nasogastric suction is indicated only in cases of intractable vomiting or ileus, and some enteral feeding should begin as soon as it is tolerable. Some evidence suggests that early enteral nutrition may improve outcomes even in severe pancreatitis61; however, if oral feedings are not tolerated or are inadequate, then parenteral feedings should be initiated.3
Chapter 89 / Disorders of the Pancreas
Abdominal Disorders Perforated viscus Peptic ulcer disease Cholecystitis, gallbladder colic Cholangitis Gastroenteritis Nephrolithiasis or pyelonephritis Bowel obstruction Mesenteric ischemia Abdominal aortic aneurysm Ectopic pregnancy
Objective four is reevaluation for complications of pancreatitis. Hypotension should be corrected with large volumes of normal saline (up to 6 L). Invasive hemodynamic monitoring may become necessary. Airway control is appropriate for respiratory failure or continued shock. Hyperglycemia should be treated cautiously because it may self-correct as the pancreatitis resolves. Hypocalcemia may be the result of decreased albumin or hypomagnesemia, so ionized calcium and magnesium levels should be checked before replacement therapy is initiated. If true hypocalcemia is present and the patient is experiencing symptoms, then treatment is appropriate. Calcium gluconate should be used if the calcium must be replenished. The serum potassium should be normalized before calcium replacement, however, because calcium will cause intravascular potassium shifts. In the case of gallstone pancreatitis, gastroenterology consultation is appropriate to discuss the use of ERCP. Early operative removal of gallstones and the gallbladder has been shown to increase mortality62; however, early removal of common bile duct stones by ERCP may reduce morbidity. At present, consensus is lacking regarding the optimal timing of ERCP in the presence of gallstone pancreatitis.63 Early endoscopic sphincterotomy (in 24 to 48 hours) and stone removal are recommended in the setting of cholangitis, sepsis, and severe obstructive pancreatitis.5,52,64 In mild pancreatitis, early ERCP has not consistently been shown to decrease morbidity. In addition, there is approximately a 5% rate of pancreatitis with ERCP and papillotomy, as well as other complications associated with the procedure (bleeding and perforation). In view of the ongoing controversy, it is appropriate to involve the consultant early in the case so that a well-coordinated plan can be created. Theoretically, the following medications should moderate the course of pancreatitis: Histamine H2 receptor blockers decrease the release of secretin by inhibition of gastric acid, glucagon directly suppresses pancreatic exocrine secretion, and octreotide inhibits pancreatic secretion. However, these therapies have not been shown to be clinically effective.5,6 Other approaches using inhibitors of inflammatory mediators also have failed to show clinical improvement.5,6,20 For patients with severe pancreatitis, an H2 blocker, although not helpful for the acute disease, may decrease stress-induced ulcers. Use of antibiotics in those patients with severe pancreatitis with or without evidence of necrosis of pancreatic tissue is controversial. Prophylactic antibiotics have been reported to be effective in reducing subsequent infection.5,8,65 These positive effects have not been seen in all studies, however, and some evidence indicates that use of prophylactic antibiotics in any patient with severe pancreatitis may increase the risk of fungal infection.66 A recent study with good methodology compared intravenous ciprofloxacin plus metronidazole with placebo and found no difference with respect to the development of infected pancreatic necrosis.67 At this time, the literature regarding the use of early antibiotics remains unsettled, although there appears to be growing acceptance of with holding antibiotics until clear evidence of infection is present.3,16,37,47,61,68 On the basis of this reality, early discussion with a consultant would be prudent, and in the absence of expert opinion, it remains reasonable to begin broad-spectrum antibiotics in those patients with severe acute pancreatitis. Surgical intervention or percutaneous drainage may be necessary for cases of infected pancreatic necrosis, infected pseudocyst, or unresolved pseudocyst. Surgery is preferred when percutaneous drainage is not effective or not possible (as with extensive pancreatic necrosis or deteriorating clinical status).18,39
1180
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Disposition The course of acute pancreatitis is unpredictable, and complications may occur hours or days after the onset of illness; therefore, almost all patients with acute pancreatitis should be admitted to the hospital. Patients with evidence of severe pancreatitis should be admitted to the intensive care unit (ICU), especially if findings include agitation, confusion, rising hematocrit, pulmonary insufficiency, or cardiovascular problems such as hypotension, progressive tachycardia, or poor urine output. Specifically, patients with a score greater than 2 on Ranson’s criteria on admission or other evidence of organ failure, local complications, or significant comorbid illness will also require ICU management.5 In hospitals without appropriate intensive care facilities, patients with evidence of severe pancreatitis should be transferred to an appropriate treatment center. If the cause of pancreatitis is suspected or confirmed to be gallstones, discussion with a specialist (gastroenterologist or surgeon) is necessary. Pediatric patients have increased morbidity and mortality from pancreatitis and should be considered for early transfer to a pediatric specialty center.
Rarely, pseudocysts can erode into vascular structures or can become infected. Narrowing of the bile duct from extrinsic pressure or strictures may lead to elevation of liver enzymes and jaundice. In approximately 5% of patients, duodenal obstruction develops secondary to inflammation around the head of the pancreas. Thus, chronic pancreatitis in an acutely ill patient may be a manifestation of the primary disease or a complication. The most common endocrine complication is the presence of glucose intolerance in many patients with chronic pancreatitis. Over years, insulin-dependent diabetes develops in 50 to 75% of patients.69,72 Patients with chronic pancreatitis have an increased risk for the development of pancreatic carcinoma of approximately 4%, as do patients with hereditary pancreatitis. Chronic pancreatitis is associated with considerable morbidity in terms of pain and complications. In addition, patients with chronic pancreatitis have an excess mortality of approximately 20%; however, the cause of death is more likely to be related to other consequences of alcoholism than to pancreatitis.
Clinical Features ■ CHRONIC PANCREATITIS Principles of Disease Chronic pancreatitis is an ongoing inflammatory process leading to irreversible structural damage and impairment of exocrine and endocrine pancreatic function. Normal pancreatic structure is replaced with fibrotic tissue, resulting in pancreatic ducts that are strictured in some areas and dilated in others. The incidence is approximately 4 per 100,000.69 In 70 to 80% of cases, the cause is chronic alcohol use. The risk increases with the duration and amount of alcohol consumption. The ingestion of more than 150 g of alcohol per day for an average of 5 to 15 years is associated with the development of chronic pancreatitis; this form of the disease affects 3 to 15% of chronic alcoholics.69-71 It is possible that chronic pancreatitis also may develop in persons sensitive to small amounts of alcohol. Three theories exist as to the mechanisms by which alcohol causes chronic pancreatitis have been proposed: (1) direct cellular toxicity, (2) alcohol-induced precipitation of proteinaceous fluid in the ductules, which causes obstruction and calcification, and (3) injury caused by recurrent acute pancreatitis leading to irreversible damage and chronic inflammation.69,71,72 Chronic pancreatitis can continue even after the cessation of alcohol use, although it is more commonly associated with alcoholic relapse. Other, less common causes of chronic pancreatitis include ductal obstruction, autoimmune pancreatitis, hereditary pancreatitis, cystic fibrosis, trauma, autoimmune, hyperparathyroidism, α1-antitrypsin deficiency, hyperlipidemia, and tropical pancreatitis (cassava fruit is implicated).69 Idiopathic chronic pancreatitis occurs in approximately 10% of patients. In the 25% of cases of unknown cause, occult alcohol use may be the culprit. In children the most common causes are cystic fibrosis and hereditary pancreatitis.69,71 The pathophysiology of chronic pancreatitis includes chronic calcific pancreatitis, usually seen in alcoholics and characterized by patchy fibrosis, ductal injury, intraductal protein plugs, stones, and chronic inflammatory pancreatitis with diffuse fibrosis and inflammatory changes. As in acute pancreatitis, chronic inflammation can cause local injury resulting in lesions such as pseudoaneurysms, splenic vein thrombosis, pancreatic ascites, or pancreatic fistulas. Pancreatic pseudocyst formation is seen in up to 25% of patients with chronic pancreatitis.
Patients with chronic pancreatitis may present with nearconstant and intractable chronic pain, or with complications from chronic pancreatitis, or may experience an acute flare of chronic underlying disease. These flares may be severe, and if the patient is relatively well during the interim it may be difficult to distinguish this episode from a recurrent bout of acute pancreatitis.71 When present, the pain is epigastric, usually radiating to the back and associated with nausea and vomiting. Often, the pain is similar to that in previous attacks of acute pancreatitis or flares of chronic pancreatitis. Use of alcohol or eating exacerbates the pain. No correlation has been found between remaining pancreatic function and the degree of pain, although a few studies have shown that the pain may diminish over years of chronic disease. Patients may experience weight loss because of malabsorption or decreased intake secondary to nausea and vomiting or to the recognition that eating may precipitate the pain. In approximately 15% of patients with chronic pancreatic disease, signs and symptoms of pancreatic exocrine function insufficiency, including malabsorption, diarrhea, steatorrhea, and weight loss, may develop over time. Malabsorption occurs after approximately 90% of the pancreas is nonfunctional. Functional endocrine insufficiency will develop in approximately one third to two thirds of patients with chronic disease. The symptoms primarily manifest as hyperglycemia; however, the development of diabetic ketoacidosis is rare. Hypoglycemia is multifactorial in origin, with causes including an insufficiency of glucagon, decreased liver glucose stores, malnourishment, and hypoglycemic medications. On physical examination, patients may appear to be in significant discomfort. They frequently appear chronically ill from alcoholism, poor nutrition, and malabsorption. On abdominal examination, palpation frequently reveals tenderness without peritoneal signs. The abdomen should be carefully palpated for a mass that may represent a pseudocyst or tumor. Stigmata of chronic alcohol abuse also may be noted; however, jaundice may be from pressure on the common bile duct or from alcohol-related liver injury.
Diagnostic Strategies The diagnosis of chronic pancreatitis frequently is made clinically rather than on the basis of laboratory testing. The serum
Figure 89-3. Pancreatic calcifications (arrowheads) throughout the pancreas as seen in chronic pancreatitis. (Incidental finding of feeding tube in main bronchus.) (Image courtesy of Ronald Arildsen.)
1181
creatic duct abnormalities and measuring pancreatic function. Some gastroenterologists consider the ductal abnormalities seen on ERP and endoscopic ultrasound to be pathognomonic for chronic pancreatitis. In the future, MRCP may assume a greater role in the evaluation of chronic pancreatitis.72
Differential Considerations The diagnosis of chronic pancreatitis usually is straightforward in the alcoholic patient with hyperamylasemia who has chronic abdominal pain and a history of similar, previous flares of pancreatitis. The diagnosis may be more difficult when amylase and lipase levels are normal. However, the clinician should not be lulled into complacency and forget that other abdominal processes, unrelated to either the pancreas or the complications of pancreatitis, are legitimate considerations in the differential diagnosis (see Box 89-5). In addition, other chronic abdominal disease such as peptic ulcers, irritable bowel, gallstones, and endometriosis may manifest with recurrent abdominal pain. Finally, narcotic-dependent patients in withdrawal may experience vomiting and abdominal pain that may be difficult to differentiate from chronic pancreatitis.
Management The initial management of chronic pancreatitis is supportive. Depending on the patient’s clinical status and electrolyte values, replenishment of fluids and electrolytes may be necessary. An “alcohol cocktail” with thiamine, multivitamins, and folate often is indicated because patients frequently are malnourished. Antiemetics should be used to manage recurrent emesis. Management of pain is one of the most important and most difficult aspects of treatment. Laboratory values may be normal despite significant pain, and patients with chronic pain syndromes may not exhibit signs of autonomic hyperactivity when experiencing exacerbations of their underlying disease. This lack of correlation may lead to a concern that the expressed need for pain medication constitutes drug-seeking behavior. Physicians should err on the side of treatment in most patients, except those with documented abuse. Nonsteroidal analgesics and acetaminophen are the preferred drugs for control of pain but often are not adequate. Either morphine or meperidine may be used and should be titrated to effect. Tramadol (Ultram) also has been used effectively.73 The use of narcotics over extended periods may be necessary. Non-narcotic modulators of pain, such as selective serotonin reuptake inhibitors or gabapentin, may be helpful for chronic pain. In the ideal medical system, the primary care physician or a pain management specialist monitors the narcotic prescription, because narcotic dependence may become an issue. The removal of inciting factors, especially alcohol, is important. Smoking also plays a role in the development of chronic pancreatitis that is independent of alcohol.71 Patients with significant pain should have nothing by mouth, although as in acute pancreatitis, a nasogastric tube is not indicated. The use of oral pancreatic replacement enzymes increases the amount of trypsin in the duodenum and may decrease stimulation of the pancreas, with subsequent decrease in pain. Studies of the effectiveness of pancreatic enzyme replacement on pain have yielded contradictory results.71 These enzymes should be used in patients with malabsorption or steatorrhea. In theory, proton pump inhibitors or H2 receptor blockers also may reduce pancreatic stimulation; however, these agents have not been shown to decrease pain or hasten recovery. Octreotide lowers cholecystokinin levels and may inhibit pancreatic secretion.71
Chapter 89 / Disorders of the Pancreas
levels of amylase and lipase initially are mildly elevated in chronic pancreatitis, but as the disease progresses, these levels normalize. As in acute pancreatitis, the degree of elevation of amylase and lipase is not prognostic. In the patient with the appropriate clinical picture, normal amylase and lipase levels are consistent with a diagnosis of chronic pancreatitis. Blood work should include a CBC and complete metabolic profile. The white blood cell count usually is normal. There may be elevations of hepatic enzymes (alkaline phosphatase, bilirubin, or transaminases) either from alcoholic hepatitis or from compression on the biliary duct from pancreatic inflammation or a mass in the head of the pancreas. Elevations in serum glucose also may be seen; hypoglycemia is less common. Decreases in albumin and calcium are common because of the chronic nature of the disease. If the diagnosis is in question, stool may be tested for fecal fat and pancreatic enzymes such as elastase, chymotrysinogen, and trypsinogen.71 Although abdominal radiographs are not necessary, presence of pancreatic calcifications is pathognomonic. Such calcifications are seen in 30 to 50% of patients, usually related to chronic alcohol-induced pancreatitis (Fig. 89-3). Patients with calcifications have had pancreatitis for an average of several years; therefore, these patients should be evaluated for longterm complications such as diabetes and malabsorption. In patients with the clinical manifestations of chronic pancreatitis, including pain, malabsorption, and diabetes, the diagnosis is made by CT scan, ERCP, or endoscopic ultrasound.71 In the ED setting, patients with known chronic pancreatitis do not need imaging except when the underlying cause of the pain is in question or if the pain is prolonged, significantly increased, or unresponsive to treatment. CT shows dilated intrapancreatic ducts, microcalcifications, pseudocysts, or other complications. CT is 90% sensitive for chronic pancreatitis and is the preferred modality when imaging is indicated. Although ultrasound imaging is useful in diagnosing the cause of acute pancreatitis, it is less useful in chronic pancreatitis, with a sensitivity of 75% and specificity of 80 to 90%. The primary ultrasound findings are pancreatic calcifications and ductal abnormalities. Although endoscopic retrograde pancreatography (ERP) is not an ED procedure, it can be helpful in diagnosing pan
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1182
In general, patients with chronic pancreatitis are managed as outpatients and present to the ED with exacerbations or complications. Because acute pancreatitis can occur in patients with chronic pancreatitis, the same prognostic indicators for severity of acute pancreatitis should be considered. Patients with severe disease should be admitted to the ICU. Patients with dehydration, abdominal pain unresponsive to medications, or questionable diagnosis should be hospitalized for evaluation and treatment. After a careful ED evaluation, in the absence of dehydration, unstable vital signs, or uncontrolled pain, the patient may be managed on an outpatient basis with close follow-up. After hospital discharge, it is important to stress lifestyle modifications of abstinence from alcohol and tobacco, as well as intake of frequent, small low-fat meals.
is diagnosed; therefore, patients may present with pain of long duration or with one of the many complications of the disease.80 One of the most common presentations is weight loss, which usually is the result of anorexia rather than malabsorption. The patient may complain of dull, constant abdominal pain in the epigastrium that may radiate to the back. Alternatively, the patient may present with painless jaundice from common bile duct obstruction, and progressive jaundice develops in approximately 75% of patients. An enlarged, palpable, but painless gallbladder in the presence of jaundice is most commonly associated with pancreatic cancer (termed Courvoisier’s sign). Glucose intolerance also may develop. As the tumor enlarges, patients may exhibit evidence of bowel obstruction. Pancreatic cancer (as well as other cancers) may render patients hypercoagulable, resulting in thromboembolic presentations. Varices and gastrointestinal bleeding may be caused by compression of the portal system. Neuroendocrine tumors of the pancreas are rare and manifest with symptoms that reflect the hormones they produce. For example, with insulinomas, the presenting manifestation may be hypoglycemia. Gastrinomas are related to Zollinger-Ellison syndrome and recurrent peptic ulcers. VIPomas (also known as Verner-Morrison syndrome) manifest with extreme watery diarrhea, hypokalemia, and achlorhydria. Glucagonomas manifest with glucose intolerance and necrolytic migratory erythema. Some tumors produce multiple hormones.79 Other, nonfunctional tumors also may be noted incidentally on CT scans as small pancreatic masses. Diagnosis is made by measurement of abnormal levels of hormones and recognition of the appropriate clinical syndrome. Fifty percent of pancreatic neuroendocrine tumors are malignant.81
■ PANCREATIC CANCER
Diagnostic Strategies
Perspective
The diagnosis of pancreatic cancer may be made by ultrasonography, although CT scan provides better imaging of the cancer. MRI also may provide information. Percutaneous ultrasound-guided biopsy, endoscopic ultrasound examination and biopsy, or CT-guided biopsy can be used to obtain tissue diagnosis.77 Histologic samples are needed to differentiate ductal adenocarcinoma from islet cell tumors, other metastatic cancers, and lymphoma. Serologic markers have not proved satisfactory for diagnosis or follow-up evaluation, although several oncogenes and tumor markers are under study (CA19-9 and CEA).
Beyond ED treatment, helpful adjuncts may include endoscopic dilation, ductal stone removal, extracorporeal shock wave lithotripsy, or stenting of the pancreatic ducts. Common bile duct stenting also may be necessary, because obstruction occurs in approximately 5 to 10% of cases.71 Surgery such as pancreatic head resection, lateral pancreaticojejunostomy, or Whipple pancreatic duodenectomy is sometimes an option when conservative treatment has failed. One recent study showed improved pain and physical health summary scores in patients randomized to undergo surgery rather than endoscopy.74 Pancreatic pseudocysts in chronic pancreatitis are less likely to resolve spontaneously and should be drained either endoscopically under ultrasound guidance or in an open procedure. Celiac plexus blocks also have been used with minimal success for pain control.
Disposition
Pancreatic cancer is a particularly lethal cancer, with a 5-year survival rate of 4% despite aggressive surgery and advances in chemotherapy. It is the fourth most common cause of cancerrelated death in the United States. The disease is diagnosed in approximately 11 people per 100,000 per year, and the incidence has increased three-fold over the past 40 years. Because early symptoms are few, a minority of patients (less than 20%) are diagnosed at an early stage.28,75,76
Principles of Disease Little is known about the etiology of pancreatic cancer. The most consistently reported risk factors are smoking, advanced age, and positive family history. Chronic alcoholism, chronic pancreatitis, and diabetes have been shown to be risk factors in some studies.76-78 Ductal adenocarcinomas account for 95% of malignant pancreatic tumors. The pancreatic head is the location of origin in 70% of cases. The tumor extends locally into adjacent structures and can metastasize by hematogenous or lymphatic spread to liver, peritoneum, lungs, bones, and brain. Neuroendocrine tumors, such as gastrinomas, vasoactive intestinal peptide (VIP)-omas, and glucagonomas make up the remaining cases.79 These types of tumors carry a better prognosis.
Clinical Features The presentation of pancreatic adenocarcinoma is variable because progression of the disease is indolent. The tumor usually has been present for several months before the cancer
Management Complete resection of the carcinoma is the only effective treatment. Unfortunately, few tumors (less than 20%) are diagnosed at a stage at which this may be possible.77 In patients with unresectable tumors, the median survival period is approximately 6 months. Palliative surgery may be performed to relieve obstruction. Biliary drainage by percutaneously placed or ERCP-placed stents also may help to relieve jaundice. Chemotherapy and radiation therapy may decrease tumor size to ease pain and prolong survival in some patients.77,82 Treatment of neuroendocrine tumors is aimed at limiting tumor growth by both excision and reversal of hormone excess.82 Patients may present to the ED with complications of the cancer such as bowel obstruction, jaundice, or problems with pain control. In view of the grim prognosis for this disease and the significant associated pain, narcotics should not be withheld, and end-of-life issues should be addressed by the oncologist.
1183
KEY CONCEPTS Most cases of acute pancreatitis are caused by gallstones (45%) and alcoholism (35%). Other causes include medications, toxins, and trauma. ■ The clinical spectrum of acute pancreatitis ranges from mild (epigastric discomfort often associated with vomiting) to life-threatening (severe abdominal pain in the presence of an acute abdomen and hemodynamic instability due to systemic complications). The mortality rate for severe pancreatitis approaches 30%. ■ There is no perfect test for diagnosing acute pancreatitis. The most useful tests include serum amylase and lipase assays. Unfortunately, both tests can yield normal results in up to 25% of cases, and mild elevations are not specific for acute pancreatitis and can be seen in many other acute surgical disorders causing abdominal pain. Both tests are highly specific for pancreatitis when serum
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
levels are elevated 5 times above the upper limits of normal. ■ Emergent abdominal CT should be performed in patients with clinically suspected pancreatitis who appear acutely ill (to exclude peripancreatic complications such as hemorrhage, pseudocyst, or abscess) and in patients with an uncertain diagnosis (to exclude other surgical causes of acute abdominal pain). ■ Because the course of acute pancreatitis is unpredictable, patients should be hospitalized for pain control, hydration, observation, and the management of complications. Patients with severe pancreatitis (i.e., those who have more than two of Ranson’s criteria, an APACHE score over 7, or evidence of systemic complications) should be cared for in an ICU.
Chapter 89 / Disorders of the Pancreas
■
Chapter 90
Disorders of the Small Intestine
Susan P. Torrey and Philip L. Henneman
■ SMALL BOWEL OBSTRUCTION Perspective The signs and symptoms of intestinal obstruction have been recognized for centuries. This clinical entity has historically been treated with a variety of interventions, including enemas and inflation of the rectum, metallic mercury ingestion, therapeutic bleeding, and percutaneous intestinal puncture. By the late 19th century, proximal intestinal decompression was reliably used to provide temporary symptomatic relief of intestinal obstruction. Advances in the 20th century and beyond, including the development of antibiotics and improved surgical techniques, have significantly improved the prognosis for patients with small bowel obstruction (SBO). Obstruction of the small bowel accounts for 15% of hospital admissions for acute abdominal complaints.1 Approximately 300,000 operations are performed in the United States each year for relief of intestinal obstruction. Aggressive treatment has resulted in a current mortality rate of less than 5%, a substantial improvement over the expected 60% mortality rate for this disease in 1900. When strangulation complicates SBO, however, the mortality rate increases to as much as 30%. Death from SBO occurs most often in the elderly or in patients with significant underlying illness. The term mechanical obstruction implies a physical barrier to the flow of intestinal contents. Within this definition, simple obstruction refers to the situation in which the intestinal lumen is partially or completely occluded at one or more points, thus producing proximal intestinal distention, but without compromise of the intestinal vascular supply. A closed-loop obstruction implies that a segment of bowel is obstructed at two sequential sites, usually by twisting about a constricting adhesive band or hernia opening. This mechanism of obstruction is associated with a high risk for compromise of intestinal blood flow with resulting intestinal ischemia, a condition referred to as strangulation obstruction. Not all closed-loop obstructions are associated with intestinal ischemia, and other types of obstruction may eventually involve vascular compromise. In contrast with mechanical obstruction, neurogenic or functional obstruction occurs when intestinal contents fail to pass through the bowel lumen because of disturbances in gut motility rather than actual blockage. This entity also is commonly referred to as an adynamic ileus. When intestinal peristalsis fails, dilatation of the involved intestinal tract develops. Adynamic ileus most commonly is seen after abdominal surgery but can be caused by other common medical conditions (Box 90-1). 1184
Focal decrease in peristaltic activity may occur because of a localized inflammatory process (e.g., pancreatitis, cholecystitis, appendicitis) and result in gas and fluid accumulation in an isolated segment of bowel. This segmental ileus is called a sentinel loop. Pseudo-obstruction refers to a poorly understood disorder of intestinal motility associated with a number of medical conditions, including amyloidosis, collagen vascular disease, diabetes, hypothyroidism, and several metabolic disorders— hypokalemia, hypocalcemia, and uremia. The signs and symptoms of intestinal obstruction are present, but evidence of an underlying lesion or cause is absent on diagnostic evaluation. Correction of any underlying disease process and supportive care are recommended, but the results of treatment often are disappointing.
Principles of Disease A relationship between the progressive physiologic changes that occur in patients with SBO and the corresponding clinical manifestations is well documented. Mechanical SBO initially causes mild proximal intestinal distention that results from the accumulation of normal gastrointestinal secretions and swallowed air above the obstructing lesion. This distention stimulates peristalsis above and below the obstruction, which accounts for the frequent loose bowel movements that may accompany partial and even complete SBO in the early stages.1 Early bowel distention stimulates epithelial cell secretory activity, resulting in the addition of more fluid, increasing bowel dilatation, and the creation of a self-perpetuating process. This situation is worsened by the inability of the distended bowel to absorb fluid and electrolytes at a normal rate. Further increases in intraluminal pressure result in capillary and lymphatic obstruction with subsequent edema of the bowel wall. Perforation occurs if this process continues uninterrupted. In addition, vomiting and intraperitoneal fluid sequestration further compound volume losses, leading to extracellular fluid depletion, hypovolemia, and, eventually, shock. The rise in intraluminal pressure is much more abrupt with a closed-loop obstruction because the intestinal contents also are prevented from retrograde flow. Strangulation occurs, with subsequent development of venous congestion, small vessel rupture, intramural and mesenteric hemorrhage, and arterial insufficiency. It also is not uncommon for the loop of distended bowel to twist on itself further, resulting in large artery
1185
BOX 90-2 Lesions Causing Small Bowel Obstruction
Abdominal trauma Infection (retroperitoneal, pelvic, intrathoracic) Laparotomy Medications (e.g., narcotics) Metabolic disease (hypokalemia) Renal colic Skeletal injury (rib fracture, vertebral fracture)
Intrinsic Congenital (atresia, stenosis) Inflammatory (Crohn’s disease, radiation enteritis) Neoplasms (metastatic or primary) Intussusception Traumatic (hematoma)
occlusion. Either sequence of events then progresses rapidly from intestinal ischemia to infarction. Necrosis of the bowel and leakage of contaminated contents cause bacterial peritonitis and sepsis. Although the proximal small bowel normally contains few bacteria, this changes quickly during times of intestinal stasis. Simple intestinal obstruction has been shown to be associated with increased bacterial translocation to mesenteric lymph nodes. In one series, 59% of the patients undergoing laparotomy for simple SBO had bacteria (most commonly Escherichia coli) cultured from mesenteric lymph nodes, compared with only 4% of patients operated on for other reasons.1,2 The most common causes of SBO are listed in Box 90-2. In developed countries, postoperative adhesions are now responsible for more than 50% of all cases. It is estimated that as many as 15% of abdominal surgeries eventually result in SBO from adhesions. A particularly high incidence of SBO is found after gynecologic or intestinal surgery, as well as in patients who have previously undergone surgery in the presence of peritonitis or significant abdominal trauma.3 Other important causes of SBO include hernias and neoplasms, each with an incidence of approximately 15%.3,4 The incidence of obstruction related to hernias has been steadily decreasing in developed countries because of elective treatment of external hernias.5 Although hernias account for a relatively small proportion of bowel obstructions, they are associated with a high rate of strangulation (28% with hernias versus 8% with adhesive obstruction). Anatomically, strangulation occurs because many obstructions caused by hernias are of the closed-loop type. When neoplasm is associated with SBO, the cause most often is colon cancer, followed by pancreatic, gastric, and gynecologic malignancies. Several less common causes of SBO are pertinent to the practice of emergency medicine. Gallstone ileus is rare in the general population but accounts for 25% of nonstrangulated SBOs in patients older than 65 years.6,7 In this entity, a gallstone erodes through an inflamed gallbladder wall into a loop of adjacent small bowel. The stone then passes through the bowel lumen until it meets some narrowing, typically at the distal ileum, where it produces mechanical obstruction. This problem occurs predominantly in elderly patients, so the associated 15 to 18% mortality rate is not surprising. Another unique cause of SBO is an obturator hernia. This hernia typically occurs in elderly emaciated women with significant concomitant medical illness but no previous abdominal surgery. It is believed that women with a wider pelvis and more oblique obturator canal are predisposed to the development of obturator hernia in the presence of decreased preperitoneal fat related to emaciation and chronic increased intra-abdominal pressure related to associated medical disease. This hernia is difficult to detect and often is diagnosed only when it arises as SBO. Both of these uncommon causes of SBO occur in the elderly, a group that is becoming an increasing percentage of the emergency department population. Another uncommon but noteworthy cause of SBO is small bowel volvulus.8 This condition results from abnormal twisting
Extrinsic Hernias (internal and external) Adhesions Volvulus Compressing masses (tumors, abscesses, hematomas) Intraluminal Foreign body Gallstones Bezoars Barium Ascaris infestation
of a loop of bowel around the axis of its own mesentery. Although volvulus of the colon (sigmoid and cecum) is common, volvulus of the small bowel is rare. Primary small bowel volvulus occurs in an otherwise normal abdominal cavity and is seen most often in adult patients in Africa, the Middle East, and the Indian subcontinent. It is rarely seen in Europe and North America. Secondary causes of small bowel volvulus include malformation and malrotation of the intestine and tethering of the loop of bowel at its apex as a result of postoperative adhesions. Early surgical intervention is important because this classic form of closed-loop obstruction is associated with a high incidence of strangulation. Intussusception occurs in all age groups but is primarily a disease of infancy and early childhood, constituting the most common cause of SBO in early childhood. Only 5% of all intussusceptions occur in adults, and intussusception accounts for only 5% of cases of SBO in adults.9,10 An intussusception occurs when a segment of bowel telescopes into an adjacent segment, resulting in obstruction and ischemic injury to the intussuscepting segment. In contrast with the idiopathic nature of most childhood intussusceptions, a mechanical cause is present in more than 90% of adult cases. Tumors, either benign or malignant, act as the lead point of intussusception in more than 65% of adult cases. Several reports have described adult intussusception associated with acquired immunodeficiency syndrome (AIDS). In this setting, the lesions generally are in the ileum, and intussusception has been associated with lymphoma or unusual inflammatory processes, including atypical mycobacterial infection. Clinical manifestations of intussusception in the adult patient are nonspecific and may occasionally be chronic or recurrent in nature. Abdominal pain is a prominent complaint, often associated with symptoms and signs suggestive of obstruction (nausea, vomiting, and abdominal distention). Radiographic features of intussusception also are nonspecific. Plain films may reveal evidence of partial or complete bowel obstruction. It has been recommended that ultrasonography may be useful in the diagnosis of adult and pediatric cases. The mainstay of diagnosis, however, remains contrast studies, typically abdominal computed tomography (CT) with an oral contrast agent. Although reduction of the intussusception may occur during contrast studies, surgery is recommended for adult patients because of the high incidence of pathologic lesions as a cause of intussusception.10
Chapter 90 / Disorders of the Small Intestine
BOX 90-1 Causes of Adynamic Ileus
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1186
Finally, a fascinating and lengthy list of unusual causes of SBO has been reviewed and includes pharmacobezoars, Ascaris lumbricoides infection, and endometriosis. These reportable causes of SBO account for less than 6% of all cases.11
Clinical Features History. Patients with SBO typically complain of regularly recurrent bouts of poorly localized abdominal pain lasting from seconds to minutes. The painful spasms occur every few minutes with proximal intestinal obstruction and less frequently with more distal obstruction. The pain is described as crampy in nature, and each episode has a characteristic crescendo-decrescendo pattern. A change in the description of the pain from intermittent and colicky to constant and severe may signal the development of complications, such as intestinal ischemia or perforation. In general, the more proximal the obstruction, the greater the patient’s discomfort and the shorter the delay between onset of symptoms and presentation. Several hours of severe colicky pain in association with bilious vomiting and mild abdominal distention are typical of proximal intestinal obstruction, whereas a day or two of progressively worsening pain and more prominent abdominal distention is typical of distal intestinal obstruction. When vomiting does occur with distal intestinal obstruction, it often is feculent from bacterial proliferation. With complete intestinal obstruction, obstipation eventually develops, whereas with early or partial obstruction, the patient may continue to pass stool or flatus. Physical Examination. The physical examination should begin with a brief but thorough assessment of the patient’s degree of distress, vital signs, and general condition. These important parameters determine the urgency of the evaluation and management of the patient. Examination of the abdomen should include inspection for distention and a careful search for surgical scars and external hernias. Auscultation may reveal hyperactive bowel sounds— in particular, rushes or high-pitched “tinkles” produced by forceful peristaltic efforts. Late in the course of bowel obstruction, bowel sounds may become hypoactive or may be absent. Percussion may elicit tympany with distal obstruction and ileus. Palpation may reveal a tender mass, especially with a closed-loop cause of obstruction. A rectal examination should be performed to evaluate the rectal vault and to check for gross or occult blood. A prospective study of patients with abdominal pain defined six clinical variables that had high sensitivity and positive predictive value for the diagnosis of bowel obstruction12: a history of previous surgery, history of constipation, age older than 50 years, vomiting, abdominal distention, and increased bowel sounds. The presence of peritoneal signs usually indicates late obstruction with complications, including strangulation. Of note, however, aggressive abdominal palpation in the setting of bowel dilatation can give the false impression of peritonitis, because quick compression-decompression of dilated bowel may elicit a significant pain response. Determining the presence of pain with cough or gentle shaking of the patient’s pelvis along with percussion tenderness often is helpful in differentiating peritonitis from pain related to rapid decompression of dilated loops of bowel. Other clues to serious complications include alterations of the vital signs with tachycardia, hypotension, and fever indicating early sepsis. Unfortunately, a number of studies document that even experienced physicians cannot reliably distinguish between strangulation and simple intestinal obstruction on the basis of examination alone.1,3 Complications. Complications associated with SBO include hypovolemia, intestinal ischemia and infarction, peritonitis
and sepsis, and respiratory compromise from elevation of the diaphragm or aspiration of gastrointestinal material. The incidence of complications will reflect in part the degree of intestinal dilation, which in most cases is directly related to delays in presentation, diagnosis, and initiation of appropriate treatment. Early diagnosis and management can prevent or minimize much of the preoperative morbidity associated with this disease. Complications related to surgical management of SBO include recurrence of the obstruction, hemorrhage, wound infection, abscess formation, sepsis, and short bowel syndrome.13-15 Much of the eventual postoperative morbidity and mortality is related to the patient’s underlying medical condition. Whether the condition is treated surgically or conservatively, the long-term recurrence rate for adhesive SBO is significant—40% with nonoperative treatment and 27% with operative treatment.16
Diagnostic Strategies Routine laboratory testing yields nonspecific findings. Leukocytosis is common with both simple and strangulated obstructions and is not a reliable marker for intestinal compromise. Serum markers of intestinal compromise and ischemia, including creatine kinase (CK), amylase, and lactate, are elevated late in the course of bowel obstruction. An electrolyte panel and renal function testing are appropriate if significant volume loss is evident. Conventional and special radiographic examinations of the abdomen are the most useful diagnostic adjuncts for the evaluation of patients with suspected SBO. These studies may confirm or exclude the presence of bowel obstruction; identify the site, severity, and cause of the obstruction; and help distinguish simple obstruction from strangulation.17-20 Characterization of the obstruction in this manner will determine the need for urgent surgical intervention versus a period of conservative, nonoperative management. An adequate plain radiographic examination of the abdomen requires at least two films, one with the patient supine and the other with the patient in the upright or decubitus position. An upright chest film may be added to exclude the presence of free subdiaphragmatic gas, an uncommon finding in bowel obstruction. Plain radiographs demonstrate the presence of SBO in 50 to 60% of cases and delineate features suggestive of obstruction in another 20 to 30%.17 The cause of obstruction is rarely demonstrated on conventional abdominal radiographs. The ability to predict the site of obstruction correctly often is limited by fluid-filled loops or abnormal positioning of small bowel. Despite these potential limitations, plain radiography is still the appropriate starting point for the diagnostic evaluation of a patient with suspected SBO. Typical plain radiographic findings with SBO are distended loops of small bowel proximal to the site of obstruction followed by normal or collapsed bowel distal to the obstruction. The supine view may show dilated loops of bowel that are sharply angulated or arranged in a series of parallel segments reminiscent of a stepladder. Upright or decubitus films may demonstrate multiple intraluminal air-fluid levels (Fig. 90-1A and B). In general, the greater the number of dilated loops of bowel, the more distal the site of obstruction. Colonic gas usually is negligible in amount unless the films are obtained early in the course of the obstruction or in the presence of a partial SBO. When the obstructed intestine contains more fluid than gas, the classic findings just described may be absent. In this setting, small pockets of gas may become trapped between the valvulae conniventes of the small bowel and may appear as an
Supine
A
B
Figure 90-1. A, Supine film showing dilated loops of small bowel in a patient with small bowel obstruction. B, Upright abdominal film revealing multiple air-fluid levels and small bowel dilation, consistent with a diagnosis of small bowel obstruction. oblique series of round radiolucencies on the upright film—the so-called string of pearls or string of beads sign, which is very suggestive of SBO. In patients with adynamic ileus, plain film findings may be similar to those in patients with intestinal obstruction. With the former entity, however, the radiologic findings tend to involve the entire gastrointestinal tract, including the colon, and air-fluid levels are not as prominent as with mechanical obstruction. The air-filled loops of bowel also are not dilated in gastroenteritis or other causes of adynamic ileus. Since the first reports describing the role of CT in bowel obstruction in the early 1990s, this modality has been increasingly used. It is considered complementary to standard radiography in the evaluation of SBO. CT has been shown to be an excellent modality for demonstrating intussusception, volvulus, and extraluminal lesions such as abscesses and tumors.19 This modality is especially helpful and should be used as an early imaging technique in the setting of known abdominal malignancy or inflammatory bowel disease or when an abdominal mass is discovered on examination. CT scans have high sensitivity, specificity, and accuracy in the diagnosis of SBO.17,18 In high-grade obstructions, in particular, these numbers are greater than 90%. CT can demonstrate both closed-loop obstruction and features suggestive of strangulation. In the vast majority of cases, CT is not required to make the diagnosis of bowel obstruction. Its main use is in better defining the site of obstruction and possible cause.17-20 Abdominal CT examination has increasingly been advocated for early diagnosis of complete obstruction and ischemia in SBO, both of which dictate early surgical intervention.21,22 One recent study demonstrates the high sensitivity for CT for ischemia and further suggests that CT findings consistent with partial SBO predict a clinical condition that will resolve without surgery in more than 90% of cases.22 Another radiologic test for small bowel obstruction that may influence clinical outcome is the use of water-soluble contrast. Several studies have shown that water-soluble contrast radiography successfully predicts nonoperative resolution of adhesive SBO.23,24 Although some early studies suggested that water-soluble contrast actually hastened resolution of the
obstruction, a recent meta-analysis does not support a therapeutic benefit for this technique.24
Differential Considerations The diagnosis of SBO should be considered in a patient with abdominal pain and vomiting, especially if the history includes previous abdominal surgery. It often is difficult to distinguish among mechanical obstruction, adynamic ileus, and pseudoobstruction on clinical grounds alone. Other clinical diagnoses that should be considered range from benign to life-threatening in nature and include pregnancy, gastroenteritis, cholelithiasis and cholecystitis, pancreatitis, peptic ulcer disease, appendicitis, ischemic bowel syndromes, and myocardial infarction. Each of these clinical entities has typical signs, symptoms, and diagnostic findings that help to differentiate it from SBO, but doing so may be challenging when patients present during the early stages of their particular disorder.
Management and Disposition The initial management of SBO has remained largely unchanged for several decades and consists of aggressive fluid resuscitation, bowel decompression, and timely surgical consultation. All patients with SBO should be admitted to the hospital. Intravenous hydration should be initiated with an isotonic crystalloid solution administered through a large-bore catheter. Enteral decompression by nasogastric suction should take place early in the clinical course to remove accumulated gas and fluid proximal to the obstruction. No convincing argument has been made for the use of a long intestinal tube (e.g., Cantor, Miller-Abbott) over a nasogastric tube.1,16 Placement of a nasogastric tube is a noxious procedure for the patient. Application of topical anesthetic to the nasopharynx and posterior pharynx may improve tolerability of the procedure. There is no convincing research to recommend routine use of antibiotics in the conservatively managed patient. However, the demonstration of bacterial proliferation during intestinal
Chapter 90 / Disorders of the Small Intestine
Upright
1187
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1188
stasis and obstruction suggests that broad-spectrum antibiotics are appropriate when surgery is planned and when the clinical picture suggests vascular compromise or intestinal perforation.1 Antibiotic use should provide coverage for gramnegative and anaerobic organisms that colonize the intestinal contents (e.g., second-generation cephalosporins). “Never let the sun set or rise on a bowel obstruction” is an oft-quoted surgical adage that has stood the test of time because of the preoperative difficulty in distinguishing strangulation from simple bowel obstruction. Proponents of early surgical intervention cite the similar clinical and radiographic presentations of simple and strangulated obstructions and argue that any delay in surgical therapy may increase morbidity. Although there is no debate about the need for surgery in patients with signs of peritoneal irritation or fever, most surgeons advocate a trial of conservative therapy in the absence of findings suggestive of strangulation. Up to 75% of patients with partial SBO and 35 to 50% of those with complete obstruction experience resolution of symptoms when treated with intravenous fluid and bowel decompression alone.16,23 Patients with early postoperative bowel obstruction, adhesive obstruction, and obstruction secondary to Crohn’s disease are more likely to respond to nonoperative management. Surgical intervention should be planned if substantial relief is not obtained within a short time after nasogastric tube placement or if symptoms persist after 48 hours of conservative treatment. A practical point is that obstruction occurring in a patient without a previous history of laparotomy is not likely to be caused by peritoneal adhesions. Such de novo obstruction and the underlying cause usually are not resolved without surgery. Neither advanced age nor known abdominal malignancy is a contraindication to operative intervention. Patients with abdominal cancer who do not have widespread intraabdominal metastases should be managed as for any other patients with SBO. They should receive a trial of bowel decompression followed by surgery if resolution of the symptoms is not evident. From 20 to 40% of patients with abdominal neoplasms and SBO have a benign cause of the obstruction.1 In addition, the incidence of strangulation with obstruction related to malignancy is low. Therefore, a trial of tube decompression is a safe and often successful option. A therapeutic approach that is gaining support for the management of SBO is laparoscopy.25-28 Bowel obstruction traditionally has been a relative contraindication to laparoscopy because of the potential for bowel distention and the risk of enteric injury. However, as experience with this surgical approach has increased, surgeons have begun to demonstrate that this is a safe and effective method of diagnosing and treating acute bowel obstruction in selected patients, particularly in those with obstruction caused by adhesions.
KEY CONCEPTS More than 50% of SBO cases are caused by postoperative adhesions. Two other leading causes are various hernias (15%) and neoplasms (15%). ■ The diagnosis of SBO usually is made on the basis of plain radiographic findings, with the upright abdominal radiograph revealing air-fluid levels and dilated loops of small bowel in a majority of cases. ■ Initial management of SBO should include volume assessment and resuscitation, plain radiography, bowel decompression, and surgical consultation. A significant percentage of SBO cases caused by adhesions may be managed without surgery.
■ ACUTE MESENTERIC ISCHEMIA Perspective Acute mesenteric ischemia primarily affects patients older than 50 years, especially those with significant cardiovascular or systemic disease. The acute form of this disease results in the rapid development of intestinal injury and is much more common than the chronic form of mesenteric ischemia. Chronic mesenteric ischemia results when splanchnic blood flow is inadequate to support fully the functional demands of the intestines yet not so compromised as to threaten bowel viability. The incidence of acute mesenteric ischemia is difficult to determine but has been reported as 0.1% of hospital admissions, and increasing occurrence in today’s aging population has been noted in several studies.29 Acute vascular compromise of the intestine remains an important and life-threatening cause of acute abdominal pain in patients presenting to the emergency department. Acute mesenteric ischemia was described in the 18th and 19th centuries in sporadic reports; however, an understanding of the underlying pathophysiology awaited the classical experimental work of Litten in 1875, when he described the results of ligation of mesenteric vessels in animals. In 1895, Elliot described the first patient to recover after resection of an infarcted intestine that probably was the result of mesenteric venous thrombosis. He created two stomas and reanastomosed the bowel segments 2 weeks later. Thus, the diagnosis of gangrenous bowel by laparotomy and its treatment by resection with anastomosis, a sequence of events that is still common today, were first performed more than 100 years ago. The concept of mesenteric revascularization as the treatment for acute mesenteric ischemia was introduced in the 1950s. Even with the advent of this significant surgical advance, however, morbidity and mortality rates remained high. Today, most physicians use an aggressive approach to the patient with suspected acute mesenteric ischemia, as first proposed in the 1970s. The single most important step in this approach is early diagnosis. The etiology of acute mesenteric ischemia actually has four distinct categories, each associated with a group of risk factors, signs and symptoms on presentation, and varying nuances in the evaluation and management of the patient. The most common cause of acute mesenteric ischemia is arterial embolus, which accounts for 40 to 50% of cases. Arterial thrombosis accounts for 25% of acute presentations, nonocclusive mesenteric ischemia for 20%, and the remaining 5 to 10% of cases are caused by mesenteric venous thrombosis.29 The importance of early diagnosis and aggressive intervention in patients with suspected acute mesenteric ischemia is underscored by mortality rates that climb to 70% once intestinal infarction has occurred. The mortality associated with this lethal disease has changed little in the last several decades and probably will not decrease until a reliable screening test for the disease is found.30
■
Principles of Disease The severity of intestinal injury is inversely proportional to mesenteric blood flow and is a function of the state of the systemic circulation, the number and caliber of involved vessels, the status of the collateral circulation in the region, and the duration of the ischemia.29 The extent of the damage ranges from reversible impairments in mucosal function to transmural infarction and necrosis of part or all of the bowel served by the compromised vasculature. Blood supply to the abdominal organs derives from three major vessels: the celiac artery trunk, the superior mesenteric
Mesenteric Arterial Embolism The median age of patients presenting with mesenteric arterial emboli is 70 years. Approximately two thirds of these patients are women. The vast majority of arterial emboli resulting in acute mesenteric ischemia involve the SMA. The source of SMA emboli usually is the heart, with fragmentation of either left atrial or ventricular thrombi during or after a dysrhythmia or valvular lesions. Emboli consisting of tumor and cholesterol also have been described. Emboli typically lodge 4 to 7 cm from the vessel’s origin at a point of anatomic narrowing such as the takeoff of a major arterial branch. More than 50% of SMA emboli are found immediately distal to the origin of the middle colic artery. Risk factors for mesenteric arterial emboli include coronary artery disease, valvular heart disease, and arrhythmias—in particular, atrial fibrillation. A recent review of autopsy findings of thromboembolic occlusion of the SMA found a common association with acute myocardial infarction, cardiac thrombi, and synchronous emboli to other organs.31 Risk factors are listed in Box 90-3. Recognition of these contributing factors is important to improve early diagnosis of this disease.
1189
Factors Associated with Mesenteric
BOX 90-3 Arterial Embolism
Coronary artery disease Post–myocardial infarction mural thrombi Congestive heart failure Valvular heart disease Rheumatic mitral valve disease Nonbacterial endocarditis Arrhythmias Chronic atrial fibrillation Aortic aneurysms or dissections Coronary angiography
Factors Associated with Nonocclusive
BOX 90-4 Mesenteric Ischemia
Cardiovascular disease leading to low-flow states Congestive heart failure Arrhythmias Cardiogenic shock Post–cardiopulmonary bypass Preceding hypotensive episode Septic shock Drug-induced splanchnic vasoconstriction Digoxin Vasopressors Ergot alkaloid poisoning Cocaine abuse
by atherosclerosis. This is the most common site for thrombus formation in the mesenteric circulation. In contrast with arterial embolism, the more proximal nature of thrombus formation results in greater visceral damage and a less favorable prognosis. SMA thrombosis usually occurs in patients with chronic, severe visceral atherosclerosis. As many as 50% of these patients give a history of “abdominal angina,” or abdominal pain after meals.32 Thus, risk factors associated with mesenteric arterial thrombosis include older age, diffuse atherosclerosis (coronary, cerebral, or peripheral vascular disease), and hypertension.
Nonocclusive Mesenteric Ischemia Nonocclusive mesenteric ischemia has been defined only in the last 50 years, as intraoperative and postmortem examinations have revealed ischemic bowel without obvious vascular obstruction.29,33 The pathogenesis of nonocclusive mesenteric ischemia is multifactorial, but a common pathway involves mesenteric vasoconstriction, usually in response to low-flow states associated with decreased cardiac output or the administration of vasoactive medications. Factors contributing to the development of nonocclusive mesenteric ischemia include hypotension, which may be associated with many systemic diseases, and splanchnic vasoconstriction from various medications (Box 90-4). Nonocclusive mesenteric ischemia is seen in patients of all ages and often develops during hospitalization for other medical or surgical problems.
Mesenteric Arterial Thrombosis
Mesenteric Venous Thrombosis
The SMA, which originates from the ventral surface of the abdominal aorta at a 45-degree angle, commonly is narrowed
Mesenteric venous thrombosis is the least common cause of acute mesenteric ischemia. It occurs in a younger population
Chapter 90 / Disorders of the Small Intestine
artery (SMA), and the inferior mesenteric artery (IMA). Abdominal organs receive their blood supply on the basis of embryologic development. The esophagus, stomach, proximal duodenum, liver, gallbladder, pancreas, and spleen are supplied by the celiac trunk. The SMA supplies the distal duodenum, jejunum, ileum, and colon to the splenic flexure. The descending and sigmoid colon and rectum are supplied by the IMA. There is an abundant system of collateral vessels with significant territorial overlap of blood flow that can be clinically significant. Approximately 25% of the cardiac output is delivered to the small and large intestines, with two thirds going to the SMA distribution and one third to the IMA.29 Eighty percent of this flow is destined for perfusion of the mucosa because of its high metabolic requirement. Accordingly, the visceral mucosa is very sensitive to decreased perfusion. With the onset of hypoperfusion, a redistribution of intramural blood flow favoring the superficial layers of the mucosa takes place. Below a critical level of blood flow, however, the intestinal villi become ischemic, and significant alterations in mucosal function occur. The countercurrent exchange mechanism in the small intestinal villi initiates and perpetuates ischemic damage to the tissue. As epithelial cells become necrotic, endothelial factors are released that lead to the attraction and activation of neutrophils and macrophages into the ischemic tissue. These cells release protease enzymes, tissue necrosis factor, platelet activating factor, arachidonic acid by-products, and toxic oxygen radicals that cause further endothelial damage, increased vascular permeability, vasoconstriction, inflammation, and further necrosis. This initial ischemic insult is compounded if and when perfusion is reestablished, because restoration of blood flow permits further recruitment of inflammatory cells to the area. Ischemic disruption of the normally impenetrable mucosal barrier allows the release of bacteria, toxins, and vasoactive mediators into the systemic circulation. Cardiac depression, multisystem organ failure, septic shock, and death may occur even before the development of intestinal ischemia. Necrotic changes can be seen as early as 10 to 12 hours after the onset of symptoms but may develop in a more delayed fashion.
1190
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Factors Associated with Mesenteric
BOX 90-5 Venous Thrombosis
Hypercoagulable states Polycythemia vera Sickle cell disease Antithrombin III deficiency Protein C or S deficiency Malignancy Myeloproliferative disorders Estrogen therapy/oral contraceptive pills Pregnancy Inflammatory conditions Pancreatitis Diverticulitis Appendicitis Cholangitis Trauma Operative venous injury Postsplenectomy Blunt or abdominal trauma Miscellaneous Congestive heart failure Renal failure Decompression sickness Portal hypertension
of patients, and the mortality rate is lower than that associated with other causes, ranging from 20 to 50%.34,35 Mesenteric venous thrombosis rarely can be a primary diagnosis, but it often occurs in association with an underlying medical condition, including hypercoagulable states, inflammatory processes within the abdomen, local trauma, and conditions associated with relative venous stasis, including oral contraceptive use and inflammatory bowel disease36,37 (Box 90-5). Historically, up to 60% of patients with mesenteric venous thrombosis have a history of peripheral deep vein thrombosis.
Clinical Features History. The clinical findings associated with acute mesenteric ischemia, regardless of the cause of the vascular compromise, are fairly nonspecific. Nonetheless, the presentation is sufficiently characteristic that acute mesenteric ischemia should be considered in the population of patients at risk: In patients older than 50 years of age with any of the previously discussed risk factors for mesenteric ischemia, the sudden onset of abdominal pain severe enough to warrant medical attention and more than 2 hours in duration is highly suggestive of acute mesenteric ischemia.30 On initial presentation, the patient with acute mesenteric ischemia typically complains of severe, poorly localized, colicky abdominal pain. Associated symptoms and signs may include nausea, vomiting, and frequent bowel movements as the bowel attempts to empty itself. The most consistent finding is pain that is out of proportion to the physical findings. This characteristic finding is noted because only visceral structures are initially ischemic, and the parietal peritoneum is spared. Mesenteric ischemia also can be more subacute in its presentation, with the insidious onset of less severe and vague abdominal pain, abdominal distention, and occult gastrointestinal bleeding. Physical Examination. In the early phases of mesenteric ischemia, physical examination findings may be nondiagnostic. As the disease process continues, abdominal distention develops
and palpation reveals diffuse abdominal tenderness without guarding. Transmural intestinal injury leads to peritoneal signs (involuntary guarding and rebound tenderness). Late in the ischemic episode, the abdomen is grossly distended, with absence of bowel sounds and exquisite tenderness to palpation. Heme-positive stool is noted in 25% of patients and often is a relatively late finding.29 Historical details and physical findings referable to other parts of the body may suggest the etiology of the acute impairment in mesenteric blood flow. Complications. Delays in diagnosis permit progression of the disease process and the development of transmural intestinal ischemia, with its correspondingly high associated morbidity and mortality rates. Even with early diagnosis and aggressive management, however, a complicated course is to be expected. Secondary reperfusion injury is common, and bowel initially believed to be viable at the time of operation may become ischemic and infarct in the postoperative period. Other postoperative complications include wound infections, intraabdominal abscesses, sepsis, and pneumonia. This population of patients also is at risk for many life-threatening complications (including myocardial infarction, pulmonary embolism, and renal failure) because of significant concurrent illness.
Diagnostic Strategies Routine laboratory and standard radiographic evaluations usually are not helpful in the diagnosis of mesenteric ischemia. An increase in the peripheral white blood cell count is a common but nonspecific finding, and although a normal count makes the diagnosis of acute intestinal ischemia less likely, it does not exclude the diagnosis. Hemoconcentration, metabolic acidosis with base deficit, and hyperamylasemia are present in more than one half of the cases of acute mesenteric ischemia but likewise are nonspecific findings. A significant emphasis has been placed on the role of serum lactate level determination in detection of ischemia, but a consensus regarding the utility of this test has yet to be reached. The sensitivity of this serum marker is high, approaching 100% when bowel infarction is present, but it has a disappointing specificity. A retrospective review of data on preoperative assessment for mesenteric ischemia noted an elevated serum lactate level at the time of diagnosis to be most useful as a significant predictor of mortality and suggested that the presence of unexplained acidosis in patients at risk should prompt a search for reversible causes of mesenteric ischemia. Levels of the seromuscular enzyme creatine kinase (CK) rise 3 to 4 hours after vascular occlusion, but the CK assay has limited specificity and sensitivity. Other seromuscular enzymes (lactate dehydrogenase, aspartate transaminase) and mucosal enzymes (alkaline phosphatase) are even less sensitive and specific than CK. Although it also has been suggested that a negative result on D-dimer assay may be useful for the exclusion of suspected mesenteric artery occlusion,38 further evaluation of this laboratory screen needs to be accomplished before widespread application. The first radiologic examination that should be done in a patient with suspected mesenteric ischemia is a plain abdominal radiograph series (supine and upright) to rule out bowel obstruction or presence of free air. Plain radiographs most often are normal in appearance in the presence of acute mesenteric vascular compromise. By the time any changes characteristic of acute intestinal ischemia are apparent, transmural damage has already taken place. Subtle signs of acute mesenteric ischemia on plain abdominal radiographs include adynamic ileus, distended air-filled loops of bowel, and bowel wall thickening from submucosal edema or hemorrhage (Fig. 90-2).39 In advanced stages of ischemia, pneumatosis of the
1191
Chapter 90 / Disorders of the Small Intestine
A
B
Figure 90-2. A, CT of abdomen showing pneumatosis intestinalis (note dependent air within wall of small bowel). B, CT of abdomen with air within portal venous system.
bowel wall may be detected as intraluminal gas dissects into the submucosa.40 Another late, and often preterminal, sign of necrotic bowel is the presence of gas within the portal venous system. Further radiographic examinations should be selected judiciously. Intraluminal barium contrast evaluations are contraindicated because residual contrast material can limit visualization of the mesenteric vasculature during diagnostic angiography. Duplex ultrasonography may be of some benefit in visualizing blood flow in the SMA and celiac axis. Unfortunately, many patients suspected of having mesenteric ischemia often have dilated, air-filled loops of bowel, which makes ultrasonography extremely difficult. Because of the availability, improved quality, and speed of CT, this radiologic test often is used for assessing undiagnosed abdominal pain in high-risk patients. In the setting of intestinal ischemia, a CT scan is capable of demonstrating edema of the bowel wall and mesentery, abnormal gas patterns, presence of intramural gas, ascites, and, occasionally, direct evidence of mesenteric venous thrombosis. The diagnosis of acute mesenteric ischemia often is made by CT imaging, reflecting the prevalence of this test in the evaluation of abdominal pain. Finally, an important consideration is that, as with plain radiography, a significant percentage of patients may have normal or nonspecific CT findings, so the diagnosis of mesenteric ischemia cannot be ruled out on the basis of normal findings on this study. Angiography remains the “gold standard” modality for the diagnosis of mesenteric ischemia and is unique among imaging techniques in that it may assist with both diagnosis and therapy. Preoperative angiography is useful in the diagnosis of either mesenteric artery embolus or thrombus. It allows identification of the site and type of occlusion as well as evaluation of the splanchnic circulation, thus facilitating plans for prompt revascularization. Thrombosis of the SMA typically reveals an occlusion just distal to the origin of the vessel. In addition,
angiography provides a definitive diagnosis of nonocclusive mesenteric ischemia. Arteriographic signs of nonocclusive disease include diffuse or focal tapering of mesenteric arterial branches, alternating segments of narrowing and dilatation of intestinal branches (“sausage sign”), poor intramural vessel filling, and mesenteric arcade vasospasm. Broad criteria for selection of patients must be used if early diagnosis and effective intervention are to be possible. Therefore, a significant number of “negative” angiograms should be accepted. Helical CT scanning and CT angiography have progressed significantly in recent years and now offer reasonable alternatives to catheter angiography. Helical CT angiography should be considered the primary diagnostic modality for patients in whom clinical suspicion for mesenteric ischemia is high.41,42 Conventional catheter angiography is reserved for equivocal cases of noninvasive imaging and also can be used for transcatheter therapeutic techniques.
Differential Considerations Mesenteric ischemia occurs most often in patients older than 50 years, but the diagnosis should be considered in all patients, regardless of age, who experience sudden onset of severe abdominal pain. The severe and colicky nature of the pain also may suggest cholecystitis, peptic ulcer disease, perforation of bowel, nephrolithiasis, diverticulitis, or bowel obstruction. The significant pain, often out of proportion to the physical findings, also may suggest the possibility of pancreatitis and abdominal aortic aneurysm rupture. The urgency of efficiently identifying acute mesenteric ischemia would recommend that this diagnosis be considered in a large population of patients, particularly in patients at risk because of underlying illness or chronic medical therapy that produces vasoconstriction. In practice, the diagnosis of acute mesenteric ischemia often is made after other disorders have been excluded.
1192
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Management and Disposition Early diagnosis achieved by aggressive use of angiography remains the key to a successful outcome. Therapeutic intervention should take place as soon as the diagnosis of acute mesenteric ischemia is made if tissue salvage is to be maximized and mortality minimized. Initial resuscitative efforts should include correction of hypovolemia and hypotension as well as any accompanying metabolic abnormalities. In the population of patients at risk for mesenteric ischemia, successful resuscitation may require invasive hemodynamic monitoring. Control of arrhythmias, congestive heart failure, and other factors contributing to relative hypoperfusion of the bowel is a priority. Medications with vasoconstrictive properties should be discontinued immediately. If vasopressors must be used to support blood pressure, the lowest possible dose should be infused, and α-agonists should be avoided, with inotropes being the preferred agents. Enteral decompression by nasogastric tube placement is recommended. Broad-spectrum antibiotic therapy that provides coverage for bowel flora should be initiated early, particularly when surgery is anticipated. When the patient’s condition is stabilized, routine laboratory and plain radiographic examinations can be performed to exclude other, more common causes of abdominal pain. If an expeditious evaluation does not reveal an alternative diagnosis, angiography should be performed. Even when the decision to operate has been made on clinical grounds, preoperative radiologic studies, including some form of angiography, may improve management of the patient at laparotomy. In addition, when the diagnosis of acute mesenteric arterial compromise is confirmed, an infusion of papaverine through the angiography catheter directly into the SMA reduces or eliminates mesenteric vasoconstriction.43 Papaverine is a potent inhibitor of phosphodiesterase, the enzyme necessary for degradation of cyclic adenosine monophosphate (cAMP). Increased cAMP levels cause vascular smooth muscle relaxation and relief of vasoconstriction. Because papaverine is 90% metabolized by the liver on its first pass, few if any systemic effects are noted during its use. Dosing consists of delivery of a 60-mg bolus into the SMA, followed by continuous infusion of 30 to 60 mg/hour at a concentration of 1 mg/mL. Use of this vasodilator in both nonocclusive and occlusive forms of mesenteric ischemia has improved survival. The surgical management of acute mesenteric ischemia is both challenging and controversial. Treatment principles range from pharmacologic manipulation without operation to revascularization procedures to bowel resection. The underlying cause of intestinal hypoperfusion often is not amenable to surgical correction, as with mesenteric venous occlusion and nonocclusive disease, and the role of operative intervention may be limited to resection of already infarcted bowel. If a revascularization procedure is to be undertaken in the presence of arterial occlusive disease, it is completed before any evaluation of bowel viability is performed. The reasoning behind this therapeutic sequence is that bowel that initially appears irreversibly damaged may exhibit significant recovery on restoration of blood flow. Obviously necrotic bowel is resected, but in the presence of extensive ischemic damage, the surgeon may choose to leave bowel of questionable viability in place and to reevaluate its viability during a subsequent operation.44 This “second-look” operation, typically performed
12 to 24 hours after the initial procedure, may permit a more limited resection. Percutaneous transluminal angioplasty has been described for both acute and chronic mesenteric ischemia from thrombosis of the SMA. In the acute setting, it appears to be associated with an increased risk of recurrence and potential for extensive bowel loss. With chronic intestinal ischemia, particularly in elderly patients who are poor surgical candidates, mesenteric angioplasty is a good option, with complete symptomatic improvement and continued relief of symptoms during follow-up observed in a majority of patients. Intra-arterial infusion into the SMA of thrombolytic agents has been used successfully for mesenteric ischemia after acute embolism,45 but only on a limited basis. The patients in the reported series were selected with emboli confirmed by angiography, with no peritoneal signs and no abdominal radiographic abnormalities including ileus. Close monitoring and frequent clinical reassessment, as well as serial angiograms, are necessary after the thrombolytic infusion. The main drawbacks to the use of thrombolytic agents are the difficulty in assessing bowel viability without laparotomy, the possible time delay of 12 to 18 hours before clot resolution, and the potential for clot fragmentation with involvement of more distal branches less amenable to surgical revascularization. In patients surviving the initial episode of acute mesenteric ischemia, recurrent thrombosis is a potential problem requiring long-term anticoagulation. Warfarin (Coumadin) is started after mesenteric arterial embolism and mesenteric venous thrombosis. Antiplatelet therapy is begun after mesenteric arterial thrombosis and nonocclusive mesenteric ischemia. The 2-year mortality rate after mesenteric ischemia is as high as 70%. This grave prognosis, however, is related mainly to cardiovascular comorbidity, rather than to recurrent mesenteric ischemic events.
KEY CONCEPTS ■
Four separate acute mesenteric ischemia syndromes are recognized. A majority of cases are caused by embolic occlusion of the SMA. The remainder are due to SMA thrombosis, venous thrombosis, and nonocclusive arterial ischemia. Each of the syndromes has a specific set of risk factors or associated medical conditions that are helpful in differentiating one from another. ■ The diagnosis of acute mesenteric ischemia may be suggested by pain out of proportion to examination findings, heme-positive stool, elevated serum lactate levels, and classic findings on plain film or CT scan, but none of these provide enough sensitivity to ensure recognition of this entity before bowel infarction occurs. ■ An aggressive approach to diagnosis and management, including early use of angiography, has provided some improvement in the prognosis for acute mesenteric ischemia, although the mortality rate for this disease is still greater than 50%. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 91
Acute Appendicitis
Jeannette M. Wolfe and Philip L. Henneman
■ PERSPECTIVE Appendicitis is a common condition requiring emergency surgery. Approximately 7% of people will experience acute appendicitis during their lifetime. Most cases occur in adolescents and young adults, with a slightly higher incidence in men than in women.1,2 Historically, the earliest evidence of appendicitis is suggested by the presence of right lower quadrant adhesions in an Egyptian mummy from the Byzantine era. In 1492, Leonardo da Vinci drew pictures of the colon and the appendix, referring to the latter structure as an orecchio, which literally means “ear.” Claudius Amyand removed the first appendix incidentally in 1735 during the repair of a scrotal hernia in an 11-year-old boy. The appendix had perforated, and a cutaneous fecal-draining fistula had developed.3 The half-hour operation was done without anesthesia, and the boy fully recovered. In the early 1800s during the Lewis and Clark expedition, the only death during the trip was that of Charles Floyd, who is thought to have died from a ruptured appendix.4 In 1880 in Europe, Lawson Tait performed the first successful planned appendectomy by removing a gangrenous appendix from a 17-year-old girl. Six years later, Reginald Fitz, a pathologist, coined the term appendicitis when he presented his classic paper at the first meeting of the Association of American Physicians. Fitz correctly described many of the pathophysiologic changes associated with appendicitis and advocated early surgery. Three years later, Charles McBurney described a point “determined by the pressure of one finger” between “one and a half and two inches from the anterior spinous process” that, when palpated, was associated with maximal discomfort in patients with acute appendicitis (McBurney’s point). The general acceptance that appendicitis was a surgical disease did not come until several decades later. Early surgical intervention became popular in the early 1900s around the time that King Edward VII was found to have a perforated appendix and was operated on days before his coronation.5
■ PRINCIPLES OF DISEASE: PATHOPHYSIOLOGY The appendix is a hollow, muscular, closed-ended tube arising from the posterior medial surface of the cecum, approximately 3 cm below the ileocecal valve. Its average length is approximately 10 cm, and its normal capacity is 0.1 to 0.3 mL. The role of the appendix in human physiology is unclear, but some
recent studies on biofilms suggest that the appendix may act as a repository for commensal bacteria that inoculate the large bowel and protect it against pathogens.6 Innervation of the appendix is derived from sympathetic and vagus nerves from the superior mesenteric plexus. Afferent fibers that conduct visceral pain from the appendix accompany the sympathetic nerves and enter the spinal cord at the level of the tenth thoracic segment. This allows referral of pain to the umbilical area. In a majority of affected patients, appendicitis is due to an acute obstruction of the appendiceal lumen. The obstruction often is from an appendicolith but also can be caused by a calculus, tumor, parasite, or enlarged lymph node. Of historical note, one of the more common causes of acute appendicitis from foreign objects in the early 19th century was ingestion of lead shells buried in quail flesh.7 More recently, a lumen obstruction from a swallowed tongue stud has been reported.8 After acute obstruction, intraluminal pressures rise and mucosal secretions are unable to drain. The resulting distention stimulates visceral afferent pathways and is perceived as a dull, poorly localized pain. Abdominal cramping may occur as a result of hyperperistalsis. Next, ulceration and ischemia develop as the intraluminal pressure exceeds the venous pressure and bacteria and polymorphonuclear cells begin to invade the appendiceal wall. The appendix may appear grossly normal at this time, with evidence of pathology apparent only by microscopic examination. With time, the appendix becomes swollen, and factors elaborated in the pathologic process begin to irritate surrounding structures, including the peritoneal wall. The pain now becomes more localized to the right lower quadrant. If swelling does not abate, hypoxia leads to gangrene (presence of necrosis) and, ultimately, perforation through the appendiceal serosal layer. This can lead to abscess formation or diffuse peritonitis. The time required for the appendix to perforate is highly variable (and controversial—some experts believe that unless a virulent organism or genetic predisposition exists, many cases will spontaneously resolve), but perforation usually occurs within 24 to 36 hours. Elderly patients may be more prone to earlier perforation because of anatomic changes in the appendix associated with aging, such as a narrowed appendiceal lumen, thinner mucosal lining, decreased lymphoid tissue, and atherosclerosis.9 In approximately one third of cases, no direct cause of obstruction is noted. In these cases, it is surmised that inflammation is caused by viral, bacterial, or parasitic infection with subsequent mucosal ulceration or lymphoid hyperplasia.2 1193
1194
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
■ CLINICAL FEATURES History Appendicitis is classically described as starting with the vague onset of dull periumbilical pain and the development of anorexia, nausea, and vomiting. The pain then migrates to the right lower quadrant, and a low-grade temperature may develop. In most instances, the patient has not previously experienced pain similar to this episode.2 Unfortunately, the presentation may be highly variable. If the appendix is retrocecal or retroiliac, the pain may be blunted by the presence of overlying bowel. If the appendix is elongated, the pain may be referred to the flank, pelvis, or right upper quadrant. Other, less typical symptoms seen with appendicitis are increased urinary frequency and the desire to defecate.2
Physical Examination The most common finding on physical examination is localized abdominal tenderness, usually in the right lower quadrant. The pain may be noted over McBurney’s point. The base of the appendix is within 5 cm of this point in only 35% of patients, however, so the pain of appendicitis also can be localized in other areas of the abdomen.10 Other physical findings include abdominal guarding and rigidity to palpation. Guarding usually is voluntary, and the patient often can be persuaded to relax. Rigidity is involuntary and implies more significant underlying pathology.2 Both of these findings reflect the tensing of the abdominal wall musculature to protect the underlying bowel. Rovsing sign is present when tenderness is referred to the right lower quadrant with palpation of the left lower quadrant. The psoas sign is the increase in pain when the psoas muscle is stretched as the patient is asked to extend the hip. The obturator sign is the elicitation of pain as the hip is flexed and externally rotated. Rebound tenderness to palpation is a late finding in patients with appendicitis and usually is noted only after the appendix is significantly inflamed or ruptured. Rebound tenderness is detected by gradually pressing over the area of tenderness for 5 to 10 seconds and then quickly withdrawing the hand to just above the skin level. A positive response is when the patient reports increased pain as the hand is removed. Patients with rebound tenderness are very uncomfortable with this maneuver, and it should not be repeated unnecessarily. The sensitivity and specificity of rebound tenderness in diagnosing acute appendicitis range from 63 to 82% and from 69 to 90%, respectively.2,11 The presence of peritoneal irritation also can be elicited by other maneuvers that cause the visceral and parietal peritonei to rub against each other, such as having the patient cough while observing for evidence of acute discomfort. A positive cough sign has been found to be 80 to 95% sensitive in diagnosing acute peritonitis.12,13 Isolated rectal tenderness rarely may be the only site of localized pain in patients with a low-lying or retrocecal appendix. In general, however, rectal tenderness has a very limited diagnostic value, especially if concurrent right lower quadrant pain and tenderness are present.14,15 Although a single rectal exam may provide other important information, such as the discovery of a rectal mass or occult blood, multiple exams are not justified. Although any of the foregoing signs may be present in patients with acute appendicitis, certain findings have a high positive likelihood ratio for identifying patients with appendicitis. These include right lower quadrant pain, rigidity, and migration of initial periumbilical pain to the right lower quad-
rant.2 Conversely, the presence of pain for more than 48 hours, a history of previous episodes of similar pain, the lack of migration and of right lower quadrant pain, and the lack of worsening pain with movement or cough make appendicitis less likely.2,14 A similar review of data on children with appendicitis concluded that fever and rebound tenderness were the most common associated findings.16 Vital signs often are normal, particularly early in the clinical course. A low-grade fever is present in approximately 15% of patients; this increases to approximately 40% if perforation has occurred.17
Special Considerations Children Acute appendicitis in young children often is diagnosed after perforation has occurred. Many common childhood illnesses are associated with nausea, anorexia, and vomiting, and young children may have difficulty communicating their discomfort.18 Anatomically, children have a thinner appendiceal wall and a less developed omentum, which may predispose them to perforation and diffuse peritonitis.
Women The diagnosis of acute appendicitis in women of childbearing age can be particularly challenging. Before the advent of imaging, as many as 45% of women with symptoms suggestive of appendicitis had a normal appendix at surgery, and as many as one third of women with true appendicitis were initially misdiagnosed. Gynecologic disease can easily masquerade as appendicitis because of the close proximity of the appendix to the right ovary and fallopian tube and the uterus.19,20 Findings that may be more suggestive of abdominal pain of gynecologic origin are listed in Table 91-1. Of note, although cervical motion tenderness is more common in patients with pelvic inflammatory disease, it may be present in up to one fourth of women with appendicitis.21 Because accurately diagnosing appendicitis in women is difficult, the use of ancillary imaging should be strongly considered.
Pregnant Women In pregnant women, the overall risk for the development of appendicitis is similar to that in the general population.22 Appendicitis appears to occur slightly more often in the second trimester than in the other two, for unknown reasons. The
Table 91-1 Abdominal Pain in Women LIKELY ETIOLOGY
CLINICAL FINDINGS
More suggestive of appendicitis
Migration of pain and tenderness localized to the right lower quadrant Anorexia Normal or minimally abnormal findings on pelvic examination (e.g., isolated right adnexal tenderness) Symptoms of several days’ duration History of pelvic inflammatory disease Hunger Diffuse lower abdominal pain Bilateral adnexal tenderness Cervical motion tenderness Vaginal discharge
More suggestive of pelvic inflammatory disease
Elderly Patients Elderly patients are three times more likely than those in the general population to have a perforated appendix at surgery. Multiple causes for this predisposition are recognized, including anatomic age-related changes in the appendix, delay in seeking medical care, atypical presentations, and minimally elevated values on laboratory tests.9
Complications The complication rate after the removal of a normal or acutely inflamed appendix is approximately 3%, but this increases by three- to four-fold if perforation occurs.17 The most common complication is infection. Localized wound infection occurs in approximately 2 to 7%, and deep intra-abdominal abscess occurs in 0.8 to 2%, with the higher percentages representing cases in which perforation had occurred.25 Other complications include a prolonged ileus, small bowel obstruction, pneumonia, and urinary retention and infection. In young women, perforation may cause obstruction of the fallopian tubes with subsequent fertility problems, although recent studies suggest that this sequela is not as prevalent as was once believed.26,27 Pregnant patients with appendicitis have an increased risk of premature labor (15 to 45%) and fetal death. The mortality rate for uncomplicated appendicitis is less than 0.1% but increases to 3 to 4% with perforation in patients with comorbid illness or advanced age. Although reported perforation rates vary significantly, the overall average is approximately 20 to 30%. This increases greatly at the extremes of age. Elderly persons have perforation rates as high as 60%, and children younger than 3 years of age can have perforation rates as high as 80 to 90%.9,28 The identification of factors that may increase a patient’s risk for perforation is evolving. The traditional belief has been that the natural course of appendicitis is inflammation that, if surgery is delayed, ultimately progresses to necrosis and perforation. Many experts now feel that in most patients, the natural course of appendicitis is spontaneous resolution without perforation.28 This view is supported by general autopsy reports from the presurgical era showing that in up to one third of cases not previously diagnosed, evidence of periappendiceal scarring was present, and by studies that report successful resolution of early appendicitis with nonoperative management.29 It is hypothesized that a subset of the population is genetically predisposed to perforation by a characteristic early aggressive and exaggerated inflammatory response.30 The evidence for such a predisposition is the relatively consistent population perforation rates, even with the advent of increased imaging and earlier diagnosis. In most patients so affected,
1195
perforation of the appendix occurs before medical evaluation, and hospital delays in operative management rarely appear to increase perforation rates.31
■ DIAGNOSTIC STRATEGIES Laboratory Testing Leukocyte Count Approximately 80 to 90% of patients with acute appendicitis will have an elevated white blood cell (WBC) count above 10,000/mm3. Unfortunately, the leukocyte count is nonspecific and often is elevated with other causes of abdominal pain.
C-reactive Protein A meta-analysis suggests that the overall sensitivity of Creactive protein is approximately 62% and its specificity is 66%. Accordingly, its usefulness as a diagnostic tool in patients with appendicitis is limited.32
Urinalysis A urinalysis is helpful in differentiating urinary tract disease from acute appendicitis and is suggested in all patients. Mild sterile pyuria may be seen if the appendix is irritating the ureter. Significant pyuria (i.e., presence of more than 20 WBCs per high-power field) is highly suggestive of urinary tract pathology.
Pregnancy Test A pregancy test should be performed in all women of childbearing age, because a positive result broadly expands the scope of the differential diagnosis for right lower quadrant pain.
Diagnostic Scores Some retrospective studies have shown a benefit for use of diagnostic scoring systems for appendicitis that assign a numerical value to different aspects of the history and physical exam. Of note, however, these scoring systems yield inconsistent results when prospectively evaluated and appear to be particularly inaccurate when applied to female patients.2,33
Imaging Studies Plain Radiography Plain radiography is not useful in diagnosing appendicitis owing to its very low sensitivity and specificity and is not recommended in the evaluation of appendicitis unless there is a significant concern of bowel obstruction, free air, or pneumonia.34
Barium Enema Barium enema has a sensitivity of approximately 80 to 90% for detecting appendicitis, and the diagnosis is essentially ruled out if the entire appendix is filled with contrast.35 Unfortunately, a normal appendiceal lumen often is not visualized with this technique. Barium enema is most helpful when other colon pathologic processes are high on the list of considerations in the differential diagnosis.
Chapter 91 / Acute Appendicitis
diagnosis of appendicitis during pregnancy can be particularly difficult. Nausea and vomiting occur frequently in a normal pregnancy, and the accuracy of the physical exam may be compromised because enlargement of the uterus may alter the location of the appendix. Nevertheless, a study of pregnant patients with appendicitis found that even when appendicitis occurred late in pregnancy, most women still had right lower quadrant pain.23 Lab values are not helpful, because leukocytosis is common during pregnancy. Although maternal death from appendicitis is extremely rare, spontaneous fetal abortion occurs in approximately 5 to 15% cases of simple appendicitis and in up to 37% of complicated cases.24 Because of this high morbidity, extra caution should be taken in pregnant women with abdominal pain.
1196
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Nuclear Medicine Scan Periappendiceal fat streaking
The use of nuclear imaging with tagged WBCs has been well studied as a diagnostic tool for acute appendicitis.36,37 The sensitivity of nuclear scans depends on the radiolabel used, and reported values ranges between 88% and 98%. The overall usefulness of these scans is limited because of poor specificity—any process causing inflammatory changes in the lower abdomen can lead to a false-positive result.
Ultrasonography Graded-compression ultrasound imaging has been prospectively shown to improve the clinical accuracy of the diagnosis of acute appendicitis.38 The reported sensitivity and specificity of ultrasound imaging for acute appendicitis in most studies are 75 to 90% and 85% to 95, respectively.39 Some recent advances in ultrasound techniques have been described. One group of investigators reported an astonishing 98% rate for visualization (compared with the often cited 2 to 45%) of the appendix by adding simple maneuvers that repositioned this structure.40 Similarly, small studies using contrast-enhanced Doppler or harmonic waves (which allow for better resolution) show promise of higher sensitivities while sparing the patient radiation exposure.41-43 On ultrasound examination, a noncompressible appendix with a diameter greater than 6 to 7 mm is considered diagnostic for appendicitis. Ultrasonography is inexpensive, requires no patient exposure to radiation or dye, adds no extra time for contrast filling, and has had long-standing success in diagnosing pelvic pathology in women. It also allows correlation of the patient’s pain with direct visualization of underlying abdominal contents. The major disadvantage of ultrasonography is that visualization of the appendix—normal or abnormal—is operator-dependent, and the technique can be especially difficult if the patient is obese, has strictures from previous surgeries, or has a retrocecal appendix. A diagnostic ultrasound study also becomes more difficult to achieve after the appendix has perforated, and patients with significant right lower quadrant pain may not tolerate the graded compression. Ultrasound findings suggestive of appendicitis have a very high positive predictive value (approximately 90%). Absence of abnormality on the ultrasound study, however, is not helpful unless the appendix is clearly visualized or alternative pathology is identified. Accordingly, with negative ultrasound findings, either in-hospital observation or a CT scan is indicated if the patient’s symptoms have not abated.
Computed Tomography Abdominal pelvic CT scanning has been prospectively studied and shown to improve the clinical accuracy of the diagnosis of appendicitis.44,45 CT findings individually suggestive of appendicitis include an enlarged appendix (diameter greater than 6 mm), pericecal inflammation, the presence of an appendicolith (Fig. 91-1), and a periappendiceal phlegmon or abscess (Fig. 91-2). The sensitivity (87 to 100%) and specificity (89 to 98%) of CT scan vary by study and technique and by how investigators categorize inconclusive scans in their statistical analyses. Of the different CT techniques, thin-cut, helical CT with rectal contrast appears to be the most sensitive, with sensitivity rates as high as 98%.46 Use of rectal contrast confers a number of advantages: it may be better tolerated in nauseated patients; there is no delay in scanning due to contrast transit time; and more consistent cecal opacification is obtained, which aids in scan interpretation.46-48 Although rectal contrast
Appendiceal wall thickening
Figure 91-1. Oral contrast computed tomography scan showing evidence of acute appendicitis with periappendiceal fat streaking. (Courtesy of Jefferson Radiology.)
Fat stranding
Discrete abscess from perforated appendix
Figure 91-2. Oral contrast computed tomography scan showing discrete
abscess from appendiceal perforation, with periappendiceal fat streaking. (Courtesy of Jefferson Radiology.)
may not be conceptually or practically acceptable to some patients, at least in one study patient discomfort and satisfaction rates were similar regardless of type of contrast received.47 Oral contrast abdominal pelvic CT is an alternative choice but requires a 60- to 90-minute delay after contrast administration for distal small bowel opacification and may be poorly tolerated in the patient with nausea or an ileus. A novel way to decrease transit time is to add polyethylene glycol to the oral contrast; in one study, this allowed good cecal opacification by 1 hour after ingestion.49 Although certain institutions have published high sensitivity and specificity rates with noncontrast CT scans, these
Magnetic Resonance Imaging MRI is emerging as a useful tool in the evaluation of suspected appendicitis, with reported sensitivity similar to that for CT scan. Access to MRI is currently limited in most EDs, however, so its use often is confined to the pregnant patient with indeterminate ultrasound findings as an alternative to CT.58-61
Laparoscopy Laparoscopy can be performed for diagnosis or definitive treatment. Historically, its greatest advantage was in the clarifi cation of the diagnosis of appendicitis versus gynecologic disease in young female patients. Because an enteric contrast CT scan can now usually visualize the appendix, the use of diagnostic laparoscopy with its anesthetic risks has decreased significantly.
In-Hospital Observation Despite the current increased tendency to pursue diagnostic imaging in patients with right lower quadrant abdominal pain, the recent literature suggests that most cases of appendicitis
1197
Table 91-2 Differential Diagnosis for Appendicitis ALL PATIENTS
WOMEN
CHILDREN
Nonspecific abdominal pain Gastroenteritis Ascending diverticulitis Gallbladder disease Inflammatory bowel disease Renal colic
Ovarian cyst Ovarian torsion Pelvic inflammatory disease Ectopic pregnancy
Henoch-Schönlein purpura Testicular torsion Epiploic appendagitis Mesenteric adenitis/ileocolitis Meckel’s diverticulum
can still be accurately diagnosed by performing serial physical examinations.62,63 A review of studies using active inpatient observation in patients with an equivocal diagnosis of appendicitis found a negative appendectomy rate of approximately 6% without an increase in perforation rates.62
■ DIFFERENTIAL DIAGNOSIS The differential diagnosis for appendicitis includes essentially any pathologic process that can cause abdominal pain. The more common diseases that can mimic appendicitis are listed in Table 91-2. Of note, the diagnosis of gastroenteritis should be made with caution and only in patients with vomiting and diarrhea.
■ MANAGEMENT A strategy to manage patients with possible appendicitis is depicted in Figure 91-3. Patients should be kept on NPO (nil per os) status and undergo a complete physical examination, including a rectal and pelvic examination. Dehydrated patients should receive intravenous crystalloid fluids, parenteral antiemetics should be given to patients with nausea or vomiting, and patients with more than mild discomfort should be offered pain medication. Multiple studies have shown that giving opiate pain medicine to adults or children with signs of appendicitis does not mask important physical exam findings or impair surgical decision making.64,65 Furthermore, a recent meta-analysis concluded that any changes in the physical examination due to administration of medications did not appear significant enough to alter management in patients with acute abdominal pain.66 Depending on local surgical preference or institutional policy, surgical consultation before medication administration may be indicated, but only if it can be done in a timely fashion. Controversy exists regarding when and how to best utilize advanced imaging techniques. Some experts have shown that CT scanning significantly decreases the rate of negative laparotomies, even in patients in whom clinical suspicion for appendicitis is high.67-69 Others feel that diagnostic imaging is overused and has not improved patient care.70,71 In any case, diagnostic imaging is likely to be most helpful if done in a select group of patients. After initial physical exam and laboratory tests, patients should be stratified by risk. Excessive imaging in patients who are at low risk for appendicitis will increase the frequency of false-positive results on relevant studies, because the prevalence of disease in this population is low. Patients may be considered to be at low risk if they have minimal physical findings and strong evidence for an alternative diagnosis, or if they have had multiple previous episodes of similar pain. The other distinct group of patients
Chapter 91 / Acute Appendicitis
findings have not been reproducible in other settings.50,51 Two recent studies found a 20 to 25% rate for inconclusive scan interpretation when no contrast was used.52,53 Periappendiceal fat streaking can easily be missed on a noncontrast CT scan obtained in thin patients or children. Intravenous contrast may help diagnose very early appendicitis by enhancing appendiceal wall inflammation, but in most instances it adds little additional information and increases the risk of adverse dye reaction. CT scanning has some advantages over ultrasound in the diagnostic evaluation of appendicitis. With enteric contrast CT, the appendix usually can be visualized, the technique is standardized, and alternative pathology often is identified. An added benefit is that the identification of CT signs of appendicitis is relatively straightforward and can be easily learned. This is an important consideration because the initial interpretation of the CT usually dictates patient disposition, and in an academic teaching center, this scan often may be read by a junior radiology resident after hours.54 The biggest disadvantages of CT scanning are the radiation exposure and the expense. A routine full abdominal and pelvic CT study results in about 10 millisieverts of radiation exposure, which theoretically is carcinogenic. From analysis of data for World War II atomic bomb survivors, it has been suggested that the radiation from a single abdominal CT scan could cause a fatal cancer in 1 of every 500 children scanned.55 Radiation exposure can be decreased by doing a focused 15-cm scan limited to the cecum and pelvis and by size-adjusting the scanner. Finally, CT is not 100% accurate. Overall, approximately 5 to 10% of CT scans are considered inconclusive for appendicitis—for example, the appendix diameter may be enlarged but without wall thickening, or the appendix may not be visualized although surrounding fat streaking is evident. Care should be taken not to label these studies as “negative,” because approximately 30% of patients with such equivocal findings will have histologic confirmation of appendicitis.56 In general, even patients with a true negative result on the CT scan should be explicitly told to return for reevaluation if their symptoms worsen or do not resolve in the next 24 to 36 hours. Such follow-up is particularly important in patients evaluated within the first few hours of symptoms, because early appendicitis may be missed on the initial CT scan.57
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1198 Physical examination including rectal and pelvic; NPO, IV fluids, urine dipstick, ± hCG1, CBC Low risk
Moderate risk
High risk
± Treat alternate diagnosis. Home with education. Recheck 12–24 hours, earlier if worse.
Consult surgery2
Consult surgery
Morphine 0.05–1 mg/kg q5 min
Pregnant women Women with abnormal pelvic exam Slender children and women
US
Negative US for appendicitis4 Negative for appy6
In hospital observation
Morphine 0.05–0.1 mg/kg q5 min
Women with unremarkable pelvic exam Overweight children Men Positive US for appendicitis Oral or enteric contrast CT5
Positive for appy
IV antibiotics
OR 1. Consult Ob/Gyn immediately if unstable pregnant patient. 2. Timing and type of surgical consultation will be dependent on local preferences. 3. Pain medication is considered safe and humane in patients with acute abdominal pain and should be offered to the uncomfortable patient as quickly as possible; if the surgeon requests to examine the patient unmedicated, the physical examination should be done in a timely fashion. 4. Failure to visualize a normal appendix on US does not rule out appendicitis. Unless other convincing pathology is found or the patient’s exam has improved, enteric CT or in-hospital observation should be considered. 5. Rectal contrast abdominal pelvic CT is the preferred study. 6. Hospitalization should be considered in patients who have significant pain or have required substantial amounts of opiates.
Figure 91-3. Suggested emergency department management of patients with possible appendicitis. appy, appendectomy; CBC, complete blood count; CT, computed tomography; hCG, human chorionic gonadotropin; NPO, nil per os (“nothing by mouth”); OR, operating room; US, ultrasonography.
who may not benefit from imaging are patients who present within the first few hours of onset of symptoms. Imaging studies in these patients are more likely to yield false-negative results, and the “negative” scan may provide false reassurance. In both subsets, the best course of action probably is patient education about worsening signs of appendicitis, along with arrangements for immediate reevaluation if their symptoms
progress, or for reexamination in 12 to 24 hours if they have not improved. Ideally, this conversation should be documented in the medical record. Patients with “equivocal” signs of appendicitis should be considered for diagnostic testing or active observation. Surgical input by phone or formal consultation may be appropriate before imaging, depending on the preference of the consulting surgeon. “Equivocal” patients include most women (especially those of childbearing age). Historically, it has been very difficult to make an accurate diagnosis of appendicitis in women on clinical grounds alone, and negative laparotomy rates of 40 to 45% were common. Fortunately, imaging can dramatically decrease these rates and should be strongly considered in all women. Ultrasonography may be the most appropriate initial study if the history is strongly suggestive of gynecologic disease or if the pelvic examination reveals any abnormality. It also should be considered in very thin patients who have not previously had abdominal pelvic surgery; otherwise, an enteric contrast CT study is recommended. If the patient is pregnant, ultrasound imaging is the first-line test to exclude other obstetric diagnoses. If the ultrasound findings are equivocal, MRI should be considered if it is readily available. Although CT is less desirable, owing to its considerable associated radiation exposure, after surgical and obstetric consultation, its use may still be justified when this drawback is balanced against the risks associated with potentially unnecessary anesthesia and surgery. Other subgroups of “equivocal” patients are more difficult to define because of variable inclusion criteria among published studies. It appears reasonable that men and children with a clinical presentation suggestive of appendicitis but who lack at least one of its classic features (such as pain is new onset, migrates from periumbilical to right lower quadrant, and is associated with anorexia and tenderness) and have no clear alternative diagnosis be considered “equivocal.” Ultrasonography often is recommended as the preferred initial imaging study in children with equivocal findings of appendicitis because it does not expose them to radiation. This is a significant consideration because children are particularly vulnerable to the risks of radiation owing to their increased cell division and longer life expectancy.55 A limited-cut helical CT study with enteric contrast (administered either rectally or orally) is an option in overweight children. If ultrasound imaging is chosen and the scan is interpreted as negative, a follow-up CT scan or admission for serial observation is indicated, unless the patient’s clinical status has improved.72 Male patients with an equivocal presentation for appendicitis should be considered candidates for an enteric contrast CT examination (although US is a reasonable alternative for thin patients who can tolerate the US compression). Men with classic signs of appendicitis, however, are likely to have the disease more than 90% of the time, and imaging adds little to their workup.73,74 Although the current practice in some institutions has evolved such that CT is almost mandated before surgical involvement, a recent well-done study suggests otherwise: with a decision for surgical involvement before CT, 65% of male patients went to the operating room without a CT scan, with a 4% negative appendectomy rate.75 Emergency physicians should advocate institutional polices that prevent unnecessary radiation exposure in this group and in children with classic presentations; in these subsets, appendicitis is a clinical diagnosis. Once the decision to operate has been made, prophylactic antibiotics should be given to provide coverage for gramnegative and anaerobic organisms; this strategy has been proved to decrease both superficial and deep postoperative
1199
Some institutions have begun to develop extensive operative and postoperative guidelines for the care of patients with appendicitis.79,80 The use of these guidelines has decreased postoperative complications and costs and appears to be most helpful in the subgroup of patients with perforation. For patients with evidence of obvious perforation and abscess formation, many surgeons prefer to drain the abscess nonoperatively and treat the condition with intravenous antibiotics and then perform an interval appendectomy 6 weeks later.81 Recently, it has even been suggested that the appendix may not have to be removed after successful abscess resolution.82
■ DISPOSITION If clinical suspicion for appendicitis is low, the patient may be sent home after extensive education has been provided and arrangements have been made for appropriate follow-up care. Discharged patients should be encouraged to start on a liquid diet and advance to solids if their symptoms abate. Patients with abdominal pain of unclear etiology who require significant doses of opiates to control their pain should be considered for hospital admission.83 In addition, if follow-up cannot be arranged, if patient or family reliability is in question, or if a significant language or transportation barrier exists, hospitalization for observation should be considered.
KEY CONCEPTS ■ ■
Classic appendicitis is a clinical diagnosis. Patients with a low risk for appendicitis may be sent home with close follow-up and education about progressive symptoms. ■ Patients with equivocal findings should undergo advanced diagnostic imaging or in-hospital serial examinations. ■ Men and children with classic signs and symptoms of appendicitis should undergo prompt surgical evaluation, because imaging may be unnecessary. ■ Ultrasound examination is an appropriate initial test in pregnant patients, in women with a clinical presentation
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
suggestive of pelvic pathology, and in thin children and women. ■ Helical CT with enteric contrast is considered the initial imaging study of choice in all male patients with equivocal signs of appendicitis and in female patients without findings suspicious for gynecologic disease (although US is a reasonable alternative for thin patients who can tolerate the US compression). ■ Pain medication should be offered to all patients with suspected appendicitis. ■ Antibiotics should be given preoperatively.
Chapter 91 / Acute Appendicitis
wound infections. Intravenous second-generation cephalosporins such as cefotetan or cefoxitin provide good coverage. In cases with a high likelihood of perforation, the traditional treatment has been a broad-spectrum triple antibiotic regimen; however, recent studies suggest that “single coverage” with a second-generation cephalosporin, mirapenam, or combination drug like pipercillin and tazobactam provides similar coverage, with easier administration.76 The timing of surgical intervention has recently been challenged, with several studies suggesting that complication rates are not adversely affected if surgery is delayed until daylight hours.77 If this practice is to be adopted, it needs to be balanced by considerations including the condition of the individual patient and the potential for disrupting the morning operating room schedule and increasing the overall length of hospital stay. The appendix can be surgically removed using either the traditional open technique or a laparoscopic approach. A Cochrane review of 45 randomized studies favored laparoscopic removal.78 The investigators concluded that laparoscopic appendix removal resulted in less frequent wound infections, less postoperative pain on day 1, shorter length of hospital stay, shorter time to return to normal activity, and decreased overall costs. Laparoscopy may be most helpful in female patients, in whom it allows inspection for pelvic pathology that may masquerade as acute appendicitis.
Chapter 92
Gastroenteritis
Sandy A. Craig and David K. Zich
■ BACKGROUND Gastroenteritis is defined as an inflammation of the stomach or intestines resulting in some combination of nausea, vomiting, and diarrhea. Although most affected patients will prove to have an infectious cause for their symptoms, the clinician also must consider noninfectious causes of vomiting and diarrhea, which are considered in detail in Chapters 20 and 23, respectively. Infectious gastroenteritis is associated with dozens of causative pathogens. Most of these organisms cause benign and self-limited illness, but in a few cases, morbidity may be significant, necessitating prompt diagnosis and treatment. Diagnostic testing should be targeted to those patients who have clinically significant illness or suspected reportable disease. It is neither feasible nor necessary to identify the specific pathogen during the emergency department (ED) evaluation. Instead, the clinician should classify the gastroenteritis as acute or chronic and invasive or noninvasive, based on readily available clinical information. This classification narrows the differential diagnosis considerably and is a valuable guide to further testing. Acute gastroenteritis lasts less than 2 weeks. The vast majority of ED patients will have acute gastroenteritis, and viral and bacterial pathogens should be the primary considerations in the differential diagnosis. Chronic gastroenteritis persists longer than 2 weeks. In such cases, the possibility of a parasitic pathogen should be added to the differential diagnosis. Invasive gastroenteritis is a clinical diagnosis made in the presence of signs or symptoms of intestinal mucosal invasion, such as fever, gross or occult blood in the stool, tenesmus, or abdominal pain. When invasive disease is suspected, further diagnostic testing is indicated. Patients with noninvasive gastroenteritis do not exhibit fever, produce bloody stools, or experience significant abdominal pain. Noninvasive gastroenteritis suggests the presence of a viral pathogen or toxin-producing bacteria. This illness typically is brief and self-limited, and diagnostic testing is not likely to be of benefit.
■ ACUTE INVASIVE BACTERIAL ENTERITIS Potential pathogens in acute invasive bacterial enteritis are summarized in Table 92-1.
Campylobacter Enteritis Epidemiology. Campylobacter is the most common documented cause of bacterial enteritis in developed countries. In Canada, 1200
where the disease is reportable, campylobacteriosis was found in 30.2 persons per 100,000 in 2004, compared with 16 cases of salmonellosis per 100,000 population.1 Most cases occur in children younger than 5 years of age or in male patients between the ages of 20 and 29 years, but people of all ages are affected. The disease is more common during the summer months. Opportunistic infections with Campylobacter species often are found in homosexual men or patients with acquired immunodeficiency syndrome (AIDS), even in the absence of symptoms of diarrhea or proctitis. Campylobacter species are a common cause of “backpacker’s diarrhea,” along with Giardia, both of which are frequently acquired by drinking from wilderness water sources. Pathophysiology. Campylobacter organisms are small, spiralshaped gram-negative bacteria. The most common species isolated are Campylobacter jejuni (94%), Campylobacter coli (1%), and Campylobacter fetus.1 Campylobacter cinaedi and Campylobacter fennelliae are isolated almost exclusively from homosexual men. Campylobacter species produce disease primarily by direct invasion of the colonic epithelium and may induce inflammatory changes that are endoscopically indistinguishable from inflammatory bowel disease. Most infections are acquired by handling raw poultry or eating raw or undercooked poultry meat. The primary reservoirs for Campylobacter organisms are chickens, with more than one half of U.S. flocks silently infected and well over one half of the chicken in U.S. supermarkets contaminated.2 Other causes include consumption of tainted beef, pork, raw milk, or untreated water, or contact with infected pets and farm animals.1 Clinical Presentation. The incubation period for Campylobacter enteritis is approximately 2 to 5 days. Disease onset usually is rapid, with signs and symptoms of fever, cramping abdominal pain, and diarrhea. Constitutional symptoms of anorexia, malaise, myalgias, and headache are the rule, and some patients experience backache, arthralgias, and vomiting. The clinical picture can mimic that in acute appendicitis. Onset of diarrhea often lags 24 to 48 hours after the onset of fever and abdominal pain. Typically, the stools are loose and bile-colored but then become watery, grossly bloody, or melanotic approximately 40% of the time. Either gross or occult blood is found in the stool of 60 to 90% of patients with Campylobacter gastroenteritis. At the height of the illness, patients usually pass 8 to 10 stools or more per day.3 Most patients are well within a week or less; however, diarrhea can persist for weeks. Relapses are common although generally milder than the original episode. Fatalities are rare;
1201
Table 92-1 Epidemiologic Aspects of Invasive Bacterial Enteritis SOURCES/RISK FACTORS OR GROUPS
Campylobacter
Contaminated food/water, wilderness waters (backpacker’s diarrhea), chickens, animals
I: 2-5 days D: 5-14 days
Salmonella
Grade A shell eggs, poultry, unpasteurized milk, domestic pets
I: 8-24 hr D: 2-5 days
Shigella
Person-to-person, confined populations, poor hygiene, water-borne Food/water/milk, person-to-person, dogs, cats, pigs
I: 24-48 hr D: 4-7 days
Raw or inadequately cooked seafood, especially shrimp Raw ground beef, raw milk, meats, person-to-person, water-borne, travel Uncooked shellfish, travel
I: 8-24 hr D: 1-2 days I: 3-8 days D: 5-10 days
Yersinia Vibrio parahaemolyticus Escherichia coli O157:H7 Plesiomonas Bacillus anthracis
Infected herbivores, undercooked meat, bioterrorism
the Centers for Disease Control and Prevention (CDC) estimates 124 fatal cases in the United States each year.3 Diagnostic Strategies. Because the clinical presentation is similar to that with other invasive bacterial pathogens, the diagnosis of campylobacteriosis cannot be made on the basis of clinical presentation alone. Identification of the pathogen will require stool culture; specimens should be obtained in patients who present with acute enteritis associated with fever, abdominal pain, occult blood, or hematochezia. In borderline cases, stool methylene blue stain for fecal leukocytes is readily available and may help identify those patients who are likely to harbor an invasive pathogen. One study of outpatients with acute diarrhea found that the presence of fecal leukocytes (greater than 5 white blood cells per high-power field) was associated with a likelihood ratio (LR) of 5.0 for the presence of an invasive pathogen (95% confidence interval [CI], 2.9 to 8.6).4 Blood culture results are rarely positive, so these studies are not routinely indicated. Sigmoidoscopy reveals a nonspecific inflammatory colitis, and Campylobacter infection must be considered before a new diagnosis of inflammatory bowel disease is made. Differential Considerations. The differential diagnosis for suspected campylobacteriosis includes all infections with organisms that produce invasive diarrhea or fecal leukocytes, particularly salmonellosis, shigellosis, yersiniosis, and Escherichia coli O157:H7 infection. Management. Empirical antibiotic therapy is not recommended for otherwise healthy patients who present with acute invasive diarrhea. (Travel-related diarrhea is an exception and is discussed later on.) Initial treatment of invasive diarrhea should focus on rehydration, and the decision to initiate antibiotic therapy should be deferred pending the result of stool culture. Treatment with antibiotics is not needed for patients who demonstrate clinical improvement by the time stool culture results become available. For those who are not improving, antibiotic therapy does shorten the duration of campylobacteriosis by approximately 1.3 days.5 Erythromycin 500 mg
I: 12-48 hr D: 5-14 days
I: 1-2 days D: 5-20 days I: 1-6 days D: weeks
UNTREATED FEATURES
May cause bloody diarrhea May mimic acute appendicitis or inflammatory bowel disease; recurrence common Family and cafeteria-type food poisoning outbreaks common; increased incidence in patients with cancer or immunodeficiency Toxigenic watery diarrhea, followed by invasive picture; may produce severe dysentery Appendicitis/terminal ileitis-like syndrome; postinfection polyarthritis; long duration of fecal excretion of the organism High attack rates, summer months; self-limited Bloody diarrhea/hemorrhagic colitis; hemolytic uremic syndrome or thrombotic thrombocytopenic purpura Severe abdominal cramps and vomiting, with dehydration Oral ulcers, neck swelling, lymphadenopathy, fever, gastrointestinal hemorrhage, possible ascites
twice a day for 5 days is the recommended first-line therapeutic regimen. Azithromycin 500 mg daily for 3 days is acceptable as well. Ciprofloxacin 500 mg twice a day can be used and was previously the treatment of choice, but alarming resistance to the fluoroquinolones has emerged, thought to be due mainly to antibiotic use in the poultry industry. Roughly 10% of Campylobacter strains are now resistant in the United States, and greater than 80% resistance has been documented in Thailand. Campylobacter organisms generally are resistant to trimethoprim-sulfamethoxazole (TMP-SMX) as well. Suggested antibiotic regimens for treatment of diarrhea are listed in Table 92-2. Relapses can occur, but the likelihood is decreased with appropriate antibiotic treatment.3 Because Campylobacter infection is an invasive enteritis, antimotility agents are not recommended unless treatment with antibiotics also is given. Complications of Campylobacter infection are rare. Cholecystitis, pancreatitis, and massive gastrointestinal bleeding all have been documented, as have meningitis, endocarditis, and osteomyelitis. In addition, a definite association has been made with Guillain-Barré syndrome. Guillain-Barré syndrome associated with Campylobacter infection tends to be more severe than Guillain-Barré syndrome from other triggers and can occur even with asymptomatic infections. Luckily, the incidence is estimated at less than 1 per 1000 cases.2
Salmonellosis Epidemiology. Salmonella is the second most common cause of documented bacterial enteritis in the United States, with 36,184 cases (a rate of 12.2 cases per 100,000 population) reported in 2005. The actual number of cases is estimated to be greater than 1.4 million annually. U.S. surveillance systems found a 12% decrease in documented Salmonella infections in 2005 compared with 1995. Enteritis caused by this organism affects people of all age groups but particularly children, with those younger than 5 years of age accounting for 20% of cases.6
Chapter 92 / Gastroenteritis
INCUBATION PERIOD (I); DURATION (D)
PATHOGEN
1202
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Table 92-2 Antibiotic Therapy for Diarrhea in Immunocompetent Adults PATHOGEN
ANTIBIOTIC*†
DOSE
Campylobacter
1. Erythromycin 2. Azithromycin 1. Ciprofloxacin 2. Azithromycin 1. Ciprofloxacin 2. TMP-SMX 1. TMP-SMX 2. Ciprofloxacin Not recommended 1. Ciprofloxacin 2. Doxycycline 3. TMP-SMX None recommended Mild symptoms: ciprofloxacin 1. Ciprofloxacin 2. TMP-SMX 1. TMP-SMX 2. Ciprofloxacin
500 mg PO bid × 5 days 500 mg PO qd × 3 days 500 mg PO bid × 7 days 1 g PO × 1, then 500 mg qd × next 6 days 500 mg PO bid × 3 days 160 mg/800 mg PO bid × 3 days 160 mg/800 mg PO bid × 3 days 500 mg PO bid × 3 days
1. Metronidazole 2. Vancomycin 1. Metronidazole or vancomycin 2. Metronidazole 1. TMP-SMX 2. Ciprofloxacin 1. Ciprofloxacin 2. Doxycycline 1. Metronidazole 2. Furazolidone 1. Paromomycin 2. Iodoquinol 3. Diloxanide furoate Severe symptoms: start treatment with metronidazole 1. Paromomycin 2. Indomethacin 1. TMP-SMX 1. TMP-SMX 1. Ivermectin 2. Thiabendazole 1. Mebendazole or pyrantel pamoate 2. Albendazole
250 mg PO qid × 10–14 days 125 mg PO qid × 10–14 days 500 mg PO qid × 10–14 days 500 mg IV qid × 10–14 days 160 mg/800 mg PO bid × 3 days 500 mg PO qid × 3 days 400 mg IV q12h 100 mg IV q12h 250 mg PO tid × 5 days 100 mg PO qid × 7–10 days 500 mg PO tid × 7 days 650 mg PO tid × 20 days 500 mg PO tid × 10 days 750 mg PO tid × 10 days followed by above
Salmonella Shigella Yersinia Vibrio parahaemolyticus Vibrio cholerae Escherichia coli O157:H7 Enterotoxigenic Escherichia coli Plesiomonas hominis Clostridium difficile Diarrhea Colitis Aeromonas Bacillus anthracis Giardia lamblia Entamoeba histolytica (confirmed by PCR assay or other test to be distinct from Entamoeba dispar) Cryptosporidium Isospora belli Cyclospora cayetanensis Strongyloides stercoralis Enterobius vermicularis
1 g PO × 1 + IV fluids 300 mg PO × 1 + IV fluids 160 mg/800 mg PO bid × 3 days + IV fluids 750 mg PO × 1 dose 500 mg PO bid × 3 days 160 mg/80 mg PO bid × 3 days 160 mg/80 mg PO bid × 3 days 500 mg PO bid × 3 days
500–750 mg PO qid × 14–21 days 500 mg PO tid 160 mg/800 mg PO qid × 10 days, then bid × 3 wk 160 mg/800 mg PO bid × 7 days 200 µg/kg/PO day × 1–2 days 50 mg/kg/day in two doses × 2 days (max 3 g/day) 100 mg PO × 1 dose, repeated after 2 wk 11 mg/kg PO × 1 dose (max 1 g), repeated after 2 wk 400 mg PO × 1 dose, repeated after 2 wk
*Another quinolone agent, norfloxacin, can be substituted for ciprofloxacin in the treatment of diarrheas. The equivalent dosage is 400 mg bid. † Listed drugs are numbered in order of preference: 1 indicates the drug of first choice; 2 and 3 indicate alternative drugs. PCR, polymerase chain reaction; TMP-SMX, trimethoprim-sulfamethoxazole.
Almost all Salmonella infections are acquired from ingestion of contaminated food or drink.7 Direct person-to-person transmission can occur, but most human infections are related to the vast reservoir of salmonellae in lower-order animals. Poultry products and beef constitute the most common sources of Salmonella. Unpasteurized milk, eggs, fish and domestic pets are other sources. Outbreaks also have been associated with consumption of fruits, vegetables, baked goods, rattlesnake meat, and medicinal preparations.7 Approximately 10% of household dogs and cats excrete salmonellae, and pet reptiles, such as turtles, snakes, and iguanas, have been responsible for outbreaks of salmonellosis.8 Cooking contaminated foods decreases the possibility of infection but does not eliminate it. Salmonellae can survive cooking deep inside certain foods, where temperatures may not reach the lethal range. Very large outbreaks of Salmonella infection have been traced to contaminated, unbroken,
grade A eggs.8 Although the organism is present in the uncracked egg, thorough cooking usually eradicates or reduces the inoculum to clinically insignificant levels. Common raw egg–based sources of Salmonella infections include homemade hollandaise sauce, eggnog, Caesar salad dressing, ice cream, mayonnaise, tiramisu, cookie dough (often consumed unbaked), frosting, and French toast mix.8 Salmonella enterica subsp. enterica serovar Enteritidis (i.e., “S. Enteritidis”) is the species universally associated with eggrelated infections. Patients convalescing from Salmonellarelated enterocolitis and persons with asymptomatic infection may continue to excrete Salmonella organisms for weeks or months, thus serving as ongoing sources of infection.7 Pathophysiology. Approximately 2000 Salmonella serotypes are known to cause human illness. Based on 2005 U.S. surveillance figures, the most common isolates are the S. enterica serovars Typhimurium (19%), Enteritidis (18%), Newport
1203
cally is effective for outpatient management of Salmonella gastroenteritis: ciprofloxacin, 500 mg twice a day for 5 to 7 days; norfloxacin, 400 mg twice a day for 5 to 7 days; or azithromycin, 1 g by mouth and then 500 mg a day for the next 6 days.3 TMP-SMX also can be used if the organism is susceptible. Ciprofloxacin is effective in the treatment of chronic S. Typhi carriers. However, treatment with fluoroquinolones can actually prolong shedding of non-Typhi organisms. Patients requiring hospitalization are best treated with intravenous ceftriaxone until results of sensitivity studies become available.3 Follow-up with the patient’s primary care physician should be arranged. Food handlers and health care personnel should not be allowed to work until the carrier state has been eradicated. Repeated stool cultures and further decisions regarding job or school situations will be required. Personal hygiene should be emphasized, because untreated patients may continue to shed infective organisms in the stool for weeks or even months. As with other invasive pathogens, the use of antimotility drugs alone is contraindicated. These drugs prolong fever and diarrhea, increase the incidence of bacteremia, and promote development of a carrier state in patients with Salmonella enteritis. However, administration of loperamide is safe when it is given concomitantly with an appropriate antibiotic. Prevention of salmonellosis depends on adequate cooking and minimizing the time that foods are allowed to stand at room temperature to reduce the chance of bacterial growth to an infectious inoculum. Careful personal hygiene, including hand washing, also is important. Salmonella infection is a nationally notifiable disease (Box 92-1). Although most patients recover fully without long-term sequelae, up to 30% (primarily adults) will experience transient reactive arthritis. Reiter’s syndrome, consisting of reactive arthritis, conjunctivitis, and urethritis, is a well-known complication and occurs in approximately 2% of patients.7
Notifiable Foodborne Diseases and
BOX 92-1 Related Conditions*
Bacterial Botulism Brucellosis Cholera Escherichia coli O157:H7—now listed as “Shiga toxin– producing Escherichia coli (STEC)” Hemolytic uremic syndrome, postdiarrheal Salmonellosis Shigellosis Typhoid fever Vibriosis (with non-cholera Vibrio species) Viral Hepatitis A Parasitic Cryptosporidiosis Cyclosporiasis Trichinosis *In the United States, additional reporting requirements may be mandated by state and territorial laws and regulations. Details on specific state reporting requirements are available from the Council of State and Territorial Epidemiologists (http://www.cste.org) and the Centers for Disease Control and Prevention (http://www.cdc.gov). From Centers for Disease Control and Prevention: Nationally Notifiable Infectious Diseases—United States, 2007, revised. Available at: http://www.cdc.gov/epo/dphsi/phs/infdis2007r.htm (accessed 2007 November 8).
Chapter 92 / Gastroenteritis
(9%), and Heidelberg (5%).6 Different Salmonella serotypes show marked variations in invasive potential and are associated with particular presentations: S. enterica serovar Typhi with enteric fever (typhoid fever), S. enterica serovar Choleraesuis with septicemia, S. Typhimurium with acute gastroenteritis, and S. Enteritidis infections from grade A shell eggs.8 Relatively large numbers of salmonellae must be ingested to produce illness. However, a carrier state can be induced with ingestion of 10 to 100 times fewer bacteria needed to induce carrier state relative to number needed to induce illness. In infants and adults with certain underlying diseases, a much smaller inoculum may produce illness. Decreased gastric acidity or an alteration of intestinal flora resulting from the administration of antibiotics can impressively reduce the size of the required inoculum. Rates of invasive infection and disease severity are increased in infants, the elderly, and people with hemoglobinopathies such as sickle cell anemia, malignant neoplasms, or AIDS.7 The CDC estimates that more than 500 fatal cases occur each year.6 Clinical Presentation. Family outbreaks and sporadic cases are more common than large epidemics. Ingested salmonellae penetrate the intestinal mucosal cells and lodge in the lamina propria. After an incubation period of 8 to 48 hours, the typical patient with Salmonella gastroenteritis presents with fever, colicky abdominal pain, and loose, watery stools, occasionally containing mucus and blood. Nausea and vomiting are common but rarely are severe or protracted. Mild to moderate diffuse abdominal tenderness can be elicited in most patients, but occasionally severe tenderness and even rebound tenderness may be noted. Symptoms usually abate within 2 to 5 days, and recovery typically is uneventful. Sustained or intermittent bacteremia may occur, especially in those with sickle cell anemia, malignancy, or AIDS. Focal infections are identified in 10% of those with Salmonella bacteremia.7 Diagnostic Strategies. The diagnosis of salmonellosis cannot be made on the basis of clinical presentation alone; stool cultures are needed for confirmation. Stool methylene blue staining for fecal leukocytes may help identify those patients who are likely to harbor an invasive pathogen. Blood culture results occasionally are positive, and samples should be obtained from severely ill or immunocompromised patients. The possibility of an underlying disease or immunodeficiency state should be considered in every patient with a severe Salmonella infection. Differential Considerations. Family or communal outbreaks can suggest Staphylococcus-related food poisoning, but staphylococcal enteritis has a shorter incubation period, is not associated with fever, and produces the typical toxigenic, noninvasive, diarrheal picture. Vomiting is also much more prominent in cases of staphylococcal food poisoning than in most cases of Salmonella infection. The differential diagnosis for salmonellosis includes all organisms that produce invasive diarrhea or fecal leukocytes, particularly campylobacteriosis, shigellosis, yersiniosis, and Escherichia coli O157:H7 infection. Management. Empirical antibiotic therapy is not recommended for otherwise healthy patients who present with suspected Salmonella enteritis. Antibiotic therapy does not shorten the duration of the disease and may prolong the duration of the carrier state. Although unproven, antibiotic therapy is recommended for patients with severe colitis and for infants younger than 3 months of age, adults older than 50 years, and other groups at risk for severe disease. Persons who represent a public health risk also should be treated in an attempt to eradicate the carrier state and prevent spread of the organism.3,7 The choice of antibiotic should be based on sensitivities of the isolate. Any of the following antibiotic regimens typi-
1204
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Shigellosis Epidemiology. Shigellosis, or bacillary dysentery, is worldwide in distribution and particularly common in countries lacking effective sanitation. In the United States, a total of 10,484 cases of shigellosis were documented in 2005, representing a rate of 3.5 cases per 100,000 population.9 The CDC estimates that there are actually approximately 448,000 cases each year.10 Shigella sonnei is responsible for approximately 75% of the infections occurring in this country; Shigella flexneri causes most of the remaining cases, with Shigella boydii and Shigella dysenteriae responsible for less than 2% of cases.9 Shigella infections are common in confined populations, such as those in mental or penal institutions, in nursing homes or day care centers, or on Native American Indian reservations. Children younger than 5 years of age account for 30% of cases. An increased incidence has been documented among men who have sex with men and in the AIDS population.11 Spread is by the fecal-oral route, and humans are the only natural hosts. Shigellae can be found in large numbers around the base of toilets used by infected persons, and the organism readily passes through toilet tissue onto the fingers. Shigellae can be recovered in cultures of samples taken as long as 3 hours after contamination. In the last few years, a number of large outbreaks have been associated with recreational water venues such as swimming pools, water parks, fountains, hot tubs, and spas.10 Pathophysiology. Unlike Salmonella, which requires a very large inoculum to produce disease, as few as 50 to 100 Shigella bacilli can cause infection. No other enteric pathogen is so efficient in producing overt disease in humans. Infection generally is superficial, localized to the epithelial lining of the colonic mucosa; therefore, bowel perforation or invasion into the bloodstream is extremely rare. Bleeding occurs from superficial ulcerations of the mucosa. Clinical presentation varies among Shigella species. S. sonnei typically causes high-volume, watery diarrhea with relatively few systemic signs. Infection with S. flexneri, S. dysenteriae, or S. boydii typically causes low-volume bloody diarrhea and more severe systemic symptoms.7 Clinical Presentation. The usual incubation period is 24 to 48 hours, and the clinical manifestations vary considerably, often appearing in a bimodal fashion. Mild, watery diarrhea with few if any constitutional symptoms or asymptomatic infection occurs in a significant proportion of infected persons. It is estimated that fever occurs in 58 to 100%, abdominal pain in 75 to 100%, tenesmus in 55 to 96%, bloody stool in 46 to 73%, and nausea or vomiting in 63 to 100%.3 When true dysentery develops, it ordinarily is preceded by a recognizable period of watery diarrhea lasting a few hours to a few days. Patients with dysentery have grossly bloody diarrhea, tenesmus, and significant constitutional symptoms and signs, such as fever, nausea, vomiting, headache, and myalgias. If symptoms are severe enough, profound dehydration and even circulatory collapse can occur. Children younger than 2 years of age may have associated neurologic manifestations, most commonly seizures, although lethargy or frank coma develops in a small percentage of patients. S. dysenteriae type 1 infection, rarely diagnosed in developed countries, is associated with the hemolytic uremic syndrome. Generally, shigellosis is a self-limited disease. Patients become afebrile in 3 to 4 days, and the abdominal cramping and diarrhea resolve within 1 week. A significant number of untreated patients will continue to shed organisms in the stool for 2 or more weeks, and approximately 10% of patients will have a relapse unless the infection is treated with antibiotics.
Diagnostic Strategies. As with other invasive pathogens, most cases of shigellosis remain undiagnosed. Patients with mild, watery diarrhea and few if any constitutional symptoms can be sent home with conservative management and no investigative procedures. However, shigellosis should be considered in every patient with an acute febrile illness associated with diarrhea, especially those patients who appear ill or who have dysenteric stools. Fecal white blood cells are present, usually in large numbers, in 85 to 95% of the cases, regardless of the gross appearance of the stool.3 Thus, finding leukocytes in watery stools can help identify shigellosis even in the absence of classic dysenteric stools. Occult blood usually is present in the stools of infected patients. Blood leukocytosis is common, and a significant leftward shift in the differential count is almost always seen. Results of blood cultures for Shigella are rarely positive. Sigmoidoscopic examination reveals diffuse mucosal inflammation, often with multiple ulcerations. A definitive diagnosis of shigellosis is made with stool culture. Stool culture results are positive in more than 90% of cases when samples are obtained during the first 3 days of illness; however, results are positive in only approximately 75% if samples are obtained more than 1 week after the onset of diarrhea.12 Differential Considerations. Considerations in the differential diagnosis include salmonellosis, Campylobacter enteritis, yersiniosis, E. coli O157:H7 infection, amebic dysentery, and ulcerative colitis. Management. Treatment primarily involves the correction of fluid and electrolyte abnormalities. If S. sonnei or S. flexneri is cultured from the stool, the decision to administer antibiotics is based on the patient’s clinical condition and the feasibility of sanitary control. Asymptomatic or recovering patients do not need to be treated with antibiotics unless treatment is necessary for public health measures. Patients whose condition is not improving and those who are immunocompromised should be treated. Antibiotics shorten the clinical course and eradicate the pathogen from the stool, often within 48 hours.12 Whenever S. dysenteriae is isolated, the patient should be treated to prevent outbreaks of dysentery, even if the patient is asymptomatic when the culture result returns from the lab. In the United States, more than 80% of Shigella organisms are resistant to ampicillin, and 47% are resistant to TMPSMX.7 Significant resistance has not yet been found to the quinolone agents ciprofloxacin and norfloxacin, and one of these should be considered the drug of choice unless sensitivity studies demonstrate that the organism is sensitive to either ampicillin or TMP-SMX. Treatment is required for only 3 days in immunocompetent patients but should be extended to 7 to 10 days in the immunocompromised.3 Antimotility agents may prolong the fever, diarrhea, and excretion of Shigella in the stools and are contraindicated in patients with invasive shigellosis. However, they may be safe when used simultaneously with antibiotics. Follow-up stool cultures should be done in patients treated for S. dysenteriae infection to ensure eradication of the organism. Follow-up cultures, however, are not necessary after treatment for S. sonnei or S. flexneri infection, provided that the patient’s condition improves clinically. Shigellosis is a nationally notifiable disease.
Yersinia enterocolitica Gastroenteritis Epidemiology. Yersinia enterocolitica, a gram-negative facultatively anaerobic bacterium, is a member of the family Enterobacteriaceae. Y. enterocolitica is a relatively infrequent cause of enteritis in the United States, with FoodNet surveillance systems
1205
also should be considered in the differential diagnosis of regional enteritis, which it can closely mimic. Management. Generally, Y. enterocolitica infection is selflimited at the diarrheal stage and resolves without treatment. As with other invasive gastrointestinal pathogens, antiperistaltic drugs are not recommended unless the patient is simultaneously treated with antibiotics. Treatment with antibiotics has not been proved to be essential or efficacious in the management of uncomplicated Yersinia enterocolitis or in the pseudoappendicitis syndrome. However, because Yersinia organisms take a long time to grow on culture, in most studies the duration of illness before antibiotics were started was 1 to 2 weeks. Yersinia organisms usually are susceptible to TMP-SMX, which is the agent of choice when antibiotic therapy is indicated.3 Drug treatment does decrease the fecal shedding of the organism. Doxycycline in combination with an aminoglycoside is an alternative regimen, as is single-agent therapy with a quinolone.3 In immunocompetent adults, a 3-day course is sufficient; the course is extended to 7 to 10 days if the patient is immunocompromised. Treatment should be considered in patients who are still significantly ill at the time culture results return, particularly if the patients are immunocompromised or have a significant underlying medical illness, or in cases in which the fecal shedding could represent a public health hazard. In those patients who interact with potentially susceptible persons, appropriate steps should be taken to ensure that they do not spread their infection.
Vibrio parahaemolyticus Gastroenteritis Epidemiology. Vibrio parahaemolyticus is a halophilic (saltrequiring) gram-negative bacillus found naturally in warm marine environments such as the coastal seawaters of Japan, the United States, and other temperate-zone nations. In Japan, V. parahaemolyticus is the most common cause of bacterial enteritis, being responsible for approximately 70% of cases. The typical source is raw fish. In the United States, V. parahaemolyticus disease is much less common, with approximately 190 documented cases annually, or 0.25 per 100,000 population. The CDC estimates that approximately 3000 cases, 40 hospitalizations, and 7 deaths occur each year. U.S. cases typically are related to consumption of raw or undercooked shellfish, especially oysters (49%), although clams (38%), shrimp, lobsters, mussels, cockles, crabs, and scallops all have been implicated. Many cases occur as outbreaks on cruise ships or in persons who have patronized a common restaurant or seafood market. V. parahaemolyticus enteritis is much more common in the summer months, with 70% of cases occurring between the months of May and October, when warm seawater temperatures favor replication of the organism.16 At such times, 100% of oysters in local beds have been shown to harbor the organism. Attack rates from a common-source exposure are fairly high, but little evidence is available for human-to-human spread among family members of infected patients.17 Pathophysiology. The mechanism by which V. parahaemolyticus causes human enteritis is poorly understood, but seems to be related to production of a thermostable direct hemolysin (TDH) enterotoxin. Serotypes that produce TDH attach to the colonic epithelium and induce a secretory diarrhea as well as local hemolysis. An infectious dose of V. parahaemolyticus is thought to be 100,000 colony-forming units (CFU) or more, and the sale of oysters with 10,000 CFU or less per gram of oyster product is allowed in the United States. Still, transmission has occurred from oyster beds in which the colony count was less than 200 CFU per gram of oyster meat.17
Chapter 92 / Gastroenteritis
documenting approximately 1 case of culture-verified yersiniosis per 100,000 population.13 Y. enterocolitica infections are much more common throughout Scandinavia and Europe. In 2004, the incidence of human yersiniosis was 13.1 per 100,000 in Finland and 7.5 per 100,000 in Germany.14 Yersiniosis is more prevalent in children and more common in the winter months. Pathophysiology. After oral ingestion, the bacterium invades the intestinal epithelium and localizes to lymphoid tissue of the intestinal mucosa, particularly Peyer’s patches. It then invades the regional mesenteric lymph nodes. Invasive enteritis is the clinical presentation in approximately two thirds of patients. Pseudoappendicitis and mesenteric adenitis account for the remainder of presentations. Infection originates from contaminated food or drink. The consumption of contaminated milk or contaminated raw pork has accounted for sporadic cases and several large outbreaks. Fecal-oral transmission to humans from a variety of animals, particularly dogs, cats, and pigs, and direct person-to-person spread probably occur, but communicability appears to be low.13 Clinical Presentation. The clinical picture with Yersinia enterocolitis often resembles that with infection by other invasive intestinal organisms: fever (68%); colicky abdominal pain (65%); watery, greenish, and sometimes bloody (26%) diarrhea; and constitutional symptoms of anorexia, vomiting (39%), and malaise.3 However, in cases of Y. enterocolitica gastroenteritis, the abdominal pain and diarrhea usually persist for 10 to 14 days or longer. In a substantial number of patients with yersiniosis, particularly adolescents and young adults, an ileocecitis may develop. In these cases, lower abdominal pain with little or no diarrhea is the predominant symptom, and the clinical presentation may perfectly mimic that in acute appendicitis. Large outbreaks have been traced to contaminated milk, largely because in the relevant series, physicians noticed an extraordinary rise in the number of negative appendectomies.15 Postinfection manifestations, such as erythema nodosum or a persistent polyarthritis, occur in as many as 2 to 5% of patients, mainly adults. Other presentations include sacroiliitis, ankylosing spondylitis, Reiter’s syndrome, exudative pharyngitis, pneumonia, empyema, and lung abscess. Y. enterocolitica septicemia is rare but is known to occur, most often in patients with diabetes mellitus, severe anemia, hemochromatosis, cirrhosis, or malignancy.15 Diagnostic Strategies. Approximately 70% of patients with Y. enterocolitica infection will present with signs and symptoms of invasive enteritis. The diagnosis of yersiniosis cannot be made on the basis of clinical presentation; a positive stool culture is required. Methylene blue staining of stool for fecal leukocytes yields a positive result in approximately 48% of cases of yersiniosis.3 Most laboratories do not routinely include Y. enterocolitica culture in the standard stool culture; Yersinia culture can be done by special request if clinically indicated (e.g., by history of Yersinia exposure, prolonged invasive enteritis despite a negative result on standard stool culture, or right lower quadrant pain with a normal appendix on imaging studies). Stool cultures require special techniques and a long time for growth. Patients with Yersinia enterocolitis often continue to shed organisms in the stools well into convalescence, long after the diarrhea subsides. The mean duration of fecal shedding is approximately 6 weeks. Differential Considerations. The diagnosis should be suspected in a patient with prolonged abdominal pain and diarrhea after what appears to be a common, usually self-limited gastroenteritis syndrome, or in a patient with symptoms similar to those of appendicitis or mesenteric adenitis. Y. enterocolitica infection
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1206
Although enteritis is the most common clinical presentation, accounting for 60 to 80% of cases, V. parahaemolyticus infections also manifest as wound infections (34%) and septicemia (5%). Serious wound infections and septicemia occur primarily in persons with underlying liver disease, alcoholism, or diabetes mellitus.15 Clinical Presentation. Signs and symptoms usually appear 8 to 12 hours after the ingestion of contaminated food, but the incubation period can range from 4 to 48 hours. The predominant manifestation is acute diarrhea, but the volume of fluid lost generally is not large. Moderately severe abdominal cramps occur in 88%, nausea in 52%, vomiting in 39%, and fever in 33%. Vomiting generally is not prominent. The illness is almost invariably self-limited and seldom lasts longer than 24 to 48 hours.17 V. parahaemolyticus infection should be suspected when a common-source outbreak of acute diarrheal disease occurs in persons exposed to fresh or frozen seafood. It also should be considered when fecal white blood cells are present in cases of acute invasive diarrhea linked to food poisoning. Diagnostic Strategies. As with other types of acute invasive enteritis, the diagnosis of V. parahaemolyticus cannot be made on the basis of clinical presentation alone. The diagnosis is made by stool culture. Although blood agar and other nonselective media support the growth of this vibrio, isolation from the stool usually requires the use of a selective medium containing thiosulfate, citrate, bile salts, and sucrose (TCBS agar). This selective culture procedure is not part of the standard stool culture in most U.S. hospitals but can be obtained by special request in cases of outbreaks related to consumption of raw or undercooked shellfish, especially in coastal areas of the United States.16 Management. Because the disease is self-limited, most patients require no therapy. Although data on efficacy of antibiotic therapy are lacking, patients who still have diarrhea when culture results become available may benefit from treatment with tetracycline or fluoroquinolones or another antibiotic as guided by susceptibility testing.15 An occasional patient may require fluid replacement. Antimotility agents are not indicated. Because V. parahaemolyticus is widely present in coastal waters, the only effective preventive measures are adequate cooking, refrigeration, and hygienic practice in the preparation of seafood for human consumption.
Enterohemorrhagic (Shiga Toxin–Producing) Escherichia coli Gastroenteritis Epidemiology. Enterohemorrhagic E. coli was first recognized as a human pathogen in 1982 after two outbreaks of hemorrhagic colitis were traced to undercooked ground beef contaminated with E. coli serotype O157:H7 and distributed at a fast food restaurant chain. It is now recognized that E. coli O157:H7 is 1 of more than 30 serotypes of E. coli known to produce Shigella-like toxins and that these “Shiga toxin–producing E. coli” (STEC) as a group constitute a major cause of hemorrhagic colitis, hemolytic uremic syndrome, and thrombotic thrombocytopenic purpura in humans.15 The CDC estimates 110,000 cases and 2,100 hospitalizations per year in the United States and reports a 42% decrease in the incidence from 1996 to 2004, probably owing to improvements in food safety protocols.18 STEC infections have been documented in more than 30 countries on 6 continents. Inadequately cooked hamburger has caused many large outbreaks.15 STEC, present in the intestines of healthy cattle, contaminates the meat during slaughter, and the grinding process then transfers the organisms from the surface of the
meat to the interior. U.S. Department of Agriculture food safety regulations now require that hamburger be cooked thoroughly, to the point that the juices are no longer pink, to effectively kill E. coli organisms. Outbreaks also have occurred from consumption of venison, salami, pepperoni, cheese curds, apple cider, raw milk, and fruits and vegetables; from contamination of municipal water supplies; from animal contact in petting zoos; and from person-to-person spread in day care centers.7 Food handlers with STEC-related diarrhea have contaminated meals responsible for institutional outbreaks. Enterohemorrhagic E. coli enteritis is more common in the summer months. Pathophysiology. Escherichia coli O157:H7 is 1 of more than 30 serotypes of E. coli known to produce Shigella-like toxins called verotoxins, which are cytotoxic to the intestinal vascular endothelial cells and cause hemorrhagic colitis. STEC does not cause an invasive infection, but the clinical presentation is quite similar to that with an invasive bacterial infection. Bacteria attach to the surface epithelium of the cecum and colon and elaborate verotoxin. Clinical signs and symptoms correlate with presence of free verotoxin in the colon. Histologic changes include apoptosis in the surface epithelium, mucin depletion, and neutrophilic infiltration of the lamina propria and epithelium. The verotoxin-induced syndrome of pain and (often) bloody diarrhea closely mimics that caused by other invasive pathogens. It appears that STEC serotypes are associated with two different Shiga-type toxins. The development of hemolytic uremic syndrome is associated primarily with serotypes that produce Shiga toxin 2. The CDC estimates that greater than 90% of cases of hemolytic uremic syndrome are associated with the serotype E. coli O157:H7.19 Clinical Presentation. After an incubation period of 3 to 4 days, patients initially produce watery diarrhea that becomes bloody hours to days later. Approximately 80 to 90% of patients report bloody stools.3,19 The amount of blood varies, but stools passed may appear to consist wholly of blood, and the infection may masquerade as gastrointestinal bleeding from noninfectious causes. The bloody diarrhea typically is accompanied by severe abdominal cramps, pain, and often vomiting. Fever is a feature in fewer than one third of cases and, if present, usually is low grade.7 This helps differentiate STEC infection from that due to other invasive organisms. Fecal leukocytes are found in approximately 50% of cases, but in small numbers, in contrast with the sheets of white blood cells seen in Shigella dysentery.3 Endoscopic, histologic, and radiographic studies demonstrate only nonspecific changes consistent with an inflammatory hemorrhagic colitis and do not accurately distinguish STEC infection from other causes of colitis.20 Uncomplicated infection resolves spontaneously over 7 to 10 days. A carrier state may last another 1 to 2 weeks, but it also resolves spontaneously. Chronic diarrhea has very rarely been described.21 Complications. E. coli O157:H7 hemorrhagic colitis has been associated with two serious complications: hemolytic uremic syndrome and thrombotic thrombocytopenic purpura. These clinically similar disorders share the following features: microangiopathic hemolytic anemia, thrombocytopenia, fever, neurologic deficits, and renal dysfunction. In thrombotic thrombocytopenic purpura, neurologic findings predominate, and renal dysfunction is unusual. The opposite is seen with hemolytic uremic syndrome, which is more common in children, especially those younger than 4 years, occurring in approximately 8% of cases.7,18,19 Of these, 3 to 5% are fatal.18 Approximately 22 to 40% of elderly persons in nursing home outbreaks acquire hemolytic uremic syndrome, and 50 to 80% of these patients die. Thrombotic thrombocytopenic purpura
Aeromonas Gastroenteritis Epidemiology. Aeromonas organisms are gram-negative, facultatively anaerobic, rod-shaped bacteria of the family Vibrionaceae. Aeromonas species are ubiquitous worldwide in fresh and brackish water and also contaminate the food and water supply.15,22 Aeromonas organisms grow at a range of temperatures but are isolated with increased frequency in summer months. No figures are available on the incidence of Aeromonas gastroenteritis in the United States. An association with human enteritis is difficult to confirm because no major outbreaks have been linked to Aeromonas species and the organism often is found in the feces of asymptomatic persons. Still, several small outbreaks have been documented in travelers and by
1207
case-control studies; it is likely that certain serotypes do cause gastroenteritis in humans.23,24 Drinking untreated water, usually from private wells or springs, causes most cases of diarrhea from Aeromonas bacteria.22 Aeromonas infection has not been associated with consumption of shellfish. Aeromonas hydrophila causes wound infection; septicemia occurs in persons of advanced age or with underlying gastrointestinal disease such as colon cancer, cirrhosis, or hepatobiliary or inflammatory bowel disease or history of recent hospitalization or antibiotic treatment. In some published series, Aeromonas infection causes 2 to 10% of all cases of diarrhea in children.22,23 Pathophysiology. The exact mechanism by which Aeromonas species produce diarrhea has not yet been explained. Both enterotoxins and cytopathic toxins may be produced, and the organisms may have some invasive characteristics.
Clinical Presentation Typical signs and symptoms are watery diarrhea, abdominal cramps (50%), vomiting (25%), and fever in approximately one half of the patients.24 Children tend to have a more acute, severe illness than that typical for adults. In untreated patients, diarrhea persists for 2 to 10 weeks, with more prolonged illness in adults than in children. Generally, leukocytes and occult blood are absent from the stool; however, patients can have a severe colitis, including fever, fecal leukocytes, and bloody diarrhea, which can mimic Crohn’s disease or ulcerative colitis. Stool should be examined for the presence of Aeromonas infection before establishing a diagnosis of inflammatory bowel disease. Diagnostic Strategies. Diagnosis is made by stool culture, but culture for Aeromonas species is not part of the standard stool culture, and the clinician must request that the laboratory culture specifically for this organism. Aeromonas infection should be suspected in children or immunocompromised patients with diarrhea associated with a history of drinking from untreated water sources. Management. No controlled trials have been conducted that clearly demonstrate a benefit to antibiotic therapy. In most patients, Aeromonas enteritis is mild and self-limited, and antibiotic therapy probably is not warranted. In the case of severe disease, prolonged diarrhea, or an immunocompromised host, double-strength TMP-SMX is the drug of choice, but the quinolones also are effective.3 A 3-day course is recommended.
Plesiomonas shigelloides Gastroenteritis Epidemiology. Plesiomonas shigelloides is a gram-negative, facultatively anaerobic bacterium of the family Vibrionaceae. The organism is found in a variety of settings, including animals, soil, and in freshwater and dilute saltwater bodies including the Gulf of Mexico. Humans usually are infected by contaminated food or water or after contact with colonized animals. Worldwide, P. shigelloides enteritis is well documented in the tropical and subtropical areas of Africa, Asia, and Australia. In the United States and Europe, most cases occur after travel to tropical areas or after consumption of raw shellfish, especially oysters.15 Sporadic diarrheal illness occurs in both normal and immunocompromised hosts. Large outbreaks have occurred, usually resulting from oyster consumption.15 Pathophysiology. Evidence supporting a pathogenic role for P. shigelloides includes documented outbreaks associated with contaminated shellfish, a very low asymptomatic carrier rate, and recovery from diarrheal illness after antibiotic therapy. The mechanism of disease production remains poorly under-
Chapter 92 / Gastroenteritis
is seen in 2 to 3% of cases, most often in immunosuppressed patients. Hemolytic uremic syndrome and thrombotic thrombocytopenic purpura typically appear 5 to 20 days after the onset of infection, and the diarrhea can be totally resolved and forgotten by the time a diagnosis is established. Death from E. coli O157:H7 hemorrhagic colitis alone or from one of the complications occurs primarily among the elderly.15 Diagnostic Strategies. The CDC recommends that all patients who present with bloody diarrhea be tested for STEC infection and that E. coli O157:H7 be included in the standard stool culture in all clinical laboratories. Diagnosis requires specific stool culture techniques. In addition to the routine battery of media, specimens should be plated onto sorbitol-MacConkey (SMAC) medium. The O157:H7 strains of E. coli are sorbitolnegative at 18 to 24 hours of growth on this medium and can be rapidly identified using various serologic tests, such as latex agglutination or fluorescent antibody testing. A commercial Shiga toxin enzyme immunoassay (EIA) was introduced in 1995 in order to assist clinical laboratories in identifying nonO157 Shiga toxin–producing strains of E. coli. The CDC now recommends that all stools submitted for culture also be tested for STEC-associated verotoxin using EIA, and that non-O157 STEC be sent to the CDC for typing. As of 2007, only 9% of clinical laboratories in the United States test for STEC using EIA technology.19 Differential Considerations. Hemorrhagic E. coli infections may be misdiagnosed as ischemic colitis, inflammatory bowel disease, intussusception, or another infectious colitis. The examining physician should test for STEC when considering a diagnosis of one of these entities. Management. Antibiotic therapy does not shorten the clinical course or eradicate the organism. Moreover, treatment with antibiotics to which the organism is resistant may increase the risk for hemolytic uremic syndrome by eliminating competing bowel flora. However, the degree to which antibiotic treatment truly increases the risk for hemolytic uremic syndrome remains controversial.3,7,18 Retrospective studies in adults suggest an association between antibiotic use and the development of hemolytic uremic syndrome but may have been biased due to preferential use of antibiotics in the more serious cases. Because antibiotic treatment is of no clinical benefit and may increase the risk of hemolytic uremic syndrome, it is not recommended for patients with known infection with E. coli O157:H7. Empirical antibiotic treatment for bloody diarrhea should be approached with caution. Empirical treatment is not recommended in children because of the increased incidence of hemolytic syndrome. In adults, empirical treatment is recommended only for those with a temperature above 38.5° C, since the presence of significant fever suggests a pathogen other than E. coli O157:H7.
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1208
stood but probably involves enteroinvasion and elaboration of cytotoxins. Clinical Presentation. The incubation period is only 1 to 2 days in duration. Abdominal pain is noted in 72% of patients. Diarrhea is watery in 73% of patients and grossly bloody in 24%. Occult blood is noted in 44%, emesis in 38%, and fever in 51%.25 Symptoms resolve after 1 to 3 days in most cases but occasionally may persist for up to 8 weeks. The duration of illness usually is shorter in children. Diagnosis of P. shigelloides infection should be considered in patients with a typical invasive-appearing diarrhea, especially if the stools are bloody or when the onset of illness occurs shortly after the ingestion of raw shellfish or foreign travel, particularly to Mexico. P. shigelloides is increasingly recognized as a pathogen in immunocompromised patients. Diagnostic Strategies. Definitive diagnosis is obtained by stool culture. The laboratory must be notified when this organism is considered. Unless oxidase testing is done, Plesiomonas may be indistinguishable from Enterobacteriaceae on nonselective culture media. Patients with Plesiomonas infections should be evaluated for possible immunodeficiency. Management. Antibiotic therapy usually is not necessary for P. shigelloides enteritis because of the brief duration of illness. In patients who are immunocompromised and those with severe or prolonged symptoms, antibiotic treatment may be of benefit. P. shigelloides usually is resistant to ampicillin but susceptible to TMP-SMX, the quinolones, cephalothin, gentamicin, and chloramphenicol. The current recommended treatment regimen is TMP-SMX, 160 mg/800 mg twice daily for 3 days; ciprofloxacin, 500 mg twice daily; or norfloxacin, 400 mg twice daily for 3 days.3 Follow-up evaluation is not necessary unless the patient is immunodeficient or does not respond clinically.
Bacillus anthracis Infection Epidemiology. Although gastrointestinal anthrax is rare in developed countries, large outbreaks still occur in agricultural regions throughout the world. Anthrax also has been used as a weapon of bioterrorism in regions not normally susceptible to the disease. As early as the 1930s, groups experimented with anthrax-impregnated chocolate as a weapon against their enemies.26 Awareness of the spectrum of clinical presentations of gastrointestinal anthrax is important not only for the welfare of the patient but for prompt recognition of potential terrorist activity. Traditionally, gastrointestinal infection has been estimated at less than 1% of all human anthrax cases. It was thought that 95% of cases involved cutaneous symptoms and 5% were limited to the respiratory tract. However, because gastrointestinal anthrax can cause only mild and self-limited symptoms, many persons with the gastrointestinal form may not seek medical treatment. Studies of several large outbreaks in Uganda and Thailand suggest that the gastrointestinal form is observed in 74 to 92% of patients who contract anthrax after eating the meat of infected herbivores.26 Areas endemic for anthrax exist in all continents containing tropical and subtemperate regions. Thailand, India, Iran, Gambia, and Uganda all have reported deaths from gastrointestinal anthrax.26 Within the United States, naturally occurring anthrax exposure has been documented in several areas, including Minnesota.26 Although no predilection for time of year in endemic areas has been documented, the disease is more common in animals after substantial rainfall following a period of drought, a pattern known as “anthrax weather.”27
In the United States, gastrointestinal anthrax has not been reported, although cutaneous anthrax is reported once or twice each year.28 Populations at risk for naturally occurring anthrax are persons living in rural, agricultural areas who have ingested undercooked meat contaminated with anthrax spores. As with many other organisms, the pediatric population seems to be most at risk for serious or fatal illness. Pathophysiology. Bacillus anthracis is a nonmotile, rodlike, gram-positive anaerobic bacillus that produces central ovalshaped spores. It is introduced into the food chain most often after ingestion by herbivores such as cattle. The animal usually becomes visibly ill, and the meat often is identifiable as abnormal after slaughter, thereby preventing human exposure. Even when the meat reaches a consumer, adequate cooking usually reduces the inoculum to harmless levels. However, if the meat is undercooked, a high rate of infectivity results. Intentional placement of anthrax spores in the food or water supply could theoretically cause outbreaks of gastrointestinal anthrax, although this has not been reported. When swallowed, anthrax spores stick to the gastrointestinal epithelium, where they germinate and create multiple superficial ulcerations. Lesions have been identified from the oral cavity to the cecum.26 The vegetative cells may at times migrate into the bloodstream, where they rapidly multiply and can cause septicemia. Bacillus anthracis protects itself with an antiphagocytic capsule and produces two exotoxins, lethal and edema toxins.28 Clinical Presentation. A minority of patients who ingest anthrax spores remain asymptomatic, and in endemic areas, adults often are thought to have some natural immunity acquired through previous exposure. Of those in whom symptoms do develop, presentation can vary widely, ranging from mild watery diarrhea to fulminant upper and lower gastrointestinal bleeding, septicemia, and death. The incubation period for gastrointestinal symptoms ranges from 1 to 6 days, with larger inocula and more severe disease developing earlier. For disease confined to the oropharynx, patients usually present with complaints of sore throat, fever, dysphagia, hoarseness, and painful neck swelling. The swelling results from marked lymphadenopathy and tissue edema and can become severe enough to compromise breathing.26 Much information concerning intestinal anthrax has been derived from two large outbreaks in Uganda and Thailand, respectively.26 A majority of the patients presented with isolated diarrhea. Nausea, vomiting, and severe abdominal pain with distention also developed in a minority of the patients. Most patients were febrile, with temperatures above 39° C, and blood in both the vomitus and diarrheal stool was common. With this form of anthrax, lesions throughout the gastrointestinal tract often are surrounded by significant edema and can lead to obstruction, necrosis, and perforation. Intraabdominal lymphadenopathy and splenomegaly develop as well. The lymphatic tissue often becomes hemorrhagic, and ascites may form, with fluid shifts large enough to cause shock and even death. In cases of primary gastrointestinal anthrax, the superficial mucosa is always involved, with ulcerations visible on endoscopic examination. These findings are in contrast with those with disseminated infection from pulmonary anthrax, in which the lesions begin submucosally as a result of seeding from the bloodstream. These lesions can then secondarily ulcerate to the surface of the gastrointestinal tract epithelium. Untreated, gastrointestinal anthrax may last weeks and can be, but is not always, fatal. Diagnostic Strategies. Diagnosis of B. anthracis in cases of oropharyngeal disease is best accomplished by swabbing the oral
1209
Cephalosporins should not be used. Even with aggressive antibiotic therapy, the disease carries a 25 to 60% mortality rate, depending on host factors and the size of the inoculum.28
■ TOXIN-INDUCED BACTERIAL GASTROENTERITIS Pathogens associated with toxin-induced bacterial enteritis are summarized in Table 92-3. In general, gastroenteritis due to toxin-forming bacteria and viral agents will manifest as an acute noninvasive enteritis, with watery diarrhea, minimal fever, little or no abdominal cramping, and absence of fecal leukocytes and erythrocytes. Treatment is primarily supportive, and diagnostic testing generally is not indicated in otherwise healthy patients. A specific diagnosis may be of help in attempting to identify a common source during large outbreaks.
Staphylococcal Food Poisoning Epidemiology. Staphylococcus-related food poisoning occurs after multiplication of an enterotoxin-forming strain of Staphylococcus that is present in the food before its ingestion. Food contamination with Staphylococcus is extremely common because the organism is ubiquitous in the environment. It can be isolated from the hands of approximately 50% of persons in the general population. Most protein-rich foods support the growth of staphylococci, especially ham, eggs (even hard-boiled), custard-filled pastries, mayonnaise, and potato salad.29 Temperatures between 45° and 140° F for only a few hours will allow proliferation of the organism in contaminated food and production of sufficient enterotoxin to cause disease. Foods
Table 92-3 Epidemiologic Aspects of Toxin-Induced Bacterial Enteritis PATHOGEN/ILLNESS
Preformed Toxins Staphylococcus
Bacillus cereus Emetic toxin Diarrheal toxin
SOURCE(S)
I: 1-6 hr D: 6-10 hr
Very high attack rates, large outbreaks
Fried rice
I: 2-4 hr D: 10 hr I: 6-14 hr D: 24-36 hr I: 5-60 min D: 6 hr I: 2-6 hr D: 7-14 days
High attack rate, almost always after consumption of fried rice Food reheated or sitting out for long periods Peppery or bitter taste, histamine intoxication, high attack rates High attack rates, gastrointestinal and neurologic symptoms with paresthesias, cold allodynia, worse with alcohol
I: 6-24 hr D: 24 hr I: 24-48 hr D: 6-8 days I: 24-72 hr D: 1-7 days I: 5-14 days D: Variable I: 1-5 days D: 2-10 wk
Food reheated or sitting out for long periods Summer months, dehydration common
Scombroid fish poisoning Ciguatera fish poisoning: ciguatoxin
Large, predacious coral reef fish
Clostridium difficile
Meat, poultry, gravies, “steam table” meats Seafood, especially raw shellfish Usually unsanitary drinking water Overgrowth of normal flora
Aeromonas
Untreated drinking water
Vibrio Escherichia coli
COMMENTS/UNTREATED FEATURES
Food handler–related; potato salad, mayonnaise, confections
Vegetables; meats, especially gravies Mahi mahi, tuna, bluefish
Toxins Produced after Colonization Clostridium perfringens
INCUBATION PERIOD (I); DURATION (D)
Travelers; dehydration common in children Antibiotic-associated colitis, cytopathic toxin Common and severe in children, chronic watery diarrhea in adults, occasionally mimics inflammatory bowel disease
Chapter 92 / Gastroenteritis
lesions for culture. Blood cultures are recommended for all suspected cases of anthrax, but culture results usually are negative in patients with isolated oropharyngeal involvement. A Gram-stained smear from the lesions will demonstrate numerous polymorphonuclear leukocytes and gram-positive bacilli. Studies of serum antibody to anthrax antigens may confirm the diagnosis, but serologic testing is not widely available. Diagnosis of the gastrointestinal variant of anthrax relies more on identifying B. anthracis DNA by PCR (polymerase chain reaction) assay and cultures of blood and, if possible, ascitic fluid. Culture of the diarrheal stool yields positive results in a minority of cases. Serum antibody tests for anthrax antigens can be helpful. Differential Considerations. Oropharyngeal anthrax lesions are sometimes confused with peritonsillar abscess, although cervical swelling is unusually severe in cases of oropharyngeal anthrax. The marked swelling of the oral lesions is secondary to edema and should not yield pus if incision and drainage are performed. Gastrointestinal anthrax can cause enough upper gastrointestinal bleeding to be confused with variceal rupture. In patients with severe disease manifested as marked ascites and abdominal pain, the clinical presentation may be similar to that in patients with end-stage liver disease with peritonitis. Management. Traditionally, penicillin has been used to treat gastrointestinal anthrax. Cases of penicillin resistance have been documented, however, and a logical assumption is that a resistant strain would be chosen for use as a weapon of bioterrorism. Therefore, current recommendations from the CDC are to treat gastrointestinal cases in the same manner as for cases of respiratory anthrax, with ciprofloxacin 400 mg IV every 12 hours or doxycycline 100 mg IV every 12 hours.
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1210
containing sufficient enterotoxin to produce violent illness usually are normal in appearance, odor, and taste. Large outbreaks are common worldwide, particularly in institutions such as school or hospital cafeterias, military bases, airlines, and restaurants.29 Pathophysiology. Although the bacterium itself is killed by cooking at temperatures above 140° F, Staphylococcus enterotoxin is heat-stable. Thus, once it is present in food, reheating or even boiling will not prevent illness. The toxin has no local effect on the digestive tract. It is a potent stimulator of T lymphocytes in the host, stimulating their proliferation and the release of various cytokines. The gastrointestinal effects are believed to be mediated by the release of histamine and leukotrienes from mast cells.30 Clinical Presentation. The illness has an explosive onset, beginning 1 to 6 hours after ingestion of the contaminated food. Cramping and abdominal pain, with violent and often-repeated retching and vomiting, are the predominant symptoms. Diarrhea is a variable feature; it usually is mild, occasionally absent entirely, and infrequently profuse. Fever occasionally is present. Staphylococcal food poisoning is short-lived, usually subsiding in 6 to 8 hours and rarely lasting as long as 24 hours. Patients often are recovering by the time they seek medical attention. Attack rates are very high, often greater than 75% of the population at risk.31 The short incubation period and multiple cases among persons eating the same meal are highly suggestive of this disease. Examination of the stool is noncontributory, and no practical laboratory test is clinically available to confirm the diagnosis. The epidemiologic circumstances, however, usually provide adequate suggestive evidence. Management. Rapid, uncomplicated, spontaneous recovery is the rule. Parenteral antiemetic agents help control vomiting. Intravenous fluids should be given to patients with significant dehydration and ongoing vomiting, particularly the very young, the elderly, and debilitated patients. Antibiotics are of no value because staphylococcal food poisoning is caused by preformed enterotoxin and not by viable microorganisms. Adherence to strict personal hygiene practices by food handlers and immediate refrigeration of foods not due for immediate consumption are the most important preventive measures. Ordinary refrigerator temperatures prevent production of the enterotoxin. Food should not be allowed to stand at room temperature for long periods before being served.
Clostridium perfringens Food Poisoning Epidemiology. Clostridium perfringens food poisoning is one of the most commonly reported food-borne illnesses in the United States, with at least 10 to 20 outbreaks reported annually. Most cases occur in large groups, with dozens or even hundreds of persons affected. Illness is caused by the ingestion of meat or poultry heavily contaminated with C. perfringens type A heatresistant spores.32 The organism also is ubiquitous in the environment and in human and animal feces. Typically, poisoning results from ingesting food that is cooked more than 24 hours before consumption, allowed to cool slowly at room temperature, and then served either cool or rewarmed. During this period of incubation, spores that survived cooking germinate, and clostridia multiply to reach sufficient numbers to constitute an infectious inoculum. Pathophysiology. Ingestion of live organisms is required to produce disease, but illness is not caused by infection; rather, it is from an enterotoxin produced by sporulation of the organism in the gastrointestinal tract. The enterotoxin is responsible for all of the symptoms of C. perfringens food poisoning. Clinical Presentation. Symptoms usually appear within 6 to 12 hours but can occur up to 24 hours after ingestion of the con-
taminated food. Frequent passage of watery diarrheal stools and moderately severe abdominal cramping are the major symptoms. Fever, nausea, and vomiting are rare. The illness is self-limited and rarely lasts for more than 24 hours. C. perfringens food poisoning should be considered in a patient who experiences an acute onset of abdominal cramps and watery diarrhea shortly after eating a suspect meat or poultry dish and when others who ate the same meal are similarly ill. Leukocytes and erythrocytes are not present on stool examination. Complications. A rare type of Clostridium food poisoning termed enteritis necroticans (also known as “pig-bel”) occurs after the ingestion of foods heavily contaminated with the type C strain of C. perfringens. The illness is characterized by an acute onset of severe abdominal pain, vomiting, diarrhea, prostration, and shock and may be rapidly fatal. Postmortem examination reveals a diffuse, hemorrhagic, necrotizing enteritis of the jejunum, ileum, and colon.33 Management. Occasionally, a patient will need intravenous fluid replacement. Antibiotics are of no value because of the toxigenic nature and brief duration of the disease. Food poisoning from C. perfringens can be prevented by avoiding long periods of warming or cooling of foods that have already been cooked.
Bacillus cereus Food Poisoning Epidemiology. Bacillus cereus is an aerobic, spore-forming, grampositive rod that is a common cause of food-borne illness. The organism is one of the most frequently isolated soil bacteria. Because of its abundance and the hardiness of its spores, B. cereus contaminates nearly all agricultural products and plays a major role in the spoilage of food items, including pasteurized milk and milk products. It commonly is isolated from pasta, rice, dairy and dried milk products, spices, dried foods, meat, chicken, vegetables, seafood, fruits, and grains. Because it is ubiquitous and tolerates extremes of temperature, control of this bacterium in the food-processing environment is very difficult to achieve.34 It is estimated that B. cereus causes more than 27,000 cases of food poisoning in the United States each year. B. cereus causes two distinct clinical syndromes: an emetic form produced by a heat-stable, Staphylococcus-like enterotoxin known as cereulide and a diarrheal form resulting from a heatlabile enterotoxin known as HBL, which is similar to that of E. coli. The emetic form usually is caused by the ingestion of contaminated fried rice, although beef, poultry, vanilla sauce, pasteurized cream, milk pudding, pasta, and infant formula also have been implicated. The diarrheal syndrome usually is associated with ingestion of HBL in meats or vegetables, but reported outbreaks also have involved fish, soups, sauces, and dairy products.34 Pathophysiology. The heat-resistant spores of B. cereus survive boiling and then germinate when boiled foods such as fried rice are left unrefrigerated. The vegetative forms then multiply and produce toxin. Flash-frying or brief rewarming of the food before serving often is not sufficient to destroy the preformed, heat-stable emetic toxin. Improper holding temperatures for cooked food is the most common feature of B. cereus food-borne illness. Clinical Presentation. The emetic syndrome is clinically indistinguishable from that caused by staphylococcal enterotoxin. After an incubation period of 1 to 5 hours, profound vomiting and abdominal cramping occur in all patients. Diarrhea is present in approximately 25 to 30% of persons affected. The duration is short, usually less than 10 hours, and patients recover uneventfully.
Cholera and Gastroenteritis Due to Non-cholera Vibrios Epidemiology. In addition to V. parahaemolyticus, other halophilic marine Vibrio species have increasingly been determined to be causes of acute gastroenteritis associated with seafood. Their epidemiology is identical to that of V. parahaemolyticus: ubiquitous presence in coastal seawater, outbreaks associated with the eating of raw or inadequately cooked shellfish, and an incidence markedly limited to the warmer months of the year.15 Outbreaks of true cholera continue to occur sporadically along the Gulf Coast of the United States from inadequately cooked crabs or oysters. Other implicated foods include imported seafood, cooked rice, frozen or fresh coconut milk, and commercially prepared cut cantaloupe.15 Cholera outbreaks in South America and India have led to an increasing number of cases of cholera imported into the United States. The CDC reports 0 to 5 cases annually.36 Pathophysiology. The difference between these species and V. parahaemolyticus lies in the mechanism of pathogenesis. V. parahaemolyticus produces disease directly by an invasive intestinal infection, whereas these strains produce an enterotoxin in vivo that is responsible for the diarrhea. Therefore, symptoms resemble those of other forms of enterotoxin-induced gastroenteritis and not those caused by invasive pathogens. The enterotoxin of the non-cholera vibrios is antigenically similar to V. cholerae enterotoxin and produces a similar diarrheal illness, although it is much less severe.15 Clinical Presentation. Patients with classic epidemic cholera experience copious “rice water” diarrhea, abdominal cramps, and often nausea and vomiting within 24 to 48 hours after ingesting contaminated seafood. A low-grade fever may be present. In these severe cases (cholera gravis), rates of diarrheal fluid loss can reach 1 L per hour. Nearly one half lose enough fluids to necessitate hospitalization, and fatality rates can reach 25 to 50% in untreated populations. The median duration of illness is approximately 7 days, quite unlike the 1- to 2-day course of V. parahaemolyticus infection.15
1211
Despite the notoriety of the classic form of cholera, the CDC estimates that only 1 in 20 cases are associated with cholera gravis. A majority of affected patients experience a relatively mild diarrheal illness that may go undocumented. Cholera is extremely rare in the United States, even among returning travelers, with a documented incidence of 1 case per 1 million population annually.36 Another Vibrio species, Vibrio vulnificus, also is associated with eating raw seafood, especially raw oysters. V. vulnificus can cause self-limited gastroenteritis with onset approximately 16 hours after ingestion of contaminated food by healthy persons. In the compromised host, this organism causes serious wound infections after contact of seawater with open wounds, or a syndrome of primary septicemia characterized by hemorrhagic bullae of the skin and rapidly progressive septic shock. V. vulnificus infection is the leading cause of death in the United States associated with consumption of seafood. Septicemia carries a mortality rate of approximately 50% in patients with significant underlying disease, particularly chronic liver disease.15 All patients with chronic liver disease, alcoholism, AIDS, other immunodeficiency states, and any significant chronic disease should be advised to avoid all raw shellfish.37 Diagnostic Strategies. Because these are noninvasive vibrios, unlike V. parahaemolyticus, stained fecal smears will not show leukocytes or erythrocytes. Stool cultures will quickly identify the organisms if plated on appropriate TCBS medium.15 Management. Patients with classic cholera often will lose enough fluids to require rehydration therapy. The World Health Organization oral rehydration formula has been used successfully to treat cholera worldwide.7 The use of either oral or intravenous fluid hydration is dictated by the clinical picture. The role of antibiotics in the treatment of intestinal infections caused by noncholera vibrios is not clearly established. However, appropriate antibiotic regimens have been shown to decrease both the severity and the duration of cholera and may have the same effect on the diarrheal disease caused by these marine vibrios.15 Choices include a single oral dose of either ciprofloxacin 1 g or doxycycline 300 mg; and a 3-day regimen of double-strength TMP-SMX twice a day.3 Prevention, as with V. parahaemolyticus infection, depends on proper handling and avoidance of inadequately cooked seafood. Cholera is a nationally reportable infection. A recently developed oral cholera vaccine is licensed and available in other countries. The oral vaccine appears to provide better immunity with fewer adverse effects than the previously available parenteral vaccine. The CDC does not recommend this vaccine for travelers, and it is not available in the United States.36
Scombroid Fish Poisoning Epidemiology. Scombroid fish poisoning is a growing problem in the United States. The disease takes its name from the family Scombroidea (e.g., tuna, mackerel, skipjack, bonito, and related species) and results from the ingestion of a wide variety of dark-meat fish, including nonscombroid species such as herring, bluefish, anchovy, sardine, amberjack, black marlin, and mahi mahi. The fish species most commonly implicated are mahi mahi, tuna, and bluefish.38 Restaurants serve these fish under various names such as mackerel, swordfish, bonito, dolphin, or amberjack, or they may make their appearance in the generic “tuna salad sandwich.” Most U.S. cases occur in Hawaii and Florida, followed in frequency by California, New York, Washington, and Connecticut. However, scombroid poisoning can occur in any location where “fresh fish” are flown in on a regular basis.
Chapter 92 / Gastroenteritis
The diarrheal syndrome begins after an incubation period of 6 to 14 hours and is characterized by diarrhea in all patients and by abdominal cramps in approximately 75%. Vomiting occurs in only 20% of cases. The duration of illness ranges from 12 to 36 hours. Symptoms are essentially the same as for food poisoning produced by C. perfringens, although vomiting is less common with C. perfringens. Bacillus cereus food poisoning should be suspected whenever an illness localized predominantly to the upper gastrointestinal tract develops less than 6 hours after eating, or whenever a predominantly lower intestinal tract illness occurs 6 to 24 hours after a suspect meal, usually of meats or vegetables. Diagnostic Strategies. Because of the brief and noninvasive nature of the illness, diagnostic testing typically is not performed. In response to large outbreaks, public health authorities may elect to test common food sources. Isolation of 105 colony forming units per gram from incriminated foods confirms the diagnosis. More recently, a real-time PCR assay has been developed that can detect the presence of B. cereus emetic toxin within 2 hours. This assay is not yet commercially available.35 Management. Both syndromes generally are mild and selflimited. Antibiotics are not indicated because symptoms are mediated by enterotoxins. Parenteral antiemetic agents provide effective relief in patients presenting with violent vomiting. B. cereus food poisoning is preventable if boiled rice or cooked foods are promptly eaten or refrigerated and not left to sit at room temperature.
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1212
Pathophysiology. Implicated species naturally contain unusually high levels of histidine. Scombroid fish poisoning results from the ingestion of heat-stable toxins produced by bacterial action on the histidine present in the dark meat of the fish. The responsible bacteria are normal constituents of the surface marine flora, rather than contaminants. The histidine decarboxylase activity of these organisms produces histamine and histamine-like substances, which cause the symptoms of scombroid fish poisoning. High levels of histamine in the fish correlate directly with the occurrence of the illness. Formation of the scombrotoxins is directly related to improper preservation and refrigeration of the fish from the time they are caught until the time they are cooked. Generally, the problem is caused by improper refrigeration by the supplier, rather than being the fault of the restaurant serving the fish. Clinical Presentation. The symptoms of scombroid fish poisoning resemble those of histamine intoxication. While eating the fish, the patient may note a metallic, bitter, or peppery taste, although many affected fish do not have an abnormal odor or taste. Symptoms usually develop abruptly within 20 to 30 minutes and consist of facial flushing, diarrhea, severe and throbbing headache, palpitations, and abdominal cramps. Other manifestations may include dizziness, dry mouth, nausea and vomiting, and urticaria. The facial flushing resembles a sunburn and can extend over the entire skin surface. The conjunctivae usually are injected. The duration of the major symptom complex generally is less than 6 hours, and although weakness and fatigue persist longer, the clinical course usually is benign. The attack rate is very high; most persons sharing the same toxic fish will become ill. Management. Parenteral antihistamine therapy, such as diphenhydramine 50 mg IM or IV or cimetidine 300 mg IM or IV, usually promptly relieves all symptoms. Rarely, intravenous fluids are necessary. The disease is preventable if fish are properly handled, especially if they are refrigerated early and adequately. This is not an allergic reaction, so patients should not be told they are allergic to these fish, nor should they be prohibited from eating them again in the future.
Ciguatera Fish Poisoning Epidemiology. Ciguatera fish poisoning is a common public health problem, with appreciable economic significance. It is endemic in tropical regions but is found worldwide. The CDC estimates that 50,000 to 100,000 people per year become ill from ciguatera poisoning.38 Fish caught around Hawaii and Florida cause most of the cases, but because the responsible ocean fish are now commonly transported inland, cases can be seen virtually anywhere in the country.38 Ciguatoxin is produced by the marine dinoflagellate Gambierdiscus toxicus, which attaches itself to marine algae and is passed up the food chain. The lipid-soluble toxin accumulates in the tissues of the larger predacious coral reef fish, with the highest concentrations in the liver, intestines, head, and roe. It does not affect the fish in any way. Only humans suffer its ill effect when the toxin is ingested. More than 400 fish species that frequent coral reefs have been implicated as ciguatoxin carriers, but fewer than 50 are commercially important. Those implicated in ciguatera poisoning include amberjack, barracuda, grouper, king mackerel, parrotfish, sea bass, snapper, sturgeon, surgeonfish, and ulua.38 Pathophysiology. Ciguatera fish poisoning results from the ingestion of the ciguatoxin neurotoxin. Ciguatoxin is heat- and acid-stable, odorless, and tasteless. It is not deactivated by cooking or freezing, nor is it eliminated by drying, salting, smoking, marinating, or pickling. It is not possible to predict
whether a fish contains sufficient amounts of the toxin to produce illness.39 Ciguatoxin has both anticholinesterase and cholinergic properties, but its neurotoxicity is mediated by its effect on sodium channels. Ciguatoxins cause a hyperpolarizing shift of the voltage dependence of channel activation such that sodium channels are open at resting membrane potential. Spontaneous firing of neurons occurs as tetrodotoxin-sensitive sodium channels are activated, giving rise to the typical neurologic signs and symptoms.39 Clinical Presentation. Ciguatera fish poisoning most commonly is seen in the spring and summer months. The incubation period is approximately 2 to 6 hours, but a delay of 12 to 24 hours is not unusual. Attack rates are very high; 80 to 90% of persons exposed become ill. Symptoms tend to be related to the amount of toxin ingested and vary considerably in their severity. If not fully recovered from an initial ingestion of ciguatoxin, affected persons are likely to have much more serious symptoms from a second ingestion.39 Classically, patients exhibit both gastrointestinal and neurologic symptoms. The gastrointestinal symptoms (e.g., nausea, vomiting, profuse watery diarrhea, crampy abdominal pain, and diaphoresis) tend to appear first and resolve over the first 24 hours. The constellation of neurologic symptoms consists largely of dysesthesias and paresthesias around the throat and the perioral area; “burning feet,” which may resemble alcoholic peripheral neuropathy; “loose, painful teeth”; and sometimes central nervous system changes, such as ataxia, weakness, vertigo, visual hallucinations, and even confusion and coma.39 Distortion of temperature perception is vividly described by patients with ciguatera poisoning. Cold allodynia, defined as dysesthesia experienced on contact with cold water or cold objects, is almost pathognomonic of ciguatera poisoning and often is incorrectly referred to as “cold-hot temperature reversal.” Another classic feature is either a return or a worsening of all of the symptoms after ingestion of alcohol.39 Ciguatera poisoning lasts an average of 1 to 2 weeks, but at least one half of its victims are still symptomatic at 8 weeks. The neurologic symptoms, particularly the paresthesias and dysesthesias, tend to persist longer than the gastrointestinal symptoms and have been reported up to years later. Differential Considerations. Ciguatera fish poisoning should be strongly considered in patients with a combination of gastrointestinal and neurologic symptoms, particularly dysesthesias. Cold allodynia and marked worsening of the symptoms with alcohol ingestion are highly suggestive of ciguatera toxicity. The disease sometimes is misdiagnosed as acute gastroenteritis with “hyperventilation syndrome” because of the combination of gastrointestinal symptoms and paresthesias, particularly when they occur about the mouth and acral areas. Similarly, manifestations of ciguatera toxicity sometimes have been ascribed to malingering because the paresthesias often are transient and vague and lack traditional dermatome patterns. Other disorders that should be considered in the differential diagnosis include paralytic or neurotoxic shellfish poisoning, eosinophilic meningitis, botulism, organophosphate insecticide poisoning, and tetrodotoxin poisoning.39 Management. Treatment is primarily supportive. Intravenous fluids are given to replace volume losses from vomiting and diarrhea, and analgesics are given as needed. In severe cases, the toxin may exhibit some anticholinesterase activity, manifested as bradycardia and hypotension, which can be treated with atropine and dopamine. Patients must be told to abstain from alcohol in any amount until symptoms have completely resolved. Pruritus may be managed with a histamine H1 receptor antagonist such as cetirizine (Zyrtec) at a dosage of 10 mg once
Enterotoxigenic Escherichia coli Epidemiology. Enterotoxin-producing E. coli, or enterotoxigenic E. coli (ETEC), is recognized as a major cause of acute diarrheal disease throughout most of the world. It is a major cause of diarrhea in persons traveling to underdeveloped areas. The disease has been most intensely studied in North American visitors to Latin America, where it occurs in 17 to 70% of travelers studied, often incapacitating them or forcing a change in their plans.40 ETEC is increasingly recognized as a cause of food-borne illness in developed countries, including the United States.41 Infection is acquired from contaminated food or drink. Unpeeled fruits, leafy vegetables, unsanitary drinking water, and ice prepared from impure water are the most common sources. Most tourists are careful about their food and drink, but there seems to be a poor correlation between individual eating habits and the incidence of traveler’s diarrhea. Pathophysiology. For an E. coli strain to cause diarrhea, it must possess both a surface factor that allows colonization (although not invasion) of the small intestine and the ability to secrete an enterotoxin that causes the outpouring of fluids and electrolytes into the small bowel lumen. The enterotoxin-induced secretion occurs in the absence of any demonstrable histologic damage to intestinal epithelial cells or to the capillary endothelial cells.42 Escherichia coli produces both heat-labile and heat-stable toxins. The heat-labile enterotoxin is similar to cholera toxin in that it binds to specific receptors on the surface of the intestinal epithelial cell, allowing translocation of an A subunit into the cell. The A subunit modifies host cell signals, resulting in the dysregulation of sodium and chloride secretion, disturbances of ion transport, and intestinal water loss. The heat-stable toxin exerts its effect through the stimulation of guanylate cyclase in mucosal cells and tends to have a more rapid onset of action. Either or both toxins can be produced by any enterotoxic strain of E. coli. The intestinal fluid losses are qualitatively identical to those in cholera and other toxigenic diarrheas.42 Clinical Presentation. After an incubation period of 24 to 72 hours, an abrupt onset of watery diarrhea occurs. Severity varies, with the illness ranging from a fulminant, cholera-like disease to the much more common and milder turista, in which the symptoms of mild, watery diarrhea and abdominal cramps are more troublesome than life-threatening. Fever is unusual. Vomiting occurs in fewer than one half of affected adults and is seldom responsible for significant fluid losses. Even in severe cases, the diarrhea seldom lasts longer than 48 to 72 hours, and the response to either oral or intravenous fluids is uniformly good. Milder disease generally subsides more gradually, occasionally persisting for 1 week or longer. Virtually all persons recover completely without long-term sequelae.
1213
E. coli enterotoxin–induced disease should be suspected when a child or adult has frequent, watery diarrhea and few other symptoms. It often is passed off as “mild, nonspecific gastroenteritis” and resolves spontaneously. Anyone who acquires toxigenic diarrhea while visiting a developing nation probably has this disease. ETEC is by far the most common cause of traveler’s diarrhea. Diagnostic Strategies. No easy, rapid means of laboratory diagnosis of ETEC infection exists. Methods that rely on identification of specific E. coli serotypes are unreliable because E. coli is part of the normal colonic flora, and its ability to produce enterotoxin is not restricted to any specific serotype. Methods based on detection of the heat-stable and heat-labile toxins using real-time PCR assay have been developed but are not clinically available. Stool preparations show no erythrocytes or leukocytes. Management. Because ETEC infection is almost always a selflimited disease, no treatment other than maintaining hydration is required. However, if the organism is identified while symptoms are still active, or if the patient is traveling in an endemic area, antibiotics can give afford clinical relief. For milder symptoms, a single dose of ciprofloxacin 750 mg by mouth in addition to loperamide should be effective. For more severe symptoms, TMP-SMX 160 mg/800 mg or standard doses of a fluoroquinolone for 3 days should eradicate the organism.3
Clostridium difficile Antibiotic-Associated Enterocolitis Epidemiology. Clostridium difficile is an anaerobic spore-forming gram-positive bacillus that was first linked to enteritis in 1978. It has been associated with a range of illnesses from asymptomatic colonization to severe diarrhea, pseudomembranous colitis, toxic megacolon, intestinal perforation, and death.43 C. difficile can be cultured from the stool of up to 50% of healthy infants and approximately 3% of healthy adults.7 The disease is unique in that an organism normally found in the colon causes illness primarily during or after the administration of antimicrobial agents. Colitis can occur as a result of oral or parenteral administration of most antimicrobial drugs, including quinolones, clindamycin, lincomycin, ampicillin, cephalosporins, tetracycline, penicillin, chloramphenicol, sulfa products, and erythromycin. Other risk factors, in addition to recent use of antimicrobials, include age older than 65 years, severe underlying illness, nasogastric intubation, use of antiulcer medications, and prolonged hospital stay.43 Infection with this organism occurs primarily in adults. Most cases of C. difficile colitis occur in patients who are hospitalized or reside in long-term care facilities. Stool carriage rates reach 16 to 35% in hospitalized patients owing to transfer of organisms by the hands of hospital personnel or from patient to patient.44 Occasional nosocomial infections occur in the absence of antibiotic therapy.7 During the past several years, an increase in the incidence and severity of C. difficile–associated colitis has been noted, along with recognition of a new strain designated NAP1/027 (toxinotype III, North American pulsed field gel electrophoresis type 1, PCR ribotype 027). This strain has been associated with outbreaks in Canada, Europe, and the United States. Data from the CDC indicate that hospitalizations with a discharge diagnosis of C. difficile colitis have increased, from 31 per 100,000 population in 1997 to 61 per 100,000 in 2003.43 Pathophysiology. C. difficile bacteria proliferate when the normal bowel flora is substantially reduced by antibiotic therapy. The organisms must then produce sufficient quantities of toxin for the disease to occur. C. difficile’s primary virulence factors are known as toxin A and toxin B. Toxin A attracts neutrophils
Chapter 92 / Gastroenteritis
daily. Amitriptyline, 25 mg twice a day, can bring about a dramatic reduction in both the pruritus and the dysesthesias, two of the most disturbing and protracted symptoms. Intravenous mannitol has historically been an accepted treatment for ciguatera poisoning. The rationale for its use was based on data compiled from one uncontrolled study of 24 patients, one nonrandomized study, one nonblinded study, and anecdotal reports. More recently, a controlled, randomized double-blinded study of 50 adults found no difference in symptomatic improvement between patients who received normal saline and those who received mannitol. In addition, studies of ciguatoxin-intoxicated animals showed that mannitol did not reverse the effects of ciguatoxin. On the basis of this information, mannitol probably is not an effective therapy for ciguatera poisoning.39
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1214
and monocytes, and toxin B disrupts colonic epithelial cells, both leading to watery diarrhea, colitis, and pseudomembrane formation. In persons who are colonized with C. difficile without experiencing colitis, higher levels of antibody to toxin A have been documented, and development of clinical disease is less likely. In those in whom colitis does develop, higher levels of anti–toxin A antibody are associated with a shorter duration of illness and decreased recurrence rate.44 The incubation period has not been well established.43 On histologic examination, the mucosa is hyperemic and edematous. Raised, yellowish-white plaques, loosely adherent to the mucosa, occur in patches, primarily in the rectosigmoid area, but can occur in any part of the colon. The disease was previously named “pseudomembranous enterocolitis” because of these pseudomembrane-like plaques. Clinical Presentation. Symptoms may appear during the course of antimicrobial therapy or commonly up to 3 or 4 weeks after discontinuation of antibiotics. Of interest, C. difficile has been reported to produce illness as late as 6 months after completion of antibiotic therapy. The clinical presentation is highly variable. Most often, patients present with mild to moderate nonbloody diarrhea associated with lower abdominal cramping and few systemic symptoms. Because the toxin alters the intestinal mucosa, the illness can manifest more like an invasive diarrhea than a toxigenic one, with fever, nausea, dehydration, severe crampy abdominal pain, distention, and profuse watery stools. Occult blood in the stool may be a feature, although hematochezia is uncommon. Fecal leukocytes usually are present. Children tend to have more severe infections than those seen in adults. C. difficile colitis is associated with a mortality rate of 6 to 30% when pseudomembranous colitis is present.44 Diagnostic Strategies. A variety of diagnostic tests are available for the detection of C. difficile. The historical “gold standard” modality is the cell cytotoxicity assay, in which stool is cultured on suitable medium and observed for the cytotoxic effect of toxin B. This test is 94 to 100% sensitive but requires 48 to 72 hours for completion. Stool toxin assays are now the primary method used to diagnose C. difficile infection.7,43,44 Enzyme-linked immunosorbent assay (ELISA) for toxins A and B initially showed low sensitivity, but this has improved sufficiently that now most hospitals use ELISA testing, with a 2-hour turnaround time, as the preferred method. Stool cultures can confirm the presence of C. difficile in the feces of patients with antibiotic-associated enterocolitis (AAC). However, a positive stool culture result is not diagnostic because C. difficile often is present in the feces of normal subjects (especially infants) or in persons receiving antibiotics who do not have an enteritis. Cultures are seldom used clinically but often are part of epidemiologic studies.7,43,44 In patients with typical history and physical examination findings, a tentative diagnosis can be made by sigmoidoscopy or colonoscopy. Differential Considerations. It is important to differentiate C. difficile colitis from simple antibiotic-associated diarrhea. Three percent to 10% of all patients treated with antibiotics, particularly children, develop diarrhea not associated with C. difficile toxin. These patients experience mild, watery diarrhea and no associated constitutional symptoms or evidence of a cytopathic toxin–induced colitis. Management. In early studies, 15 to 23% of patients who had C. difficile colitis experienced spontaneous resolution within 48 to 72 hours of discontinuing the offending antibiotic.44 If discontinuing the antibiotic does not resolve the diarrhea, or if the diarrhea is severe, antibiotic treatment should be started promptly. Either oral metronidazole or oral vancomycin can be used. The dosage of metronidazole is 250 mg orally four times daily for 10 to 14 days. Metronidazole also is effective when
administered by the intravenous route. Doses of vancomycin ranging from 125 to 500 mg orally four times daily have been equally effective.44 Intravenous vancomycin generally is not effective because it does not reach effective intraluminal concentrations. Initial trials comparing metronidazole and vancomycin found similar response rates of approximately 90%. In more recent studies, approximately 78% of patients treated with metronidazole had an initial response to therapy, and approximately one third of these had recurrence of disease, for an overall response rate of approximately 50%. Because vancomycin is much more expensive than metronidazole and because of concerns about emerging vancomycin resistance, oral metronidazole is the agent of choice for treatment in patients with mild disease, and oral vancomycin is reserved for patients who do not respond to metronidazole therapy or for those who are moderately to extremely ill at the time of presentation.43 Patients generally become afebrile and show clinical improvement within 36 to 72 hours; the diarrhea resolves over 5 to 7 days, even though results of toxin assays and stool cultures may remain positive for weeks. From 8 to 50% (average, 25%) of patients suffer a relapse regardless of the antibiotic chosen, its dosage, or the duration of treatment, and the recurrence rate is increasing. Risk factors for recurrent disease include new exposure to antibiotics, age older than 65 years, severity of underlying disease, low serum albumin, need for admission to an intensive care unit, and hospital stay of 16 to 30 days. Nearly all of these patients will respond to another course of antibiotic therapy.44 Adding the yeast Saccharomyces boulardii, 500 mg orally twice daily for 4 weeks, to antibiotic treatment has been shown to dramatically decrease the number of recurrences of C. difficile–associated disease in patients with previous episodes. However, no benefit is seen when S. boulardii is given to patients with an initial episode.44 No serious adverse reactions have occurred with the use of S. boulardii.44 Although toxicity from parenteral vancomycin is common, no adverse effects have been reported with its oral use in the treatment of C. difficile colitis. Antimotility or constipating agents are contraindicated in these patients because of the risk of toxic megacolon and the possibility of increasing the level of cytopathic toxin in the colon.44
■ ACUTE VIRAL GASTROENTERITIS Etiology and Epidemiology. Viral gastroenteritis is the second leading cause of illness in the United States. Although several virus families have been implicated, including caliciviruses, coronaviruses, and parvoviruses, two have predominated in the past decade. Noroviruses, which include the Norwalk virus, are primarily responsible for disease in adults and older children, whereas human reovirus–like agents, also called rotaviruses, cause most diarrheal disease in infants and young children. Worldwide, rotavirus infection causes 600,000 to 875,000 deaths annually and is responsible for 6% of deaths in children younger than 5 years of age. In the United States, rotavirus infection is the leading cause of hospitalization for gastroenteritis and precipitates 600,000 physician visits, 50,000 admissions, and approximately 20 deaths each year. Rotaviruses also cause epidemics in adults, especially those in contact with sick children.7 These viruses have a low infectious dose, and attack rates may reach 50%. The incubation period is short, so explosive outbreaks are common. Norovirus is transmitted by several routes and is relatively stable in the environment, making this infection particularly prone to spread of infection. It is esti-
1215
In adults, the diagnosis of viral gastroenteritis usually is one of exclusion. Viral enteritis is the likely diagnosis when a patient has mild intestinal symptoms, does not appear ill, and further history and physical examination uncover no reason to suspect a bacterial pathogen, inflammatory disease, or any other cause. No investigation beyond the physical examination usually is required. Some of these patients may actually have mild bacterial infections, but these also generally are selflimited, and treatment is not different. A laboratory diagnosis can be made by demonstration of the viruses in stools by electron microscopy or various methods used for the detection of viral antigens, such as latex agglutination, ELISA, or reverse transcriptase–PCR techniques. With rotavirus illness, large numbers of viruses are present in the stools, and these antigen detection methods are quite sensitive and specific. Rotavirus testing probably is indicated only in more serious cases of diarrhea. Fecal leukocytes and erythrocytes are not found in cases of viral gastroenteritis.7 Management. The most important aspect of therapy for acute viral gastroenteritis is fluid replacement. Many children require hospitalization for intravenous fluid and electrolyte repletion. No specific antiviral therapy is indicated. Use of antidiarrheal agents is not recommended in children. In adults, they generally are not needed but may provide some symptomatic relief. Because viral spread is primarily by the fecal-oral route, scrupulous hand washing and other hygienic practices are the best preventive measures. A live tetravalent vaccine against rotavirus was approved for use in infants in 1998 but was withdrawn from the market in 1999 owing to a temporal association with intussusception. Rotateq is a newer oral, liquid live pentavalent vaccine that was licensed in the United States in 2006 and is now part of the annual recommended immunization schedule for all infants at ages 2, 4, and 6 months.
■ PARASITIC GASTROINTESTINAL INFECTION Organisms that commonly cause protozoal enteritis are summarized in Table 92-4. As a group, parasitic pathogens are associated with more prolonged illness than that characteristic for bacterial or viral pathogens. The clinician should consider the possibility of a parasitic infection in patients with diarrhea that persists for longer than 2 weeks, especially in immunocompromised persons, travelers, and residents of developing countries.
Coccidial Infections Cryptosporidium and Isospora belli Infections Epidemiology. Cryptosporidium and Isospora are intestinal protozoal parasites that commonly cause diarrhea in the young of many animal species. In humans, cryptosporidiosis is a worldwide problem, most often seen in persons who handle animals, children in day care centers, healthy homosexual men, and immunocompromised patients.7,46 Cryptosporidium has been the most common cause of chronic diarrhea in persons with AIDS, although the incidence has decreased in the U.S. AIDS population with the use of highly active antiretroviral therapy (HAART).47 Congenital immunodeficiency and treatment with cancer chemotherapeutics or other immunosuppressive drugs are additional predisposing factors. The organism is highly infectious and is easily transmitted among nosocomial, household, and day care contacts or in any facility in which personal hygiene is poor.47 In addition to zoonotic and personto-person spread, indirect transmission occurs by exposure to
Chapter 92 / Gastroenteritis
mated that noroviruses are responsible for 68 to 80% of all outbreaks of gastroenteritis in industrialized countries.7 Both viruses can be transmitted from person to person by the fecaloral route, but water- or food-borne outbreaks also are common. Sources in large reported outbreaks of norovirus infection have included municipal or semipublic water supplies, bodies of water used in recreational swimming, stored water on cruise ships, cafeteria sandwiches, food handlers, and shellfish.7,45 The ingestion of raw oysters has caused many large outbreaks. The CDC estimates that 50% of all food-borne outbreaks of gastroenteritis are attributable to noroviruses.7 Nosocomial spread of infection also is common. Astrovirus and picornaviruses are common causes of diarrhea in HIV-infected patients.7 Pathophysiology. Viruses distort the absorptive cells of the microvilli of the small bowel, decreasing their absorptive surface and causing diarrhea from decreased absorption of fluid and electrolytes. Viral nonstructural proteins also may act as enterotoxins, promoting active chloride secretion into the bowel lumen. The histologic picture resembles that with tropical sprue, and transient malabsorption of fats and sugars, which may persist for a week or more after infection, occurs in patients with viral gastroenteritis.45 Diarrheal stools in viral disease contain more sodium, chloride, and bicarbonate than do normal stools, but these abnormalities are not comparable with the almost isotonic fluid loss of bacterial toxin–induced diarrhea. Potassium loss usually is not significant unless symptoms are prolonged. Clinical Presentation. Viral gastroenteritis occurs primarily in two epidemiologically distinct clinical forms. Outbreaks caused by rotaviruses usually are sporadic, occasionally are epidemic, and typically occur in the winter months in infants and children 6 to 24 months of age. The incubation period is 24 to 72 hours, followed by abrupt onset of vomiting, watery diarrhea, and low-grade fever but little or no associated abdominal pain. Vomiting is a prominent and constant early manifestation of rotavirus enteritis but rarely persists beyond the first 36 hours. The diarrhea generally lasts for 4 to 7 days and may be followed by steatorrhea in approximately 20 to 40% of patients. Many children become significantly dehydrated, requiring hospitalization and intravenous fluid replacement, but the disease rarely is life-threatening in developed countries. Overt clinical disease can occur among family and adult contacts of ill children, but it is uncommon. Most adults with rotavirus infections are asymptomatic. When symptoms do occur, they usually are mild, perhaps because these episodes represent reinfections; 60 to 90% of older children and adults have antibodies to rotaviruses.45 The second clinical entity is characteristically epidemic and is responsible for family and community-wide outbreaks of gastroenteritis among school-aged children, family contacts, and adults. This form generally is caused by the Norwalk virus. After an incubation period of 20 to 36 hours, diarrhea, nausea, and mild abdominal cramps occur. Vomiting is not prominent. Fever typically is absent. Anorexia, headache, malaise, and myalgias may be present. The illness is selflimited, usually lasting only 24 to 48 hours. Most affected adults have mild symptoms and do not seek medical attention.7,45 Diagnostic Strategies. The diagnosis of rotavirus gastroenteritis should be considered in children with significant vomiting, diarrhea, low-grade fever, moderate dehydration, and a normal white blood cell count, especially in those 6 to 24 months of age who become symptomatic during the winter months. An elevated blood urea nitrogen level and a compensated metabolic acidosis are common findings. Serum electrolyte studies indicate that the dehydration usually is isotonic.
1216
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Table 92-4 Epidemiologic Aspects of Protozoal Gastroenteritis PATHOGEN
SOURCES/RISK FACTORS OR GROUPS
INCUBATION PERIOD (I)
FEATURES
Infection may be commensal or intermittently symptomatic or produce severe dysentery 5-10% of U.S. population, malabsorption syndromes or commensal
Entamoeba histolytica
Fecally contaminated food and water sources
3 wk to 4 mo
Giardia lamblia
Water-borne organisms, fecal-oral contact, day care centers, travelers, backpackers, AIDS, homosexual men
1-3 wk
Fecal-oral contact, water-borne organisms, animals, day care centers, AIDS
5-10 days
Fresh fruit, berries, lettuce, water supply Occupational exposure to soil, travel to endemic areas in United States (Kentucky, Tennessee, West Virginia) or overseas Fecal-oral contact, homosexual men
1 wk
Coccidia Cryptosporidium and Isospora Cyclospora cayetanensis Strongyloides stercoralis
Enteromonas hominis
Weeks to months
?
Profuse watery diarrhea, self-limited in the immunocompetent, persistent in the immunocompromised Explosive, protracted, watery diarrhea; fatigue, weight loss Eosinophilia, sepsis, and hyperinfection syndrome in AIDS patients Chronic watery diarrhea, especially in children
AIDS, acquired immunodeficiency syndrome.
fecally contaminated environmental surfaces, toys, food, and recreational water.47 Cryptosporidium oocysts are highly resistant to chlorine and common disinfectants. Large outbreaks have originated from community swimming pools.46 The oocysts also are small (2 to 6 µm in diameter) and may not be removed from contaminated water by standard filtration systems used in the treatment of public water supplies. Ingestion of as few as 10 to 20 oocysts can cause illness in healthy persons. In a recent report, thousands of people contracted cryptosporidiosis from contamination of filtered public water systems that met federal and state standards for drinking water.46 Contamination of rivers and streams by Cryptosporidium has been reported in a number of states.47 In children, cryptosporidiosis is more common in the late summer and early fall and often is associated with intestinal infection with other organisms, particularly Giardia. Isosporiasis generally is an opportunistic infection. It occurs primarily in patients with AIDS, especially Haitians with AIDS, and in homosexual men. In the homosexual community, isosporiasis is a sexually transmitted disease similar to giardiasis and amebiasis.48 Pathophysiology. The pathophysiology is the same for both Cryptosporidium and Isospora. Disease is acquired by ingestion of oocysts. Excystation occurs, and sporozoites attach to the surface of intestinal epithelial cells of the terminal ileum and proximal colon; no tissue invasion occurs. Profuse fluid loss results from a combination of active chloride secretion and malabsorption. A specific enterotoxin has not been identified, although up-regulation of a variety of proinflammatory cytokines, chemokines, and neuropeptides has been observed. In cryptosporidiosis, a biliary reservoir may contribute to chronicity of the infection and inability to eradicate the organism.46 Clinical Presentation. The clinical presentations of cryptosporidiosis and of isosporiasis are indistinguishable and highly variable. After an incubation period of approximately 1 week, symptoms may develop insidiously or suddenly. Infection is characterized by mild to profuse watery diarrhea, crampy abdominal pain, anorexia, nausea, malaise, weight loss, and
flatulence. The diarrhea and abdominal pain often are exacerbated by eating. Immunocompromised patients (especially HIV-infected patients with CD4+ counts less than 200/mm3) can experience enormous stool fluid losses: 3 to 4 L/day is common, and losses may reach 10 to 20 L/day. Physical examination usually reveals only signs of dehydration. Minimal diffuse abdominal tenderness may be elicited by palpation, and fever and leukocytosis are uncommon. Eosinophilia is not a feature. Findings on stool examination for blood or leukocytes are almost uniformly negative in adults but occasionally are positive in children.46,47 The patient’s immune status is the primary determinant of whether the infection is self-limited or persistent. Diarrhea in immunocompetent persons usually resolves after 1 to 3 weeks, but it can continue longer or become chronic. In immuno deficient patients, especially those with AIDS, chronic, persistent diarrhea is common, causing significant discomfort and morbidity unless the infection is responsive to treatment.7,46,47 Asymptomatic infections can occur from either Cryptosporidium or Isospora, and a carrier state has been demonstrated for Cryptosporidium.47 In a U.S. study of immunocompetent patients who underwent upper endoscopy for a variety of reasons, 13% were found to harbor cryptosporidia in the second portion of the duodenum. None of the patients had diarrhea. Diagnostic Strategies. The diagnosis of coccidial infection with either Cryptosporidium or Isospora is made by documenting the oocysts in the stool. Yeasts are similar in size and shape to coccidia but are not acid-fast. Before the development of antigen-based assays, acid-fast staining techniques were the preferred means of diagnosis. Such methods are insensitive for the diagnosis of cryptosporidiosis and require a high oocyst concentration and an experienced examiner. Immunofluorescent assays using monoclonal antibodies to the Cryptosporidium oocyst are commonly used and are 10 times more sensitive than acid-fast staining. ELISA-based detection kits also are commercially available, highly sensitive, and easy to use. Sensitivities and specificities of 94 to 100% have been documented
Cyclospora cayetanensis Infection Epidemiology. Cyclospora is a coccidial parasite widely distributed throughout tropical and subtropical areas of the world that produces disease similar to that caused by Cryptosporidium and Isospora. In Nepal, Cyclospora organisms have been found in up to 11% of patients with diarrheal disease.49 North American outbreaks have been traced to contaminated foods, primarily fresh fruit such as raspberries grown in Guatemala, fresh salads
1217
containing snow peas from Guatemala, fresh basil preparations, and mesclun lettuce; foreign travel; and contaminated water supplies.50 Cyclospora infects all classes of vertebrates, reptiles, and rodents, and the vast majority of cases occur during the spring and summer seasons, often in large outbreaks. Pathophysiology. The exact mode of Cyclospora transmission and mechanism of disease is not completely understood. The organism passes from the human gastrointestinal tract as an oocyst that must sporulate in order to become infective. Sporulation requires several days to weeks outside the human host; therefore, the disease is not transmitted from person to person and is not likely to be transmitted by infected food handlers. Recent work in Nepal suggests that domestic pets and farm animals may serve as intermediate hosts.51 Histologic examination shows flattening of the villi and thickening of the basement membrane in the duodenum with intracellular parasites identified in jejunal biopsies, but it is unknown whether symptoms are toxin-induced or secondary to direct infection of the small bowel.50 As with other parasitic diseases, immunocompromised patients are affected more frequently than immunocompetent hosts.51 Clinical Presentation. The average incubation period is 1 week.50 Typically, the patient presents with an acute onset of explosive watery diarrhea and abdominal cramps. Thirty percent of patients report fever during this initial phase. Severe diarrhea subsides within 1 to 3 days and is followed by a period of intermittent mild diarrhea and marked anorexia and fatigue, which distinguishes this from other parasitic disorders. Weight loss is almost invariable, as is malabsorption of d-xylose.50 Constitutional symptoms are relatively mild, and fever is uncommon. Fecal leukocytes are absent. The disease generally is self-limited in immunocompetent persons but may last as long as 2 to 6 weeks, and relapsing diarrhea is common. Sustained fatigue and weight loss are especially characteristic of cyclosporiasis. These findings are noted in approximately 90% of patients and should prompt consideration of the diagnosis in the setting of outbreaks or history of foreign travel.51 Diagnostic Strategies. The clinical picture may be suggestive, but finding oocysts in the stool is required to confirm the diagnosis. Stool specimens submitted for ova and parasite testing are not routinely examined for Cyclospora unless specifically requested. The oocysts measure 8 to 10 µm in diameter, approximately twice the size of Cryptosporidium oocysts. Oocysts can be identified by modified acid-fast stain, yet the inexperienced microscopist may overlook them or find Cyclospora difficult to distinguish from Cryptosporidium. False-negative and false-positive results on microscopic examination are common. Microscopic examination is greatly facilitated by stool concentration techniques.50 Cyclospora also can be identified by stool PCR testing, although this technique is not widely available in U.S. hospitals. Differential Considerations. The primary confusing organism is Cryptosporidium, because the symptoms of infection with the two pathogens often are indistinguishable. Cryptosporidium tends to be associated with animal contact, exposure in day care centers, and immunocompromised status. Aeromonas hydrophila infection also should be considered in the differential diagnosis. Management. The disease tends to be self-limited in immunocompetent persons, but treatment with sulfa medications is very effective for cyclosporiasis, in contrast with cryptosporidiosis. The drug regimen of choice consists of double-strength TMP-SMX one tablet twice daily for 7 days in immunocompetent patients, and one tablet four times daily for 10 days, followed by one tablet three times per week, in patients with AIDS. The pediatric dose is TMP 5 mg/kg plus SMX
Chapter 92 / Gastroenteritis
in several studies. A PCR-based assay for Cryptosporidium has been developed but is not yet available in clinical laboratories.7,46 Management. In immunocompetent persons, Cryptosporidium infection generally is self-limited; symptomatic therapy and fluid replacement are sufficient. Initial management should focus on replacement of fluid and electrolytes, by the oral route if possible. Antimotility agents such as loperamide or the combination diphenoxylate/atropine (Lomotil) can decrease stool output in persons with mild to moderate illness. For immunocompetent patients with persistent disease, nitazoxanide 100 to 500 mg every 12 hours has been approved by the U.S. Food and Drug Administration (FDA) and has proved to be effective in adults and children older than 1 year of age.46 Immunocompetent patients shed oocysts in the stools when symptomatic and continue to do so for up to 6 weeks after resolution of their clinical illness, creating a public health risk. Oocyte shedding resolves more rapidly in those treated with nitazoxazide.46 Treatment of cryptosporidiosis in immunocompromised patients is more challenging. The most successful interventions occur when the underlying immunodeficiency can be reversed. In patients taking immunosuppressive agents, immune function generally recovers if the drugs can be discontinued. Patients with AIDS and CD4+ counts below 200/µL sometimes respond dramatically if immune function is restored through institution of HAART. 46 Patients severely infected with Cryptosporidium often are fecally incontinent, with large numbers of infectious oocysts in the stool, so strict enteric precautions are necessary to prevent nosocomial spread. If enhancing the immune system fails and symptoms continue to be severe, nitazoxanide 0.5 to 1 g twice daily combined with antidiarrheals may be used, although evidence is lacking for the efficacy of nitazoxanide in HIV-infected patients. Paromomycin (Humatin, 500 to 750 mg three or four times daily) also is sometimes used to ameliorate cryptosporidiosis, particularly in combination with azithromycin, but no good evidence exists that these drugs are routinely effective.46 In contrast with cryptosporidiosis, isosporiasis responds promptly to antibiotic therapy. The treatment regimen of choice for isosporiasis in immunocompetent adults is TMPSMX 160 mg, 800 mg twice daily for 10 days. In immunocompromised adults, the dosage is increased to four times daily for 10 days and then extended with twice-daily dosing for 3 weeks. In patients with sulfonamide sensitivity, pyrimethamine 50 to 75 mg daily may be effective.3 Chronic suppressive therapy with either twice-daily doses of TMP-SMX or daily doses of pyrimethamine often is required, because Isospora infection recurs in more than 50% of patients.3 Patients seen in the ED in whom stool examination for parasites later reveals Cryptosporidium or Isospora should be contacted for appropriate management. Recovering patients require only an explanation of the diagnosis and its ramifications. Patients who are not recovering and those who are known or thought to be immunosuppressed should receive appropriate treatment and referral for follow-up care. Cryptosporidiosis is a nationally notifiable disease.
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1218
25 mg/kg twice daily for 7 days. Thorough washing of fresh produce before consumption decreases but does not eliminate the risk of transmission. Irradiation may be a future solution.3,50,51 Cyclosporiasis is a reportable disease.
Giardiasis Epidemiology. Giardia is one of the most common causes of parasitic diarrheal outbreaks in the United States.7,46 The mode of transmission in large outbreaks is contamination of municipal water supplies with cyst-infested feces from humans or animals, particularly beavers or muskrats, but also dogs, raccoons, and other animals. Campers and backpackers commonly acquire giardiasis, called “backpacker’s diarrhea,” from drinking fecally contaminated water from the “pristine” mountain streams.46 Only rarely is Giardia infection transmitted by contaminated food. Giardia can be spread by sexual or other close person-toperson contact in which fecal contamination may occur, particularly among homosexual men, and in day care centers and institutions for the mentally challenged. The prevalence is 4 to 7% in the general U.S. population, 5 to 20% in homosexual men, and 25 to 50% in children attending day care centers.46 Many cases of acute symptomatic giardiasis in the United States are found in persons returning home from travel elsewhere. Travelers to any developing country can acquire giardiasis, but it is especially feared in those who visit the republics of the former Soviet Union, the Caribbean states, and Latin America, where the water supplies appear to be heavily contaminated with Giardia cysts.52 Small groups traveling to the former Soviet Union, particularly the city of Leningrad, have experienced attack rates approaching 60%, and the disease is ruefully known among its victims as “the Trotskys.” More recently, a Swedish study of giardiasis in returning travelers found that highest attack rates occurred in those returning from the Indian subcontinent, east Africa, and west Africa.53 Patients with decreased gastric acidity, for any reason, are more susceptible to Giardia infection. Giardiasis also is more frequent in patients with various immunoglobulin deficiencies; a relative deficiency of intestinal immunoglobulin A (IgA) may be the reason. Pathophysiology. Giardia trophozoites infect the duodenum, jejunum, and upper ileum. Encystation occurs in the gut lumen, and cysts passed in the feces remain viable for long periods. After the cysts are ingested by the next host, excystation to the active trophozoites occurs in the proximal small bowel, completing the parasite’s life cycle. The trophozoites multiply rapidly. A single diarrheal stool may contain billions of parasites or hundreds of millions of cysts. The trophozoites are capable of superficial invasion of the mucosa, but malabsorption probably causes most symptoms. Clinical Presentation. Many patients harboring Giardia are asymptomatic. The most common symptoms of acute infection are abdominal distention (69%), colicky pain (70%), flatulence (74%), and frequent episodes of explosive diarrhea (89%) producing pale, loose stools that often are offensive-smelling. Audible borborygmi is a classic feature. The serum white blood count usually is normal, and eosinophilia is not seen. The onset typically is sudden, following an incubation period of 1 to 3 weeks. Symptoms may resolve within 7 to 10 days or continue intermittently and produce a malabsorption-like illness, particularly in patients with an immunoglobulin deficiency. The infection resolves spontaneously within 6 weeks in 85% of patients. Diagnostic Strategies. Routine tests (e.g., blood counts, electrolyte panel, radiographic studies) generally are not helpful.
Eosinophilia is not seen. Stool examination is the primary means of diagnosis. In the acute phase of Giardia infection, rapid bowel transit allows trophozoites as well as the more hardy cystic form of the parasite to appear in the stool. Trained observers using standard stool examination techniques will readily identify Giardia in more than 95% of acute cases if three or more stool specimens are studied. Detecting Giardia in cases of subacute, chronic, or asymptomatic infection, however, can be difficult. The trophozoites may be passed only intermittently and in small numbers. Concentration techniques should be used to improve the chance of finding the cysts in the stools. Diagnosis may require small bowel sampling techniques such as duodenal-jejunal aspiration by endoscopy or biopsy of duodenal-jejunal tissue. Giardia antigen tests using immunofluorescence ELISA, nonenzymatic immuno assays, or direct fluorescent antibody (DFA) techniques are replacing microscopic examination as the strategy of choice. The per-specimen cost is similar to that for microscopy; sensitivity is 85 to 98% and specificity 90 to 100%.46 Differential Considerations. Enteromonas hominis, a flagellate parasite like Giardia, can produce an intestinal infection that mimics giardiasis. Enteromonas infection most commonly affects children and homosexual men. Even when all techniques fail to confirm a clinically suspected case of giardiasis, an empirical diagnosis can be supported by a successful trial of appropriate antibiotics. Management. The treatment regimen of choice has historically been metronidazole, a 5-nitroimidazole drug, 250 mg three times daily for 7 days for adults or 5 mg/kg three times daily (maximum 250 mg three times daily) for 7 days for children.3,46 Metronidazole has an efficacy rate of 80 to 95%. Tinidazole is another 5-nitroimidazole agent that is FDA approved for the treatment of giardiasis. Some experts consider tinidazole a first-line agent because it can be given as single-dose therapy. When given in a single 2-g dose (50 mg/kg in children), tinidazole achieves a cure rate of approximately 90%. Furazolidone is the only alternative drug available as a suspension, which may be helpful in treating children. Its cure rates average only 80%, however. The recommended dosage is 100 mg four times daily for adults or 1.5 mg/kg four times daily for children, up to the adult dose, for a total of 7 to 10 days. Nausea and vomiting are common side effects, and rarely a hemolytic anemia develops in patients with glucose-6-phosphate dehydrogenase deficiency.46 Treatment of asymptomatic infections is controversial and best determined on a case-by-case basis. Asymptomatic cyst passers, especially children and food handlers, pose a threat of infection to others and are at risk for the development of intermittent chronic symptoms. Treatment of asymptomatic carriers can theoretically reduce the risk of spread. In heavily infected endemic populations, reinfection is practically universal after 3 months, and treatment is not useful or cost-effective.54 Giardiasis must be considered a family infection. Strict adherence to hand washing is important, especially after using the toilet, playing with pets, or changing diapers. To prevent reinfections, other household members and sexual contacts should be examined and, if found to harbor the parasite, treated appropriately.46
Acute Intestinal Amebiasis Epidemiology. Entamoeba histolytica is a ubiquitous organism infecting at least 10% of the world’s population but originally thought to cause clinical disease in only 10% of persons who carry it. This number probably is an underestimate, however, because recent studies have confirmed the presence
1219
disease is complicated by liver abscess, which is the most common serious complication of amebic colitis.56 Diagnostic Strategies. Definitive diagnosis of intestinal amebiasis historically relied on microscopic identification of the organisms in the stools. Microscopic examination limited to a single stool sample can miss almost one half of infections; at least three fresh samples should be submitted. The examination must precede the administration of antibiotics, antidiarrheal agents, antacids, or enemas, or before the performance of radiographic procedures using barium sulfate. All of these agents destroy trophozoites or distort cysts, thereby interfering with the recovery of amebas. A rectal biopsy specimen or mucosal exudate obtained at sigmoidoscopy may reveal the amebas, even when multiple previous stool examinations have yielded negative results. To obtain mucosal exudate, a glass or metal pipette must be used, because amebas adhere to cotton swabs. Microscopy cannot distinguish Entamoeba histolytica from the nonpathogenic E. dispar. The presence of amebas on stool microscopy in an asymptomatic patient should suggest the presence of nonpathogenic E. dispar; treatment is not indicated.55 The diagnosis of amebiasis has been greatly improved by the development of stool antigen assays. Monoclonal antibody–based enzyme immunoassay for E. histolytica antigens has a sensitivity and specificity of approximately 95% and can distinguish E. histolytica from E. dispar. PCR techniques also have been developed but are not clinically available.55,57 Serologic tests are quite sensitive and specific for active amebic infection. Because administration of steroids to patients with amebic colitis is potentially fatal, and because identification of the parasite in stools is difficult, a serologic test for amebiasis should be done in all newly diagnosed cases of inflammatory bowel disease before initiation of steroid therapy.57 Differential Considerations. Amebiasis should always be considered in cases of acute dysentery-like colitis and in the differential diagnosis for any chronic diarrhea, especially when the stool contains blood-streaked mucus. Amebiasis also should be suspected in homosexual men with acute or chronic colitis. Among patients with AIDS, however, amebic dysentery is rare.58 Patients with nondysenteric amebiasis often are mis diagnosed as having irritable bowel syndrome, diverticulitis, or regional enteritis. Management. Substantial controversy previously existed over whether asymptomatic cyst passers should be treated. Today, however, advances in testing permit distinction between nonpathogenic E. dispar and E. histolytica, so an accurate diagnosis is now possible. Accordingly, it seems prudent to treat for E. histolytica infection in all patients in whom this organism is detected, even asymptomatic carriers. If only E. dispar is identified, treatment is unnecessary. When differentiation is not possible and the patient is asymptomatic, treatment is not recommended unless the clinical picture suggests an increased likelihood of E. histolytica infection. Such would be the case with patients with high specific antibody titers, a history of close contact with a patient with invasive amebiasis, or a patient with symptoms during an outbreak of amebiasis. In symptomatic patients who are diagnosed with E. histolytica or E. dispar infection, other pathogens should be ruled out before E. histolytica can be assumed to be the cause.55 For treatment of benign cyst passers, paromycin (aminosidine) 500 mg orally three times a day for 7 days should be effective. Other regimens include oral iodoquinol, 650 mg orally three times a day for 20 days, and diloxanide furoate, 500 mg orally three times a day for 10 days. For mild to moderate intestinal infection, metronidazole is added (see Table 92-2).57 Treatment with metronidazole should precede treat-
Chapter 92 / Gastroenteritis
of a morphologically indistinguishable ameba, E. dispar, as a separate nonpathogenic species that also colonizes the human gut and probably is responsible for a majority of asymptomatic infections originally attributed to E. histolytica. Distinguishing between the two organisms can be accomplished using ELISA or PCR assay, with PCR testing slightly more reliable.55 Worldwide, approximately 50 million symptomatic cases occur each year, resulting in 100,000 fatalities. In the United States, high-risk groups include travelers, homosexual men, patients with AIDS, and institutionalized persons. Most cases acquired in the United States are asymptomatic; acute amebic dysentery is rare and most often occurs in travelers returning from developing countries in which the disease is endemic.56,57 Entamoeba histolytica exists in trophozoite and cystic forms. The trophozoites infect the colon and may produce symptomatic disease. Infectious cysts are passed in the stool and are highly resistant to environmental factors. Transmission usually occurs through ingestion of cysts present in fecally contaminated food or water. Homosexual men commonly acquire amebic infection from cysts ingested through anal-oral sexual practices. When these patients present with diarrhea, a diligent search for other organisms should be completed before symptoms are ascribed to amebic infection. Co-infection with other enteric pathogens occurs frequently in homosexual men.56,57 Pathophysiology. The factors that determine whether infection with E. histolytica will be commensal or invasive are poorly understood. Variable strain virulence and host susceptibility are determinants. In young children, pregnant women, persons with malnutrition or underlying systemic disease, or persons taking corticosteroids, amebiasis often is more fulminant.57 Invasive trophozoites characteristically produce colonic ulcerations that on histologic examination are seen to have rounded or punched-out margins and are elevated by a submucosal inflammatory reaction from the advancing trophozoites. The ulcer bases are covered with whitish or yellowish exudate. Usually, no diffuse mucosal inflammation exists in areas between ulcers. Should diffuse inflammation occur, however, the picture becomes indistinguishable from that of idiopathic ulcerative colitis or Crohn’s disease. Less than 1% of infections will spread outside the intestines. Gastrointestinal complications can include severe bleeding, toxic megacolon, intussusception, stricture, obstruction, and perforation. Extraintestinal complications include liver and brain infection, as well as pleural or pericardial effusions.55,57 Clinical Presentation. In many patients, E. histolytica lives as a commensal, without producing symptoms. Acute amebic dysentery manifests after an incubation period as short as 1 week or as long as 1 year. The onset is abrupt, with fever; severe abdominal cramps; profuse, bloody diarrhea; and tenesmus. Chronic amebic colitis is the common symptomatic form, for which the onset is gradual. Usually, intermittent diarrhea is present, with two to four foul-smelling stools daily, often containing blood-streaked mucus. Vague abdominal cramps, flatulence, weight loss, and low-grade fever are present. Symptomatic periods may alternate with asymptomatic periods lasting for months to years. The only physical finding may be slight right lower quadrant tenderness to palpation and occasional tender hepatomegaly. The diagnosis is elusive because cysts or trophozoites are difficult to detect; inflammatory bowel disease is a common misdiagnosis, and steroid therapy may exacerbate the symptoms.57 The stools of patients with symptoms contain mucus and leukocytes, although not in large quantity and numbers. Eosinophilia is not seen except in rare cases of ameboma. Liver function test results generally are normal unless the
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1220
ment with aminosidine. Aminosidine can cause diarrhea as a side effect, making it difficult to assess the patient’s response to the metronidazole if both agents are given together.59 Patients with severe infections and systemic illness such as dehydration or hypotension should be hospitalized. Therapy usually is effective, but relapses can occur. Standard precautions to prevent fecal-oral spread are the best preventive measures.
Enterobiasis Epidemiology. Enterobius vermicularis, also known as pinworm or seatworm, is perhaps the most prevalent parasite in the United States. It is estimated that 20 to 30% of all children are infected with pinworms; 200 million people are infected annually wordwide, 40 million in the United States alone. Adult worms are small, spindle-shaped, white to yellowish roundworms that live in the cecum and adjacent portions of the large and small bowel. The female averages 10 mm in length, and the male is 3 mm long. The gravid female migrates through the anal canal at night and oviposits her eggs (up to 50,000) onto the perineal area. The eggs become infective larvae 4 to 6 hours after deposition. Once ingested, the eggs hatch in the duodenum, and the larvae mature as they migrate down to the cecum. Approximately 1 month from the time of ingestion, newly developed, gravid females are again discharging eggs.57 The human body is the only natural host of E. vermicularis. The most common means of infection, particularly in children, is by the direct transfer of eggs from the anus to the mouth on contaminated fingers. Retrograde infection, seen primarily in adults, may sometimes occur. In such cases, larvae hatch in the perineal region, reenter the anus, and migrate to the cecum. Spread within family and children’s groups occurs readily, either by direct transfer of eggs or by airborne transmission. The eggs, which are relatively resistant to desiccation, also contaminate night clothes and bed linens, where they remain viable and infective for 2 to 3 weeks. Pathophysiology. Because E. vermicularis does not penetrate the mucosa, no specific anatomic lesions are initially seen. The movement of the worms or the presence of the eggs on the perineum usually causes local tingling or itching. Scratching causes irritation of the skin, which can lead to excoriations, eczematous dermatitis, and secondary bacterial infections. In women, gravid female worms can migrate through the vagina and uterus into the fallopian tubes, where they may evoke vaginitis, endometritis, or salpingitis. Young girls with pinworms may have a much higher incidence of urinary tract infections than that typical for persons not infected.60 Clinical Presentation. The most common symptom is pruritus ani. This usually occurs at night in relation to the nocturnal migration and oviposition. Scratching may lead to secondary skin changes and bacterial infection. Restlessness, insomnia, and enuresis probably are a result of the pruritus. Alternative presentations include abdominal pain, rectal bleeding, diarrhea, and weight loss. In the absence of host autoinfection, infection typically lasts 4 to 6 weeks. Diagnostic Strategies. Adult worms may be recognized in the perineal area, and in suspected cases, nocturnal examination of this area using a flashlight may confirm infection. Worms sometimes can be seen on the surface of stool as well. The most reliable way to diagnose infection is to examine material taken from the perineal area for ova. The cellophane tape test is simple and reliable. The tape is folded, sticky side out, over the end of a tongue blade, pressed firmly against the perineal area, and then spread on a glass slide with toluene and examined under the low-power objective lens of a microscope. The typical eggs are identified easily.
A single cellophane tape test will detect approximately 50% of infections. If done daily for 3 days, the test will detect 90% of infections; after 5 days, it will detect 99%. Examining stool specimens for ova is rarely helpful; only 5% of patients will have eggs in the stool. Scrapings from under the fingernails may reveal the ova.57 Eosinophilia is not found because the worm does not have a tissue phase. Management. All infected persons in a family or communal group should be treated simultaneously. It is accepted practice to treat empirically all other members of the same group at the same time, even if they are not infected. The drugs of choice are albendazole, 400 mg by mouth once; mebendazole (Vermox), in a single oral dose of 100 mg chewed well; and pyrantel pamoate (Antiminth), in a single oral dose of 11 mg/kg (maximum dose, 1 g). With all of these treatments, a second dose should be administered 2 weeks later. The repeat dose is needed because mature worms seem to be more vulnerable than young worms. Although the maturation process takes 1 to 2 months, a second dose is effective in eradicating the organism in all stages of the life cycle in 90 to 95% of infections.57 The ease of airborne dissemination of the eggs, their resistance to desiccation, and the poor hygienic practices of children all increase the likelihood of reinfection. Ova also are resistant to ordinary fumigants and disinfectants, making control in schools, institutions, and the home very difficult.
Diarrhea in Patients with AIDS Epidemiology. Diarrhea is the most common manifestation of gastrointestinal disease in patients with AIDS and may be the presenting symptom or a life-threatening complication of the disease. The occurrence rate is greater than 90% in developing countries and historically 50 to 60% in the United States. With the widespread use of HAART, the incidence of infectious
Table 92-5 Causes of Diarrhea in Patients with AIDS FREQUENCY
ORGANISM
Most common
Cryptosporidium Cytomegalovirus Entamoeba histolytica (probably commensal, not causative) Giardia lamblia Mycobacterium avium-complex Salmonella species, especially Salmonella enterica subsp. enterica serovar Typhimurium* Aeromonas hydrophila Microsporidium Astrovirus/picornavirus Clostridium difficile Campylobacter jejuni Viruses—herpes simplex virus, rotavirus, adenovirus, Norwalk agent Cyclospora Isospora belli Enteromonas hominis Strongyloides stercoralis Blastocystis hominis Shigella species Yersinia enterocolitica
Common
Less common
*Formerly Salmonella typhimurium. AIDS, acquired immunodeficiency syndrome.
1221
counts less than 100/mm3. Fever, weight loss, and abdominal pain are prominent; diarrhea is mild to moderate. Many patients die within 6 months of diagnosis. Microsporidia have emerged as a common cause of diarrhea in patients with AIDS in whom CD4+ counts are less than 100/mm3. Worldwide, these organisms have been identified in 10 to 20% of patients with AIDS who are experiencing diarrhea. In one prospective study in the United States, microsporidia were identified in 39% of patients with AIDS undergoing small bowel biopsy for evaluation of diarrhea.63 Salmonella infections, especially with S. Typhimurium, are common in immunocompromised hosts.63 Patients with AIDS who acquire Salmonella enteritis are at increased risk for bacteremia and metastatic focal infection compared with normal hosts. Clostridium difficile enteritis occurs more commonly in patients with AIDS owing to the common use of prophylactic antibiotic therapy and frequent hospitalizations. It is the most common bacterial enteritis in the AIDS population. Clinical presentation, response rate, and relapse rate are no different from those in the healthy host.63 Giardiasis occurs with a frequency and severity that are unrelated to the degree of immune compromise.55 Entamoeba histolytica infection is neither more prevalent nor more severe in HIV-infected patients. When amebae are present, it is rare for these organisms to cause invasive disease in patients with AIDS.63 For unknown reasons, Yersinia, V. parahaemolyticus, viruses (non-cytomegalovirus), Neisseria gonorrhoeae, and Chlamydia trachomatis are relatively unusual causes of diarrhea in patients with AIDS. Shigellosis and campylobacteriosis are more common among male patients with AIDS who have sex with men than among HIV-negative men who have sex with men. Clinical Presentation. In patients with AIDS, diarrhea presents in one of three ways. First, at the time of HIV seroconversion, patients usually experience diarrhea, nausea, anorexia, and malaise in association with an acute infectious mononucleosis–like syndrome. Second, diarrhea may be the presenting symptom of full-blown AIDS, with associated fever, malaise, anorexia, and significant weight loss. The most common presentation, however, is for diarrhea to start well after AIDS has become clinically apparent and when the CD4+ count falls below 300/mm3. In these cases, the patient typically has a chronic debilitating infection that rarely remits spontaneously unless CD4+ counts are normalized using HAART. Diarrhea often is accompanied by profound weight loss, major nutritional impairments, and a diminished sense of well-being. Many cases are refractory to antimicrobial treatment and persist until death, or may even be the cause of death.55,63 In patients with AIDS, the presenting signs and symptoms generally do not allow consistent classification of diarrheas, as is done for the immunocompetent host. This is so in part because many patients with AIDS have multiple, concomitant enteric pathogens. However, some clinical pictures are typical. Patients with a fulminating clinical course usually have a disseminated infection, such as infection with cytomegalovirus or M. avium-complex. Massive weight loss also is associated with diarrhea due to infection with those two organisms and the coccidia Cryptosporidium and Isospora. Voluminous, watery diarrhea usually is due to one of the coccidial organisms. Patients with a proctocolitis-like picture most often have herpes simplex virus or cytomegalovirus infection. Strongyloidiasis should be considered in any immunocompromised patient who experiences sudden clinical deterioration with eosinophilia, polymicrobial sepsis, meningitis, or adynamic ileus.63 Complications. The most common complications are dehydration, electrolyte abnormalities, and malnutrition resulting from
Chapter 92 / Gastroenteritis
diarrhea has decreased dramatically in the United States and other developed countries, even among those with CD4+ counts less than 200/mm3. It is believed that most cases of diarrhea in patients with AIDS in developed countries are now noninfectious in nature, most often related to the use of protease inhibitors; hence, the clinician should ask about recent changes in the HAART regimen in addition to considering infectious causes.46,55 Worldwide, patients who are HIV-positive and those with active AIDS are more susceptible to infection both from the usual enteric organisms and from opportunistic organisms. Diarrheal diseases are much more problematic in patients with AIDS because of their diminished immunity and underlying poor nutritional status.58 A vastly expanded profile of pathogens may potentially cause diarrhea in patients with AIDS compared with immunocompetent persons. Also, the ramifications of the disease are significantly more serious, necessitating a much more aggressive diagnostic evaluation and treatment regimen. Pathophysiology. The early concept of mucosal transmission of HIV was that it occurred through mucosal trauma that created breaches in the physical barrier of the genital or rectal epithelium, allowing introduction of viral particles directly to the bloodstream. It is now known that infection also occurs after atraumatic inoculation of mucous membranes. With gastrointestinal exposure, virions of certain HIV strains bind to particulate-sampling M cells (membranous epithelial cells) in the gastrointestinal mucosa and undergo receptor-mediated transport from intestinal lumen to lamina propria, with subsequent infection of CD4+ lymphocytes located in the gastrointestinal mucosa. Mucosal biopsy specimens in up to 40% of patients with AIDS demonstrate HIV virions, with depletion of CD4+ lymphocytes in the lamina propria and apoptosis of both HIVinfected and non–HIV-infected lymphocytes. It also has been observed that in a significant number of HIV-infected patients, diarrhea develops in the absence of an identifiable gastrointestinal pathogen, and that this pathogen-negative HIV-related diarrhea subsides with institution of HAART. These findings suggest that HIV itself can induce a primary enteropathy, although specific mechanisms have yet to be elucidated.61 In the population of homosexual men, unprotected receptive anal intercourse and anal-oral contact among multiple partners provide exposure to a diverse spectrum of enteric pathogens. In the heterosexual intravenous drug–abusing population, infection spreads primarily by water- and foodborne transmission of the organisms. Patients with AIDS are unable to combat these intestinal pathogens, probably because of a combination of T lymphocyte functional deficiency and underlying HIV-induced enteropathy. Etiology. A known enteric pathogen can be identified in approximately 85% of patients with AIDS who are experiencing diarrhea. Multiple pathogens may be present in as many as 20 to 25% of patients. Homosexual men with AIDS experience diarrhea more often than do other patients with AIDS. Diarrhea also develops more often in patients who do not have access to or are not compliant with HAART and in those infected with resistant strains of HIV than in other AIDS patients.62 Cryptosporidium and cytomegalovirus infections are the two most common causes. The incidence of each is approximately 20%.62,63 Chronic high-volume watery diarrhea most often is from one of the coccidia Cryptosporidium and Isospora belli. Although self-limited in the healthy host, coccidial disease often is persistent in patients with CD4+ counts less than 200/mm3. Cytomegalovirus and Mycobacterium aviumintracellulare also produce a chronic illness in those with CD4+
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1222
both fluid loss and malabsorption. Cytomegalovirus infection can produce gastrointestinal hemorrhage or perforation or toxic megacolon. M. avium-complex infection often manifests with severe anemia, weight loss, and a rapid downhill course of persistent weakness, malaise, and malabsorption. Bacteremia historically was found in as many as 40 to 45% of patients with AIDS and diarrhea, usually caused by M. aviumcomplex or Salmonella species. The incidence has decreased markedly over the past decade as the use of HAART has increased.63 Diagnostic Strategies. The diagnostic approach to patients with AIDS who are experiencing diarrhea is entirely different from that to an immunocompetent host. In 80 to 90% of cases, one or more enteric pathogens are found; of the pathogens identified, a majority are susceptible to appropriate antimicrobials.55,62 Diarrhea in patients with AIDS in whom CD4+ counts are less than 300/mm3 generally is not self-limited and requires medical intervention to effect resolution. Therefore, each patient merits a diagnostic evaluation (Box 92-2). Infectious causes are detected less frequently in patients being treated with HAART.62 Patients with AIDS who are experiencing diarrhea should have up to three stool specimens cultured for enteric bacteria, because fecal shedding of pathogenic bacteria can be intermittent. Stool also should be examined for parasites, using antigen detection assays where available or microscopy of multiple samples for ova, parasites, and mycobacteria. The usual bacte-
Diagnostic Protocol for Evaluating Diarrhea in
BOX 92-2 Patients with AIDS Initial Evaluation Indicated in all patients.
1. Stool cultures for enteric bacteria—Salmonella, Shigella, Campylobacter, Yersinia, STEC 2. Stool examination for Giardia, Cryptosporidium, Isospora, Cyclospora, amebas, and mycobacteria; Clostridium difficile toxin assay 3. Blood cultures 4. Proctosigmoidoscopy in patients with clinically severe colitis or a proctocolitis picture, especially homosexual men Further Evaluation Indicated if initial study results are negative or to look for multiple organisms present if a patient fails to respond to appropriate therapy for an identified pathogen. 1. Repeat stool cultures and examinations, possibly add culture for viruses 2. Colonoscopy or esophagogastroduodenoscopy performed to obtain duodenal fluid and small bowel and colonic biopsies, which are examined for: a. Duodenal fluid examination for ova and parasites b. Duodenal and colonic biopsy specimens cultured for mycobacteria, cytomegalovirus, and herpes simplex virus; colonic biopsy tissue also is cultured for bacterial enteric pathogens (add testing for gonorrhea and chlamydial infection in patients with acute proctitis) c. Biopsy specimens examined using multiple stains (e.g., acid-fast, hematoxylin-eosin, Giemsa, silver, periodic acid–Schiff ) for protozoa, mycobacteria, and cells containing viral inclusion bodies STEC, Shiga toxin–producing Escherichia coli.
rial enteric pathogens are readily identified, but the protozoal infections can be difficult to detect if microscopy-based strategies are used. In patients with severe diarrhea from Cryptosporidium, the first stool examination usually yields a positive result, but in less severe cases, additional samples may have to be studied to make the diagnosis. Specialized techniques are necessary to detect most of the viral agents. Blood cultures are a valuable diagnostic adjunct. Bacteremia may be found in up to 43% of patients, and a blood culture may yield a positive result when the stool cultures and examinations fail to reveal a pathogen.64 The most common organisms detected by means of blood culture are M. avium-complex, cytomegalovirus, Salmonella, and occasionally herpes simplex virus. Most cases of M. avium-complex bacteremia occur in heterosexual intravenous drug abusers. A routine complete blood cell count demonstrating eosinophilia suggests parasitic infection with Strongyloides stercoralis or Isospora.65 If results of these diagnostic tests are negative, endoscopy should be performed to obtain mucosal biopsy samples from affected areas of the rectum or colon. Rectal biopsy, which can be performed easily even in seriously ill patients, is an indispensable tool in the diagnosis of cytomegalovirus infection. Viral inclusion bodies with clear halos typical of cytomegalovirus can be demonstrated or cytomegalovirus DNA can be detected using PCR techniques. Special staining examinations, DNA detection techniques, or culture of the biopsy tissue also may diagnose M. avium-complex, Cryptosporidium, or Giardia infections that were missed on stool examination. Herpes simplex virus infection can be identified microscopically by detecting multinucleated giant cells or through cultures obtained at the time of endoscopy. Small bowel biopsy and duodenal aspiration are indicated when stool examination, cultures, and sigmoidoscopy fail to yield a definitive diagnosis. Small bowel studies are most helpful for detecting infection due to Cryptosporidium, cytomegalovirus, M. avium-complex, Giardia, or I. belli.63 Differential Considerations. Kaposi’s sarcoma, even if it involves the bowel, rarely produces diarrhea. Symptomatic oral and esophageal candidiasis is common in patients with AIDS, but diarrhea from Candida has not been reported. Diarrhea can be a side effect of antiretroviral drugs used to treat AIDS, especially the protease inhibitors nelfinavir and the combination agent lopinavir plus rotinavir.55 Antibiotic-associated colitis from C. difficile should be considered when onset of the diarrhea follows antibacterial therapy or hospitalization. Ulcerative colitis can mimic or be mimicked by cytomegalovirus colitis.63 Aphthous ulceration, particularly of the colon, should be considered when cultures and endoscopy biopsy specimens fail to reveal evidence of infection with common infectious pathogens, herpes simplex virus, or cytomegalovirus. Empirical corticosteroid therapy may lead to dramatic improvement in these patients.63 Acute proctitis should be differentiated from diarrhea or acute colitis because the investigative evaluation and treatment regimens are distinctly different. Management. The best strategy for treatment and prevention of acute and chronic infectious diarrhea is maintenance of CD4+ counts through HAART. Treatment of diarrhea in patients with AIDS also includes diet, antimotility agents, and antimicrobial agents. Diets that are lactose-free and low in fat often diminish the diarrhea caused by malabsorption. Patients should avoid intestinal stimulants such as caffeine, raw or inadequately cooked seafood, rattlesnake preparations, and untreated water. Variable success has been reported with standard antimotility agents such as diphenoxylate or loperamide. In patients with cryptosporidiosis, these agents commonly cause a marked increase in crampy abdominal pain; long-acting
■ TRAVELER’S DIARRHEA Epidemiology. It is said that “travel expands the mind and loosens the bowels.” Diarrhea is by far the most common health problem among the 12 million people who travel from an industrialized nation to a developing country each year. Travel to high-risk areas is associated with diarrheal attack rates of 30 to 50%. Among travelers to industrialized countries, however, the development of diarrhea is infrequent. For visitors to the United States, the attack rate is less than 4%.55 Traveler’s diarrhea is more common in young adults than in elderly persons. Travelers with inflammatory bowel disease, diabetes, or immune compromise and those taking antacids also are at increased risk.67 Pathophysiology. The syndrome is caused by an infection acquired by ingesting fecally contaminated food or water. High-risk items include raw leafy vegetables, raw or undercooked meats or seafood, unpeeled fruits, unpasteurized dairy products, tap water, and ice. Once an organism is ingested, rapid and dramatic change occurs in the traveler’s intestinal flora. When the ingested inoculum overcomes host defense mechanisms, diarrhea develops. Most often, the pathomechanism involves elaboration of enterotoxins that produce a secretory diarrhea. When organisms are invasive, rather than toxigenic, a typical infectious enteritis results. Etiology. Pathogens that commonly cause traveler’s diarrhea are listed in Table 92-7. Bacterial enteropathogens cause approximately 80% of cases of traveler’s diarrhea. Enterotoxigenic E. coli is the most common pathogen and can be acquired
1223
anywhere in the world, although the incidence of infection is higher in Latin America than in Africa, Asia, or the Middle East. V. parahaemolyticus is an increasingly common cause because of its association with consumption of raw or inadequately cooked seafood. The organism commonly causes diarrhea in persons traveling to Japan or Asia or in those vacationing on cruise ships.15,55,68 Plesiomonas shigelloides typically is associated with uncooked shellfish, especially oysters. It also is associated with travel to Mexico. Plesiomonas produces a typical invasive enteritis.15 E. coli O157:H7 and enteroinvasive E. coli each cause up to 5% of cases of traveler’s diarrhea.68 Another type of E. coli, enteroadherent E. coli, may be the cause of a significant number of the previously undiagnosed cases of traveler’s diarrhea. Viral agents cause up to 10% of cases of traveler’s diarrhea.68 Parasitic pathogens are implicated in approximately 10% of cases of traveler’s diarrhea. Giardia is the most common parasite acquired by travelers; travelers to the former Soviet Union, particularly Leningrad and Moscow, have a very high risk for acquiring giardiasis. Entamoeba histolytica, Cyclospora, and Cryptosporidium are each implicated in less than 2% of cases.68,69 Clinical Presentation. Traveler’s diarrhea typically begins abruptly, with production of four or five loose or watery stools per day for 1 to 3 days. Approximately one third of patients are temporarily confined to bed, and the symptoms last more than a week in 10% of patients.68 Onset usually is within the first 3 to 4 days of travel but can occur at any time, including after the patient arrives home. Patients may not associate their diarrhea with recent travel because the incubation time for the infection, particularly if it is parasitic, may have been long enough to allow them to return home before the symptoms began. Associated symptoms include abdominal cramps, nausea, bloating, urgency, and occasionally vomiting, fever, chills, headache, malaise, tenesmus, and bloody stools. Symptoms and signs depend on whether the pathogen is noninvasive or invasive. Traveler’s diarrhea may ruin a trip, but it is rarely life-threatening. Prevention. Diet. Traditionally, instruction regarding food and beverage preparation has been touted to prevent traveler’s diarrhea. Most travelers, however, do not follow such advice. Ideally, tourists should eat only foods that are freshly prepared and served piping hot. High-risk foods should be assiduously avoided, and travelers should follow the Peace Corps adage of “boil it, cook it, peel it, or forget it.” Thirsty travelers should be advised to avoid ice and to drink beverages such as tea and coffee that are made with boiled water, canned or bottled carbonated beverages, and wine.67,68 Boiling water is by far the most reliable method to make it safe for drinking and brushing teeth. Travelers and outdoor enthusiasts should be advised to bring the water to a vigorous boil and allow it to cool without adding ice. Boiling destroys virtually all bacteria, viruses, and parasitic cysts. A pinch of salt in each quart improves the taste. When boiling is not feasible, water can be chemically disinfected with 2% tincture of iodine drops or tetracycline hydroperiodide tablets, such as Globaline or Potable Aqua, available from pharmacies and sporting goods stores. Nonantimicrobial Medications. The nonantimicrobial agent most studied in the prevention of traveler’s diarrhea is bismuth subsalicylate (e.g., Pepto-Bismol). Taking two tablets, or 2 ounces, four times per day decreases the incidence of traveler’s diarrhea by 65%. This dosage, however, contains the daily equivalent of eight 5-grain aspirin tablets. Bismuth subsalicylate should not be used in patients who are allergic to salicylates, those taking large doses of salicylates for arthritis, or
Chapter 92 / Gastroenteritis
morphine sulfate derivatives have provided better clinical relief.46 Specific antimicrobial therapy can lead to marked symptomatic improvement in 55 to 75% of patients in whom a pathogen has been identified (Table 92-6). Many patients show substantial improvement after therapy, even when the organism has not been eliminated or the diarrhea persists. Antimicrobial therapy should be dictated by the results of the diagnostic evaluation. Empirical use of antimicrobial agents is not indicated; no one antimicrobial agent can possibly provide reasonable coverage for the wide variety of causative organisms found in these patients. The possibility of infection by multiple organisms must always be considered, especially when patients do not respond to antimicrobial agents known to be effective aganst an identified pathogen. Ganciclovir, a nucleoside analogue similar to acyclovir, is effective in inducing clinical remission in up to 90% of patients with gastrointestinal cytomegalovirus infection.66 Foscarnet also is effective and typically is used in ganciclovir-resistant infections. The dosage of both drugs should be reduced in patients with low creatinine clearance. Treatment for the other most common cause of diarrhea in patients with AIDS, cryptosporidiosis, has improved with the use of nitazoxanide: in one study of HIV-positive patients with CD4+ counts greater than 50/mm3, 90% responded to a dose of 2 g/day, compared with 20% of those who received placebo.46 Infection with I. belli and C. cayetanensis can be cured with TMP-SMX.3 M. avium-complex infection is poorly responsive to therapy, and death generally ensues within 6 to 8 months of diagnosis. Virtually all of the other organisms listed in Table 92-5 are susceptible to the usual therapeutic agents, although higher dosages and longer courses of treatment often are required. Recurrences of diarrhea due to either the opportunistic organisms or the usual enteric organisms are common. Chronic suppressive antimicrobial therapy may be indicated to prevent relapse or reinfection.3
Table 92-6 Treatment Directed against Pathogens Causing Diarrhea in Patients with AIDS ORGANISM/TREATMENT REGIMEN
Cytomegalovirus Foscarnet 90 mg/kg IV q12h × 14–21 days (diluted in 100 mL D5W, infused over at least 1 hr) for 3–6 weeks Ganciclovir 5 mg/kg q12h IV × 14 days (diluted in 100 mL D5W, infused over 1 hr) for at least 21 days Cryptosporidium Paromycin 25–35 mg/kg/d PO in 2–4 doses × 4 wk Azithromycin 600 mg PO qd × 4 wk in combination with paromycin Nitazoxanide 1 g PO bid × 4 wk
COMMENTS
Effective, but 75% recurrence rate within 8 to 9 weeks; maintenance therapy may be warranted
Disease generally chronic, despite treatment
Cyclospora cayetanensis TMP-SMX 160 mg/800 mg qid PO × 10 days, then 1 tab PO 3×/wk Entamoeba histolytica Iodoquinol 650 mg PO tid × 20 days Diloxanide 500 mg PO tid × 10 days severe symptoms, add metronidazole 750 mg PO tid × 10 days Paromycin 25–35 mg/kg divided tid × 7 days
Treat only if distinguished from Entamoeba dispar, or for severe symptoms, which warrant search for other causes as well
Giardia lamblia Metronidazole 250–750 mg PO tid × 5–10 days Tinidazole 2 g PO as a single dose Furazolidone 100 mg PO qid for 7–10 days
Patient’s symptoms may resolve despite continued enteric presence of Giardia
Mycobacterium avium-complex Clarithromycin 500 mg PO bid or azithromycin 500 mg PO qd plus ethambutol 15 mg/kg/d with or without rifabutin 300 mg PO qd until response Levaquin 500 mg PO qd
Mixed results with antituberculosis drugs, more drug-resistant than tuberculous strains of mycobacteria; little evidence that treatment prolongs life
Salmonella species Ciprofloxacin 500–750 mg PO bid × 10–14 days or azithromycin 1 g day 1, then 500 mg × 10 days or ceftriaxone 1–2 g IV q12h × 7–10 days Herpes simplex virus Acyclovir 5 mg/kg PO or IV tid × 7–10 days Famciclovir 250 mg PO tid × 7–10 days Valacyclovir 1 gm PO bid × 7–10 days Campylobacter jejuni Erythromycin 500 mg bid × 7–10 days or azithromycin 500 mg qd × 5–7 days Isospora belli TMP-SMX 160 mg/800 mg PO qid × 10 days, then bid × 3 wk Pyrimethamine 50–75 mg qd × 10 days Aeromonas hydrophila Ciprofloxacin 500 mg PO bid × 3 days TMP-SMX 160 mg/800 mg PO bid × 3 days Enteromonas hominis Metronidazole 250–750 mg PO tid × 10 days
Blastocystis hominis Metronidazole 750 mg PO tid × 10 days or furazolidone 100 mg PO qid × 7–10 days Shigella species Ciprofloxacin 500 mg PO bid × 7 days Yersinia Ciprofloxacin 500 mg PO bid × 7 days Strongyloides stercoralis Ivermectin 200 µg/kg/day × 1–2 days or thiabendazole 25 mg/kg PO bid × 2 days
Bacteremia common; maintenance therapy often required
Proctitis picture, especially in homosexual men
40% recurrence rate; high rate of fluoroquinolone resistance
50% recurrence rate; chronic suppressive therapy usually recommended Associated with drinking untreated water
Increasingly common in homosexual men, possibly commensal; treatment indicated when no other pathogens found in the presence of appropriate symptoms Possibly more common in children
Most species resistant to ampicillin, and increasing resistance is found to TMP-SMX Appendicitis-like picture; bacteremia possible Migration of the larvae through the bowel wall may be accompanied by gram-negative bacteremia and a hyperinfection syndrome in patients with AIDS; in disseminated strongyloidiasis, therapy should be continued for at least 5 days
AIDS, acquired immunodeficiency syndrome; D5W, dextrose 5% in water; HAART, highly active antiretroviral therapy; TMP-SMX, trimethoprim-sulfamethoxazole.
1225
Table 92-7 Causes of Traveler’s Diarrhea
Bacteria (in approximately 80–85% of cases) Enterotoxigenic Escherichia coli Shigella Campylobacter Enteroinvasive E. coli (hemorrhagic strain of O157:H7) Salmonella Others, such as Vibrio, Aeromonas, Plesiomonas shigelloides, Yersinia, other types of E. coli
ESTIMATED INCIDENCE (%)*
45–50 8–12 7–9 5–6 3–5 1–5
Viruses (in approximately 5–10% of cases) Rotavirus Norwalk agent and others
5–10 0–5
Parasites (in approximately 5–6% of cases) Giardia lamblia Cryptosporidium Entamoeba histolytica Strongyloides stercoralis Unknown
4–5 3–4 0–1 0–1 5–10
*Rough estimates, which vary depending on locales visited.
patients taking oral anticoagulants, uricosuric drugs, or methotrexate. Salicylates have antiplatelet effects, inhibit the activity of uricosuric drugs, and increase the toxicity of methotrexate by decreasing its renal clearance. Bismuth subsalicylate also turns the tongue and stool black and may cause mild tinnitus and interfere with the bioavailability of doxycycline. Antiperistaltic agents such as diphenoxylate and loperamide are not effective prophylactic agents. Controlled studies have indicated that the prophylactic use of diphenoxylate actually increases the incidence of traveler’s diarrhea: slowing of the gut allows more time for organisms to colonize and elaborate toxin or produce infection.68 Antimicrobials. The risks associated with prophylactic treatment with antibiotics include allergic reactions, skin rashes, photosensitivity reactions, serious hematologic reactions, Stevens-Johnson syndrome, and antibiotic-induced infections such as antibiotic-associated colitis or Candida vaginitis.68 The main argument, however, against the widespread use of prophylactic antibiotics is the risk of emergence of resistant organisms. Resistance to doxycycline is now found in many parts of the world. TMP-SMX resistance is common in tropical areas. The quinolones are the most effective agents, but resistance rates of up to 80% have been reported for some organisms in parts of Southeast Asia.68,69 In addition, if the traveler is provided with a prescription for a suitable antibiotic based on regional resistance patterns, taking one dose in combination with loperamide immediately after the onset of diarrhea often will halt symptoms within hours. For these reasons, prophylactic use of antibiotics before the onset of symptoms is no longer recommended, except in cases of significant concurrent illness or immunocompromised status.68,69 If prophylaxis is indicated, a single daily dose of prophylactic antibiotics such as ciprofloxacin (500 mg), norfloxacin
Chapter 92 / Gastroenteritis
AGENT
(400 mg), or rifamixin (200 mg), in combination with bismuth subsalicylate two tablets (262 mg/tablet) or 30 mL four times daily can effectively prevent traveler’s diarrhea in up to 90% of persons. The regimen is started on the day before travel and continued until 2 days after returning home. TMP-SMX also can be used but is not recommended as a first-line agent because of significant resistance. Management. Diet. In most patients, fluid and electrolyte balances can be maintained by drinking potable fruit juices, bottled beverages, or caffeine-free soft drinks. Nonantimicrobial Medications. Adsorbents such as kaolin or pectin are ineffective for treating traveler’s diarrhea. They may give the stools more consistency but have not been shown to decrease cramping or the frequency of stooling, or to shorten the course of an infectious diarrheal illness. Nonantimicrobial agents such as bismuth subsalicylate, paregoric, codeine, diphenoxylate, or loperamide may provide prompt but temporary symptomatic relief. Loperamide (initial dose 4 mg followed by 2 mg after each unformed stool for 2 days; total dosage no more than eight 2mg capsules/day) has been shown to provide significantly more relief than that obtained with bismuth subsalicylate (30 mL orally every 30 minutes for 3 to 5 hours on each of 2 days, for a total dose of 240 mL/day).68 Antimotility agents alone, however, should never be given to patients when an invasive bacterial infection is suspected, and they should be discontinued if symptoms persist longer than 48 hours. Antimicrobials. Antibiotic therapy can provide prompt relief of symptoms, decrease the rate of stooling, and shorten a typically 1- to 3-day illness to a few hours.67-69 In cases of mild toxigenic, nondysenteric traveler’s diarrhea, a single dose of ciprofloxacin 750 mg PO in combination with loperamide often resolves the symptoms within an hour. When the course is extended to 3 days, response rates of up to 98% are attained. In patients with high fever, bloody stools, or the typical bacterial or invasive picture, the treatment of choice is norfloxacin, 400 mg twice daily, or ciprofloxacin, 500 mg twice daily, in combination with loperamide. The duration of treatment generally is 3 days, although one double dose may be all that is necessary.68 Persons with dysentery who fail to respond to one antibiotic should promptly be switched to another. Azithromycin may be effective in cases in which fluoroquinolones either cannot be used or are ineffective. The quinolone agents are not recommended for use in children and pregnant women.68 The antibiotic rifamixin was FDA-approved in 2004 for the treatment of traveler’s diarrhea in adults and children older than 12 years of age. It is a nonabsorbable antibiotic that acts by inhibiting RNA synthesis, and the potential for development of bacterial resistance to this agent is believed to be low. Several studies demonstrate that oral rifamixin 200 mg three times daily for 3 days is an effective treatment for traveler’s diarrhea.68 A summary of current recommendations for the prevention and treatment of traveler’s diarrhea is outlined in Box 92-3. Further detailed information on traveler’s diarrhea and the other medical problems of travelers can be found at the Center for Preventive Services section of the CDC’s website, http:// wwwn.cdc.gov/travel/default.aspx. The CDC’s “Yellow Book,” Health Information for International Travel, can be downloaded for free at http://wwwn.cdc.gov/travel/contentYellowBook. aspx, and the CDC provides up-to-the-minute travel infor mation through its traveler’s hotline telephone number, 877-FYI-TRIP.
1226
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
BOX 92-3 Current Recommendations for the Prevention and Treatment of Traveler’s Diarrhea 1. Provide instruction regarding sensible dietary practices and drinking water management. 2. Reserve prophylaxis for patients with special circumstances, including severe concurrent illness or immunosuppression. The basis of this recommendation is threefold: first, the potential for aspirin-related complications; second, the ramifications of widespread use of prophylactic antibiotics in terms of adverse medication reactions or the emergence of resistant organisms; and third, the availability of highly successful treatment strategies. 3. Other patients who request prophylaxis should be steered toward the use of bismuth subsalicylate rather than antibiotics.
4. For those requiring prophylaxis, one of the quinolones— norfloxacin, ciprofloxacin, or rifamixin—is recommended. 5. For all patients requiring prophylaxis, it is reasonable to institute prompt antimicrobial therapy once traveler’s diarrhea occurs. a. Toxigenic/nondysentery: loperamide combined with one dose of ciprofloxacin 750 mg orally. b. Infectious/dysentery: norfloxacin, ciprofloxacin, or rifamixin—alone or in combination with loperamide twice daily for 3 days. c. The rare traveler with persistent symptoms, particularly fever, chills, or blood or mucus in stools, unresponsive to antimicrobial therapy within 24 to 48 hours, should seek immediate medical attention.
KEY CONCEPTS ■
Contact isolation precautions should be followed for all patients with significant diarrhea that continues during the ED stay. ■ Most otherwise healthy patients presenting with diarrhea do not require laboratory testing or antibiotic therapy. ■ Fecal testing should be reserved for those patients who are systemically ill, are febrile, have significant comorbid disease, have bloody stools, or report recent antibiotic use, or in whom exposure to a treatable or reportable pathogen is strongly suspected. ■ Empirical treatment of diarrhea in adults with a fluoroquinolone or TMP-SMX generally is not recommended but is acceptable for traveler’s diarrhea and when severe inflammatory diarrhea is present with a low risk of E. coli O157:H7 as its cause. This indication includes bloody diarrhea in patients with a temperature greater than 38.5° C. “False-positive” blood in stools may come from aggravation of hemorrhoids or perianal irritation secondary to copious stooling. Fecal leukocytes or lactoferrin testing may help distinguish inflammatory diarrhea from that of other causes but has a low specificity. Stool testing should be performed before institution of all empirical treatment. ■ Milk and other lactose-containing foods should be avoided because gastroenteritis often leads to transient lactose intolerance. Caffeine products also may worsen diarrhea. Bowel rest is not necessary or advantageous, however, and patients should be encouraged to eat and drink what they find palatable during the illness. ■ Risks of empirical antibiotic treatment include adverse and allergic reactions, possible increase in the risk of hemolytic uremic syndrome in the case of E. coli O157:H7 infection, prolonged shedding of salmonellae of nonTyphi serotypes, and increased organism resistance. ■ Treatment should be considered in patients who are still significantly ill when culture results are reported, particularly if they are immunocompromised or have a
significant underlying medical illness, or in cases in which the fecal shedding could represent a public health hazard. ■ Patients with sickle cell anemia, other hemolytic anemias, or AIDS are unusually susceptible to Salmonella bacteremia. ■ Although risk associated with antimotility agents may be overstated, these agents are still not recommended for treatment of invasive enteritis unless antibiotics also are used. ■ In patients with Y. enterocolitica gastroenteritis, the abdominal pain and diarrhea usually persist for 10 to 14 days or longer. in a substantial number of patients with yersiniosis, in particular adolescents and young adults, an ileocecitis may develop. In these cases, lower abdominal pain with little or no diarrhea predominates and may perfectly mimic the pain of acute appendicitis. ■ The symptoms of scombroid fish poisoning, which resemble those in histamine intoxication and usually develop abruptly within 20 to 30 minutes of eating the fish, consist of facial flushing, diarrhea, severe, throbbing headache, palpitations, and abdominal cramps, and generally last less than 6 hours. The mainstay of therapy is antihistamine administration. ■ Many cases of C. difficile antibiotic–associated enterocolitis are self-limited, provided that the offending agent is discontinued. When stopping the antibiotic does not resolve the diarrhea, or when the diarrhea is severe, empirical antibiotic treatment should be started promptly. Either oral metronidazole or oral vancomycin can be used to eradicate C. difficile colitis. ■ Giardia is the most common cause of water-borne diarrheal outbreaks in the United States. Most patients harboring Giardia organisms are asymptomatic. The most common symptoms of acute infection are abdominal distention, colicky pain with audible borborygmi, flatulence, and frequent stools. All patients
1227
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
intervention to effect resolution. Therefore, each patient deserves a diagnostic evaluation. ■ ETEC is responsible for 40 to 50% of all cases of traveler’s diarrhea, and these organisms can be acquired anywhere in the world. Prophylactic antibiotics are no longer recommended for otherwise healthy patients, but quick resolution of diarrhea is possible with prompt treatment.
Chapter 92 / Gastroenteritis
with proven Giardia infection should receive treatment even if they are asymptomatic. ■ Diarrhea is the most common manifestation of gastrointestinal disease in patients with AIDS and may be either the presenting symptom or a life-threatening complication of the disease. Diarrhea in patients with AIDS in whom CD4+ counts are less than 300/mm3 generally is not self-limiting but requires medical
Chapter 93
Disorders of the Large Intestine
Michael A. Peterson
■ IRRITABLE BOWEL SYNDROME Perspective Irritable bowel syndrome (IBS) is a chronic non–lifethreatening disorder characterized by abdominal pain and alteration in bowel habits. IBS is an extremely common disorder; estimates put the prevalence in the North American population at 10 to 15%, with women affected twice as often as men.1 Although only one third of patients who have the clinical syndrome ever seek medical attention, IBS accounts for more than 10% of all visits to primary care physicians and more than 25% of all visits to gastroenterologists.2 IBS is said to contribute more impairment to quality of life than does either diabetes or renal failure. No specific physical or laboratory abnormalities defining IBS have been identified. IBS is a functional asomatic syndrome that is defined by criteria and usually is diagnosed after other, more serious diagnoses are excluded. A new diagnosis of IBS may be difficult to make in the emergency department (ED) because many of the studies required to exclude other conditions are not readily available. Initially, patients usually are discharged with a diagnosis of “abdominal pain of unclear etiology” or the equivalent. Such patients and even patients with known IBS presenting with acute symptoms pose a diagnostic challenge. Symptoms of IBS overlap with those of other conditions, including some that are life-threatening, which must be excluded before hospital discharge.
Principles of Disease Although the cause of IBS is unknown, it is associated with several pathophysiologic findings suggesting that it is a disorder of altered gut motility, gut sensation, and perception of intestinal activity. IBS initially was thought to be primarily a psychiatric disorder because there are no visible anatomic abnormalities, and because stress is an exacerbating factor. Physiologic testing has since shown that patients with IBS have disturbances in the rhythmic pattern of electrical activity in the intestine and in how the intestine responds to stimulation. Patients with IBS seem to be more attuned to activity in the abdomen as well, sensing different phases of intestinal motor activity and intestinal content movement more than people without IBS. Psychiatric conditions often coexist with IBS, ranging from generalized anxiety disorder to major depression. An association with previous sexual abuse also has been reported.2 In 1228
women, symptoms often are related to the menstrual cycle, suggesting a hormonal influence. A familial predisposition for the symptoms of IBS has been described, suggesting a genetic component.
Clinical Features The diagnosis of IBS is defined by clinical criteria in a patient who has no other organic explanation for the symptoms. Several sets of clinical criteria have been published, one set of which is the Rome II criteria (Box 93-1). Patients with IBS experience symptoms intermittently, with the typical patient averaging symptoms on 1 of every 3 days. Complaints include abdominal pain, bloating, and constipation or diarrhea. Pain typically is relieved with defecation; pain that persists suggests another diagnosis. A mucoid discharge from the rectum often accompanies diarrhea. Upper gastrointestinal symptoms such as nausea and dyspepsia also can occur. Patients may present to the ED with an exacerbation of their previous symptoms or with a new abdominal complaint and often report that they are undergoing a period of stress. Physical examination may reveal mild abdominal tenderness that may be focal, often varying in location, or diffuse. IBS is subdivided into two categories: constipation-predominant and diarrhea-predominant. Pain that is progressive, keeps the patient awake at night, or is associated with anorexia or significant abdominal tenderness suggests an alternate diagnosis. Fever, abdominal mass, and rectal bleeding also are atypical for IBS. In the absence of symptoms suggesting another diagnosis, the clinical criteria have a specificity ranging from 87 to 100%, although sensitivity may be only 60%.2 In patients determined to have IBS through correct use of the clinical criteria, follow-up evaluation over many years rarely leads to a change in the diagnosis.2
Diagnostic Strategies A final diagnosis of IBS usually is made in the primary care setting and not in the ED. A typical primary care evaluation for IBS may include a complete blood count, thyroid studies, stool exam for ova and parasites, evaluation for lactose intolerance, and possibly lower gastrointestinal endoscopy, although current approaches try to limit testing in cases in which the clinical picture is clearly IBS.3 The ED evaluation seeks to exclude other, more urgent causes for the patient’s symptoms. In this setting, testing for pancreatitis, hepatitis, biliary colic, or urologic disorders, including urolithiasis, may
1229
BOX 93-1 Rome II Criteria for Irritable Bowel Syndrome
1. Relief of discomfort with defecation 2. Association of discomfort with altered stool frequency 3. Association of discomfort with altered stool form From Brandt LJ, et al: Systematic review on the management of irritable bowel syndrome in North America. Am J Gastroenterol 97:S7, 2002.
Osmotic laxatives Antidiarrheals Antispasmodics (anticholinergics and calcium channel blockers) Tricyclic antidepressants Serotonin receptor antagonists (e.g., alosetron) Prokinetic agents Female hormone–modulating agents Tranquilizers
Differential Diagnosis in Irritable
BOX 93-2 Bowel Syndrome
Constipation-Predominant Bowel obstruction Cancer Adult-onset Hirschsprung’s disease Rectocele Paradoxical closure of the anus during defecation Diarrhea-Predominant Bacterial/parasitic intestinal infection Inflammatory bowel disease Lactose intolerance Malabsorption Radiation proctocolitis Painful Inflammatory bowel disease Ureteral colic Bowel obstruction Diverticular disease Gastroesophageal reflux/ulcer Liver or pancreatic disease Lead toxicity Porphyria
cramping, and peripherally acting narcotics, such as loperamide, are used to reduce diarrhea. Osmotic laxatives such as lactulose sometimes are helpful in constipation. Tricyclic antidepressants have been effective in certain classes of patients with IBS. Serotonin receptor antagonists, such as alosetron, and prokinetic agents also are used (Box 93-3). Nonsteroidal anti-inflammatory drugs (NSAIDs) may worsen symptoms.3 Behavioral therapy may benefit patients who are unresponsive to medication, but high-quality evidence of its effectiveness is lacking. A role for nontraditional therapies such as arrowroot, artichoke leaf, and some Chinese traditional herbal medicines is supported by limited scientific evidence.2
Disposition IBS is not a life-threatening disease and can be managed on an outpatient basis so long as other disorders have been excluded. The search for optimal therapy usually involves empirical trials until the correct fit is found. This process is best accomplished through a well-established primary care relationship. The disorder is chronic, but with appropriate therapy, many patients experience significant improvement in their quality of life.
be appropriate as indicated by the pattern of the presenting complaints.
■ DIVERTICULAR DISEASE
Differential Considerations
Diverticular disease is an affliction of middle age that seems to be a direct consequence of the diet of modern Western civilization. Diverticular disease was virtually unknown in the Western world before the 20th century and is still rare in other cultures. In 1925, only 9% of people older than 50 years of age in the United States had diverticula; by 1968, the percentage had increased to 30%.4 Today it is estimated that 5 to 10% of people older than 45 years and 80% of people older than 85 years have diverticula.5 Diverticula are less common in people younger than age 40, representing only approximately 2 to 5% of all patients with the disease. The proliferation of this disease was coincident with the invention and widespread use of the flour rolling mill, which removes the fiber-containing outer part of the wheat kernel. This coincidence has prompted the labeling of diverticulosis as a “modern deficiency disease,” which is supported by the fact that adding fiber back into the diet seems to be protective against the development of diverticulosis.4 In rural Africa and Asia, where the diet is high in fiber, diverticular disease is virtually unknown.5 Diverticulosis denotes the presence of diverticula in the colon. Most patients with this condition are asymptomatic. Diverticulitis denotes inflammation of diverticular tissue, which usually is painful. Complicated diverticulitis is defined by the
The differential diagnosis of symptomatic IBS depends on the predominant symptoms and includes a host of disorders (Box 93-2). Patients may present with pain, constipation, or diarrhea or any combination of the three.
Management Not all patients with IBS require treatment. It is recommended that therapy be initiated only if symptoms diminish the quality of life.1 Because no curative therapy is available, treatment is directed toward the relief of symptoms. Diet, behavioral, and pharmacologic therapies all are used in IBS. Specific therapy will be determined by the type of IBS: constipationpredominant, diarrhea-predominant, or constipation-plusdiarrhea combination. Dietary suggestions include a low-fat diet, reduced nondigestible sugars, and avoidance of gasforming foods, although none of these has any proven benefit. Fiber supplementation may aid constipation-predominant IBS. Medications with antispasmodic activity, such as anticholinergics and calcium channel blockers, are used for abdominal
Perspective
Chapter 93 / Disorders of the Large Intestine
Abdominal pain or discomfort for 12 weeks or longer within the past year and two of the following:
Medications Used in the Treatment of Irritable
BOX 93-3 Bowel Syndrome
1230
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
presence of more extensive disease, including abscess formation, peritonitis, intestinal obstruction, or fistula formation.
Principles of Disease The wall of the colon is penetrated at regular intervals by blood vessels, collectively known as the vasa recta, that supply the internal intestinal layers. The site of vessel penetration is apparently the weakest part of the colon wall, because it is at these sites that diverticula form. Although the exact pathogenic mechanism is unknown, the current theory is that diverticula form in response to increased intracolonic pressures generated when the colon is processing smaller, non–fibercontaining stools. Higher pressures lead to a herniation of colonic mucosa through the intestinal wall at the vasa recta, creating small, sac-like appendages. These appendages (diverticula) typically measure 5 to 10 mm in diameter but on rare occasions can grow into huge sacs measuring many centimeters across (giant colonic diverticula). Diverticula are asymptomatic in most affected persons, but symptoms are believed to develop when diverticula become obstructed, presumably with inspissated stool. When obstruction occurs, inflammation sets in and microperforations of the sac develop, resulting in inflammation of pericolonic structures and abdominal pain. In the Western world, 85% of diverticular disease occurs in the left colon, usually the sigmoid.5 This is not the case in Japan, where right-sided diverticular disease is more common. Japanese-Hawaiians consuming a low-fiber Western diet have a significantly increased incidence of diverticular disease, but it remains in the right colon.4 This finding suggests that diet plays a significant role in the formation of diverticula but that the location of diverticula is genetically determined. A spectrum of disease is observed in relation to diverticular inflammation. In uncomplicated diverticulitis, only the pericolonic fat is inflamed. With time, a phlegmon, abscess, or gross perforation may develop. Any extension of disease beyond the pericolonic fat is considered to represent complicated diverticulitis. The involved colonic segment may fistulize to any adjacent organ, most commonly the bladder (accounting for 65% of all fistulas).6 Adjacent bowel may become obstructed by mass effect from an abscess or may incur an inflammatory ileus. Recurrent episodes of diverticulitis can lead to strictures in the colon with subsequent colonic obstruction. Diverticula also can bleed, presumably from erosion into the mucosal wall by dried stool trapped in the diverticular sac. Severe hemorrhage occurs in 3 to 5% of all patients with diverticulosis and accounts for approximately 40% of all instances of lower gastrointestinal hemorrhage.4 Bleeding notably occurs in the absence of inflammation and typically is painless. NSAID use is known to be associated with this complication.6
Clinical Features Diverticulosis Commonly asymptomatic, patients with diverticulosis sometimes have nonspecific abdominal complaints including bloating, crampy pain, excessive gas, or a change in bowel habits.7 Diverticulitis will develop in approximately 10 to 30% of patients with diverticulosis; approximately 75% remain asymptomatic throughout their lifetime.4,8
Diverticulitis Because most diverticula found in persons in the Western world form in the left colon, the typical presentation of diverticulitis is persistent left lower quadrant pain and tenderness.
Pain may be felt first in the hypogastrium before localizing to the left lower quadrant. Referred pain may occur in the penis, scrotum, or suprapubic region. Right-sided diverticulitis may manifest as right lower quadrant pain and is impossible to distinguish clinically from appendicitis. Additional findings suggest various complications: Diffuse tenderness is associated with gross perforation or abscess rupture; dysuria, with a colovesical fistula; mass, with an abscess; and vomiting or abdominal distention, with intestinal obstruction. Fecal matter or gas emanating from the vagina suggests a colouterine fistula. Almost any adjacent organ can be involved in the inflammatory process. Patients recently diagnosed with diverticulitis who are being treated on an outpatient basis with oral antibiotics and who present to the ED with continuing or worsening symptoms should be evaluated for the possibility of an abscess. Special care must be taken with elderly or immunocompromised patients, because clinical signs and symptoms are much less dramatic, even with more severe disease. Perforation is more common in these patients, manifests with less significant findings, and carries a high mortality rate.5
Diagnostic Strategies Uncomplicated Diverticulitis The clinical diagnosis of uncomplicated diverticulitis can be made in a patient in the appropriate age range exhibiting focal left lower quadrant pain and tenderness in the absence of symptoms or signs that suggest an alternative diagnosis. No mass or peritoneal irritation should be encountered on examination, and the patient should otherwise appear well. If the patient fits this clinical picture, treatment can be initiated on an empirical basis, and no laboratory tests or diagnostic imaging is required. Ancillary tests are performed primarily to exclude alternative diagnoses or the presence of complicated diverticulitis. When the diagnosis is unclear, studies to exclude gynecologic, renal, hepatic, biliary, or pancreatic disease may be indicated, depending on the patient’s presentation and degree of distress. Computed tomography (CT) of the abdomen should be considered for elderly and immunocompromised patients to exclude the possibility of a subtle presentation of complicated diverticulitis.
Complicated Diverticulitis Abdominal Computed Tomography. Abdominal CT is the preferred method of evaluation in complicated diverticulitis. CT has the advantage of evaluating the colon and the structures around it, so it can make the diagnosis of diverticulitis and simultaneously evaluate the extent of disease. CT also can be used to guide percutaneous drainage of diverticular abscesses. Findings on CT consistent with diverticulitis include the presence of diverticula, inflammation of pericolonic fat, thickening of the bowel wall to more than 4 mm, free abdominal air, and an abscess.5,8,9 CT also can help make an alternative diagnosis when diverticulitis is absent. CT is relatively noninvasive and is well tolerated by ill patients. Sensitivity and specificity for diverticulitis range from 69 to 95% and from 75 to 100%, respectively.6 Negative findings on CT scan cannot absolutely exclude diverticulitis. Small abscesses within the colon or mesocolon can be missed, as can the diverticula themselves. It also may be difficult to differentiate between carcinoma and diverticulitis on CT scan. Marked bowel wall thickening associated with diverticulitis looks like cancer; contrast enema or endoscopy may be required to differentiate between the two. Barium Enema. Although double-contrast barium examination is the standard for the diagnosis of asymptomatic diverticula,
Differential Considerations In a patient who presents with more serious disease, making the correct diagnosis usually poses no great dilemma because more extensive laboratory testing and diagnostic imaging typically have been performed. In a patient with milder disease, in whom a tentative diagnosis of diverticulitis may be based on clinical grounds alone, the differential diagnosis presents more of a challenge. An important consideration is whether or not the patient has colonic carcinoma; however, it usually is safe to wait until after the acute episode has resolved to investigate this possibility. Additional diagnoses to consider include colitis (either inflammatory or ischemic); ureteral stones; inguinal hernia; and pelvic pathology, including ectopic pregnancy or pelvic inflammatory disease, and ovarian pathology with or without ovarian torsion. Appendicitis should be suspected when symptoms are predominantly right-sided. Diffuse abdominal pain should prompt an evaluation for other lifethreatening problems, including leaking abdominal aortic aneurysm, peritonitis, hemoperitoneum from ectopic pregnancy, and bowel obstruction.
Management Diverticulosis All patients diagnosed with diverticulosis should be placed on a high-fiber diet, which has been shown to reduce abdominal
1231
symptoms and recurrent bouts of diverticulitis.6 It is not known if the common advice to avoid foods that may obstruct diverticula, such as nuts and small seeds, is of any real merit.
Uncomplicated Diverticulitis Uncomplicated diverticulitis in an immunocompetent, nonelderly patient can be managed on an outpatient basis with oral antibiotics11 (Box 93-4). Coverage for gram-negative aerobic and anaerobic bacteria is required. Patients may be placed on a liquid diet for comfort, although this is not mandatory. NSAIDs or narcotics are appropriate for pain control, but many experts recommend avoiding morphine sulfate because it increases intraintestinal pressure, which theoretically can precipitate perforation.5 A high-fiber diet prevents recurrent diverticulitis for 5 years in 70% of patients. Patients with significant comorbid illness or other problems, including inability to tolerate oral liquids, poor social support, and inability to comply with follow-up in a reasonable time frame (2 to 3 days), should be considered for hospital admission. Hospitalized patients generally are treated with intravenous antibiotics (Box 93-5) and placed on bowel rest, although patients hospitalized for psychosocial reasons can be treated with oral medications.
Complicated Diverticulitis Patients with complicated diverticulitis should be admitted to the hospital and treated with intravenous antibiotics. Emergent surgical intervention is indicated for all patients with peritonitis or perforation. Newer techniques using a laparoscopic approach with lavage and biologic glue have supplanted open surgical techniques in some patients.12 Continuing clinical decline, sepsis resistant to medical management, or a high level
Oral Antibiotic Therapy for
BOX 93-4 Uncomplicated Diverticulitis ■ Trimethoprim-sulfamethoxazole, double-strength
tablets bid, and metronidazole, 500 mg q6h or ■ Ciprofloxacin, 750 mg bid, and metronidazole, 500 mg q6h or ■ Amoxicillin-clavulanate extended-release, 1000/62.5 mg, 2 tabs bid or All oral regimens should be taken for 7 to 10 days. From Gilbert DN, Moellering RC Jr, Eliopoulos GM, Sande MA (eds): The Sanford Guide to Antimicrobial Therapy, 37th ed. Sperryville, Va, Antimicrobial Therapy, Inc, 2007.
Intravenous Antibiotic Coverage for
BOX 93-5 Bowel Flora
Mild to Moderate Infection ■ Ticarcillin-clavulanate, 3.1 g IV q6h or ■ Ampicillin-sulbactam, 3 g IV q6h or ■ Ciprofloxacin, 400 mg IV q12h, and metronidazole, 1 g IV q12h Severe Infection ■ Ampicillin, 2 g IV q6h, and metronidazole, 500 mg IV q6h, and (gentamicin, 7 mg/kg q24h, or ciprofloxacin) 400 mg IV q12h or ■ Imipenem, 500 mg IV q6h From Gilbert DN, Moellering RC Jr, Eliopoulos GM, Sande MA (eds): The Sanford Guide to Antimicrobial Therapy, 37th ed. Sperryville, Va, Antimicrobial Therapy, Inc, 2007.
Chapter 93 / Disorders of the Large Intestine
it should be avoided in the setting of diverticulitis. The potential for preexisting occult perforation and subsequent risk of barium peritonitis limits its usefulness. Barium enema may be used after the acute episode to exclude the diagnosis of carcinoma. Water-Soluble Contrast Enema. A water-soluble contrast enema is the preferred method of imaging if a contrast enema study is needed in the acute setting. Water-soluble contrast material shows less detail than barium, but this modality is still useful. Findings consistent with diverticulitis include the presence of diverticula along with extravasation of contrast material into an abscess cavity or into the peritoneum. This study also can show a fistula or evidence of compression of the colon by an extrinsic mass. Because contrast material usually collects only in the intestinal lumen, contrast enemas give less information than CT about the extent of disease outside of the colon. Ultrasonography. Ultrasound examination can detect various pathologic features characteristic of diverticulitis, including fluid collections around the colon, thickened hypoechoic bowel wall, or hyperechoicity adjacent to the bowel wall that suggest pericolonic inflammation. Tenderness to palpation over an abnormal-appearing colon suggests that the colon is the source of the patient’s pain. Diverticula can occasionally be visualized by ultrasound exam. As is often the case, the sensitivity of ultrasound imaging for these findings varies significantly with the experience of the operator. Because gas interferes with ultrasound imaging, adequate visualization of the bowel can be a problem. Currently, the role for ultrasonography in the evaluation of diverticulitis is not well defined. Endoscopy. Endoscopy is limited in the acute setting by its more invasive nature, the risk of perforation, and the logistics of arranging this procedure emergently.10 Although the endoscope affords visualization of diverticula and other pathologic processes within the lumen of the colon, it does not permit evaluation of the extent of extracolonic disease. Plain Radiography. Plain films of the abdomen are not likely to be helpful in the diagnosis of diverticulitis unless either intestinal obstruction or perforation is suspected.
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1232
of suspicion for carcinoma warrants urgent surgical consultation.8 Small abscesses (less than 5 cm in diameter) are treated with intravenous antibiotics alone (see Box 93-5), whereas larger abscesses are drained either percutaneously with imaging guidance or surgically.13 Bowel obstruction during an attack of diverticulitis usually is self-limited and resolves with conservative management. Chronic recurrent diverticulitis can result in stricture, which necessitates surgical intervention. Fistulas usually are repaired surgically. A significantly dilated cecum (to greater than 10 cm in diameter) or gas in the bowel wall should prompt early consultation with a surgeon about the possibility of bowel necrosis and impending perforation.5
because it frequently is associated with malignant disease. One half of all operative cases involving LBO in the United States are the result of colorectal cancer. Adhesions, a common cause of small bowel obstruction, cause only 1 to 8% of LBOs.15 Other causes of LBO include volvulus, diverticular disease, fecal impaction, strictures (often related to inflammatory bowel disease or chronic colon ischemia), adhesions, hernia, and pseudo-obstruction. Most causes are managed surgically, but pseudo-obstruction responds well to medical management alone.
Definitive Management
When mechanical obstruction is secondary to an obstructing lesion, either inside the bowel (carcinoma) or outside the bowel (diverticular abscess, volvulus), the bowel becomes increasingly dilated with air, and fluid that cannot be passed distally. As the distention increases, the intraluminal pressure increases. When intraluminal pressure approaches systolic blood pressure, blood flow to the bowel wall is compromised and edema sets in, with subsequent transudation of fluid into the lumen. Transudation along with decreased resorption of intraluminal fluid leads to dehydration. Eventually, as arterial flow to the bowel wall is compromised, ischemia and gangrene develop. Translocation of bacteria from compromised bowel can lead to sepsis. Perforation of the bowel wall follows if the process is not interrupted.15 Pseudo-obstruction, also called Ogilvie’s syndrome, occurs through a completely different mechanism. Pseudoobstruction is defined as LBO in which no obstructing lesion can be identified. This condition usually is found in patients with significant acute comorbid conditions.16 Patients typically have a history of significant spine or retroperitoneal trauma, severe electrolyte disturbances, or narcotic exposure. Although the exact mechanism is unknown, it is believed to involve malfunction of autonomic control of the bowel. Normal balance between parasympathetic and sympathetic input is disrupted, resulting in changes in motility that lead to obstruction. The pathophysiologic changes observed with pseudoobstruction are the same as those described for mechanical obstruction.
It is not known whether medical or dietary treatment is of benefit in diverticulitis. The only proven way to eradicate diverticula is to remove the affected segment of colon surgically. Most patients who recover from their first attack of diverticulitis are likely to remain asymptomatic for many years. With subsequent attacks, the likelihood of recurrence increases. Elective resection of diverticula generally is reserved for patients who have had more than one attack of diverticulitis. According to some experts, younger patients (i.e., younger than 40 years of age) should undergo elective resection after their first bout of diverticulitis because of concerns about a higher risk for a second attack, but this recommendation is controversial, with recent evidence suggesting that outcomes are better without surgery.8,14 Most resections can be done laparoscopically with a single-stage procedure (no colostomy).14 Estimates on recurrence of diverticular disease after resection vary, ranging from 3 to 27%.5-7,14
Disposition Uncomplicated Diverticulitis Nonelderly and immunocompetent patients may be sent home on oral antibiotics with referral for follow-up evaluation in 2 to 3 days to determine the success of treatment. Patients are cautioned to return to the ED if their condition worsens. Patients not significantly improved at follow-up should undergo diagnostic imaging to look for an abscess and be hospitalized for intravenous antibiotic therapy. Of patients treated medically for their first attack of diverticulitis, 95% remain symptom-free for the next 2 years, and 80 to 90% remain symptom-free permanently.5,14 Patients with recurrent episodes of diverticulitis should be referred to a surgeon for outpatient consultation for elective resection. All patients should undergo an evaluation for colon cancer when the acute episode has resolved, because the incidence of coexistent cancer has been reported to be as high as 9%.
Complicated Diverticulitis All patients require hospitalization for intravenous antibiotic therapy and bowel rest. Most patients (65 to 85%) recover with medical management alone; the rest require surgical intervention. Outcomes generally are good, with mortality rates ranging from 1 to 6% for all patients, increasing to 12 to 18% for patients requiring surgery.4
Principles of Disease
Clinical Features The typical presenting complaints in LBO are abdominal pain, abdominal distention, obstipation, and vomiting. The time frame within which these symptoms develop varies in accordance with the rapidity of onset of the obstruction. LBO associated with a volvulus can develop rapidly, whereas obstruction from cancer tends to be of gradual onset. Patients presenting later in the course of obstruction may be significantly dehydrated. Significant fever or tachycardia should prompt an investigation for gangrene and perforation. A palpable abdominal mass may represent a tumor, an abscess, or simply distended bowel. A rectal examination is helpful to look for an obstructing rectal mass or a large volume of hard stool in the rectal vault consistent with fecal impaction.
Diagnostic Strategies ■ LARGE BOWEL OBSTRUCTION Perspective Large bowel obstruction (LBO) is much less common than small bowel obstruction, but LBO is a more ominous condition
Electrolyte measurements may be helpful in guiding fluid and electrolyte replacement therapy. A significantly elevated white blood cell (WBC) count should raise suspicion for gangrenous bowel, whereas anemia suggests the possibility of colorectal cancer.
1233
Chapter 93 / Disorders of the Large Intestine
A
B
Figure 93-1. Plain radiographs showing large bowel obstruction at the sigmoid colon caused by carcinoma. A, Erect view. B, Supine view.
Plain Radiography. A distended colon is the hallmark of LBO (Fig. 93-1), although small bowel may be distended as well if the ileocecal valve is incompetent. In some cases, gas-filled small bowel may obscure visualization of the colon, leading to the misdiagnosis of small bowel obstruction. An abrupt cutoff at the distal end of the obstructed colonic segment suggests a possible pseudo-obstruction. A cecal diameter exceeding 12 cm is of concern because this finding is associated with a higher risk of perforation.16 The actual location and cause of the LBO usually are not evident on plain films. Computed Tomography. CT is a valuable tool for determining the cause of the obstruction, especially if the cause is a diverticular abscess or intussusception.15 CT typically is less helpful in pseudo-obstruction, in which either colonoscopy or a watersoluble contrast enema study is needed to make the diagnosis. Colonoscopy and Water-Soluble Contrast Enema Patients in whom the cause of obstruction is not known and who are not candidates for urgent surgical intervention should undergo either a water-soluble contrast enema study or colonoscopy to determine the etiology of the obstruction. This diagnostic strategy is much more accurate in ruling out pseudo-obstruction than imaging.
Differential Considerations The most common causes of LBO are colorectal cancer (53%), volvulus (17%), diverticulitis (12%), and compression from metastatic disease (6%). Other, less common causes are strictures, incarcerated hernia, fecal impaction, adhesions, and pseudo-obstruction.
Management Management in the ED is directed at relief of symptoms. Rehydration, electrolyte replacement, and pain management are the first concerns. Gastric decompression with a nasogastric tube may be helpful in cases in which vomiting is prominent or when there is evidence of significant fluid or gas buildup in the small intestine. No additional fluid or solids should be administered by mouth. Antibiotics are indicated if gangrene or perforation is suspected (see Box 93-5). Definitive management depends on the cause of the obstruction, which may or may not be determined in the ED. Select diverticular abscesses may be drained percutaneously, whereas a sigmoid volvulus or pseudo-obstruction can be decompressed endoscopically. Diverticular disease and sigmoid volvulus eventually necessitate an elective surgical procedure to prevent recurrence, although this often can be delayed. Carcinoma, cecal volvulus, strictures, intussusception, adhesions, and hernias are dealt with primarily surgically. So long as the possibility of perforation is not an immediate concern, pseudo-obstruction is managed for the first 24 hours with bowel rest, hydration, and management of any acute comorbid conditions. If the colon fails to decompress, colonoscopic or pharmacologic intervention (neostigmine)16 may be attempted, with surgery reserved for refractory cases. Fecal impaction generally is managed definitively in the ED through digital disimpaction or an enema. Particularly helpful are retention enemas, for which the patient retains the enema fluid in the rectum for 15 minutes or longer. Occasionally disimpaction is technically difficult enough to warrant general anesthesia.
1234
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Disposition Most cases of LBO require procedural intervention (surgical, endoscopic, or percutaneous abscess drainage) to achieve resolution. All patients require hospitalization and consultation with a specialist capable of performing the appropriate procedure. Emergent surgical consultation is warranted in patients with evidence of gangrenous bowel or perforation.
■ VOLVULUS Perspective Volvulus of the colon occurs when a loop of bowel twists and obstructs the intestinal lumen. If severe enough, the twist may involve and compromise the vascular supply to the loop of colon. The incidence of colonic volvulus is 2.65 cases per 100,000 population per year, accounting for 1 to 7% of all LBOs.17 Volvulus occurs in all age groups, but older adults are affected most often, with a mean age of 60 to 70 years. One third of cases in the developed world involve institutionalized patients.17 Most cases are divided roughly equally between the sigmoid colon and the cecum, although volvulus can occur in all other areas of the colon. Sigmoid volvulus typically is a disease of older persons. Mortality rates with sigmoid volvulus exceed 50% in patients who present with gangrenous bowel. In the absence of gangrenous bowel, the risk of death is approximately 10%.
Principles of Disease Sigmoid Volvulus The anatomic requirement for a sigmoid volvulus is a long, redundant section of sigmoid that is attached to the abdominal wall by a narrow strip of mesentery. The narrow attachment allows the mesentery to twist on itself, thereby obstructing the intestinal lumen. It is not clear whether this is a congenital condition or occurs as part of the aging process. After the colon twists on itself, the proximal colon continues to force gas and liquid into the obstructed segment, causing a sometimes massive dilation of the distal colon. Significant electrolyte disturbances can occur secondary to third spacing, and respiratory compromise occasionally occurs from massive abdominal distention. If the condition is left untreated, the vascular supply can become compromised, resulting in gangrene and perforation. The exact precipitator of an acute episode of volvulus is not clear. A high-fiber diet has been implicated, because a significant increase in the disease is noted in patients who are switched to a high-fiber diet. Chronic constipation has been associated with volvulus, but it is unclear how the two conditions are related. Residents of long-term care facilities and patients with neurologic or psychiatric diseases also are predisposed to sigmoid volvulus, possibly as a result of alterations in colonic motility. No association with previous surgery has been observed. Women seem to be at a higher risk for cecal volvulus during pregnancy, presumably because of crowding of the abdominal cavity by the enlarged uterus. The condition is still rare, however, occurring in approximately 1 per 1 million pregnancies.18
Cecal Volvulus As in sigmoid volvulus, a mobile segment of cecum is a prerequisite to the disease. This mobility seems to be due to a congenitally incomplete fusion of the cecal mesentery to the posterior abdominal wall. Cadaver studies show that 10% of
the adult population have ceca that are mobile enough to cause torsion.17 In 10% of the cases, cecal volvulus is due to a variant called cecal bascule, in which the cecum does not twist but merely folds over on itself; symptoms and management are the same.15 The tendency for cecal volvulus may be related to “maneuvering room” available for the colon within the abdomen. Persons with less space in the abdomen for the colon to move about seem to be more predisposed to volvulus in general. Gangrene of the bowel is common and occurs in 20% of patients with cecal volvulus.19
Clinical Features Sigmoid Volvulus The hallmark of sigmoid volvulus is the triad of abdominal pain, distention, and constipation. The extent to which the sigmoid colon can twist on itself is recognized to vary, so the presentation of sigmoid volvulus will vary accordingly, from subtle to dramatic. The clinical picture may range from one of minor abdominal discomfort that has been present for many days to an acute onset of severe abdominal pain associated with gross abdominal distention and unstable vital signs. Sometimes the diagnosis of sigmoid volvulus is not made until the patient has been hospitalized for some time. In many instances, the history may be suggestive of previous episodes of volvulus that self-reduced. The physical examination may reveal a distended tympanitic abdomen, often with most of the distention in the upper abdomen but primarily on one side. Patients may look remarkably well for the amount of distention that is encountered. Significant abdominal pain, fever, lack of bowel sounds, peritonitis, or cardiovascular instability suggests gangrenous bowel and should prompt immediate surgical consultation. The absence of these findings does not exclude gangrene, however. The duration of symptoms alone is not predictive of gangrene of the bowel.
Cecal Volvulus The clinical triad of abdominal pain, distention, and constipation seen in sigmoid volvulus also is seen in cecal volvulus, but many patients with cecal volvulus lack one or more of these findings. Vomiting is seen in only approximately 50% of patients.
Diagnostic Strategies Sigmoid Volvulus The diagnosis of sigmoid volvulus can be made on the basis of findings on plain radiographs in most cases. A grossly distended loop of colon lacking haustral markings is typical and is seen just as often on the right side of the abdomen as on the left (Fig. 93-2). The bowel may have the appearance of a “bent inner tube.” Free air may be seen on an upright chest film or lateral decubitus radiograph of the abdomen in patients who have a perforation. Gas backing up into the rest of the colon may obscure the typical appearance of sigmoid volvulus on plain radiographs, leading to a significant number of nondiagnostic studies. Cecal volvulus and bowel obstruction from other causes may have a similar radiographic appearance. When the diagnosis is in doubt, contrast enema may be helpful. Contrast material fills up the colon to the tapering point of torsion, giving a “bird’s beak” appearance to the column of contrast material (Fig. 93-3). Sigmoidoscopy is diagnostic in many cases, visualizing a spiral sphincter–like twist
1235
Chapter 93 / Disorders of the Large Intestine
Figure 93-2. Plain film of abdomen shows a large dilated loop characteristic of sigmoid volvulus.
Figure 93-4. Plain film shows distended colon characteristic of cecal
volvulus. Note presentation in the left lower quadrant and absence of right-sided gas shadows.
cases. The cecum should be markedly dilated and may contain an air-fluid level. The small bowel often is distended as well. In contrast with the picture in sigmoid volvulus, the distal colon should have a paucity of gas (Fig. 93-4). The classic “coffee bean” sign, a large oval gas shadow with a line down the middle representing bowel bent over on itself, may be seen in the midabdomen. Free air suggests perforation and requires emergent surgical consultation. A common mistake is misinterpreting the plain radiograph as showing a sigmoid volvulus. If the diagnosis is unclear, a contrast enema is helpful in showing the site of torsion. Ultrasound imaging generally is unhelpful.18 On CT, a mesocolon “whirl sign” may be seen, indicating a twisted segment of mesentery.21 In many cases, cecal volvulus is definitively diagnosed only at surgery.19
Differential Considerations Any process that causes LBO may mimic volvulus. Figure 93-3. Characteristic “bird’s beak” sign of volvulus (arrowhead)
shown on this image from a barium enema study.
Management Sigmoid Volvulus
in the colonic mucosa.20 CT scan, when used, also is highly accurate, but most diagnoses can be made without it.20
Cecal Volvulus Plain radiographs often are helpful in establishing a diagnosis of cecal volvulus, but the findings are not definitive in 50% of
Although spontaneous reduction of a sigmoid volvulus can occur, it is infrequent enough to mandate a proactive approach to treatment. If clinical evidence of gangrenous bowel is lacking, endoscopic detorsion should be attempted by an experienced operator. Using the endoscope, the bowel is first examined for any signs of gangrene. If the bowel is healthy,
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1236
the twisted, obstructed proximal end of the bowel lumen is identified, and a lubricated rubber tube is inserted through the obstruction. With decompression of gas and liquid stool, the bowel is able to undergo self-detorsion. Endoscopic decompression is successful in 50 to 90% of cases.15,20 If the patient has gangrenous bowel or the volvulus does not respond to endoscopic decompression, surgery is indicated. Recurrence rates are estimated at 60%; elective resection of the redundant sigmoid is recommended after resolution of the acute episode. The mortality rate for sigmoid volvulus is 20% overall and exceeds 50% in the subpopulation of patients with gangrene.
Cecal Volvulus The proximal nature of the cecum makes it unavailable for endoscopic manipulation, so detorsion is done surgically. After detorsion, the cecum typically is fixed to the abdominal wall, or the redundant section is resected.21 Recurrence is rare after resection.
Disposition All patients with volvulus require hospitalization for detorsion and surgical intervention to prevent recurrence.
■ INTUSSUSCEPTION Perspective Intussusception in adults is rare, accounting for only 1 to 5% of cases of adult bowel obstruction. Most adult intussusceptions (80%) are of the small bowel. Although only 10% of children have a pathologic lesion as the cause of the intussusception, 90% of adults do. In the colon, these lesions are malignant 50 to 80% of the time, as opposed to the small bowel, in which malignant lesions are present approximately one third of the time.22 In adults, the intussusception often is unsuspected before being revealed on a CT scan or during laparotomy. The condition occurs over a wide variety of ages, with a mean age at presentation of 65 years.23
Principles of Disease The exact mechanism of intussusception is unknown, but it is believed that a lesion (the “lead point”) changes the motility properties of the intestine, allowing a proximal segment to invaginate into a more distal segment. As peristaltic activity pushes the invaginated segment along with its mesentery and mesenteric blood vessels down the bowel, the blood supply to the segment can be compromised, and ischemia may occur. Edema associated with the intussusception can lead to a mechanical obstruction of the bowel.
Clinical Features Intussusception in adults manifests in one of two patterns. The most common is that of acute partial intestinal obstruction. Less than 20% of intussusceptions cause complete obstruction.24 With this pattern, the typical presenting complaint is abdominal pain. Vomiting, bleeding, and constipation may be present but often are not. The abdomen may be distended, and bowel sounds often are decreased. A mass is seldom palpated; the classic triad of abdominal pain, mass, and heme-positive stools noted in children is rarely found in adults. The second presentation is much more subtle, with intermittent abdominal pain for months or years. The diagnosis usually is made only when the pain becomes unrelenting or has been recurrent enough to prompt imaging.
Diagnostic Strategies Plain Radiography. Plain radiography is a reasonable screening test in a patient suspected of having bowel obstruction, but the radiographs usually show only nonspecific large bowel dilatation. Computed Tomography. Typically used in the evaluation of abdominal pain and bowel obstruction, CT usually is the most useful test in suspected intussusception but may not detect the actual intussusception in as many as one half of the cases.22,23 Ultrasound Examination. Ultrasonography also is helpful in detecting intussusception but is not as useful as CT in excluding other diagnoses. A transverse view of the intussusception has a donut or target shape, with multiple concentric rings. A longitudinal view of the intussuscepted segment has an ultrasound appearance similar to that of a kidney (“pseudo-kidney sign”), with a bright central area surrounded by a darker outer layer. Barium Enema. Although a barium enema study can demonstrate intussusception and even reduce it, it is a less desirable study than either CT or ultrasound examination for initial diagnosis. In contrast with that in children, reduction of intussusception in adults is not desired before surgery because of concerns about spreading malignant cells from potentially malignant lead points. Barium enema should not be performed in patients suspected of having a bowel perforation. Colonoscopy. Colonoscopy is helpful in defining the lesion causing intussusception but does not usually detect the intussusception itself.
Differential Considerations The differential diagnosis includes other causes of bowel obstruction.
Management Surgery is required in most cases. ED care is supportive and aimed at optimizing fluid status, recognizing gangrene or perforation, administering antibiotics if compromised bowel is suspected, and securing surgical consultation in the appropriate time frame. Because of the high incidence of malignancy, reduction often is not attempted in adults before surgical exploration.23 Occasionally, intussusception may resolve spontaneously, but an evaluation to exclude a pathologic lead point still must be undertaken.
Disposition Because of the surgical nature of this disease, all patients require hospitalization. Operative mortality tends to be minimal.22
■ INFLAMMATORY BOWEL DISEASE Perspective Inflammatory bowel disease (IBD) includes two clinically similar but distinct diseases: Crohn’s disease (CD) and ulcerative colitis (UC). Both diseases are characterized by chronic and unpredictable relapsing inflammation of the gastrointestinal tract from causes that have not been definitively identified. Significant morbidity occurs from acute exacerbations of inflammation. It is estimated that more than 1 million people in the United States are affected by IBD.25 Cases are divided approximately equally between CD and UC, with a combined
Principles of Disease Ulcerative Colitis UC causes inflammation and ulceration throughout the colon and rectum, but spares the small intestine. Inflammation is more superficial than that found in CD. Typically, the inflammation exists as one continuous lesion originating in the rectum and extending a variable distance into the colon, although more recently, cases of discontinuous disease (“skip lesions”) similar to that in CD have been reported in UC.28 The concordance rate between identical twins is low (6 to 14%), suggesting that factors other than genetics are involved in the development of UC.29 Stress can trigger exacerbations, and cigarette smoking has a protective effect, suggesting environmental factors at work. Appendectomy at an early age is protective, suggesting that the immune system may play a role. In animal models of IBD, the disease does not occur in animals that are devoid of normal bowel flora, suggesting that bowel flora are a necessary ingredient for disease.30 One unifying theory is that UC represents a genetic predisposition to development of an inflammatory reaction to normal intestinal flora—in essence, losing the normal tolerance to these bacteria.
Crohn’s Disease The cause of CD is unknown, but genetic, environmental, immunologic, and infectious processes all have been implicated as possible causative or contributory factors.26 Concordance between identical twins is 45 to 50%, suggesting a strong genetic predisposition that is modified by other factors.29 Africans have a low incidence of CD, but African Americans have an incidence similar to that in white Americans.26 The first genetic mutation associated with CD was described in 2001 and is associated with 10 to 20% of cases of CD.30 An association between a strain of Mycobacterium and CD has been hypothesized, with the evidence being convincing enough to spur legislation in the United Kingdom to eradicate this potential pathogen from the food chain, but recent failures in a randomized controlled trial of antituberculosis medications in CD cast serious doubts on any causal relationship.31,32 Although the onset of the disease can be at any time of life, CD affects primarily young patients, with onset of disease typically in the teens and 20s. Inflammation in CD is deep, involving the entire colonic wall. The disease is not limited to the colon and rectum as it is in UC but may affect any part of the gastrointestinal tract. CD most often involves the distal small intestine and colon and less commonly the esophagus, duodenum, or stomach.33 Because of the transmural nature of the inflammation, the development of intestinal strictures or fistulas to adjacent organs is a potential complication.
Clinical Features Typical presenting complaints in patients with IBD include abdominal pain, often crampy, and tenesmus with loose or
1237
diarrheal stools. Blood may be present in the stool. Patients with CD may have a history of nocturnal diarrhea, a complaint that helps differentiate CD from patients who have IBS. Weight loss is common. The physical examination may reveal significant abdominal tenderness or an abdominal mass representing an abscess. Patients with CD may have fissures, ulcerated hemorrhoids, strictures, or cutaneous abscesses around the anus. Extraintestinal manifestations include inflammatory conditions of the skin, eyes, and joints. In children, growth and sexual development may be affected. Onset of symptoms usually occurs before the age of 30 years,25 although the diagnosis can be difficult to make in the early stages. Patients often present to the ED with a known diagnosis of IBD and worsening abdominal symptoms. A common reason for relapse is interruption of the medications that have kept the disease in remission. Many patients become complacent during quiescent periods and stop taking such medications. IBD requires continuous, lifelong maintenance therapy. Adherence to therapy has been shown to reduce the risk of acute attacks and cancer.25 Common complications of IBD include formation of fistulas, strictures, and abscesses; less common but life-threatening complications include fulminant colitis, toxic megacolon, and intestinal perforation.
Toxic Megacolon Toxic megacolon is a pathologic dilatation of the colon resulting from inflammation of the smooth muscle layers of the intestine. Muscle inflammation leads to paralysis, dilation, and eventually perforation if left untreated. The hallmark of toxic megacolon is colonic dilatation in a patient with a known inflammatory condition of the colon who appears systemically toxic. Presence of inflammation and toxicity differentiates toxic megacolon from other disorders that cause colon dilatation, including mechanical obstruction, pseudo-obstruction, and congenital or acquired megacolon. Toxic megacolon typically is associated with IBD or infectious colitis. The triggering event may be recent ingestion of anticholinergics, antimotility agents, narcotics, or antidepressants. Patients usually have experienced symptoms of colitis, which often are severe, for several days before the onset of toxic megacolon. Abdominal pain, fever, tachycardia, and abdominal distention are present. Plain radiographs are diagnostic and show a colon with a diameter of 6 cm or greater, although this feature may not be present in early stages. Treatment includes aggressive fluid hydration, intravenous corticosteroids, antibiotics covering bowel flora (see Box 93-5), and an evaluation for potential intestinal infections, especially in immunocompromised patients. The mortality rate has decreased over the past 4 decades to less than 2% as a result of the early recognition and aggressive treatment.
Diagnostic Strategies No specific laboratory tests available to diagnose IBD are available, although recent tests targeting antibodies to Saccharomyces cerevisiae or antineutrophil cytoplasm help to differentiate between CD and UC.27 Laboratory abnormalities may be due to a variety of reasons. Electrolyte abnormalities may be secondary to significant diarrhea, or anemia may occur from bloody stools. The erythrocyte sedimentation rate can be elevated and useful for categorizing the severity of the disease. Stools contain fecal leukocytes, but findings on stool cultures and ova and parasite examinations should be normal. Plain radiography is not helpful in the diagnosis of uncomplicated disease but may show bowel obstruction, toxic megacolon, or free air from a perforation (Fig. 93-5). The use of plain films should be limited to patients suspected of having
Chapter 93 / Disorders of the Large Intestine
annual incidence of approximately 10 cases per 100,000.26 The long-term management of IBD is a complex, stepwise process that involves multiple medications and surgery. The goals of the ED evaluation are to (1) recognize potential new cases of IBD, (2) consider and exclude serious complications in patients with IBD, and (3) identify those patients with IBD who need in-hospital care. Treatment plans are best developed in consultation with a physician experienced in the long-term management of IBD. Although life expectancy is slightly decreased for patients with CD, it is normal for patients with UC.27
1238
Medications Used in the Treatment of
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
BOX 93-6 Inflammatory Bowel Disease 5-Aminosalicylic Acid Agents Sulfasalazine Mesalamine Antibiotics Metronidazole Ciprofloxacin Rifaximin Tobramycin Corticosteroids Prednisone Hydrocortisone Methylprednisolone Budesonide Antimetabolites Azathioprine 6-Mercaptopurine Methotrexate Immunosuppressants Cyclosporine Anti–Tumor Necrosis Factor Antibodies Infliximab Figure 93-5. Toxic megacolon secondary to ulcerative colitis. The smooth
indentations seen along the margin of the colon represent pseudopolyps.
Disease Severity Criteria in Inflammatory
these complications. Contrast studies can reveal lesions suggestive of IBD, including ulcerations of the mucosal surface, fistulas, and strictures. In Europe, where ultrasound technicians are more experienced in its application, ultrasonography is used to identify active disease and to look for complications.28 Ultrasound imaging is used much less commonly in the United States. Magnetic resonance imaging can locate affected bowel segments and identify fistulas, stenoses, and abscesses. CT is the best study routinely available to evaluate extraluminal complications. CT colonography (“virtual colonoscopy”), although good at identifying cancerous lesions of the colon, does not show the typical lesions of IBD.28 Endoscopic evaluation with biopsy usually is required to confirm the diagnosis.27
Differential Considerations Symptoms and signs are protean and overlap with those of many common abdominal conditions, including appendicitis, infectious colitis, ischemic colitis, radiation colitis, diverticular disease, cancer, and bowel obstruction.
Management Medical management is the mainstay of therapy for most patients with IBD. In general, patients are maintained on 5-aminosalicylic acid (5-ASA) agents while asymptomatic and then steroids are added once symptoms recur. Once remission is obtained, steroids are discontinued and the patient is once more maintained on 5-ASA agents. If remission is not obtained with steroids, other agents such as antimetabolites and immunosuppressants are used (Box 93-6). The choice of agents depends on classification of the disease as either mild to moderate or severe (Box 93-7). Surgery is reserved
BOX 93-7 Bowel Disease
Ulcerative Colitis Mild Disease ■ >4 stools per day ■ Stools may contain some blood ■ No systemic signs of toxicity (fever, tachycardia, anemia, elevated erythrocyte sedimentation rate) Moderate Disease ■ >4 stools per day ■ Minimal signs of toxicity Severe Disease ■ >6 bloody stools per day ■ Signs of systemic toxicity Crohn’s Disease Mild ■ Patient ambulatory and able to eat ■ No toxicity ■ No significant abdominal pain or mass Moderate ■ Mild disease that has failed to respond to treatment ■ Patient may have some systemic toxicity Severe ■ Persistence of symptoms during corticosteroid therapy ■ High fever, persistent vomiting ■ Intestinal obstruction ■ Rebound tenderness ■ Cachexia ■ Abscess Adapted from Hanauer SB, Present DH: The state of the art in the management of inflammatory bowel disease. Rev Gastroenterol Disord 3:81, 2003.
Disposition Consultation with a gastroenterologist is recommended before patient disposition. Most patients with an uncomplicated mild to moderate exacerbation of IBD need only to restart their maintenance therapy if it was interrupted, or to add oral corticosteroids to their regimen. Patients with severe disease or those in whom oral corticosteroids have failed to effect improvement need hospitalization for administration of parenteral corticosteroids.27 Bowel rest does not seem to be beneficial except as preparation for surgical intervention.37 Emergent surgical consultation should be sought for life-threatening hemorrhage, evidence of perforation, or toxic megacolon. Urgent surgical intervention is indicated if the bowel is obstructed. Abscesses may be treated percutaneously with imaging guidance or surgically.33 Chronic fistulas initially are treated medically.35 After hospital discharge, close follow-up
1239
by the physician monitoring the patient’s disease is indicated to ensure that remission is achieved in a timely fashion and that the patient complies with the suppressive therapy after the acute event. For patients in remission, endoscopic monitoring for cancer is required on an ongoing basis, although the optimal frequency of examination has not yet been defined.35 The estimated prevalence of cancer among patients with CD is significant at 2%,33 and patients with UC have a 15 times greater risk for the development of colorectal cancer than in the general population.29
■ COLONIC ISCHEMIA Perspective Colonic ischemia is the most common of the intestinal ischemic disorders. Estimates place the incidence of colonic ischemia at 1 of every 2000 hospitalizations. Its presentation overlaps with many other significant abdominal diseases, and it is difficult to diagnose without endoscopic visualization of the colonic mucosa. Although elderly persons are most at risk, with 90% of cases in those older than 60 years of age, the condition can occur in all age groups.38 Both sexes are equally affected. In one study, more than 50% of persons admitted with colonic ischemia initially were diagnosed with IBD. Because there is no specific treatment and outcomes usually are good, the difficulty in making the diagnosis does not cause significant morbidity.
Principles of Disease The exact cause of colonic ischemia is unknown. Isolated ischemia without small bowel involvement usually is due to nonocclusive microvascular disease of the colon and not to large vessel (mesenteric artery) occlusion. The primary insult is a low-blood-flow state associated with a variety of factors including congestive heart failure, vasoactive drugs, atherosclerosis, renal failure, and recent cardiac or vascular surgery.38-40 Younger patients may suffer colonic ischemia in the setting of collagen vascular disease, hematologic disorders, long-distance running, or cocaine abuse. The colonic vascular system generally enjoys significant collateral flow, but in some patients this protective mechanism is tenuous, predisposing the affected person to ischemia from low-flow events.38 In addition, colonic arterioles seem to be particularly sensitive to vasoconstrictive influences, and the rapid-growing intestinal mucosa is especially vulnerable to interruptions in blood flow. High intraluminal pressures that normally develop within the colon also can alter intestinal perfusion significantly. Colonic ischemia can occur in any part of the colon, including the rectum, but for unknown reasons it occurs most often in the left colonic segment. Colonic ischemia represents a spectrum of disease whose manifestations vary with the extent of the ischemic insult. In most cases, the ischemic episode is self-limited and the condition resolves completely with conservative therapy, but in one third of patients, a prolonged or severe insult results in scarring or stricturing of the colon and chronic symptoms.35 If the ischemia is transmural, gangrene and intestinal perforation are possibilities. Chronic mild inflammation results in intermittent symptoms similar to those of IBD.
Clinical Features The presentation of colonic ischemia may vary but typically involves the acute onset of mild crampy abdominal pain in the left lower quadrant with abdominal distention and almost
Chapter 93 / Disorders of the Large Intestine
for patients with severe disease who do not respond to medical therapy or for complications such as obstruction or fistula formation. 5-ASA agents are the first line of therapy for disease that is not severe. These agents can be administered orally, or rectally if the disease is in or near the rectum. Sulfasalazine, one of the original drugs in this category, is limited by sulfa toxicity at higher doses and has serious side effects including bone marrow suppression. A newer 5-ASA derivative, mesalamine, has less toxicity, allowing higher dosages. When IBD goes into remission after an acute flare, the patient may be continued on 5-ASA derivatives for maintenance therapy. Antibiotics may be used for the primary treatment of IBD as well, but their use in IBD is controversial. Evidence supporting antibiotic use is stronger in CD than in UC. Metronidazole and ciprofloxacin are the most common antibiotics used, with some evidence suggesting that tobramycin or rifaximin may be beneficial as well.34 Oral corticosteroids are used in patients with moderate to severe disease or patients whose IBD is unresponsive to a 5-ASA agent. Steroids should be tapered when remission is achieved to avoid typical steroid side effects. Intravenous corticosteroids are reserved for hospitalized patients with severe disease. Budesonide, a newer oral corticosteroid, is degraded on its first pass through the bloodstream and has fewer systemic side effects. The immunomodulating drugs azathioprine and 6mercaptopurine are used in patients resistant to other therapies or to wean steroid-dependent patients off steroids. Patients on these medications should be assessed for bone marrow suppression and pancreatitis.35 The immunosuppressant agent cyclosporine is used in severe cases, often when patients are not surgical candidates. Although most patients tolerate it well, cyclosporine has significant potential toxicity, including myelosuppression, electrolyte disturbances, and hepatic and nephrotoxicity.25 Opportunistic infections including Pneumocystis pneumonia have been known to occur.29 Infliximab, an antibody to human tumor necrosis factor-α, is useful in advanced cases of IBD. It generally has a benign side effect profile26 but carries an increased risk of opportunistic infections, including tuberculosis and fungal infections. Surgery is reserved for patients with severe disease refractory to medical management or for patients with complications such as intestinal obstruction, significant bleeding, abscess, or fistula. A colectomy is curative for UC and improves quality of life, but there is no curative surgery for CD. Extraintestinal manifestations usually respond to therapy for intestinal disease.27,36
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
1240
always blood in the stool.41 The typical patient has had recent surgery or has a significant medical illness. Some patients present without pain. Nausea and vomiting can occur with obstruction secondary to a stricture or an ileus. Tenderness over the affected colon may be present but often is not dramatic. Peritoneal findings, fever, and a significantly elevated WBC count suggest gangrenous bowel and perforation. Toxic megacolon is a recognized complication.
Angiography usually is not helpful in either the diagnosis or the management of colonic ischemia. In most cases, the blood flow defect is at the microvascular level and has resolved by the time the patient presents for evaluation.42 The exception is the case in which only the ascending colon is affected, suggesting a superior mesenteric artery thrombosis.
Diagnostic Strategies
Differential Considerations
No sensitive or specific biochemical markers for colonic ischemia are recognized, although biochemical abnormalities such as elevated serum lactate, phosphate, and alkaline phosphate levels may be present. These abnormalities may be absent in milder disease and often are not observed in more significant disease until after irreversible damage has occurred. A complete blood count to exclude significant anemia and to look for a leukocytosis suggestive of perforation is appropriate. Serum electrolytes should be checked if diarrhea or vomiting has been significant or prolonged. Stool blood and WBCs are common findings in several of the entities that present similarly to colonic ischemia, including IBD and infectious colitis. A positive occult stool guaiac test result should ensure that the patient is eventually evaluated for colonic carcinoma. Unfortunately, the definitive diagnosis of colonic ischemia rarely is made in the ED.41
The symptoms of colonic ischemia are nonspecific and overlap with those of numerous other disorders including IBD, radiation proctocolitis, and infectious colitis, and of other causes of nonprofuse lower gastrointestinal bleeding. If strictures are present, the possibility of diverticulitis or colon cancer should be considered.
Plain Radiography Plain radiographs often show only nonspecific dilated bowel. Findings specific for colonic ischemia occur in only 20% of patients. The classic findings are intraluminal prominences, known as thumbprinting, representing submucosal hemorrhage and swelling. Thumbprinting also can be seen occasionally with other disorders, including IBD, colonic infections, or hemorrhage secondary to anticoagulants. Other findings consistent with colonic ischemia include wall thickening and ahaustral segments. Air in the portal venous system or bowel wall suggests imminent intestinal infarction.
Barium Enema Thumbprinting is detected more often by barium enema than by plain radiography, but this study has largely been replaced by colonoscopy.
Colonoscopy Colonoscopy with colonic biopsy is the preferred method to diagnose colonic ischemia because it visualizes the abnormal colonic mucosa better than barium enema and affords the opportunity to take biopsy specimens to differentiate between cancer and other non-ischemic causes of colitis. Colonoscopy also can detect necrotic bowel by its distinct cyanotic or black appearance. If colonoscopy is delayed, pathologic changes consistent with colonic ischemia may have already improved or resolved. The diagnosis may be missed on colonoscopy in up to one third of cases.39
Computed Tomography Although CT does not allow the definitive diagnosis of colonic ischemia, it can exclude other disorders. CT features suggestive of colonic ischemia include thumbprinting, wall thickening, and luminal narrowing and inner wall hypoperfusion (“double halo sign”).38
Angiography
Management In the absence of surgical complications, the treatment of colonic ischemia is supportive and includes hospitalization for bowel rest, hydration, and pain management, with some authors recommending antibiotic coverage.38 NSAIDs are best avoided, as are oral cathartics or bowel preparation regimens that may lead to perforation. Broad-spectrum antibiotics covering bowel flora are indicated for patients with more significant symptoms (see Box 93-5). If colonic ischemia is precipitated by an episode of hypotension, the underlying cause of the hypotension must be sought and treated aggressively, and cardiac output must be maximized.43 Vasopressors should be avoided to prevent worsening of ischemia, as should steroids, which may facilitate bowel perforation. Colonic distention if present can be relieved acutely through the use of a rectal tube; surgical consultation is recommended in these cases. Decompression of the colon may result in a lowering of transmural pressure and improved colonic perfusion. Sepsis, peritoneal changes, free abdominal air, significant fever, massive bleeding, and a significant leukocytosis suggest bowel necrosis or perforation and should prompt emergent surgical consultation.38,40
Disposition Patients with mild symptoms and no significant abdominal tenderness or bleeding can be managed on an outpatient basis and referred for colonoscopy. Stool studies including cultures for bacteria, microscopy to look for ova and parasites, and a Clostridium difficile titer are helpful if the diagnosis is uncertain. Patients with more significant findings, especially if the diagnosis of gangrenous bowel cannot be excluded, require hospitalization. A high mortality rate (60%) is expected for patients undergoing emergent surgery, although deaths before the age of 50 years are rare.40 Most patients improve without surgical intervention, and only 5% have a recurrence of colonic ischemia. Colectomy usually is curative.
■ RADIATION PROCTOCOLITIS Perspective Radiation proctocolitis is a common side effect of radiation therapy, occurring in 50 to 75% of patients receiving radiation to the pelvis. The disease has two distinct presentations: acute and chronic. Acute radiation proctocolitis begins during or shortly after a course of radiation therapy, usually is easily diagnosed, and is self-limited. Chronic radiation proctocolitis occurs in 5 to
Principles of Disease Radiation causes tissue injury through the creation of oxygen free radicals, which damage cellular DNA. The faster the growth rate of cells, the more this DNA damage affects their function. For this reason, radiation is an effective treatment for neoplastic disease, but it also damages rapidly growing normal tissue, such as intestinal epithelium. Radiation damage also may adversely affect anal sphincter function, leading to bowel incontinence.
Acute Radiation Proctocolitis Intestinal epithelium normally is sloughed and replaced at a rapid rate. After the start of radiation therapy, growth of replacement epithelium is slowed, but sloughing continues at the preexposure rate. This mismatch leads to gaps in the epithelium, which over time coalesce into ulcerations. In addition, edema and inflammatory changes of the submucosa cause excessive mucus secretion and bleeding. When radiation therapy has ended, the cycle of damage stops, and healing occurs over the next few weeks.
Chronic Radiation Proctocolitis The pathologic mechanism in chronic radiation proctocolitis is entirely different from that in acute radiation proctocolitis. Chronic radiation proctocolitis results from a progressive endarteritis with abnormal tissue collagen deposition. Affected intestine has a decreased microvascular density, with subsequent decreased perfusion.45 Over time, affected bowel gradually becomes more ischemic, leading to ulceration, scarring, and narrowing of the bowel lumen. Frank necrosis and perforation, although uncommon, can occur. Long-term outcomes in chronic radiation proctocolitis have not been well studied, but it seems that patients in whom fistulas and persistent bleeding strictures develop have the poorest prognosis.44
Clinical Features Acute radiation proctocolitis manifests with abdominal pain, bleeding, and tenesmus. Onset during the course of radiation therapy, typically after several treatments, suggests the diagnosis. Fecal urgency and incontinence can be devastating to quality of life.45 Chronic radiation proctocolitis has a more insidious onset with a variety of presentations, including ulcerative disease, stricture with or without obstruction, fistulas, and bowel perforation. Symptoms may be similar to those in acute disease, with tenesmus, diarrhea, and urgency. Bleeding can occur but usually is not hemodynamically significant. Decreased caliber of stool with increased straining or constipation suggests a stricture. Fistulas can develop between affected bowel and
1241
any adjacent organ, but the most common fistulas are rectovaginal. Some patients may exhibit anal sphincter dysfunction and loss of bowel control. Symptoms tend to have a significant negative impact on the quality of life.44
Diagnostic Strategies The diagnosis of acute radiation proctocolitis is made clinically on the basis of the development of typical symptoms in the setting of radiation therapy. Further evaluation usually is not warranted. Chronic radiation proctocolitis typically is a diagnosis of exclusion. Endoscopy can be suggestive, revealing pale, thickened, and friable mucosa with prominent telangiectasias. Biopsy specimens often show only nonspecific chronic inflammation. In some cases, endoscopy may be technically difficult because of scarring and reduced mobility of the intestine. Barium enema is an acceptable alternative when endoscopy is problematic, so long as bowel perforation is not a concern.
Differential Considerations In chronic radiation proctocolitis, the possibility that symptoms are due to recurrence of the initial malignancy or a new malignancy induced by radiation exposure must be entertained. Symptoms of chronic radiation proctocolitis generally are clinically indistinguishable from those of other causes of bowel inflammation, including inflammatory bowel disease, infectious colitis, and ischemic colitis.46
Management Treatment of acute radiation proctocolitis is symptomatic, and a therapeutic plan should be developed in conjunction with the patient’s radiation therapist. Steroid enemas to reduce inflammation and water-absorbing stool softeners to reduce mucus-containing diarrhea are helpful. Reduction of the daily radiation dose also can reduce symptoms significantly. Chronic radiation proctocolitis treatment also is symptomatic. If rectal involvement is significant, stool softeners, analgesics, anti-inflammatory agents (e.g., sulfasalazine, balsalazide), and sucralfate enemas are helpful. Metronidazole is beneficial when added to anti-inflammatory therapy. Minimally symptomatic strictures can be managed initially with stool softeners and enemas as needed. Some strictures have a reversible edema component, so the extent of narrowing may lessen after treatment. Fistulas and significant strictures generally require surgical repair. Approximately 20% of all patients with chronic radiation injury to the intestinal tract require some type of surgical intervention. Biopsy specimens from ulcerations associated with chronic injury should be obtained to exclude malignancy.
Disposition Suspected perforation mandates emergent surgical consultation, and signs of bowel obstruction should prompt urgent surgical consultation. Unless symptoms are severe, patients with acute or chronic radiation proctocolitis usually can be managed on an outpatient basis under the care of their radiation therapist or gastroenterologist. With acute disease, symptoms typically resolve several weeks after radiation treatments have been completed. Mild chronic disease typically resolves with medical therapy, but more severe symptoms often require aggressive intervention.
Chapter 93 / Disorders of the Large Intestine
10% of patients who have had pelvic radiation therapy and typically begins any time up to 2 years after the end of radiation therapy, although onset of clinical manifestations may be delayed beyond 2 years. Some cases have occurred decades later. Patients with more severe acute radiation proctocolitis seem to be prone to chronic proctocolitis.44 Because of its nonspecific presentation and delayed appearance, the diagnosis of chronic radiation proctocolitis can be challenging. Patients at risk for chronic radiation proctocolitis seem to be those with more severe acute disease.
1242
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
KEY CONCEPTS ■
New or atypical symptoms in a patient with known IBS should prompt an evaluation for other abdominal pathology. ■ A new diagnosis of IBS should be left to the primary care setting; the purpose of the ED evaluation is to exclude other abdominal disorders. ■ A positive result on testing for fecal occult blood should never be assumed to be due to diverticula. An appropriate investigation to exclude malignancy is essential. ■ Uncomplicated diverticulitis can be diagnosed and treated without imaging in many patients. ■ An LBO should prompt an evaluation for malignancy. ■ Gangrene or perforation should be suspected in any patient with persistent unexplained tachycardia, fever, or remarkable abdominal tenderness that is associated with intestinal disease. ■ Volvulus often appears as a nonspecific large bowel obstruction on plain radiographs. ■ Although typically a disease of older persons, volvulus can occur at any age.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
■
Intussusception in adults most commonly is associated with a significant cause, often a malignancy. ■ Intussusception in adults usually manifests as a partial small bowel obstruction and rarely is associated with the classic triad of abdominal pain, mass, and heme-positive stool that is seen in children. ■ IBD is a lifelong relapsing disorder that can be treated with a variety of therapies. Management decisions are best made in consultation with the physician who will be providing ongoing care for the patient. ■ Treatment of uncomplicated IBD depends on the clinical classification of disease severity (see Box 93-7). ■ A new diagnosis of IBD in an elderly patient should be made only after the exclusion of colonic ischemia. ■ Any evidence of blood in the stool should prompt an evaluation for colon cancer. ■ Chronic radiation proctocolitis should be considered in any patient with a history of irradiation of the pelvis or abdomen who presents with symptoms of gastrointestinal inflammation. This is true regardless of how long ago the radiation therapy was received.
Chapter 94
Disorders of the Anorectum
Wendy C. Coates
■ PERSPECTIVE Patients present to the emergency department (ED) with a variety of anorectal complaints. Such problems may be selflimited or may signify the presence of an underlying medical condition. A high degree of sensitivity and a professional demeanor should be maintained in interactions with these patients, who may find it difficult to discuss historical details openly and to describe physical complaints related to this area of the body and its function.
■ PRINCIPLES OF DISEASE The anorectum marks the end of the alimentary canal. From its beginning at the rectosigmoid junction at the level of the third sacral vertebra (S3), the rectum follows the sacral curvature for 12 to 15 cm and then sharply turns posteriorly and inferiorly at the puborectalis muscle (Fig. 94-1). Here the anal canal begins its 4-cm course to the anal verge, the orifice whereby stool exits the body. It is supported by three muscle groups, the levator ani and the internal and external anal sphincters. Anal valves are located 2 cm proximal to the anal verge at the dentate line. Above the valves are the anal crypts, which contain mucous glands to provide lubrication during defecation. These constitute a nidus for abscess and fistula formation if occluded. Proximal to the crypts are the columns of Morgagni, where the epithelium of the anal canal changes from pink columnar (as in the rectum) to squamous.1-3 The superior, middle, and inferior hemorrhoidal arteries provide the blood supply to the anorectum. They arise from the inferior mesenteric, internal iliac, and internal pudendal arteries, respectively. The superior hemorrhoidal veins drain into the portal system, and the inferior hemorrhoidal veins drain into the caval system. Lymphatic drainage is to the inferior mesenteric nodes above the dentate line and to the inguinal nodes from all areas of the anorectum.2 Sympathetic and parasympathetic nervous systems function together to retain the contents of the rectum until evacuation is desired. Continence is maintained when sympathetic fibers from L1 to L3 (upper rectum) and presacral nerves (lower rectum) inhibit contraction of rectal smooth muscle and L5 fibers cause the internal sphincter to contract. Elimination occurs when parasympathetic fibers from the anterior roots of S2 to S4 cause the rectal wall to contract and the internal sphincter to relax. Voluntary external sphincter control is mediated by motor branches of the pudendal nerve (S2, S3)
and the perineal branch of S4. The levator ani is supplied by the pudendal nerve and pelvic branches of S3 to S4 fibers. Sensory perception of rectal distention involves a signal pathway from extramural receptors to parasympathetic fibers from S2 to S4. The abundant sensory nerve endings of the distal anal epithelium perceive sensations that are transmitted by the pudendal nerve.2 Defecation begins as the rectum becomes distended, the internal sphincter relaxes, and stool enters the anal canal. At an appropriate time and place, the external sphincter is relaxed to complete the process of elimination. Sometimes voluntary straining is needed to assist in the passage of stool. When the Valsalva maneuver is performed, the abdominal muscles contract, the rectal angle straightens, and the pelvic floor descends. To postpone defecation, the external sphincter contracts voluntarily. This contraction relaxes the rectal wall and quells the urge to defecate unless there is an underlying sphincter disorder or an overwhelming volume of stool.4
■ CLINICAL FEATURES History A complete history of anorectal and gastrointestinal (GI) symptoms and the presence of systemic disease elucidates the diagnosis of most anorectal disorders (Box 94-1; Fig. 94-2). Common complaints include bleeding, swelling, pain, itching, and discharge. Standard historical questions about time and circumstances of onset, duration, quality, and exposure to radiation should be asked. Alterations in bowel habits should be noted. These include changes in color, frequency, or consistency of the stool and the presence of straining, flatus, and incontinence of solid or liquid stool. Persons with underlying GI disorders (e.g., Crohn’s disease, cancer, polyps) are predisposed to atypical presentations of anorectal problems. Similarly, those with underlying systemic diseases such as acquired immunodeficiency syndrome (AIDS), cancer, diabetes mellitus, and coagulopathy are prone to develop more serious complications of standard anorectal conditions. Finally, patients should be asked directly about sexual practices involving the anus.5
Rectal Bleeding The color, amount, and relationship to defecation are important factors in establishing the cause of rectal bleeding. Approximately 10 to 20% of the population experiences rectal bleeding at some time.6 Pain and bright red blood signify anal 1243
1244
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Medical History in Diagnosis of
BOX 94-1 Anorectal Disorders Rectosigmoid junction S3
Middle valve of Houston (level of peritoneal reflection)
Levator ani muscle Puborectalis muscle
GI History Change in bowel habits (straining, flatus, color, consistency, frequency) Nausea or vomiting Incontinence of stool Underlying GI disease (Crohn’s disease, cancer, polyps) Systemic Disease History Diabetes mellitus Coagulopathy Cancer HIV infection
Anorectal junction Columns of Morgagni
External anal sphincter
Anorectal History Pain Bleeding Swelling Itching Discharge Urgency
Sexual History of the Anus Penetration Known STDs Assault
Dentate line Internal anal sphincter Anal verge
GI, gastrointestinal; HIV, human immunodeficiency virus; STD, sexually transmitted disease.
Figure 94-1. Anorectal anatomy.
Pain? No
Yes Bleeding?
Bleeding?
Yes
No
Swelling? Yes
No
Yes
No
Swelling? Yes
Swelling? No
Yes
No
Itch?
Itch? Yes
Yes Episodic pain?
Anal fissure
Abscess Pilonidal cyst Hidradenitis
Proctalgia fugax
D/C?
Urgency? Yes
Yes Hemorrhoids
No
No
Yes Incontinence
Polyps Adenoma Hemorrhoids
Condylomata acuminata
Procidentia
Proctitis
Pruritus ani
Figure 94-2. Algorithm for anorectal complaints. D/C, discharge.
fissures or hemorrhoids. Fissure pain is sharp, sudden in onset, and not associated with swelling, whereas pain from a prolapsed or thrombosed hemorrhoid is gnawing, continuous, and of more gradual onset. Painless rectal bleeding occurs with internal hemorrhoids, cancer, or precancerous lesions. The relationship of bleeding to defecation is important. Visible blood on the toilet paper usually is caused by anal fis-
sures or external hemorrhoids; however, minute quantities can result from any irritating condition. Bright red blood that drips into the toilet bowl or streaks around the stool is caused by internal hemorrhoids. Blood mixed with stool originates proximal to the rectum, whereas melena indicates a very proximal source. Bloody mucus is associated with cancer, inflammatory bowel disease, and proctitis.1,6
1245
Swelling and Masses
Pain and Itching Severe, episodic anorectal pain that is not associated with bleeding or swelling may represent proctalgia fugax or levator ani syndrome. Perianal itching (pruritus ani) is caused by any lesion that makes hygiene difficult to maintain or may be attributed to certain foods or medications.
Physical Examination The physical examination should take place in private, respecting the patient’s modesty. The patient can then relax the external sphincter to facilitate a complete examination. The patient is placed in the left lateral decubitus position and covered with a sheet. The buttocks are inspected for dermatologic manifestations of disease and then gently spread apart to expose the anal orifice. Elements of personal hygiene are noted, in addition to anatomic disruptions such as fissures, skin tags, lesions, protruding hemorrhoids, or abscesses. The patient is asked to strain to assess the integrity of the pelvic floor and note prolapse of hemorrhoids or rectal mucosa. Next, a welllubricated gloved finger is placed flat against the anal opening, exerting gentle pressure until the external sphincter relaxes and allows the finger to enter the anus. Anal sphincter tone can be assessed by asking the patient to squeeze the anal muscles against the examining finger. By sweeping the finger in a circumferential manner, accessible areas of the anorectum can be examined for masses and areas of tenderness. The cervix or prostate is palpated through the rectal wall. A bidigital examination reveals masses and tender areas at the distal portion of the anal canal and perineum. On withdrawal, the contents on the glove can be assessed for frank or occult blood, mucus, or pus.1 Direct visualization can be accomplished by anoscopy. With the patient positioned as described, the lubricated anoscope is inserted into the anus with the obturator in place. The obturator is removed to allow a circumferential view of the rectal mucosa. Attention is directed to sites of bleeding, hemorrhoids, masses, or abnormal tissue and finally the dentate line and anal epithelium.
■ SPECIFIC ANORECTAL PROBLEMS Hemorrhoids Perspective. When the Philistines defeated the Israelites, the book of I Samuel reports the fate of the avengers: “A deadly panic had seized the whole city, since the hand of God had been very heavy upon it. Those who escaped death were afflicted with hemorrhoids, and the outcry from the city went up to the heavens.”7 In 1815 the battle of Waterloo marked the defeat of Napoleon’s army. Speculation purports that the great leader suffered from hemorrhoids at the time of his defeat.8 Hemorrhoidal disease continues to afflict modern
Chapter 94 / Disorders of the Anorectum
Patients who complain of a swelling near the anus or have the sensation of rectal fullness often list hemorrhoids as their chief complaint. Painful swellings that bleed usually are thrombosed hemorrhoids, but other painful lesions such as abscesses, pilonidal disease, and hidradenitis suppurativa must be considered. Painless, itchy swellings may be caused by condylomata acuminata or secondary syphilis. A mass protruding through the anal orifice may signal rectal prolapse.1 Perianal and rectal carcinoma should be considered in older persons and those with long-standing anorectal complaints.5
humans, with a 4.4% incidence in the U.S. population. Both sexes are affected, and an increased frequency has been documented among whites, rural dwellers, and those of high socioeconomic status.9-11 Principles of Disease. The cause of hemorrhoids is controversial. The anal vascular cushion theory is the most widely accepted. Rather than forming a continuous ring around the anal canal, the submucosa forms three distinct cushions of tissue that are richly supplied with small blood vessels and muscle fibers. Blood supply to these cushions is from the superior rectal artery, with some contribution from the middle and inferior hemorrhoidal arteries, which explains why hemorrhoidal bleeding is bright red. The muscularis submucosa cushions the anal canal during defecation to prevent injury and to aid in fecal continence.12 As the supportive tissue deteriorates, often starting in the third decade of life, venous distention, prolapse, bleeding, and thrombosis may occur. Some controversy exists about whether straining and constipation cause these changes by producing venous backflow when intra-abdominal pressure increases.12,13 In pregnant women, direct pressure on a hemorrhoidal vein can produce symptomatic hemorrhoids. Up to one third of pregnant women experience hemorrhoids in the last trimester of pregnancy or the postpartum period. An increased incidence of thrombosed hemorrhoids is associated with traumatic deliveries.14,15 Some familial predisposition is recognized, but whether this is a result of genetics or acquired factors such as diet is unknown. Hemorrhoids are not varicose veins; they are normal structures that manifest symptoms when the muscularis submucosa weakens and the anal cushions are displaced distally.12 Conditions that increase sphincter tone correlate with a higher prevalence of hemorrhoids.10,13 Portal hypertension does not cause hemorrhoids. The incidence of symptomatic hemorrhoids is similar in patients with and in those without portal hypertension. Rectal bleeding in patients with portal hypertension may be caused by rectal varices, which are vascular communications between the superior and middle hemorrhoidal veins.10,11 A major exception to this observation occurs in the pediatric population; children with portal hypertension are susceptible to hemorrhoidal exacerbations.16,17 Clinical Features. A careful history is needed to confirm the presence of hemorrhoids, because many patients use this term to refer to any perianal condition. Bleeding with defecation is the most common complaint, and unless the hemorrhoids are thrombosed, it usually is painless. Patients report variable amounts of bright red blood on the toilet paper or in the toilet bowl. Many complain of swelling, itching, mucoid discharge, or simply the presence of a moist perianal area. Further history should address recent stool patterns, such as diarrhea or constipation; chronic medical problems, such as portal hypertension or bleeding disorders; and a dietary and family history. Hemorrhoidal symptoms are exacerbated by frequent bowel movements, prolonged sitting, heavy lifting, and straining while defecating. Although straining is cited as a cause of hemorrhoids, it also may be a result of them when the patient is constipated from delaying defecation because of fear of pain. Physical examination should ascertain the type and degree of hemorrhoids. This can be accomplished by a visual inspection at rest and during straining. Nonprolapsing hemorrhoids can be visualized on anoscopy as a focus of bleeding or as they bulge when the patient is asked to strain while the anoscope is removed. Anoscopy is painful and not useful in cases of prolapsed or thrombosed hemorrhoids. Hemorrhoids are classified according to their location and severity (Table 94-1). External hemorrhoids originate below the dentate line and receive their blood supply from the inferior
1246
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
Table 94-1 Types of Hemorrhoids TYPE
ORIGIN
EPITHELIUM
External
Inferior hemorrhoidal plexus Proximal to dentate line Superior hemorrhoidal plexus Distal to dentate line Superior and inferior hemorrhoidal plexus
Modified squamous epithelium (anoderm)
Internal Mixed
Transitional or columnar epithelium (mucosa) Transitional, columnar, or modified squamous epithelium (mucosa and anoderm)
hemorrhoidal plexus. They are covered with modified squamous epithelium (anoderm) and resemble the surrounding skin. Two syndromes are common. First, the veins beneath the skin of the hemorrhoid become dilated and the surrounding subcutaneous tissue becomes engorged, causing swelling or pressure after defecation. Painless, bright red bleeding may occur. Second, the veins can become thrombosed as clots form within them (Fig. 94-3A). This produces acute pain and tenderness to palpation. A bluish discoloration often is noted. Internal hemorrhoids originate above the dentate line and receive their blood supply from the superior hemorrhoidal plexus (Fig. 94-3B). They are covered with a mucosal surface consisting of transitional or columnar epithelium that looks very different from the surrounding anoderm. They are classified according to severity (Table 94-2). Symptoms and signs range from mild, painless bleeding with defecation to irreducible prolapse with unremitting and debilitating pain. Firstdegree internal hemorrhoids protrude into the lumen of the anal canal, causing a feeling of fullness. Because the mucosal wall lacks sensory nerve endings, these lesions do not cause pain. Second-degree internal hemorrhoids temporarily prolapse outside the anal canal during defecation but spontaneously return to their normal position at the end of the bowel movement. Both of these are amenable to medical management. Third-degree internal hemorrhoids prolapse spontaneously or during defecation and remain outside the body until they are manually replaced into the anal canal. A throbbing, pressure-like pain may accompany bleeding and subsides when the hemorrhoids are reduced. Fourth-degree internal hemorrhoids cannot be reduced and are permanently prolapsed. Continued prolapse leads to the formation of a thrombus with possible progression to gangrene. Definitive treatment for the intense pain and thrombosis is surgical.1,13
Management. The symptoms of nonthrombosed external and nonprolapsing internal hemorrhoids can be ameliorated by the standard regimen—warm water, analgesics, stool softeners, and high-fiber diet (WASH)—aimed at combating the problems that led to their formation (Box 94-2). Anal canal pressures decrease in warm water (40° C).2 Patients can direct a shower stream at the area for several minutes or take sitz baths. Mild oral analgesic agents reduce the pain. Several over-thecounter preparations are available for the treatment of hemorrhoidal symptoms; however, their use is directed at improved hygiene and temporary symptom relief rather than correcting the condition.18 The use of topical anesthetics, corticosteroids, astringents (e.g., witch hazel), mineral oils, and cocoa butter is controversial. Prolonged use of topical corticosteroids produces atrophic skin changes and is discouraged.12 Stool softeners can make the passage of stool easier, to avoid straining. A highfiber diet (consumption of 20 to 30 g of dietary fiber per day) produces stool that is passed more easily.19 Patients with second- or third-degree internal hemorrhoids also benefit from this regimen; however, permanent resolution
of Internal Hemorrhoids by Table 94-2 Classification Severity TYPE
PROLAPSE
MODE OF REDUCTION
Firstdegree Seconddegree Thirddegree
None
N/A
During defecation May be spontaneous or during defecation Permanently
Spontaneous
Fourthdegree
Manual
Irreducible
TREATMENT
Medical management Medical management Medical management Optional surgical repair Surgical repair
N/A, not applicable.
BOX 94-2
The WASH Regimen for Management of Hemorrhoids
Warm water Analgesic agents Stool softeners High-fiber diet
Figure 94-3. Thrombosed
hemorrhoids. A, External. B, Internal. Note the engorged external hemorrhoids surrounding the thrombosed internal hemorrhoids. (A, Courtesy of Michelle Lin, MD, Harbor-UCLA Medical Center; B, courtesy of Gershon Effron, MD, Sinai Hospital of Baltimore. From Seidel HM, et al: Mosby’s Guide to Physical Examination, 4th ed. St. Louis, Mosby, 1999.)
A
B
1247
Table 94-3 Surgical Management of Hemorrhoids MANAGEMENT
Thrombosed external hemorrhoids Second- and third-degree internal hemorrhoids
Excision in emergency department Elective surgical repair Banding Sclerotherapy Hemorrhoidectomy Nonemergent hemorrhoidectomy Emergent hemorrhoidectomy
Fourth-degree hemorrhoids (nonthrombosed) Thrombosed or gangrenous fourth-degree internal hemorrhoids
Chapter 94 / Disorders of the Anorectum
CLASSIFICATION
Figure 94-5. Lateral anal fissure. (Courtesy of Gershon Effron, MD, Sinai Hospital of Baltimore. From Seidel HM, et al: Mosby’s Guide to Physical Examination, 4th ed. St. Louis, Mosby, 1999.)
Anesthetic
Elliptical incision
A
sphincter tone and thereby reduce the associated pain and inflammation.21
Anal Fissures Remove clot
B
C Figure 94-4. Excision of thrombosed external hemorrhoid. A, Field block
with local anesthetic. B, An elliptical incision is made around the hemorrhoid. C, The thrombosed hemorrhoid is removed. (From Larson S, et al [eds]: Atlas of Emergency Procedures. St. Louis, Mosby, 2001.)
of their symptoms may require surgical intervention (Table 94-3). These patients can be discharged from the ED with the WASH regimen and referred to a surgeon for banding, sclerotherapy, or elective hemorrhoidectomy. Patients with acute, gangrenous, thrombosed fourth-degree internal hemorrhoids should be referred for emergent hemorrhoidectomy. Acutely thrombosed external hemorrhoids can be excised (not incised and drained) in the ED to provide prompt relief within the first 48 hours after the onset of symptoms (Fig. 944). Incision results in incomplete evacuation of the clot, subsequent rebleeding, and swelling. Excision provides long-term relief and prevents subsequent formation of skin tags.13,20 If not excised, the thrombosed external hemorrhoid will resolve spontaneously after several days when it ulcerates and leaks the dark accumulated blood, with relief of associated symptoms. Residual skin tags may persist. In the ED setting, this procedure is not commonly performed in pediatric patients, pregnant women, and immunocompromised patients. Nonsurgical therapy with topical nifedipine (0.3%) with lidocaine (1.5%) gel has been shown to alleviate symptoms when applied twice daily for 2 weeks. The purported effectiveness of this regimen for treatment of thrombosed hemorrhoids is related to the ability of nifedipine to modulate resting
Principles of Disease. The development of an anal fissure is the most common cause of intensely painful rectal bleeding of sudden onset. A superficial tear in the anoderm results when a hard piece of feces is forced through the anus, usually in patients who are constipated. Although anyone can experience an anal fissure, it is most common in the 30- to 50-year age bracket.22 It is the most commonly encountered anorectal problem in pediatric patients, especially infants.23,24 Males and females are affected equally. Most fissures occur along the posterior midline, where the skeletal muscle fibers that encircle the anus are weakest. Anterior midline fissures are more common in women than in men.1,22 Fissures that occur elsewhere are more likely to be associated with systemic disease such as leukemia, Crohn’s disease, human immunodeficiency virus (HIV) infection, tuberculosis (TB), or syphilis.25 Fissures not treated promptly may become chronic, with development of a classical “fissure triad” of deep ulcer, sentinel pile, and enlarged anal papillae (Fig. 94-5). A sentinel pile forms when the skin at the base of the fissure becomes edematous and hypertrophic. A resolving sentinel pile can form a permanent skin tag and may be associated with a fistulous tract. Clinical Features. The patient complains of a sudden, searing pain during defecation that may be accompanied by a small amount of bright red blood on the stool or on the toilet paper. This is followed by a nagging, burning sensation that lasts for a few hours from internal sphincter spasm. Subsequent bowel movements are excruciating, and the external sphincter can exhibit a reflex spasm. Physical examination must be performed cautiously to avoid further spasm and pain. The depth of the fissure, its orientation to the midline, and the presence of a coexisting sentinel pile or edema are noted. Rectal examination during an acute exacerbation often is impossible because of pain and sphincter spasm.22 Management. Specific measures for the treatment of anal fissures are summarized in Box 94-3. Treatment using the WASH regimen (see Box 94-2) focuses on eliminating constipation
1248
focuses on reducing resting anal pressures and may require anal dilation performed with the patient under anesthesia or surgical correction to reduce the tone of the internal sphincter.29,35
PART III ■ Medicine and Surgery / Section Five • Gastrointestinal System
BOX 94-3 Treatment of Anal Fissures ■ WASH regimen* ■ Nitroglycerin ointment (0.4%) bid or tid ■ Nifedipine gel (0.2%) bid with lidocaine (1.5%) ■ Botulinum toxin (Botox) 0.1–0.2 mL ■ Anal dilation performed with the patient under
Abscesses and Fistulas general
anesthesia excision
■ Surgical
*See Box 94-2.
with a bulking agent, stool softener, and high-fiber diet. Warm sitz baths and limited use of topical anesthetics may be helpful. Parental encouragement to pediatric patients helps prevent encopresis that can result from a fear of painful bowel movements. Most acute, uncomplicated fissures resolve in 2 to 4 weeks. For adult patients who suffer from chronic anal fissures, application of various topical agents aimed at reducing sphincter pressures has been effective.26,27 Hyperbaric oxygen has been used successfully for adjunctive therapy.28 Application of lidocaine ointment is effective in treating chronic anal fissures. In addition, nitroglycerin ointment applied topically to the anoderm two or three times daily has been shown to relieve the pain from anal fissures. Although it was not associated with more rapid healing, patients receiving this therapy reported a higher level of comfort during the healing process. The typical side effect of a vasodilatory headache may be experienced by some patients.29,30 Nifedipine gel (0.2%) in combination with lidocaine (1.5%) applied to the anal area twice daily is effective in promoting healing and reducing discomfort in the management of anal fissures. The mechanism of healing is thought to be the reduction of anal canal pressures by local calcium channel blockade.31-33 When the efficacy of calcium channel blockers was compared directly with that of topical nitrates, the rates of healing and recurrence were similar; however, the incidence of side effects was lower in one study.34 Injection of botulinum toxin by colorectal surgeons (2.5 to 5.0 Units, 0.1 to 0.2 mL of Botox preparation) is effective in relaxing the sphincter tone by inhibiting acetylcholinesterase release but may cause temporary, reversible fecal incontinence.32,35-39 Injection into the external (rather than internal) sphincter muscles may reduce this undesirable side effect. It has not yet been studied as a primary treatment in the ED or primary care setting. In comparison with the topical treatments, botulinum toxin is superior in its rate of permanent healing40; however, the first line of therapy is still topical agents because of their cost, ease of application, and benign side effect profiles.39 Long-term treatment of recurrent fissures
Principles of Disease. Anorectal abscesses and fistulas are most common in adults 30 to 50 years of age, and men are afflicted more often than women.1,41 An increased incidence in infants (85% in male babies) has been reported to be associated with congenital abnormalities.42-44 One probable cause of anorectal abscesses is occlusion of the ducts of the mucus-producing anal glands at the base of the anal crypts (the cryptoglandular theory). Abscesses also are caused by inflammatory bowel disease, trauma, cancer, radiation injury, and infection (TB, lymphogranuloma venereum, actinomycosis).1,41,43 Common causative bacteria are Staphylococcus aureus, Escherichia coli, Streptococcus, Proteus, and Bacteroides. Management General Approach. The various types of abscesses are the acute manifestations of a continuum of anorectal infections, whereas fistulas are the chronic sequelae. Symptoms vary depending on the site of infection, but incision and drainage constitute the curative treatment in all cases (Table 94-4). Delay of medical management may allow extension of the infection and eventual compromise of the sphincter mechanism.45 Adjunctive antimicrobial therapy is indicated in patients who are immunocompromised or diabetic or have valvular heart disease. Tetanus immunization status should be verified. The sites of anorectal abscess formation are depicted in Figure 94-6. The difficulty in diagnosis is that pain often precedes physical findings of a mass or fluctuance. Approximately 34% of patients with AIDS develop anorectal abscesses and fistulas. In addition to the usual organisms, many are infected with opportunistic ones. HIV-infected patients appear to be more likely than their seronegative cohorts to have an incomplete fistulous tract. This condition prevents adequate spontaneous drainage, highlighting the urgency of treating these patients promptly. A small incision is desirable when possible because wound healing in general may be impaired. Treatment of Specific Abscesses Perirectal and Perianal Abscesses. Perirectal and perianal abscesses are the most common (40 to 45%) and produce painful swelling at the anal verge that is worsened by defecating or sitting. Most patients are afebrile. Physical examination reveals localized tenderness, erythema, swelling, and fluctuance. If tolerated, anoscopy may reveal pus in the anal crypts. ED management by incision and drainage with same-day discharge is possible in patients who do not have comorbidity (e.g., dia-
Table 94-4 Types of Abscesses of the Anorectum FEATURE
PERIANAL
ISCHIORECTAL
INTERSPHINCTERIC
SUPRALEVATOR
POSTANAL
Incidence Location Symptoms
40–45% Outside and verge Painful perianal mass
20–25% Buttocks Buttock pain
20–25% Lower rectum Rectal fullness
Fever, ↑WBCs Associated fistula ED incision and drainage
− ++ +
± + ±
± +++ −
40
4.5
Trichomonas infection
>4.5
Candida infection
102 CFUS/ML)
POSITIVE FLUID OR CERVICAL SMEAR
POSITIVE CULTURE OF GENITAL LESIONS, CERVIX, OR URETHRA*
+ +
+ ±
+ +
− −
− −
+ + + −
− − − −
− − − −
± + + +
+ + + ±
−
−
−
−
−
MICROSCOPIC HEMATURIA OR PYURIA
Acute pyelonephritis Acute cystitis Urethritis caused by sexually transmitted disease Herpes simplex virus Neisseria gonorrhoeae Chlamydia trachomatis Vulvovaginitis (bacterial vaginosis, trichomoniasis, yeast, genital herpes simplex) Noninflammatory dysuria (trauma, irritant, allergy)
*Positive for herpes simplex virus, Neisseria gonorrhoeae, or Chlamydia trachomatis. From Stamm WE: Protocol for diagnosis of urinary tract: Reconsidering the criterion for significant bacteriuria. Urology 32(2 Suppl):6, 1988.
Table 97-2 Differential Diagnosis of Dysuria Syndromes: Physical Examination SYNDROME/DISORDER
VAGINAL OR CERVICAL DISCHARGE, VULVAR LESIONS
SUPRAPUBIC TENDERNESS
FLANK TENDERNESS, FEVER
− −
± ±
− −
+ + + +
− − − −
− − − −
−
−
−
Acute pyelonephritis Acute cystitis Urethritis caused by sexually transmitted disease Herpes simplex virus Neisseria gonorrhoeae Chlamydia trachomatis Vulvovaginitis (bacterial vaginosis, trichomoniasis, yeast, genital herpes simplex) Noninflammatory dysuria (trauma, irritant, allergy)
From Stamm W: Protocol for diagnosis of urinary tract: Reconsidering the criterion for significant bacteriuria. Urology 32(2 Suppl):6, 1988.
Differentiation among Major Causes Table 97-3 Clinical of Dysuria CAUSE
CLINICAL FEATURES
Urinary tract infection
Internal dysuria Frequency, urgency, voiding small volumes Abrupt onset Suprapubic pain Often associated with diaphragm use Presence of pyuria Presence of hematuria (50% of patients) Internal dysuria Occasional history of frequency, urgency, voiding small volumes Gradual onset History of new or multiple sexual partners Vaginal discharge External dysuria Gradual onset Vaginal discharge Vaginal odor Pruritus
Sexually transmitted disease
Vaginitis
From Stamm W: Protocol for diagnosis of urinary tract: Reconsidering the criterion for significant bacteriuria. Urology 32(2 Suppl):6, 1988.
therapy is recommended. Seven- to 10-day therapy generally offers no benefit over shorter courses in uncomplicated UTIs; however, it remains the standard of care in complicated infections (in patients with pregnancy, diabetes, or sickle cell anemia) for which cure rates are lower with shorter regimens. The fluoroquinolones are considered first-line agents in regions in which the incidence of TMP-SMX resistance has approached 10 to 20%.27 Ciprofloxacin is the most commonly used drug and requires twice-daily dosing. Although more expensive than ciprofloxacin, gatifloxacin and levofloxacin offer once-daily dosing, have the broadest activity, and have same-dose bioequivalence between oral and parenteral administration. Fluoroquinolones damage developing cartilage in animal studies and should be avoided in children.31 Nitrofurantoin and trimethoprim are excellent drugs for the treatment of acute bacterial cystitis. Nitrofurantoin is inexpensive and maintains low serum and high urine levels, with a bacterial resistance pattern that remains unchanged. Adverse reactions are primarily secondary to gastrointestinal disturbance, but they may be alleviated by using the macrocrystalline form (Macrodantin). Folate antagonists such as trimethoprim have a broader spectrum of activity than that of nitrofurantoin. The addition of sulfamethoxazole further broadens the spectrum to include coverage for Proteus and Klebsiella. Folate antagonists carry a higher incidence of adverse
1303
Woman with dysuria
Yes
No Evaluate lower tract symptoms/signs
Evaluate for upper tract symptoms/signs
Differentiate cystitis from urethritis and vaginitis
Assess for costovertebral angle tenderness Order urinalysis, urine Gram’s stain and culture
No
Is pyuria present?
Treat empirically
Yes Is hematuria present?
Consider STD or vaginitis
Yes
UTI likely
No Do a pelvic exam Order cervical culture and Gram’s stain; pH and microscopic exam of vaginal fluid
Does the patient have vaginal discharge?
Yes
Yes
No Does the patient have a new or multiple sexual partners? No
Consider chlamydial infection
Are UTI symptoms/signs clear-cut? Yes
No Order a urine culture
No
Does culture confirm presence of a pathogen >102 CFUs/mL?
Select a first-line antimicrobial and treat empirically
Yes
Figure 97-1. Diagnostic protocol for women with dysuria.
CFUs, colony-forming units; STD, sexually transmitted disease; UTI, urinary tract infection. (From Stamm WE: Protocol for diagnosis of urinary tract: Reconsidering the criterion for significant bacteriuria. Urology 32[2 Suppl]:6, 1988.)
Complex Urinary Tract Infection Mild to moderate pyelonephritis can be safely treated on an outpatient basis with a fluoroquinolone for 10 to 14 days (firstline agent) or TMP-SMX (second-line agent) so long as the patient is able to eat and drink, has achieved adequate pain control, and has approrpiate pychosocial support in the home. In many clinical centers, observation units have evolved to offer a short-stay (less than 24 hours) option for moderate cases in which immediate discharge for outpatient therapy may not be the optimal approach to management. Severe upper tract UTI necessitating hospitalization initially should be treated with parenteral antibiotics, with transition to oral therapy after the patient has been afebrile for 24 to 48 hours. Oral therapy should be continued for 2 weeks. Because 20% of cultures are resistant to ampicillin, cephalothin, and sulfonamides, antibiotic therapy should be initiated with a fluoroquinolone. Hospitalization is required in the presence of clinical toxicity (fever, tachycardia, hypotension, vomiting), inability to take oral medications, an immunocompromised state, third trimester pregnancy, inadequate social circumstances, failure of oral outpatient therapy, or urologic abnormalities or in patients with significant comorbid conditions including heart failure, renal insufficiency, and compromised immune status. A subgroup of patients suffering from upper tract UTI do not require immediate hospital admission but may benefit from intravenous hydration and pain and fever control, along with a first dose of an intravenous fluoroquinolone. If these patients do not have any contraindications as previously discussed and they improve clinically and are able to tolerate food and drink, they can be safely discharged home on a 10- to 14day course of an oral fluoroquinolone with close primary physician follow-up. Urine culture with sensitivity testing and further diagnostic evaluation are not necessary in this patient population.
■ URINARY TRACT INFECTION IN CHILDREN Perspective UTI is a major bacterial disease of childhood; it is estimated that 0.8 to 1.5% of children have bacteriuria. The risk for development of UTI before 11 years of age is 3% in girls and 1% in boys.9 The incidence of UTI in the neonatal period is higher in boys, but the infection becomes more prominent in girls during infancy and thereafter. In children aged 1 to 3 months, UTI is associated with a high incidence of sepsis (30%).10 After the age of 3 months, the incidence of sepsis associated with UTI decreases (to 5%).10 Vesicoureteral reflux is a common risk factor for UTI and renal scarring in children.19
Chapter 97 / Selected Urologic Problems
Does the patient have fever, malaise, chills, or flank pain?
effects than nitrofurantoin, predominantly gastrointestinal upset, yeast vaginitis, and rash. Addition of the sulfa component further increases the likelihood of side effects. Except in pregnancy, ampicillin and amoxicillin should not be used empirically as first-line drugs for the treatment of uncomplicated UTI. Recurrence rates with use against ampicillin-resistant strains are high, and neither agent effectively eradicates the vaginal reservoir of pathogenic bacteria. A useful adjunctive therapy for UTIs is phenazopyridine (Pyridium). It produces topical analgesia in the urinary tract and helps relieve dysuria. Patients should be cautioned that body secretions and excretions (e.g., tears, urine) will turn orange. This side effect can stain contact lenses and alarm unknowing patients.
1304
PART III ■ Medicine and Surgery / Section Six • Genitourinary and Gynecologic Systems
Table 97-4 Treatment Regimens for Bacterial Urinary Tract Infections (UTIs) CONDITION
CHARACTERISTIC PATHOGENS
MITIGATING CIRCUMSTANCE(S)
RECOMMENDED EMPIRICAL TREATMENT*
Acute uncomplicated cystitis in women
Escherichia coli, Staphylococcus saprophyticus, Proteus mirabilis, Klebsiella pneumoniae
None
Acute uncomplicated pyelonephritis in women
E. coli, P. mirabilis, K. pneumoniae, S. saprophyticus
Mild to moderate illness, no nausea or vomiting— outpatient therapy Severe illness or possible urosepsis—hospitalization required
3-day regimens: oral TMP-SMX, trimethoprim, norfloxacin, ciprofloxacin, ofloxacin, lomefloxacin, or enoxacin† Consider 7-day regimen: oral TMP-SMX, trimethoprim, norfloxacin, ciprofloxacin, ofloxacin, lomefloxacin, or enoxacin† Consider 7-day regimen: amoxicillin, macrocrystalline nitrofurantoin, cefpodoxime proxetil, or TMP-SMX† Oral‡ TMP-SMX, norfloxacin, ciprofloxacin, ofloxacin, lomefloxacin, or enoxacin for 10–14 days Parenteral§ TMP-SMX, ceftriaxone, ciprofloxacin, ofloxacin, or gentamicin (with or without ampicillin) until fever gone; then oral‡ TMP-SMX, norfloxacin, ciprofloxacin, ofloxacin, lomefloxacin, or enoxacin for 14 days Parenteral§ ceftriaxone, gentamicin (with or without ampicillin), aztreonam, or TMP-SMX until fever gone; then oral‡ amoxicillin, a cephalosporin, or TMPSMX for 14 days Oral‡ norfloxacin, ciprofloxacin, ofloxacin, lomefloxacin, or enoxacin for 10–14 days
Diabetes, symptoms for >7 days, recent UTI, use of diaphragm, age >65 yr Pregnancy
Pregnancy—hospitalization recommended
Complicated UTI
E. coli, Proteus species, Klebsiella species, Pseudomonas species, Serratia species, enterococci, staphylococci
Mild to moderate illness, no nausea or vomiting— outpatient therapy
*Treatments listed are those to be prescribed before the etiologic agent is known (Gram staining can be helpful); they can be modified once the agent has been identified. These recommendations are limited to drugs currently approved by the U.S. Food and Drug Administration, although not all of the regimens listed are approved for these indications. Fluoroquinolones should not be used in pregnancy. TMP-SMX, although not approved for use in pregnancy, has been widely used. Gentamicin should be used with caution in pregnancy because of its possible toxicity to eighth nerve development in the fetus. † Multiday oral regimens for cystitis are as follows: TMP-SMX, 160 to 800 mg every 12 hours; trimethoprim, 100 mg every 12 hours; norfloxacin, 400 mg every 12 hours; ciprofloxacin, 250 mg every 12 hours; ofloxacin, 200 mg every 12 hours; lomefloxacin, 400 mg every day; enoxacin, 400 mg every 12 hours; macrocrystalline nitrofurantoin, 100 mg four times a day; amoxicillin, 250 mg every 8 hours; and cefpodoxime proxetil, 100 mg every 12 hours. ‡ Oral regimens for pyelonephritis and complicated UTI are as follows: TMP-SMX, 160 to 800 mg every 12 hours; norfloxacin, 400 mg every 12 hours; ciprofloxacin, 500 mg every 12 hours; ofloxacin, 200 to 300 mg every 12 hours; lomefloxacin, 400 mg every day; enoxacin, 400 mg every 12 hours; amoxicillin, 500 mg every 8 hours; and cefpodoxime proxetil, 200 mg every 12 hours. § Parenteral regimens are as follows: TMP-SMX, 160 to 800 mg every 12 hours; ciprofloxacin, 200 to 400 mg every 12 hours; ofloxacin, 200 to 400 mg every 12 hours; pentamicin, 1 mg/kg of body weight every 8 hours; ceftriaxone, 1 to 2 g every day; ampicillin, 1 g every 6 hours; imipenem-cilastatin, 250 to 500 mg every 6 to 8 hours; ticarcillin-clavulanate, 3.2 g every 8 hours; and aztreonam, 1 g every 8 to 12 hours. TMP-SMX, trimethoprim-sulfamethaxozole. Modified from Stamm W, Hooton TM: Management of urinary tract infections in adults. N Engl J Med 329:1328, 1993.
Data suggest that the incidence of scar formation after acute pyelonephritis may be as high as 37%.
Principles of Disease As in adults, E. coli is the predominant pathogen. Age-related differences are recognized: In older boys, Proteus often is isolated during UTI, whereas in newborn children, Klebsiella often is the causative agent. The route of infection also is agerelated. In the newborn period, it is thought that the bacteria are blood-borne (and often associated with generalized sepsis). In the older age group, as in adults, the ascending urethral route is primarily responsible for generating infection of the urinary tract.
Clinical Features UTI often is overlooked in children because of inappropriate emphasis on classic signs and symptoms, with little regard to age variables. Nonspecific findings should be considered the
rule and not the exception (Table 97-5). Pyelonephritis may be present without overt symptoms. A UTI in a febrile patient usually indicates pyelonephritis. An elevated BUN level or hypertension in a child older than 2 months strongly suggests bilateral hydronephrosis or advanced renal parenchymal disease.
Neonates Generalized septicemia often is the major manifestation of neonatal UTI.2 Classically, feeding difficulties, irritability, and sluggishness are seen in this age group. Bacteremia is present in nearly 50% of cases.9
Age 1 Month to 3 Years This age group has the most deceptive clinical presentation of UTI. Nonspecific findings are typical: fever, irritability, abdominal pain, vomiting, and failure to thrive. Occasionally, gross hematuria may be present.
1305
Table 97-5 Signs and Symptoms of Urinary Tract Infection by Age Group INFANT
PRESCHOOLER
SCHOOL-AGE CHILD
Poor feeding Vomiting Jaundice Hypothermia Fever Failure to thrive Sepsis
Poor feeding Vomiting Diarrhea Fever Strong-smelling urine
Abdominal pain Vomiting Strong-smelling urine Fever Enuresis Increased frequency of urination Dysuria Urgency
Fever Enuresis Increased frequency of urination Dysuria Urgency Costovertebral angle tenderness (flank pain)
High clinical suspicion? • White • Uncircumcised male • Fever ≥2 days • Fever ≥39° C
• Male 220, DBP 1–2 minutes. May repeat or double >120, or MAP† >130 mm Hg labetalol every 20 minutes to a maximum dose of 150 mg. 3. SBP 230 or DBP Step 1: Give 10 mg labetalol‡ IV push 121–140 mm Hg over 1 to 2 minutes. May repeat or double labetalol every 10 minutes to a maximum dose of 150 mg, or give the initial labetalol bolus and then start a labetalol drip at 2–8 mg/minute. Step 2: If BP is not controlled by labetalol, consider sodium nitroprusside. 10 mg labetalol‡ IVP. May repeat or 4. SBP 180–230 or double labetalol every 10–20 minutes DBP 105– 120 mm Hg to a maximum dose of 150 mg or give initial labetalol bolus and then start a labetalol drip at 2–8 mg/minute. *All initial blood pressures should be verified before treatment by repeating reading in 5 minutes. † As estimated by one-third the sum of systolic and double diastolic pressures. ‡ Labetalol should be avoided in patients with asthma, cardiac failure, or severe abnormalities in cardiac conduction. For refractory hypertension, alternative therapy with sodium nitroprusside or enalapril may be considered. BP, blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; SBP, systolic blood pressure; TPA, tissue plasminogen activator.
stroke.69 Normally normotensive stroke patients with low blood pressure or normally hypertensive stroke patients with low or even low-normal blood pressure should be given a fluid bolus to try to increase cerebral perfusion. This is especially important in patients who present in a dehydrated state. If initial fluid challenge is ineffective, the patient may require vasopressor therapy (e.g., with dopamine) to gradually increase MAP and improve cerebral perfusion.67 Acute Drug Therapy. To date, only the use of intravenous t-PA has been approved by the U.S. Food and Drug Administration (FDA) for treatment of patients with acute ischemic stroke. These recommendations initially were based on the results of the National Institute of Neurological Disorders and Stroke
Fibrinolytic Therapy for Acute Ischemic Stroke:
BOX 99-4 Inclusion and Exclusion Criteria
Inclusion Criteria 1. Age 18 years or older 2. Clinical diagnosis of ischemic stroke causing a measurable neurologic deficit 3. Time of symptom onset well established to be less than 180 minutes before treatment would begin Exclusion Criteria 1. Evidence of intracranial hemorrhage on noncontrast head CT 2. Only minor or rapidly resolving stroke symptoms 3. High clinical suspicion of subarachnoid hemorrhage even with normal CT findings 4. Active internal bleeding (e.g., gastrointestinal or urinary bleeding within last 21 days) 5. Known bleeding diathesis, including but not limited to: ■ Platelet count 15 seconds 6. Within 3 months of intracranial surgery, serious head trauma, or previous stroke 7. Within 14 days of major surgery or serious trauma 8. Recent arterial puncture at noncompressible site 9. Lumbar puncture within 7 days 10. History of intracranial hemorrhage, arteriovenous malformation, or aneurysm 11. Witnessed seizure at stroke onset 12. Recent acute myocardial infarction 13. On repeated measurements, systolic pressure >185 mm Hg or diastolic pressure >110 mm Hg at time of treatment, requiring aggressive treatment to reduce blood pressure to within these limits CT, computed tomography.
(NINDS) trial, although subsequent analysis of other studies has supported its use.70-73 Concern has emerged regarding the safety of the use of t-PA in community practice.74 However, a meta-analysis of non–trial-related use of t-PA (n = 2639) in community practice demonstrated efficacy and safety for t-PA similar to those reported in the NINDS trial.75 The current recommendation for recombinant t-PA (rt-PA) is that it be administered intravenously at a dose of 0.9 mg/kg to a maximum of 90 mg (10% of the dose given as a bolus followed by an infusion lasting 60 minutes). Treatment must be initiated within 3 hours of the onset of ischemic symptoms in patients who meet strict inclusion and exclusion criteria (Box 99-4). Intravenous t-PA is not recommended when the time of stroke onset cannot be ascertained reliably, including strokes recognized on awakening. In addition, caution is warranted in treating patients with large strokes (NIHSS score of 20 or higher) or early CT changes from a recent major infarction (e.g., acute hypodensity or mass effect), because they are at increased risk for symptomatic hemorrhage.76 Earlier studies have demonstrated the importance of adhering to the inclusion-exclusion criteria established by the NINDS trial.74,77 A recent study suggests that patients with mild or rapidly resolving symptoms may still benefit from the use of intravenous tPA.78 The use of intravenous t-PA beyond the 3-hour window
1343
Management of Intracranial Hemorrhage. The patient with a potential ICH requires rapid assessment and transport to a care center with CT scanning capability and intensive care management facilities. Out-of-hospital management is similar to that for ischemic stroke. The circumstances surrounding the event, as well as other concomitant medical conditions, also should be ascertained. The initial level of consciousness, GCS score, any gross focal deficits, difficulty with speech, clumsiness, gait disturbance, or facial asymmetry should be noted. Supportive care involving attention to airway management and perfusion is of the highest priority. Patients with hemorrhagic stroke are more likely to have an altered level of consciousness that may rapidly progress to unresponsiveness requiring emergent endotracheal intubation. Intravenous access should be established and cardiac monitoring should be initiated. Evaluation of blood glucose and appropriate dextrose and naloxone administration should be considered in any patient with altered mental status. Considerable disagreement exists regarding optimal blood pressure management in the patient with ICH. Hypertension may cause deterioration by increasing ICP and potentiating further bleeding from small arteries or arterioles. On the other hand, hypotension may decrease CBF, thereby worsening brain injury. In general, recommendations for treatment of hypertension in patients with ICH are more aggressive than those for patients with ischemic stroke. The current consensus regarding management of ICH is to recommend antihypertensive treatment with parenteral agents for systolic pressures higher than 160 to 180 mm Hg or diastolic pressures higher than 105 mm Hg. Treatment for lower pressures remains controversial.23,60 Nitroprusside is the agent most commonly recommended because it can provide rapid and consistent lowering of the blood pressure to the desired level, and adjustments can be rapidly made. Nitroprusside provides a rapid onset, is titratable, and has no effect on mental status. Disadvantages include the need for careful monitoring (ideally with an indwelling arterial catheter) and the theoretical risk of worsening the hemorrhage due to the vasodilatory effects of nitroprusside on cerebral vessels. Labetalol is another therapeutic option. More recently, nicardipine has been proposed as an optimal antihypertensive agent in the setting of cerebrovascular emergencies, owing to both its good titration profile, which may create less need for adjunctive antihypertensive drugs, and its favorable cerebral hemodynamic effects.93 Hyperventilation and diuretics such as mannitol have been used when ICH is complicated by signs of progressively increasing ICP, clinical deterioration associated with mass effect, or impending uncal herniation.23 These interventions should not be used prophylactically. Mannitol moves fluid from the intracranial compartment, thereby reducing cerebral edema. Although this effect may be temporarily helpful in the acute setting, the brain tissue will reequilibrate and rebound swelling can occur and worsen the patient’s clinical status. The effectiveness of mannitol in the setting of ICH is questionable.94-96 Hypertonic saline is being investigated as an alternative agent.97 Use of steroids in cerebral hemorrhage, once a common practice, appears to be harmful and is not recommended. Other experimental modalities include barbiturate coma and hypothermia. Seizure activity can cause neuronal injury, elevations in ICH, and destabilization of an already critically ill patient. In addition, nonconvulsive seizure may contribute to coma in up to 10% of patients in a neuro–intensive care unit.23 Seizure prophylaxis (fosphenytoin 18 mg/kg) should be considered for patients with ICH, especially those with lobar hemorrhage. Surgery is not beneficial in most cases of ICH. Selected patients with sizable lobar hemorrhage and progressive neuro-
Chapter 99 / Stroke
has not been demonstrated to be of clinical benefit, although a meta-analysis of these studies suggests a beneficial effect in a specific subset of patients.27-29,70 Intra-arterial thrombolysis is an alternative treatment for eligible patients presenting beyond the 3-hour time window but within 6 hours of symptom onset.37,79,80 In the Prolyse in Acute Cerebral Thromboembolism II (PROACT II) study, patients with middle cerebral artery strokes were 58% more likely than those who received a placebo to have little or no neurologic disability at 90 days when treated with prourokinase up to 6 hours after stroke onset.80 In addition, the use of intra-arterial thrombolysis in patients with posterior circulation strokes and those unresponsive to initial treatment with intravenous t-PA also is being evaluated and has shown favorable outcomes.37,81-84 A variety of mechanical clot retrieval devices are being investigated. The best studied is the Mechanical Embolus Removal in Cerebral Ischemia (MERCI) retrieval device. This corkscrew-like device has been shown to be successful in recanalizing intracranial lesions of the internal carotid artery and has demonstrated improved clinical outcomes and survival compared with patients without successful recanalization of the artery.85 Successful recanalization has been achieved when performed within an 8-hour window of symptom onset.86,87 In 2004, the FDA cleared the MERCI retrieval device for use in the setting of acute ischemic stroke. Previous studies have focused on the use of antiplatelet agents in acute ischemic stroke. Data from two large trials involving almost 40,000 patients indicate that early use of aspirin in patients with acute ischemic stroke who were not treated with a fibrinolytic agent was associated with a small but significant reduction in rates of stroke recurrence and mortality.88 These studies in combination suggest a number needed to treat of 77 (i.e., 77 stroke patients would need to be treated with daily aspirin therapy to prevent a poor outcome, such as death, dependency at discharge or at 6 months after stroke, in 1 patient). The need for acute administration of aspirin in the ED is unclear, because patients were given aspirin up to 48 hours after stroke onset. Aspirin should not be given for the first 24 hours in patients who have received a fibrinolytic agent, because this has been associated with an increased risk of ICH and death.89 The use of low-molecular-weight or unfractionated heparin is common in patients with acute ischemic stroke or TIAs, but its value is unproved. Some studies suggest that heparin may reduce the risk of subsequent ischemic stroke but increase the risk of hemorrhagic stroke. To date, no studies have definitively established the efficacy of anticoagulants in the management of acute ischemic stroke.90-92 However, heparin sometimes is considered in patients at high risk for stroke progression, including patients with crescendo TIAs or TIA due to a cardioembolic source, patients with a high-grade carotid artery stenosis, patients with posterior circulation TIA, and patients with evolving strokes. Heparin is recommended for the treatment of carotid and vertebral artery dissection unless a contraindication such as intracranial extension is present. If a dissection is diagnosed and the patient has no symptoms of ischemia, treatment with antiplatelet therapy alone may be an option.13 Heparin therapy should not be initiated in patients with suspected endocarditis or in any patient until a CT scan has ruled out intracranial bleeding. Owing to the lack of consistent evidence of efficacy, the most prudent course in the ED setting is to determine the need for heparin therapy in conjunction with the patient’s neurologist or the admitting physician. Other innovative approaches to stroke care, including mild to moderate hypothermia and early hemicraniectomy, are currently being investigated.
PART III ■ Medicine and Surgery / Section Seven • Neurology
1344
logic deterioration may benefit from surgical drainage. Surgery is more efficacious in patients with cerebellar hemorrhage. The clinical course in cerebellar hemorrhage is notoriously unpredictable. Patients with minimal abnormalities may experience sudden deterioration, with progression to coma and death, with little warning. For this reason, most neurosurgeons will consider emergent surgery for patients with cerebellar hemorrhage within 48 hours of onset. In cases of severe intraventricular hemorrhage or hemorrhages in the posterior fossae, the normal circulation of cerebrospinal fluid (CSF) can become interrupted, leading to the development of hydrocephalus. This condition is characterized by an abnormal rise in CSF volume. In such cases a ventricular catheter should be inserted by a neurosurgeon.
■ DISPOSITION Ischemic Stroke and Transient Ischemic Attacks “Stroke center” definitions have been proposed, and a national certification process for primary stroke centers is now under way despite considerable political controversy. In broad terms, institutional certification as a primary stroke center requires the establishment of a stroke infrastructure (such as a stroke team, stroke unit, patient care protocols, and support services including CT scanning and laboratory testing availability), as well as institutional administrative support and strong leadership.98 Additionally, recommendations also have been established for comprehensive stroke centers (CSCs). CSCs are expected to have the capability to provide the full spectrum of care to patients with stroke and other cerebrovascular diseases. More specifically, CSCs should offer advanced imaging modalities, perform surgical and endovascular interventions, and maintain a core infrastructure such as a stroke unit and stroke registry.99 The establishment of PSCs and CSCs is intended to improve outcomes for stroke patients by providing them with a high level of coordinated care. It has been recommended that patients with symptoms consistent with an acute stroke should be transported to emergency facilities capable of initiating fibrinolytic therapy within 1 hour of hospital arrival. At a minimum, this requires emergent CT capabilities, an institutional “acute stroke protocol,” and availability of a physician versed in the use of thrombolytic therapy. Intensive care monitoring and neurosurgery capabilities should be available within 2 hours of drug initiation, either at the treating hospital or by helicopter or ground transport to an appropriate health care facility.100 In most cases, once the diagnosis of an acute stroke or stroke syndrome is established and the patient is stabilized, hospitalization will be necessary for further evaluation and treatment. Patients may deteriorate over the first 24 hours and require close in-hospital monitoring. Most patients can be managed on a general medical or telemetry unit. Patients with large acute hemispheric strokes (associated with increased risk for herniation) or with significant posterior circulation–related changes and those treated with a fibrinolytic agent should be monitored in a step-down or intensive care unit for at least 24 hours. Some hospitals have specialized stroke units that provide organized, multidisciplinary care to stroke inpatients. A recent evidence-based medicine review amalgamated the results of 26 trials that compared the outcomes for more than 5500 stroke patients treated in either specialized stroke units or general
wards. This review found a 14% reduction in the odds of poststroke mortality for patients treated in a stroke unit compared with patients managed on a general ward (odds ratio [OR], 0.86; 95% confidence interval [CI]: 0.76 to 0.98). A similar reduction was found for the composite outcomes of death or institutionalized care (OR, 0.82; 95% CI: 0.73 to 0.92) and death or dependency (OR, 0.82; 95% CI: 0.73 to 0.92). All outcomes were independent of age, sex, or stroke severity, and evidence for increase in hospital length of stay by treatment in a stroke unit was lacking.101 In select cases, patients with multiple previous strokes who have been thoroughly evaluated or those who experience mild new episodes or have a completed stroke days to weeks after the event may be treated in the ED and discharged home after discussion with their primary care physician or neurologist. Close follow-up should be arranged. Patients with new-onset TIAs should be hospitalized for evaluation and workup owing to the substantial short-term risk of stroke and other adverse events.21 Exception can be made for patients with only minimal anterior circulation symptoms, for whom an extensive ED evaluation constitutes appropriate management. This evaluation must include a CT scan, a carotid Doppler study, MRA or CTA of the anterior circulation, and an echocardiogram (if indicated). A medically or surgically treatable cause for TIAs (e.g., high-grade carotid stenosis, mural thrombus) should be sought, which would require in-hospital treatment such as anticoagulation, stenting, or carotid endarterectomy. If the patient’s symptoms have completely resolved, results of the workup are negative, and close neurologic follow-up can be arranged, outpatient therapy may be appropriate. The decision to start the patient on an antiplatelet agent should be made in conjunction with the neurologist. The ABCD (age, blood pressure, clinical features, duration of TIA symptoms) score has been validated as a good predictor of future stroke risk in patients with TIA treated in the ED. ABCD scores range from 0 to 6, with a higher score indicating a higher future stroke risk; this scoring system is based on the following criteria: age 60 years or older = 1 point; systolic blood pressure greater than 140 mm Hg and/or diastolic greater than 90 mm Hg = 1 point; unilateral weakness = 2 points; speech disturbance without weakness = 1 point; symptom duration 10 to 59 minutes = 1 point; symptom duration 60 minutes or longer = 2 points. Patients with an ABCD score of 5 or 6 in the ED have a 30-day risk of stroke eight times that of patients with an ABCD score less than 5 (hazard ratio 8.01; 95% CI: 3.21 to 19.98).102
Hemorrhagic Stroke All patients with an acute hemorrhagic stroke in whom surgical intervention is a consideration should be admitted to an intensive care unit under the care of a neurologist or a neurosurgeon. If this is unavailable at the evaluating institution, the patient should be transported to an appropriate institution.
Acknowledgment The authors gratefully acknowledge the assistance of Stephen M. Davis, MPA, MSW, Co-Director of Clinical Research at the West Virginia University Department of Emergency Medicine, in the preparation of this chapter.
1345
KEY CONCEPTS Patients presenting with the signs and symptoms of an acute ischemic stroke within 3 hours of symptom onset should be evaluated for thrombolytic therapy within the NINDS-recommended time frames (see Table 99-3). ■ Carotid Doppler, MRA, or CTA studies are recommended before discharge of a patient with TIA from the ED. ■ Overly aggressive blood pressure management should be avoided in patients with acute ischemic stroke.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
■
Accurate time of symptom onset should be documented in all patients with stroke. ■ Assessment of gait is essential to rule out posterior circulation stroke in patients presenting with vertigo. ■ The possibility of carotid or vertebral dissection should be considered in young patients with stroke and in patients with headaches and neck pain with acute stroke.
Chapter 99 / Stroke
■
Chapter 100
Seizures
Evelyn H. Duvivier and Charles V. Pollack, Jr.
■ PERSPECTIVE A seizure is the clinical manifestation of excessive, abnormal cortical neuron activity. The physical manifestation depends on the area of brain cortex involved and, to a lesser extent, on the specific underlying abnormality. Patients who have recurring seizures without consistent provocation have epilepsy, although this term encompasses many disparate clinical syndromes. Seizures also may occur as a predictable response to certain toxic, pathophysiologic, or environmental stresses; these are reactive or secondary seizures, and patients who experience them do not have epilepsy. In the United States, 10% of people experience at least one seizure in their lifetime; the cumulative incidence of epilepsy is 3%.1 The evaluation of patients with seizures, whether ongoing or recent, in the emergency department (ED) may be complex and difficult. A careful history must be elicited to determine the presence of ictal events that represent epilepsy or exposure to ictogenic stimuli (e.g., alcohol, cocaine), significant underlying illness (e.g., meningitis, hypoxemia, hypoglycemia, intracranial mass), or contributing causes (e.g., sleep deprivation in an epileptic). The physical examination should focus on the identification of focal neurologic abnormalities, systemic illness, and signs of toxic exposure. If the patient continues to experience seizure activity, airway protection and abortive therapy must be provided. Laboratory and radiographic evaluation that is guided by historical and physical findings may be limited or unnecessary in some cases. Finally, the appropriate disposition of a patient presenting to the ED with a seizure or with a history of recent seizure requires an understanding of the underlying illness, likelihood of recurrence, indications for maintenance pharmacologic therapy, and state reporting regulations. In addition to the distinction between primary (epileptic) and secondary (reactive) seizures, many other classifications of ictal events have been proposed.2-7 Seizures are termed generalized or focal (partial) depending on their clinical manifestations. The former type of seizure results from the abnormal electrical event that simultaneously involves both cerebral hemispheres and is accompanied by loss of consciousness; in the latter, abnormal activity is limited to part of one cerebral hemisphere only. Generalized seizures usually are characterized by rhythmic, tonic-clonic muscle contractions, or convulsions, although nonconvulsive generalized seizures also occur. Partial seizures can be differentiated further into seizures during which cognition is maintained (simple partial) and seizures during which cognition is impaired (complex partial). The 1346
term cognition is defined as involving at least two of the five features—perception, attention, emotion, memory, and executive function8,9—and replaces the previously used term consciousness, which is both difficult to define and difficult to document. Finally, partial seizures may become generalized (partial with secondary generalization). Inexperienced witnesses may provide histories that are insufficient for accurate categorization of seizures. However, when an accurate history is available, secondary (reactive) seizures typically are generalized, not partial, in nature. The definitive differentiation among these classifications may require electroencephalogram (EEG) recording during the seizure, sometimes in association with simultaneous video recording. Seizures in children, as in adults, are classified as primary (idiopathic) and secondary (symptomatic or reactive). The term cryptogenic is used sometimes when seizures are thought to be secondary but no cause has been identified. The history is the most important diagnostic tool in evaluating seizures in children. The actual seizure activity usually is not observed, and the emergency physician must rely on a detailed and accurate history for diagnosis. Other important terms to describe ictal events include status epilepticus, in which seizures occur serially without an intervening return to a normal neurologic condition; spasm, which is a specific, debilitating seizure syndrome that occurs in infants; and myoclonus, which refers to rhythmic, shock-like muscle contractions also typical for specific seizure syndromes. The postictal period is an interval after a seizure, of variable duration, usually characterized by impaired consciousness but sometimes also marked by self-limited focal paralysis or neurogenic pulmonary edema.
■ PRINCIPLES OF DISEASE The pathophysiology of seizures at the neuronal level is incompletely understood, with most of what is known coming from animal studies in which either electrical or pharmacologic stimulation is applied directly to brain cortex. To produce generalized ictus, stimuli must be applied to both hemispheres simultaneously. Some studies show the concept of recruitment, which occurs when the initiating neurons’ abnormal, increased electrical activity activates adjacent neurons and propagates until the thalamus and other subcortical structures are recruited. The clinical seizure activity typically, but not always, reflects the focus of initiation.9-11
■ CLINICAL FEATURES Primary Seizures in Adults Primary ictal events in adults include events of genetic and of idiopathic origin. Onset is typically during childhood or adolescence, but occasionally idiopathic seizures may begin de novo in adulthood. Because idiopathic seizures are rare, a firsttime seizure in an adult requires a thorough ED evaluation. Focal seizures in adults may be classified as simple partial or complex partial. Simple partial seizures are limited in electrical focus to one cerebral hemisphere and do not cause loss of cognition. Although the specific function of the initiating neurons determines the clinical manifestation of the ictal event (i.e., motor, somatosensory, special sensory, autonomic, or psychic), such clinical manifestations are not sufficiently specific for anatomic localization without an EEG. Typical features of simple partial seizures include focal clonic movements; paresthesias; visual, auditory, olfactory, or gustatory experiences; sweating and flushing; dysphasia; a sense of déjà vu; or a sense of unwarranted fear.9,11 Motor signs, which by definition remain ipsilateral in simple partial seizures, may spread contiguously in a stepwise fashion (Jacksonian march) as neuron recruitment occurs in the motor cortex. There is generally no postictal state after a simple partial seizure. Complex partial seizures are ictal events that involve impairment of cognition, either at onset or evolving from focal activity. Amnesia for the ictal event is a consistent feature of complex partial seizures, although during the episode the patient may remain responsive to the surroundings. Complex partial seizures typically involve automatisms that are specific to the affected person, such as lip smacking, repeated swallowing or uttering verbal phrases, or picking at clothing. Complex partial seizures generally are associated with an aura,
1347
such as a specific smell, taste, visual hallucination, or intense emotional feeling. In contrast with those experiencing generalized seizures, these patients may continue with ongoing motor activity, such as driving an automobile, riding a bicycle, or playing a musical instrument (reactive automatisms), and they may react to their surroundings in a semiappropriate manner.9 Partial seizures may progress rapidly to generalized seizures. A postictal state is common after complex partial seizures and may persist for hours.9,11 Generalized seizures in adults may be convulsive or nonconvulsive. By definition, patients lose consciousness in a generalized seizure, and no aura is present. Some patients may experience a brief, vague prodrome or dysphoric state just before the ictal event. Convulsive generalized seizures are typified by the tonic-clonic, or grand mal, seizure, in which the patient loses consciousness, stiffens with generalized muscular hypertonus, and then rhythmically and violently contracts multiple, bilateral, and usually symmetrical muscle groups. The muscular force may be sufficiently vigorous to result in posterior shoulder dislocation or fractures of thoracic spine vertebral bodies; significant tongue and buccal injuries also may be incurred from biting with repeated jaw muscle contractions. Dysautonomia, including transient apnea, is a potential manifestation of convulsive generalized seizures; urinary incontinence is more common than fecal incontinence. A generalized convulsive seizure generally lasts 1 to 2 minutes and is followed by a postictal state, headache, and drowsiness that may persist for hours. This state must be differentiated in the ED from altered consciousness attributable to other causes. Nonconvulsive generalized seizures include absence, or petit mal, seizures; myoclonic seizures; tonic seizures; and atonic seizures. Absence seizures in adults are subclassified further as typical or atypical. Typical absence ictus is characterized by the sudden cessation of normal, conscious activity followed by a nonconvulsive, dissociative state that persists for a few seconds to several minutes before suddenly terminating. Eye movements, blinking, or automatisms may be present. There is no aura and no postictal state. If the seizure occurs midsentence, then the patient typically will resume speaking at precisely the point of interruption without awareness of the intervening event. Absence seizures typically begin in childhood but occasionally develop in adults. Atypical absence seizures are marked by more complicated motor signs, coexistence with other forms of generalized seizures, inconsistent postictal confusion, and irregular EEG abnormalities.9,11 Atonic seizures are characterized by focal diminution of muscle tone (limb or head) or generalized loss of postural tone in which the head falls forward and then the body slumps to the ground (“drop attack”), usually landing buttocks first (although this can vary depending on the axis of gravity at the time of the fall). Recovery occurs immediately, and there is either no loss or an extremely brief loss of consciousness. In myoclonic-atonic seizures, a brief (less than 100 msec) myoclonic jerk of muscle group of variable anatomy occurs before the episode of atonia.9 Because typically no postictal state is associated with these episodes, an altered level of consciousness in a patient presenting to the ED after an atonic or myoclonicatonic seizure should prompt an investigation for head trauma or a toxic or metabolic abnormality. Status epilepticus is defined as serial seizure activity without interictal recovery or prolonged, continuous seizure activity. Traditionally, status epilepticus was defined as seizure activity lasting longer than 30 minutes, which is the estimated duration necessary for neuronal injury.14,15 However, because an isolated tonic-clonic seizure rarely lasts more than a few minutes, an operational definition of status epilepticus has been
Chapter 100 / Seizures
What prompts such initiation is unclear. Proposed mechanisms include disruption of normal structure—whether congenital, maturational, or acquired (as with scar tissue)—and disruption of local metabolic or biochemical function. The latter mechanism is better elucidated because the roles of two neurotransmitters—acetylcholine, which is excitatory to cortical neurons, and γ-aminobutyric acid (GABA), which is inhibitory—have been more fully characterized. In sensitive neurons, such as those at an ictogenic focus, subtle changes in the local concentrations of these neurotransmitters can produce sustained membrane depolarization, ultimately followed by local hyperpolarization and recruitment. Recruitment may follow contiguous paths or extend along diverse integrated circuits that are deep and cross the midline.9-11 When the ictal discharge extends below the cortex to deeper structures, the reticular activating system in the brainstem may be affected, altering consciousness. In generalized seizures, the focus often is subcortical and midline, which explains the prompt loss of consciousness and bilateral involvement.9,12 Seizures typically are self-limited; at some point, the hyperpolarization subsides, and the electrical discharges from the focus terminate. This termination may be related to reflex inhibition, loss of synchrony, neuronal exhaustion, or alteration of the local balance of acetylcholine and GABA in favor of inhibition.9,12 The systemic manifestations of convulsive ictal activity include hypertension, tachycardia, tachypnea, and hyperglycemia from sympathetic stimulation. With more prolonged convulsions, skeletal muscle damage, lactic acidosis, and, rarely, frank rhabdomyolysis may ensue.9,10,13 Autonomic discharge and bulbar muscle involvement may result in urinary or fecal incontinence, vomiting (with significant aspiration risk), tongue biting, and airway impairment.
PART III ■ Medicine and Surgery / Section Seven • Neurology
1348
advocated as either a continuous seizure lasting more than 5 minutes, or more than two discrete seizures without intervening recovery of consciousness.16 Although it is recognized that the underlying cause of status epilepticus is the predominant factor determining morbidity and mortality, prolonged seizure activity does cause neuronal injury and therefore warrants prompt abortive therapy. Furthermore, status epilepticus may become refractory to treatment over time.14,17 The most common cause of status epilepticus is discontinuation of anticonvulsant medication. This situation may be compounded by barbiturate withdrawal when phenobarbital therapy is abruptly withdrawn. Patients may present for the first time with a primary seizure disorder in status. Many other causes of status epilepticus have been documented9,15,18,19 (Box 100-1). After prolonged status epilepticus or after incomplete treatment, the patient may exhibit very subtle manifestations of continued seizure activity, such as small-amplitude twitching of the extremities or jerking of the eyes, or any visible motor activity may cease while seizure activity detectable on the EEG continues.9,20-22 Recognition of the latter scenario, termed nonconvulsive status epilepticus, requires a high index of clincial suspicion. Prompt treatment is essential; otherwise, neuronal damage can result. All classes of primary seizures may recur sporadically, randomly, or predictably. Cyclic recurrence has been reported with awakening, sleep deprivation, emotional or physical stress, alcohol, and menses, among other factors. Seizures also may be triggered by specific senory stumuli, the most common of which is visual stimulation in the form of flashing lights, such as strobe lights, television, and video games.9,23 Seizures also can be caused by auditory, gustatory, tactile or startle triggers that are specific to the affected person. The most common cause of recurrent primary seizures is medication noncompliance.9,11
Reactive Seizures in Adults Reactive or secondary seizures do not result from genetic or idiopathic causes. The conditions that cause reactive seizures may be static (e.g., anatomic scarring), progressive (e.g., degenerative cortical disorders), or transient (e.g., acute electrolyte derangements).
Seizures Caused by Metabolic Derangements Hypoglycemia is a common metabolic cause of reactive seizures. Ictal activity can occur when the plasma glucose level is less than 45 mg/dL, although some patients may manifest neurologic disturbances even at higher levels.24 A rapid bedside glucose test should be an integral part of the ED evaluation of the patient exhibiting seizure activity. Convulsive and nonconvulsive seizures and generalized and partial seizures all may occur during hypoglycemia.24 Patients at the extremes of age are particularly susceptible to glucose stress during acute illness. Hypoglycemia also may result from insulin reaction, a deliberate insulin or hypoglycemic agent overdose, alcoholism, poor nutrition, and sepsis. Hypoglycemic seizures respond to glucose therapy; anticonvulsants are unnecessary. Cation derangements, notably hyper- and hyponatremia, hypomagnesemia, and hypocalcemia, are other common metabolic causes of ictal activity.25,26 Hypo-osmolar and hyperosmolar states can precipitate seizures. Disorders of sodium—the primary cation in the extracellular fluid compartment and the primary determinant of serum osmolarity—are most common. Hyponatremia is the most frequently identified electrolyte disorder in hospitalized patients, and sodium levels less than 120 mEq/L often are associated with seizures.27,28 The rate at
Etiology of Status Epilepticus: Common
BOX 100-1 Causative Disorders
Metabolic Disturbances Hepatic encephalopathy Hypocalcemia Hypoglycemia or hyperglycemia Hyponatremia Uremia Infectious Processes CNS abscess Encephalitis Meningitis Withdrawal Syndromes Alcohol Antiepileptic drugs Baclofen Barbiturates Benzodiazepams CNS Lesions Acute hydrocephalus Anoxic or hypoxic insult Arteriovenous malformations Brain metastases Cerebrovascular accident Chronic epilepsy Eclampsia Head trauma Intracerebral hemorrhage Neoplasm Neurosurgery Posterior reversible leukoencephalopathy Remote structural injury Intoxication Bupropion Camphor Clozapine Cyclosporine Flumazenil Fluoroquinolones Imipenem Isoniazid Lead Lidocaine Lithium Metronidazole Theophylline Tricyclic antidepressants CNS, central nervous system.
which the sodium level decreases, and not the absolute magnitude of the decrease, determines the risk for neurologic manifestation.27,29 Correcting hyponatremia should be undertaken slowly in the ED, to avoid osmotic demyelination. If seizures are persistent, administration of hypertonic (3%) saline may be indicated.27 Hypernatremia will result in cerebral edema and seizures in the setting of rapid elevation of serum sodium to greater than 160 mEq/L or during aggressive correction of subacute hyponatremia.25,29 Hypercalcemia reduces neuronal excitability and rarely causes seizures; significant hypocalcemia (7.5 mEq/L) is associated, however, with ictal activity. Hypocalcemia may result
Seizures Caused by Infectious Diseases Infectious diseases can cause seizures independent of a purely febrile mechanism. These seizures generally result from primary central nervous system (CNS) infections but occasionally arise from other septic sources. The most important ictogenic infections are meningitis, encephalitis, cerebral abscess, cerebral parasitosis, and human immunodeficiency virus (HIV) disease and associated opportunistic infections, with their protean CNS manifestations. Seizures can occur as a result of the acute inflammatory response or as sequelae to bacterial or viral meningitis. During the acute course of their illness, up to 40% of patients with meningitis will have at least one seizure; this is more common at the extremes of age but is rarely associated with residual epilepsy.9,37,38 By contrast, seizures occur in up to 50% of patients with a brain abscess, and epilepsy develops in 40% of the survivors.9,39 After meningitic seizures are terminated with benzodiazepines, phenytoin should be initiated temporarily.9 Viral meningoencephalitides, the most common of which are caused by the herpes simplex virus, also are associated with seizures. These seizures may be generalized or partial, often recur during the acute phase of the illness, and may persist after the illness resolves.9 The parasitic CNS infection neurocysticercosis is relatively common in areas of the United States in which the population includes immigrants from Latin America. Seizures complicate 50 to 90% of neurocysticercosis cases.40 Latent syphilis also may be a cause of adult-onset seizures. Primary HIV disease of the CNS, its attendant infectious and mass lesion complications, such as from toxoplasmosis and lymphoma, and the demyelinating infection progressive multifocal leukoencephalopathy constitute a significant cause of generalized and partial seizures.41 Choosing an antiepileptic drug for an HIV-infected patient with seizures should be done in consultation with infectious disease and neurology specialists, because of the well-recognized increase in adverse effects of and interactions between antiepileptic drugs and antiviral medications.
Seizures Caused by Drugs and Toxins The list of substances reported to cause seizures either as an idiosyncratic side effect of therapeutic use or as a manifesta-
1349
tion of toxic overdose is extensive.18,42 The recognition of this etiologic category is crucial in the ED setting. Seizure activity should be viewed as a dire sign of toxicity and may herald the onset of life-threatening instability. Seizures may occur after therapeutic doses of antimicrobials, cardiovascular agents, neuroleptics, and sympathomimetics.43 Seizures also may result from exposure to plant toxins, insecticides and rodenticides, and hydrocarbons. Certain over-thecounter supplements also have been associated with seizures, either alone or through adverse interactions with prescription medications.44,45 The most common drug-associated and toxinassociated seizures occur, however, in conjunction with illicit drugs, such as cocaine, amphetamines, and phencyclidine; with overdoses of anticholinergic agents, such as cyclic antidepressants and antihistamines; as a manifestation of withdrawal from ethyl alcohol and sedative-hypnotics; and with toxic levels and deliberate overdoses of diverse medications including aspirin, theophylline, meperidine, isoniazid, lithium, and the anticonvulsants phenytoin and carbamazepine.42,46 Standard ED therapeutic measures usually are effective for management of toxic seizures. In some cases, specific antidotal therapy is available, such as alkalinization for cyclic antidepressant and salicylate overdoses, pyridoxine (vitamin B6) for isoniazid overdose, and hemodialysis for salicylate and lithium toxicity. Because of its prevalence in urban ED patient populations, cocaine toxicity warrants special mention.47 Seizures may occur after isolated recreational use or chronic abuse, after overdose, and in “body packers” and “body stuffers.”48 Cocaine-related seizures may be a manifestation of direct CNS toxicity or an indirect result of hypoxemia from cardiac toxicity.49 Seizures in cocaine-intoxicated patients must be managed as part of the overall toxic reaction, which often includes high fever, rhabdomyolysis, and cardiac arrhythmias. A benzodiazepine is the appropriate initial therapeutic agent. Ethyl alcohol is another common toxic cause of seizures. Ictal events may occur with acute inebriation but are more common during withdrawal from alcohol.50 Withdrawal seizures typically are generalized, are recurrent, and may begin within 6 hours of cessation of or decrease in alcohol consumption. Through a phenomenon termed kindling, the risk and severity of seizures increase with each episode of withdrawal. Kindling implies that with each episode of alcohol withdrawal, the seizure threshold is lower. Alcoholic patients with seizures must be evaluated for other related, concomitant ictogenic problems (e.g., hypoglycemia, electrolyte derangements, head trauma, co-ingestion of other toxins, pregnancy). The preferred treatment for alcohol-associated seizures is with benzodiazepines; these drugs substitute for the GABA-enhancing effect of ethanol in the CNS.
Seizures Caused by Trauma Post-traumatic seizures can occur acutely as a result of blunt or penetrating head trauma or as a post-traumatic sequela. Immediate post-traumatic seizures occur within 24 hours of injury. Epidural, subdural, and intracerebral hematomas and traumatic subarachnoid hemorrhages all can be acutely ictogenic, particularly as intracranial pressure rises. More often, however, the onset of seizure activity is delayed for at least several hours. Early post-traumatic seizures occur within 1 week of injury, whereas late post-traumatic seizures occur after 1 week. Immediate and early post-traumatic seizures are more common in children than in adults, and children also are more likely than adults to present in status epilepticus in the immediate or early post-traumatic phase.51,52 Within the first year after significant head trauma, the incidence of seizures is at least 12 times that in the general population.53
Chapter 100 / Seizures
from hypoparathyroidism, renal failure, or acute pancreatitis and typically is associated with hypomagnesemia, which also can precipitate seizures, particularly at serum levels less than 1 mEq/L. Hypomagnesemia is seen most often as a result of poor nutrition, especially in alcoholic patients. Patients with significant hypomagnesemia or hypocalcemia should be treated empirically for both disorders.25,26 Nonketotic hyperosmolar hyperglycemia also is associated with seizure activity. Partial seizures, including partial status, predominate. These seizures do not respond to anticonvulsants; rather, they are best managed with gradual correction of fluid deficits and glucose excess.30-32 Seizures may complicate the course and treatment of renal failure.33 Ictal activity occasionally complicates uremic encephalopathy, is more common in conjunction with acute fluid and electrolyte shifts during dialysis (dialysis disequilibrium syndrome), and can occur as a complication of immunosuppressive therapy after renal transplantation. Thyroid hormones lower seizure threshold, and consequently Graves’ disease and thyrotoxicosis may occasionally manifest as seizures, including status epilepticus.9,34,35 Seizures also occur with hypoparathyroidism as a direct result of secondary hypocalcemia.36
PART III ■ Medicine and Surgery / Section Seven • Neurology
1350
The severity of head injury correlates with the likelihood of post-traumatic seizures. The incidence of seizures after injury with neurologic deficit without dural violation is 7 to 39%; when the dura is disrupted, the incidence is 20 to 57%.53 Imaging studies should be performed urgently because the likelihood of identifying significant cerebral edema, cerebral contusions, hematomas, and depressed skull fractures is relatively high.52,54 Antiepileptic drugs are recommended for prophylaxis against post-traumatic seizures occurring within the first 7 days after severe brain injury in adults; however, they have not been shown to be effective in preventing late posttraumatic seizures.51,55,56
Seizures Associated with Malignancy or Vasculitis Seizures are a common manifestation of primary and metastatic CNS neoplasms. They also may complicate cancer treatment as a result of postsurgical scarring or chemotherapyrelated electrolyte derangements, hematologic abnormalities, or immunosuppression. Although any CNS tumor can be ictogenic, low-grade and slow-growing primary neoplasms (e.g., well-differentiated gliomas and oligodendrogliomas) are implicated most commonly.57 In such cases, seizures, which most often are partial with secondary generalization, may be the initial clinical manifestation. A new-onset seizure in a patient with a non-CNS primary malignancy, such as melanoma and tumors of the lung, breast, colon, germ cells, or renal cells, should prompt consideration of CNS metastasis and warrants neuroimaging. Seizures also may be the presenting manifestation of CNS vasculitis in patients with systemic lupus erythematosus and polyarteritis nodosa. These commonly are complex partial seizures that give a general indication of the acute inflammatory focus. Sometimes secondary generalization follows.58
Seizures Caused by Strokes, Arteriovenous Malformations, and Migraines Ischemic or hemorrhagic stroke is the cause of new-onset seizures in 40 to 54% of elderly patients.59 The overall incidence of seizures with stroke ranges from 4 to 15%; more than one half occur within the first week after stroke. The incidence of epilepsy after stroke is 4 to 9%.60,61 Seizures that occur acutely with stroke are thought to result from local metabolic alterations in the CNS; these events are transient, and the seizures often are focal and self-limited. Seizures that develop later are more likely to be generalized. Seizures also occur in conjunction with unruptured cerebrovascular aneurysms and arteriovenous malformations.9 Arteriography may be required to confirm the diagnosis; unruptured arteriovenous malformations are easier to detect on an enhanced cranial computed tomography (CT) scan than are smaller, unruptured aneurysms. Seizures also may arise in concert with vascular headaches, either coincidentally, by migrainous activation of an epileptic focus, or after vascular headache has induced cerebral infarction that becomes an epileptic focus.62
Seizures Caused by Degenerative Disease of the Central Nervous System In approximately 5% of patients with multiple sclerosis, focal or generalized seizures develop during the course of their illness. These seizures must be differentiated from the tonic spasms that may occur in multiple sclerosis. Patients with demyelinating disease also should be evaluated for the other types of reactive seizures.9
CNS degeneration associated with aging, including dementia and Alzheimer’s disease, increases the risk of reactive seizures and epilepsy.59,63 The elderly also are more likely to have other ictogenic problems (e.g., stroke, brain neoplasm, toxic and metabolic disturbances, blunt head trauma from falls). Maintenance treatment of elderly patients with epilepsy often is complicated by drug-drug interactions, and breakthrough seizures may result even when patients are compliant. Although the incidence of unprovoked seizures increases after age 60 years, ED management of these patients must include a thorough evaluation for causes of secondary seizures.59
Gestational Seizures Seizures associated with pregnancy are divided into two categories: gestational epilepsy, in which hormonal and metabolic changes exacerbate underlying epilepsy or adversely influence serum levels of anticonvulsants, and eclampsia or toxemia, which is a gestational hypertensive encephalopathy manifested by seizures, hypertension, coma, proteinuria, and edema. For the former, antiepileptic therapy should be tailored by the patient’s neurologist and obstetrician to maximize seizure control and minimize the risk of teratogenic effects.64 Convulsive generalized status epilepticus in pregnancy jeopardizes both mother and fetus. The definitive treatment for eclamptic seizures is magnesium sulfate. Simultaneous reduction in blood pressure using hydralazine, labetalol, or nifedipine is recommended.65
Psychogenic Nonepileptic Seizures Psychogenic seizures, or pseudoseizures, are functional events that may be associated with alterations in consciousness, abnormal movements and behaviors, and autonomic changes. They are not the result of abnormal CNS electrical activiy. Psychogenic seizures may be primarily motor and mimic convulsive generalized seizures, including refractory status epilepticus, or they may be nonconvulsive and mimic either absence or complex partial seizures. Although certain features of convulsive psychogenic seizures may suggest the diagnosis, no clinical criteria are 100% specific; simultaneous video and EEG recordings may be required to confirm the diagnosis.9 The ED evaluation of these patients is difficult, because seizures and pseudoseizures can coexist. All but obviously functional abnormalities should be treated as for true ictus pending formal neurologic evaluation. Many patients with pseudoseizures are not deliberately attempting to mislead the examining physician. The long-term treatment of patients with confirmed pseudoseizures may include direct confrontation, intensive psychotherapy, and a placebo.
Postictal States The postictal state that follows most generalized seizures typically is characterized by a decreased level of arousal and responsiveness, disorientation, amnesia, and headache. These conditions may persist for only a few minutes or for many hours and may not be consistent from seizure to seizure. The most important consideration in ED management of the postictal state is to monitor and investigate the altered mental status after a seizure; otherwise, dangerous underlying metabolic or toxic abnormalities may be overlooked. At the minimum, airway positioning maneuvers, pulse oximetry, rapid glucose determination, and cardiac rhythm monitoring are necessary. Two unusual postictal manifestations may provoke particular consternation in the ED: postictal paralysis and neurogenic pulmonary edema. Postictal paralysis, or Todd’s paralysis, may
■ DIAGNOSTIC STRATEGIES First-Time Seizures The essential components of the seizure evaluation in the ED setting are discussed in Chapter 16. An accurate and thorough history of the ictal event, any known or potential precipitants or exposures, and the patient’s medical problems must be obtained. A thorough physical examination, including a complete neurologic examination, is essential. Any identified focal neurologic deficits must be monitored for progression or resolution. Appropriate ancillary studies may be comprehensive, but if precipitants (e.g., hypoglycemia, intoxication) are known, studies may be comparatively limited. Although the Academy of Neurology recommends neuroimaging, by either CT or magnetic resonance imaging (MRI), for all adults presenting with an apparent unprovoked first seizure,68 the usefulness of emergent imaging depends on the clinical situation. An emergent cranial CT scan is indicated when a serious structural lesion is suspected on clinical grounds, including presence of a new focal deficit, persistent altered mental status, fever, recent trauma, persistent headache, history of cancer, anticoagulant use, suspicion or known history of AIDS, age older than 40 years, and partial-complex seizure.69 A reason able approach may be to obtain scans on an outpatient, followup basis in patients who have recovered completely from the ictal event and in whom no apparent cause has been elucidated; if reliable follow-up care is unlikely or even questionable, the CT scan should be obtained in the ED to ensure its completion. In patients with known epilepsy and recurrent seizures, the same considerations apply, but in addition, epileptic patients with a change in seizure pattern, prolonged postictal state, or persistent abnormal mental status should be scanned in the ED. The decision to initiate anticonvulsant therapy after a single seizure depends on the etiology of the seizure. Seizures due to structural lesions, such as stroke, tumor, or head injury, are likely to recur and may warrant antiepileptic medication. However, such patients also are likely to be admitted to the hospital if the lesion is newly discovered. For patients with a single unprovoked seizure, most authorities now agree that antiepileptic therapy should not be initiated; rather, the patient should be discharged with referral for neurologic consultation.9,70 The rationale for this approach is threefold. First, the
1351
diagnosis may be incorrect, especially if the seizure-like activity was not witnessed by ED personnel. It is estimated that 20 to 25% of patients diagnosed as having seizures are found to have been misdiagnosed, with the most frequent alternative diagnoses being cardiovascular and psychopathologic.71-74 Second, the patient may not have a recurrent seizure. It is estimated that less than 50% of patients who have had a single unprovoked seizure will experience a recurrent seizure within 2 years.9,70,75,76 Furthermore, whereas treatment decreases the risk of early recurrent seizure, it does not affect longterm prognosis of epilepsy,70,77,78 nor does it have an impact on patient quality of life,79 with the exception of driving limitations, which are prolonged in a patient with a recurrent seizure. Third, antiepileptic medications have side effects that may outweigh the benefit of treatment, especially in women of childbearing age, owing to the teratogenic risk of anti epileptic drugs, and in patients with liver, kidney, or hematologic disorders and patients already receiving multiple medications.
Recurrent Seizures The initial approach to stabilization of a patient with a known seizure disorder does not differ from that for a new-onset patient; this includes a rapid blood glucose determination. The most common cause of seizures in a patient with a diagnosed seizure disorder is noncompliance with medications.9 However, supratherapeutic and toxic levels of some anticonvulsants, such as carbamazepine, phenytoin, and lamotrigine, whether attained chronically or after acute overdose, can also cause seizures.80-83 Accordingly, it is prudent to check the serum drug level, if this test is available, before giving a full loading dose of anticonvulsants to patients on long-term therapy. Meanwhile, a thorough history and physical examination should focus on intercurrent illness or trauma, drug or alcohol use, potential adverse drug-drug interactions with anticonvulsants, a recent change in anticonvulsant dosing regimens, and any change in ictal pattern or characteristics. Clinical indications should dictate the selection of other laboratory or radiographic tests.
■ DIFFERENTIAL CONSIDERATIONS Even when a “seizure” is witnessed in the ED, other abnormal movements and states of consciousness can be confused with ictal activity. The most common misdiagnoses are cardiovascular (syncope) and psychogenic, but other considerations in the differential diagnosis include hyperventilation and breathholding, certain toxic and metabolic states, transient ischemic attacks, narcolepsy, and some movement disorders.9,11,72,73,84 Syncope—whether vasodepressive (e.g., “vagal” or micturition syncope), orthostatic, or arrhythmogenic (e.g., paroxysmal ventricular tachycardia or fibrillation, long Q-T syndrome)— may be confused with ictal events; differentiating among these may be particularly difficult when episodes are recurrent— hence the consideration “fit versus faint.” Generally, ictal tonic-clonic movements are much more forceful and are more prolonged than the “twitches” sometimes associated with fainting. In addition, most seizures are characterized by a postictal state, which, with the important exception of atonic drop attack ictus, is not a feature of syncope. The cause of an unwitnessed, unprovoked loss of consciousness with a fall, after which the patient presents to the ED, may be difficult to classify. Retrograde amnesia suggests an ictal diagnosis. Hyperventilation syndrome can be associated with mood disturbances, paresthesias, and posturing movements of the distal extremities. Manifestations of toxic and metabolic disorders
Chapter 100 / Seizures
follow generalized or complex partial seizures and is a focal motor deficit that may persist up to 24 hours. Weakness of one extremity or a complete hemiparesis may occur; in the latter case, the patient must be safely restrained to avoid falls caused by a combination of weakness and diminished responsiveness resulting from the postictal state. Todd’s paralysis is associated with a high likelihood of an underlying structural cause for the seizure. Neurogenic pulmonary edema is a relatively common, although often subclinical, complication of any structural CNS insult, including seizure, trauma, and hemorrhage.66 Neurogenic pulmonary edema probably is caused by centrally mediated sympathetic discharge and generalized vasoconstriction, coupled with increased pulmonary capillary membrane permeability. After a seizure, neurogenic pulmonary edema can be confused clinically and radiographically with aspiration pneumonia. Neurogenic pulmonary edema is managed with ventilatory support, including positive end-expiratory pressure and other aggressive measures to reduce intracranial pressure. Hypoxia or other clinical evidence of pulmonary congestion after a seizure should prompt consideration of neurogenic pulmonary edema.67
PART III ■ Medicine and Surgery / Section Seven • Neurology
1352
that may mimic ictus include delirium tremens and alcoholic blackouts, the alteration in consciousness associated with hypoglycemia and acute intermittent porphyria, the buccolingual spasms of phencyclidine intoxication, and the tonic spasms caused by tetanus, strychnine, and camphor.11 Nonictal CNS events, such as transient ischemic attacks, transient global amnesia, and atypical migraines, may manifest in a manner similar to that in absence seizures and postictal states such as Todd’s paralysis. Carotid sinus hypersensitivity, which can even result from a too-tight necktie, may cause drop attacks.85 Narcolepsy (recurrent irresistible daytime sleepiness), especially when it occurs with cataplexy (sudden falls), may be associated with hallucinations and abnormal movements. It can be differentiated from seizure activity by the history and response to stimulation. Movement disorders, such as hemiballismus and tics, usually are associated with other neurologic problems. Finally, dissociative states such as fugue and panic attacks can be confused with seizures. An EEG is an appropriate diagnostic option in unclear cases.
■ MANAGEMENT Immediate Management ED management of a patient experiencing a seizure begins with active, anticipatory airway management. In generalized ictus, the gag reflex is suppressed, and vomiting often is complicated by aspiration of gastric contents. The patient should be placed in a left lateral decubitus position, and any dentures should be removed. A bite-block should be placed to protect the tongue and allow access for suctioning. If the patient is persistently apneic or if an unavoidable airway threat is present, endotracheal intubation is warranted for definitive protection. A benzodiazepine should be used as an induction agent in the hope that its action may terminate the seizure or obviate the need for tracheal intubation. Trismus may necessitate use of a short-acting neuromuscular blocking agent to facilitate intubation. In general, the first-line pharmacologic agent for treatment of any active seizure is a parenteral benzodiazepine. Because benzodiazepines directly enhance GABA-mediated neuronal inhibition, they affect clinical and electrical manifestations of seizures. Benzodiazepines are effective in terminating ictal activity in a majority of patients and have been shown to be more effective than phenytoin in terminating status epilepticus.86,87 Although phenobarbital appears to be as effective as the benzodiazepine lorazepam in terminating status epilepticus, the associated high risk of hypoventilation and hypotension limits its use as a first-line agent.87 Benzodiazepines available in the ED setting include diazepam (Valium), lorazepam (Ativan), and midazolam (Versed) (Table 100-1). All three may be used in patients of any age, and all share the following characteristics: rapid efficacy (seconds to minutes), relatively short duration of anticonvulsant action, a sedative effect, and the potential for hypotension and respiratory depression. Lorazepam has emerged as the drug of choice for the initial management of epilepsy, because it terminates seizure rapidly (within 2 minutes) and has a longer duration of action (4 to 6 hours, compared with 20 minutes for diazepam), thus necessitating fewer repeat doses.88-91 For this reason, it also is the preferred agent for control of alcohol withdrawal seizures.92,93 Lorazepam is available intramuscularly and as a sublingual preparation for outof-hospital control of seizures in children.9 An advantage of diazepam is that it is in liquid form at room temperature and is therefore available premixed in resuscitation kits, and it can be administered quickly and without a need for reconstitution
by the intravenous, endotracheal, or intraosseous route. It also is available in a rectal gel formulation. Its onset of action with intravenous administration is within 10 to 20 seconds, but a 50% chance of recurrent seizure within 2 hours if diazepam is used alone has been noted.87 Midazolam’s onset of action is within 1 minute; it is available in both intranasal and buccal formulations,94 and among the benzodiazepines it has the least cardiovascular effect.91 Second-line abortive anticonvulsant therapy consists of phenytoin (Dilantin) and phenobarbital. Phenytoin reduces the repetitive firing of action potentials through sodium channel blockade, thereby stabilizing neuronal membranes.9 Phenytoin neither sedates patients nor causes respiratory depression, but rapid intravenous administration of phenytoin in its propylene glycol diluent may cause hypotension and cardiac bradydysrhythmias, as well as local vascular injury, including venous thrombosis and localized tissue necrosis (purple glove syndrome).95-97 It should therefore be administered through a 20 gauge or larger line proximal to the forearm, at a rate no faster than 50 mg/minute, and the patient should have a cardiac monitor. Phenytoin’s onset of action is within 10 to 30 minutes, and intravenous administration typically requires at least 20 minutes.98 The duration of action is approximately 24 hours. Continued benzodiazepine dosing is appropriate until phenytoin achieves adequate brain levels. Fosphenytoin is a water-soluble prodrug form of phenytoin, with a more physiologic pH. Its main advantages are that it is not likely to precipitate during intravenous infusion and that it also can be administered intramuscularly, although the volume required for full loading by the intramuscular route may be in the range of 20 mL or more.99 Although fospheny toin can be infused more rapidly, the time to therapeutic concentration of the active drug is the same as for intravenous phenytoin.100 The hemodynamic advantages of fosphenytoin over intravenous phenytoin have not proved to be significant.101,102 Its use is most appropriate when intravenous access is not obtainable or when the intravenous line is of small gauge, as is often the case in children or the elderly.88 If levels of phenytoin or phenobarbital are subtherapeutic in a patient already being treated for seizures, loading doses can be given intravenously; alternatively, an adjusted oral dosing schedule can be prescribed to boost the serum level over 24 to 48 hours. Oral loading of phenytoin is associated with fewer adverse events than those noted with loading with either intravenous phenytoin or fosphenytoin,101 but its use may be limited when therapeutic activity is required urgently. Phenobarbital is similar to benzodiazepines in that it binds to and enhances the inhibitory neurotransmitter GABA, thereby acting as a CNS depressant that decreases ictal and physiologic cortical electrical activity. Sedation and depression of respiratory drive and blood pressure must be anticipated,91 and, for this reason, nonsedating phenytoin is preferred. The onset of action of phenobarbital is within 15 to 30 minutes, and the duration of action is 48 hours. Valproic acid administered intravenously has recently been recognized as a safe and effective treatment for seizures, especially in patients with allergies to phenytoin, the elderly, and patients with cardiorespiratory instability who might be at increased risk of adverse events from phenytoin.103-105 Valproic acid administered intravenously has been shown to be as effective as phenytoin given intravenously in patients with benzodiazepine-refractory status epilepticus, with fewer cardiopulmonary side effects.106 Hyperammonemic encephalop athy after valproic acid loading has been reported and should be evaluated by determination of serum ammonia level in a patient who does not regain consciousness after seizure resolution.19,107 Appropriate ED dosing regimens for the
1353
Table 100-1 Drugs Used in the Abortive Treatment of Status Epilepticus in the Emergency Department BRAND NAME
ADULT DOSE
COMMENTS
Diazepam
Valium
Lorazepam
Ativan
May be given per rectum in pediatrics (0.3–0.5 mg/kg) Preferred benzodiazepine owing to its longer duration of action
Midazolam
Versed
Phenytoin
Dilantin
5–10 mg IV every 10 minutes, up to 30 mg per 8-hour period 0.1 mg/kg IV (usually 4 mg in adult); may repeat in 10 minutes, then 0.01–0.1 mg/kg per hour infusion 0.2 mg/kg IV bolus, then 0.05–0.6 mg/kg per hour infusion 20 mg/kg IV at 200 mm H2O in nonobese patients, >250 mm H2O in obese patients) measured by lumbar puncture in the recumbent position or by epidural or intraventricular pressure monitoring 3. Normal CSF chemistry (low CSF protein acceptable) and cellularity 4. Intracranial disease (including venous sinus thrombosis) ruled out by appropriate investigations 5. No metabolic, toxic, or hormonal cause of intracranial hypertension CSF, cerebrospinal fluid; ICHD-II, International Classification of Headache Disorders, 2nd ed. Available at http://ihs-classification.org/en/.
ischemia of the visual pathways. These episodes can be followed by prolonged periods of visual loss, which can become permanent in up to 10% of patients.89 Patients also may complain of nausea, vomiting, and dizziness. The physical examination will reveal papilledema and visual field defects, including an enlarged blind spot initially, followed by loss of peripheral vision. Occasionally, a sixth nerve palsy is noted. Diagnosis. The diagnosis of IIH should not be made without neuroimaging and measurement of ICP. The diagnostic criteria are listed in Box 101-4. Treatment. Predisposing factors (e.g., discontinuation of implicated medications) should be corrected. Symptomatic treatment often includes lowering ICP and managing the headache. Acetazolamide (a carbonic anhydrase inhibitor) can be used to decrease CSF production alone or with a loop diuretic such as furosemide. Steroids also have been used, although their mechanism of action is unclear. Prolonged therapy is problematic, and rebound IIH often occurs when doses are tapered. Repeated lumbar punctures can be attempted, but most patients find this approach objectionable. In patients with impending visual loss or incapacitating symptoms, placement of a ventricular shunt or optic nerve sheath fenestration may be indicated.
Post-traumatic Headache Headache is the most common symptom after minor head injury. It often is part of a complex syndrome that can include dizziness, fatigue, insomnia, irritability, memory loss, and difficulty with concentration. The prevalence of headache with post-traumatic syndrome is not known, because most patients are not admitted for this condition. There are approximately 2 million closed head injuries per year, and post-traumatic headache (PTHA) occurs in an estimated 30 to 50% of patients with these injuries.90 Acute PTHA develops hours to days after the injury and resolves within 3 to 6 months.91 Chronic PTHA may last from several months to years and may mimic other forms of headache, including tension and migraine headaches. The presence of headache, dizziness, or nausea on initial presentation is strongly associated with the development of chronic PTHA.92 Patients in whom PTHA develops after minor head injuries have normal findings on neurologic examination and neuroim-
aging studies. The pathophysiologic mechanism for their symptoms is unclear and may have both anatomic and functional components. Most patients are more concerned about the cause of the headache rather than the headache itself. Treatment is symptomatic. For acute PTHA, analgesics such as acetaminophen or NSAIDs are adequate for pain control. For chronic PTHA, treatment must be individualized depending on the type of headache and associated symptoms the patient is experiencing. Novel therapies, such as antidepressants and beta-blockers, may be effective in selected patients.
Acute Glaucoma Patients with acute angle closure glaucoma present with sudden onset of severe pain localized to the affected eye that may radiate to the ear, sinuses, teeth, or forehead.69 Visual symptoms, including blurriness, halos around lights, and scotomas, typically are present, and many patients also experience nausea and vomiting. The underlying pathophysiologic mechanism is congenital narrowing of the anterior chamber angle that, under certain conditions, closes, resulting in a significant rise in intraocular pressure (IOP). Episodes can be precipitated by entering a low-light environment such as a movie theater, with resultant pupillary dilatation, or by the use of medications such as mydriatics (e.g., for dilated ocular examination), sympathomimetics (e.g., pseudoephedrine), or agents with anticholinergic properties (e.g., antiemetics, antihistamines, antipsychotics, antidepressants).86,93 Physical examination reveals a red eye with a fixed, middilated pupil, corneal clouding, and shallow anterior chamber. The diagnosis is confirmed by demonstrating markedly elevated IOP in the range of 60 to 90 mm Hg (normal is less than 21 mm Hg). Treatment includes topical miotics, topical beta-blockers, oral carbonic anhydrase inhibitors (e.g., acetazolamide, 250 mg four times daily), intravenous osmotic agents (e.g., mannitol), and prompt referral to an ophthalmologist. The potential for diagnostic confusion between acute glaucoma, iritis, and cluster headache must be recognized. Although cluster headache may arise with pain, nausea, and a red eye, vision is not affected and the pupil generally is small and the eyelid is ptotic (from an oculosympathetic paresis).94 Acute iritis also arises with a painful red eye, but only acute angle closure glaucoma is associated with markedly elevated IOP.
Post–Dural Puncture Headache Principles of Disease and Pathophysiology. Headache is the most common complication of lumbar puncture, occurring in up to 40% of patients.95 The incidence is highest in the 18- to 30-year age group, but this complication is uncommon in young children and in adults older than 60. Although the onset often is immediate, patients may not report symptoms for several days. In a majority of affected persons, the duration of headache is less than 5 days.95,96 The cause of post–dural puncture headache (PDPH) is not entirely clear. The most likely explanation is a persistent CSF leak that exceeds CSF production, resulting in CSF hypotension. If sufficient CSF is lost, the brain descends in the cranial vault when the patient assumes the upright position, leading to increased traction on the pain fibers.97 Thus, the headache is characteristically positional and increases with the upright position and decreases with recumbency. The amount of time a patient remains recumbent after lumbar puncture does not appear to affect the incidence of headache.12
Intracranial Infection Headache is common among patients with intracranial infections, including meningitis, brain abscess, encephalitis, and acquired immunodeficiency syndrome. The severity and type of headache vary depending on the specific infection. With acute bacterial meningitis, the patient often has a severe bursting headache that rapidly increases in severity over a short period.102 These patients typically have significant meningismus, with both Kernig’s and Brudzinski’s signs. With viral meningitis, patients also may complain of severe headache and nuchal rigidity, but the course is more indolent than with bacterial meningitis. The severity of headache associated with encephalitis depends on the type of virus involved. For example, the headache is usually mild with mumps encephalitis. With herpes simplex infection, however, the headache is abrupt and severe and frequently is associated with confusion, fever, altered level of consciousness, seizures, and focal neurologic signs. Patients with brain abscess often have headache as their presenting complaint.103 As the infection progresses, vomiting, focal neurologic signs, and depressed level of consciousness typically develop. Headache is a frequent complaint in patients with human immunodeficiency virus infection and can be caused by a number of conditions, including aseptic meningitis, toxoplasmosis, cryptococcal or tuberculous meningitis, and cytomegalovirus encephalitis. In a majority of cerebral infections, the mechanism of head pain includes meningeal irritation and increased ICP. In addition, headache may be a general reaction to fever or the toxic products of the infecting agent.104
Hypertensive Headache Contrary to common belief, hypertension is not an important cause of headache, and the occurrence of headache and hypertension in the same patient is often coincidental.105 Whether some patients with mild to moderate hypertension suffer from headache caused by elevated blood pressure is uncertain. The
1365
rate of blood pressure increase is more important as a cause of headache than the absolute blood pressure value. Diastolic pressures lower than 130 mm Hg are rarely the cause of headache.72 Nonetheless, the association of headache with severe hypertension is well documented. Acute, severe headache is a prominent symptom of hypertensive encephalopathy, and most patients have blood pressure readings in the range of 250/150 mm Hg. Other conditions include headache secondary to toxic agents (e.g., drug-induced hypertension), pheochromocytoma, and eclampsia. The headache of severe hypertension typically is diffuse and is worse when the patient awakes in the morning and gradually subsides over the course of the day.105 Treatment is directed at lowering the blood pressure; in most cases, the headache is relieved within 24 hours. In patients with hypertensive encephalopathy, the headache may persist for days until brain edema has resolved.
Cervicogenic Headache Cervicogenic headache refers to headache originating from disorders of the neck. Diagnosis is based on the presence of one of the following three distinct sets of symptoms106: 1. Unilateral headache triggered by movements of the head or neck or certain head positions 2. Unilateral headache triggered by pressure on the neck 3. Unilateral headache spreading to the neck or possibly the ipsilateral shoulder or arm Many of these headaches are reported after a whiplash injury. Even though neck structures play a primary role in the pathophysiology of some headaches, clinical patterns indicating a neck-headache relationship have not been adequately defined.
Medication-Induced Headache Medication use, abuse, or withdrawal can be a cause of headache, and the term medication-induced headache is used to describe these conditions. Medication-induced headache is underdiagnosed and often difficult to manage.107 Although not well understood, it tends to occur in patients with a primary headache disorder (e.g., migraine, tension-type) who use immediate-relief medications, often in excessive quantities.108 Medications that have been implicated include NSAIDs, aspirin or acetylsalicylic acid (ASA), acetaminophen, barbitur ate-analgesic combinations plus caffeine with or without codeine, opioids, caffeine, and ergotamine. A key factor in medication overuse headache is preemptive use of drugs, in anticipation of—rather than for—headache.33 Women are affected more commonly than men, and the most frequently affected age group is that of persons between 30 and 40 years.109 The headache itself is variable and may be accompanied by asthenia, nausea, anxiety, depression, and difficulty with concentration. Typically, it is worse on awakening in the morning and after physical exertion. The symptomatic medication that leads to the development of this disorder initially provides some pain relief to the patient, but over time tolerance develops, and larger doses are required to obtain symptomatic improvement.109 Treatment typically requires complete withdrawal of the medication being overused, to achieve long-term results. In addition, these patients require a comprehensive education and follow-up program with pharmacologic, dietary, and behavioral components.109
Chapter 101 / Headache
Certain factors have been implicated as causes of PDPH, including the size or diameter of the spinal needle, the orientation of the bevel during the procedure, and the amount of fluid withdrawn. Smaller-diameter needles cause less leakage, and it is postulated that inserting the needle with the bevel up (i.e., bevel pointing up when the patient is in the lateral position) minimizes damage to the dural fibers. Using atraumatic needles or pencil-point needles (e.g., Whitaker98 or Sprotte99) also has been shown to reduce significantly the incidence of PDPH.100,101 Clinical Features. PDPH typically is bilateral, throbbing, and exacerbated by the upright position. Associated signs and symptoms include neck stiffness; nausea; vomiting; auditory disturbances, including tinnitus and hearing loss (hypoacusis); and ocular symptoms, including blurred vision and diplopia.97 Treatment. Most PDPHs resolve spontaneously within a few days with bedrest, adequate hydration, and mild analgesics. For persistent headaches, methylxanthine agents have been found to help some patients. Oral caffeine (300 mg every 4 to 6 hours), caffeine sodium benzoate (500 mg in 1 L of fluid), or theophylline (300 mg PO every 8 hours) may be effective.95 For severe headaches lasting longer than 24 hours, an epidural blood patch (autologous blood clot) relieves the headache in the majority of patients.101
1366
PART III ■ Medicine and Surgery / Section Seven • Neurology
Trigeminal Neuralgia Trigeminal neuralgia is a painful unilateral affliction of the face, characterized by brief electric shock–like (lancinating) pains limited to the distribution of one or more divisions of the trigeminal nerve. Pain is commonly evoked by trivial stimuli (e.g., washing, shaving, smoking, talking, brushing the teeth) but also may occur spontaneously.110 Individual attacks are brief, lasting a few seconds to less than 2 minutes, and are stereotypic in the individual patient. The lightning-like pains and unilateral grimaces characteristic of trigeminal neuralgia led to the designation of the term tic douloureux.111 The diagnosis is straightforward in most patients on the basis of clinical criteria. However, because these symptoms also can be caused by an underlying mass lesion, CT or MRI is indicated in previously undiagnosed patients and when sensory loss or motor dysfunction is present. Several drugs have been effective in treating trigeminal neuralgia, including carbamazepine, phenytoin, and baclofen; however, approximately 30% of patients fail to respond to medical therapy.111 In these patients, surgical management, by alcohol or glycerol injection or microvascular decompression, may be indicated.110,112
They typically are occipital in location and may increase in severity with mounting sexual excitement. Their duration can be from minutes to hours. Occasionally, some patients experience a sudden, explosive headache that occurs during orgasm. In these patients, SAH should be ruled out.113–115
High-Altitude Headache Headache is one of the cardinal manifestations of acute mountain sickness and can occur at altitudes higher than 5000 feet above sea level in unacclimatized persons. The headache is throbbing in nature, located in the temporal or occipital areas, and probably is caused by a mild increase in ICP secondary to brain swelling.116 It is worse at night or in the early morning and exacerbated by the Valsalva maneuver or bending forward.117 Other findings associated with high-altitude illness include fatigue, nausea, vomiting, dizziness, insomnia, and an altered mental status. Pulmonary edema and cerebral edema develop in severe cases. The treatment for these conditions includes supplemental oxygen and descent to a lower altitude.
KEY CONCEPTS
Cough and Exertional Headache In some patients, severe headache can be provoked by rapid increase in intra-abdominal pressure such as coughing, sneezing, laughing, heavy lifting or exertion, and the Valsalva maneuver. The pain starts within a few seconds of the precipitant and typically is brief when associated with cough but can last as long as 24 hours when associated with exertion. The headache is bilateral and throbbing in nature and in a majority of patients resolves spontaneously without persistent neurologic symptoms (e.g., neck stiffness or photophobia). In some patients, the headache may be secondary to structural lesions, especially in the posterior fossa113; therefore, all previously undiagnosed patients require CT, or preferably MRI, followed by lumbar puncture to rule out intracranial disease including SAH. For patients with recurrent benign exertional headache, treatment includes avoidance of the underlying triggering mechanism and use of analgesics as necessary. For patients with exertional headache, NSAIDs including indomethacin have been effective.114
Coital Headache Coital cephalgia is a recurrent, benign headache associated with sexual activity and is more common in men than in women. Different types have been described, including headaches that occur before, during, or immediately after orgasm.
■
Headache is a common presenting complaint in the ED. The goal of evaluation in this setting is to distinguish between benign primary headache disorders and the more serious and potentially lifethreatening secondary causes of headache. ■ A majority of patients do not have abnormal neurologic findings; therefore, the key to a successful diagnosis is a thorough and systematic history. ■ Patients with the following headache presentations are at risk for serious underlying disease: sudden explosive headache; first or “worst-ever” headache; new-onset headache after age 50 years; headache associated with papilledema, alteration in or loss of consciousness, or focal neurologic symptoms; headache after head trauma; subacute headache with increasing frequency or severity; headache associated with fever, cancer, or immunosuppression; and headache triggered by exertion, sexual activity, or Valsalva maneuver.118 ■ The need for diagnostic studies is dictated by the suspected secondary cause of headache.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 103
Brain and Cranial Nerve Disorders
Brian A. Stettler
This chapter discusses cranial nerve problems, cerebral venous thrombosis, and multiple sclerosis—neurologic disorders that often provide significant diagnostic and therapeutic challenges in management in the emergency department (ED) setting (Table 103-1).
■ TRIGEMINAL NEURALGIA Perspective Trigeminal neuralgia, or tic douloureux, is a syndrome featuring painful paroxysms in one or more distributions of the trigeminal nerve. Trigeminal neuralgia is relatively uncommon, with an annual incidence of 4 to 13 cases per 100,000 population.1,2 It is more common in women than in men, with a female-tomale ratio of 1.7 : 1. Affected persons typically are between 50 and 69 years of age, and symptoms occur more frequently on the right side of the face.3
Pathophysiology Trigeminal neuralgia is an idiopathic disorder, although significant evidence points to vascular compression of the trigeminal nerve root in many cases. This compression commonly is caused by a tortuous arterial or venous loop in the posterior fossa, an arteriovenous malformation, or rarely a tumor. In surgical case series, vascular compression of the trigeminal nerve root is found in 80 to 90% of cases.4,5 Of note, however, structural lesions are not found in all patients with trigeminal neuralgia.6
Clinical Features Trigeminal neuralgia manifests with unilateral facial pain, typically characterized as lancinating paroxysms of pain in the lips, teeth, gums, or chin. The pain of trigeminal neuralgia commonly is associated with physical triggers such as chewing, brushing the teeth, shaving, washing or touching the affected area of the face, swallowing, or exposure to hot or cold temperature in the affected area. The maxillary and mandibular divisions of the trigeminal nerve are most commonly involved; rarely, the ophthalmic division alone is involved. Patients tend to experience the pain in clustered episodes that last a few seconds to several minutes. The attacks can occur during the day or night but rarely arise during sleep.6,7
Diagnostic Strategies A careful history and physical examination should be performed to rule out other painful facial conditions including odontogenic infections, sinus disease, otitis media, acute glaucoma, temporomandibular joint disease, and herpes zoster. Patients who lack local pathologic findings to explain the painful syndrome require a very careful neurologic examination. The presence of a neurologic deficit should prompt suspicion of a structural lesion, such as aneurysm, tumor, or other intracranial lesion such as from multiple sclerosis (MS). Of note, 2 to 4% of patients with trigeminal neuralgia also have MS.8 Patients with normal findings on the head and neck examination and no neurologic deficits who have episodic, unilateral facial pain associated with nonpainful triggers are likely to have trigeminal neuralgia.
Management Since the 1960s the medical treatment of choice for trigeminal neuralgia has been use of the anticonvulsant carbamazepine. The purported effectiveness of this treatment is, however, based on uncontrolled studies, and the mechanism of action of anticonvulsant therapy for trigeminal neuralgia is unclear. The true efficacy of medical therapy is difficult to assess owing to a very high rate of spontaneous remission. Nonetheless, carbamazepine appears to be an effective and well-tolerated agent for treatment of trigeminal neuralgia. The initial dosage of carbamazepine is 100 mg twice daily; this dose is then increased to three times daily after one week. The dose may be increased by 100 mg per day, up to a maximum of 1200 mg per day. A complete blood count and liver function studies should be performed periodically in these patients to monitor for hematologic and hepatic side effects. Additional agents that have been used for treatment of trigeminal neuralgia include phenytoin, baclofen, valproate sodium, lamotrigine, and gabapentin. None of these drugs have been shown to be more effective than carbamazepine.7 Surgical management has been a theraputic option since the 1950s. Surgical procedures include both peripheral approaches and central procedures. Peripheral strategies include medication injection and cryotherapy techniques designed to temporarily block, or permanently ablate, branches of the peripheral trigeminal nerve. Although these procedures are relatively effective initially, recurrence is common. Repeated nerve 1379
1380
PART III ■ Medicine and Surgery / Section Seven • Neurology
Table 103-1 The Cranial Nerves: Normal Function and Pathologic Considerations CLINICAL FUNCTION RELEVANT TO EMERGENCY MEDICINE
PATHOLOGIC FEATURES
POSSIBLE CAUSES
Cranial nerve I: Olfactory nerve
Sense of smell
Unilateral anosmia
Cranial nerve II: Optic nerve
Vision
Unilateral vision loss
Trauma: Skull fracture or shear injury interrupting olfactory fibers traversing the cribriform plate Tumor: Frontal lobe masses compressing the nerve Trauma: Traumatic optic neuropathy Tumor: Orbital compressive lesion Inflammatory: Optic neuritis (MS) Ischemic: Ischemic optic neuropathy Trauma: Herniation of the temporal lobe through the tentorial opening causing compression and stretch injury to the nerve Ischemic: Especially in diabetes Microvascular ischemic injury to nerve causes extraocular muscle paralysis but usually is papillary-sparing (often painful) Vascular: Intracranial aneurysms may press on the nerve, leading to dysfunction Myasthenia gravis can lead to atraumatic ocular muscle palsy Trauma is the most common cause of nerve dysfunction
CRANIAL NERVE
Cranial nerve III: Extraoculomotor function via Oculomotor nerve motor fibers to levator palpebrae, superior rectus, medial rectus, inferior rectus, inferior oblique muscles Pupillary constriction via parasympathetic fibers to constrictor pupillae and ciliary muscles
Ptosis caused by loss of levator palpebrae function Eye deviated laterally and down Diplopia Dilated, nonreactive pupil Loss of accommodation
Cranial nerve IV: Trochlear nerve
Motor supply to the superior oblique muscle
Cranial nerve V: Trigeminal nerve
Motor supply to muscles of mastication and to tensor tympani Sensory to face, scalp, oral cavity (including tongue and teeth) Motor supply to the lateral rectus muscle
Inability to move eye downward and laterally Diplopia Patients tilt head toward unaffected eye to overcome inward rotation of affected eye Partial facial anesthesia Trauma: Episodic, lancinating facial pain Facial bone fracture may injure one associated with benign triggers section, leading to area of facial such as chewing, brushing teeth, anesthesia light touch Tic douloureux
Cranial nerve VI: Abducens nerve
Cranial nerve VII: Facial nerve
Motor supply to muscles of facial expression Parasympathetic stimulation of the lacrimal, submandibular, and sublingual glands Sensation to the ear canal and tympanic membrane
Cranial nerve VIII: Hearing and balance Vestibulocochlear nerve
Inability to move affected eye Tumor: Lesions in the cerebellopontine laterally angle Diplopia on attempting lateral gaze Any lesion, vascular or otherwise, in the cavernous sinus may compress nerve Elevated intracranial pressure (ICP): Because of its position and long intracranial length, increased ICP from any cause may lead to injury and dysfunction of the nerve Hemifacial paresis: Lower motor neuron: Lower motor neuron lesion leaves Infection (viral): The likely cause of Bell’s entire side of face paralyzed palsy Upper motor neuron lesion leaves Lyme disease: The most common cause of bilateral cranial nerve VII palsy in areas forehead musculature where Lyme disease is endemic functioning Bacterial infection extending from otitis Abnormal taste media Sensory deficit around ear Intolerance to sudden loud noises Upper motor neuron: Stroke, tumor Unilateral hearing loss Tumors: Acoustic neuroma Tinnitus Mimics Ménière’s disease, perilymphatic Vertigo, unsteadiness fistula
1381
Table 103-1 The Cranial Nerves: Normal Function and Pathologic Considerations—cont’d
PATHOLOGIC FEATURES
POSSIBLE CAUSES
Cranial nerve IX: General sensation to posterior Glossopharyngeal third of tongue nerve Taste for posterior third of tongue Motor supply to the stylopharyngeus
Clinical pathology referable to the nerve in isolation is very rare Occasionally painful paroxysms beginning in the throat and radiating down the side of the neck in front of the ear but behind the mandible
Brainstem lesion Glossopharyngeal neuralgia
Cranial nerve X: Vagus nerve
Unilateral loss of palatal elevation: Patients complain that on drinking liquids, the fluid refluxes through the nose Unilateral vocal cord paralysis: Hoarse voice
Brainstem lesion Injury to the recurrent laryngeal nerve during surgery
Downward and lateral rotation of the scapula and shoulder drop
Trauma to the nerve
Tongue deviations: Upper motor neuron lesion causes the tongue to deviate toward the opposite side Lower motor neuron lesion causes the tongue to deviate toward the side of the lesion, and the affected side atrophies over time
Stroke or tumor can cause upper motor neuron lesion Amyotrophic lateral sclerosis (ALS) can cause bilateral lower motor neuron lesion with atrophy Metastatic disease to the skull base may involve the nerve
Cranial nerve XI: Spinal accessory nerve Cranial nerve XII: Hypoglossal nerve
Motor to striated muscles and muscles of the pharynx, larynx, and tensor (veli) palatini Motor to smooth muscles and glands of the pharynx, larynx, thoracic and abdominal viscera Sensory from larynx, trachea, esophagus, thoracic and abdominal viscera Motor supply to the sternocleidomastoid and trapezius muscles Motor supply to the intrinsic and extrinsic muscles of the tongue
blocks are not recommended owing to a high risk of permanent facial anesthesia. Central procedures can be divided into percutaneous approaches and open approaches. Percutaneous destruction of the trigeminal ganglion can be done by means of radiofrequency ablation, thermal ablation, glycerol injection, or balloon microcompression. These procedures carry the risk of corneal anesthesia, oculomotor paresis, or masticatory weakness.9 Open surgical management is the surgical option of choice in most treatment centers. Open surgical procedures include microvascular decompression of the nerve with or without partial ablation. Although the open microvascular decompression procedure has proved to be very effective, with pain relief achieved in 80 to 95% of patients, the surgery is associated with the risk of significant complications, including hearing loss, facial anesthesia, cerebrospinal fluid leak, brainstem or cerebellar injury, headaches, meningitis, and death.10,11 Gamma knife radiosurgery, a minimally invasive, precision-directed stereotactic radiosurgery, also has been associated with good outcomes. This highly specialized technique requires extremely sophisticated stereotactic radiofrequency equipment and is available only in specialized centers.12,13
Disposition Patients with suspected trigeminal neuralgia should be referred for specialty evaluation. Patients with a neurologic deficit require urgent imaging studies, typically magnetic resonance imaging (MRI), to rule out a mass or vascular abnormality.
KEY CONCEPTS ■
Patients with unilateral, intermittent, lancinating facial pain without abnormalities on physical examination are likely to have trigeminal neuralgia. ■ Carbamazepine is the first-line agent for medical treatment. ■ Patients who do not tolerate or whose pain is refractory to medical management may be candidates for microvascular decompression or ablation.
■ FACIAL NERVE PARALYSIS Perspective The acute onset of facial nerve paralysis often will prompt an ED visit, when early diagnosis and therapy can improve the patient’s chance for recovery of function of the facial nerve. Facial nerve paralysis of acute onset affects approximately 20 to 25 persons per 100,000 population per year, without geographic, gender, or race predilection.14,15
Principles of Disease The facial nerve innervates the muscles of facial expression and the muscles of the scalp and external ear, in addition to the buccinator, platysma, stapedius, stylohyoid, and posterior belly of the digastric muscles. The sensory portion of the nerve supplies the anterior two thirds of the tongue with taste and
Chapter 103 / Brain and Cranial Nerve Disorders
CRANIAL NERVE
CLINICAL FUNCTION RELEVANT TO EMERGENCY MEDICINE
PART III ■ Medicine and Surgery / Section Seven • Neurology
1382
sensation to portions of the external auditory meatus, soft palate, and adjacent pharynx. The parasympathetic portion supplies secretomotor fibers for the submandibular, sublingual, lacrimal, nasal, and palatine glands.16 The nerve originates from the pontomedullary junction of the brainstem. The nerve enters the internal auditory meatus with cranial nerve VIII. Within the temporal bone the facial nerve has four major branches: the greater and lesser superficial petrosal nerves, the nerve to the stapedius muscle, and the chorda tympani. The facial nerve exits the temporal bone at the stylomastoid foramen. The nerve then enters the parotid gland, where it divides to supply the muscles of facial expression.16,17
Pathophysiology Although a complete list of possibilities in the differential diagnosis for facial nerve paralysis would be a long one, the causes pertinent to emergency medicine can be grouped into three specific categories: infectious, traumatic, and neoplastic.
Infection Bell’s Palsy Bell’s palsy, also commonly called idiopathic facial paralysis, has long been postulated to have a viral cause. This disease entity is characterized by an abrupt onset of a lower motor neuron paresis that can progress over 1 to 7 days to complete paralysis. A prodromal illness is described by 60% of patients. Symptoms and signs frequently associated with the facial paresis include ear pain, a perception of sensory change on the involved side of the face, decreased tearing, an overflow of tears on the cheek (epiphora), abnormally acute hearing (hyperacusis), and an impairment or perversion of taste (dysgeusia).18 Treatment approaches can be medical or surgical. The primary medical therapies for Bell’s palsy center on reducing inflammatory changes to the nerve with corticosteroids and treating the presumed viral cause. If these therapies are unsuccessful then surgical decompression may be considered. The use of corticosteroids for Bell’s palsy has been controversial. The rationale for this application of steroid therapy is that edema of the nerve, confined within the facial canal, is thought to cause or contribute to the nerve injury. On the basis of this theory, most experts currently recommend a course of prednisone with an initial dose of 1 mg/kg per day for 7 to 10 days, with or without a short taper.14,17,19,20 The most definitive randomized, double-blind, placebo-controlled trial involving 496 patients showed an improvement in complete recovery of facial nerve function at 3 months from 64% with placebo to 83% with the use of prednisolone in a dose of 25 mg by mouth twice daily.21 Therapy should be started as soon as possible, ideally within the first 24 hours, but is still recommended for patients without contraindications who seek treatment within 1 week of symptom onset.19 A number of publications have advanced the belief that Bell’s palsy may be caused by herpesvirus infection. One study demonstrated herpes simplex virus type 1 DNA in the endoneural tissue of 11 of 14 patients with Bell’s palsy but not in that of control subjects.22 In a trial of prednisone and acyclovir in 99 patients, patients treated with prednisone and acyclovir had a more favorable recovery than that observed in patients receiving prednisone alone.23 A study of 296 patients with Bell’s palsy treated with valacyclovir or placebo in addition to a fixed dose of prednisolone found significant benefit to the addition of valacyclovir, particularly in the setting of severe palsy or in those treated within 24 hours of symptom onset.24 Other studies have found conflicting results. Despite
a lack of overwhelming evidence, the addition of an antiviral agent should be considered in the treatment of Bell’s palsy, especially with severe loss of function. The most commonly recommended antiviral regimens include valacyclovir, 1000 mg orally two times daily for 10 days. Valacyclovir and famciclovir have better oral absorption, are better tolerated, and are dosed less frequently, resulting in higher compliance. Accordingly, they have been recommended as alternatives to acyclovir.17,19,20,22,25 As with steroid therapy, although earlier treatment is preferred, treatment should be considered for patients presenting within 1 week of symptom onset.
Ramsay Hunt Syndrome Ramsay Hunt syndrome (herpes zoster oticus) is characterized by unilateral facial paralysis, a herpetiform vesicular eruption, and vestibulocochlear dysfunction. The vesicular eruption may occur on the pinna, external auditory canal, tympanic membrane, soft palate, oral cavity, face, and neck as far down as the shoulder. The pain is considerably more severe than that associated with Bell’s palsy, and it frequently is out of proportion to physical findings. In addition, outcomes are worse than with Bell’s palsy, with a lower incidence of complete facial recovery and the possibility of sensorineural hearing loss. Therapy is similar to that for Bell’s palsy. Both prednisone and antiviral therapy for 7 to 10 days are advocated.17,26,27
Lyme Disease Lyme disease is the most frequent vector-borne infection in the United States. It is caused by the spirochete Borrelia burgdorferi and is spread by the bite of Ixodes genus ticks. Neurologic manifestations can arise in any phase of the disease, and the incidence of facial palsy in patients with neurologic involvement is 35 to 51%. In regions in which Lyme disease is endemic, it has been shown to be the leading cause of facial paralysis in children, responsible for one half of all pediatric cases of facial nerve paralysis.28,29 Bilateral facial nerve paralysis is rare but can occur with systemic infections. The two diseases most commonly associated with bilateral simultaneous onset of facial paralysis are Lyme disease and infectious mononucleosis. Bilateral facial paralysis should be considered to be a manifestation of Lyme disease until further testing excludes this diagnosis.20,28-30 The evaluation and treatment of Lyme disease are discussed in Chapter 132.
Bacterial Infections Facial paralysis can be caused by acute bacterial infections of the middle ear, mastoid, or external auditory canal. In the preantibiotic era, facial paralysis was associated with acute otitis media in approximately 2% of cases; today, however, it occurs in only 0.2% of cases. Treatment consists of intravenous antibiotics and myringotomy for decompression. Malignant otitis externa can be associated with facial paralysis. This disease entity is most commonly seen in immunocompromised patients and usually is caused by Pseudomonas infection. Treatment involves prolonged intravenous anti pseudomonal antibiotic therapy and may require surgical débridement.20,31
Trauma In patients with head trauma, the facial nerve is the most commonly injured cranial nerve. The cause generally is a temporal bone fracture with nerve transection. Surgical exploration is warranted if there is firm evidence that the nerve has been transected, indicated by a sudden onset of complete unilateral facial paralysis, loss of electrical activity, and evidence of a displaced fracture involving the facial canal.
1383
Neoplasm
Clinical Features and Differential Considerations The medical history should focus on onset of the paralysis, concentrating on timing and rapidity of onset and looking for any associated signs and symptoms. A rapid onset of facial paralysis with dysgeusia and hyperacusis preceded by a viral prodrome is suggestive of Bell’s palsy. A history of recurrent ipsilateral paralysis or slow progression of symptoms is more characteristic for a tumor. Associated cranial nerve abnormalities, although occasionally seen with Bell’s palsy, also point to the possibility of a tumor or ischemic insult. The Ramsay Hunt syndrome causes significant pain and a vesicular rash, although the rash may follow the facial paresis by a few days. Significant anatomic abnormalities on visual or otoscopic inspection of the ipsilateral ear will be found with bacterial otitis media and otitis externa. Finally, systemic symptoms or bilateral facial paresis, especially in endemic areas, should raise the possibility of Lyme disease.
Diagnostic Strategies The diagnostic workup of acute facial nerve paresis is based on whether the clinical picture is suggestive of a disease process other than Bell’s palsy. If the clinical history is classic for Bell’s palsy, then no imaging or laboratory studies are required. Of note, any history of possible exposure warrants serologic evaluation for Lyme disease. Although outpatient testing including electroneurography may ultimately be performed, this usually is not part of the initial evaluation. The physical examination finding of a “central” seventh nerve paralysis (upper face–sparing) should prompt imaging with computed tomography (CT) or MRI, and consideration should be given to the possibility of an acute stroke or other hemispheric lesion. History or physical examination findings suggestive of a possible tumor require imaging to rule out a neoplasm. The study of choice will depend on the institution and preferences of the consultant.
Disposition The vast majority of patients who have a seventh nerve paralysis will have a clinical diagnosis of Bell’s palsy and may be discharged with referral for short-term follow-up. Patients with a possible hemispheric process such as stroke or tumor should be hospitalized for further evaluation. Patients suspected of having Lyme disease require immediate initiation of appropriate antibiotic therapy. In patients with a peripheral facial nerve paralysis, the ipsilateral eye should be patched, and consideration should be given to ophthalmologic follow-up, because there is a high rate of corneal abrasions and corneal dryness associated with the inability to properly blink or completely close the eye.
■
Recent literature highlights significant potential benefit for patients with clinical evidence of Bell’s palsy when they are treated early in the course with corticosteroids. The additional benefit of adding antiviral medication is controversial, but this treatment probably is warranted in patients with severe loss of function. ■ Slowly progressive facial paralysis is suggestive of a neoplasm. Recurrent unilateral paralysis may occur with Bell’s palsy but frequently (30%) is seen in patients with tumor. ■ Simultaneous bilateral facial paralysis is suggestive of Lyme disease, which must be considered as a possible cause, especially in endemic regions. ■ Patients who have facial muscle paresis with intact forehead movement should be considered to have an upper motor neuron lesion until the diagnostic investigation proves otherwise.
■ VESTIBULAR SCHWANNOMA Perspective Vestibular schwannoma, formally referred to as acoustic neuroma, is a rare but important cause of sensorineural hearing loss. The annual incidence of VS is 1 case per 100,000 population, with a mean age at the time of detection of 46 to 58 years.33 The female-to-male ratio is 1.5 : 1. Vestibular schwannoma is very rarely bilateral, occurring in this form in approximately 5% of cases and generally associated with type II neurofibromatosis. Although histologically benign, vestibular schwannoma can cause neurologic damage by direct compression on the eighth cranial nerve and the other structures in the cerebellopontine angle.34
Principles of Disease Vestibular schwannoma arises from the Schwann cells covering the vestibular branch of the eighth cranial nerve as it passes through the internal auditory canal. The tumor may compress the cochlear (acoustic) branch of the eighth cranial nerve, causing hearing loss, tinnitus, and dysequilibrium. Continued growth of the tumor may result in compression of structures in the cerebellopontine angle, where the facial and trigeminal nerves may be compressed and damaged. Larger tumors may further encroach upon the brainstem and if large enough may compress the fourth ventricle, ultimately resulting in signs of increased intracranial pressure (ICP).35
Clinical Features Asymmetrical sensorineural hearing loss is the hallmark of vestibular schwannoma. Up to 15% of patients with this tumor, however, will have normal results on an audiogram. These patients typically have symptoms such as unilateral tinnitus, imbalance, headache, fullness in the ear, otalgia, or facial nerve weakness. Thus, patients with asymmetrical symptoms should be further evaluated for vestibular schwannoma even with normal findings on the audiogram.36 Vestibular schwannomas are extremely slow-growing tumors, averaging an approximately 1-mm increase per year, although many do not grow at all on serial examinations.37 Symptom onset is therefore generally quite gradual. In one series of 126 cases, the average time from symptom onset to discovery of a vestibular schwannoma was approximately 4 years.38
Chapter 103 / Brain and Cranial Nerve Disorders
Tumors of the facial nerve itself, or tumors anywhere along the course of the facial nerve that invade or compress the nerve, may lead to facial paralysis. Typically the course is progressive over at least 3 weeks. A sudden onset of paralysis, however, does not rule out an underlying tumor, because facial paralysis secondary to a neoplasm is of sudden onset in approximately 25% of cases.32 A neoplastic cause should be suspected in patients who suffer from recurrent ipsilateral facial paralysis, significant pain, prolonged symptoms, or any other concomitant cranial nerve abnormality.
KEY CONCEPTS
1384
PART III ■ Medicine and Surgery / Section Seven • Neurology
Diagnostic Strategies When vestibular schwannoma is suspected, the patient should be evaluated with an audiogram or a gadolinium-enhanced MRI. This imaging technique is extremely sensitive and has led to earlier diagnosis and a decrease in mean size at detection of vestibular schwannoma. CT lacks the necessary sensitivity in the posterior cranial fossa to reliably rule out the presence of vestibular schwannoma. The smaller the tumor at the time of diagnosis, the more options there are for therapy and the better the potential prognosis.34
Differential Considerations A majority of disease entities included in the differential diagnosis for acoustic neuroma cause symmetrical sensorineural hearing loss. Asymmetrical sensorineural hearing loss has few causes other than vestibular schwannoma. Ménière’s disease may present a diagnostic dilemma because it can be asymmetrical. Ménière’s disease may be differentiated from vestibular schwannoma in that the tinnitus of Ménière’s disease usually is intermittent, whereas the tinnitus of vestibular schwannoma typically is continuous. In addition, patients with Ménière’s disease typically describe true vertigo, whereas patients with a vestibular schwannoma are more likely to describe imbalance or dysequilibrium. Vestibular schwannomas account for 80% of all cerebellopontine angle tumors. Among all other lesions, meningioma is the most common. Meningiomas more frequently cause symptoms of facial palsy or trigeminal nerve abnormality. Of note, however, considerable similarity between the clinical picture of a meningioma and that of vestibular schwannoma in the cerebellopontine angle has been described.39
KEY CONCEPTS ■
The onset of unilateral auditory symptoms, especially sensorineural hearing loss, requires evaluation and referral to an ear, nose, and throat specialist. ■ Neurologic symptoms of lower cranial nerve dysfunction, ataxia, or raised ICP may be caused by a benign tumor of the cerebellopontine angle. ■ The smaller the tumor at diagnosis, the better the long-term outcome with definitive treatment. followed in order by the trochlear and abducens nerves. In one large series in Japan, the incidence of cranial nerve palsies was 1.0% among diabetics and 0.1% among nondiabetics.40,41 Whereas ophthalmoplegia appears to be closely related to diabetes, facial palsy is less strongly correlated with this disease.40
Principles of Disease The pathologic basis of diabetic mononeuropathy appears to be ischemia of the affected cranial nerve caused by occlusion of an intraneural nutrient artery serving the nerve. This occlusion leads to injury located primarily in the center of the nerve, because the core fibers are more dependent on the supply from such nutrient arteries. The peripheral fibers are less affected because they also are supplied by collateral vessels. In the oculomotor nerve, the preservation of the circumferentially located parasympathetic fibers explains the pupillary sparing that usually is found in this syndrome. In two studies, the microvascular changes in the intraneural arteries that lead to occlusion were noted in diabetic patients but absent in nondiabetics.42,43
Management
Clinical Features
Vestibular schwannoma may be removed surgically or ablated with stereotactic radiation. In general, tumors larger than 3 cm are recommended for microsurgery, because radiation treatments, such as with the Gamma Knife or linear accelerator, are less effective for local control and growth arrest in larger masses. Smaller tumors are amenable to use of stereotactic radiation, which may have greater salvage rates of facial nerve function and hearing. Stereotactic radiation therapy generally has good long-term outcomes of local growth arrest, with nerve salvage approaching 90% or greater. Injuries to the trigeminal, facial, and acoustic nerves, and to the cerebellum, are possible complications of both procedures. In patients who are minimally symptomatic with small tumors, serial monitoring with MRI is a viable nonsurgical option. All patients should be evaluated by a specialist in the evaluation and treatment of vestibular schwannoma, because smaller tumor size at detection is associated with a better long-term outcome.33,37
Patients typically describe acute onset of unilateral retro-ocular and supraorbital pain, diplopia, and ptosis.41 The physical manifestations of a third cranial nerve palsy include the inability to move the eye superiorly and medially, accompanied by ptosis. The pupillary light reflex usually is present. Although a less common finding, the fourth and sixth cranial nerves may be affected. Patients with a fourth cranial nerve palsy are unable to move the eye inferolaterally, and those with a sixth cranial nerve palsy are unable to move the eye laterally. Because of the long intracranial course of the sixth nerve, a patient with an isolated sixth nerve palsy should be evaluated for an intracranial lesion or increased ICP.44
Disposition Patients with suspected acoustic neuroma should be referred for an audiogram or MRI and evaluation by a specialist in either otolaryngology or neurosurgery.
■ DIABETIC CRANIAL MONONEUROPATHY Perspective Cranial mononeuropathies occur uncommonly, usually are a complication of diabetes, and most often affect the extraocular muscles. The oculomotor nerve is most commonly affected,
Differential Considerations Evaluating cranial nerve dysfunction requires a thorough history and physical examination and cranial imaging, usually with MRI. Diabetic mononeuropathy should be considered a diagnosis of exclusion, with considerations in the differential diagnosis including trauma, tumor, vertebrobasilar ischemia, aneurysm, and hemorrhage into the brainstem.45
Management Treatment consists of patching the affected eye and administration of analgesics and antiplatelet therapy. The prognosis is good. If the neuropathy does not begin to resolve within 3 to 6 months, or if more than one nerve is affected, another cause should be sought. Complete resolution is expected within the first year. Antioxidant preparations, including α-lipoic acid, have been used therapeutically and have not shown harm, but
such agents have yet to be shown to have convincing clinical effect.46
■
Diabetic neuropathy is a diagnosis of exclusion because no definitive diagnostic testing is available. ■ Both ischemic and hemorrhagic brainstem lesions must be ruled out in the case of an acute ophthalmoplegia. ■ Extraocular mononeuropathy is sufficiently common in patients with diabetes mellitus that its occurrence in isolation warrants evaluation of the patient for previously undiagnosed diabetes.
■ CEREBRAL VENOUS THROMBOSIS Perspective No precise studies of the epidemiology of cerebral venous thrombosis (CVT) have been performed. In case series, the median patient age is approximately 37 years, with a femaleto-male ratio of 3 : 1.47
Principles of Disease Cerebral blood is drained by several major veins that lead into the dural sinuses. The major dural sinuses are the superior sagittal sinus, the inferior sagittal sinus, the straight sinus, the lateral sinuses, and the sigmoid sinuses. The variability in symptoms and signs in patients who present with CVT stems from differences in thrombus location and acuity of thrombus formation. Symptoms of intracranial hypertension are present in most patients with sinus thrombosis, whereas those with thrombosis of the cerebral veins are thought to be more prone to hemorrhagic infarction and localizing neurologic deficits.48 As with venous thrombosis in other locations, multiple causes and predisposing factors for CVT are recognized. Underlying causes often are divided into infectious and noninfectious categories. Infectious causes include local infections, such as sinusitis, otitis media, cellulitis on the face, and systemic infections. Noninfectious causes include direct injury to the cerebral venous system from trauma, surgery, tumor, dehydration, or any other condition that may predispose the patient to development of a hypercoagulable state.49
Clinical Features The symptoms and signs associated with CVT are quite varied. Headache is the primary feature of CVT in 74 to 92% of affected patients.49,50 Papilledema is noted in 28 to 45% of cases.47,50,51 Lethargy, decreased level of consciousness, or mental status changes may be noted. Seizures occur in 35 to 50% of patients in the acute phase.47,49,51 In addition to the location and acuity of thrombosis formation, a patient’s symptom onset will vary in accordance with the extent of collateral vessel growth in the venous territory. Early thrombotic changes may be well compensated for by the collateral venous drainage. Symptoms will appear only when the compensation for venous thrombosis is no longer sufficient. Variability in collateralization between patients also adds to the variability and time course of symptomatology. Two national and international observational studies document an average time from symptom onset to diagnosis of 7 days, reflecting the difficulty in diagnosing this rare disease entity.47,51,52 The reported incidence of focal neurologic signs, including seizures, on clinical
Diagnostic Strategies The gold standard modality for the diagnosis of CVT has shifted in recent years from cerebral angiography to magnetic resonance venography (MRV). CT scanning is useful in the initial workup of the patient with possible CVT, but noncontrast CT is neither sensitive nor specific enough to reliably confirm or exclude the diagnosis. Findings on CT that are consistent with CVT include hyperdensity of a thrombosed sinus, brain edema, and hemorrhage secondary to venous congestion. CT venography is both more sensitive and more specific in diagnosing CVT. Similar to CT scanning, MRI also can demonstrate local changes secondary to venous congestion, such as brain edema or hemorrhage. In addition, MRI can demonstrate the possibility of CVT based on the lack of a “flow void.” On conventional MRI, a flow void indicates the presence of blood flow within the sinus, whereas the absence of a flow void indicates a possible thrombus. Diagnostic accuracy, however, is greatly improved through use of MRV. This technique takes advantage of the MRI signal characteristics of flowing blood to create images of venous structures. Combining these imaging techniques further enhances diagnostic accuracy. For imaging a particular dural sinus, presence of the sinus on conventional MRI and lack of flow on MRV are diagnostic of a sinus thrombosis. This combined approach has diagnostic sensitivity similar to that of conventional angiography.49,53 Two small studies show similar sensitivity between MRV and CT venography for the diagnosis of CVT when the CT study is performed on a multidetector row CT scanner. Both studies, involving a total of 69 patients, showed 100% sensitivity of CT venography for CVT in comparison with MRV.54,55 The sensitivity of CT venography performed by scanners that do not use multidetector row technology is unknown. Several small studies have attempted to evaluate the usefulness of the D-dimer assay as a screening tool to exclude CVT, particularly when MRI or CT venography is not available. Although the reported sensitivity rates are fair at 83 to 100%, larger prospective studies need to be done to further define the role of D-dimer in the evaluation of CVT, because several case reports have noted normal D-dimer levels in the setting of documented CVT.56-59 In general, although a normal Ddimer level does not exclude the diagnosis of CVT, it does appear to make this diagnosis much less likely, particularly in a patient with symptoms of less than 2 weeks in duration.
Differential Considerations Considerations in the differential diagnosis of CVT include the conditions that cause patients to present with the new onset of neurologic deficits, alteration in consciousness, or severe headache. A diagnosis of CVT should be considered in a patient with such symptoms when the etiology is unclear, presence of having a hypercoagulable state is likely, and the head CT scan is normal in appearance or shows subtle signs of CVT.
Management CVT is a relatively rare disease, and controlled studies evaluating its treatment are lacking. Current therapeutic consensus
Chapter 103 / Brain and Cranial Nerve Disorders
KEY CONCEPTS
1385
examination varies between series, ranging from 25 to 71%.49,50 Because of the broad spectrum of possible clinical features, the diagnosis of CVT may be difficult but should be a consideration in any patient with unexplained headache, especially in combination with focal neurologic deficit, papilledema, or seizures.
PART III ■ Medicine and Surgery / Section Seven • Neurology
1386
strongly recommends systemic anticoagulation with lowmolecular-weight or unfractionated heparin to prevent further clot formation and to promote recanalization, even in patients with intracranial hemorrhage on initial imaging.49,50,60,61 In one placebo-controlled randomized trial comprising 20 patients, anticoagulation with heparin to a target partial thromboplastin time (PTT) of 80 to 100 seconds demonstrated benefit, even in patients in which evidence of intracranial hemorrhage was seen on the CT scan before anticoagulation.62 In another study of 60 patients randomized to receive placebo or low-molecularweight heparin, no statistical benefit was shown for treatment.63 Two large observational trials showed improvement in modified Rankin scale at follow-up in the anticoagulated groups, although the trials were not randomized.47,51 Despite a paucity of randomized controlled trials, expert opinion favors anticoagulation in all groups unless another contraindication is present.64 Catheter-based intervention with thrombolysis has been attempted in multiple case series using either urokinase or tissue plasminogen activator. Thrombolysis was shown to be relatively safe and relatively successful in very small case series.61 In one nonrandomized study of 40 patients, 20 received systemic heparin and 20 received catheter-based infusion of urokinase followed by systemic heparin. Despite initially worse neurologic function in the thrombolysis group, a significant difference in neurologic function favoring thrombolysis was observed at discharge.65 Although this therapy is promising, it should be considered only for patients with symptoms of decreased level of consciousness, elevated ICP, or rapid deterioration on neurologic examination.
Disposition All patients with suspected CVT should be admitted to a unit capable of providing a high level of care with neurologic consultation. Patients should be anticoagulated if no contraindication exists, and catheter-based thrombolysis should be considered in patients with depressed mental status or focal findings on neurologic exam.
KEY CONCEPTS ■
CVT is a relatively rare entity, and only awareness of and familiarity with the clinical presentation will lead to the correct diagnosis. ■ The onset may be insidious with a considerable delay between onset and arrival in the treatment setting. ■ The differential diagnosis for CVT should consider other conditions that cause patients to present with new-onset neurologic deficits, alteration in consciousness, or severe headache. CVT is more likely to be present in such patients when the etiology is unclear, the patient is suspected of having a hypercoagulable state, and the head CT is normal in appearance or shows subtle signs of CVT. ■ Noncontrast CT scanning is not adequate to rule out CVT. An MRI with MRV is recommended, although multidetector row CT venography is an acceptable alternative. ■ Treatment of most patients with CVT should include systemic anticoagulation, even in the setting of hemorrhagic cerebral infarcts, unless another contraindication exists.
■ MULTIPLE SCLEROSIS Perspective Multiple sclerosis (MS) is an inflammatory disease that affects the central nervous system (CNS). Although the exact etiology remains uncertain, the pathologic manifestation of this inflammatory disease is a demyelination of discrete regions (plaques) within the CNS with a relative sparing of axons. The clinical picture is highly variable but is classically characterized by episodes of neurologic dysfunction that evolve over days and resolve over weeks. MS has an overall prevalence in the United States of 0.1%. The peak age at onset is 25 to 30 years, with women being slightly younger at onset than men. The incidence in women exceeds that of men by a ratio of 1.8 : 1. The worldwide prevalence is greatest in the United Kingdom, Scandinavia, and North America. Epidemiologic studies indicate that both genetic and environmental factors are associated with an increased incidence of this disease. MS has a 30% concordance rate between monozygotic twins, and 20% of patients with MS have at least one affected relative. MS is more common in temperate climates. It is rare between 23 degrees north and south latitudes but has a rising incidence above and below 50 degrees north and south latitudes, respectively. Although no exact environmental factor has been identified, if a person emigrates from an area of high prevalence to an area of low prevalence before the age of 20, the risk is diminished. MS is rare in Africans and Asians, but African Americans have a higher incidence than their relatives who remain in Africa.66 In addition, reports of clusters or miniepidemics support environmental factors. Thus, an environmental cause superimposed on genetic susceptibility appears to be a likely etiologic scenario.67,68
Principles of Disease MS is considered to be an organ-specific autoimmune disease. One theory proposes that genetic factors interact with an environmental trigger or infection to establish pathologically autoreactive T cells in the CNS. After a long and variable latency period (typically 10 to 20 years), a systemic trigger, such as a viral infection or superantigen, activates these T cells. The activated T cells, on reexposure to the autoantigen, initiate the inflammatory response. This sets off a complex immunologic cascade that leads to the demyelination characteristic of MS. This demyelination process releases CNS antigens that are hypothesized to initiate further episodes of autoimmuneinduced inflammation. The mechanisms underlying this autoimmunity in MS are unknown.69
Clinical Features The clinical picture in MS is one of marked heterogeneity. The classic clinical syndrome consists of recurring episodes of neurologic symptoms that rapidly evolve over days and slowly resolve. Variability occurs in age at onset, location of CNS lesions, frequency and severity of relapses, and the degree and time course of progression. The clinical features of MS can be divided into areas of specific CNS impairment: cognition, cranial nerves, motor pathways, sensory pathways, cerebellar pathways, and bowel, bladder, and sexual dysfunction.66 Patients with MS have frequent complaints of poor memory, distractibility, and a decreased capacity for sustained mental effort. Formal neuropsychological testing suggests that cognitive involvement is common and underreported. Specifically,
Diagnostic Strategies Although no laboratory tests are diagnostic for MS, one clinical feature remains relatively unique to this disease: Uhthoff’s phenomenon, temporary worsening of current or preexisting signs or symptoms of MS secondary to small increases in the patient’s body temperature. Accordingly, exercise, a hot bath,
1387
exposure to a warm environment, or fever can bring about Uhthoff’s phenomenon. This phenomenon reflects subclinical demyelination or preexisting injury to nerves without obvious significant clinical involvement before heat exposure or temperature elevation.66 The clinical diagnosis rests on occurrence of at least two clinical episodes with different neurologic symptoms at different times. Thus, MS commonly has been characterized as a disorder with lesions that differ in time and space. It also has been described as a relapsing-remitting disorder with symptoms that fluctuate over time. Findings on cerebrospinal fluid (CSF) analysis are abnormal in 90% of the cases. Fifty percent of patients will have pleocytosis, with more than 5 lymphocytes per high-power field in the CSF. Approximately 70% of patients will have an elevated gamma globulin level, with immunoglobulin G (IgG) ranging from 10 to 30% of the CSF total protein. Electrophoresis of the CSF demonstrates oligoclonal bands of IgG in 85 to 95% of patients who carry a diagnosis of MS; however, oligoclonal bands of IgG also are seen with neurosyphilis, fungal meningitis, and other CNS infections. Lumbar puncture should be considered for all patients with suspected MS, but mass lesions and elevated ICP should be ruled out before lumbar puncture.77 The initial imaging test to aid in the diagnosis of multiple sclerosis is MRI. MRI is a sensitive test for the detection of lesions consistent with MS and also is useful to assess disease severity.78 The lesions of MS typically appear hyperintense, or bright white, on T2-weighted or fluid-attenuated inversion recovery (FLAIR) MRI studies. Lesions usually are multiple and commonly are found in the periventricular white matter.79 In patients with an initial neurologic event consistent with CNS demyelination and an MRI cranial study showing multiple white matter lesions, the 5-year risk of developing MS is 60%. Patients with similar clinical syndromes and a normal MRI appearance have less than a 5% 5-year risk.80
Differential Considerations Other diseases that affect the CNS white matter may be clinically and radiographically similar to MS. Considerable care must be taken to exclude these disease processes before making a diagnosis. These include CNS tumors (especially lymphomas and gliomas), spinal cord compression, vasculitides, Behçet’s disease, neurosarcoidosis, postinfectious and postvaccinal encephalomyelitis, human immunodeficiency virus (HIV) encephalopathy, Lyme disease, and vitamin B12 deficiency.
Management Management of patients with MS has essentially three aspects: (1) therapies aimed at halting the progression of the disease, (2) treatment for acute exacerbations, and (3) therapies designed to modify complications. Therapies aimed at halting disease progress are based primarily on the use of either β-interferon or glatiramer acetate. The interferons are a group of natural compounds with antiviral and immunomodulatory actions, which are retained by the recombinant preparations used in therapy for MS, interferon beta-1a and interferon beta-1b. Side effects include flulike symptoms, depression, anxiety, and confusion. In one study, 560 patients with MS were randomly assigned to receive subcutaneous interferon beta-1a or placebo (n = 187) three times a week for 2 years. The relapse rate was significantly lower at 1 and 2 years with interferon beta-1a than with placebo. The time to first relapse was prolonged significantly and the accu-
Chapter 103 / Brain and Cranial Nerve Disorders
neuropsychological testing has shown that 43 to 65% of patients with MS have some degree of cognitive impairment.70,71 Of note, a correlation has been found between the MRI-based total lesion load and presence of cognitive impairment.72 Cranial nerve dysfunction is common in MS. The most common associated cranial nerve abnormality is optic neuritis, a unilateral syndrome characterized by pain in the eye and a variable degree of visual loss affecting primarily central vision. Within 2 years of an attack of optic neuritis, the risk of MS is approximately 20%, and within 15 years, it is approximately 45 to 80%.73,74 Optic neuritis often is the first symptom of MS.75,76 As a result of lesions in the vestibulo-ocular connections, the oculomotor pathways also may be affected. The deficit may manifest as diplopia or nystagmus. The nystagmus may be severe enough that the patient may complain of oscillopsia (a subjective oscillation of objects in the visual field). Cranial nerve impairment also may include impairment of facial sensation, which is relatively common. Unilateral facial paresis also may occur. In addition, the occurrence of trigeminal neuralgia in a young person may be an early sign of MS. Motor pathways also are commonly involved. Specifically, corticospinal tract dysfunction is common in patients with MS. Paraparesis or paraplegia is all too common and occurs with greater frequency than upper extremity lesions, owing to the common occurrence of lesions in the motor tracts of the spinal cord. In patients with significant motor weakness, spasms of the legs and trunk may occur on attempts to stand from a seated position. This dysfunction is manifested on physical examination as spasticity that typically is worse in the legs than in the arms. The deep tendon reflexes are markedly exaggerated, and sustained clonus may be demonstrated. Although these symptoms frequently are bilateral, they generally are asymmetrical.66 Sensory manifestations are a frequent initial feature of MS and will be present in nearly all patients at some point during the course of the disease. Sensory symptoms are commonly described as numbness, tingling, “pins and needles” paresthesias, coldness, or a sensation of swelling of the limbs or trunk.66 Impairment of the cerebellar pathway results in significant gait imbalance, difficulty with coordinated actions, and dysarthria. Physical examination reveals the typical features of cerebellar dysfunction, including dysmetria, dysdiadochokinesis (an impairment of rapid alternating movements), a breakdown in the ability to perform complex movements, an intention tremor in the limbs and head, truncal ataxia, and dysarthria.66 Impairment of bowel, bladder, and sexual function also is common. The extent of sphincter and sexual dysfunction usually parallels the motor impairment in the lower extremities. Urinary frequency may progress to urinary incontinence with progression of the disease. An atonic bladder may develop, which empties by simple overflow and often is associated with the loss of perception of bladder fullness and with anal and genital hypoesthesia. Constipation becomes common over time, and almost all patients with paraplegia require special measures to maintain effective bowel habits. Sexual dysfunction, although frequently overlooked, is very common in MS. Approximately 50% of patients become completely sexually inactive as a result of this disease.66
PART III ■ Medicine and Surgery / Section Seven • Neurology
1388
mulation of brain lesions on MRI was lower in the treatment group than in the placebo group. The investigators concluded that subcutaneous interferon beta-1a is a well-tolerated and effective treatment for relapsing-remitting MS in terms of relapse rate, defined disability, and all MRI outcome measures.81 β-Interferon also has been shown to retard progression to clinically definite MS and to decrease the total number of brain lesions seen on subsequent MRI studies in patients who have their first demyelinating episode with MRI abnormalities at initial presentation.82-85 This finding highlights the importance of early evaluation and treatment. Glatiramer acetate also has successfully been used in the treatment of MS. This agent is a mixture of synthetic polypeptides designed to mimic myelin basic protein. The mechanism of action by which glatiramer acetate exerts its effect is unknown, but it is thought to modify the immune processes responsible for the pathogenesis of MS. In one study, 251 patients with relapsing-remitting MS were randomized to receive daily subcutaneous injections of glatiramer acetate (previously called copolymer 1) or placebo for 24 months. Patients receiving glatiramer acetate experienced significantly fewer relapses and were more likely to demonstrate neurologic improvement, whereas those receiving placebo were more likely to worsen. This drug generally is quite well tolerated.86 Current recommendations for management of relapsingremitting MS are to initiate treatment with β-interferon or glatiramer acetate. Such regimens have been demonstrated to decrease the volume of plaques seen on MRI and to diminish relapses.69 Immunosuppressive agents, including mitoxantrone and azathioprine, also have been shown to be effective in reducing progression of disease but, in view of concerns over side effects, generally are used as second-line agents.87,88 Acute exacerbations of MS also should be targets for therapy. Although most such episodes will resolve without therapy, steroids have been demonstrated to diminish the duration of acute exacerbations. More than 85% of patients with relapsing-remitting MS show improvement with intravenous methylprednisolone. Intravenous steroids have been shown in controlled trials to speed the recovery from the visual loss of optic neuritis when compared with placebo. In addition, when patients with acute optic neuritis are treated with high-dose intravenous steroids, the 2-year rate of development of MS is reduced, although this effect diminishes over time.74,89 Of interest, oral prednisone was not found to be helpful in the optic neuritis trials and was associated with a potential increase in the number of optic neuritis episodes. The current standard therapy for an acute exacerbation in MS is intravenous methylprednisolone. A typical dose administered intravenously is 250 to 500 mg every 12 hours for 3 to 7 days. Whether this should be followed by an oral prednisolone taper remains controversial. Potential adverse effects of methylprednisolone therapy include fluid retention, gastrointestinal hemorrhage, anxiety, psychosis, infection, and osteoporosis. Several therapies directed toward the complications of MS may be helpful. The associated spasticity generally is treated with baclofen. This is a highly effective therapy aimed at
reducing the painful flexor and extensor spasms. A major side effect is drowsiness, which generally diminishes with continued use. Higher-dose therapy can cause confusion, especially in the setting of baseline cognitive impairment. For patients with intractable spasticity, baclofen is available for intrathecal administration by either bolus therapy or continuous implanted pump therapy. Additional therapeutic agents for control of spasticity include tizanidine, diazepam, and dantrolene. The tremor and ataxia associated with MS occasionally are treated with propranolol, diazepam, or clonazepam. The results of these therapies, however, generally are unsatisfactory. Pain often is associated with MS and affects the shoulders, pelvic girdle, and face. The facial pain may be indistinguishable from that of trigeminal neuralgia. Treatment options include carbamazepine, baclofen, and tricyclic antidepressants. Fatigue, which is common, may be ameliorated with amantadine. This agent produces partial relief for a minority of patients. In controlled studies, the effect is only slightly better than placebo.69
Disposition Patients with a history of MS who seek treatment for significant symptoms must first be evaluated to rule out other, non– MS-related pathology. Also, the presence of other systemic illnesses, especially infections, which can worsen the symptoms of MS, should be excluded. If the problem is thought to be an exacerbation of MS, most patients will require hospital admission for intravenous steroid therapy. An alternative to hospitalization may be to initiate intravenous steroids in the ED and to arrange for a next-day follow-up visit with the primary care physician or neurologist if outpatient intravenous steroid administration is an option. Patients with the new onset of symptoms suggestive of MS should be admitted or referred to a neurologist, depending on the type and severity of symptoms.
KEY CONCEPTS ■
Any patient with a long-term illness, such as MS, must be evaluated to rule out pathologic processes not related to that illness before an exacerbation of the illness can be assumed to be the cause of any problems experienced by the patient. ■ Therapy for patients with MS will require consultation with the patient’s primary care provider or neurologist to provide consistent disease management. ■ Intravenous methylprednisolone effectively promotes earlier resolution of recurrences. ■ Intravenous methylprednisolone has been shown to speed the recovery from vision loss from optic neuritis.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 103
Brain and Cranial Nerve Disorders
Brian A. Stettler
This chapter discusses cranial nerve problems, cerebral venous thrombosis, and multiple sclerosis—neurologic disorders that often provide significant diagnostic and therapeutic challenges in management in the emergency department (ED) setting (Table 103-1).
■ TRIGEMINAL NEURALGIA Perspective Trigeminal neuralgia, or tic douloureux, is a syndrome featuring painful paroxysms in one or more distributions of the trigeminal nerve. Trigeminal neuralgia is relatively uncommon, with an annual incidence of 4 to 13 cases per 100,000 population.1,2 It is more common in women than in men, with a female-tomale ratio of 1.7 : 1. Affected persons typically are between 50 and 69 years of age, and symptoms occur more frequently on the right side of the face.3
Pathophysiology Trigeminal neuralgia is an idiopathic disorder, although significant evidence points to vascular compression of the trigeminal nerve root in many cases. This compression commonly is caused by a tortuous arterial or venous loop in the posterior fossa, an arteriovenous malformation, or rarely a tumor. In surgical case series, vascular compression of the trigeminal nerve root is found in 80 to 90% of cases.4,5 Of note, however, structural lesions are not found in all patients with trigeminal neuralgia.6
Clinical Features Trigeminal neuralgia manifests with unilateral facial pain, typically characterized as lancinating paroxysms of pain in the lips, teeth, gums, or chin. The pain of trigeminal neuralgia commonly is associated with physical triggers such as chewing, brushing the teeth, shaving, washing or touching the affected area of the face, swallowing, or exposure to hot or cold temperature in the affected area. The maxillary and mandibular divisions of the trigeminal nerve are most commonly involved; rarely, the ophthalmic division alone is involved. Patients tend to experience the pain in clustered episodes that last a few seconds to several minutes. The attacks can occur during the day or night but rarely arise during sleep.6,7
Diagnostic Strategies A careful history and physical examination should be performed to rule out other painful facial conditions including odontogenic infections, sinus disease, otitis media, acute glaucoma, temporomandibular joint disease, and herpes zoster. Patients who lack local pathologic findings to explain the painful syndrome require a very careful neurologic examination. The presence of a neurologic deficit should prompt suspicion of a structural lesion, such as aneurysm, tumor, or other intracranial lesion such as from multiple sclerosis (MS). Of note, 2 to 4% of patients with trigeminal neuralgia also have MS.8 Patients with normal findings on the head and neck examination and no neurologic deficits who have episodic, unilateral facial pain associated with nonpainful triggers are likely to have trigeminal neuralgia.
Management Since the 1960s the medical treatment of choice for trigeminal neuralgia has been use of the anticonvulsant carbamazepine. The purported effectiveness of this treatment is, however, based on uncontrolled studies, and the mechanism of action of anticonvulsant therapy for trigeminal neuralgia is unclear. The true efficacy of medical therapy is difficult to assess owing to a very high rate of spontaneous remission. Nonetheless, carbamazepine appears to be an effective and well-tolerated agent for treatment of trigeminal neuralgia. The initial dosage of carbamazepine is 100 mg twice daily; this dose is then increased to three times daily after one week. The dose may be increased by 100 mg per day, up to a maximum of 1200 mg per day. A complete blood count and liver function studies should be performed periodically in these patients to monitor for hematologic and hepatic side effects. Additional agents that have been used for treatment of trigeminal neuralgia include phenytoin, baclofen, valproate sodium, lamotrigine, and gabapentin. None of these drugs have been shown to be more effective than carbamazepine.7 Surgical management has been a theraputic option since the 1950s. Surgical procedures include both peripheral approaches and central procedures. Peripheral strategies include medication injection and cryotherapy techniques designed to temporarily block, or permanently ablate, branches of the peripheral trigeminal nerve. Although these procedures are relatively effective initially, recurrence is common. Repeated nerve 1379
1380
PART III ■ Medicine and Surgery / Section Seven • Neurology
Table 103-1 The Cranial Nerves: Normal Function and Pathologic Considerations CLINICAL FUNCTION RELEVANT TO EMERGENCY MEDICINE
PATHOLOGIC FEATURES
POSSIBLE CAUSES
Cranial nerve I: Olfactory nerve
Sense of smell
Unilateral anosmia
Cranial nerve II: Optic nerve
Vision
Unilateral vision loss
Trauma: Skull fracture or shear injury interrupting olfactory fibers traversing the cribriform plate Tumor: Frontal lobe masses compressing the nerve Trauma: Traumatic optic neuropathy Tumor: Orbital compressive lesion Inflammatory: Optic neuritis (MS) Ischemic: Ischemic optic neuropathy Trauma: Herniation of the temporal lobe through the tentorial opening causing compression and stretch injury to the nerve Ischemic: Especially in diabetes Microvascular ischemic injury to nerve causes extraocular muscle paralysis but usually is papillary-sparing (often painful) Vascular: Intracranial aneurysms may press on the nerve, leading to dysfunction Myasthenia gravis can lead to atraumatic ocular muscle palsy Trauma is the most common cause of nerve dysfunction
CRANIAL NERVE
Cranial nerve III: Extraoculomotor function via Oculomotor nerve motor fibers to levator palpebrae, superior rectus, medial rectus, inferior rectus, inferior oblique muscles Pupillary constriction via parasympathetic fibers to constrictor pupillae and ciliary muscles
Ptosis caused by loss of levator palpebrae function Eye deviated laterally and down Diplopia Dilated, nonreactive pupil Loss of accommodation
Cranial nerve IV: Trochlear nerve
Motor supply to the superior oblique muscle
Cranial nerve V: Trigeminal nerve
Motor supply to muscles of mastication and to tensor tympani Sensory to face, scalp, oral cavity (including tongue and teeth) Motor supply to the lateral rectus muscle
Inability to move eye downward and laterally Diplopia Patients tilt head toward unaffected eye to overcome inward rotation of affected eye Partial facial anesthesia Trauma: Episodic, lancinating facial pain Facial bone fracture may injure one associated with benign triggers section, leading to area of facial such as chewing, brushing teeth, anesthesia light touch Tic douloureux
Cranial nerve VI: Abducens nerve
Cranial nerve VII: Facial nerve
Motor supply to muscles of facial expression Parasympathetic stimulation of the lacrimal, submandibular, and sublingual glands Sensation to the ear canal and tympanic membrane
Cranial nerve VIII: Hearing and balance Vestibulocochlear nerve
Inability to move affected eye Tumor: Lesions in the cerebellopontine laterally angle Diplopia on attempting lateral gaze Any lesion, vascular or otherwise, in the cavernous sinus may compress nerve Elevated intracranial pressure (ICP): Because of its position and long intracranial length, increased ICP from any cause may lead to injury and dysfunction of the nerve Hemifacial paresis: Lower motor neuron: Lower motor neuron lesion leaves Infection (viral): The likely cause of Bell’s entire side of face paralyzed palsy Upper motor neuron lesion leaves Lyme disease: The most common cause of bilateral cranial nerve VII palsy in areas forehead musculature where Lyme disease is endemic functioning Bacterial infection extending from otitis Abnormal taste media Sensory deficit around ear Intolerance to sudden loud noises Upper motor neuron: Stroke, tumor Unilateral hearing loss Tumors: Acoustic neuroma Tinnitus Mimics Ménière’s disease, perilymphatic Vertigo, unsteadiness fistula
1381
Table 103-1 The Cranial Nerves: Normal Function and Pathologic Considerations—cont’d
PATHOLOGIC FEATURES
POSSIBLE CAUSES
Cranial nerve IX: General sensation to posterior Glossopharyngeal third of tongue nerve Taste for posterior third of tongue Motor supply to the stylopharyngeus
Clinical pathology referable to the nerve in isolation is very rare Occasionally painful paroxysms beginning in the throat and radiating down the side of the neck in front of the ear but behind the mandible
Brainstem lesion Glossopharyngeal neuralgia
Cranial nerve X: Vagus nerve
Unilateral loss of palatal elevation: Patients complain that on drinking liquids, the fluid refluxes through the nose Unilateral vocal cord paralysis: Hoarse voice
Brainstem lesion Injury to the recurrent laryngeal nerve during surgery
Downward and lateral rotation of the scapula and shoulder drop
Trauma to the nerve
Tongue deviations: Upper motor neuron lesion causes the tongue to deviate toward the opposite side Lower motor neuron lesion causes the tongue to deviate toward the side of the lesion, and the affected side atrophies over time
Stroke or tumor can cause upper motor neuron lesion Amyotrophic lateral sclerosis (ALS) can cause bilateral lower motor neuron lesion with atrophy Metastatic disease to the skull base may involve the nerve
Cranial nerve XI: Spinal accessory nerve Cranial nerve XII: Hypoglossal nerve
Motor to striated muscles and muscles of the pharynx, larynx, and tensor (veli) palatini Motor to smooth muscles and glands of the pharynx, larynx, thoracic and abdominal viscera Sensory from larynx, trachea, esophagus, thoracic and abdominal viscera Motor supply to the sternocleidomastoid and trapezius muscles Motor supply to the intrinsic and extrinsic muscles of the tongue
blocks are not recommended owing to a high risk of permanent facial anesthesia. Central procedures can be divided into percutaneous approaches and open approaches. Percutaneous destruction of the trigeminal ganglion can be done by means of radiofrequency ablation, thermal ablation, glycerol injection, or balloon microcompression. These procedures carry the risk of corneal anesthesia, oculomotor paresis, or masticatory weakness.9 Open surgical management is the surgical option of choice in most treatment centers. Open surgical procedures include microvascular decompression of the nerve with or without partial ablation. Although the open microvascular decompression procedure has proved to be very effective, with pain relief achieved in 80 to 95% of patients, the surgery is associated with the risk of significant complications, including hearing loss, facial anesthesia, cerebrospinal fluid leak, brainstem or cerebellar injury, headaches, meningitis, and death.10,11 Gamma knife radiosurgery, a minimally invasive, precision-directed stereotactic radiosurgery, also has been associated with good outcomes. This highly specialized technique requires extremely sophisticated stereotactic radiofrequency equipment and is available only in specialized centers.12,13
Disposition Patients with suspected trigeminal neuralgia should be referred for specialty evaluation. Patients with a neurologic deficit require urgent imaging studies, typically magnetic resonance imaging (MRI), to rule out a mass or vascular abnormality.
KEY CONCEPTS ■
Patients with unilateral, intermittent, lancinating facial pain without abnormalities on physical examination are likely to have trigeminal neuralgia. ■ Carbamazepine is the first-line agent for medical treatment. ■ Patients who do not tolerate or whose pain is refractory to medical management may be candidates for microvascular decompression or ablation.
■ FACIAL NERVE PARALYSIS Perspective The acute onset of facial nerve paralysis often will prompt an ED visit, when early diagnosis and therapy can improve the patient’s chance for recovery of function of the facial nerve. Facial nerve paralysis of acute onset affects approximately 20 to 25 persons per 100,000 population per year, without geographic, gender, or race predilection.14,15
Principles of Disease The facial nerve innervates the muscles of facial expression and the muscles of the scalp and external ear, in addition to the buccinator, platysma, stapedius, stylohyoid, and posterior belly of the digastric muscles. The sensory portion of the nerve supplies the anterior two thirds of the tongue with taste and
Chapter 103 / Brain and Cranial Nerve Disorders
CRANIAL NERVE
CLINICAL FUNCTION RELEVANT TO EMERGENCY MEDICINE
PART III ■ Medicine and Surgery / Section Seven • Neurology
1382
sensation to portions of the external auditory meatus, soft palate, and adjacent pharynx. The parasympathetic portion supplies secretomotor fibers for the submandibular, sublingual, lacrimal, nasal, and palatine glands.16 The nerve originates from the pontomedullary junction of the brainstem. The nerve enters the internal auditory meatus with cranial nerve VIII. Within the temporal bone the facial nerve has four major branches: the greater and lesser superficial petrosal nerves, the nerve to the stapedius muscle, and the chorda tympani. The facial nerve exits the temporal bone at the stylomastoid foramen. The nerve then enters the parotid gland, where it divides to supply the muscles of facial expression.16,17
Pathophysiology Although a complete list of possibilities in the differential diagnosis for facial nerve paralysis would be a long one, the causes pertinent to emergency medicine can be grouped into three specific categories: infectious, traumatic, and neoplastic.
Infection Bell’s Palsy Bell’s palsy, also commonly called idiopathic facial paralysis, has long been postulated to have a viral cause. This disease entity is characterized by an abrupt onset of a lower motor neuron paresis that can progress over 1 to 7 days to complete paralysis. A prodromal illness is described by 60% of patients. Symptoms and signs frequently associated with the facial paresis include ear pain, a perception of sensory change on the involved side of the face, decreased tearing, an overflow of tears on the cheek (epiphora), abnormally acute hearing (hyperacusis), and an impairment or perversion of taste (dysgeusia).18 Treatment approaches can be medical or surgical. The primary medical therapies for Bell’s palsy center on reducing inflammatory changes to the nerve with corticosteroids and treating the presumed viral cause. If these therapies are unsuccessful then surgical decompression may be considered. The use of corticosteroids for Bell’s palsy has been controversial. The rationale for this application of steroid therapy is that edema of the nerve, confined within the facial canal, is thought to cause or contribute to the nerve injury. On the basis of this theory, most experts currently recommend a course of prednisone with an initial dose of 1 mg/kg per day for 7 to 10 days, with or without a short taper.14,17,19,20 The most definitive randomized, double-blind, placebo-controlled trial involving 496 patients showed an improvement in complete recovery of facial nerve function at 3 months from 64% with placebo to 83% with the use of prednisolone in a dose of 25 mg by mouth twice daily.21 Therapy should be started as soon as possible, ideally within the first 24 hours, but is still recommended for patients without contraindications who seek treatment within 1 week of symptom onset.19 A number of publications have advanced the belief that Bell’s palsy may be caused by herpesvirus infection. One study demonstrated herpes simplex virus type 1 DNA in the endoneural tissue of 11 of 14 patients with Bell’s palsy but not in that of control subjects.22 In a trial of prednisone and acyclovir in 99 patients, patients treated with prednisone and acyclovir had a more favorable recovery than that observed in patients receiving prednisone alone.23 A study of 296 patients with Bell’s palsy treated with valacyclovir or placebo in addition to a fixed dose of prednisolone found significant benefit to the addition of valacyclovir, particularly in the setting of severe palsy or in those treated within 24 hours of symptom onset.24 Other studies have found conflicting results. Despite
a lack of overwhelming evidence, the addition of an antiviral agent should be considered in the treatment of Bell’s palsy, especially with severe loss of function. The most commonly recommended antiviral regimens include valacyclovir, 1000 mg orally two times daily for 10 days. Valacyclovir and famciclovir have better oral absorption, are better tolerated, and are dosed less frequently, resulting in higher compliance. Accordingly, they have been recommended as alternatives to acyclovir.17,19,20,22,25 As with steroid therapy, although earlier treatment is preferred, treatment should be considered for patients presenting within 1 week of symptom onset.
Ramsay Hunt Syndrome Ramsay Hunt syndrome (herpes zoster oticus) is characterized by unilateral facial paralysis, a herpetiform vesicular eruption, and vestibulocochlear dysfunction. The vesicular eruption may occur on the pinna, external auditory canal, tympanic membrane, soft palate, oral cavity, face, and neck as far down as the shoulder. The pain is considerably more severe than that associated with Bell’s palsy, and it frequently is out of proportion to physical findings. In addition, outcomes are worse than with Bell’s palsy, with a lower incidence of complete facial recovery and the possibility of sensorineural hearing loss. Therapy is similar to that for Bell’s palsy. Both prednisone and antiviral therapy for 7 to 10 days are advocated.17,26,27
Lyme Disease Lyme disease is the most frequent vector-borne infection in the United States. It is caused by the spirochete Borrelia burgdorferi and is spread by the bite of Ixodes genus ticks. Neurologic manifestations can arise in any phase of the disease, and the incidence of facial palsy in patients with neurologic involvement is 35 to 51%. In regions in which Lyme disease is endemic, it has been shown to be the leading cause of facial paralysis in children, responsible for one half of all pediatric cases of facial nerve paralysis.28,29 Bilateral facial nerve paralysis is rare but can occur with systemic infections. The two diseases most commonly associated with bilateral simultaneous onset of facial paralysis are Lyme disease and infectious mononucleosis. Bilateral facial paralysis should be considered to be a manifestation of Lyme disease until further testing excludes this diagnosis.20,28-30 The evaluation and treatment of Lyme disease are discussed in Chapter 132.
Bacterial Infections Facial paralysis can be caused by acute bacterial infections of the middle ear, mastoid, or external auditory canal. In the preantibiotic era, facial paralysis was associated with acute otitis media in approximately 2% of cases; today, however, it occurs in only 0.2% of cases. Treatment consists of intravenous antibiotics and myringotomy for decompression. Malignant otitis externa can be associated with facial paralysis. This disease entity is most commonly seen in immunocompromised patients and usually is caused by Pseudomonas infection. Treatment involves prolonged intravenous anti pseudomonal antibiotic therapy and may require surgical débridement.20,31
Trauma In patients with head trauma, the facial nerve is the most commonly injured cranial nerve. The cause generally is a temporal bone fracture with nerve transection. Surgical exploration is warranted if there is firm evidence that the nerve has been transected, indicated by a sudden onset of complete unilateral facial paralysis, loss of electrical activity, and evidence of a displaced fracture involving the facial canal.
1383
Neoplasm
Clinical Features and Differential Considerations The medical history should focus on onset of the paralysis, concentrating on timing and rapidity of onset and looking for any associated signs and symptoms. A rapid onset of facial paralysis with dysgeusia and hyperacusis preceded by a viral prodrome is suggestive of Bell’s palsy. A history of recurrent ipsilateral paralysis or slow progression of symptoms is more characteristic for a tumor. Associated cranial nerve abnormalities, although occasionally seen with Bell’s palsy, also point to the possibility of a tumor or ischemic insult. The Ramsay Hunt syndrome causes significant pain and a vesicular rash, although the rash may follow the facial paresis by a few days. Significant anatomic abnormalities on visual or otoscopic inspection of the ipsilateral ear will be found with bacterial otitis media and otitis externa. Finally, systemic symptoms or bilateral facial paresis, especially in endemic areas, should raise the possibility of Lyme disease.
Diagnostic Strategies The diagnostic workup of acute facial nerve paresis is based on whether the clinical picture is suggestive of a disease process other than Bell’s palsy. If the clinical history is classic for Bell’s palsy, then no imaging or laboratory studies are required. Of note, any history of possible exposure warrants serologic evaluation for Lyme disease. Although outpatient testing including electroneurography may ultimately be performed, this usually is not part of the initial evaluation. The physical examination finding of a “central” seventh nerve paralysis (upper face–sparing) should prompt imaging with computed tomography (CT) or MRI, and consideration should be given to the possibility of an acute stroke or other hemispheric lesion. History or physical examination findings suggestive of a possible tumor require imaging to rule out a neoplasm. The study of choice will depend on the institution and preferences of the consultant.
Disposition The vast majority of patients who have a seventh nerve paralysis will have a clinical diagnosis of Bell’s palsy and may be discharged with referral for short-term follow-up. Patients with a possible hemispheric process such as stroke or tumor should be hospitalized for further evaluation. Patients suspected of having Lyme disease require immediate initiation of appropriate antibiotic therapy. In patients with a peripheral facial nerve paralysis, the ipsilateral eye should be patched, and consideration should be given to ophthalmologic follow-up, because there is a high rate of corneal abrasions and corneal dryness associated with the inability to properly blink or completely close the eye.
■
Recent literature highlights significant potential benefit for patients with clinical evidence of Bell’s palsy when they are treated early in the course with corticosteroids. The additional benefit of adding antiviral medication is controversial, but this treatment probably is warranted in patients with severe loss of function. ■ Slowly progressive facial paralysis is suggestive of a neoplasm. Recurrent unilateral paralysis may occur with Bell’s palsy but frequently (30%) is seen in patients with tumor. ■ Simultaneous bilateral facial paralysis is suggestive of Lyme disease, which must be considered as a possible cause, especially in endemic regions. ■ Patients who have facial muscle paresis with intact forehead movement should be considered to have an upper motor neuron lesion until the diagnostic investigation proves otherwise.
■ VESTIBULAR SCHWANNOMA Perspective Vestibular schwannoma, formally referred to as acoustic neuroma, is a rare but important cause of sensorineural hearing loss. The annual incidence of VS is 1 case per 100,000 population, with a mean age at the time of detection of 46 to 58 years.33 The female-to-male ratio is 1.5 : 1. Vestibular schwannoma is very rarely bilateral, occurring in this form in approximately 5% of cases and generally associated with type II neurofibromatosis. Although histologically benign, vestibular schwannoma can cause neurologic damage by direct compression on the eighth cranial nerve and the other structures in the cerebellopontine angle.34
Principles of Disease Vestibular schwannoma arises from the Schwann cells covering the vestibular branch of the eighth cranial nerve as it passes through the internal auditory canal. The tumor may compress the cochlear (acoustic) branch of the eighth cranial nerve, causing hearing loss, tinnitus, and dysequilibrium. Continued growth of the tumor may result in compression of structures in the cerebellopontine angle, where the facial and trigeminal nerves may be compressed and damaged. Larger tumors may further encroach upon the brainstem and if large enough may compress the fourth ventricle, ultimately resulting in signs of increased intracranial pressure (ICP).35
Clinical Features Asymmetrical sensorineural hearing loss is the hallmark of vestibular schwannoma. Up to 15% of patients with this tumor, however, will have normal results on an audiogram. These patients typically have symptoms such as unilateral tinnitus, imbalance, headache, fullness in the ear, otalgia, or facial nerve weakness. Thus, patients with asymmetrical symptoms should be further evaluated for vestibular schwannoma even with normal findings on the audiogram.36 Vestibular schwannomas are extremely slow-growing tumors, averaging an approximately 1-mm increase per year, although many do not grow at all on serial examinations.37 Symptom onset is therefore generally quite gradual. In one series of 126 cases, the average time from symptom onset to discovery of a vestibular schwannoma was approximately 4 years.38
Chapter 103 / Brain and Cranial Nerve Disorders
Tumors of the facial nerve itself, or tumors anywhere along the course of the facial nerve that invade or compress the nerve, may lead to facial paralysis. Typically the course is progressive over at least 3 weeks. A sudden onset of paralysis, however, does not rule out an underlying tumor, because facial paralysis secondary to a neoplasm is of sudden onset in approximately 25% of cases.32 A neoplastic cause should be suspected in patients who suffer from recurrent ipsilateral facial paralysis, significant pain, prolonged symptoms, or any other concomitant cranial nerve abnormality.
KEY CONCEPTS
1384
PART III ■ Medicine and Surgery / Section Seven • Neurology
Diagnostic Strategies When vestibular schwannoma is suspected, the patient should be evaluated with an audiogram or a gadolinium-enhanced MRI. This imaging technique is extremely sensitive and has led to earlier diagnosis and a decrease in mean size at detection of vestibular schwannoma. CT lacks the necessary sensitivity in the posterior cranial fossa to reliably rule out the presence of vestibular schwannoma. The smaller the tumor at the time of diagnosis, the more options there are for therapy and the better the potential prognosis.34
Differential Considerations A majority of disease entities included in the differential diagnosis for acoustic neuroma cause symmetrical sensorineural hearing loss. Asymmetrical sensorineural hearing loss has few causes other than vestibular schwannoma. Ménière’s disease may present a diagnostic dilemma because it can be asymmetrical. Ménière’s disease may be differentiated from vestibular schwannoma in that the tinnitus of Ménière’s disease usually is intermittent, whereas the tinnitus of vestibular schwannoma typically is continuous. In addition, patients with Ménière’s disease typically describe true vertigo, whereas patients with a vestibular schwannoma are more likely to describe imbalance or dysequilibrium. Vestibular schwannomas account for 80% of all cerebellopontine angle tumors. Among all other lesions, meningioma is the most common. Meningiomas more frequently cause symptoms of facial palsy or trigeminal nerve abnormality. Of note, however, considerable similarity between the clinical picture of a meningioma and that of vestibular schwannoma in the cerebellopontine angle has been described.39
KEY CONCEPTS ■
The onset of unilateral auditory symptoms, especially sensorineural hearing loss, requires evaluation and referral to an ear, nose, and throat specialist. ■ Neurologic symptoms of lower cranial nerve dysfunction, ataxia, or raised ICP may be caused by a benign tumor of the cerebellopontine angle. ■ The smaller the tumor at diagnosis, the better the long-term outcome with definitive treatment. followed in order by the trochlear and abducens nerves. In one large series in Japan, the incidence of cranial nerve palsies was 1.0% among diabetics and 0.1% among nondiabetics.40,41 Whereas ophthalmoplegia appears to be closely related to diabetes, facial palsy is less strongly correlated with this disease.40
Principles of Disease The pathologic basis of diabetic mononeuropathy appears to be ischemia of the affected cranial nerve caused by occlusion of an intraneural nutrient artery serving the nerve. This occlusion leads to injury located primarily in the center of the nerve, because the core fibers are more dependent on the supply from such nutrient arteries. The peripheral fibers are less affected because they also are supplied by collateral vessels. In the oculomotor nerve, the preservation of the circumferentially located parasympathetic fibers explains the pupillary sparing that usually is found in this syndrome. In two studies, the microvascular changes in the intraneural arteries that lead to occlusion were noted in diabetic patients but absent in nondiabetics.42,43
Management
Clinical Features
Vestibular schwannoma may be removed surgically or ablated with stereotactic radiation. In general, tumors larger than 3 cm are recommended for microsurgery, because radiation treatments, such as with the Gamma Knife or linear accelerator, are less effective for local control and growth arrest in larger masses. Smaller tumors are amenable to use of stereotactic radiation, which may have greater salvage rates of facial nerve function and hearing. Stereotactic radiation therapy generally has good long-term outcomes of local growth arrest, with nerve salvage approaching 90% or greater. Injuries to the trigeminal, facial, and acoustic nerves, and to the cerebellum, are possible complications of both procedures. In patients who are minimally symptomatic with small tumors, serial monitoring with MRI is a viable nonsurgical option. All patients should be evaluated by a specialist in the evaluation and treatment of vestibular schwannoma, because smaller tumor size at detection is associated with a better long-term outcome.33,37
Patients typically describe acute onset of unilateral retro-ocular and supraorbital pain, diplopia, and ptosis.41 The physical manifestations of a third cranial nerve palsy include the inability to move the eye superiorly and medially, accompanied by ptosis. The pupillary light reflex usually is present. Although a less common finding, the fourth and sixth cranial nerves may be affected. Patients with a fourth cranial nerve palsy are unable to move the eye inferolaterally, and those with a sixth cranial nerve palsy are unable to move the eye laterally. Because of the long intracranial course of the sixth nerve, a patient with an isolated sixth nerve palsy should be evaluated for an intracranial lesion or increased ICP.44
Disposition Patients with suspected acoustic neuroma should be referred for an audiogram or MRI and evaluation by a specialist in either otolaryngology or neurosurgery.
■ DIABETIC CRANIAL MONONEUROPATHY Perspective Cranial mononeuropathies occur uncommonly, usually are a complication of diabetes, and most often affect the extraocular muscles. The oculomotor nerve is most commonly affected,
Differential Considerations Evaluating cranial nerve dysfunction requires a thorough history and physical examination and cranial imaging, usually with MRI. Diabetic mononeuropathy should be considered a diagnosis of exclusion, with considerations in the differential diagnosis including trauma, tumor, vertebrobasilar ischemia, aneurysm, and hemorrhage into the brainstem.45
Management Treatment consists of patching the affected eye and administration of analgesics and antiplatelet therapy. The prognosis is good. If the neuropathy does not begin to resolve within 3 to 6 months, or if more than one nerve is affected, another cause should be sought. Complete resolution is expected within the first year. Antioxidant preparations, including α-lipoic acid, have been used therapeutically and have not shown harm, but
such agents have yet to be shown to have convincing clinical effect.46
■
Diabetic neuropathy is a diagnosis of exclusion because no definitive diagnostic testing is available. ■ Both ischemic and hemorrhagic brainstem lesions must be ruled out in the case of an acute ophthalmoplegia. ■ Extraocular mononeuropathy is sufficiently common in patients with diabetes mellitus that its occurrence in isolation warrants evaluation of the patient for previously undiagnosed diabetes.
■ CEREBRAL VENOUS THROMBOSIS Perspective No precise studies of the epidemiology of cerebral venous thrombosis (CVT) have been performed. In case series, the median patient age is approximately 37 years, with a femaleto-male ratio of 3 : 1.47
Principles of Disease Cerebral blood is drained by several major veins that lead into the dural sinuses. The major dural sinuses are the superior sagittal sinus, the inferior sagittal sinus, the straight sinus, the lateral sinuses, and the sigmoid sinuses. The variability in symptoms and signs in patients who present with CVT stems from differences in thrombus location and acuity of thrombus formation. Symptoms of intracranial hypertension are present in most patients with sinus thrombosis, whereas those with thrombosis of the cerebral veins are thought to be more prone to hemorrhagic infarction and localizing neurologic deficits.48 As with venous thrombosis in other locations, multiple causes and predisposing factors for CVT are recognized. Underlying causes often are divided into infectious and noninfectious categories. Infectious causes include local infections, such as sinusitis, otitis media, cellulitis on the face, and systemic infections. Noninfectious causes include direct injury to the cerebral venous system from trauma, surgery, tumor, dehydration, or any other condition that may predispose the patient to development of a hypercoagulable state.49
Clinical Features The symptoms and signs associated with CVT are quite varied. Headache is the primary feature of CVT in 74 to 92% of affected patients.49,50 Papilledema is noted in 28 to 45% of cases.47,50,51 Lethargy, decreased level of consciousness, or mental status changes may be noted. Seizures occur in 35 to 50% of patients in the acute phase.47,49,51 In addition to the location and acuity of thrombosis formation, a patient’s symptom onset will vary in accordance with the extent of collateral vessel growth in the venous territory. Early thrombotic changes may be well compensated for by the collateral venous drainage. Symptoms will appear only when the compensation for venous thrombosis is no longer sufficient. Variability in collateralization between patients also adds to the variability and time course of symptomatology. Two national and international observational studies document an average time from symptom onset to diagnosis of 7 days, reflecting the difficulty in diagnosing this rare disease entity.47,51,52 The reported incidence of focal neurologic signs, including seizures, on clinical
Diagnostic Strategies The gold standard modality for the diagnosis of CVT has shifted in recent years from cerebral angiography to magnetic resonance venography (MRV). CT scanning is useful in the initial workup of the patient with possible CVT, but noncontrast CT is neither sensitive nor specific enough to reliably confirm or exclude the diagnosis. Findings on CT that are consistent with CVT include hyperdensity of a thrombosed sinus, brain edema, and hemorrhage secondary to venous congestion. CT venography is both more sensitive and more specific in diagnosing CVT. Similar to CT scanning, MRI also can demonstrate local changes secondary to venous congestion, such as brain edema or hemorrhage. In addition, MRI can demonstrate the possibility of CVT based on the lack of a “flow void.” On conventional MRI, a flow void indicates the presence of blood flow within the sinus, whereas the absence of a flow void indicates a possible thrombus. Diagnostic accuracy, however, is greatly improved through use of MRV. This technique takes advantage of the MRI signal characteristics of flowing blood to create images of venous structures. Combining these imaging techniques further enhances diagnostic accuracy. For imaging a particular dural sinus, presence of the sinus on conventional MRI and lack of flow on MRV are diagnostic of a sinus thrombosis. This combined approach has diagnostic sensitivity similar to that of conventional angiography.49,53 Two small studies show similar sensitivity between MRV and CT venography for the diagnosis of CVT when the CT study is performed on a multidetector row CT scanner. Both studies, involving a total of 69 patients, showed 100% sensitivity of CT venography for CVT in comparison with MRV.54,55 The sensitivity of CT venography performed by scanners that do not use multidetector row technology is unknown. Several small studies have attempted to evaluate the usefulness of the D-dimer assay as a screening tool to exclude CVT, particularly when MRI or CT venography is not available. Although the reported sensitivity rates are fair at 83 to 100%, larger prospective studies need to be done to further define the role of D-dimer in the evaluation of CVT, because several case reports have noted normal D-dimer levels in the setting of documented CVT.56-59 In general, although a normal Ddimer level does not exclude the diagnosis of CVT, it does appear to make this diagnosis much less likely, particularly in a patient with symptoms of less than 2 weeks in duration.
Differential Considerations Considerations in the differential diagnosis of CVT include the conditions that cause patients to present with the new onset of neurologic deficits, alteration in consciousness, or severe headache. A diagnosis of CVT should be considered in a patient with such symptoms when the etiology is unclear, presence of having a hypercoagulable state is likely, and the head CT scan is normal in appearance or shows subtle signs of CVT.
Management CVT is a relatively rare disease, and controlled studies evaluating its treatment are lacking. Current therapeutic consensus
Chapter 103 / Brain and Cranial Nerve Disorders
KEY CONCEPTS
1385
examination varies between series, ranging from 25 to 71%.49,50 Because of the broad spectrum of possible clinical features, the diagnosis of CVT may be difficult but should be a consideration in any patient with unexplained headache, especially in combination with focal neurologic deficit, papilledema, or seizures.
PART III ■ Medicine and Surgery / Section Seven • Neurology
1386
strongly recommends systemic anticoagulation with lowmolecular-weight or unfractionated heparin to prevent further clot formation and to promote recanalization, even in patients with intracranial hemorrhage on initial imaging.49,50,60,61 In one placebo-controlled randomized trial comprising 20 patients, anticoagulation with heparin to a target partial thromboplastin time (PTT) of 80 to 100 seconds demonstrated benefit, even in patients in which evidence of intracranial hemorrhage was seen on the CT scan before anticoagulation.62 In another study of 60 patients randomized to receive placebo or low-molecularweight heparin, no statistical benefit was shown for treatment.63 Two large observational trials showed improvement in modified Rankin scale at follow-up in the anticoagulated groups, although the trials were not randomized.47,51 Despite a paucity of randomized controlled trials, expert opinion favors anticoagulation in all groups unless another contraindication is present.64 Catheter-based intervention with thrombolysis has been attempted in multiple case series using either urokinase or tissue plasminogen activator. Thrombolysis was shown to be relatively safe and relatively successful in very small case series.61 In one nonrandomized study of 40 patients, 20 received systemic heparin and 20 received catheter-based infusion of urokinase followed by systemic heparin. Despite initially worse neurologic function in the thrombolysis group, a significant difference in neurologic function favoring thrombolysis was observed at discharge.65 Although this therapy is promising, it should be considered only for patients with symptoms of decreased level of consciousness, elevated ICP, or rapid deterioration on neurologic examination.
Disposition All patients with suspected CVT should be admitted to a unit capable of providing a high level of care with neurologic consultation. Patients should be anticoagulated if no contraindication exists, and catheter-based thrombolysis should be considered in patients with depressed mental status or focal findings on neurologic exam.
KEY CONCEPTS ■
CVT is a relatively rare entity, and only awareness of and familiarity with the clinical presentation will lead to the correct diagnosis. ■ The onset may be insidious with a considerable delay between onset and arrival in the treatment setting. ■ The differential diagnosis for CVT should consider other conditions that cause patients to present with new-onset neurologic deficits, alteration in consciousness, or severe headache. CVT is more likely to be present in such patients when the etiology is unclear, the patient is suspected of having a hypercoagulable state, and the head CT is normal in appearance or shows subtle signs of CVT. ■ Noncontrast CT scanning is not adequate to rule out CVT. An MRI with MRV is recommended, although multidetector row CT venography is an acceptable alternative. ■ Treatment of most patients with CVT should include systemic anticoagulation, even in the setting of hemorrhagic cerebral infarcts, unless another contraindication exists.
■ MULTIPLE SCLEROSIS Perspective Multiple sclerosis (MS) is an inflammatory disease that affects the central nervous system (CNS). Although the exact etiology remains uncertain, the pathologic manifestation of this inflammatory disease is a demyelination of discrete regions (plaques) within the CNS with a relative sparing of axons. The clinical picture is highly variable but is classically characterized by episodes of neurologic dysfunction that evolve over days and resolve over weeks. MS has an overall prevalence in the United States of 0.1%. The peak age at onset is 25 to 30 years, with women being slightly younger at onset than men. The incidence in women exceeds that of men by a ratio of 1.8 : 1. The worldwide prevalence is greatest in the United Kingdom, Scandinavia, and North America. Epidemiologic studies indicate that both genetic and environmental factors are associated with an increased incidence of this disease. MS has a 30% concordance rate between monozygotic twins, and 20% of patients with MS have at least one affected relative. MS is more common in temperate climates. It is rare between 23 degrees north and south latitudes but has a rising incidence above and below 50 degrees north and south latitudes, respectively. Although no exact environmental factor has been identified, if a person emigrates from an area of high prevalence to an area of low prevalence before the age of 20, the risk is diminished. MS is rare in Africans and Asians, but African Americans have a higher incidence than their relatives who remain in Africa.66 In addition, reports of clusters or miniepidemics support environmental factors. Thus, an environmental cause superimposed on genetic susceptibility appears to be a likely etiologic scenario.67,68
Principles of Disease MS is considered to be an organ-specific autoimmune disease. One theory proposes that genetic factors interact with an environmental trigger or infection to establish pathologically autoreactive T cells in the CNS. After a long and variable latency period (typically 10 to 20 years), a systemic trigger, such as a viral infection or superantigen, activates these T cells. The activated T cells, on reexposure to the autoantigen, initiate the inflammatory response. This sets off a complex immunologic cascade that leads to the demyelination characteristic of MS. This demyelination process releases CNS antigens that are hypothesized to initiate further episodes of autoimmuneinduced inflammation. The mechanisms underlying this autoimmunity in MS are unknown.69
Clinical Features The clinical picture in MS is one of marked heterogeneity. The classic clinical syndrome consists of recurring episodes of neurologic symptoms that rapidly evolve over days and slowly resolve. Variability occurs in age at onset, location of CNS lesions, frequency and severity of relapses, and the degree and time course of progression. The clinical features of MS can be divided into areas of specific CNS impairment: cognition, cranial nerves, motor pathways, sensory pathways, cerebellar pathways, and bowel, bladder, and sexual dysfunction.66 Patients with MS have frequent complaints of poor memory, distractibility, and a decreased capacity for sustained mental effort. Formal neuropsychological testing suggests that cognitive involvement is common and underreported. Specifically,
Diagnostic Strategies Although no laboratory tests are diagnostic for MS, one clinical feature remains relatively unique to this disease: Uhthoff’s phenomenon, temporary worsening of current or preexisting signs or symptoms of MS secondary to small increases in the patient’s body temperature. Accordingly, exercise, a hot bath,
1387
exposure to a warm environment, or fever can bring about Uhthoff’s phenomenon. This phenomenon reflects subclinical demyelination or preexisting injury to nerves without obvious significant clinical involvement before heat exposure or temperature elevation.66 The clinical diagnosis rests on occurrence of at least two clinical episodes with different neurologic symptoms at different times. Thus, MS commonly has been characterized as a disorder with lesions that differ in time and space. It also has been described as a relapsing-remitting disorder with symptoms that fluctuate over time. Findings on cerebrospinal fluid (CSF) analysis are abnormal in 90% of the cases. Fifty percent of patients will have pleocytosis, with more than 5 lymphocytes per high-power field in the CSF. Approximately 70% of patients will have an elevated gamma globulin level, with immunoglobulin G (IgG) ranging from 10 to 30% of the CSF total protein. Electrophoresis of the CSF demonstrates oligoclonal bands of IgG in 85 to 95% of patients who carry a diagnosis of MS; however, oligoclonal bands of IgG also are seen with neurosyphilis, fungal meningitis, and other CNS infections. Lumbar puncture should be considered for all patients with suspected MS, but mass lesions and elevated ICP should be ruled out before lumbar puncture.77 The initial imaging test to aid in the diagnosis of multiple sclerosis is MRI. MRI is a sensitive test for the detection of lesions consistent with MS and also is useful to assess disease severity.78 The lesions of MS typically appear hyperintense, or bright white, on T2-weighted or fluid-attenuated inversion recovery (FLAIR) MRI studies. Lesions usually are multiple and commonly are found in the periventricular white matter.79 In patients with an initial neurologic event consistent with CNS demyelination and an MRI cranial study showing multiple white matter lesions, the 5-year risk of developing MS is 60%. Patients with similar clinical syndromes and a normal MRI appearance have less than a 5% 5-year risk.80
Differential Considerations Other diseases that affect the CNS white matter may be clinically and radiographically similar to MS. Considerable care must be taken to exclude these disease processes before making a diagnosis. These include CNS tumors (especially lymphomas and gliomas), spinal cord compression, vasculitides, Behçet’s disease, neurosarcoidosis, postinfectious and postvaccinal encephalomyelitis, human immunodeficiency virus (HIV) encephalopathy, Lyme disease, and vitamin B12 deficiency.
Management Management of patients with MS has essentially three aspects: (1) therapies aimed at halting the progression of the disease, (2) treatment for acute exacerbations, and (3) therapies designed to modify complications. Therapies aimed at halting disease progress are based primarily on the use of either β-interferon or glatiramer acetate. The interferons are a group of natural compounds with antiviral and immunomodulatory actions, which are retained by the recombinant preparations used in therapy for MS, interferon beta-1a and interferon beta-1b. Side effects include flulike symptoms, depression, anxiety, and confusion. In one study, 560 patients with MS were randomly assigned to receive subcutaneous interferon beta-1a or placebo (n = 187) three times a week for 2 years. The relapse rate was significantly lower at 1 and 2 years with interferon beta-1a than with placebo. The time to first relapse was prolonged significantly and the accu-
Chapter 103 / Brain and Cranial Nerve Disorders
neuropsychological testing has shown that 43 to 65% of patients with MS have some degree of cognitive impairment.70,71 Of note, a correlation has been found between the MRI-based total lesion load and presence of cognitive impairment.72 Cranial nerve dysfunction is common in MS. The most common associated cranial nerve abnormality is optic neuritis, a unilateral syndrome characterized by pain in the eye and a variable degree of visual loss affecting primarily central vision. Within 2 years of an attack of optic neuritis, the risk of MS is approximately 20%, and within 15 years, it is approximately 45 to 80%.73,74 Optic neuritis often is the first symptom of MS.75,76 As a result of lesions in the vestibulo-ocular connections, the oculomotor pathways also may be affected. The deficit may manifest as diplopia or nystagmus. The nystagmus may be severe enough that the patient may complain of oscillopsia (a subjective oscillation of objects in the visual field). Cranial nerve impairment also may include impairment of facial sensation, which is relatively common. Unilateral facial paresis also may occur. In addition, the occurrence of trigeminal neuralgia in a young person may be an early sign of MS. Motor pathways also are commonly involved. Specifically, corticospinal tract dysfunction is common in patients with MS. Paraparesis or paraplegia is all too common and occurs with greater frequency than upper extremity lesions, owing to the common occurrence of lesions in the motor tracts of the spinal cord. In patients with significant motor weakness, spasms of the legs and trunk may occur on attempts to stand from a seated position. This dysfunction is manifested on physical examination as spasticity that typically is worse in the legs than in the arms. The deep tendon reflexes are markedly exaggerated, and sustained clonus may be demonstrated. Although these symptoms frequently are bilateral, they generally are asymmetrical.66 Sensory manifestations are a frequent initial feature of MS and will be present in nearly all patients at some point during the course of the disease. Sensory symptoms are commonly described as numbness, tingling, “pins and needles” paresthesias, coldness, or a sensation of swelling of the limbs or trunk.66 Impairment of the cerebellar pathway results in significant gait imbalance, difficulty with coordinated actions, and dysarthria. Physical examination reveals the typical features of cerebellar dysfunction, including dysmetria, dysdiadochokinesis (an impairment of rapid alternating movements), a breakdown in the ability to perform complex movements, an intention tremor in the limbs and head, truncal ataxia, and dysarthria.66 Impairment of bowel, bladder, and sexual function also is common. The extent of sphincter and sexual dysfunction usually parallels the motor impairment in the lower extremities. Urinary frequency may progress to urinary incontinence with progression of the disease. An atonic bladder may develop, which empties by simple overflow and often is associated with the loss of perception of bladder fullness and with anal and genital hypoesthesia. Constipation becomes common over time, and almost all patients with paraplegia require special measures to maintain effective bowel habits. Sexual dysfunction, although frequently overlooked, is very common in MS. Approximately 50% of patients become completely sexually inactive as a result of this disease.66
PART III ■ Medicine and Surgery / Section Seven • Neurology
1388
mulation of brain lesions on MRI was lower in the treatment group than in the placebo group. The investigators concluded that subcutaneous interferon beta-1a is a well-tolerated and effective treatment for relapsing-remitting MS in terms of relapse rate, defined disability, and all MRI outcome measures.81 β-Interferon also has been shown to retard progression to clinically definite MS and to decrease the total number of brain lesions seen on subsequent MRI studies in patients who have their first demyelinating episode with MRI abnormalities at initial presentation.82-85 This finding highlights the importance of early evaluation and treatment. Glatiramer acetate also has successfully been used in the treatment of MS. This agent is a mixture of synthetic polypeptides designed to mimic myelin basic protein. The mechanism of action by which glatiramer acetate exerts its effect is unknown, but it is thought to modify the immune processes responsible for the pathogenesis of MS. In one study, 251 patients with relapsing-remitting MS were randomized to receive daily subcutaneous injections of glatiramer acetate (previously called copolymer 1) or placebo for 24 months. Patients receiving glatiramer acetate experienced significantly fewer relapses and were more likely to demonstrate neurologic improvement, whereas those receiving placebo were more likely to worsen. This drug generally is quite well tolerated.86 Current recommendations for management of relapsingremitting MS are to initiate treatment with β-interferon or glatiramer acetate. Such regimens have been demonstrated to decrease the volume of plaques seen on MRI and to diminish relapses.69 Immunosuppressive agents, including mitoxantrone and azathioprine, also have been shown to be effective in reducing progression of disease but, in view of concerns over side effects, generally are used as second-line agents.87,88 Acute exacerbations of MS also should be targets for therapy. Although most such episodes will resolve without therapy, steroids have been demonstrated to diminish the duration of acute exacerbations. More than 85% of patients with relapsing-remitting MS show improvement with intravenous methylprednisolone. Intravenous steroids have been shown in controlled trials to speed the recovery from the visual loss of optic neuritis when compared with placebo. In addition, when patients with acute optic neuritis are treated with high-dose intravenous steroids, the 2-year rate of development of MS is reduced, although this effect diminishes over time.74,89 Of interest, oral prednisone was not found to be helpful in the optic neuritis trials and was associated with a potential increase in the number of optic neuritis episodes. The current standard therapy for an acute exacerbation in MS is intravenous methylprednisolone. A typical dose administered intravenously is 250 to 500 mg every 12 hours for 3 to 7 days. Whether this should be followed by an oral prednisolone taper remains controversial. Potential adverse effects of methylprednisolone therapy include fluid retention, gastrointestinal hemorrhage, anxiety, psychosis, infection, and osteoporosis. Several therapies directed toward the complications of MS may be helpful. The associated spasticity generally is treated with baclofen. This is a highly effective therapy aimed at
reducing the painful flexor and extensor spasms. A major side effect is drowsiness, which generally diminishes with continued use. Higher-dose therapy can cause confusion, especially in the setting of baseline cognitive impairment. For patients with intractable spasticity, baclofen is available for intrathecal administration by either bolus therapy or continuous implanted pump therapy. Additional therapeutic agents for control of spasticity include tizanidine, diazepam, and dantrolene. The tremor and ataxia associated with MS occasionally are treated with propranolol, diazepam, or clonazepam. The results of these therapies, however, generally are unsatisfactory. Pain often is associated with MS and affects the shoulders, pelvic girdle, and face. The facial pain may be indistinguishable from that of trigeminal neuralgia. Treatment options include carbamazepine, baclofen, and tricyclic antidepressants. Fatigue, which is common, may be ameliorated with amantadine. This agent produces partial relief for a minority of patients. In controlled studies, the effect is only slightly better than placebo.69
Disposition Patients with a history of MS who seek treatment for significant symptoms must first be evaluated to rule out other, non– MS-related pathology. Also, the presence of other systemic illnesses, especially infections, which can worsen the symptoms of MS, should be excluded. If the problem is thought to be an exacerbation of MS, most patients will require hospital admission for intravenous steroid therapy. An alternative to hospitalization may be to initiate intravenous steroids in the ED and to arrange for a next-day follow-up visit with the primary care physician or neurologist if outpatient intravenous steroid administration is an option. Patients with the new onset of symptoms suggestive of MS should be admitted or referred to a neurologist, depending on the type and severity of symptoms.
KEY CONCEPTS ■
Any patient with a long-term illness, such as MS, must be evaluated to rule out pathologic processes not related to that illness before an exacerbation of the illness can be assumed to be the cause of any problems experienced by the patient. ■ Therapy for patients with MS will require consultation with the patient’s primary care provider or neurologist to provide consistent disease management. ■ Intravenous methylprednisolone effectively promotes earlier resolution of recurrences. ■ Intravenous methylprednisolone has been shown to speed the recovery from vision loss from optic neuritis.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 104
Spinal Cord Disorders
Andrew D. Perron and J. Stephen Huff
■ PERSPECTIVE Spinal cord disorders encompass a wide range of pathologic entities and affect all age groups. Some spinal cord disorders may have catastrophic outcomes if not recognized early in the clinical course. The ultimate neurologic outcome with many of these disorders may depend on expeditious recognition in the emergency department (ED), with appropriate initial investigations, neuroimaging, management, and consultation for definitive therapy. Diagnosis of these disorders may be extremely challenging, and certain disorders may mimic other disease processes until late in the clinical course, when there is clear neurologic impairment. As with many disease processes affecting the nervous system, correct diagnosis and appropriate management require knowledge of the anatomic organization of the spinal cord and skill in taking the history and in performing the neurologic examination. This chapter generally is concerned with processes affecting the spinal cord and its vascular supply, as well as processes compressing the spinal cord. Direct trauma and mechanical instability of the spinal column are discussed in Chapter 40.
■ PRINCIPLES OF DISEASE Anatomy In adults, the spinal cord is approximately 40 cm long and extends from the foramen magnum, where it is continuous with the medulla oblongata, to the body of the first or second lumbar vertebra. Similar to the brain, the spinal cord is covered by three meningeal layers: the inner pial layer, the arachnoid, and the outer dural layer. At its lower end, the spinal cord tapers into the conus medullaris, where several segmental levels are represented in a small area. The lumbar and sacral nerve roots form the cauda equina as they descend caudally in the thecal sac before exiting the spinal canal at the respective foramina. The non-neural filum terminale runs from the tip of the conus and inserts into the dura at the level of the second sacral vertebra. Two symmetrical enlargements of the spinal cord contain the segments that innervate the limbs. The cervical enlargement (cord level C5 to T1) gives rise to the brachial plexus and subsequently to the peripheral nerves of the upper extremity. The lumbar enlargement (L2 to S3) gives rise to the lumbosacral plexus and peripheral nerves of the lower extremity. The space surrounding the spinal cord within the spinal canal is
reduced in the area of the enlargements, potentially leaving the cord more vulnerable to compression in these regions. At each segmental level, anterior (ventral) and posterior (dorsal) roots arise from rootlets along the anterolateral and posterolateral surfaces of the cord. At each level, the anterior root conveys the outflow of the motor neurons in the anterior horn of the spinal cord, and the posterior root contains sensory neurons and fibers that convey sensory inflow. The arterial supply of the spinal cord is derived primarily from two sources. The single anterior spinal artery arises from the paired vertebral arteries. This anterior spinal artery runs the entire length of the cord in the midline anterior median sulcus and supplies roughly the anterior two thirds of the spinal cord. Blood supply to the posterior third of the spinal cord derives from the smaller paired posterior spinal arteries. The anterior and the posterior spinal arteries receive segmental contributions from radicular arteries, the largest being the radicular artery of Adamkiewicz, which typically originates from the aorta between T8 and L4. The venous drainage of the cord largely parallels the arterial supply. The internal anatomy of the spinal cord is divided into central gray matter, which contains cell bodies and their processes, and surrounding white matter, where the ascending and descending myelinated fiber tracts are located. These fiber tracts are organized into discrete bundles, with the ascending tracts conveying sensory information and the descending tracts conveying the efferent motor impulses and visceral innervation. For clinical purposes, neuroanatomy of the spinal cord may be greatly simplified, as depicted in Figure 104-1. Major ascending sensory tracts are represented on the right side of the figure, with motor tracts on the left side. The posterior columns carry afferent ascending proprioceptive and vibratory information on the ipsilateral side of the cord to the area stimulated; decussation of these fibers occurs in the medulla so that contralateral cortical representation ultimately occurs. In a portion of the lateral column of white matter, the lateral spinothalamic tract conveys afferent information about pain and temperature. (Tracts are named with their point of origin first—the spinothalamic tract, for example, arises in the spinal cord and travels to the thalamus.) The tract is laminated so that sacral fibers are represented most laterally. Crossing of fibers from this tract occurs near the level of entry of the spinal nerve; a cord lesion affecting one lateral spinothalamic tract results in decreased or absent pain and temperature perception below the level of injury on the contralateral side of the body. 1389
PART III ■ Medicine and Surgery / Section Seven • Neurology
1390
Lateral corticospinal tract Descending tract Voluntary movement
Posterior columns Ascending proprioceptive and vibratory senses
Lateral spinothalamic tract Ascending pain and temperature information
Figure 104-1. Simplified spinal cord anatomy showing clinically essential motor and sensory tracts. (Photomicrograph courtesy of John Sundsten, Digital Anatomist Project, University of Washington.)
For clinical purposes, the major descending motor tract is represented in the lateral corticospinal tract (which, as the name implies, originates in the cortex and flows toward the spinal cord). This tract also is anatomically organized, with efferent motor axons to the cervical area located medially and the sacral efferent axons located laterally. Decussation of this descending tract occurs in the medulla. The cell bodies of the lower motor neurons (anterior horn cells) are in the ventral portion of the gray matter of the spinal cord.
■ CLASSIFICATION OF SPINAL CORD SYNDROMES The anatomic organization of the spinal cord lends itself to a corresponding anatomic-pathophysiologic classification of cord dysfunction. Any of the different anatomic syndromes may be the final clinical picture of a variety of clinical processes either intrinsic or extrinsic to the spinal cord. The syndromes frequently exist in partial or incomplete forms.
Complete (Transverse) Spinal Cord Syndrome Complete spinal cord lesions may occur as either acute or subacute pathologic processes. A complete spinal cord lesion is defined as a total loss of sensory, autonomic, and voluntary motor innervation distal to the spinal cord level of injury. Reflex responses mediated at the spinal level, such as muscle stretch (“deep tendon”) reflexes, may persist, although they also may be absent or abnormal. Autonomic dysfunction may manifest with hypotension (neurogenic shock) or priapism. The most common cause of the complete transverse cord syndrome is trauma, although this anatomic syndrome is nonspecific as to etiology.1,2 Other causes of acute complete cord syndrome include infarction, hemorrhage, and entities causing extrinsic compression. Of patients in whom complete transverse syndromes develop and persist for more than 24 hours, functional recovery does not occur in 99%.3,4 Any evidence of cord function below the level of injury denotes a partial rather than a complete lesion. Signs such as persistent perineal sensation (“sacral sparing”), reflex rectal sphincter tone or voluntary rectal sphincter contraction, or even slight voluntary toe move-
ment suggest a partial cord lesion, which usually carries a better prognosis than a complete lesion.1 Spinal shock refers to the loss of muscle tone and reflexes with complete cord syndrome during the acute phase of injury. The intensity of the spinal shock increases with affected spinal cord level.5 Spinal shock typically lasts less than 24 hours but has been reported occasionally to last days to weeks.5,6 A marker of spinal shock is loss of the bulbocavernosus reflex, which is a normal cord-mediated reflex that also may be preserved in complete cord lesions. The bulbocavernosus reflex involves involuntary reflex contraction of the anal sphincter in response to a squeeze of the glans penis or a tug on the Foley catheter. The termination of the spinal shock phase of injury is heralded by the return of the bulbocavernosus reflex; increased muscle tone and hyper-reflexia follow later.1,5,6
Incomplete Spinal Cord Lesions Incomplete spinal lesions are characterized by preservation of function of various portions of the spinal cord. Of all incomplete spinal lesions, most can be classified generally as one of three clinical syndromes: central cord syndrome, BrownSéquard syndrome, or anterior cord syndrome (Table 104-1).
Central Cord Syndrome Central cord syndrome, first described by Schneider and colleagues in 1954, is the most prevalent of the partial cord syndromes.7,8 Because of the anatomic organization of the spinal cord, a central cord injury is characterized by bilateral motor paresis, with upper extremities affected to a greater degree than lower extremities, and distal muscle groups affected to a greater degree than proximal muscle groups. Sensory impairment and bladder dysfunction are variable features. At times, burning dysesthesias in the upper extremities may be the dominant feature.9 Central cord injury affects the central gray matter and the central portions of the corticospinal and spinothalamic tracts. It is caused most often by a hyperextension injury, with the postulated pathomechanism being squeezing or pinching of the spinal cord anteriorly and posteriorly by inward bulging of the ligamentum flavum. The most common cause of such injuries is a fall, followed in frequency by a motor vehicle crash.7,8 The result is contusion to the spinal cord, with the central portion being most affected. This injury classically occurs in elderly individuals with degenerative arthritis and spinal stenosis in the cervical area, but may affect any patient with cervical canal narrowing of any etiology (e.g., congenital narrow canal as seen in achondroplasia or canal narrowing from disk protrusion or tumor). The prognosis with central cord syndrome depends on the degree of injury at presentation and patient age.10,11 In patients younger than 50 years of age, more than 80% regain bladder continence, and approximately 90% return to ambulatory status. In patients older than 50, only 30% regain bladder function, with approximately 50% regaining the ability to ambulate.11
Brown-Séquard Syndrome Brown-Séquard syndrome, first described in 1846 by the one physician for whom it is named,12 is the result of an anatomic or functional hemisection of the spinal cord. Usually associated with penetrating injuries,13 Brown-Séquard syndrome also may be seen with compressive or intrinsic lesions. The syndrome has been reported in association with spinal cord tumors, spinal epidural hematoma, vascular malformations, cervical spondylosis, degenerative disk disease, and radiation injury and as a
1391
Table 104-1 Spinal Cord Syndromes SENSORY
MOTOR
SPHINCTER INVOLVEMENT
Central cord syndrome
Variable
Upper extremity weakness, distal > proximal
Variable
Brown-Séquard syndrome
Ipsilateral position and vibration sense loss Contralateral pain and temperature sensation loss
Motor loss ipsilateral to cord lesion
Variable
Anterior cord syndrome
Loss of pin and touch sensation Vibration, position sense preserved
Motor loss or weakness below cord level
Variable
Transverse cord syndrome—complete
Loss of sensation below level of cord injury
Loss of voluntary motor function below cord level
Sphincter control lost
Conus medullaris syndrome
Saddle anesthesia may be present, or sensory loss may range from patchy to complete transverse pattern Saddle anesthesia may be present, or sensory loss may range from patchy to complete transverse pattern
Weakness may be of upper motor neuron type
Sphincter control impaired
Weakness may be of lower motor neuron type
Sphincter control impaired
Cauda equina syndrome
complication of spinal instrumentation.13 The syndrome in its pure form is characterized by ipsilateral loss of motor function and proprioception or vibration, with contralateral loss of pain and temperature sensation, below the spinal cord level of injury. Because fibers associated with the lateral spinothalamic tract ascend or descend one or two spinal cord segments before crossing to the contralateral side, ipsilateral anesthesia (pain and temperature modalities) may be noted one or two segments above the lesion, although this observation is variable. Most patients with Brown-Séquard syndrome incur only partial sensory and motor impairment, and the classic pattern is not seen.11,13,14 Brown-Séquard syndrome carries the best prognosis of any of the incomplete spinal cord syndromes. Fully 80 to 90% of patients with Brown-Séquard syndrome regain bowel
and bladder function, 75% regain ambulatory status, and 70% become independent in their activities of daily living.11
Anterior Cord Syndrome Anterior cord syndrome is characterized by loss of motor function, pinprick, and light touch below the level of the lesion, with preservation of posterior column function, including some touch, position, and vibratory sensation. Although most reported cases of anterior spinal syndrome follow aortic surgery,15 the syndrome also may occur after severe hypotension, infection, myocardial infarction, vasospasm from drug reaction, and aortic angiography.16 The anatomic lesion may be the result of a cervical hyperflexion injury resulting in a
Chapter 104 / Spinal Cord Disorders
SYNDROME
PART III ■ Medicine and Surgery / Section Seven • Neurology
1392
cord contusion or from protrusion of bony fragments or herniated cervical disk material into the spinal canal. Rarely, it is produced by laceration or thrombosis of the anterior spinal artery or a major radicular feeding vessel.11 Patients present with the characteristic neurologic findings noted earlier. Functional recovery varies; most improvement occurs over the first 24 hours, but little improvement is expected thereafter.4 Although anterior cord lesions from ischemia usually are incomplete, patients without motor function at 30 days have little or no likelihood of regaining any motor function by 1 year.15,17 Overall, only 10 to 20% of patients with this entity regain some muscle function, and even in this group, there is little power or coordination.11
Conus Medullaris Syndrome and Cauda Equina Syndromes The separation of conus medullaris and cauda equina lesions in clinical practice is difficult because the clinical features of the disorders overlap. Additionally, a combined lesion may occur that masks clear clinical symptoms or signs of either an upper or a lower motor neuron type of injury. The conus medullaris is the terminal end of the spinal cord, located at approximately the L1 level in adults. The conus medullaris syndrome may involve disturbances of urination (usually from a denervated, autonomic bladder that manifests clinically with overflow incontinence) and sphincter impairment or sexual dysfunction. Sensory involvement may affect the sacral and coccygeal segments, resulting in saddle anesthesia. Pure lesions of the conus medullaris are rare.18 Upper motor neuron signs, such as increased motor tone and abnormal reflexes, may be present, but their absence does not exclude the syndrome. The conus medullaris syndrome can be caused by central disk herniation, neoplasm, trauma, or vascular insufficiency. Because the conus is such a small structure, with lumbar and sacral segments represented in a small area, a lesion usually causes bilateral symptoms. This finding may help distinguish lesions of the conus from lesions of the cauda equina, which often are unilateral.18 The cauda equina (Latin for “horse’s tail”) is the name given to the lumbar and sacral nerve roots that continue on within the dural sac caudal to the conus medullaris. Not a true “cord syndrome,” cauda equina syndrome represents dysfunction at the level of nerve roots, but the anatomic clustering of nerve roots with the lumbar dural sac allows injury to several nerve roots to occur simultaneously. The etiologic lesion in the cauda equina syndrome usually is a midline rupture of an intervertebral disk, most commonly at the L4-5 level. Tumors and other compressive masses also may cause the syndrome. As in the conus medullaris syndrome, patients generally present with progressive symptoms of fecal or urinary incontinence, impotence, distal motor weakness, and sensory loss in a saddle distribution. Muscle stretch reflexes also may be reduced. The presence of urinary retention is the most consistent finding, with a sensitivity of 90%.19 Low back pain may or may not be present.
■ CLINICAL FEATURES History Weakness, sensory abnormalities, and autonomic dysfunction are the cardinal manifestations of spinal cord dysfunction. The tempo and degree of impairment often reflect the disease process. Past medical history is vital because a history of coagulopathy or other systemic processes may be elicited. A history of cancer should suggest the possibility of metastatic disease.
Recent trauma raises the possibility of vertebral fracture or disk protrusion.
Physical Examination The physical examination pertinent to spinal cord dysfunction involves testing in three areas: (1) motor function, (2) sensory function, and (3) reflexes. Each component is best tested with the anatomic organization of the spinal cord in mind to help determine the level of the spinal cord dysfunction.
Motor Function Testing of motor function encompasses examination of muscle bulk, tone, and strength. Muscle bulk is easily examined in large motor groups, such as the thigh or calf muscles, the biceps, or the triceps. Inspection of the intrinsic hand muscles also may be helpful for determining muscle bulk; wasting may be evident as hollowed or recessed regions of the hand. Decreased mass, asymmetry, or fasciculations should be noted. Tone is tested with repeated passive knee, elbow, or wrist flexion, with the examiner feeling for abnormally increased or decreased resistance. Rapid pronation-supination of the forearm is another useful method to assess tone. Increased tone may indicate spasticity or an upper motor neuron lesion, whereas decreased tone corresponds with lower motor neuron, motor endplate, or muscular problems. Finally, motor strength is graded in the upper and the lower extremities. Motor grading for the neurologic examination is relatively straightforward. Scored on a scale of 0 to 5, neuromuscular functioning is graded as follows: 0: No firing of the muscle is present. 1: The muscle fires but is unable to move the intended part. 2: The muscle is able to move the intended part with gravity eliminated. 3: The muscle is able to move the intended part against gravity. 4: The muscle is able to move the intended part, but not at full strength. 5: Full muscular strength is present. A rectal examination is performed to assess voluntary sphincter contraction, resting tone, and, as described previously, the bulbocavernosus reflex.
Sensory Function Sensory testing requires a cooperative patient and an attentive examiner. The spinal cord–related modalities that may be clinically useful in the ED setting include testing for pinprick, light touch (contralateral lateral spinothalamic tract), and proprioception (ipsilateral posterior column). Assessment of the patient’s response to pinprick, light touch, and proprioception in all four extremities is necessary if a neurologic injury is suspected. Testing of sacral dermatomes may be an important part of the examination in some patients. As previously noted, sacral sparing is an important finding indicating that spinal cord dysfunction may be incomplete. The sensory fibers from sacral dermatomes are more peripherally located in the ascending fiber bundles; central or partial cord lesions may ablate sensation in the extremities yet allow some perception of sensation in the sacral area.
Reflexes Muscle stretch (“deep tendon”) reflexes may be tested rapidly at the bedside. Responses are graded on a scale of 0 to 4+, with
■ DIAGNOSTIC STRATEGIES Historical or physical examination findings that suggest spinal cord dysfunction prompt further investigations. The basic strategy is to detect or exclude extrinsic compressive lesions or other potentially treatable entities. Magnetic resonance imaging (MRI) has changed the diagnostic approach to patients with suspected spinal cord dysfunction. Plain radiographs and computed tomography (CT) scans may show bone and some soft tissue abnormalities. Conventional radiographs and CT scans are required in patients with trauma or suspected bone involvement by tumor or degenerative processes, but MRI shows many of these abnormalities and defines the spinal cord as well as the soft tissue structures associated with it. Tissue damage patterns within the cord, such as hemorrhage and edema, also may be detected with MRI. CT myelography may be able to answer some of these questions in patients in whom implanted metal precludes MRI but generally does not yield the same level of detail. After imaging studies exclude compressive lesions or other masses affecting the spinal cord, the possibility of inflammatory or demyelinating disorders remains, and lumbar puncture may be useful in diagnosis.
■ DIFFERENTIAL CONSIDERATIONS The prime principle in management of spinal cord dysfunction is to consider and exclude potentially treatable problems. The clinical assessment of spinal cord dysfunction is limited to detecting weakness, sensory alterations, sphincter dysfunction, and perhaps reflex abnormalities. Pain in the back may be present depending on the pathologic process but generally is not helpful in formulating a list of considerations for the
1393
differential diagnosis. Because potential functional loss and impact on quality of life are great, the detection of a process for which some intervention is possible assumes great importance. A likely diagnosis of spinal cord infarction may be entertained, but the pursuit of a treatable process, such as spinal cord compression from an epidural hematoma, should be seriously considered.21 This discovery process may involve specialty consultation or obtaining studies not readily available in many ED settings, such as MRI. As a general rule, liberal use of consultation and imaging is recommended when the possibility of spinal cord dysfunction is considered. The history may suggest a specific cause and will guide the tempo of investigation. The caveat is that spinal cord diseases may mimic many other disease processes, and neither the history nor physical examination may allow diagnosis until appreciable neurologic dysfunction has developed. The picture of a complete transverse spinal cord syndrome with paraplegia, sensory loss at a clear anatomic level, and sphincter dysfunction cannot be fully simulated by other anatomic lesions. Incomplete or evolving spinal cord syndromes may be imitated by other disease processes. It is always prudent to focus the differential diagnosis on anatomic considerations—the classic “where is the lesion?” approach (Table 104-2). Progressive lower extremity weakness and sensory alteration may represent cord dysfunction but could reflect an intracranial vertex mass with bilateral cortical dysfunction. Ataxia may be a finding in cerebellar disease but also has rarely been reported as an isolated finding with spinal cord compression. Another example is rapidly progressive paralysis in a patient with areflexia and quadriplegia; ascending paralysis (Landry-Guillain-Barré syndrome) at times may mimic an acute cord lesion. Generally, pathologic processes involving the spinal cord may be divided into processes affecting the cord or its blood supply primarily, such as demyelination, infection, or infarction, and processes that compress the cord, most often originating outside the dura (Box 104-1). Myelitis is a comprehensive term for spinal cord inflammation with dysfunction, and the potential causes are legion. The clinical presentation often is similar across the variety of entities that may cause cord compression. The tempo of the process may yield a different clinical picture. In chronic compression, muscle wasting and abnormal reflexes may be present, whereas both of these may be lacking in acute compression. A neurologic deficit in concert with back pain strongly suggests a spinal cord lesion, necessitating prompt investigation to identify a specific cause. Atypical presentations for these lesions are the rule, and additional diagnostic studies should be pursued as appropriate.
Table 104-2 Clinical Characteristics of Neuromuscular Diseases HISTORY
STRENGTH
DTR
SENSATION
WASTING
Myelopathy
Trauma, infection, cancer
Normal to decreased
Increased
No
Motor neuron disease (ALS) Neuropathy
Progressive difficulty with swallowing, speaking, walking Recent infection Ascending weakness Food (canned goods) Tick exposure Easy fatigability Thyroid disease Previous similar episodes
Decreased
Increased
Normal to decreased Normal
Normal or decreased Distal > proximal Normal to fatigue
Decreased
Decreased
Yes
Normal
Normal
No
Decreased Proximal > distal
Normal
Normal
Yes
Neuromuscular junction disease Myopathy
ALS, amyotrophic lateral sclerosis; DTR, deep tendon reflex.
Yes
Chapter 104 / Spinal Cord Disorders
2 being normal. Hyperactive reflexes suggest upper motor neuron disease (affecting the neurons or their outflow from the brain or spinal cord), as do sustained clonus and a Babinski’s sign. The absence of these reflex changes does not constitute evidence that a myelopathy is not present. In fact, one small series noted a low incidence of extensor planar responses, as well as a lack of hyper-reflexia, in patients presenting to the ED with acute or progressive cord compression or myelopathies.20 Reflexes also may be diminished or absent when sensation is lost, or when spinal shock is present, or when lower motor neuron disease is present. Diseases of muscles or neuromuscular junctions also may decrease reflexes. In acute cord injury, reflexes may be diminished in the acute phase. The bulbocavernosus reflex may be helpful in this assessment.
1394
PART III ■ Medicine and Surgery / Section Seven • Neurology
Nontraumatic Etiologies of Spinal
BOX 104-1 Cord Dysfunction
Processes Affecting the Spinal Cord or Blood Supply Directly Multiple sclerosis Transverse myelitis Spinal arteriovenous malformation/subarachnoid hemorrhage Syringomyelia HIV myelopathy Other myelopathies Spinal cord infarction Compressive Lesions Affecting the Spinal Cord Spinal epidural abscess Spinal epidural hematoma Diskitis Neoplasm Metastatic Primary CNS HIV, human immunodeficiency virus; CNS, central nervous system.
■ MANAGEMENT Just as the clinical manifestations of spinal cord dysfunction are nonspecific with respect to etiology, the treatment for many of the disease entities often is nonspecific. Steroid administration has been recommended as therapy in spinal cord trauma, although this use of steroids has been seriously questioned in the literature.22-24 Steroids also have been used with many nontraumatic causes of cord compression, despite the lack of rigorous clinical studies supporting this use. Radiation treatment is recommended for cord compression by tumor. Surgical consultation for decompression may be considered, although the indications for surgery and timing of surgery are controversial. A specific diagnosis is needed to guide therapy. Accordingly, involvement of appropriate consultants and discussion of what may be understudied therapies are suggested.
■ SPECIFIC DISEASE PROCESSES As noted earlier, spinal cord disorders may be grouped into lesions resulting from processes intrinsic to the cord and vasculature and lesions causing extrinsic compression. The order of the following discussion roughly corresponds with the organization of Box 104-1.
Intrinsic Cord Lesions Multiple Sclerosis Principles of Disease. Demyelination denotes a disease process with the prominent feature of partial or complete loss of the myelin surrounding the axons of the central nervous system. Multiple sclerosis (MS) is the most common example of such a process; spinal cord involvement may dominate the clinical picture. Clinical Features. Central nervous system lesions that are “scattered in time and space” are the hallmark of MS. The demyelinated segments do not transmit action potentials normally, resulting in a wide variety of spinal cord–related abnormalities, depending on the location and extent of the demyelination.
In addition to patchy motor and sensory deficits, patients with MS may complain of bladder dysfunction or tremor or demonstrate evidence of a transverse partial or complete cord syndrome mimicking a compressive spinal lesion.25,26 The history may include a previous episode of optic neuritis or transient visual problems. Spinal cord lesions in MS primarily involve the lateral corticospinal tracts, the posterior columns, and the lateral spinothalamic tracts. Motor system dysfunction is the most frequent manifestation of MS involvement of the spinal cord, usually as a result of lesions in the lateral corticospinal tracts. The examination of patients with MS often reveals paresis, increased muscle tone, hyper-reflexia, clonus, and a Babinski’s response. Spinal cord involvement also may result in dysautonomias. Diagnostic Strategies. Spinal MRI is the diagnostic imaging modality of choice because it can exclude cord compression and show lesions suggesting MS.26-28 Cranial MRI may be helpful in showing other central nervous system lesions. Cerebrospinal fluid (CSF) testing for myelin basic protein and oligoclonal bands also is a diagnostic option, but no CSF abnormalities are entirely specific for MS.29,30 Oligoclonal bands in the CSF may aid in the diagnosis, but they are significant only if not present in the serum as well.29 Differential Considerations. Considerations in the differential diagnosis include systemic lupus erythematosus, Lyme disease, neurosyphilis, human immunodeficiency virus (HIV) myelopathy, and other disorders. Management. MS exacerbations may be treated with highdose methylprednisolone followed by a tapering dose of prednisone. Corticosteroids have been shown to be useful in shortening the time required for recovery from an exacerbation of MS.26 Consultation and referral to a neurologist usually are indicated. Immunosuppressive therapy in patients with the chronic progressive form of the disease has met with variable success.25,26 Because numerous disorders can mimic MS, the definitive diagnosis of this disease usually is not made in the ED.31
Transverse Myelitis Principles of Disease. Acute transverse myelitis refers to acute or subacute spinal cord dysfunction characterized by paraplegia, a transverse level of sensory impairment, and sphincter disturbance. It is relatively rare, with a reported annual incidence of 1 case per 1.3 million population. The presentation may be mimicked by compressive lesions, trauma, infection, or malignant infiltration. The exact pathogenesis is unknown, although it is noted to follow viral infection in approximately 30% of patients and commonly is termed “postinfectious myelitis.”32 Other postulated etiologic categories include infectious, autoimmune, and idiopathic.33,34 No apparent cause for acute transverse myelitis is identified in 30% of the patients.32 Progression of symptoms usually is rapid, with 66% of the cases reaching maximal deficit by 24 hours.35 Symptoms may progress, however, over days to weeks. The thoracic cord region is affected most often by this process (60 to 70%)35; the cervical spinal cord is rarely affected.36 Clinical Features. In addition to motor, sensory, and urinary disturbances, patients with acute transverse myelitis may complain of back pain and may have low-grade fever, raising concern for spinal epidural abscess. As with MS, the examination may reveal weakness progressing to paresis, hypertonia, hyper-reflexia, clonus, and a Babinski’s response. Spinal cord involvement also can result in dysautonomias. Diagnostic Strategies. Evaluation for acute transverse myelitis is done primarily with emergent MRI to exclude compressive
Spinal Subarachnoid Hemorrhage Principles of Disease. Intraspinal hemorrhage is rare and occurs in the same anatomic locations as intracranial hemorrhages: epidural, subdural, subarachnoid, and intramedullary hemorrhages are all possible.5 Spinal subarachnoid hemorrhage usually is caused by an arteriovenous malformation.41 Hemorrhage from tumors and cavernous angiomas and spontaneous hemorrhage secondary to anticoagulation therapy also have been reported.42,43 Bleeding may occur exclusively in the subarachnoid space or within the substance of the spinal cord itself. Clinical Features. Patients present with excruciating back pain, of paroxysmal onset, at the level of the hemorrhage. This pain also may be in a radicular distribution or extend into the flank. Patients may complain of headache and exhibit cervical rigidity if the blood migrates into the intracranial subarachnoid space, simulating an intracranial subarachnoid hemorrhage. Variable neurologic deficits are present, depending on the magnitude and anatomic location of the hemorrhage. These deficits typically include extremity numbness, weakness, and sphincter dysfunction.42,44 Nuchal rigidity or signs of meningeal irritation also may be present. Diagnostic Strategies. The diagnostic study of choice is MRI. Lumbar puncture also can confirm the presence of blood in the CSF. Differential Considerations. Considerations in the differential diagnosis include epidural abscess, tumor, transverse myelitis, ischemia from an aortic catastrophe such as dissection, and anterior spinal artery thrombosis. Management. Treatment depends on the etiology of the hemorrhage. Neurosurgical referral is obtained for further evaluation and for clot evacuation if compression is present. Angiography may be recommended if arteriovenous malformation is suspected.
Syringomyelia Principles of Disease. Syringomyelia is the presence of a cavitary lesion within the substance of the spinal cord. A syrinx usually is a chronic progressive lesion, and its location within the cord determines the constellation of neurologic findings on examination. Clinical Features. Headache and neck pain are the most common complaints, followed by sensory disturbance, gait disorder, and lower cranial nerve dysfunction.45 The classic pattern of sensory
1395
deficit involves a loss of pain and temperature sensation in the upper extremities, with preservation of proprioception and light touch. This phenomenon is described as a “disassociative anesthesia” because of the discrepant loss of sensory modalities. The sensory deficit often is described as being in a “capelike” distribution over the shoulders and arms. The anatomic basis for the neurologic features of a syrinx is its location near the central canal. Here it may compress the crossing fibers of the lateral spinothalamic tract that carries pain and temperature fibers. Crude touch, position, and vibratory sensation typically are unaffected. Sensory fibers from the lower limbs are similarly spared. The symptoms of syringomyelia develop and progress in accordance with the intracavitary pressure and location of the syrinx. The most common features on physical examination are lower limb hyper-reflexia, weakness and wasting in the hands and arms, dissociated sensory loss, and gait disorder. Symptoms may be exacerbated by a sneeze, cough, or Valsalva maneuver.46 Ninety percent of patients in whom this process develops have Arnold-Chiari I malformation (projection of cerebellar tonsils and medulla into the spinal canal).47 Syringomyelia also may result from spinal cord trauma (often months to years later) or compressive tumors, or may be a sequela of meningitis.48 Diagnostic Strategies. Syrinx is best seen on MRI. No other study currently in widespread use is equal to MRI in diagnostic ability. Differential Considerations. Considerations in the differential diagnosis for syrinx include intrinsic spinal tumor and demyelination. Management. When the diagnosis of syringomyelia is considered, emergent imaging in the ED is not necessary if follow-up evaluation can be arranged, because this condition usually is a slowly progressive process. In patients for whom MRI studies are obtained and the diagnosis is made, referral to a neurosurgeon is indicated, because symptoms progress in approximately two thirds of patients.49
Idiopathic Spastic Paraparesis Idiopathic spastic paraparesis is a progressive disorder characterized by progressive weakness and signs of spasticity of the lower extremities. This condition sometimes also is referred to as primary lateral sclerosis, which describes the demyelination pattern in the lateral column of the spinal cord. This disorder typically occurs in older men. Sometimes a heritable form may be discovered. It is a diagnosis of exclusion.50-52
Human Immunodeficiency Virus Myelopathy HIV myelopathy typically occurs in patients with advanced HIV disease. Weakness, gait disturbance, sphincter dysfunction, sensory abnormalities, and signs of spasticity are features of this progressive process. This is a diagnosis of exclusion, because disorders such as toxoplasmosis, lymphoma, varicella zoster, and cytomegalovirus infection may produce a similar clinical picture in immunocompromised patients. Pathologically, vacuolization of myelin sheaths in the cord may be found. Treatment is directed at the retroviral infection, although there is no proven treatment.53,54
Spinal Cord Infarction Spinal cord infarction is another diagnosis of exclusion. Aortic dissection, surgery, and global ischemia are the more common causes, although this disorder may occur as a complication of systemic lupus erythematosus or may be cryptogenic. An
Chapter 104 / Spinal Cord Disorders
lesions. Results of CSF studies are normal in 40% of the cases, with only mildly elevated protein or pleocytosis in the remaining 60%.37 The most essential aspect of the evaluation is to eliminate a potentially treatable cause, such as spinal epidural abscess, neoplasm, or hematoma. Differential Considerations. Considerations in the differential diagnosis for transverse myelitis include MS, spinal epidural abscess, spinal neoplasm, and hematoma. Management. Treatment with steroids is of unknown benefit. Anecdotal reports of improvement after steroid administration exist,35,38 but some studies have found no benefit to their use.37 Neurologic consultation is suggested, and hospitalization usually is required. The clinical course of acute transverse myelitis varies widely, ranging from complete recovery to death from progressive neurologic compromise.34 Maximal improvement usually is obtained within 3 to 6 months.39,40 At 5-year follow-up evaluation of one series of patients with this disease, 30% had a good recovery, 25% had a fair recovery, 30% had a poor outcome, and 15% had died as a result of complications of the disease.40
PART III ■ Medicine and Surgery / Section Seven • Neurology
1396
anterior spinal cord syndrome is the most common clinical picture. Some recovery is possible, although it generally is less complete than in cerebral stroke. The site of clinical dysfunction may be distant from the site of vascular occlusion.55
Extrinsic Cord Lesions Spinal Epidural Hematoma Principles of Disease. Spinal epidural hematoma is a relatively rare condition resulting from a variety of etiologic disorders. Its incidence is 0.1 per 100,000 population.56,57 The etiology may be traumatic, as after lumbar puncture or epidural anesthesia or associated with spinal surgery.1,5,58 Spinal epidural hematoma is more likely to occur in anticoagulated or thrombocytopenic patients or in patients with liver disease or alcoholism.59 Spontaneous bleeding is rare but may arise from spinal arteriovenous malformation or vertebral hemangioma. Approximately one fourth to one third of all cases are associated with anticoagulation therapy, including low-molecular-weight heparin.60,61 Clinical Features. The patient usually presents with sudden, severe, constant back pain with a radicular component. It may be noted to follow a straining episode. The pain may be worsened by percussion over the spine and maneuvers that increase intraspinal pressure, such as coughing, sneezing, or straining.62 The pain often causes the patient to seek care before the development of neurologic signs, possibly leading to delays in diagnosis.5 Neurologic deficits follow and may progress over hours to days.57 Anticoagulant use or an intrinsic coagulation abnormality may be present. The patient usually is in significant distress from the pain. Motor and sensory findings depend entirely on the level and size of the hematoma but can include weakness, paresis, loss of bowel or bladder function, and virtually any sensory deficit. Diagnostic Strategies. MRI, as with virtually all suspected intrinsic spinal disorders, is the diagnostic study of choice.62 Differential Considerations. Considerations in the differential diagnosis include abscess, epidural neoplasm, acute disk herniation, and spinal subarachnoid hemorrhage. Management. Recovery without surgery is rare, and surgical consultation for consideration of emergent decompressive laminectomy must be considered. The overall mortality rate is low (at approximately 8%).57,62 Functional recovery is related primarily to the length of time the symptoms are present. Recovery after 72 hours of symptoms is rare63 but has been reported even without surgery.64
Spinal Epidural Abscess Principles of Disease. Spinal epidural abscess is an infectious process usually confined to the adipose tissue of the dorsal epidural space, where there is a rich venous plexus. It is an uncommon disease, with an overall frequency of 0.2 to 1.2 per 10,000 hospital admissions.65,66 Major risk factors include diabetes, intravenous drug abuse, chronic renal failure, alcoholism, and immunosuppression.66-68 Although the disease may manifest in subacute or chronic forms, the acute presentation is seen most frequently in the ED. Thoracic and lumbar sites of infection predominate, with cervical epidural abscess being much less common.69,70 Infection typically extends over four to five spinal vertebral segments.71 The dura mater limits the spread of an epidural infection, making subdural or intraspinal spread uncommon. Hematogenous spread of infection to the epidural space is the most common source (seen in 26 to 50% of cases),5,65,71 either to the epidural space or to the vertebra with
extension to the epidural space. Skin and soft tissue infections are the most frequently reported identified source (in 15%)67,71; Staphylococcus aureus is the most prevalent organism, being cultured in more than 50% of cases.65,72 Other frequently identified pathogens include aerobic and anaerobic streptococci, Escherichia coli, and Pseudomonas aeruginosa. Multiple organisms are identified in approximately 10% of cases; no organism is identified in 40%.65,72 Clinical Features. The classic clinical presentation of spinal epidural abscess begins with a backache that progresses to localized back pain often associated with tenderness to percussion. Fever, sweats, and rigors are common, being reported in 30 to 75% of patients.65,72 The classic triad of back pain, fever, and progressive neurologic deficits is present in only a few patients, however, and delayed clinical diagnosis is common.65 Radicular symptoms may not be present initially but usually develop as the disease progresses. Without treatment, myelopathic signs will develop, usually beginning with bowel and bladder disturbance. Weakness ensues, followed by paraplegia or quadriplegia. Approximately 10% of patients with spinal epidural abscess present with encephalopathy.65,71 Diagnostic Strategies. MRI is the imaging modality of choice and needs to be performed emergently if the diagnosis is entertained. Other diagnostic testing is nonspecific for spinal epidural abscess, but a complete blood count may support the diagnosis, because leukocytosis commonly is present, with an average white blood cell count of 13,000/µL to 16,000/µL.71 The erythrocyte sedimentation rate, although not specific for epidural abscess, is virtually always elevated with this condition.65,66,71 Plain films usually are normal in appearance, unless evidence of osteomyelitis of an adjacent vertebral body is seen. Lumbar puncture is relatively contraindicated with known epidural abscess but often is performed as part of the evaluation for meningitis. CSF findings are consistent with a parameningeal infection, showing elevation of protein and some cellular response. Differential Considerations. Any compressive spinal lesion, including tumor or hematoma, can mimic spinal epidural abscess. Management. Urgent surgical consultation for decompression usually is required. Antibiotics effective against the most common pathogens (particularly S. aureus) should be started empirically. One such regimen that covers gram-positive and gram-negative organisms consists of a third-generation cephalosporin plus vancomycin, both given intravenously, plus rifampin given orally. Outcome is related to the speed of diagnosis before the development of myelopathic signs. The disease is fatal in 18 to 23% of cases, and patients with neurologic deficit rarely improve if surgical intervention is delayed more than 12 to 36 hours after onset of paralysis.65,71 Patients operated on before development of neurologic symptoms have an almost universally good outcome.66,67
Diskitis Principles of Disease. Diskitis is an uncommon primary infection of the nucleus pulposus, with secondary involvement of the cartilaginous endplate and vertebral body. It may occur after surgical procedures or spontaneously, the latter being more common in pediatric patients.73,74 An increased incidence of diskitis has been noted in immunocompromised patients and in patients with systemic infections. Both an acute and a chronic disease course have been described, with the acute course being more common.73 Clinical Features. Patients present with moderate to severe pain, localized to the level of involvement and exacerbated by
Neoplasm Principles of Disease. Spinal cord tumors are classified according to their relationship to the dura and spinal cord (extradural, intradural or extramedullary, and intradural or intramedullary). Spinal cord tumors produce neurologic symptoms by compression, invasion, or destruction of myelinated tracts. The resulting neurologic symptoms are directly related to the growth rate and the location of the tumor. Spinal cord tumors account for 4 to 10% of central nervous system tumors but for only 1% of all cancers. Primary tumors occur with an incidence of 1 per 1 million population.75 Most tumors of the spinal cord are metastatic in origin, however. Approximately 10% of patients with known cancer are diagnosed with a spinal metastasis at some point in the course of their disease, and 5 to 10% of patients ultimately diagnosed with cancer first present with a spinal metastasis.2 Lung cancer, breast cancer, and lymphoma account for more than 50% of the primary malignancies that subsequently develop spinal metastasis, spreading by the hematogenous route and direct extension. Most metastases occur in the thoracic spine, and nearly 20% of patients with tumor spread to the spine will be found to have disease at multiple levels.2,76,77 Clinical Features. In 95% of patients with spinal neoplasm, the initial complaint is pain, either in the back at the level of the tumor or in a radicular distribution. Pain often is characterized
1397
as dull, constant, and aching and commonly is said to worsen with recumbency (in contrast with the pain of herniated disk).5 Nighttime pain that is severe is characteristic of spinal neoplasm.78 Any action that increases intraspinal pressure (Valsalva maneuver, sneeze, cough) may be associated with increased pain. Neurologic deficits vary, depending on the location of the lesion. Besides a thorough neurologic examination, a search for possible primary sites should be done on the physical examination. Diagnostic Strategies. Plain radiographs are usually the initial diagnostic test, and 70 to 85% of patients with spinal column involvement show some abnormality on these films.78 Patients with neurologic abnormalities and suspicious findings on plain films are candidates for emergent MRI or CT myelography. In patients with a known history of neoplasm and new back pain, some authorities recommend foregoing plain films and proceeding directly to MRI, because plain films can be misleading or nondiagnostic.79 Differential Considerations. Considerations in the differential diagnosis include any of the compressive lesions (blood, infection). Tumor also can mimic intrinsic lesions, such as transverse myelopathy and cord infarction. Management. Acute compressive myelopathy from neoplasm constitutes an oncologic emergency. Immediate treatment is required to preserve function and prevent deterioration. With onset of paraplegia and incontinence, less than 5% of patients regain ambulatory status.1,80 Of patients who are ambulatory at the time of diagnosis, 60% remain ambulatory.5 High-dose steroids, radiotherapy, and surgery all may be necessary acute interventions, and consultation with neurosurgeons, neurologists, oncologists, and therapeutic radiologists may be required.
KEY CONCEPTS ■
Spinal cord disorders may manifest in subtle fashion and with nonspecific clinical signs and symptoms. In the absence of neurologic abnormalities or complaints, diagnosis of these disorders can be extremely difficult. ■ Patients with rapid onset and progression of spinal cord symptoms should receive specialized imaging and consultation in the ED. ■ MRI frequently is required to make a definitive diagnosis for spinal syndromes. ■ With compressive lesions of the spinal cord, duration of neurologic dysfunction is directly related to ultimate neurologic outcome. The diagnosis must be made expeditiously and definitive therapy begun as soon as possible.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 104 / Spinal Cord Disorders
almost any movement of the spine. Radicular symptoms are present in 50 to 90% of cases.73,74 The lumbar spine is the most common site of disease. Elevated temperature is noted in more than 90% of patients.73 Patients experience pain with range of motion. Neurologic deficits are the exception with diskitis. Diagnostic Strategies. Plain radiographs usually are not helpful for early diagnosis, but destruction of the disk space is highly suggestive if present. The radiographic findings become positive after 2 to 4 weeks of disease. In addition to disk space narrowing, plain films may show irregular destruction of the vertebral body endplates. Often there is a latent period (2 to 8 weeks) between the onset of back pain and the development of other clinical symptoms or abnormalities on the physical examination. MRI is the radiographic study of choice because it not only can diagnose diskitis but also can rule out paravertebral or epidural abscess. Laboratory studies often show an elevated erythrocyte sedimentation rate, but the white blood cell count usually is normal.73,74 S. aureus is the most common pathogen, but gram-negative, fungal, and tuberculous infections all have been recognized. Differential Considerations. Considerations in the differential diagnosis include spinal epidural abscess, neoplasm, and hematoma. Management. With timely diagnosis and treatment, outcome generally is good, and medical treatment with intravenous antibiotics usually is curative. Surgery often is not necessary.73,74
Chapter 105
Peripheral Nerve Disorders
E. Bradshaw Bunney and E. John Gallagher
■ PERSPECTIVE Background The nervous system is traditionally divided into central nervous system (CNS) and peripheral nervous system (PNS) components. The PNS can be further subdivided into 12 cranial and 31 spinal nerves. Disorders of the cranial nerves are discussed in Chapter 103. Because diseases of the neuromuscular junction and the myopathies are located distal to the neuron itself, they are also considered separately in Chapter 104. Radiculopathies, which are disorders of the roots of the PNS, are so commonly associated with musculoskeletal neck and back pain that they are mentioned only briefly here and are discussed in detail in Chapter 51. The simplest approach to diseases of the PNS parallels the CNS model of separating focal from nonfocal disease. In the PNS, the first broad category is the focal group, which can be divided into those with evidence of single versus multiple lesions of peripheral nerves, known respectively as simple mononeuropathies and multiple mononeuropathies (or mononeuropathy multiplex). The second broad category, which constitutes the nonfocal group of peripheral neuropathies, contains the polyneuropathies. These tend to produce bilaterally symmetrical symptoms and signs, reflecting the widespread nature of the underlying pathologic process. The evaluation of PNS disease involves a goal-directed history and physical examination targeted at answering the following three questions, each of which corresponds to a stratum of the algorithm presented in Fig. 105-1: 1. Are the sensorimotor signs and symptoms symmetrical or asymmetrical? 2. Are the sensorimotor signs and symptoms distal or both proximal and distal? 3. Is the modality involved exclusively motor, sensory, or mixed sensorimotor? By systematically combining responses to these questions, one can identify seven discrete categories of peripheral neuropathy, each of which contains a finite set of possible diagnoses. Because pure motor or sensory findings tend to occur mainly in an asymmetrical, distal distribution, this is the only category in Figure 105-1 subdivided into pure motor and pure sensory abnormalities.
Epidemiology Although Guillain-Barré syndrome (GBS) is the most commonly encountered emergent peripheral neuropathy in devel1398
oped countries, its annual incidence is just over 1 to 2 cases per 100,000 population.1 In contrast to the low incidence of acute peripheral neuropathies, several of which are associated with short-term mortality, the vast majority of peripheral neuropathies seen in the emergency department (ED) are subacute or chronic and are associated not with mortality but with long-term morbidity. Current estimates suggest that about 1.5% of the U.S. population suffers from peripheral neuropathy.2 Over 7% of the population has diabetes mellitus, with a prevalence rate of 20% in individuals older than 60 years. Roughly 50% of these individuals have peripheral neuropathy.3
■ PRINCIPLES OF DISEASE Anatomy The spinal component of the PNS is shown schematically in Figure 105-2. The anterior and posterior nerve roots exit the spinal cord at each segmental level. Just distal to the dorsal root ganglion they converge to form a mixed (motor and sensory) spinal nerve, of which there are 31 pairs: 8 cervical, 12 thoracic, 5 lumbar, 5 sacral, and 1 coccygeal. The spinal nerves immediately bifurcate into anterior (ventral) and posterior (dorsal) rami. The posterior ramus travels to the back. The anterior ramus innervates the anterolateral portion of the body and supplies all peripheral nerves for the upper and lower extremities through the brachial and lumbosacral plexus, respectively. Interweaving of fibers occurs within a plexus, producing a mixed sensorimotor innervation of peripheral nerves exiting the plexus. In addition to the motor and sensory modalities of the PNS, the autonomic nervous system has a peripheral component. Anatomically and functionally, the autonomic nervous system is divided into two parts: a sympathetic (thoracolumbar) component and a parasympathetic (craniosacral) component. Autonomic dysfunction may cause systemic abnormalities, such as orthostasis, or local problems, such as atrophic, dry skin.
Pathophysiology The PNS has only three basic responses to a wide array of pathologic stimuli. As shown in Figure 105-2, these are (1) the myelinopathies, where the primary site of involvement is limited to the myelin sheath surrounding the axon; (2) the axonopathies, where the primary site of involvement is the axon, with or without secondary demyelination; and (3)
1399 SYMMETRICAL
peripheral neuropathy in the emergency department. AIDP, acute inflammatory demyelinating polyneuropathy (Guillain-Barré); CIDP, chronic inflammatory demyelinating polyneuropathy; DSPN, distal symmetrical polyneuropathy. *A proximal distribution of sensorimotor findings may dominate the clinical picture in patterns 3, 4, and 5, depending on the location of the lesion(s).
Proximal/distal
Distal
Proximal/distal
Mixed
Mixed
Mixed
1. AIDP/CIDP Primary myelinopathy
2. DSPN Primary axonopathy
3. Plexopathy/ radiculopathy*
Distal
Mixed
Pure
6. Motor neuronopathy
4. Mononeuropathy*
7. Sensory neuronopathy
5. Mononeuropathy multiplex*
Posterior cord Dorsal root
Neuromuscular junction Dorsal root ganglion
Ventral root
Peripheral nerve axon Myelin sheath
Anterior cord
CNS
Posterior ramus
Anterior ramus Mixed spinal nerve
Muscle
PNS
Figure 105-2. Schematic representation of macroscopic and microscopic anatomy of the peripheral nervous system and its interface with the central nervous system. See text for explanation.
the neuronopathies, where the cell body of the neuron itself is the primary site of involvement, ultimately affecting the entire peripheral nerve. Although overlap occurs, each of these prototypes has a distinctive clinical presentation, electrophysiologic profile, and microscopic appearance. Electrophysiologic testing, that is, nerve conduction studies (NCSs) and needle electromyography (EMG), detects underlying pathologic abnormalities. Because neither test is readily available in the acute care setting, they are discussed only briefly here. Information gathered from NCSs and EMG can be used to obtain objective information on the anatomic distribution of involvement (symmetrical vs. asymmetrical and distal vs. proximal and distal) and the modalities involved
(sensory, motor, or mixed). NCSs and EMG can also identify the level of the neuraxis affected by the disease process (i.e., root, plexus, or nerve); if the nerve is affected, electrophysiologic testing can help determine whether the lesion is mononeuropathic (either caused by an isolated mononeuropathy or mononeuropathy multiplex) or polyneuropathic. Finally, EMG and NCSs can distinguish axonal from myelinopathic disease, further narrowing the differential diagnosis. Prognosis is determined by the nature of pathologic involvement of the PNS. Primary demyelination spares the axon and thus carries the best prognosis. The prognosis is worse in axonopathies because reestablishing nerve function is dependent on the much slower process of axonal regeneration. Neuronopathies, which begin
Chapter 105 / Peripheral Nerve Disorders
Figure 105-1. An approach to
ASYMMETRICAL
1400
PART III ■ Medicine and Surgery / Section Seven • Neurology
Causes of Acute, Emergent Weakness and
BOX 105-1 Possible Respiratory Compromise
Autoimmune Demyelinating Guillain-Barré syndrome Chronic inflammatory demyelinating polyneuropathy (CIDP) Myasthenia gravis Toxic Botulism Buckthorn Seafood Paralytic shellfish toxin Tetrodotoxin (puffer fish, newts) Tick paralysis Metals Arsenic Thallium Metabolic Dyskalemic syndromes Acquired (especially with thyrotoxicosis) Familial Hypophosphatemia Hypermagnesemia Porphyria Infectious Poliomyelitis Diphtheria
and Prototypes of Table 105-1 Patterns Peripheral Neuropathies TYPE
PATTERN DISTRIBUTION
1
Proximal and distal, symmetrical, sensorimotor polyneuropathy Proximal and distal Motor > sensory Distal, symmetrical, sensorimotor polyneuropathy Distal Sensory > motor Proximal and distal, asymmetrical, sensorimotor neuropathy Proximal and distal Sensory and motor Distal, asymmetrical, sensorimotor mononeuropathy Distal Sensory and motor Distal, asymmetrical, sensorimotor mononeuropathy multiplex Distal Sensory and motor Distal, asymmetrical, pure motor neuronopathy Distal Motor Distal, asymmetrical, pure sensory neuronopathy Distal Sensory
2
3
4
5
6
with primary destruction of the nerve cell body, produce pure motor or pure sensory syndromes. Eventually the entire nerve is affected, resulting in the worst prognosis of the three.
■ CLINICAL FEATURES The differential diagnosis for any patient presenting with sensory, motor, or sensorimotor complaints, particularly if localized to the extremities, should include a peripheral neuropathy. Within this group, patients with focal weakness are most concerning because they are at greatest risk for respiratory compromise. Box 105-1 lists the causes of acute, emergent weakness that may affect respiration. Although several of the disorders listed are myopathies (see Chapter 106) rather than peripheral neuropathies, they are lumped together because it is important to identify patients at risk for respiratory failure early in the course of evaluation. As soon as the emergent causes of weakness have been excluded—which is possible in the majority of patients—the individuals with focal weakness should be assessed next to exclude CNS disease (e.g., stroke) (see Chapter 99). One can then proceed through the systematic approach to peripheral neuropathy outlined in Figure 105-1. Another way to look at the algorithm displayed in Figure 105-1 is shown in Table 105-1, with the distinguishing features of each of the seven peripheral neuropathic patterns described by distribution and modality and represented by a disease prototype.
Type 1: Demyelinating Polyneuropathies The pattern of symmetrical weakness, usually worse distally, accompanied by variable sensory findings is characteristic of acute GBS. This pattern is discussed first because it is the most common cause of weakness associated with acute respiratory failure seen in emergency practice.
7
PROTOTYPICAL DISEASE MODALITIES
GBS Symmetrical Diabetic DSPN Symmetrical Brachialplexopathy Asymmetrical CTS (median mononeuropathy) Asymmetrical Vasculitic mononeuropathy multiplex Asymmetrical ALS Asymmetrical Pyridoxine toxicity Asymmetrical
ALS, amyotrophic lateral sclerosis; CTS, carpal tunnel syndrome; DSPN, distal symmetrical polyneuropathy; GBS, Guillain-Barré syndrome.
Guillain-Barré Syndrome GBS is a heterogeneous and unpredictable disorder, with marked variation in latency between antecedent infection and symptom onset. The clinical signs, cadence of disease progression, degree of respiratory compromise, laboratory findings, and time required for convalescence are also highly variable. The most common form of GBS is an acute inflammatory demyelinating polyneuropathy (AIDP), comprising 90% of the cases seen in the United States.1 Less common variants include acute motor axonal neuropathy (AMAN), acute motor and sensory axonal neuropathy (AMSAN), and the Miller Fisher syndrome. AMAN accounts for most of the remaining cases seen in the United States afflicting those of Asian decent. Miller Fisher syndrome is a rare form of GBS characterized by the triad of ophthalmoplegia, ataxia, and areflexia (Box 105-2).1,4 The majority of patients seek treatment days to weeks after resolution of an upper respiratory or gastrointestinal illness, presenting with progressive, symmetrical distal (and usually to a lesser extent proximal) weakness. Signs and symptoms are usually worse in the lower extremities and are associated with diminution or loss of deep tendon reflexes (DTRs), variable sensory findings, and sparing of the anal sphincter. Up to 32% will have all four extremities affected at the time of presentation and 10% will have weakness that begins in the upper
1401
BOX 105-2 Demyelinating Polyneuropathies
extremities.1 However, the ocular muscles are usually spared. Urinary retention secondary to autonomic dysfunction may occur, contributing to a clinical picture easily mistaken for a spinal cord lesion or conus medullaris syndrome. The most commonly infectious organisms associated with GBS are Campylobacter jejuni (in patients with a history of diarrhea), cytomegalovirus, Epstein-Barr virus, and Mycoplasma pneumoniae. AIDP is caused in part by macrophage invasion of the myelin sheath. The macrophage is believed to detect antigens in the myelin that are nearly identical to the antigens present on certain infectious organisms.1,5,6 In practice, patients with symmetrical weakness of relatively acute onset, decreased or absent DTRs, and variable degrees of sensory loss should be managed as if they have GBS or one of its variants, which places them at risk for respiratory compromise. Conversely, patients with predominantly sensory signs and symptoms are less likely to develop acute respiratory distress and have a more favorable prognosis.7 About half of patients with GBS have autonomic dysfunction, experience a peak of disease severity within a week of onset, have some form of cranial nerve involvement (usually VII), and suffer long-term sequelae of their illness. Nearly one third require ventilatory support. Both the mortality and the recurrence rate are about 3%.8 In addition to electrophysiologic testing, there are three ancillary tests that may be helpful in the diagnosis of GBS. Cerebrospinal fluid (CSF) analysis is useful when it demonstrates the characteristic picture of markedly elevated protein with only a mild pleocytosis. In the clinical setting of suspected GBS, this finding is highly specific. Early in the disease, however, patients may have normal CSF values. Consequently, a normal CSF cannot be used to exclude GBS because of the limited sensitivity of this test. Selective enhancement of the anterior spinal nerve roots on magnetic resonance imaging (MRI) is suggestive, but not diagnostic, of GBS.9 The GBS disability score, which combines age, presence or absence of diarrhea, and a score of the patient’s ability to ambulate independently at 2 weeks, has been shown to be predictive of prognosis at 6 months, particularly related to independent activity.10 Management. Individuals with suspected GBS must have their respiratory function tested. A decrease in forced vital capacity (FVC) has been shown to correlate with the need for intubation in patients with GBS. An FVC of less than 20 mL/kg was associated with pending respiratory failure and the need for intubation, whereas patients with an FVC of greater than 40 mL/kg did not require intubation.11,12 Likewise, patients with a negative inspiratory force (NIF) of less than 30 cm water are more likely to require mechanical ventilation.12 Other tests, such as the forced expiratory volume in 1 second
Type 2: Distal Symmetrical Polyneuropathies Most polyneuropathies are characterized by a pattern of distal, symmetrical sensorimotor findings, worse in the lower than upper extremities, with a stocking-glove distribution of sensory abnormalities that gradually diminishes as one moves proximally. The motor findings and loss of DTRs, which lag behind the sensory features, follow a similar pattern of progression from distal to proximal. The diffuse, distal, symmetrical nature of this pattern is most consistent with a toxic-metabolic disease process, as yet unidentified, that causes a length-dependent axonopathy. Distal symmetrical polyneuropathy (DSPN) is the most common type of peripheral neuropathy seen in emergency practice. Only the most common causes of DSPN are discussed, with a more complete listing of causes shown in Box 105-3.
Diabetic Distal Symmetrical Polyneuropathy The preponderance of cases of DSPN occur in diabetics, also termed diabetic polyneuropathy. Initial symptoms usually consist of “positive” sensory complaints (e.g., dysesthesias such as tingling or burning) beginning on the plantar surfaces of both feet. At the early stages of a typical DSPN, there may be some asymmetry.20 At this juncture, it may be impossible to distinguish a focal neuropathic process such as a mononeuropathy from a polyneuropathy, although in this location, prior probability strongly favors a polyneuropathy. As the process advances, the plantar surfaces of both feet become dysesthetic before the dorsum of either foot is involved. Weakness of dorsiflexion of the big toe is usually the first motor sign, followed by weakness of foot dorsiflexion, footdrop, loss of Achilles’ reflex, and later a “steppage gait.”
Chapter 105 / Peripheral Nerve Disorders
Guillain-Barré syndrome Acute inflammatory demyelinating polyradiculoneuropathy (AIDP) Acute motor axonal neuropathy (AMAN) Acute motor and sensory axonal neuropathy (AMSAN) Miller Fisher syndrome Chronic inflammatory demyelinating polyradiculoplexo-neuropathy Malignancy HIV Hepatitis B Buckthorn Diphtheria
(FEV1) or peak flow rate (PFR) can also be used to assess respiratory function. Patients unable to perform these tests and those with less than 100% of predicted values should have an arterial blood gas obtained. Evidence of alveolar hypoventilation (elevated carbon dioxide [Pco2]) in a patient with an unsecured airway requires a level of intensive monitoring that is impractical in many emergency departments. Therefore, patients with weakness, CO2 retention, or other evidence of early ventilatory failure should be considered for early, prophylactic intubation.13 Among patients with possible GBS who have normal pulmonary function, extensor neck strength can be monitored to predict impending ventilatory failure. Patients with probable GBS should receive neurologic consultation and be admitted to the hospital. Either plasma exchange or intravenous immunoglobulin (IVIG) should be administered. There is sound evidence that both are superior to placebo and that combination or sequential therapy confers no therapeutic advantage over either intervention alone. Plasma exchange is cumbersome and not available at many hospitals. IVIG is more readily available and is usually administered in a dose of 400 mg/kg/ day for 5 days. However, IVIG is quite expensive, costing roughly $50 to $80 per gram.14 Although not approved by the Food and Drug Administration, IVIG is supported in certain national guidelines.15 Corticosteroids are no longer recommended for treatment of GBS.16 Oral steroids have been shown to delay recovery. Intravenous steroids alone have been shown to impart no benefit, and though the combination of intravenous steroids and IVIG has hastened recovery, there was no affect on long-term outcome.17–19 The marked elevation in blood pressure seen in some patients with GBS should not be treated because it is typically transient and may be followed by precipitous and unpredictable hypotension.
1402
PART III ■ Medicine and Surgery / Section Seven • Neurology
BOX 105-3 Distal Sensorimotor Polyneuropathies Diabetes mellitus Alcoholism Neoplastic or paraneoplastic Hereditary motor and sensory neuropathies (Charcot-Marie-Tooth) Cryptogenic sensorimotor polyneuropathies (CSPN) HIV Toxins Organic or industrial agents Acrylamide Allyl chloride Carbon disulfide Ethylene oxide Hexacarbons Methyl bromide Organophosphate-induced delayed polyneuropathy (OPIDP) Polychlorinated biphenyls (PCBs) Trichloroethylene Vacor Metals Arsenic Gold Mercury (inorganic) Thallium Therapeutic agents Amiodarone Antiretrovirals
Dapsone Disulfiram Isoniazid Metronidazole Nitrofurantoin Paclitaxel (Taxol) Phenytoin Statins (HMG-CoA reductase inhibitors) Thalidomide Vinca alkaloids (vincristine, vinblastin) Nutritional Beriberi (thiamine or vitamin B1) Pellagra (niacin, B vitamins) Pernicious anemia (vitamin B12) Pyridoxine deficiency (vitamin B6) End-organ dysfunction Acromegaly Chronic pulmonary disease Hypothyroidism Renal failure (uremic neuropathy) Paraproteinemias Amyloidosis Monoclonal gammopathy of unknown significance (MGUS) Multiple myeloma Waldenström’s macroglobulinemia Porphyria
HMG-CoA, hydroxymethylglutaryl coenzyme A.
Sensory loss continues to move proximally, and before it reaches the knees, the fingertips are usually involved. DTRs are progressively lost, as is proprioception. If the latter becomes severe, patients may develop sensory ataxia. As the neurop athy continues to progress, sensory abnormalities ultimately involve all modalities and extend to a diamond-shaped periumbilical area. Far advanced disease may affect sensation over the skull vertex and facial midline structures. Atrophy and areflexia occur as weakness worsens. Severely impaired patients may be unable to ambulate or grasp objects. These symptoms have a significant impact on the patient’s quality of life, affecting not only physical functioning but emotional, sleep, and social functioning. Many of these patients display signs of depression or anxiety.21 Polyneuropathies can be difficult to diagnose and are best approached by performing electrodiagnostic studies on patients with a constellation of symptoms and signs suggesting a particular neuropathy.22 Management. As with virtually all peripheral neuropathies, referral is indicated for management of diabetic DSPNs. If discomfort is severe, the etiology of the neuropathy seems likely to be diabetic, and if referral is delayed, it may be necessary to provide the patient with some symptomatic relief. Because treatment of neuropathic pain has traditionally been linked to etiology rather than an underlying mechanism, the choice of pharmacologic agents is empirical, with substantial practice variation in the United States and worldwide.23 In the United States the first choice is often a nonsteroidal antiinflammatory drug, which has little proven efficacy and a high potential for renal impairment.3 Based on placebo-controlled randomized clinical trials, tricyclic antidepressants and anticonvulsants appear to have the best NNTs (number of patients needed to treat in order to provide at least 50% relief of symptoms in one patient). These are generally in the range 3 to 5,
with confidence intervals whose upper limits reach 10 in some instances.24 Imipramine or amitriptyline may be started at a dose of 25 mg at bedtime (10 mg in elderly patients) and titrated slowly up to a dose of 300 mg. Carbamazepine at a dose of 200 to 400 mg every 8 hours or gabapentin at a dose of 900 to 3600 mg/day are also effective treatments.25,26 Tramadol in two studies has shown an NNT below 5.24 Although tramadol is a mixed opioid, development of dependence in long-term use appears to be uncommon. In a recently published guideline, the following medications were recommended for the treatment of neuropathic pain: gabapentin, opioids, tramadol, and tricyclic antidepressants.3 Also being used is pregabalin at 150 to 600 mg/day, with a mechanism similar to that of gabapentin, and dulexetine at 60 mg/day, which is a selective serotonin and norepinephrine reuptake inhibitor.3 Among the selective serotonin reuptake inhibitors (SSRIs), paroxetine and bupropion appear to be effective, but fluoxetine is not.27 The summary NNT for the SSRIs has a confidence interval that reaches 50, suggesting that, pending further data, these agents should be considered second-line drugs.24 Topical capsaicin provides relief in some patients, but the burning associated with its application has limited its usage. Improving glycemic control can prevent, diminish, or reverse early diabetic DSPNs. Patients will typically spend over $1,000 per year for pain relief from diabetic DSPN.3
Alcoholic Distal Symmetrical Polyneuropathy Although the association between alcoholism and peripheral neuropathy has been well established for centuries, demonstration of a direct neurotoxic effect of alcohol remains elusive. The preponderance of evidence, from both observational studies in humans and experimental data from animal models,
Human Immunodeficiency Virus Neuropathies With the widespread use of highly active and effective antiretroviral treatment, peripheral neuropathies have become the most common neurologic complication of HIV infection. The typical HIV neuropathy is a DSPN, which appears to be triggered by a combination of dideoxynucleoside therapy and poorly characterized immune-mediated mechanisms associated with HIV.29,30 These patients require referral for specialized care. In addition to standard therapies for DSPN, lamotrigine has been shown to be moderately effective in the treatment of HIV-associated painful neuropathies.31
Toxic and Metabolic Neuropathies Many toxic agents and metabolic derangements produce a typical DSPN. Box 105-3 lists some of the most common toxic and metabolic causes of peripheral neuropathy. On the basis of preliminary results from a case-control study, the statins have been added to this list.32
Type 3: Asymmetrical Proximal and Distal Peripheral Neuropathies (Radiculopathies and Plexopathies) Radiculopathies are discussed in detail in Chapter 103. Plexopathies, which are discussed briefly in this chapter, are uncommon and often the result of trauma (Box 105-4). Generally, a plexopathy, whether brachial or lumbosacral, is identified by a process of elimination (i.e., a pattern of sensorimotor and reflex abnormalities that fit neither a radicular nor individual peripheral nerve distribution). Although this approach does not exclude a mononeuropathy multiplex on physical examination alone, a careful history should determine whether the patient is at risk for developing a mononeuropathy or plexop athy on the basis of underlying disease. Most plexopathies are often the result of blunt trauma and are usually seen in young men following motor vehicle accidents. Most present for evaluation several months after injury because of the need to recover from concurrent injuries. Therapeutic intervention is often delayed in order to maximize the potential for spontaneous recovery. Several surgical repairs exist, including neurotization or nerve transfer.33 Radiation (actinic) plexopathy occurs after a variable period of latency following treatment, which may extend to 20 years or more. Almost all series include women who received radiation treatment for breast cancer. Among neoplastic causes, most originate from the lung or breast. Patients with probable neoplastic brachial plexopathy need imaging studies and may require immediate radiation therapy. Pain control is the focus of management. Thoracic outlet syndrome remains a controversial disorder.34 Although the pendulum has swung over the past 50 years from
1403
Asymmetrical Proximal/Distal
BOX 105-4 Peripheral Neuropathies
Brachial plexopathy Open Direct plexus injury (knife or gunshot wound) Neurovascular (plexus ischemia) Iatrogenic (central line insertion) Closed Traction injuries “Stingers” Traction neurapraxia Partial or complete nerve root avulsion Radiation Neoplastic Idiopathic brachial plexitis Thoracic outlet Lumbosacral plexopathies Open Closed Traction injuries Pelvic double vertical shearing fracture Posterior hip dislocation Retroperitoneal hemorrhage Vasospastic (deep buttock injection) Neoplastic Radiation Idiopathic lumbosacral plexitis Infectious Herpesvirus (sacrococcygeal) Herpes simplex II Herpes zoster Cytomegalovirus (CMV) polyradiculopathy (HIV)
a postulated vascular cause to a neurogenic etiology, current evidence supporting the high prevalence of compression of the brachial plexus as a cause of thoracic outlet syndrome is in fact only slightly better35 than earlier evidence favoring a vascular etiology.36 Nevertheless, the disorder is currently felt to be most commonly due to compression of the medial or lower portion of the brachial plexus by a cervical rib or fibrous band.37 The syndrome is characterized by gradually progressive weakness and wasting of median and ulnar hand muscles with ulnar forearm and hand sensory signs and symptoms. Patients with this clinical picture should be referred for NCSs and EMG, which are said to be diagnostic.37 The treatment of true neurogenic thoracic outlet syndrome requires surgical removal of the rib or aberrant fibrous band to decompress the brachial plexus.38 An excellent discussion of this entity from a different perspective can be found in Chapter 85. Because of the complexity of plexopathies, there is no reason to expect that one can or should do more in the ED than localize the probable pathologic process to the brachial or lumbosacral plexus. Depending on severity and suspected etiology, one should either admit or refer the patient to a neurologist with experience in PNS disease.
Type 4: Isolated Mononeuropathies The pattern of asymmetrical, sensorimotor, usually distal, peripheral neuropathy is characteristic of a mononeuropathy. Mononeuropathies are of two main types: isolated and multiple. The isolated mononeuropathies are discussed in this section, while the multiple mononeuropathies, also termed
Chapter 105 / Peripheral Nerve Disorders
suggests that the association between alcohol and peripheral neuropathy may be confounded by nutritional status (i.e., deficiency states might be the true underlying cause of alcoholic peripheral neuropathy). The clinical and pathologic picture of alcoholic neuropathy is similar to that of the DSPN of diabetes. However, in alcoholism severe myopathy and cerebellar degeneration often complicate the clinical picture.28 Autonomic skin changes with atrophy and hair loss accompany the sensorimotor abnormalities. Often other systemic effects of alcoholism are so severe that the patient may not notice the neuropathic symptoms. All patients with suspected alcoholic DSPN should receive dietary supplements and referral for outpatient management.
1404
PART III ■ Medicine and Surgery / Section Seven • Neurology
BOX 105-5 Isolated Mononeuropathies Upper extremity Radial nerve Axilla Humerus Elbow (posterior interosseous neuropathy) Wrist (superficial cutaneous radial neuropathy) Ulnar nerve Axilla Humerus Elbow Condylar groove Cubital tunnel Wrist (Guyon’s canal) Hand Superficial terminal ulnar neuropathy Deep terminal ulnar neuropathy Proximal hypothenar Distal hypothenar Median nerve Axilla Humerus (musculocutaneous mononeuropathy) Forearm Anterior interosseus Pronator syndrome (?) Wrist (carpal tunnel) Hand (recurrent motor branch) Suprascapular mononeuropathy Axillary mononeuropathy Lower extremity Sciatic nerve Femoral nerve Iliacus compartment (proximal) Saphenous mononeuropathy (distal) Lateral femoral cutaneous (meralgia paresthetica) Peroneal nerve Common peroneal mononeuropathy (fibular head, popliteal fossa) Deep peroneal mononeuropathy (anterior compartment) Tibial nerve Popliteal fossa (proximal) Tarsal tunnel (distal) Sural nerve Popliteal fossa, calf (proximal) Fifth metatarsal base (distal) Plantar nerve Distal to tarsal tunnel Interdigital neuropathies (Morton’s neuroma) Obturator mononeuropathy
mononeuropathy multiplex, are discussed in the next section, as a type 5 peripheral neuropathy. Isolated mononeuropathies are usually caused by trauma, either blunt or penetrating (Box 105-5). If the trauma is blunt, the injury may be secondary to compression from an internal or external source. Entrapment neuropathies are a subset of compression neuropathies occurring at anatomic locations where nerves traverse potentially constricting compartments or tunnels.39 Isolated mononeuropathies may be acute, intermittent, or chronic and continuous. Antecedent peripheral neuropathy may be a risk factor for development of compression neuropathy (so-called double-crush syndrome), particularly in diabetics.40
Posterior cord, brachial plexus
Axillary nerve Posterior cutaneous nerve of forearm to triceps Intermuscular septum to Extensor carpi radialis longus to Extensor carpi radialis brevis Posterior interosseous nerve
to Latissimus dorsi to Triceps to Brachioradialis Superficial radial nerve Extensor digitorum Extensor digiti quinti to Extensor carpi ulnaris Extensor pollicis longus/brevis Abductor pollicis longus Arcade of Frohse
Supinator muscle
Figure 105-3. Radial nerve, major branches, right arm, lateral view.
(From Stewart JD: Focal Peripheral Neuropathies, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2000.)
Radial Mononeuropathy The radial nerve arises from C5-T1 roots. After exiting the brachial plexus, it passes behind the proximal humerus in the spiral groove and takes a lateral (radial) course down the upper arm (Fig. 105-3). At about the level of the antecubital fossa, it bifurcates into the posterior interosseous (pure motor) and superficial radial (pure sensory) nerves. The radial nerve controls extension of the fingers, thumb, wrist, and elbow (triceps). In contrast to the median and ulnar nerves, the radial nerve provides only extrinsic motor innervation to the hand (i.e., it does not supply motor fibers to any muscles that both originate and insert within the hand). In further contrast to the median and ulnar nerves, which supply most of the sensation to the hand, the radial nerve makes a contribution only to a cutaneous dorsal area overlying the first dorsal interosseus muscle, sometimes extending part of the way up the dorsa of the thumb, index, and long fingers. Radial mononeuropathy caused by involvement at the level of the axilla is uncommon. When it occurs, it is usually associated with other upper extremity mononeuropathies or a brachial plexopathy. Although improper use of crutches may cause this syndrome, it usually occurs after an extended period of unconsciousness during which the arm is positioned in such a way that prolonged, deep compression is applied to the axilla. Axillary radial mononeuropathy is distinguished from the more common humeral form by the finding of triceps involvement in addition to typical wrist and finger drop. Triceps involvement occurs because the innervation to the triceps is proximal to the point where the nerve is most vulnerable as it winds around the humeral shaft (see Fig. 105-3). Most radial mononeuropathies are due to so-called Saturday night palsies. The euphemism is derived from the association of radial mononeuropathy with improper positioning of the arm during deep, commonly inebriated sleep. Consequently, the radial nerve is trapped for a prolonged period between the humeral shaft and some firm surface, causing an external compression mononeuropathy. “Bridegroom’s palsy” is another eponym for radial mononeuropathy, so named because the radial nerve may be compressed by the bride’s head resting on the bridegroom’s arm during sleep. Because innervation of the wrist and finger extensors occurs distal to this area of the humeral shaft, findings are characterized by wrist and finger drop and mild numbness over the skin of the first dorsal interosseus muscle. Depending on the level,
Medial cord, brachial plexus
1405
recover spontaneously. In contrast, surgical intervention is needed to free the nerve from entrapment associated with complex fractures. While patients are waiting for spontaneous recovery to occur, the hand should be maintained in about 60 degrees of dorsiflexion. Although a simple dorsal plaster or fiberglass splint treats the wristdrop, atrophy and contractures can be minimized, and function of the hand can be improved if wide rubber bands anchored to the splint at a point proximal to the wrist are attached to individual fingers to provide passive dorsiflexion.
Ulnar Mononeuropathy The ulnar nerve includes C7-T1 roots and passes through the brachial plexus to descend medially, without branching, to the ulnar (medial) condylar groove at the elbow. It then enters the cubital canal, where it gives off branches to the ulnar wrist flexor and the deep flexors of the fourth and fifth digits. Just proximal to the wrist, two important sensory branches leave the main trunk to supply cutaneous sensation to part of the hand (Fig. 105-4). These are the palmar and dorsal cutaneous branches, which do not pass through Guyon’s canal. The palmar branch supplies sensation to the hypothenar eminence and the dorsal branch innervates the ulnar side of the dorsum of the hand, extending out nearly to the tip of the fifth and ulnar half of the fourth digit. At the wrist, the nerve enters Guyon’s canal (Fig. 105-5) between the pisiform and hook of the hamate, then bifurcates into the superficial terminal sensory branch and the deep motor branch. The superficial sensory nerve supplies ulnar sensation to the palmar side of the fifth and half of the fourth digit (see Fig. 105-5). The deep motor nerve supplies the hypothenar muscles, then crosses to the radial side of the palm to innervate
to Flexor carpi ulnaris to Flexor digitorum profundus (digits 4, 5) Palmar cutaneous branch Superficial terminal branch
Superficial terminal branch Dorsal ulnar cutaneous branch
Deep motor branch
∗
3
4 Deep terminal branch Pisiform
Hamate 2
1
Figure 105-5. Distal ulnar nerve and branches, right hand, palmar view. Figure 105-4. Ulnar nerve, major branches, right arm, anterior view. (From Stewart JD: Focal Peripheral Neuropathies, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2000.)
Numbers indicate four main sites of distal ulnar mononeuropathy in the wrist and hand. *Denotes hypothenar branches. (From Stewart JD: Focal Peripheral Neuropathies, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2000.)
Chapter 105 / Peripheral Nerve Disorders
degree, and duration of compression, some fascicles of the nerve may remain functional, resulting in a partial radial mononeuropathy. Thus, the superficial radial nerve may remain intact, resulting in no loss of sensation, or loss of wrist and finger extension may be incomplete. Because the finger drop of radial mononeuropathy places the hand at a mechanical disadvantage, examination of ulnar function by testing interossei may produce false-positive findings of weakness. To adjust for this, the examiner should ask the patient to place the palm on a horizontal supporting surface such as a stretcher. With the fingers extended and no longer “dropped” at the metacarpophalangeal joints, interosseous strength can now be fairly tested. Failure to perform this maneuver may cause misdiagnosis of a simple radial mononeuropathy as a brachial plexopathy in an effort to explain what appears to be radial and partial ulnar nerve involvement. About 90% of radial nerve palsies occurring during sleep, coma, or anesthesia recover fully, usually within 6 to 8 weeks. Evidence of denervation on EMG studies predicts a slower rate of recovery. Tourniquet injuries to the radial nerve usually recover spontaneously within 2 to 4 months. If axonal degeneration is seen on electrophysiologic testing, recovery may take longer, although virtually all radial mononeuropathies caused by tourniquets eventually resolve. About 75% of radial nerve injuries associated with a closed humeral shaft fracture
PART III ■ Medicine and Surgery / Section Seven • Neurology
1406
the ulnar intrinsics (all interossei and the ulnar lumbricals of the fourth and fifth digits), terminating in the first dorsal interosseus. The interossei abduct and adduct the fingers and are all innervated by the ulnar nerve. The lumbrical muscles flex the metacarpal phalangeal joints and are evenly divided between the ulnar (fourth and fifth) and median (second and third) digits. The ulnar nerve can be thought of as the complement to the median nerve in the hand because it supplies all of the muscles and all palmar sensation not innervated by the median nerve. The ulnar nerve may be injured at two locations near the elbow: in the ulnar condylar groove and distally in the cubital canal. Because the condylar groove is shallow, the ulnar nerve runs superficially in this location and is vulnerable to injury, usually from external pressure or from a fracture or dislocation. The ulnar nerve has a propensity to develop a “tardy ulnar palsy,” occurring years after a traumatic event. Many of these delayed ulnar mononeuropathies can be localized to the elbow on electrophysiologic testing. Some ulnar mononeuropathies occur secondary to compression just proximal to entry into the cubital canal or are entrapped within the canal itself. Transient symptoms may occur during prolonged flexion or with repeated flexion and extension at the elbow. Although distinguishing a condylar from a cubital ulnar mononeuropathy is difficult, it is usually possible to localize the problem to the region of the elbow or the wrist. In addition to prior probability heavily favoring the elbow, the presence of sensory abnormalities in an ulnar distribution in the hand and fingers (i.e., usually including the fifth digit and “splitting” the fourth digit) strongly suggests that the lesion is at the level of the elbow rather than the wrist. The ulnar cutaneous innervation to the hand branches off from the main trunk proximal to the nerve entering Guyon’s canal (see Figs. 105-4 and 105-5). Thus, a lesion at the wrist should not produce sensory abnormalities, whereas one at the elbow would be expected to do so. Compression of the ulnar nerve within Guyon’s canal is rare. When it does occur, it affects all of the ulnar intrinsics (i.e., the two ulnar [fourth and fifth] lumbricals) and all the interossei. However, the ulnar extrinsics (i.e., the deep flexors of the fourth and fifth digits) are not affected, nor is the ulnar flexor of the wrist. The only sensory abnormalities are those in the distribution of the superficial terminal sensory branch, sparing other areas of ulnar innervation (see Fig. 105-5). There are three ulnar mononeuropathies that occur distal to Guyon’s canal in the hand. The two most common ones involve the deep terminal branch, either proximal or distal to the separation of the hypothenar branches (see Fig. 105-5). If the lesion is proximal, it produces weakness of all the ulnar innervated muscles of the hand without sensory loss. If it is distal, the hypothenar ulnar intrinsics are spared but the picture is otherwise similar. Usually, this occurs secondary to a laceration or repeated compression in the hand from use of certain tools, a cane, or the handle of a crutch. Involvement of the superficial terminal branch (see Fig. 105-5) arises as pure sensory loss of the palmar surface of the fifth digit and ulnar half of the fourth digit caused by direct compression of this branch just distal to Guyon’s canal. The dorsal surface of these two digits should have normal sensation except for the distal tips. This configuration of findings is due to the intact innervation provided by the dorsal and palmar cutaneous branches that enter the hand without passing through Guyon’s canal (see Fig. 105-4). Most ulnar mononeuropathies will spontaneously resolve. However, if muscle atrophy, particularly in the hypothenar area, is detected, then surgery may be considered.41,42
Pronator teres muscle (superficial and deep heads) Anterior interosseous nerve “Sublimis bridge”
Flexor digitorum superficialis muscle
Palmar cutaneous branch Transverse carpal ligament
Figure 105-6. Median nerve, major branches, right arm, anterior view. (From Stewart JD: Focal Peripheral Neuropathies, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2000.)
Median Mononeuropathy The median nerve arises from C5-T1 spinal nerve roots and exits the brachial plexus through the lower trunk (Fig. 105-6). Median mononeuropathy is usually diagnosed as carpal tunnel syndrome (CTS), which is the most common of all entrapment neuropathies. CTS has a prevalence of 3 to 6% in the U.S. population.43 Although the patient may complain of bilateral symptoms, a careful history usually reveals that symptoms in one hand preceded those in the other. Awakening at night and shaking the hand is a common symptom of CTS. Symptoms are often worsened by activity. For unclear reasons, the pain may spread as high as the arm or shoulder, although the paresthesias are generally confined to the fingers. Many patients on initial questioning state that their entire hand is involved, although this is not supported by careful sensory examination. Complaints that the hands are clumsy or weak, especially when holding a glass or opening a screw-top container, are frequent. The skin of the fingers innervated by the median nerve may be drier and rougher to the touch than the corresponding ulnar skin, depending on the duration of entrapment.44 When there is motor involvement in CTS it is confined to the median intrinsics, which innervate the lumbricals (flexion of the metacarpal phalangeal joints), and subserve thumb opposition, abduction, and flexion, known as the LOAF muscles. However, the hallmark of CTS is sensory involvement, with motor abnormalities occurring later. The typical pattern of sensory innervation of the hand by the median, ulnar, and radial nerves shows marked individual variation.
1407
Conditions Associated with Carpal
Acromegaly Amyloid Diabetes mellitus Hypothyroidism Obesity Pregnancy Renal failure Rheumatoid arthritis
Superior gluteal nerve Sciatic nerve Inferior gluteal nerve Post. cutaneous nerve of thigh
Semitendinosus
The most specific finding for CTS is splitting of the fourth digit (i.e., normal sensation of the ring finger on the ulnar palmar side with abnormal sensation on the median [radial] palmar side of the same finger). The most sensitive finding is abnormal sensation of the distal palmar tip of the index finger. If sensory findings are absent in the presence of motor findings consistent with median nerve involvement, it is highly unlikely that the patient has CTS, and an alternative diagnosis should be sought. If neither sensory nor motor symptoms are evident, none of the provocative tests originally reported to reproduce the sensory symptoms of CTS—of which the most common are Tinel’s sign (percussion of the median nerve at the wrist) and Phalen’s sign (maximal palmar flexion at the wrist)—has shown adequate sensitivity or specificity to determine which patients should be referred for electrodiagnostic studies.45–47 As suggested earlier, the best way to examine patients for sensory findings is to touch the distal palmar tips very lightly, asking the patient whether the sensation feels “abnormal.” CTS appears to be associated with the conditions listed in Box 105-6. Of these, the two most common are diabetes mellitus and pregnancy. CTS associated with systemic illness is commonly bilateral. Although CTS in pregnancy may be selflimiting, about half the women in one series were still symptomatic at 1-year follow-up.48 All patients with suspected CTS should be referred for NCSs. However, because of the dissociation between clinical and electrodiagnostic indicators of CTS early in the disease, patients with normal electrodiagnostic findings in the presence of symptoms suggestive of CTS (with or without signs) should have an MRI49 or sonogram.50 At present, the sensitivity of MRI is good but its specificity is poor.51 Ultrasound has been shown to be useful particularly in patients with symptoms and a negative NCS. This is done by measuring the cross-sectional area of the median nerve at the end of the pisiform.52 Thus, if all diagnostic studies in a symptomatic patient are negative, or if only the MRI result is positive, they should be repeated within a few months if symptoms do not resolve.44 This recommendation is based on the theory that the CTS will progress over time to the point that an objective indicator such as the NCS will become positive. Because of the possibility of development of a disabling “median hand” after inadvertent direct injection of the median nerve, one should not inject the carpal tunnel with steroids in the ED. The physician to whom the patient is referred can decide after NCS whether to recommend splinting, injection, or surgical division of the transverse carpal ligament. Endoscopic repair appears to provide excellent results.53
Sciatic Mononeuropathy The sciatic nerve includes L4-S3 spinal nerve roots that pass through the lumbosacral plexus and divide into two terminal branches: the common peroneal and tibial nerves. The nerve
Semimembranosus Biceps, long head
Biceps, short head
Adductor magnus
Tibial nerve Common peroneal nerve
Figure 105-7. Sciatic nerve, major branches, right leg, posterior view.
(From Stewart JD: Focal Peripheral Neuropathies, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2000.)
exits the pelvis through the sciatic notch, passes behind the hip, and remains deep in the thigh until its terminal bifurcation in the proximal popliteal fossa (Fig. 105-7). Lesions of the sciatic nerve occur with posterior hip dislocation or with virtually any form of penetrating or blunt trauma that causes formation of a buttock hematoma. Other causes include deep gluteal injection and prolonged supine immobilization on a firm surface. Because the sciatic nerve innervates the hamstrings and provides all sensorimotor function distal to the knee, a complete sciatic mononeuropathy is a devastating injury. Ambulation is extremely difficult because of inability to flex the knee and a flail foot (i.e., neither flexion nor extension is possible at the ankle). Fortunately, many sciatic mononeuropathies are incomplete. For unknown reasons, a partial lesion typically involves only the trunk of the sciatic nerve, which subsequently becomes the common peroneal nerve, sometimes making the two difficult to distinguish from one another clinically. On electrophysiologic studies, evidence of involvement of gluteal muscles or of any muscles innervated by the tibial nerve readily distinguishes a partial sciatic mononeuropathy from a lesion of the common peroneal nerve. Treatment of footdrop requires a posterior splint to maintain the ankle at 90 degrees until a brace can be obtained (see later section on “Common Peroneal Mononeuropathy”).
Lateral Femoral Cutaneous Mononeuropathy Lateral femoral cutaneous mononeuropathy (meralgia paresthetica) is a common syndrome believed to be caused by injury to this pure sensory nerve as it passes through or over the inguinal ligament, where it may become entrapped or kinked. Along with facial nerve neuropathy, meralgia paresthetica is
Chapter 105 / Peripheral Nerve Disorders
BOX 105-6 Tunnel Syndrome
PART III ■ Medicine and Surgery / Section Seven • Neurology
1408
one of the most commonly reported mononeuropathies associated with HIV. External pressure and obesity may also contribute to nerve injury, causing numbness and dysesthesia over the skin of the upper lateral thigh. Regression usually occurs spontaneously, but recurrence is common and may require a release procedure for the inguinal ligament.
Common Peroneal Mononeuropathy The common peroneal nerve is a continuation of one trunk of the sciatic nerve. It is most vulnerable to injury where it winds around the fibular neck (Fig. 105-8). It then passes through the fibular canal and bifurcates into its terminal branches, the superficial and deep peroneal nerves. The superficial peroneal nerve innervates the peroneal muscles (foot everters) and supplies sensation to the lateral, distal lower leg and dorsum of the foot. The deep peroneal nerve traverses the anterior compartment and supplies innervation to the dorsiflexors of the foot and toes, plus cutaneous sensation between the first and second toes. Most common peroneal mononeuropathies are idiopathic and thought to be related to compression where the nerve is superficially located lateral to the fibular neck. Because this common neuropathy is often noted on awakening, it may be secondary to position during sleep. Leg crossing may also be a risk factor for development of this mononeuropathy. The most striking feature of a complete common peroneal mononeuropathy is footdrop caused by weakness of foot dorsiflexion. At testing, the everters of the foot are also weak, but the inverters, which are innervated by the tibial nerve, remain strong. This is the single most reliable clinical feature distinguishing sciatic from common peroneal mononeuropathy. Analogous to radial mononeuropathy in the upper extremity, sensory abnormalities in the leg and foot are inconstant and easily overlooked in peroneal mononeuropathy. Most patients
with peroneal palsy recover. Those who do not should be studied electrophysiologically to ensure that the point of compression is not proximal to the fibular neck (i.e., in the popliteal fossa). If the point of peroneal injury appears to be in the region of, or distal to, the fibular neck on EMG, patients whose footdrop does not resolve should be considered candidates for exploration to determine whether the nerve is compressed within the fibular canal. Treatment of common peroneal palsy may require a posterior splint to maintain the ankle at 90 degrees until the nerve regenerates. This splinting prevents the foot from falling into sustained equinus (plantar flexion), which in turn allows the intermalleolar distance to narrow, effectively locking the talus out of the ankle mortice. The treatment of isolated mononeuropathies depends on their etiology, location, and natural history of spontaneous recovery. All penetrating neuropathies should have surgical exploration and repair performed. Blunt trauma may cause a mononeuropathy indirectly by entrapment of a nerve within a fracture, hematoma, or compartment, requiring surgical intervention. Alternatively, nerves may be injured at a point where they are superficial, either by a single direct blow or by sustained pressure caused by immobility (pressure palsies). Most of these resolve spontaneously over time, depending on the severity of injury and length of the nerve. If entrapment can be confirmed by imaging or electrophysiologic studies, a release procedure is indicated. In many instances, when there is disagreement between clinical and EMG findings, MRI may be helpful in selecting patients for exploration by visualizing entrapment or traction.54 Characteristic sonographic findings have also been reported in several mononeuropathies.55 The mononeuropathies that do not require timely surgical exploration should be referred for further workup to confirm the location of the neuropathic lesion.
Type 5: Mononeuropathy Multiplex
Sciatic nerve Tibial nerve Common peroneal nerve Fibular tunnel Peroneus longus muscle Superficial peroneal nerve to Peroneus brevis
Mononeuropathy multiplex is characterized by an asymmetrical, sensorimotor, usually distal pattern of peripheral neuropathy (Box 105-7). As with isolated mononeuropathies, sensory abnormalities tend to be located in the same general anatomic region as the accompanying motor findings. Whether DTRs are affected depends on which nerves are involved. For example, if the process includes the femoral nerve, the patellar reflex is likely to be diminished or absent.
BOX 105-7 Mononeuropathy Multiplex Deep peroneal nerve to Tibialis anterior to Extensor digitorum longus to Extensor hallucis longus
to Extensor digitorum brevis
Figure 105-8. Common peroneal nerve, major branches, right leg,
anterolateral view. (From Stewart JD: Focal Peripheral Neuropathies, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2000.)
Vasculitis Systemic vasculitis Polyarteritis nodosa Rheumatoid arthritis Systemic lupus erythematosus Sjögren’s syndrome (keratoconjunctivitis sicca) Nonsystemic vasculitis Diabetes mellitus Neoplastic Paraneoplastic Direct infiltration Infectious Lyme disease HIV Sarcoid Toxic (lead) Transient (polycythemia vera) Cryoglobulinemia (hepatitis C)
1409
Vasculitis
Diabetes Mellitus Although the role of ischemia in diabetic neuropathies is controversial, evidence for a vascular cause is stronger in the asymmetrical diabetic multiple mononeuropathies than in the more common DSPNs seen in diabetes.
Lyme Disease The PNS manifestations of Lyme disease can be divided into early and late. The early PNS syndromes commonly include facial nerve involvement (rarely other cranial nerve palsies) and radiculoneuritis. Late PNS involvement occurs as a DSPN, mononeuropathy multiplex, or radiculoneuropathy. The most common neurologic abnormality in Lyme disease is unilateral or bilateral facial nerve palsy, usually occurring within a month of exposure. Patients may also complain of headache and constitutional symptoms. Early in the course of Lyme disease, severe neuritic pain may develop in a radicular distribution, often in or near the dermatome where the tick bite occurred. There may also be associated sensory changes, motor weakness, and decreased reflexes consistent with nerve root involvement. Patients with chronic Lyme disease present with sensory symptoms, particularly distal paresthesias in the lower extremities. Less commonly, they develop a picture consistent with mononeuropathy multiplex or a radiculopathy. The latter is much less severe than the early radiculoneuritis of Lyme disease. The most useful diagnostic tests for patients with suspected Lyme disease are a serum enzyme-linked immunosorbent assay, Western blot, and CSF examination. CSF abnormalities suggestive of Lyme disease are a lymphocytic pleocytosis, elevated protein, and normal glucose. The CSF is almost always abnormal in early radiculitis, sometimes abnormal with isolated facial palsy, and typically normal in chronic Lyme disease. Facial nerve palsy without CSF abnormalities may be treated with oral doxycycline 100 mg twice a day for 2 weeks. Intravenous ceftriaxone is the drug of choice for all other neurologic syndromes associated with Lyme disease. The adult dosage is 2 g/day, and the pediatric dosage is 75 to 100 mg/kg/day. Treatment with ceftriaxone should be continued for at least 2 weeks.
Type 6: Amyotrophic Lateral Sclerosis Although amyotrophic lateral sclerosis (ALS) and motor neuron disease (MND) are often used synonymously, the latter represents a spectrum of diseases ranging from primary lateral sclerosis, in which degeneration is confined to upper motor neurons, to progressive muscle atrophy, in which only lower motor neurons are involved. ALS, which requires the presence of both upper and lower motor neuron findings, resides in the middle of this spectrum, representing the most common form of MND. The incidence of ALS is 1.5 to 2.5 per 100,000.56 In ALS, the primary pathologic process in the PNS component of the disease is a neuronopathy of the anterior horn cell. Because this structure is located proximal to the point where
Objective Clinical Findings Consistent with
BOX 105-8 Amyotrophic Lateral Sclerosis
Upper motor neuron signs Hyperreflexia Sustained clonus, especially at ankle Finger flexors and jaw jerk Spasticity, especially of gait Positive Babinski sign Lower motor neuron signs Positive motor phenomena Fasciculations Cramps Negative motor phenomena Asymmetrical distal weakness Atrophy Combined upper and lower motor neuron signs Dysarthria Dysphagia Respiratory compromise
motor and sensory fibers merge to form mixed spinal nerve roots, the signs and symptoms of MND are purely motor (see Fig. 105-2). In the CNS, there is a loss of Betz cells from the motor cortex with secondary degeneration of the corticospinal tracts. Box 105-8 lists some representative upper, lower, and mixed motor signs. Patients typically demonstrate asymmetrical distal weakness without sensory findings. Positive motor phenomena in the form of fasciculations are found in almost all patients at diagnosis but are rarely an initial complaint. Although there is electrophysiologic evidence of autonomic involvement in ALS, this is generally subclinical. Most patients with an asymmetrical, distal, pure motor neuropathy have ALS, for which only supportive treatment is currently available. However, there are some preliminary studies of recombinant human insulin–like growth factor 1 that have shown marginal improvement in this otherwise fatal disease.57 All patients in whom this diagnosis is suspected should be referred for electrophysiologic confirmation against standardized criteria. Confirmation is particularly important because multifocal motor neuropathy, a rare disease that masquerades as ALS, responds dramatically to cyclophosphamide and immunoglobulin administration.58
Type 7: Sensory Neuronopathy (Ganglionopathy) This category of peripheral neuropathy is characterized by a selective or predominant involvement of the dorsal root ganglion, producing a relatively pure sensory syndrome analogous to the pure motor syndrome of ALS. Although all sensory modalities are affected, proprioception is profoundly altered, leading to sensory ataxia and loss of DTRs without weakness. The distribution is typically asymmetrical and distal at the outset, but depending on severity and extent of progression, it may become functionally symmetrical. Sensory ganglionopathies can now be confirmed by MRI of the spinal cord and surrounding areas, showing degeneration of central sensory projections that localize the disease process to the dorsal root ganglion.59 Some of the more common causes of this type of peripheral neuropathy are listed in Box 105-9.
Chapter 105 / Peripheral Nerve Disorders
Mononeuropathy multiplex is strongly associated with vasculitis, which is the most common indication for sural nerve biopsy in most series. However, because diabetes mellitus is far more prevalent than vasculitis, the most common cause of mononeuropathy multiplex among ED patients is diabetes.
1410
PART III ■ Medicine and Surgery / Section Seven • Neurology
BOX 105-9 Sensory Neuronopathies (Ganglionopathies) Herpes Herpes simplex I and II Varicella zoster (shingles) Inflammatory sensory polyganglionopathy (ISP) Paraneoplastic Primary biliary cirrhosis Sjögren’s syndrome (keratoconjunctivitis sicca) Toxin induced Pyridoxine (vitamin B6) overdose Metals Platinum (cisplatin) Methyl mercury Vitamin E deficiency
■ ANCILLARY DIAGNOSTIC TESTING Relatively few blood tests contribute to the diagnosis of peripheral neuropathy, and only a small number of these are available in the ED. CSF analysis may be helpful in GBS and Lyme disease. Additional tests that may be indicated in patients referred for evaluation are listed in Box 105-10, along with others that may be ordered selectively, depending on the clinical picture. Expensive batteries of tests purporting to measure a wide variety of antibodies to components of peripheral neuropathies are commercially available but have not been shown to be useful as screening tests.
KEY CONCEPTS Ancillary Diagnostic Testing in Suspected BOX 105-10 Peripheral Neuropathy Obtained in most patients Complete blood count (CBC) Erythrocyte sedimentation rate (ESR) Glucose Creatine kinase (CK) Creatinine Obtained only if indicated Human chorionic gonadotropin (HCG) Magnesium Phosphate Vitamin B12 Hemoglobin A1c Serum protein electrophoresis (SPEP) with immune fixation electrophoresis (IFE) Venereal Disease Research Laboratory (VDRL) or rapid plasma reagin (RPR) screen with fluorescent treponemal antibody-absorption (FTA-ABS) test, as appropriate Thyroid function HIV Lyme enzyme-linked immunosorbent assay (ELISA) and Western blot Rheumatoid factor and antinuclear antibody Blood, urine, hair, or nails for metal, depending on suspected chronicity of exposure Specific serum antibodies to components of peripheral nervous system Cerebrospinal fluid for cells, protein, Lyme titer Electrodiagnostic testing Nerve conduction studies Electromyography Neurodiagnostic imaging Magnetic resonance imaging Computed tomography Sonography Quantitative sensory testing Nerve biopsy Sural Intraepidermal nerve fiber density
■
In the emergency department, it is not usually possible to arrive at the diagnosis of a specific peripheral neuropathy because of the need for confirmatory ancillary testing that is beyond the scope of emergency practice. Rather, the focus should be on identifying one of seven categorical patterns of peripheral neuropathy, shown in Figure 105-1 and listed in Table 105-1, after other non-PNS causes have been eliminated. ■ One of these seven patterns can usually be identified by combining three clinical features that are readily obtainable from a goal-directed history and physical: (1) right-left symmetry or asymmetry, (2) proximal-distal location, and (3) sensorimotor modalities affected. This approach is summarized as an algorithm in Figure 105-1. ■ Identification of one of the seven types of peripheral neuropathy determines the need for ancillary diagnostic testing, therapeutic intervention, disposition, and the timing of neurologic referral. ■ Respiratory compromise is the primary life-threatening event seen in some peripheral neuropathies; GBS is by far the most common peripheral neuropathic cause of respiratory arrest. ■ Any patient with symmetrical weakness, distributed both proximally and distally, with loss or diminution of DTRs and variable sensory abnormalities should be treated as having GBS.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 106
Neuromuscular Disorders
Peter Shearer and Andy Jagoda
■ PERSPECTIVE Disorders of the neuromuscular unit can result in clinical presentations that range from subtle symptoms to acute respiratory failure. Morbidity and mortality are often related to failure of the muscles that maintain airway integrity and drive respiration. In most cases, the pathophysiology of these disorders is well understood and permits an organization and understanding that is based on the level of the nervous system affected. This facilitates an approach that is based on signs and symptoms, the findings of which direct the urgency of diagnostic testing and treatment. Processes involving the brainstem and brain can usually be differentiated from those in the spinal cord and in the peripheral nervous system on the basis of historical and physical findings. In general, lesions at the level of the brainstem or above produce unilateral weakness; bilateral weakness caused by lesions above the spinal cord is generally associated with a change in mental status or cranial nerve involvement. Lesions of the central nervous system result in upper motor neuron signs that include spasticity, hyperreflexia, and extensor plantar reflexes. As a corollary, when bilateral upper motor neuron signs are found in conjunction with normal mental status, diagnostic testing including neuroimaging should focus on looking for a lesion in the spinal cord.
■ PRINCIPLES OF DISEASE The neuromuscular unit has four components: the anterior horn cells of the spinal cord, the peripheral nerve, the neuromuscular junction, and the muscle being innervated. The level of the pathology determines associated signs and symptoms (Table 106-1). Myelopathies involve the spinal cord; radiculopathies involve the nerve roots as they leave the spinal cord; neuropathies involve the peripheral nerves; and myopathies involve the muscle. The use of physical signs to differentiate these disorders is discussed in Chapter 11. Neuropathies involve the axon itself or the myelin sheath (or the Schwann cells that make the myelin sheath) of the nerve. Nerve conduction studies can differentiate the locations of involvement. As the conduction along the axon is disrupted, the subsequent delay in transmission first causes symptoms in the muscles controlled by longer nerve axons, resulting in a history of weakness beginning in the distal extremities. As the myelin destruction or axonal degeneration progresses, patients usually note a slowly progressive course of symptoms.
The motor nerve branches into multiple terminals as it approaches the muscle. The neuromuscular junction is composed of the presynaptic membrane, the postsynaptic membrane, and the synaptic cleft. The neurotransmitter is acetylcholine (ACh). The motor synapse is a nicotinic receptor, whereas muscarinic synapses link the central nervous system with the autonomic nervous system. Disorders of the postsynaptic nicotinic receptors produce weakness. Postsynaptic ACh receptors are continually turned over at a rate that is related to the amount of stimulation. A disorder of transmission often leads to increased production of ACh receptors. Myasthenia gravis (MG) is the prototype of neuromuscular junction diseases.
■ CLINICAL FINDINGS History The history of patients with complaints of weakness focuses on the acuity and progression of onset and the potential for airway compromise. Any complaint of difficulty breathing or swallowing raises suspicion of bulbar involvement and concern for life-threatening deterioration. The history must elicit whether the weakness is muscular or nonspecific generalized fatigue. Weakness implies the inability to exert normal force, whereas fatigue implies a decrease in force with repetitive use. When muscular weakness exists, the clinician should determine whether it is focal or generalized, proximal or distal. The history of present illness must include the duration of symptoms, exacerbating and mitigating factors, and presence of associated symptoms such as fever, weight loss, and bowel or bladder changes. Historical elements might explain the presenting complaint: a preexisting neuromuscular disorder that could lead to deterioration; prior episodes or a family history of weakness suggesting periodic paralysis; a recent respiratory or diarrhea illness suggesting a postinfectious, autoimmune process such as transverse myelitis or Guillain-Barré syndrome (GBS); a cancer history suggesting a metastatic tumor as the cause of a compressive myelopathy; a food or travel history suggesting botulism or tick exposure.
Physical Examination The physical examination should first assess the patient’s airway and ventilation and then proceed to localize the level 1411
1412
PART III ■ Medicine and Surgery / Section Seven • Neurology
Table 106-1 Clinical Characteristics of Neuromuscular Diseases DISEASE
HISTORY
STRENGTH
DEEP TENDON REFLEX
SENSATION
WASTING
Myelopathy Motor neuron disease (ALS) Neuropathy
Trauma, infection, cancer Progressive difficulty swallowing, speaking, walking Recent infection Ascending weakness Food (canned goods) Tick exposure Easy fatigability Thyroid disease Previous similar episodes
Normal to decreased Decreased
Increased Increased
Normal to decreased Normal
No Yes
Normal or decreased Distal > proximal Normal to fatigue
Decreased
Decreased
Yes
Normal
Normal
No
Normal
Normal
Yes
Neuromuscular junction disease Myopathy
Decreased Proximal > distal
ALS, amyotrophic lateral sclerosis.
BOX 106-1 Grading Score for Motor Strength 5 = Normal strength 4 = Weak but able to resist examiner 3 = Moves against gravity but unable to resist examiner 2 = Moves but unable to resist gravity 1 = Flicker but no movement 0 = No movement
of the lesion. The presence of swallowing and a strong cough suggests that the patient has sufficient protective and ventilatory reserve. The muscles used to lift the head off the bed may weaken before those of respiration and should be assessed. A patient who is not yet intubated but is complaining of shortness of breath or difficulty breathing should have frequent vital capacity measurements. Normally, these values range from 60 to 70 mL/kg. When the forced vital capacity reaches 15 mL/kg, intubation is necessary. If vital capacity cannot be measured, a maximal negative inspiratory force (NIF) is easily determined. An NIF less than 15 mm Hg suggests the need for endotracheal intubation. A bedside assessment used to follow ventilatory status is to have the patient count numbers with one breath.1 With sequential performance of this test, a decline in respiratory function is detected as the patient fails to count as high as before. Arterial blood gas is not necessarily helpful because functional reserve can be severely diminished by the time a patient develops either hypercarbia or hypoxia. The assessment of vital signs is important because some causes of weakness may result in dysregulation of the autonomic system. A systematic neurologic examination should assess the patient’s mental status, cranial nerves, motor function, sensory function, deep tendon reflexes, and coordination, including cerebellar function. The motor examination begins by determining whether the weakness is unilateral or bilateral and which muscle groups are involved. Key components of the examination include motor strength, muscle bulk, and pres-
ence of fasciculations. Box 106-1 provides the grading system used in motor strength assessment. Table 106-2 provides the findings used to distinguish upper motor neuron from lower motor neuron processes.
Differential Considerations Myelopathies A patient with myelopathy shows signs of upper motor neuron dysfunction. Without upper motor neuron function, muscle weakness is present with increased spinal reflexes, including an extensor plantar reflex (Babinski’s response). Muscle tone initially ranges from normal to slightly increased, eventually leading to spasticity as a late finding. The same reflex arcs eventually create spasticity in the affected muscles. The weakness is ascending in nature, and there is often bladder and bowel involvement. When sensory findings are present, they often define the level of the lesion. The presence of bowel or bladder dysfunction, or diminished sensation, localizes the lesion to the spinal cord. The presence of pain often connotes a compressive lesion such as a herniated intravertebral disk, epidural hematoma, abscess, or tumor. Acute, painless spinal cord lesions include transverse myelitis and spinal cord infarction.
Motor Neuron Disease Amyotrophic lateral sclerosis is the prototypical disease process resulting from a degeneration of the motor neuron without sensory involvement. These patients may complain of dysarthria or dysphagia; however, the characteristic findings are those of combined upper and lower motor neuron dysfunction. Consequently, findings include hyperreflexia, muscle wasting, and fasciculation. Pain is not a component of the clinical picture. Poliomyelitis affects the anterior horn cells and results in lower motor neuron disease without sensory involvement. The
Table 106-2 Distinguishing Upper Motor Neuron (UMN) from Lower Motor Neuron (LMN) Involvement MOTOR NEURON
DEEP TENDON REFLEX
MUSCLE TONE
ATROPHY
FASCICULATIONS
BABINSKI
UMN LMN
Increased Decreased
Increased Decreased
No* Yes
No Yes
Present Absent
*Not significant but can occur.
Neuropathies Weakness from a neuropathy is often noted first in distal muscles and then ascends. Decreased grip strength or footdrop may be noted first. Muscle tone ranges from slightly diminished to flaccid. As all outflow from the spinal cord is affected, deep tendon reflexes are diminished or absent. Patients exhibit varying degrees of altered sensation, muscle wasting, and fasciculation depending on the duration of the symptoms. Disorders that should be considered include GBS, toxic neuropathies, diabetic neuropathy, and tick paralysis (which is caused by inhibition of both nerve conduction and function of the neuromuscular junction).
Diseases of the Neuromuscular Junction Disorders of the neuromuscular junction cause progressive motor fatigability. The initial depolarization of the muscle causes stimulation of a maximum number of receptors, producing a normal, or nearly normal, strength response. Repeated stimulation leads to diminishing motor strength, which is caused by the blockage of the receptors (as in MG) or by a decrease in the amount of ACh released (as in botulism) or by inactivating Ach by irreversibly binding with it (as in organophosphate poisoning). A decrease in the release of ACh may produce a combination of nicotinic and muscarinic effects leading to anticholinergic findings such as decreased visual acuity, confusion, urinary retention, tachycardia, low-grade fever, and dry, flushed skin. In the case of Lambert-Eaton myasthenic syndrome, weakness is more pronounced at the beginning of muscle use and improves with repeated use as more ACh builds up in the synaptic cleft with each stimulation. Diseases of the neuromuscular junction should be considered in patients who present with generalized weakness in association with an acute cranial nerve deficit. Muscle tone is generally diminished and sensation is preserved.
Myopathies Myopathies produce generalized, symmetrical weakness. Reflexes are present but diminished, muscle tone is usually diminished, but sensation is preserved. Myopathies caused by inflammatory disorders (polymyositis, dermatomyositis, polymyalgia rheumatica, and viral myositis) cause muscle pain and tenderness. Metabolic disorders affecting muscle strength (e.g., electrolyte and endocrine disorders) are painless in nature.
■ DIAGNOSTIC STRATEGIES Laboratory Studies Serum potassium, calcium, and phosphorus should be assessed in patients with acute weakness. Thyroid function tests are recommended in cases of suspected myopathies. A creatine kinase (CK) level assesses for muscular inflammation; a urinalysis should be performed for the presence of myoglobinuria and possible rhabdomyolysis.
1413
Special Studies Magnetic resonance imaging (MRI) is the preferred test for suspected cases of acute myelopathy. Computed tomography of the spinal cord with myelography can help to differentiate compressive (herniation, abscess, tumor) from noncompressive causes when MRI is not available. Cerebrospinal fluid analysis is indicated when GBS or transverse myelitis is suspected.
■ SPECIFIC DISORDERS Disorders of the Neuromuscular Junction Myasthenia Gravis Perspective. It is rare for the emergency physician to diagnose a new case of MG; more commonly, patients with established disease present with exacerbations of their disease that must be recognized and addressed. In addition, the emergency physician must be aware of medication interactions in patients with MG. Principles of Disease. MG is a rare disorder that affects approximately 60,000 Americans.2 Age of onset is bimodal, with the first peak among women 20 to 40 years of age and a second peak among men 50 to 70 years old. MG results from autoantibodies directed against the nicotinic acetylcholine receptor (AChR) at the neuromuscular junction. This leads to complement-mediated destruction of AChRs with a decrease in the total number of available receptors. The autoantibodies further compete with ACh for binding at remaining receptors. Thus, with repeated stimulation of the same muscle, fewer and fewer sites are available and fatigue develops. Fatigability and muscular weakness are the hallmarks of MG. Considering the slow clinical progression of MG and the low likelihood of short-term complications from its progression, the importance of suspecting the diagnosis is to facilitate proper referral for further evaluation. Clinical Features. Ocular symptoms are often the first manifestation of MG. The typical symptoms are ptosis, diplopia, or blurred vision. Ocular muscle weakness is the first sign in up to 40% of patients, although 85% of patients with MG eventually have ocular involvement. When present, ptosis is often worse toward the end of the day. Respiratory failure is rarely the initial symptom of MG. Even so, up to 17% of patients may have weakness of the muscles of respiration.3 Bulbar muscles may be involved, producing dysarthria or dysphagia. Lambert-Eaton myasthenic syndrome is a rare disorder in which almost 50% of cases are associated with small cell carcinoma of the lung. Autoantibodies cause inadequate release of ACh from nerve terminals, affecting both nicotinic and muscarinic receptors. With repeated stimulation, the amount of ACh in the synaptic cleft increases, leading to an increase in strength, the opposite of that seen with MG. The classic syndrome includes weakness that improves with use of muscles, particularly proximal hip and shoulder muscles; hyporeflexia; and autonomic dysfunction, most commonly seen as dry mouth.4 Management primarily focuses on treating the underlying neoplastic disorder, although IVIG has been reported to be useful.5 Diagnostic Strategies New-Onset Myasthenia Gravis. The diagnosis of MG is based on clinical findings and a combination of serologic testing, electromyographic testing, and the bedside edrophonium or ice bag tests. Serum testing for AChR antibodies is positive in 80 to 90% of patients with MG, but not available in the emergency department setting.
Chapter 106 / Neuromuscular Disorders
weakness can be symmetrical or more often asymmetrical. Patients initially have a clinical picture similar to that of viral meningitis, with fever and neck stiffness. Currently, most cases follow exposure of an immunocompromised host to the oral polio vaccine, and this should be sought in the history. The cerebrospinal fluid analysis resembles that of viral meningitis.
PART III ■ Medicine and Surgery / Section Seven • Neurology
1414
The edrophonium test and ice bag tests are similar to perform and the results are based on their effect on the ptosis seen in patients with suspected MG. The production of edrophonium was discontinued in early 2008 and it will no longer be available once current stores are depleted. Edrophonium is a short-acting acetylcholinesterase (AChE)-blocking agent that produces an increase of ACh in the synaptic cleft and a reduction in ptosis after intravenous administration. With the ice bag test, cooling decreases symptoms in MG while heat exacerbates symptoms. In both tests, the change in the amount of ptosis is measured before and after the application of edrophonium or an ice bag. The distance from the upper to the lower eyelid in the most severely affected eye is measured first. If edrophonium is given, an intravenous test dose of 1 to 2 mg is given first as some patients have a severe reaction. If no adverse reaction is found and the patient does not dramatically improve in 30 to 90 seconds, a second dose of 3 mg is given. If there is still no response, a final dose of 5 mg is given for a total maximum dose of 10 mg.6 Atropine should be available at the bedside during the test. Because of the potential for cholinergic-induced increased airway secretions, this test should be used with caution in asthmatics and patients with chronic obstructive pulmonary disease. If an ice pack is used, it is applied to the affected eye for approximately 2 minutes, and the distance between the lids is measured again. A prospective evaluation of the ice bag approach found the test result to be positive (an improvement in distance of at least 2 mm) in 80% of patients with MG and in no patients without MG.7 Myasthenic Crisis. Myasthenic crisis is defined as respiratory failure leading to mechanical ventilation. It occurs in 15 to 20% of patients with MG,8 usually within the first 2 years of disease onset. Although it is potentially life-threatening, the mortality from this complication of MG has declined dramatically with better and more aggressive ICU care and the use of plasmapheresis (PE) and/or immunodulatory therapy with high-dose intravenous immunoglobulin (IVIG) and corticosteroids. Underlying infection, aspiration, and medication changes— stopping anticholinergic medications or taking a new medication that precipitates weakness—most often set off crisis, but the precipitant may not be found in up to 30% of cases.9 Other precipitants can be surgery and pregnancy (Box 106-2).
BOX 106-2 Drugs That May Exacerbate Myasthenia Gravis Cardiovascular Beta-blockers Calcium channel blockers Quinidine Lidocaine Procainamide Antibiotics Aminoglycosides Tetracyclines Clindamycin Lincomycin Polymyxin B Colistin Other Phenytoin Neuromuscular blockers Corticosteroids Thyroid replacement
The initial step in managing the patient in crisis is stabilization of the airway. In less severe cases in which intubation is not imminent, it is imperative to monitor ventilatory status pending intensive care unit admission through forced vital capacity or NIF measurement. Noninvasive ventilation with biphasic positive airway pressure may be effective in managing patients who need ventilatory support.10 Signs of myasthenic crisis should be sought in all patients with MG, even when they do not complain of weakness. Many commonly used drugs can adversely affect patients with MG (see Box 106-2). A patient with stable MG who has an acute medical or surgical condition requires a full neurologic examination. The decision to admit or discharge a patient with MG from the emergency department should take into account the potential for neurologic deterioration. Management Cholinesterase Inhibitors. Pyridostigmine (60–120 mg every 4–6 hours) and neostigmine (15–30 mg every 4–6 hours) prolong the presence and activity of ACh in the synaptic cleft. They are the backbone of chronic outpatient therapy and provide symptomatic improvement. The most common side effects are those of excessive cholinergic stimulation, such as increased airway secretions and increased bowel motility. At extremes there may be bradycardia or even worsening of weakness, simulating a myasthenic crisis. These drugs are often used as adjunctive therapy to control symptoms while other therapy is being instituted, after which they are often discontinued.11 The use of intravenous pyridostigmine in the setting of acute exacerbation is controversial and not recommended because PE or IVIG is recommended. Cholinergic drug therapy is not recommended for the treatment of MG in the emergency department. Immunosuppressant Drugs. Immunosuppressant drugs are often used for the chronic control of MG. Although they have no role in the acute management of a myasthenic crisis, they may be started before extubation of a patient recovering from crisis. Cochrane Database reviews in 2005 and 2007 found support for the use of corticosteroids but only limited evidence that cyclosporine and cyclophosphamide and azathioprine improve MG.12,13 Of note, the initiation of corticosteroids in patients with moderate to severe weakness may actually precipitate a worsening of weakness or even myasthenic crisis. Thymectomy. While the association between thymoma and MG is not fully elaborated, it is well known that thymectomy for patients with thymoma can lead to remission of MG or enable a reduction in other medications. Thymectomy for patients with MG but without thymoma has been shown to have similar benefits and is recommended for patients younger than 60 with remission or improvement in up to 50% of cases and is supported by a clinical policy of the American Academy of Neurology.14 The onset of improvement after thymectomy is often delayed for 2 to 5 years. Immunomodulatory Therapy. PE and IVIG can be used for patients with exacerbations of MG or preoperatively in patients with stable MG. PE removes the AChR antibodies and other immune complexes from the blood. The fall in AChR levels is associated with improvement in symptoms of MG. There is a risk of complications from hypotension or anticoagulation. Because of safety concerns, clinical trials have not been done in children. Although there are no randomized controlled studies, a review yielded many case series with short-term benefit, especially in myasthenic crisis, and it is recommended by the American Academy of Neurology.15 A review of IVIG trials found one randomized controlled trial of IVIG versus placebo that demonstrated the benefit from IVIG. Another trial failed to show a difference between
Botulism Principles of Disease. Botulism is a toxin-mediated illness that can cause acute weakness leading to respiratory insufficiency. The Centers for Disease Control and Prevention (CDC) reports that an average of 145 cases are reported each year: 15% are foodborne, 65% are infant botulism, and 20% are wound related.17 Clostridium botulinum is an anaerobic, spore-forming bacterium. Three of eight known toxins produced by C. botulinum (types A, B, and E) cause human disease. There has been an increase in the incidence of botulism from wound infections. In a 4-day period in 2003 in Washington State, four people contracted wound botulism from black tar heroin.18 Botulism is also thought to be a potential agent for bioterrorism. The botulinum toxin works by binding irreversibly to the presynaptic membrane of peripheral and cranial nerves, inhibiting the release of ACh at the peripheral nerve synapse. As new receptors are generated, the patient improves. Clinical Features. The toxin blocks both voluntary motor and autonomic functions. Because the disorder is at the neuromuscular junction, there is no sensory deficit and no sense of pain. The onset of symptoms is 6 to 48 hours after the ingestion of tainted food. There may or may not be accompanying signs and symptoms of gastroenteritis, with nausea, vomiting, abdominal cramps, diarrhea, or constipation. The classic feature of botulism is a descending, symmetrical, flaccid paralysis. The muscles often affected first are the cranial nerves and bulbar muscles, and the patient presents with diplopia, dysarthria, and dysphagia, followed later by generalized weakness. There may be associated blurring of vision. Because the toxin decreases cholinergic output, anticholinergic signs may be seen in the form of constipation, urinary retention, dry skin and eyes, and increased temperature. Pupils are often dilated and not reactive to light. This can be a point of differentiation from MG. Deep tendon reflexes are normal or diminished. Infantile botulism results from the ingestion of C. botulinum spores that are able to germinate and produce toxin in the high pH of the gastrointestinal tract of infants. The same spores are not active in the gut of adults because of the lower pH. The CDC reports approximately 100 cases per year.19 It occurs in infants between the ages of 1 week and 11 months and has been implicated as a cause of sudden infant death syndrome. Because spores can survive in honey, it is recommended that honey not be fed to infants. The clinical presentation includes constipation, poor feeding, lethargy, and weak cry; consequently, this diagnosis must be in the differential diagnosis of the floppy infant. Diagnostic Strategies. The diagnosis is made by both clinical findings and exclusion of other processes. The toxin can be identified in serum and stool, but the assay is not commonly available in most hospitals and requires a prolonged turnaround time. If the suspected food source is available, it should also be tested for the toxin. Management. The treatment is initially focused on stabilizing the airway and supportive measures. There is an equine antitoxin that can shorten the disease course, although it is not clear that the antitoxin decreases ventilator dependence and there is a risk of anaphylaxis and serum sickness. Nevertheless, the antitoxin should be administered as soon as possible. The toxin is available through the CDC by calling (404) 3292888. An intravenous human botulism immune globulin (BIG-
1415
IV) has been developed for treatment of infantile botulism20 and is available through the California Department of Health Services Infant Botulism Treatment and Prevention Program on-call physician at (510) 231-7600.
Tick Paralysis Principles of Disease. The pathogenesis of tick paralysis, also known as tick toxicosis, is not fully understood. It is known that a toxin is injected while the tick feeds, and it is referred to as an ixovotoxin. The toxin appears to diminish the release of ACh at the neuromuscular junction and also reduces nerve conduction velocity. It may also have effects at autonomic ganglia, leading to pupillary signs. According to the CDC, the state of Colorado reports on average one case per year, though in 2006 four cases were reported during one week.21 Clinical Features. Tick paralysis is an acute, ascending, flaccid motor paralysis that can be confused with GBS, botulism, and MG. It typically begins with the development of an unsteady gait, followed by ascending, symmetrical, flaccid paralysis. Although symptoms usually begin 1 to 2 days after the female tick has attached and begun to feed, delays of up to 6 days have been reported.22 There may be associated ocular signs, such as fixed and dilated pupils, that can help distinguish it from GBS. Management. The management is supportive care and tick removal. A tick can be removed using forceps to grasp it as closely as possible to the point of attachment. Care must be taken not to leave mouth parts in the patient’s tissue. Although symptoms may resolve rapidly after removal of the tick, supportive measures such as intubation should not be withheld pending resolution of symptoms.
Disorders of the Muscles Perspective Newly acquired weakness originating at the muscular level can be divided into two types: inflammatory and toxicmetabolic. Inflammatory disorders usually produce pain and tenderness, but metabolic disorders do not.
Inflammatory Disorders Principles of Disease. The most common inflammatory myopathies are polymyositis (PM) and dermatomyositis (DM). PM may be idiopathic in nature, occur secondary to infections (viral or bacterial), or be seen in conjunction with other disorders such as sarcoidosis or hypereosinophilic syndromes. Inflammatory myopathies cause weakness, pain, and tenderness of the muscles involved. They must be distinguished from simple myalgias related to a fever or cramping that may suggest myotonia (inability to relax the muscle). Clinical Features. DM and PM can occur at any adult age, although DM may also affect children. There is a slightly increased incidence in women. An associated increased risk of malignancy, especially breast, ovary, lung, gastrointestinal, and lymphoproliferative disorders, has been noted after the diagnosis of DM or PM, although the reported rate of malignancy varies widely. Proximal muscle weakness predominates and leads to complaints of difficulty rising from a seated position or climbing stairs and weakness in lifting the arms over the head. There is often pain and tenderness in these proximal muscles as well. There is a decrease in reflexes as the weakened muscles fail to contract. Thus, the decrease in reflexes is in proportion to the decrease in strength. Fasciculations are not seen, and atrophy is a very late finding.
Chapter 106 / Neuromuscular Disorders
IVIG and PE.16 The decision to institute either therapy is based on the input of the consulting neurologist and the resources available at the admitting hospital. If PE is not readily available for a patient with myasthenic crisis, IVIG should begin with 1 g/kg.
PART III ■ Medicine and Surgery / Section Seven • Neurology
1416
DM is similar to PM, but it is also associated with classic skin findings. These are more prominent in childhood but are also found in adults. They include a periorbital heliotrope and erythema and swelling of the extensor surfaces of joints. The facial rash is usually photosensitive and may also involve the exposed areas of the chest and neck. Diagnostic Strategies Electrolyte abnormalities must be ruled out and the serum CK checked. If possible, the skeletal muscle isoform (MM) should be distinguished from the cardiac muscle isoform (MB). The CK must be interpreted in light of the entire clinical picture. The presence of an elevated CK does not establish the cause of weakness as a myopathy because some neuropathies can also produce an elevated CK. Similarly, a normal CK does not rule out a myopathy as the cause of weakness. Electromyography and muscle biopsy are used to confirm the diagnosis.23 Management. PM and DM are usually managed with oral prednisone in a dose of 1 to 2 mg/kg/day. When steroids prove ineffective and during acute exacerbations, cytotoxic drugs such as azathioprine or methotrexate are added. Fortunately, the degree of rhabdomyolysis seen with the inflammatory myopathies is not sufficient to cause renal impairment.
Metabolic Disorders Perspective. Acute, generalized muscle weakness can be seen with severe electrolyte abnormalities of any cause: hypokalemia, hyperkalemia, hypocalcemia, hypercalcemia, hypomagnesemia, and hypophosphatemia. Acute painless myopathies can also be seen with endocrine disorders involving the thyroid, parathyroid, or adrenal glands. Of particular interest are several disorders referred to collectively as the periodic paralyses. This group of entities includes familial periodic paralysis (FPP) of the hyperkalemic and hypokalemic forms and thyrotoxic periodic paralysis (TPP), which is similar to hypokalemic FPP except that it is associated with hyperthyroidism. Periodic Paralysis Principles of Disease. These are autosomal-dominant disorders of ion channels resulting in intermittent attacks of flaccid extremity weakness associated with either hyperkalemia or hypokalemia, although the latter is more common. It is most often associated with an inherited genetic mutation. Patients usually report a personal and family history of similar episodes.24 Clinical Features and Diagnostic Strategies. Patients may suffer either isolated or recurrent episodes of flaccid paralysis. The lower limbs are involved more often than the upper, although both can be affected. Bulbar, ocular, and respiratory muscles are usually not involved. Onset is rapid; a prodrome of myalgias and muscle cramps may occur but is uncommon; mental status and sensory function are typically preserved, but reports of sensory nerve involvement have been documented.25 Males are more often affected than females, and there is a higher incidence in Asians, particularly Japanese, although it occurs in other ethnic groups. Attacks may be induced by the injection of insulin, epinephrine, or glucose. The onset of symptoms often follows a high carbohydrate intake (with subsequent insulin rise) and a period of rest. A typical complaint is the acute onset of weakness noted on waking in the morning after a large meal the preceding evening. An electrocardiogram, which should be done immediately in all patients suffering from acute paralysis, demonstrates signs of hyperkalemia or hypokalemia. An imme-
diate potassium level should be ordered; in the hypokalemic form, the potassium level during an attack falls to values below 3.0 mEq/L. Management. Many cases resolve spontaneously with supportive care alone. The mainstay of management is the treatment of the underlying electrolyte imbalance. In the hypokalemic state the total body potassium is not depleted but has shifted intracellularly. Thus, in the repletion of potassium, caution is necessary to prevent overtreatment. For this reason, intravenous potassium should be used sparingly; one or two 10-mEq doses of potassium chloride (KCl), each administered over 1 hour, should be the maximum given intravenously. This can be done in parallel with 40 mEq oral potassium repletion and retesting of serum potassium levels. Intravenous hydration helps to redistribute the body’s potassium stores. Thyrotoxic Periodic Paralysis. The clinical picture of TPP is almost identical to that of hypokalemic FPP, and indeed a small number of patients with hypokalemic FPP have hyperthyroidism. In TPP, symptoms related to hyperthyroidism are often present at the same time the patient develops weakness. The relation of the hyperthyroidism to hypokalemia is probably due to increased sodium-potassium adenosine triphosphatase activity, which causes a rapid shift of potassium from the extracellular into the intracellular compartment. Treatment of the hyperthyroid symptoms, such as tachycardia, may help the treatment of the paralysis as well. There are case reports of TPP in which the patient’s weakness did not respond to potassium replacement until propranolol was given to treat tachycardia.26,27 There is probably a genetic feature underlying this disorder because there is a higher incidence of repeated attacks of hypokalemic periodic paralysis among Japanese and Chinese patients with hyperthyroidism. It is important that all patients have thyroid function testing done after a first episode of hypokalemic paralysis.
KEY CONCEPTS ■
The approach to evaluating patients with acute neuromuscular weakness is facilitated by first determining the location of the lesion (spinal cord, nerve, neuromuscular junction or muscle) and then considering the most common disorders that affect the area in question. ■ In patients with bilateral upper motor neuron signs and a normal mental status, neuroimaging of the spinal cord should be strongly considered. ■ In patients presenting with acute neuromuscular weakness, complaints of difficulty in breathing or swallowing should heighten suspicion of bulbar involvement with possible airway compromise. In such patients, a forced vital capacity less than 15 mL/ kg or a maximal NIF less than 15 mm Hg are potential indications for mechanical ventilation. ■ Botulism usually arises as a painless descending paralysis, often first affecting the cranial nerves and bulbar muscles, without sensory deficits or significant alteration of consciousness. The treatment is airway management and administration of antitoxin. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 107
Central Nervous System Infections
William J. Meurer and Frank W. Lavoie
■ PERSPECTIVE Background Central nervous system (CNS) infections have always been among the most perplexing and devastating illnesses. “Epidemic cerebrospinal fever,” classically described by Viesseux in 1805, was associated with almost universal mortality.1 The first American epidemic of meningococcal meningitis was recorded in 1806.2 Since that time, epidemiologic changes have occurred in concert with advances in understanding of disease processes and evolution of effective treatment strategies. The etiologic spectrum of CNS infection has changed considerably as a result of the development and aggressive use of antibiotics and the epidemic emergence of immune disorders such as infection with the human immunodeficiency virus (HIV). Some of the research on CNS infections has markedly increased in sophistication, which provides insights into pathogenesis, including the role of host mechanisms such as cytokines and other immune components. The pathophysiologic alterations are increasingly understood at the cellular and molecular levels. Likewise, diagnostic tools have been developed that allow precise pathogen identification, most recently using molecular technologies such as polymerase chain reaction (PCR) tests for viral nucleic acids in cerebrospinal fluid (CSF). The initial treatment methodologies began by demonstrating the efficacy of antiserum treatment by Flexner in 1913 and of antibiotics by Colebrook and Kenny in 1936.3,4 The mortality rates were decreased further with the use of high-dose penicillin by Dowling and colleagues in the 1940s.5 Unfortunately, despite historical advances, the morbidity and mortality of these disorders remain considerable, although substantial progress has been made.6 The use of pneumococcal Haemophilus influenzae type b (Hib) and meningococcal vaccines has led to dramatic changes in the incidence of meningitis caused by these bacteria.7–14
Definitions CNS infections comprise a broad spectrum of disease entities. Meningitis is defined as inflammation of the membranes of the brain or spinal cord and is also called arachnoiditis or leptomeningitis. Encephalitis denotes inflammation of the brain itself, whereas myelitis refers to inflammation of the spinal cord. The terms meningoencephalitis and encephalomyelitis describe more diffuse inflammatory processes. Collections of infective and
purulent materials may form within the CNS as abscesses. Brain abscesses may be intraparenchymal, epidural, or subdural, or may be found in intramedullary or epidural spinal locations. This chapter focuses on the more common acute and subacute CNS infections. Infections of the nervous system with HIV or human T lymphotrophic virus, rabies virus, polio or hepatitis viruses, Borrelia burgdorferi (Lyme disease), Treponema organisms (syphilis), parasites, Rickettsia, and the chronic and slow infections of the CNS (subacute sclerosing panencephalitis, progressive multifocal leukoencephalopathy, and the prion-mediated spongiform encephalopathies, such as Creutzfeldt-Jakob disease, bovine spongiform encephalopathy, and kuru) are not addressed in detail. Of note, the incidence of neurocysticercosis is on the rise in the United States.15
Epidemiology Bacterial meningitis is a common disease worldwide. Meningococcal meningitis is endemic in parts of Africa, and epidemics commonly occur in other countries, including the United States. A variety of other pathogens are also causative.16–20 The overall incidence of bacterial meningitis in the United States is 5 to 10 cases per 100,000 people per year.21 Men are affected more often than women.21 In the United States, approximately 80% of cases are caused by either Streptococcus pneumoniae or Neisseria meningitidis.22 In regions where vaccination is common, the epidemiology of bacterial meningitis has substantially changed.9–11,14,23 The incidence of bacterial meningitis increases in late winter and early spring, but the disease may occur at any time of the year. Because most cases are unreported, the actual incidence of viral meningitis is unknown. It is estimated to affect 11 to 27 individuals per 100,000 people.24 A prominent increase of cases is seen in summer months, which is concurrent with seasonal predominance of the enterovirus group of the picornaviruses. The same organisms responsible for viral meningitis may also be associated with encephalitis. Encephalitis is, however, far less common, and the ratio of cases of meningitis to encephalitis varies according to the specific pathogen. Arbovirus infection is transmitted by an insect vector, although clinical disease develops in only a small percentage of the people bitten. Before 1999, approximately 19,000 cases of encephalitis were hospitalized in the United States annually. Since then, there has been a rapid increase because of the emergence of 1417
PART III ■ Medicine and Surgery / Section Seven • Neurology
1418
West Nile virus (WNV). In 2003, more than 8000 additional individuals were hospitalized because of WNV alone.25,26 Approximately 2000 cases of brain abscess occur in the United States annually.27 Although CNS abscesses may occur at any age and any time of year, they are more commonly seen in men than women.28,29 CNS abscesses are associated with local contiguous and remote systemic infections, intravenous (IV) drug use, neurologic surgery, and cranial trauma. Brain abscess secondary to otitis media most often occurs in pediatric or older adult populations. When associated with sinusitis, it most often arises among young adults. Increasingly, CNS abscesses are seen in the immunocompromised population, particularly those with HIV infection, and among bone marrow and solid organ transplant recipients. However, antimicrobial prophylaxis of immunosuppressed patients and more aggressive treatment of otitis and sinusitis have decreased the overall incidence to 0.9 per 100,000 person-years.27
■ PRINCIPLES OF DISEASE Etiology Meningitis Meningeal inflammation may be caused by a variety of disease processes, but the infectious etiologies predominate. Among the bacterial etiologies, Streptococcus pneumoniae remains the predominant pathogen in adult patients, followed by N. meningitidis and Listeria monocytogenes.30,31 N. meningitidis is the predominant organism in adults younger than 45 years. Five major serogroups cause most meningococcal disease worldwide (A, B, C, Y, and W-135). Serogroup A accounts for the majority of cases of meningococcal meningitis in developing nations.32 Serogroup distribution for invasive disease has changed markedly in the United States, with B, C, and Y now most commonly responsible.33–36 These pathogens account for the bulk of cases in nontraumatic meningitis, although virtually any organism can be encountered, particularly among patients who are elderly, alcoholic, and immunosuppressed and those who have cancer. Interestingly, higher case fatality has been observed in N. meningitis outbreaks versus sporadic cases, likely due to increased virulence of outbreak related strains.37 Causes of aseptic meningitis, which simply defined is all cases with negative bacterial CSF cultures, are listed in Box 107-1.38 Meningeal infection may also occur in association with a dural leak secondary to neurosurgery or neurotrauma. S. pneumoniae, Staphylococcus aureus, Pseudomonas aeruginosa, and coliform bacteria are seen most commonly in this population. Viral meningitis may likewise be caused by a variety of etiologic agents.39 Enteroviruses are statistically encountered most commonly.40 Unfortunately, precise definition of the etiologic agent is often impossible. Fungal and parasitic meningitides are additional concerns, particularly among immunocompromised patients.18,19 Noninfectious meningitides include drug-induced meningitis, carcinomatous meningitis, CNS involvement in serum sickness, vasculitis, systemic lupus erythematosus, Behçet’s disease, sarcoidosis, and others. The differentiation of noninfectious from infectious etiologies can often be perplexing.
adults are also commonly affected. Epidemics of viral encephalitis have been attributed to a wide variety of viral agents. WNV, a flavivirus, first infected humans in the New York City area and rapidly spread to 47 states by 2003.19,41 Varicella, herpes zoster, HHV 6 and 7, and Epstein-Barr virus have been increasingly reported to be the cause of encephalitis in immunocompetent hosts.42,43 Vaccinia encephalitis has been recognized in those receiving vaccination for smallpox.44 Postinfectious encephalomyelitis is also induced by a variety of viral pathogens, most commonly by the measles virus.45 However, Mycoplasma pneumoniae and idiopathic causes are becoming more common in developed countries.
Central Nervous System Abscess The etiologies of CNS abscess are multiple and reflect the primary infective process and the immune state of the human host. A variety of mixed pathogens may be responsible for intracranial abscesses. Streptococci, particularly the Streptococcus milleri group, have been identified in nearly 50% of brain abscesses.46 Anaerobic bacteria, predominantly Bacteroides species, are commonly seen when the primary infectious process is chronic otitis media or pulmonary disease. S. aureus and Propionbacterium acnes are often identified, particularly after cranial penetration from surgery or trauma.47,48 The Enterobacteriaceae are an additional common isolate. Opportunistic fungal and parasitic etiologies are often seen in the immunosuppressed, including Nocardia species.46,49 Culture of epidural and subdural abscesses more often yields a single organism, with streptococci most commonly seen when associated with contiguous spread and S. aureus and gram-negative rods most commonly encountered after neurologic trauma.19 Etiologic agents in spinal abscess are similarly varied. S. aureus is most commonly encountered (Fig. 107-1).
Encephalitis Arboviruses and herpes simplex virus (HSV), a human herpes virus (HHV), are the most common causes of epidemic and sporadic cases of encephalitis, respectively. Children are the most vulnerable to infection with these viruses, although
Figure 107-1. Central nervous system abscess: computed tomography scan of an intraparenchymal abscess (arrows).
1419
BOX 107-1 Causes of Aseptic Meningitis II. Noninfectious causes 1. Postinfectious/postvaccinial Rubella Varicella Variola Rabies vaccine Pertussis vaccine Influenza vaccine Vaccinia Yellow fever vaccine 2. Drugs Nonsteroidal anti-inflammatory drugs Trimethoprim-sulfamethoxazole, amoxicillin Muromunab CD3 (OKT3) Azathioprine Intravenous immunoglobulin Isoniazid Intrathecal methotrexate Intrathecal cytosine arabinoside Allopurinol Carbamazepine Sulfasalazine 3. Systemic disease Collagen vascular disorders Systemic lupus erythematosus Wegener’s granulomatosis Central nervous system vasculitis Rheumatoid arthritis Kawasaki’s disease Sarcoidosis Leptomeningeal cancer Posttransplantation lymphoproliferative disorder Behçet’s disease Vogt-Koyabagj syndrome 4. Neoplastic disorders Leukemia Carcinomatous meningitis secondary to primary or secondary tumors of the brain 5. Inflammation of neighboring structures Brain abscess Epidural abscess 6. Miscellaneous Arachnoiditis Migraine Urinary tract infection
Reproduced from Kumar R: Aseptic meningitis: Diagnosis and management. Ind J Pediatr 72:57, 2005.
Pathophysiology Bacterial Meningitis The pathogenetic sequence in bacterial meningitis has been well characterized.18,19,50,51 The first step is nasopharyngeal colonization and mucosal invasion. Although colonization rates vary, virulent microbes use secretion of immunoglobulin A proteases and induce cilio-stasis of mucosal cells. After penetration occurs by a variety of mechanisms, bacterial intravascular survival occurs because of evasion of the complement pathway. The varying capsular properties of each organism protect the bacteria. The third step occurs when the bacteria cross the blood-brain barrier to enter the CSF. The dural venous sinuses, cribriform plate area, and choroid plexus have
all been implicated as potential sites of invasion. Although the mechanism of invasion is not completely understood, host defense mechanisms within the CSF are often ineffective; there are low levels of complement, immunoglobulin, and opsonic activity. Bacterial proliferation then occurs, which stimulates a convergence of leukocytes into the CSF. Meningeal and subarachnoid space inflammation is also associated with the release of cytokines into the CSF, most notably tumor necrosis factor and interleukins 1 and 6.50,52 This results in increased permeability of the blood-brain barrier, cerebral vasculitis, edema, and increased intracranial pressure (ICP). A subsequent decrease in cerebral blood flow leads to cerebral hypoxia. Glucose transport into the CSF is decreased concomitantly with an increased use of glucose by the brain,
Chapter 107 / Central Nervous System Infections
I. Infectious causes 1. Viruses Enteroviruses—polio, coxsackie, ECHO virus Herpes group of viruses Herpes simplex virus types 1 and 2 Varicella zoster virus Cytomegalovirus Epstein-Barr virus Human herpes virus 6 Respiratory viruses Adenovirus Rhino virus Influenza virus types A and B Arboviruses Mumps virus Lymphocytic choreomeningitis HIV 2. Bacteria Partially treated meningitis Parameningeal infection Endocarditis Mycoplasma pneumoniae Mycobacterium tuberculosis Ehrlichiosis Borrelia burgdorferi Treponema pallidum Brucella Leptospirosis 3. Fungi Cryptococcus neoformans Histoplasma capsulatum Coccidioides immitis Blastomyces dermatitides Candida 4. Parasites Toxoplasma gondii Neurocysticercosis Trichinosis Naegleria Hartmannella Bartonella henselae 5. Rickettsiae Rocky Mountain spotted fever Typhus
PART III ■ Medicine and Surgery / Section Seven • Neurology
1420
bacteria, and leukocytes, which depresses CSF glucose concentrations. The increased permeability leads to increased CSF proteins.
Viral Meningitis and Encephalitis Viruses enter the human host through the skin (i.e., insect vectors), through the respiratory, gastrointestinal, or urogenital tract, or by receipt of infected blood products or donor organs.53,54 Viral replication subsequently occurs outside the CNS, most often followed by hematogenous spread to the CNS. Additional routes into the CNS include retrograde transmission along neuronal axons and direct invasion of the subarachnoid space after infection of the olfactory submucosa.55,56 Fortunately, most systemic viral infections do not result in meningitis or encephalitis. The development and subsequent magnitude of viral infection depend on the virulence of the specific virus, the viral inoculum level, and the state of immunity of the human host. The tropism of the virus for specific CNS cell types also influences the focality of disease and its manifestations.55 Particular viruses may preferentially attack cortical, limbic, or spinal neurons, oligodendria, or ependymal cells. An example is the tropism of HSV for the temporal lobes and the development of temporal lobe seizures and behavioral changes in afflicted patients.
Fungal Meningitis Fungal meningitis probably develops in much the same way as bacterial meningitis, although this has been incompletely studied. Pulmonary exposure followed by hematogenous spread is the primary pathogenetic mechanism in most cases. Immune system defects or immunosuppressive drugs compromise host defense mechanisms, with ensuing development of CNS infection.
Central Nervous System Abscess Intraparenchymal brain abscesses, subdural empyema, or intracranial or spinal epidural abscesses form by inoculation of the CNS from contiguous spread of organisms from a sinus, middle ear, or dental infection or metastatic seeding from a distant site, usually from pulmonary infection, endocarditis, or osteomyelitis.28,47 The primary infection can be identified in 75 to 85% of cases. These conditions may also follow surgery or penetrating cranial trauma, particularly when bone fragments are retained in brain tissue. Otogenic abscesses occur most commonly in the temporal lobe in adults and cerebellum in children, whereas sinogenic abscesses typically occur in frontal areas.46 Multiple brain abscesses suggest hematogenous spread of organisms, although solitary lesions may also occur. The pulmonary system is the most common source of hematogenous spread.19
■ CLINICAL FEATURES Symptoms and Signs Numerous host factors have been implicated in the acquisition of meningitis (Box 107-2).57 Although these factors alone and in combination increase the risk of meningitis, the disease often occurs in patients with none of these factors. Many patients with meningitis present with advanced disease; in these patients, the diagnosis of acute meningitis is strongly suspected. The constellation of symptoms that may classically occur in an acute CNS infection consists of fever,
BOX 107-2 Host Factors Predisposing to Meningitis Age 60 yr Male gender Low socioeconomic status Crowding (e.g., military recruits) Splenectomy Sickle cell disease Black race Alcoholism and cirrhosis Diabetes Immunologic defects Recent colonization Dural defect (e.g., traumatic, surgical, congenital) Continuous infection (e.g., sinusitis) Household contact with meningitis patient Thalassemia major Intravenous drug abuse Bacterial endocarditis Ventriculoperitoneal shunt Malignancy
headache, photophobia, nuchal rigidity, lethargy, malaise, altered sensorium, seizures, vomiting, and chills.17,57 Unfortunately, more subtle presentations are also common. Immunosuppressed and geriatric patients present a diagnostic challenge because the classical signs and symptoms of meningitis may not be present. Although some degree of fever is present in most patients, as are headache and neck stiffness, meningitis should be carefully considered in any immunosuppressed patient with symptoms or signs of infectious disease. Often, the only presenting sign of meningitis in the elderly patient is an alteration of mental status. However, a metaanalysis suggested that the absence of fever, stiff neck, and mental status change excludes meningitis in immunocompetent adults.36 The presentation of fungal meningitis can be obscure even in the healthy adult population. Headache, low-grade fever, lassitude, and weight loss may be present but often to such a mild degree that the correct diagnosis is not initially considered.17 This is also true of tuberculous meningitis, which often has a protracted course and a vague nonspecific presentation consisting of fever, weight loss, night sweats, and malaise, with or without headache and meningismus.16 The physical findings in meningitis vary, depending on the host, causative organism, and severity of the illness. Nuchal rigidity or discomfort on flexion of the neck is common. Kernig’s and Brudzinski’s signs are present in approximately 50% of adults.19 Described in 1882 by Vladimir Kernig, Kernig’s sign is present in the patient if the examiner is unable, because of resistance and hamstring pain, to straighten the patient’s leg passively to a position of full knee extension when the patient is lying supine with the hip flexed to a right angle. Jozef Brudzinski initially described five signs, two of which are currently utilized.2 The contralateral sign is present if an attempt to flex the hip passively on one side is accompanied by a similar movement of the other leg. The neck sign is present if attempts to flex the neck passively are accompanied by flexion of the hips. The absence of jolt accentuation of headache with this maneuver may be useful in obviating the need for lumbar puncture (LP) in a patient with low suspicion for meningitis.58 Deep tendon reflexes may be increased, and ophthalmoplegia may be present, especially of the lateral rectus muscles.
Complications Bacterial Meningitis The immediate complications of bacterial meningitis include coma (with loss of protective airway reflexes), seizures, cerebral edema, vasomotor collapse, disseminated intravascular coagulation, respiratory arrest, dehydration, syndrome of inappropriate secretion of antidiuretic hormone, pericardial effusion, and death (Box 107-3).20 Various delayed complications include multiple seizures, focal paralysis, subdural effusions, hydrocephalus, intellectual deficits, sensorineural hearing loss, ataxia, blindness, bilateral adrenal hemorrhage (WaterhouseFriderichsen syndrome), peripheral gangrene, and death.57 The case fatality rate for pneumococcal meningitis averages 20 to 25%, with higher fatality rates occurring in patients with serious underlying or concomitant disease or advanced age.61,62 The prognosis is related to the degree of neurologic impairment on presentation. Overall, 20 to 30% of the survivors of pneumococcal meningitis have some residual neurologic
1421
BOX 107-3 Complications of Bacterial Meningitis Immediate Coma Loss of airway reflexes Seizures Cerebral edema Vasomotor collapse Disseminated intravascular coagulation (DIC) Respiratory arrest Dehydration Pericardial effusion Death Others Delayed Seizure disorder Focal paralysis Subdural effusion Hydrocephalus Intellectual deficits Sensorineural hearing loss Ataxia Blindness Bilateral adrenal hemorrhage Death Others deficit.57 The case fatality rate for Listeria meningitis may be as high as 40%.31 With the advent of antibiotic therapy, the mortality from meningococcal meningitis has markedly decreased to less than 20%, but it remains substantially higher in elderly patients or in those who also have meningococcemia.62 Although most of the complications and sequelae are less common than with pneumococcal disease, the incidence of WaterhouseFriderichsen syndrome is dramatically higher when meningococcemia is present.57 The overall mortality rate in community-acquired gram-negative meningitis has been less than 20% since the introduction of the third-generation cephalosporins.18
Viral Meningitis With rare exceptions, the overall prognosis for complete recovery from viral meningitis is excellent. Various complications related to the systemic effects of the particular virus include orchitis, parotitis, pancreatitis, and various dermatoses. Usually all of these complications resolve without sequelae.39
Viral Encephalitis The outcomes in viral encephalitis are dependent on the infecting agent. Encephalitis caused by Japanese encephalitis virus, Eastern equine virus, and St. Louis encephalitis virus is severe, with high mortality rates and virtually universal neurologic sequelae among survivors.63 WNV produces encephalitis in only 0.5% of those infected, yet it resulted in 120 deaths in 2003.26 Western equine virus and California encephalitis virus cause milder infections, and death is rare. The incidence of neurologic sequelae is highly variable and appears to depend on both the host and the infecting agent.63,64 The mortality rate from HSV encephalitis before the use of acyclovir was 60 to 70%. Acyclovir treatment has reduced the mortality rate to approximately 30%.45 Common sequelae observed among survivors include seizure disorders, motor deficits, and changes in mentation.
Chapter 107 / Central Nervous System Infections
The systemic findings may include an obvious source of infection such as sinusitis, otitis media, mastoiditis, pneumonia, or urinary tract infection. Various manifestations of endocarditis may be present. Arthritis may be seen with N. meningitidis and occasionally with other bacteria.57 Petechiae and cutaneous hemorrhages are widely reported with meningococcemia but also occur with Hib, pneumococcal organisms, L. monocytogenes, and echovirus infections, in addition to staphylococcal endocarditis.57 Endotoxic shock with vascular collapse often develops in severe meningococcal disease, but shock may be present in the advanced stages of any bacterial meningitis. Any determination of a serious systemic infection should encourage rather than dissuade the clinician from considering the possibility of a concomitant CNS infection. Patients with encephalitis may also have symptoms of meningeal irritation. An alteration of consciousness occurs in virtually all patients. Fever, headache, and a change of personality are also usually present.56 Hallucinations and bizarre behavior may precede motor, reflex, and other neurologic manifestations by several days, occasionally prompting an initial diagnosis of a psychiatric disorder. Because focal neurologic deficits and seizures occur much more commonly with encephalitis than meningitis, there may also be diagnostic confusion with a brain abscess. Distinguishing the etiologic agent in encephalitis is clinically difficult, although HSV encephalitis results in a higher incidence of dysphasia and seizures.59 In some patients, WNV produces a myelitis that affects the anterior horn cells of the spinal column, resulting in a flaccid paralysis with a clear sensorium, similar to findings in polio or Guillain-Barré syndrome.41 Patients with intracranial abscess may be indistinguishable from those with meningitis or encephalitis. Most patients with intraparenchymal abscess have a subacute course of illness, with symptoms progressing during the course of 2 or more weeks. However, nuchal rigidity and fever are present in fewer than 50% of cases. Focal neurologic deficits are present in most of these patients. A large number of patients exhibit papilledema, which is a rare finding in meningitis. An abrupt neurologic deterioration that results from uncal herniation or rupture into the ventricular system may occur. Patients with a subdural or epidural abscess most often have headache, fever, and focal signs, although more subtle presentations are common. Most of the patients with spinal abscess typically present with spinal pain and other symptoms and signs of cord compression but not necessarily with fever.60
1422
PART III ■ Medicine and Surgery / Section Seven • Neurology
Tuberculous Meningitis Death from tuberculous meningitis in the adult age group ranges from 10 to 50% of cases, with the incidence directly proportional to the patient’s age and the duration of symptoms before presentation. Focal ischemic stroke may result from the associated cerebral vasculitis. In advanced disease, up to 25% of patients may require some neurosurgical procedure for obstruction (ventriculoperitoneal shunt or drainage).65 In most patients some neurologic deficit develops, but severe longterm sequelae among survivors are unusual.16,65
Fungal Meningitis Common CNS complications with fungal meningitis include abscesses, papilledema, neurologic deficits, seizures, bone invasion, and fluid collections. Direct invasion of the optic nerve results in ocular abnormalities in up to 40% of patients with cryptococcal meningitis.17 The mortality rate is high but variable and is related to the timeliness of diagnosis, underlying illness, and therapeutic regimens.
Central Nervous System Abscess With the early diagnosis afforded by the use of the cranial computed tomography (CT) scan; appropriate antimicrobial therapy; and combined management approaches with surgery, aspiration, and medical therapy, the mortality rate from brain abscess has declined dramatically from approximately 50% to less than 20%.28,66 Seizure disorder is the most common sequela of intracranial abscess, occurring in 80% of patients.18 Other neurologic sequelae of intracranial abscesses, including focal motor or sensory deficits or changes in mentation, are common. Complications of spinal abscess primarily result from cord compression, including paralysis, motor and sensory deficits, and bowel and bladder dysfunction. Generalized spread of CNS infection and death may also occur.60
■ DIAGNOSTIC STRATEGIES Lumbar Puncture General Considerations Because the consequences of missing a CNS infection are devastating, CNS infection must be presumed to be present until excluded. The possibility of the diagnosis of meningitis mandates LP unless the procedure is contraindicated by the presence of infection in the skin or soft tissues at the puncture site or the likelihood of brain herniation.45 Adherence to this principle prevents a delay in diagnosis, which substantially increases the morbidity and mortality of the disease. Some patients have clinically obvious bacterial meningitis, and CSF examination serves primarily to help identify the organism, thereby facilitating the appropriate treatment. Most patients, however, present more of a diagnostic problem, and analysis of the CSF fluid constitutes the critical step in the elucidation of the presence of CNS infection.
Increased Intracranial Pressure In most patients with bacterial meningitis, LP may be safely performed without antecedent neuroimaging studies. As this may not be the case in other brain pathologies, in many circumstances it is advisable to obtain a CT scan of the head before performing an LP.67,68 These indications must be carefully weighed against the patient’s condition, the probability
of meningitis, and the availability of the CT or magnetic resonance imaging (MRI) scan.18 It has been conventionally asserted that an LP in the presence of increased ICP may be harmful or fatal to the patient. Although data to address this concern are limited, the presence of focal neurologic signs does appear to be associated with a dramatic increase in complications from LP. These patients may deteriorate precipitously during or after the procedure.69–72 Patients with a markedly depressed sensorium that precludes careful neurologic examination or those with a focal neurologic deficit, papilledema, seizures, or evidence of head trauma must be considered to be at risk for a herniation syndrome that may be exacerbated by an LP. If the presentation is an acute, fulminating, febrile illness and bacterial meningitis is the concerning diagnosis, early initiation of antimicrobial therapy is mandatory because of the association of prognosis and time to treatment.73 The algorithmic alternatives are therefore (1) immediate LP followed by initiation of antibiotic treatment before obtaining the results or (2) initiation of antibiotic treatment followed by a cranial CT scan and then an LP. The latter choice of empirical treatment with antibiotics is now the routine in many institutions, although in some cases a third option could be considered: antibiotics and no LP despite an unremarkable CT scan.72 This reflects the efficacy of current methodologies of identification of causative organisms by means other than bacteriologic cultures. The controversy emerging regarding not performing LP despite a lack of CT scan findings is based on some reviews and case reports. These describe a fulminant herniation syndrome temporally related to LP in patients with normal CT scans.70 Increased ICP may not be reliably detected using CT. Clinical signs of increased ICP, rapid change in consciousness, and recent seizures were identified as risk factors predicting deterioration despite a normal CT scan.72 The risks of ongoing empirical treatment with antibiotics without additional information from CSF analysis appears to be low, as the yield from blood cultures and other diagnostic techniques such as PCR is relatively high. Therefore, this risk may be less than the risks of performing LP in certain very high-risk patients.
Cerebrospinal Fluid Analysis Opening Pressure The normal CSF pressure in an adult varies from 50 to 200 mm H2O. This value applies only to patients in the lateral recumbent position and may increase substantially when the patient is in the sitting position. The pressure is often elevated in bacterial, tuberculous, and fungal meningitides and a variety of noninfectious processes.56 Pressure may be falsely elevated when the patient is tense or obese or has marked muscle contraction.
Collection of Fluid At least three sterile tubes each containing at least 1 to 1.5 mL of CSF should be obtained and numbered in sequence. A fourth tube may be desirable should later studies such as viral cultures or a Venereal Disease Research Laboratories (VDRL) test for syphilis become necessary. The fluid should be sent to the laboratory for immediate analysis of turbidity, xanthochromia, glucose, protein, cell count and differential, Gram’s stain, bacterial culture, and antigen testing (Table 107-1). In certain cases, an India ink stain, bacteriologic stain for acid-fast bacilli, or VDRL test should be obtained. When only a small amount of fluid can be obtained, the most important studies
1423
Table 107-1 Analysis of Cerebrospinal Fluid NORMAL VALUE
SIGNIFICANCE OF ABNORMALITY
Cell count
Increased WBC counts are seen in all types of meningitis and encephalitis; increased PMN count suggests bacterial pathogen
Gram’s stain
≤5 WBC/mm3 ≤1 PMN/mm3 ≤1 eosinophil/mm3 No organism
Turbidity
Clear
Xanthochromia
None
CSF-to-serum glucose ratio Protein
0.6 : 1 15–45 mg/dL
India ink stain Cryptococcal antigen Lactic acid Bacterial antigen tests Acid-fast stain
Negative Negative ≤35 mg/dL Negative Negative
Offending organism identified 80% of time in bacterial meningitis, 60% if patient pretreated Increased turbidity with leukocytosis, blood, or high concentration of microorganisms Presence of RBCs in spinal fluid for 4 hr before lumbar puncture; occasionally caused by traumatic tap (if protein ≥150 mg/dL) or hypercarotenemia Depressed in pyogenic meningitis or hyperglycemia; lag time if glucose given IV Elevated with acute bacterial or fungal meningitis; also elevated with vasculitis, syphilis, encephalitis, neoplasms, and demyelination syndromes Positive in one third of cases of cryptococcal meningitis 90% accuracy for cryptococcal disease Elevated in bacterial and tubercular meningitis ≥95% specific for organism tested; up to 50% false-negative rate Positive in 80% of cases of tuberculous meningitis if ≥10 mL of fluid
CSF, cerebrospinal fluid; PMN, polymorphonuclear; RBC, red blood cell; WBC, white blood cell.
are the cell count with differential, Gram’s stain, and bacterial cultures. Ideally, the cell count should be performed on both the first and third or fourth tubes to help differentiate true CSF pleocytosis from contamination of the specimen by a traumatic LP.
Turbidity The CSF should be assessed immediately for turbidity or cloudiness by the person performing the LP. Because normal CSF is completely clear and colorless and should be indistinguishable from water, any degree of turbidity is pathologic. Leukocytosis is the most common cause of CSF turbidity; counts greater than 200 cells/mm3 usually cause clinically detectable changes in CSF clarity.74
Cell Count and Differential Normal adult CSF contains no more than 5 leukocytes/mm3 with at most one granulocyte (polymorphonuclear [PMN] leukocyte)57,74,75; therefore, the presence of more than one PMN or a total cell count of more than 5 cells/mm3 should be considered evidence of CNS infection. In addition, the presence of any eosinophil in the CSF is abnormal, although occasionally basophils may be seen in the absence of disease.74 Pretreatment with a few doses of antibiotics, although possibly diminishing the yield of Gram’s staining and cultures, should not affect the CSF cell counts in meningitis.18,30,76,77 The cell counts in bacterial meningitis are usually markedly elevated, sometimes exceeding 10,000 cells/mm3, and demonstrate a dramatic granulocytic shift.57 In general, counts exceed 500 cells/mm3, with a preponderance of PMN leukocytes. However, the initial CSF analysis exhibits lymphocytosis (lymphocyte count >50%) in 6 to 13% of all cases of bacterial meningitis. When only the patients with bacterial meningitis with fewer than 1000 cells/mm3 are considered, 24 to 32% have a predominance of lymphocytes.78,79 In addition, the same population of patients often has only a mild disturbance of CSF glucose and protein levels. In well-established viral meningitis and encephalitis, counts are usually less than 500 cells/ mm3, with nearly 100% of the cells being mononuclear.40 Early
(1000 mg/dL) in the presence of a relatively benign clinical presentation should suggest fungal disease.17
Xanthochromia
India Ink Preparation
Xanthochromia refers to the yellowish discoloration of the supernatant of a centrifuged CSF specimen. Xanthochromia is abnormal and results from the lysis of RBCs and release of the breakdown pigments oxyhemoglobin, bilirubin, and methemoglobin into the CSF. This process normally begins within 2 hours, and pigments may persist up to 30 days; therefore, early analysis of the LP specimen is essential. If a traumatic tap has introduced enough plasma to raise the CSF protein level to 150 mg/dL or more, blood pigments may cause xanthochromia. If the CSF protein level is less than 150 mg/dL, however, and systemic hypercarotenemia does not exist, xanthochromia of a centrifuged CSF specimen should suggest that subarachnoid hemorrhage has occurred.74
India ink staining of the CSF should be performed when a diagnosis of cryptococcal meningitis is being considered. The demonstration of budding organisms (Fig. 107-2) is virtually diagnostic for cryptococcal disease but occurs in only one third of the cases.17 A more definitive diagnostic test is the cryptococcal antigen.
Glucose When the serum glucose is normal, the CSF glucose is usually between 50 and 80 mg/dL. The CSF glucose is normally in a ratio of 0.6 : 1 to the serum glucose, except with marked systemic hyperglycemia, when the ratio is closer to 0.4 : 1. Therefore, a CSF-to-serum glucose ratio of less than 0.5 in normoglycemic subjects or 0.3 in hyperglycemic subjects is abnormal and may represent the impaired glucose transport mechanisms and increased CNS glucose use associated with pyogenic meningitis.57,74 Mild decreases in the CSF glucose level may occur with certain viral and parameningeal processes. However, bacterial or fungal meningitis should be presumed to be the cause of low CSF glucose, termed hypoglycorrhachia, until each is clearly excluded.81 If the serum glucose level has increased rapidly—for example, after IV administration of 50% dextrose in water—equilibration in the CSF may take up to 4 hours, and therefore the interpretation of CSF-to-serum glucose ratios may be unreliable.
Protein The normal CSF protein level in adults ranges from 15 to 45 mg/dL. An elevated CSF protein, usually higher than 150 mg/dL, commonly occurs with acute bacterial meningitis.57 When a traumatic LP has occurred, the CSF protein can be corrected for the presence of blood by subtracting 1 mg/dL of protein for each 1000 RBCs.74 Elevated CSF protein concen-
Lactic Acid Although nonspecific, elevations in CSF lactic acid concentrations (>35 mg/dL) are potentially indicative of bacterial meningitis, and lactate may rise prior to the decline in glucose.82,83 Normal lactate levels ( 40 yr Sudden onset Physical examination often abnormal Vital signs may be abnormal Social immodesty Aphasia Consciousness impaired
Remote impairment Repetitive activity Posturing Rocking Auditory hallucinations Flat affect Oriented Continuous scattered thoughts Unfiltered perceptions Unable to attend Unable to focus Age < 40 yr Gradual onset Physical examination normal Vital signs usually normal Social modesty Intelligible speech Awake and alert
Modified from Frame DS, Kercher EE: Acute psychosis: Functional vs. organic. Emerg Med Clin North Am 9:123, 1991.
individuals initially given the diagnosis of schizophreniform disorder recover within 6 months, the other two thirds retain their symptoms and are diagnosed as having schizophrenia. Patients with mood disorders may develop psychotic symptoms. If these symptoms are present only during periods of mood disturbance, the diagnosis of mood disorder with psychotic features is applied; if they persist longer than 2 weeks in the absence of prominent mood symptoms, the diagnosis of schizoaffective disorder is made. Patients with personality disorders may occasionally develop brief psychotic episodes under stress. Ganser’s syndrome is a symptom complex considered to be emotional in origin in which the patient may appear to have amnesia, hallucinations, or alterations in consciousness, usually in association with physical complaints. An individual with Ganser’s syndrome may have psychotic symptoms for no apparent gain except to assume the role of the psychiatric patient. Persons with a delusional disorder experience nonbizarre delusions that may dominate their lives. They may believe that famous people are in love with them (erotomanic type), that they have extraordinary powers with a special relationship to a deity or a famous person (grandiose type), that their sexual partner is unfaithful (jealous type), that they are being malevolently treated in some way (persecutory type), or that they have some physical defect or general medical condition (somatic type). Although patients with somatic delusions may experience tactile or olfactory hallucinations related to the delusional theme (e.g., the sensation of being infested with insects), the other features associated with schizophrenia are not present.
■ MANAGEMENT General Approach Patients with thought disorders may be agitated and hyperactive, may be withdrawn but hypervigilant, or may complain of somatic delusions. In addition, they may have paranoid ide-
ation, may be angry they have been brought to the emergency department against their will, or may be frightened because they have been confronted by the police, restrained, and isolated. The presence of such patients in the emergency department may be disconcerting to staff because these patients are often irrational, erratic, and unpredictable in their behavior. Although emergency personnel must remain calm, empathetic, and reassuring in their interactions with patients exhibiting a thought disorder, they must also take steps to ensure staff safety whenever dealing with patients at risk for sudden violence. Such patients include those who have manifested violent behavior before coming to the emergency department, those who physically or verbally threaten staff, and those who demonstrate an escalating level of agitation despite verbal attempts to calm them. Each patient should have a complete history and physical examination performed, including a detailed mental status evaluation, to rule out an organic brain syndrome. Valuable information can be obtained from family members, friends, coworkers, neighbors, paramedical personnel, police, or previous medical records (see Table 108-1).24 The most important step in evaluating a patient with a suspected thought disorder is the assessment of the patient’s thought processes through the initial interview. The goals are to establish a positive physician-patient relationship, to make a correct diagnosis, and to gather information necessary to render an optimal disposition. The interview should be conducted in a quiet, comfortable room with adequate privacy. The examiner should be sitting, and if possible the interview should proceed to completion without interruption. If the patient is believed to be potentially dangerous but is not in need of immediate restraint, the interview should take place in an open area with security personnel nearby. The emergency physician should begin with an introduction and should express the desire to be “of help” to the patient. The interview should begin with open-ended questions designed to assess the patient’s complaint and understanding of the current circumstances. Good opening questions include “Do you understand why you have been brought here
Chapter 108 / Thought Disorders
Memory deficits Activity
ORGANIC
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1434
today?”; “You seem to be upset. Can you tell me why?”; and “Do you have any idea why you might be having these symptoms?” The patient’s appearance, body language, affect, and speech should be observed during the responses to these questions. A brief mental status examination should be performed. It may be initiated in a nonthreatening manner by stating, “I am now going to ask you a few questions to see how well you are concentrating.” The patient should first be asked questions regarding orientation to time, place, and person, as this is the most sensitive test for differentiating organic from functional disease. Patients who are disoriented should have a detailed medical evaluation to exclude the presence of an organic brain syndrome. Patients who are oriented should be assessed for attention, memory, intellectual functioning, and judgment in an attempt to determine their specific diagnosis, their potential for danger to themselves or others, and their degree of dysfunction and ability to care for themselves in the outpatient setting.
Rapid Tranquilization When psychotic patients exhibit behavior that is violent or so disorganized and uncooperative that clinical evaluation is impossible, the temporary use of physical restraints is indicated while rapid tranquilization is initiated (see Chapter 189). The technique of rapid tranquilization involves serial doses of a high-potency antipsychotic agent until target symptoms, such as agitation and excessive psychomotor activity, are improved. The goal is to facilitate cooperation of the patient without causing unnecessary sedation, which would inhibit further medical and psychiatric assessment. Oral, intramuscular (IM), or intravenous (IV) doses can be given every 30 to 60 minutes until the patient becomes calmer and more cooperative. If the patient is willing, an oral concentrate is preferred as it implies consent and can take effect almost as quickly as IM administration. Haloperidol (Haldol), a butyrophenone, is widely used for rapid tranquilization in the United States.25–27 The initial dose is 5 to 10 mg intramuscularly or intravenously for young to middle-aged patients and 0.5 to 2.0 mg intramuscularly or intravenously for elderly patients. Although rapid tranquilization with haloperidol quickly reduces tension, anxiety, and hyperactivity, delusions and hallucinations may not resolve for several weeks. In 2007, the U.S. Food and Drug Administration (FDA) released a non–black box warning about sudden death with the use of haloperidol in large doses or through the IV route. Droperidol (Inapsine), another butyrophenone, has also been extensively used for this indication in doses from 2.5 to 5.0 mg intramuscularly or intravenously.28,29 Compared with haloperidol, droperidol has a faster onset and shorter duration of action and causes slightly more sedation. The FDA black box warning was published in 2001 because of reports of a potential association between droperidol and prolonged QT interval, torsades de pointes, and sudden death.30 Despite subsequent studies supporting both the efficacy and safety of droperidol, the FDA warning has resulted in a significant decrease of its use in clinical practice.31,32 Neuroleptics should not be used for pregnant or lactating females, phencyclidine overdose, or anticholinergic drug-induced psychosis. In addition, they should not be used as the sole agent to manage agitation in patients with drug or alcohol withdrawal. Newer atypical antipsychotic agents appear to have a broader spectrum of response with fewer side effects than the typical agents. These medications are available in tablet form and should be considered for patients who consent to oral medication. However, oral administration of a pharmacologic agent
during an acute episode of agitation and psychosis is usually impossible. Ziprasidone (Geodon), aripiprazole (Abilify), and olanzapine (Zyprexa) are newer atypical antipsychotics currently approved in the United States for IM injection.33 For ziprasidone, the initial dose is 20 mg intramuscularly and can be repeated every 4 hours. It has been shown to be as or more effective than haloperidol for sedation and with fewer extrapyramidal side effects, but it has not yet been widely used in the emergency department setting.34,35 Benzodiazepines are effective in managing agitation in patients who have alcohol or sedative-hypnotic withdrawal, cocaine intoxication, or a contraindication to neuroleptic use. Benzodiazepines are helpful adjuncts to neuroleptic medication in providing rapid tranquilization, particularly in patients exhibiting combativeness or severe agitation. Lorazepam (Ativan), 1 to 2 mg, is frequently mixed with haloperidol, 5 mg, in the same syringe and administered intramuscularly or intravenously for this purpose.27 Benzodiazepines have also been shown to mitigate catatonic signs in schizophrenic patients.36 Disadvantages of benzodiazepines are the need for repeat dosing and close monitoring for potential respiratory depression with large doses.37
Outpatient Management The outpatient treatment of schizophrenia involves maintenance therapy using neuroleptic agents, family counseling, and social rehabilitation. Emergency physicians rarely prescribe outpatient neuroleptic medications but should be familiar with complications associated with their long-term use. Box 108-4 lists the most common neuroleptic medications currently used in the United States.38–44 The mechanism of action is related to the blockade of dopamine receptors in the central nervous system, particularly dopamine D2 receptors in the basal ganglia and limbic portions of the forebrain. The earlier, less potent drugs, of which chlorpromazine
BOX 108-4 Common Neuroleptic Agents Low potency Chlorpromazine (Thorazine) Chlorprothixene (Taractan) Intermediate potency Acetophenazine (Tindal) Loxapine (Loxitane) Mesoridazine (Serentil) Molindone (Moban) Perphenazine (Trilafon) Thioridazine (Mellaril) Triflupromazine (Vesprin) High potency Droperidol (Inapsine) Fluphenazine (Prolixin) Haloperidol (Haldol) Thiothixene (Navane) Trifluoperazine (Stelazine) Atypical Aripiprazole (Abilify) Clozapine (Clozaril) Olanzapine (Zyprexa) Paliperidone (Invega) Quetiapine (Seroquel) Risperidone (Risperdal) Ziprasidone (Geodon)
Complications of Neuroleptic Drug Therapy Dystonia Acute dystonia, the most common adverse effect seen with neuroleptic agents, occurs in 1 to 5% of patients. This reaction is caused by a disruption of the dopaminergic-cholinergic balance in the nigrostriatal pathways of the basal ganglia, resulting in cholinergic dominance.48 Dystonic reactions, which can occur at any point during long-term therapy and up to 48 hours after administration of neuroleptics in the emergency department, involve the sudden onset of involuntary contraction of the muscles of the face, neck, or back. The patient may have protrusion of the tongue (buccolingual crisis), deviation
1435
of the head to one side (acute torticollis), sustained upward deviation of the eyes (oculogyric crisis), extreme arching of the back (opisthotonos), and rarely laryngospasm. These symptoms tend to fluctuate, decreasing with voluntary activity and increasing under emotional stress, which occasionally misleads emergency physicians to believe they are factitious. Dystonic reactions should be treated with IM or IV benztropine (Cogentin) 1 to 2 mg, or diphenhydramine (Benadryl), 25 to 50 mg, which usually results in immediate reversal of symptoms. Patients should receive oral therapy with the same medication for 48 to 72 hours to prevent recurrent symptoms.48
Akathisia Akathisia is a state of motor restlessness characterized by a physical need to be moving constantly. It occurs most often in middle-aged patients during the first few months of therapy. Patients are usually pacing the room and expressing a sense of inner tension that is not relieved by activity. If asked, they do not want to be constantly moving but feel physically compelled to do so. This reaction can easily be mistaken for a decompensating psychosis, leading to a vicious circle in which more medication is given to treat a side effect caused by the same drug. This misdiagnosis can be avoided by carefully evaluating the patient for the exacerbation of positive psychotic symptoms, which are not increased by akathisia. Akathisia is treated with beta-blockers (e.g., propranolol, 30–60 mg/day) and anticholinergic drugs (e.g., benztropine, 1 mg twice to four times daily). A new potential agent for the treatment of aka thisia is glycine, a nonessential amino acid that stimulates glutamatergic neurotransmission.49 In addition, if possible, the dosage of the antipsychotic agent should be lowered or replaced with another drug.
Pseudoparkinsonism and Akinesia A clinical picture can occur that may be indistinguishable from Parkinson’s disease, particularly in elderly patients during the first month of therapy. Treatment with anticholinergic agents (e.g., benztropine) or antiparkinsonian drugs is usually effective. Akinesia, which is characterized by immobility, withdrawal, and lack of motivation, may be mistaken for a postpsychotic depression. It is responsive to antiparkinsonian drugs, but symptoms usually resolve gradually over time.
Tardive Dyskinesia Tardive dyskinesia usually appears after several years of neuroleptic drug treatment and is characterized by involuntary movements, especially of the face and tongue, that are described as writhing, grimacing, and choreoathetoid in nature. The earliest manifestation is often a curling or twisting movement of the tongue. The onset of these symptoms can be falsely attributed to psychological factors because they intensify under emotional stress, fatigue, and voluntary activity, and disappear with sleep. The reported prevalence of tardive dyskinesia ranges from 0.5 to 70%, with a mean value of 24%.48 The incidence of the disorder appears to be directly related to the duration of treatment, total cumulative dosage, evidence of preexisting brain damage, and age of the patient. It is more common in elderly women and patients with associated mood disorders. For patients with mild symptoms, discontinuing or lowering the dosage of antipsychotic agents, switching to a newer atypical neuroleptic agent, and cotreatment with benzodiazepines may reverse the symptoms. Patients with moderate to severe symp-
Chapter 108 / Thought Disorders
is the prototype, cause more pronounced sedation, orthostatic hypotension, and cardiovascular toxicity. This is the result of a combination of anticholinergic, antihistaminic, and anti-alpha-adrenergic effects. More potent agents (e.g., haloperidol) are safer, especially in older patients, because of their relative lack of these adverse effects. However, these more potent drugs are associated with a higher incidence of extrapyramidal symptoms, such as dystonias, akathisia, akinesia, and rigidity. The high frequency of severe adverse reactions, poor compliance by patients, and the large number of patients with symptoms refractory to traditional antipsychotic agents prompted the development of new alternative agents. These “atypical” neuroleptic agents block serotonin to a greater extent than dopamine, resulting in a low incidence of extrapyramidal side effects. Clozapine is particularly effective in patients who have not responded to other antipsychotic drugs.41 However, clozapine is expensive, has a side effect profile similar to that of the low-potency antipsychotic agents, and causes agranulocytosis in approximately 1% of patients.39 It is recommended only for the treatment of patients with refractory psychosis. Olanzapine, quetiapine, and the newest agent, aripiprazole, are similar to clozapine but have fewer side effects and less risk of agranulocytosis.38 Risperidone, another newer neuroleptic agent with improved effects on negative symptoms, has been found to be superior to haloperidol in several short-term trials.42–45 The most recent FDA-approved drug (2006), paliperidone (Invega), is the active metabolite of risperidone. Unfortunately, the expense and unavailability of these newer antipsychotic agents limits their utility for treating acute psychosis in the emergency department. Ziprasidone, risperidone, and quetiapine have been associated with QT interval prolongation similar to that seen with droperidol and haloperidol but have not yet been found to be associated with sudden cardiac death.46 Because of the high incidence of extrapyramidal symptoms in patients treated with high-potency neuroleptics, it is common practice to administer antiparkinsonian drugs (e.g., benztropine, procyclidine, trihexyphenidyl) at the same time, either to treat the adverse effects or to prevent them. Prophylactic treatment is most useful in patients with a history of extrapyramidal symptoms, those receiving high doses of high-potency antipsychotic agents, and those in whom the occurrence of these symptoms is likely to increase the risk of noncompliance. Noncompliance with antipsychotic medication remains a leading cause of psychiatric hospitalization. Patients with recurrent psychotic relapses caused by noncompliance are candidates for treatment with long-acting injectable antipsychotic drugs, usually given every 2 weeks. Three such agents available in the United States are fluphenazine, risperidone, and haloperidol decanoate.47
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1436
toms are difficult to treat, but some have variable degrees of improvement when treated with reserpine or tetrabenazine along with discontinuation of neuroleptic treatment.48
Orthostatic Hypotension All the antipsychotic agents can cause orthostatic hypotension, an effect related to alpha-adrenergic blockade. This complication is less common with the more potent agents (e.g., haloperidol). Typically, episodes are mild in severity and brief in duration. Symptomatic patients should be treated with oxygen, Trendelenburg’s position, and IV crystalloid fluid administration. Pressor agents (e.g., dopamine) should be used only for severe, symptomatic episodes that fail to respond to the previous measures. Agents with beta-agonist activity (e.g., epinephrine, isoproterenol) are contraindicated in these patients.
Neuroleptic Malignant Syndrome Neuroleptic malignant syndrome (NMS) is a life-threatening complication of neuroleptic drug treatment that affects 0.5 to 1% of patients.50 It is seen with both typical and atypical neuroleptic agents and usually occurs in the first few weeks after initiation of treatment, but can also be seen after a recent increase in drug dosage or after parenteral treatment with high doses of neuroleptic agents. Neuroleptic malignant syndrome is characterized by high fever, severe muscle rigidity, altered consciousness, autonomic instability, and elevated serum creatine kinase levels and can be confused with serotonin syndrome. Additional complications can include respiratory failure, gastrointestinal hemorrhage, hepatic and renal failure, coagulopathy, and cardiovascular collapse. The pathophysiology of NMS is not well understood, but it is thought to be related to dopamine depletion in the central nervous system leading to defective thermoregulation in the hypothalamus. Predisposing factors include exhaustion, dehydration, and the use of long-acting depot neuroleptics. Treatment consists of recognition and discontinuation of the neuroleptic agent, fever reduction, rehydration with IV fluids, and general supportive measures. Dantrolene, a direct-acting muscle relaxant, should be used in severe cases. It can be administered by continuous rapid IV push at a minimum initial dose of 1 mg/kg, repeated until symptoms subside or up to a maximum cumulative dose of 10 mg/kg. For severe symptoms, dopamine agonists such as bromocriptine, levodopa, and amantadine have shown encouraging results. Because of earlier
recognition and treatment, mortality rates with NMS have decreased from 30% to less than 10%.51
■ DISPOSITION The ultimate disposition of the acutely psychotic patient depends on the underlying cause of the psychosis, whether the patient is a danger to self or others, and the presence of social support in the community. Hospitalization is indicated for patients experiencing their first psychotic episode, for patients deemed to be a danger to themselves (suicidal) or others (homicidal), for patients who are grossly debilitated, for patients who are moderately debilitated but have no social support system within the community, and for patients with either functional or organic psychosis that does not clear with a brief period of treatment and observation in the emergency department. The decision to hospitalize psychotic patients is complex and imprecise and often must be made in a short period with limited information. A psychiatric short-procedure unit offers a cost-effective alternative to hospitalization.52 After stabilization, patients are moved from the emergency department to a separate treatment area, where they are treated for a period of 12 to 24 hours by a small staff of consultants.
KEY CONCEPTS ■
In patients exhibiting abnormal behavior, an organic etiology is suggested by (1) new onset of symptoms in a patient older than 35 years, (2) rapid onset of symptoms in a previously normal person, (3) visual hallucinations, (4) abnormal vital signs, and (5) lethargy or disorientation. ■ Physical restraints should be considered for patients who have violent behavior, verbally threaten staff, or demonstrate escalating agitation despite verbal attempts to calm them. Restraints should be removed after sedation has been achieved. ■ Rapid tranquilization is best accomplished with the use of high-potency neuroleptics (e.g., haloperidol, ziprasidone). Benzodiazepines (e.g., lorazepam) are a helpful adjunct for patients with severe agitation. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 109
Mood Disorders
Radu V. Saveanu, Marshall G. Vary, and Douglas A. Rund
■ PERSPECTIVE
■ PRINCIPLES OF DISEASE
Happiness and sorrow are common emotions but usually cause no impairment in functioning or threat to life. Mood disorder, by contrast, can significantly impair physical, social, and family functioning and can cause psychological pain, physical pain, and a negative perception of physical health. The term mood disorders replaced the term affective disorders in the fourth edition (2000) of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR).1 Mood usually refers to an individual’s subjective emotional state, whereas affect normally refers to the individual’s emotional state as it appears to an outside observer.2 Mood disorders refers to a group of psychiatric disorders where pathological moods are the dominant feature. According to DSM-IV-TR, mood disorders are divided into four categories: (1) depressive disorders (unipolar depression), (2) bipolar disorders, (3) mood disorder due to a general medical condition, and (4) substance-induced mood disorder. DSM-IV-TR assigns psychiatric diagnoses to patients by assessing specific observable and measurable symptoms and signs and applying straightforward operational criteria. Research regarding the neurobiology of mood disorders and genetics may ultimately permit classification of mood disorders by specific, genetically predisposed pathophysiologic derangements.
Current neurobiologic concepts provide the basis for various pharmacologic treatments. The psychosocial theories of depression consider the complex interaction of genetics, environment, and experience, providing the basis for understanding the various psychotherapeutic approaches to treatment.
■ EPIDEMIOLOGY Psychiatric problems account for at least 5.4% of all emergency department (ED) visits, and the rate of psychiatric-related visits has increased 15% since 1992.3 The World Health Organization ranks major depression as one of the most prevalent and disabling diseases in the world,4 and the lifetime prevalence of major depressive disorders in the United States is 16.2%. Depressed patients also have associated anxiety disorders (almost 60%), substance use disorders (24%), and impulse control disorder (30%).5 Depressed patients with substance abuse comorbidity have higher ED utilization than patients without such comorbidity.6 Patients with chronic illness have a much higher prevalence of undiagnosed depression than the general population. The lifetime suicide risk for persons with major untreated depressive illness is 15%.7 Mood disorders are becoming more common and have increased in every generation born after 1910.8 The prevalence of bipolar disorders (manic-depressive disorders) is substantially lower than that of major depression. The overall lifetime prevalence of a manic episode is 1.6%.
Neurotransmitters In the fourth century bce, Hippocrates believed that the body contained four essential humors: blood, phlegm, yellow bile, and black bile. Harmony in the brain required a harmony of humors. “Disharmony” produced mental illness. In the second century, Galen believed that “melancholia” resulted from an excess of black bile acting on the brain. Such excesses were thought to be caused by noxious stomach vapors, grief, anxiety, excessive wine, and advancing age. Proposed treatments included bloodletting, surgery, special diets, and exercise.9 Modern theories regarding the pathophysiology of depression have focused on the three major monoamine systems— serotonin, norepinephrine, and dopamine. Of the three neurotransmitter systems, serotonin has received the most attention. There is very strong evidence of reduced activity of serotonergic neurons in depression as assessed by cerebrospinal fluid and neuroendocrine studies. There are also very strong data demonstrating the role that norepinephrine plays in the etiology of depression. Norepinephrine reuptake inhibitors, such as nortriptyline, have been found to be effective antidepressants. Research over the past several years has increasingly focused on the role of dopamine in the etiology of depression. A number of studies have found a reduction in the synaptic availability of dopamine in patients suffering from depression.10 Complex processes in the nervous system cause depression or mania. If the concentrations of neurotransmitters at synaptic sites were the only factors responsible for alleviation of depression, the therapeutic effect of an antidepressant agent would be almost immediate. However, clinical improvement with these agents takes several weeks and is highly individually variable. Emphasis has now shifted to the study of neurobehavioral systems, intricate regulatory mechanisms, development of preferred neural circuits, the effect of environment on phenotype expression, and gene transcriptors.11 1437
1438
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
Cerebral Anatomy Certain areas of the brain are involved in processes that become abnormal during episodes of depression and mania.12,13 Stress activates neurons in the locus ceruleus, resulting in increased alertness, decreased appetite, increased heart rate, increased cortisol production, and other features of a stress response. The response can be dampened by neurons in the cerebral cortex. Prolonged stress, from which escape appears hopeless, seems to decrease the activity of neurons in the locus ceruleus. Another noradrenergic system, the median forebrain bundle, elicits reward-seeking behavior when stimulated. Prolonged stress decreases levels of norepinephrine in this region and may explain the lack of energy and interest that accompany depression.14,15 Serotonergic neurons are located in the brainstem dorsal raphe and project diffusely throughout the brain. The seroton ergic system seems to enhance sleep, appetite, libido, and circadian rhythms. Activation decreases aggressive behavior in animal models.16,17 Dopaminergic pathways include the tuberoinfundibular system, originating in the hypothalamus (prolactin secretion); the nigrostriatal system, originating in the substantia nigra (involuntary motor activity); the mesolimbic pathway, originating in the ventral tegmentum; and the mesocortical pathway, originating in the ventral tegmentum.18 These pathways regulate emotion, pleasure, learning, and reinforcement. The mesocortical pathway extends to frontal cortical regions that regulate complex cognition concentration and motivation.19
Endocrine System The cortical-hypothalamic-pituitary-adrenocortical system is affected in many patients with depression, causing increased levels of plasma cortisol, apparently from an impaired biofeedback loop regulating cortisol. The thyroid axis malfunctions in 5 to 10% of depressed patients. Thyroid-stimulating hormone levels are elevated, and thyroid replacement therapy facilitates treatment in certain patients.20
Genetics Family studies have repeatedly demonstrated a relationship between genetic inheritance and mood disorders.21 A review of twin studies in unipolar depression estimates that genetics accounts for 30 to 40% of the risk of developing depression, with the rest being attributed to environmental factors.10 Monozygotic and dizygotic twins have a high concordance rate, 70 and 35%, respectively, for mood disorder.22 The mechanisms of genetic transmission are still undetermined but may relate to the synthesis, transport, and action of serotonin and other transmitters.23,24 The inherited susceptibility to depression may manifest only during severe stress or serious illness.25
Psychosocial Theories The complex neural mechanism that regulates mood responds to and is modified by each person’s experience, including events in early childhood, reward and punishment during growth and development, interpersonal relationships, and various kinds of loss. Psychosocial theories of mood disorder form the basis for psychotherapy. Freud noted that personal loss included grief and sadness, but that depression also involved guilt and lowered self-esteem.26 Freud theorized that suicide in depressed patients is a manifestation of aggression that has been turned against the self in a person otherwise unable to express anger toward loved ones.
The interpersonal theory of depression emphasizes guilt, disputes between partners and family members, role transitions in families and relationships, and problems with social skills necessary to sustain a fulfilling relationship. The cognitive model of depression assumes that the “negative,” distorted thoughts or cognitions (thinking one is helpless, hopeless, worthless) can cause depression in certain individuals. The concept of “learned helplessness” first developed when animals were blocked from escape and subjected to repeated noxious stimuli such as electric shock. They eventually stopped trying to escape and became apathetic, even when escape became available.27 The concept was later applied to humans and adapted to construct a behavioral mode of therapy for depression.
■ CLINICAL FEATURES Major Depressive Disorder Major depressive disorder is characterized by one or more major depressive episodes, as defined by DSM-IV-TR criteria (Boxes 109-1 and 109-2), and the lifelong absence of manic episodes. A major depressive episode is characterized by disturbances in four major areas: mood, psychomotor activity,
Summary of DSM-IV-TR Criteria for a Major
BOX 109-1 Depressive Episode
A. Five or more of the following symptoms present almost every day during the same 2-week period and represent a change from previous functioning; at least one of the symptoms is either (1) depressed mood or (2) loss of interest or pleasure. Note: Do not include symptoms caused by a general medical condition, and do not include mood-incongruent delusions or hallucinations. 1. Depressed mood (can be irritable mood in children and adolescents) 2. Loss of interest or pleasure in activities 3. Significant weight loss when not dieting, or weight gain or decrease, or increase in appetite 4. Insomnia or hypersomnia 5. Psychomotor agitation or retardation 6. Fatigue or loss of energy 7. Feelings of worthlessness, or excessive or inappropriate guilt 8. Diminished ability to think or concentrate, or indecisiveness 9. Recurrent thoughts of death (not just fear of dying), recurrent suicidal ideation, or a suicide plan or attempt B. Symptoms do not meet criteria for a “mixed episode.” C. Symptoms cause clinically significant distress or impairment in social, occupational, or other functioning. D. Symptoms are not caused by direct physiologic effects of a substance (e.g., drug of abuse, medication) or a general medical condition (e.g., hypothyroidism). E. Symptoms are not better accounted for by bereavement; after the loss of a loved one, the symptoms persist for longer than 2 months or are characterized by marked functional impairment, morbid preoccupation with worthlessness, suicidal ideation, psychotic symptoms, or psychomotor retardation. Modified from American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed, Text Revision. Washington, DC, American Psychiatric Association, 2000.
1439
Factors in Evaluating Depression:
INterest Sleep Appetite Depressed mood Concentration Activity Guilt Energy Suicide
cognition, and vegetative function.28 The patient must exhibit or experience at least five symptoms for a minimum of 2 weeks.
Mood Disturbances Depressed mood is painful and referred to as being anguished, sad, gloomy, dejected, unhappy, discouraged, or in low spirits. The mood may also involve feelings of anxiety and irritability. The patient’s feelings can be so intensely painful that suicide may be seen as the only way to terminate the agony. Anhedonia refers to the inability to experience pleasure or interest in formerly pleasurable or satisfying activities. The patient must have actually stopped doing the formerly pleasurable activities, for example, an avid tennis or golf player who gives up playing entirely. Questions can help to elicit loss of interest or pleasure: “When were you last feeling well?”; “When you were feeling well, what kinds of things did you do for enjoyment?”; “Are you doing those things now?”; and “Are you enjoying them?”29
Disturbances in Psychomotor Activity Psychomotor disturbances can take the form of retardation or agitation. Psychomotor retardation includes significant slowing of thought processes and physical activity. The patient is slow to answer questions and moves slowly or not at all. Questioning such patients in the ED may be frustrating when the answers come slowly, in short words or phrases, and are low in volume and lack inflection. The “body language” of depression includes sitting slumped over, arms folded, mouth turned down, and eyes closed or downcast. Such slowing clearly affects the patient’s work, school, or family functioning. Symptoms may be erroneously attributed to worsening dementia in elders. An alternative presentation is psychomotor agitation, in which the patient fidgets, paces, rubs the skin, and is unable to sit still. Other common, almost stereotypical manifestations include hand wringing and tugging at the hair.
Vegetative Disturbances Vegetative symptoms include disturbances in three major areas: sleep, appetite, and sexual function. Depressed patients typically report some form of sleep disorder, such as difficulty falling asleep, middle insomnia, or the classic symptoms of early-morning wakening and inability to fall back to sleep. Some depressed patients may report sleeping 12 to 14 hours a day and inability to arise in the morning. This may be a more common symptom in depressed teenagers. The depressed patient may lose appetite and weight or may gain weight in a short time. Loss of interest in sexual activity and impotence can be considered vegetative symptoms; they may accompany
Cognitive Disturbances Depressed patients are unable to concentrate or think properly, which can cause significant dysfunction in a job or profession. Thought content is negative, such as recurrent thoughts of guilt, failure, worthlessness, and self-criticism. Suicide may preoccupy the patient’s thinking and may reinforce feelings of helplessness, perpetuating self-reproach. The patient may formulate a definite plan for ending life. Depressed patients must be questioned about suicidal thoughts and plans, which allow them to describe their pain and may provide them with some relief. Psychosis may accompany severe depression. Hallucinations and delusions are classified as mood congruent or mood incongruent. Mood-congruent delusions reflect the depressed mood. The patient may report, for example, being “already dead” or feeling like “my insides have rotted away.” Hallucinations typically consist of voices saying extremely unpleasant things or punishing the patient for previous wrongs. Moodincongruent delusions do not reflect the depressed mood as clearly and include the paranoid delusions of being followed and having one’s thoughts controlled by external forces.
Special Considerations Masked Depression Mood disorders may not be clear at presentation. The depressed patient may have only vague physical symptoms, such as weakness, fatigue, headache, or complaints of pain. Patients may not be aware of their depression and are often heavy utilizers of medical care. Such symptoms may be the presenting features of a masked, or hidden, depression. Clues suggesting mood disturbance include the recent onset of a set of unusual behaviors, trouble at work or job loss, marital difficulties, or self-destructive behavior (e.g., substance abuse, sexual promiscuity).
Children and Adolescents A common and overt presentation of depression in an older child or teenager is a suicide attempt. Such patients should be considered depressed and unstable until subsequent assessment differentiates depression from other conditions in these age groups, including transient psychoses, anxiety disorders, high levels of life stress, and substance abuse. Symptoms of depression in children and adolescents generally follow the same criteria as for adults. Some children are misdiagnosed as having attention deficit disorder, especially if symptoms involve poor concentration, listlessness, agitation, and withdrawal from daily activities. Depression in these age groups is often misunderstood, masked in its presentation, or simply overlooked by friends, parents, teachers, and physicians. Adequate treatment maximizes the child’s potential and minimizes the serious negative impact depression can have on multiple spheres of development.
Elders Depression is common in elders; losses and grief, serious health issues, and loss of autonomy create a setting conducive to depression. The classic symptoms of moderate to severe depression, with or without psychosis, are typically seen. Depressed patients can present with symptoms involving
Chapter 109 / Mood Disorders
BOX 109-2 “IN SAD CAGES”
depression or may be part of the anhedonia associated with depressed mood.
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1440
memory loss, inattention, withdrawal from daily activities, confusion, and lapses in personal and social hygiene that suggest dementia rather than depression. When such symptoms are from depression, the condition is called pseudodementia. Serious depression in elders is a highly treatable, reversible condition. Distinguishing it from dementia is essential for further diagnostic and therapeutic follow-up.
Other Depressive Disorders Seasonal Affective Disorder Seasonal affective disorder is not a separate mood disorder but a subclassification of major depressive disorder that is diagnosed when major depression occurs during seasons with less daylight (fall and winter), then either resolves or occasionally changes to manic episodes in seasons with more daylight, for at least 2 consecutive years. Melatonin, a hormone secreted in the brain and produced at high levels in the dark, has been implicated in the etiology of this disorder. Symptoms generally include hypersomnia, anergia, weight gain, and craving for carbohydrates. Phototherapy is an effective and safe treatment for this “winter depression,”30 and light exposure to the eyes seems to be essential but the exact mechanism of action is still unknown.
Postpartum Depression Symptoms of depression are common in the postnatal period. Up to 65% of mothers report some depressed mood after childbirth, often called “postpartum blues.” Symptoms are generally mild and transient, although in some patients it may lead to an episode of major depression. Approximately 10% of mothers will experience a full-fledged episode of depression in the postpartum period. The signs and symptoms of postpartum depression are similar to those of a major depressive disorder, but the onset is within four weeks of delivery.31 Postpartum depression is more common in those who have a history of mood disorder, experience marital conflict, and have limited assistance with infant care. Severe postpartum depression may negatively influence development in the child.32
Dysthymic Disorder Dysthymic disorder is a long-standing, fluctuating, low-grade depression. Some features of a major depressive episode may be present, but marked changes in appetite or psychomotor disturbances are not typically observed. Depressed mood typically begins early in life, and the individual may report having always been depressed. Affected individuals generally are able to carry out their work assignments, but they gain little pleasure from the leisure activities others find enjoyable, such as recreation, time with family, or sexual activity. They typically experience significant impairment in interpersonal functioning.
Bipolar Disorders Bipolar disorder is lifelong, with episodic exacerbation of symptoms and deterioration of function characterized by extreme mood swings. Patients with bipolar disorder thus require different forms and intensities of treatment at different times. Bipolar I disorder includes at least one manic episode, and patients have typically had one or more major depressive episodes. Bipolar II disorder involves a hypomanic episode and at least one major depressive episode. A hypomanic
Summary of DSM-IV-TR Criteria for a
BOX 109-3 Manic Episode
A. Distinct period of abnormally and persistently elevated, expansive, or irritable mood, lasting at least 2 weeks (or any duration if hospitalization is necessary). B. During the period of mood disturbance, three or more of the following symptoms have persisted (four, if the mood is only irritable) and have been present to a significant degree: 1. Inflated self-esteem or grandiosity 2. Decreased need for sleep (e.g., feels rested after only 3 hours of sleep) 3. More talkative than usual or pressure to keep talking 4. Flight of ideas or subjective experience that thoughts are racing 5. Distractibility (i.e., attention too easily drawn to unimportant or irrelevant external stimuli) 6. Increase in goal-directed activity (either socially, at work or school, or sexually) or psychomotor agitation 7. Excessive involvement in pleasurable activities that have a high potential for painful consequences (e.g., buying sprees, sexual indiscretions, foolish investments) C. Symptoms do not meet criteria for a “mixed episode.” D. Mood disturbance is sufficiently severe to cause marked impairment in occupational functioning or social activities or to necessitate hospitalization to prevent harm to self or others, or psychotic features are present. E. Symptoms are not caused by direct physiologic effects of a substance (e.g., drug of abuse, medication) or a general medical condition (e.g., hyperthyroidism). Modified from American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed, Text Revision. Washington, DC, American Psychiatric Association, 2000.
episode includes the features of a manic episode without psychosis, marked impairment of function, or the need for hospitalization.
Manic Episode To be considered manic (Box 109-3), the disturbance must be severe enough to cause psychosis, the need for hospitalization, or marked impairment in functioning. Bipolar disorders are much less common than major depressive disorder. The overall prevalence of a manic episode is 1.6% in both women and men.33 Patients who are experiencing a manic episode may be gregarious, humorous, and engaging. An alternative presentation is one of belligerence and irritability. Clues to mania include a history of the patient’s behavior immediately before the evaluation and any prior history of bipolar disorder or a history of taking medications almost exclusively prescribed for bipolar disorder, such as lithium. In most cases, the manic patient will be brought to the ED by someone else (e.g., family, police, emergency medical services). They often try to leave as soon as possible, display impaired judgment and impulsivity, and may need to be restrained. Pressured speech is one of the first clinical signs of mania. The patient keeps talking, with no interruption between thoughts or sentences. The speech may be loud and rapid, with creative, amusing, or trivial and irrelevant content. The patient may tell jokes, use puns, or play other word association games. A hallmark of mania is grandiosity, which involves feelings of inflated self-esteem and great personal importance. The patient may
Cyclothymic Disorder Cyclothymic disorder is characterized by a life of mood swings of insufficient severity to meet criteria for a bipolar disorder. Persons with this disorder may have a chaotic life, characterized by frequent mood swings, unstable relationships, and uneven school or work performance.
Mood Disorders Caused by a General Medical Condition Certain medical illnesses have a well-known association with mood disorder (Box 109-4). In Parkinson’s disease, electrical stimulation to a certain area of the substantia nigra alleviates symptoms of depression. Stimulation of an area only 2 mm
BOX 109-4
Medical Illnesses Associated with Onset of Depression
Neurologic Parkinson’s disease Stroke Multiple sclerosis Head trauma Sleep apnea Neoplastic Pancreatic carcinoma Brain tumor Disseminated carcinomatosis Endocrine Hypothyroidism Hyperthyroidism Cushing’s disease Addison’s disease Diabetes mellitus Infectious Human immunodeficiency virus Cardiac Coronary artery disease Myocardial infarction Renal End-stage renal disease Renal dialysis Connective tissue Lupus erythematosus Rheumatoid arthritis
1441
away can cause acute reversible symptoms of depression, such as crying, not wanting to live, and hopelessness.34 Parkinson’s disease has a well-known association with depression, with up to 40% of patients demonstrating major depression.35 Certain malignancies have a well-known association with depression, including pancreatic carcinoma, brain neoplasm, and disseminated malignancy (e.g., lymphoma).36 Coronary artery disease,37 myocardial infarction, stroke, end-stage renal disease, acquired immunodeficiency syndrome, several endocrine diseases, and connective tissue disease are also associated with major depressive disorder.38 After a myocardial infarction, patients with depression experience a 3.5-fold increase in cardiovascular mortality compared with nondepressed patients.39 Patients with depression appear to be more likely to develop stroke, diabetes, and osteoporosis than those who are not depressed.40,41 Depression related to medical conditions may be different in some respects from primary depression. For instance, the former responds less favorably to antidepressant medication than primary depression.42 Two significant issues arise in the assessment of patients with depression who have a serious medical illness. First, symptoms of depression must be distinguished from the symptoms and signs associated with serious medical illness (e.g., weight loss, loss of energy, slowing of activity, sleep disturbance, loss of ability to concentrate). Some experts have proposed that alternative criteria for depression caused by general medical condition be substituted for DSMIV-TR neurovegetative symptoms in patients with serious medical illness such as depressed appearance, social withdrawal, pessimism or self-pity, anhedonia, and nonreactive mood. Second, it is important to determine whether the depression associated with terminal, rapidly progressive, or painful illness should be considered appropriate. Although patients with such diseases may understandably be sad, most do not have major depression. The treatment of major depression in such patients should always be attempted and can greatly improve their quality of life.
Mood Disorders Caused by Medications or Other Substances Certain medications are associated with symptoms of mood disorders (Box 109-5). Intoxication or chronic heavy use of alcohol, sedatives, hypnotics, anxiolytics, narcotics, and other depressants can cause symptoms of a major depressive episode. Stimulants such as cocaine, phencyclidine, hallucinogens, and amphetamines can cause symptoms of a manic episode. Mood disorder symptoms can also develop during withdrawal. To qualify for this diagnosis, the symptoms must not occur exclusively during a course of delirium, must cause significant distress or impairment of functioning, and must develop within a month of either substance intoxication or withdrawal. When the mood disorder predates the period of substance abuse or lasts longer than 1 month after the period of abuse, the diagnosis may be an underlying mood disorder, such as a major depressive disorder or bipolar disorder, with a comorbid substance abuse or dependence diagnosis. Substance abuse is often seen in patients with underlying depressive or bipolar conditions.
■ DIAGNOSTIC STRATEGIES The initial history and physical examination should focus on the presenting complaints and evaluate the possibility that drug abuse, medications, or a general medical condition may be responsible for the patient’s condition. The diagnosis of a
Chapter 109 / Mood Disorders
describe a massive undertaking such as “uniting the world’s churches” or “solving world poverty.” Manic patients have decreased or no need for sleep and typically report being awake for days during a manic episode. They may be involved in a massive project (e.g., writing a novel), may completely disregard consequences of actions, may have difficulty with spending (e.g., credit cards revoked), and may engage in risky behavior (e.g., sexual liaisons with strangers, risky driving). An accurate history must be obtained from family or others who know the patient’s behavior. Manic patients may present to the ED as trauma patients, injured by an action reflecting the patient’s grandiosity (e.g., attempting to fly), impulsivity, or belligerence (e.g., fighting, resisting arrest). A manic episode may be punctuated by abrupt periods of tearfulness and profound depression, including suicidal ideation. When depressive and manic features occur concurrently in such a manner, the disorder is termed mixed or bipolar, mixed phase.
1442
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
BOX 109-5 Medications That Can Cause Depressive or Manic Symptoms Depressive symptoms Antihypertensives Beta-blockers Captopril Clonidine Diltiazem Enalapril Nifedipine Prazosin Thiazide diuretics Anticonvulsants Phenytoin Topiramate Valproic acid Hormones Anabolic steroids Contraceptives Corticosteroids Thyroid hormone Sedative-hypnotics Barbiturates Benzodiazepines
mood disorder is based on history and observation of the patient’s ability and style in relating to family and medical staff. The patient’s body language may be helpful. Precipitating events (e.g., loss of job or relationship), accompanying symptoms (e.g., hallucinations, delusions, anxiety disorder, mania), and suicidal intent should be assessed. The patient’s history should be confirmed through interviews with family, friends, or eyewitnesses to the events that precipitated the ED visit. A tentative diagnosis can be established using DSM-IVTR criteria. Laboratory tests to investigate medical conditions may be necessary (see Box 109-4), but no tests can confirm or exclude mood disorders.
■ DIFFERENTIAL CONSIDERATIONS Medical disorders, medications, and substance abuse or withdrawal not only can cause but also can mimic mood disorders. The patient who presents with agitation, for example, might have hypoxia, cocaine intoxication, or alcohol-sedative withdrawal. The patient with symptoms and signs of depression may have an unrecognized malignancy or sedative intoxication. Antidepressant medications are used to treat a variety of disorders, such as anxiety, obsessive-compulsive disorder, post-traumatic stress disorder, pain syndromes, smoking cessation, and vasodepressor syncope. Patients taking antidepressant medication are often not being treated for depression. Grief and bereavement are normal human reactions to the acute loss of another person, health, social position, or job. The period of mourning is characterized by sadness, diminished sense of well-being (somatic complaints), sleeplessness, and sadness triggered by thoughts of the loss. Normal grief, however, does not include guilt, loss of self-esteem, feelings of worthlessness, suicidal intent, psychomotor retardation, or occupational dysfunction.
Manic symptoms Psychiatric agents Antidepressants Antibiotics Acyclovir Chloroquine Interferon Isoniazid Norfloxacin Ofloxacin Sulfonamides Other agents Amantadine Bromocriptine Cyclobenzaprine Cycloserine Digitalis Disopyramide Levodopa Metoclopramide Nonsteroidal anti-inflammatory drugs Phenylpropanolamine Theophylline
Adjustment disorders are behavioral or emotional disorders that occur in response to an identifiable stress or stressors. The emotional component can involve sadness, low self-esteem, suicidal behavior, hopelessness, helplessness, or other selfthreatening behavior. Acute adjustment disorder occurs within 3 months of the stressor and does not last longer than 6 months. The stressors are typically not as severe as those precipitating bereavement reaction, and the responses are often more maladaptive. The teenager who ends a romantic relationship, for instance, may attempt a drug overdose in response to the stress. In such cases, adjustment disorder is a more likely diagnosis than major depressive episode. The pattern of recurrent maladaptive behavioral responses to stress may be lifelong, but the acute episode should resolve within 6 months. Borderline personality disorder is characterized by unstable personal relationships, unstable self-image, and inappropriate behaviors. The disorder may include chronic feelings of emptiness, which may be misdiagnosed as depression, or lability of mood, which may be mistaken for mania or hypomania. Borderline patients typically live lives of crisis and constant conflict. Dementia can be confused with depression. Dementia is characterized by abnormal mental status, including abnormalities in tests of memory, calculation, and judgment. Delirium with waxing and waning sensorium, hallucinations, and delusions may involve disorganization, agitation, and restlessness, which might first be considered features of mania or agitated depression. Differential considerations for manic symptoms include the manic phase of bipolar disorder, stimulant abuse (e.g., cocaine, amphetamines), hallucinogen abuse, alcohol or sedative withdrawal, delirium, hyperthyroidism, other medical conditions causing agitation, brief reactive psychosis, schizoaffective disorder, and schizophrenia.
1443
■ MANAGEMENT
The creation of a safe and stable environment for the patient must be a first priority in management. The patient with an acute manic episode may be disruptive, refuse medical evaluation, and make repeated attempts to leave the ED. The initial step in treating such a disruptive patient is to offer assistance in reducing their agitation (placing the patient in a single room, recommending medication). At times this approach does not work and the patient may need to be placed in restraints for his or her safety and that of others. Initiating treatment for a mood disorder is not typically done in the ED. An exception is the acute manic episode (or possibly a severe depressive episode with psychosis) with behavior so extreme that the patient or others are threatened. Such cases may well involve significant hallucinations, delusions, and other features of psychoses. In such cases an antipsychotic agent is often indicated. For years, clinicians have used intramuscular or oral haloperidol with or without lorazepam to calm such patients. A typical regimen for “rapid tranquilization” is an initial dose of 5 mg haloperidol with 2 mg lorazepam intramuscularly and reassessment in 30 to 45 minutes for resolution of “target” symptoms such as agitation. Another 5-mg dose is administered after 30 to 60 minutes as needed for improvement in hallucinations, delusions, agitation, or violent behavior.43 Most patients respond after one or two doses. Benztropine (Cogentin), 1 to 2 mg, is often given initially to prevent extrapyramidal symptoms. Droperidol is a popular antipsychotic drug used effectively for agitation but has a black box Food and Drug Administration (FDA) warning about prolongation of QT intervals and torsades de pointes.44 The “atypical” antipsychotic medicines include ziprasidone, risperidone, olanzapine, aripiprazole, and quetiapine. The atypical agents are favored because they produce few of the side effects associated with conventional antipsychotic agents, such as acute dystonia, other extrapyramidal symptoms, and sedation. Oral doses should be offered first, and several agents, including risperidone, olanzapine, and aripi prazole, are available in rapidly dissolving tablet form. Three are available as an intramuscular injection: ziprasidone (Geodon), olanzapine (Zyprexa), and aripiprazole (Abilify). Ziprasidone 10 to 20 mg has been shown to be effective; however, its use is limited to 40 mg per 24 hours. Olanzapine 2.5 to 10 mg has also been shown to be effective, but has been associated with postural hypotension, and is not recommended in combination with benzodiazepines due to risk of hypoventilatory syndrome. Aripiprazole is the newest agent and at doses of 9.75 to 15 mg seems to be the least sedating of the atypicals, but more likely to cause nausea and vomiting.45–50 Immediate psychiatric consultation should begin during the initiation of rapid tranquilization, since patients undergoing rapid tranquilization will generally require hospitalization (Box 109-6).
Long-Term Treatment
Psychiatric Episode
Patient demonstrates risk of suicide or homicide or of significant harm to self or others. Patient lacks the capacity to cooperate with treatment. Patient has inadequate psychosocial support for safe outpatient treatment and compliance. Patient has a comorbid condition or complication that makes outpatient treatment unsafe (e.g., acute psychosis, bizarre behavior, need for detoxification).
medication trials,51 and many patients in general medical practice are inadequately treated.52 Coexistent medical illness, psychotic or bipolar symptoms, substance use, and recurrent or refractory depressive symptoms must be considered in the choice of a medication for treatment of depression. The side effects of tricyclic antidepressants and monoamine oxidase inhibitors, along with strict dietary limitations, have led to the use of selective serotoninreuptake inhibitors or serotonin-norepinephrine reuptake inhibitors as first-line treatment for depression. Side effects of these agents may include dizziness, sedation, peripheral anticholinergic symptoms, weight gain, sexual dysfunction, neurologic symptoms, insomnia, and anxiety. Frequently another antidepressant, ECT, thyroid hormone, or other psychoactive medication may need to be added for patients with treatmentresistant depression. Psychotherapy. Brief psychotherapy is often initially employed in patients with major depression. Interpersonal psychotherapy, psychodynamic psychotherapy, and group or family therapy are also used with some patients. Psychosocial therapeutic support typically includes community-based support groups that focus on specific individual, occupational, or family/marital issues that arise in depression and are amenable to group and supportive intervention. Depressed patients benefit most from a combination of somatic therapy (medication and/or ECT) and psychotherapy. All patients with incomplete therapeutic response, recurrent depression, or comorbid conditions (e.g., anxiety/panic, substance abuse) should receive multimodal treatment.53 Electroconvulsive Therapy. ECT has a high therapeutic success rate and an excellent safety profile but is not a first-line treatment for uncomplicated major depression. In part, this is a result of an undeserved reputation among laypersons that ECT causes “permanent brain damage.” Indications for ECT include severe depression with malnutrition, severe psychosis with agitation, continuing significant suicide risk with ongoing suicidal behaviors, prolonged catatonia, and recurrent depression previously with a positive response to ECT. ECT is more often used as a second-line treatment for patients with moderate to severe depression who have not responded to trials of medication or who cannot tolerate the medication because of side effects or concurrent medical conditions.
Depression
Bipolar Disorders
Effective treatment modalities for depression are grouped into three broad categories: (1) antidepressant medication, (2) psychotherapy, and (3) electroconvulsive therapy (ECT). Antidepressant Therapy. Many effective antidepressants are available for first-episode uncomplicated major depression. After 4 to 6 weeks of therapy, the response rate is usually 60% or greater for all agents. However, 10 to 15% of patients quit
Bipolar disorder is treated primarily with mood-stabilizing drugs, including lithium, valproate, carbamazepine, and lamotrigine. Almost all bipolar patients require a mood stabilizer during exacerbation of depression or mania, and most patients benefit from a mood stabilizer for ongoing supportive maintenance treatment as well. Lithium was the first highly effective mood-stabilizing agent for the treatment of bipolar
Chapter 109 / Mood Disorders
Emergency Department Stabilization
Criteria for Hospitalization of Emergency
BOX 109-6 Department Patient with Acute
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1444
patients.54 Valproate is a very effective mood stabilizer with dose-related side effects, most of which clear after an initial period of treatment or with reduced dosage.55 Valproate can be instituted rapidly in acutely manic bipolar patients.56 Both lithium and valproate serum levels are routinely monitored during therapy. The therapeutic window is much wider for valproate than for lithium. Carbamazepine also has doserelated side effects but has rare potentially serious side effects. Mood-stabilizing medications usually take 3 or more weeks to become effective.57 Some bipolar patients require antipsychotic medications and benzodiazepines in the interim to control symptoms. Some patients with bipolar disorder have significant psychotic symptoms requiring use of a major neuroleptic, such as haloperidol, ziprasidone (Geodon), risperidone (Risperdal), olanzapine (Zyprexa), quetiapine (Seroquel),58 or aripiprazole (Abilify). After 3 weeks, a second moodstabilizing medication is often added for patients who have a partial response. Atypical antipsychotics have significant moodstabilizing properties and studies have shown that they are effective in treating both bipolar mania and bipolar depression (although not all are FDA approved for this indication).58 For bipolar patients with a depressive episode, adjunctive treatment with antidepressants should be done very cautiously because of associated risks of switching into mania or hypomania.59 In addition, some recent studies found that antidepressants may not be effective for bipolar depression and may worsen the course of the illness.60 Bipolar patients are sensitive to psychosocial stressors, sleep loss, changes in medication dosage, substance abuse, and medical illness. This sensitivity can lead to a marked
worsening in a patient’s level of adaptation and functioning. It is helpful to understand these precipitating stressors when developing a therapeutic support plan for patients. Psychosocial therapeutic support, including individual psychotherapy, supportive community groups, family/marital treatment, and occupational support, are important in both the acute phase and the maintenance phase of treatment for bipolar patients.
KEY CONCEPTS ■
Patients with apparent mood disorders should be evaluated for a medical disorder, medication effect, or drug use that can mimic both depression and mania. ■ Mood disorders should be suspected in patients with multiple, vague, nonspecific complaints and in patients who are frequent, heavy utilizers of medical care. ■ The differentiation of depression and dementia in elders can be difficult but is important, since depression often responds dramatically to treatment. ■ Patients with mood disorders should be assessed for their potential for violence or self-harm before discharge.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 110
Anxiety Disorders
Eugene E. Kercher and Joshua L. Tobias
Anxiety is essential to the human condition. Confrontation with anxiety can relieve us from boredom, sharpen our sensitivity, and create the tension which is necessary to preserve human existence. Rollo May Acute anxiety and apprehension are common in emergency department (ED) patients. However, many medical entities mimic anxiety disorders, and up to 42% of patients thought to have anxiety disorders are later found to have organic disease. Emergency physicians must thoroughly assess the anxious patient and identify and appropriately treat any underlying medical conditions.1
■ PERSPECTIVE Anxiety is a specific unpleasurable state of tension that forewarns the presence of danger. This uneasiness stems from the anticipation of some imminent danger, the source of which is unknown or unrecognized. Vigilance is a positive consequence of anxiety, helping people to recognize threats quickly, which produces more learning and more intelligence. The capacity to experience anxiety and the capacity to plan are therefore related, with anxiety accompanying intellectual activity as its “shadow.”1 Anxiety facilitates performance up to a point with the welldescribed adrenergic responses to stress that contribute to survival. When responses go beyond this point, further increases in anxiety may lead to deterioration of performance and nonadaptive responses may add to the stress of the patient. The threshold for pain decreases and the person becomes more aware of bodily discomfort. Respiratory, cardiovascular, gastrointestinal, genitourinary, and neuromuscular complaints become prominent.1 Once the normal reaction to a threat is surpassed and function is impaired, pathologic anxiety (anxiety disorders) is the result. The emergency physician should not assume anxiety is purely functional because physical discomfort and illness often trigger an anxiety attack. The anxiety state makes significant metabolic demands that may actually cause a marginally compensated organ system to fail. It is the goal of the emergency physician to be able to distinguish between the anxiety and the illness and, if necessary, treat both entities.
■ EPIDEMIOLOGY Approximately 40 million American over the age of 18 are affected by anxiety disorders each year. This is nearly 20% of
the adult population of the United States.2 Anxiety disorders are among the most prevalent psychiatric disorders, and are the most common psychiatric problem seen by primary care physicians, with 20% of these patients experiencing a type of anxiety disorder.3 Most people who use primary care services have significant mood and anxiety symptoms, such as panic disorders, generalized anxiety disorders, and depression. Unfortunately, nearly half of these patients exhibit these symptoms but never receive appropriate treatment,4 in part, because the patients would rather present with a physical complaint to the physician and try to disguise their anxiety, rather than undergo the perceived stigma that goes with psychiatric complaints.5 Patients with chronic illness and those who make frequent medical visits have higher rates of anxiety and depression. The prevalence of anxiety disorders surpasses that of any other mental health disorder, including substance abuse. In view of the close relationship between alcohol abuse and anxiety disorders, those with anxiety disorders often turn to alcohol and substance abuse as a form of self-medication and the substance abuser frequently develops underlying anxiety in relation to the use of alcohol and drugs.6
■ PRINCIPLES OF DISEASE The precise mechanism for the cause of anxiety has not been established. Noradrenergic, serotonergic, and other neurotransmitter systems all play a role in the body’s response to a stressor. The serotonin system and the noradrenergic systems are common pathways implicated in anxiety. It is believed that low serotonin system activity and elevated noradrenergic system activity are involved. Gamma butyric acid (GABA) is the principal inhibitory neurotransmitter in the central nervous system. Benzodiazepines’ principal mechanism of action is on the GABAA receptors. The well-established effectiveness of benzodiazepines in the treatment of anxiety has led to the study of the GABA system and its relationship to anxiety. Newer studies are focusing on the role that corticosteroids may play in fear and anxiety. Steroids are thought to induce chemical changes in select neurons that strengthen or weaken certain neural pathways which affect behavior under stress.7 Other investigators have found anxiety reactions are associated with aberrant metabolic changes induced by lactate infusion and hypersensitivity of the brainstem to carbon dioxide receptors. Newer research is focusing on the regulatory centers found in the cerebral hemispheres. The hippocampus and the amygdala regulate emotion and memory and are important areas in relation to an individual’s response to fear.8 Family studies suggest genetic factors are implicated in anxiety, but 1445
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1446
the precise nature of the inherited vulnerability is unknown. Psychological and environmental factors, as outlined in psychodynamic, behavioral, and cognitive theories, also play a causative role in the generation of anxiety in biologically predisposed individuals.8
■ CLINICAL FEATURES Many patients entering the unfamiliar environment of the ED are going to experience anxiety and stress, and for some this can be a significant clinical issue. The ED patient encounters a world of both internal and external dangers: assaults on bodily integrity in the form of uncomfortable procedures and forced intimacy with strangers; the atmosphere of illness, pain, and death; and separation from loved ones and familiar surroundings. The patient typically experiences uncertainty about his or her illness, the implications that the illness may have on personal relationships and employment, and the financial burden that may accompany the illness. The anxious patient can be a diagnostic challenge. The presence of anxiety may represent the patient’s reaction to medical illness or the medical setting or a manifestation of the physical disorder itself, or the anxiety may be an expression of an underlying psychiatric disorder. The distinction between anxiety as a symptom and anxiety as a syndrome may be difficult to make in the ED. There is an overlap between normal situational anxiety and fear, anxiety-like symptoms resulting from a variety of organic disease states and their treatments, and the characteristic presentation of anxiety itself. The physical symptoms of autonomic arousal (e.g., tachypnea, tachycardia, diaphoresis, light-headedness) may be the only manifestation of anxiety (Box 110-1). Patients may only complain of overall poor health or vague subjective findings when they visit the physician. Classic panic disorder symptoms of chest pain, shortness of breath, and the sense of impending doom will often lead the patient to the ED, especially if it is the primary episode.7 Anxiety associated with organic etiologies is more likely to present with physical
BOX 110-1 Somatic Symptoms of Anxiety Respiratory Hyperventilation Sense of dyspnea Cardiovascular Palpitations Chest discomfort Awareness of missed beats Gastrointestinal Dry mouth Difficulty in swallowing Epigastric discomfort Excessive flatulence Frequent or loose stools Genitourinary Frequent or urgent micturition Failure of erection Amenorrhea Menstrual discomfort Neuromuscular Tremor Aching muscles Prickling sensations Headache Dizziness, tinnitus
symptoms and less likely to be associated with avoidance behavior.9
■ DIFFERENTIAL CONSIDERATIONS Medical Illness Presenting as Anxiety Patients with anxiety disorders may present with apparent physical disease, and many physical diseases may be strongly associated with symptoms of anxiety. Differentiating between these two scenarios is a daunting task for the emergency physician. Several factors help distinguish an organic anxiety syndrome from a primary anxiety disorder10 (Box 110-2). With anxiety, the somatic symptoms can be so prominent that they occupy most of the patient’s attention, making it difficult to differentiate between a primary anxiety disorder or reactive anxiety to a situation or disease. Anxiety disorder classifications, in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), include anxiety caused by a general medical condition11 (Box 110-3). Anxious patients are frequently convinced that their problem is purely physical. The emergency physician must realize that the patient with anxiety is not in control of the symptoms and frequently cannot immediately identify the correct precipitant. Even though the patient may be uncomfortable, uncooperative, impatient, and unreasonable, triage medical personnel must recognize that the patient believes an illness truly exists and is not being consciously manipulative. Because anxiety may be the most obvious symptom of an underlying disease or condition, the patient should be evaluated for exacerbation of known preexisting disease as well as for onset of new illness. The emergency physician must keep in mind that anxiety is associated with increased medical risk in the acute exacerbation of chronic illness.12 The classic scenarios of pulmonary embolism and hyperthyroidism causing anxiety are well documented. Cardiac disease studies indicate poorer outcomes in post-myocardial infarction patients with anxiety than those without documented anxiety. Patients with respiratory diseases, such as asthma or chronic obstructive pulmonary disease, often develop anxiety with their long-standing illnesses. In addition, many of the medications used to treat the above illnesses may induce anxiety.5 The most common organic cause of anxiety is alcohol and drug use, from either intoxication or, more typically, withdrawal states.
Cardiac Diseases Various psychiatric conditions may present to the ED with complaints of chest pain. Approximately 25% of patients with chest pain that present to the ED have panic disorder. Their panic disorder often goes undiagnosed, resulting in multiple visits and expensive cardiac workups with each visit.13 Some of the symptoms of myocardial infarction and angina pectoris may
BOX 110-2 Predictors of Organic Anxiety Syndrome 1. Onset of anxiety symptoms after age 35 years 2. Lack of personal or family history of an anxiety disorder 3. Lack of childhood history of significant anxiety, phobias, or separation anxiety 4. Lack of avoidance behavior 5. Absence of significant life events generating or exacerbating the anxiety symptoms 6. Poor response to antipanic agents
1447
BOX 110-3 Definitions of Anxiety Disorders
From American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed, Text Revision. Washington, DC, American Psychiatric Association, 2000.
include crushing chest pain, shortness of breath, nausea, palpitations, heavy perspiration, and a feeling of impending death. These are also the primary symptoms of acute anxiety, but the pain is usually described as atypical, and patients are generally female and younger.14 Because of the morbidity and mortality of cardiovascular disease, a patient warrants a full cardiac evaluation when the differentiation between myocardial infarction and acute anxiety is unclear. Cardiac dysrhythmias can cause palpitations, discomfort, dizziness, respiratory distress, and fainting. An anxious patient with a panic disorder will frequently have similar symptoms. Fortunately, most dysrhythmias can be documented and characterized by an electrocardiogram. Mitral valve prolapse syndrome can be associated with palpitations and panic attacks indistinguishable from a panic disorder. Benzodiazepines can
Endocrine Diseases The DSM-IV defines the most common endocrinologic conditions associated with anxiety states as hypoparathyroidism, hyper- and hypothyroidism, hypoglycemia, pheochromocytoma, and hyperadrenocorticism.11 Anxiety is the predominant symptom in 20% of patients with hypoparathyroidism. Other symptoms include paresthesias, muscle cramps, muscle spasm, and tetany. Most cases are idiopathic or the result of surgical removal of the parathyroid glands during thyroidectomy, and studies indicate a higher incidence of anxiety in the surgically removed subset of patients.16 The diagnosis of hypoparathyroidism is suggested by a low serum calcium level, a high phosphate level, and confirmed by a parathyroid hormone assay. Approximately 14% of diabetic patients suffer from anxiety disorders, and elevated anxiety symptoms are seen in up to 40% of diabetics. There is evidence that diabetics who are treated with antianxiety medication not only experience a reduction in anxiety but also a decrease in glycosylated hemoglobin levels and high-density lipoprotein.17 Many patients with anxiety, somatoform, or characterologic disorders are convinced that they have reactive hypoglycemia. A normal fingerstick blood glucose analysis done during an attack can exclude this diagnosis. Pheochromocytomas are rare tumors that produce elevated levels of catecholamine in the body. Common symptoms include paroxysmal hypertension, headache, anxiety, sweating, flushing, abdominal and back pain, and vomiting and diarrhea. Pheochromocytoma attacks can present just like panic attacks and can be precipitated by emotional stress. While the sweating associated with pheochromocytoma attacks involves the whole body, the sweating in panic attacks is more likely to be confined to the hands, feet, and forehead. Elevated urinary catecholamine or plasma metanephrine can confirm a pheochromocytoma.18 Hyperthyroidism is one of the most frequently encountered endocrine diseases associated with anxiety. As with panic disorders, hyperthyroidism is associated with acute episodic anxiety. Thyrotoxicosis causes anxiety, palpitations, perspiration, hot skin, rapid pulse, active reflexes, diarrhea, weight loss, heat intolerance, proptosis, and lid lag.19 Psychiatric presentations are often the first sign of hypothyroidism, occurring as the initial symptom in approximately 2 to 12% of reported cases along with organic mental deficits. Anxiety and progressive mental slowing associated with diminished recent memory and speech deficits with diminished learning ability are the characteristic initial progression of symptoms. The development of severe anxiety disorders in hypothyroid states are more related to the rapidity of change of thyroid hormone levels than the absolute levels encountered. In general, checking the serum thyroidstimulating hormone and free thyroxine levels will suffice in the ED to make the diagnosis of thyroid emergencies.20
Respiratory Diseases Most conditions causing airway compromise or impairing gas exchange would never be mistaken for a psychiatric disorder. However, certain conditions that cause hypoxemia or hypercarbia may present with significant anxiety, and up to a third
Chapter 110 / Anxiety Disorders
Panic attack is a discrete period in which there is a sudden onset of intense apprehension, fearfulness, or terror, often associated with feelings of impending doom. Agoraphobia is an anxiety about, or avoidance of, places or situations from which escape might be difficult. Panic disorder with agoraphobia is characterized by both recurrent, unexpected panic attacks and agoraphobia. Agoraphobia without a history of panic disorder is characterized by the presence of agoraphobia and panic-like symptoms without a history of unexpected panic attacks. Specific phobia is characterized by clinically significant anxiety provoked by exposure to a specific feared object or situation, often leading to avoidance behavior. Social phobia is characterized by clinically significant anxiety provoked by exposure to certain types of social or performance situations, often leading to avoidance behavior. Blushing is the cardinal characteristic symptom. Obsessive-compulsive disorder is characterized by obsessions that cause marked anxiety or distress and by compulsions that serve to neutralize anxiety. Post-traumatic stress disorder is characterized by experiencing of an extremely traumatic event, accompanied by symptoms of increased arousal and by avoidance of stimuli associated with trauma. Acute stress disorder is characterized by symptoms similar to those of post-traumatic stress disorder that occur immediately in the aftermath of an extremely traumatic event. Generalized anxiety disorder is characterized by at least 6 months of persistent and excessive anxiety and worry. Anxiety disorder caused by a general medical condition is characterized by prominent symptoms of anxiety that are judged to be a direct physiologic consequence of a general medical condition. Substance-induced anxiety disorder is characterized by prominent symptoms of anxiety that are judged to be a direct physiologic consequence of a drug of abuse or medication or toxin exposure. Anxiety disorder not otherwise specified is included for coding (1) disorders with prominent anxiety or phobic avoidance that do not meet criteria for specific anxiety disorders and (2) anxiety symptoms with inadequate or contradictory information.
be used to provide symptomatic relief to patients who experience chest pain. Studies have shown that benzodiazepines reduce anxiety, pain, and cardiovascular activation. It is hypothesized that secondary to the reduction in circulating catecholamines, benzodiazepines may cause coronary vasodilatation, prevent dysrhythmias, and block platelet aggregation.15
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1448
of the patients with chronic obstructive pulmonary disease meet the criteria for anxiety disorder.21 Asthma is characterized by episodic attacks of dyspnea and often anxiety. Anxiety can also precipitate and prolong asthma attacks. Patients who have severe asthma are twice as likely to have an anxiety disorder and almost five times as likely to have a phobia compared with nonasthmatics. Severe asthmatics are almost five times as likely to have a panic disorder and about four times as likely to have a panic attack. Acute dyspnea secondary to asthma is easily differentiated from pure panic in that there is good air movement with normal lung sounds in a patient experiencing a panic attack, but studies consistently show that anxiety disorders increase asthma morbidity and mortality.22 Shortness of breath is a common complaint in the ED. When accompanied by anxiety, panic attacks or anxiety disorders may be high on the differential diagnosis. The clinician must always fully evaluate these patients, as these are often the complaints of a patient presenting with a pulmonary embolism. Acute shortness of breath in any patient should never be dismissed lightly, especially since pulmonary embolus can present with only shortness of breath as the major symptom. These patients can be distinguished by close attention to history and examination, assessing risk factors for thromboembolic disease and use of basic investigations (e.g., pulse oximetry, electrocardiogram, chest radiography, arterial blood gas analysis, and D-dimer) and further tests as indicated.23
Neurologic Disorders Many neurologic conditions are associated with anxiety symptoms.24-26 Temporal lobe seizures, complex partial seizures, tumors, arteriovenous malformation and ischemia or infarction all have been reported to present with panic attacks. Anxiety often accompanies a transient ischemic attack and may be the major symptom on presentation if the transient ischemic attack has resolved by the time the patient reaches the ED. In Huntington’s disease, anxiety has been reported as the most common prodromal symptom. Anxiety occurs in up to 40% of patients with Parkinson’s disease and up to 37% of patients with multiple sclerosis. Similarly, anxiety symptoms have been noted to be common in moderate Alzheimer’s disease. The coexistence of anxiety disorders plays an important role in the prognosis and impairment of patients who have had cerebral vascular accidents with neurologic sequelae. Anxiety and depression are associated with left-hemispheric strokes and anxiety alone with right-hemispheric strokes. And finally, anxiety disorders have also been reported in the aftermath of traumatic brain injury.27
Drug Intoxication and Withdrawal States Amphetamines, cocaine, and sympathomimetic drugs are abused for their stimulant and mind-altering properties. Amphetamine use has exploded over the past decade, and cocaine use is still the drug of choice in many large cities. Patients often present to the ED agitated, anxious, or aggressive when these drugs are taken in large doses and with prolonged use. Caffeine is a common stimulant, and energy drinks and gourmet coffee represent a constantly growing market in the United States. These drinks are packed with caffeine and the herbal equivalent guarana as well as ginkgo biloba. Studies indicate that 240 to 300 mg of caffeine per day should be the upper limit of healthy consumption. Many of these energy drinks contain that amount in a single serving.25 Lower doses of caffeine can be pleasantly stimulating, but higher doses cause hyperalertness, hypervigilance, motor tension, tremors, gastrointestinal
distress, and anxiety. The acute symptoms of caffeine intoxication and generalized anxiety disorder are almost identical. Stimulants such as Ephedra and ephedrine-based compounds were found in many dietary supplements and listed by the herbal name of Ma Huang. Despite the Food and Drug Administration ban on Ephedra-containing compounds in 2004, access still exists over the Internet. Many illicit drug users who use marijuana believe that the drug reduces their anxiety. But some experience a depersonalization that provokes severe anxiety, fearfulness, and agoraphobic symptoms. Lysergic acid diethylamide (LSD), phencyclidine (PCP), and ecstasy are hallucinogens that can produce anxiety and paranoia from chronic use or “bad trips.” Flashbacks affect some users of LSD, where the person may experience the symptoms of anxiety and paranoia weeks or months after use.26 Sedative-hypnotic drugs (e.g., benzodiazepines, barbitur ates, meprobamate, methaqualone, chloral hydrate, and par aldehyde) are taken to relieve anxiety or sleeplessness, but their discontinuation can cause sedative withdrawal and rebound anxiety. The severity of the withdrawal syndrome depends on the drug, dosage, duration of use, and speed of elimination. In general, the intermediate-acting sedative-hypnotics (4–6 hours) cause the worst withdrawal symptoms. These symptoms include hyperalertness, motor tension, muscle aches, agitation, anxiety, insomnia, hyperactive reflexes, postural hypotension, tremulousness, nausea, vomiting, convulsions, delirium, and even death. Benzodiazepine withdrawal is rarely fatal but can be very unpleasant. In anxious patients, severe rebound anxiety can occur after a few weeks’ use of recommended therapeutic doses. Lorazepam and alprazolam are short-acting agents and their abrupt discontinuation frequently causes panic attacks within 1 to 2 days. With longer-acting agents, withdrawal symptoms typically peak in about one week. Normal people may experience this rebound as stimulating. Although antidepressants are rarely abused, their abrupt withdrawal can also cause an abstinence syndrome of insomnia, vivid nightmares, and extreme anxiety.27 Alcohol withdrawal, in alcohol-dependent individuals or heavy binge drinkers, can appear 6 to 12 hours after the last drink or significant reduction in consumption of alcohol. Patients often have detectable alcohol still in their systems at this time. Anxiety is one of the first and most prominent symptoms and is seen within 24 to 48 hours of the withdrawal state.28
Anxiety in Primary Psychiatric Disorders Even in patients with known mental illness, a panic disorder is a diagnosis of exclusion because several mental illnesses cause panic attacks as a secondary manifestation. The presence of panic often influences the treatment and outcome of the primary mental illness. Panic attacks can occur as part of a bipolar (manic-depressive) disorder, in either the manic or the depressed phase. In manic and hypomanic disorders the patient’s predominant affect is usually cheerful and euphoric but may also be dysphoric with irritability and extreme anxiety of panic proportions.29 Early in the course of schizophrenia, a patient will often experience panic attacks. Fearfulness, tension, agitation, immobility, disorganized thinking, dilated pupils, extreme insecurity, suspiciousness, and delusions of reference and persecution may characterize schizophrenic panic attacks. The hallucinations often have derogatory accusative content. Social anxiety is a highly prevalent and disabling condition with schizophrenia that is unrelated to clinical psychotic symptoms.30
1449
(compulsions), such as handwashing or checking. OCD is classified as an anxiety disorder because (1) anxiety or tension is often associated with obsessions and resistance to compulsions, (2) anxiety or tension is often immediately relieved by yielding to compulsions, and (3) OCD often occurs in association with other anxiety disorders. In summary, the obsessions and intrusive thoughts increase anxiety and the compulsions and repetitive behaviors decrease anxiety but with significant disruption of one’s life.9
■ MANAGEMENT Initial Evaluation The patient should first be placed in a quiet area for evaluation. Some patients calm down when removed from the ED environment. If the emergency physician encounters difficulty in calming the patient, supportive family members may help. Often a known and trusted face helps anxious patients make order out of their inner turmoil. Prior discussion and clarification with the family are essential to elicit their support.1 Once the patient is calmed, a more formal evaluation can begin. The emergency physician should ask open-ended questions and observe the patient’s responses carefully. Questions regarding drug or alcohol use should be delayed until rapport has been established. Reassurance should not be premature, because this important interventional technique is more effective when it is delayed until after the patient’s specific concerns are clarified.1 The extent of the medical workup for significant anxiety will vary depending on the age and health status of the patient, the nature of the anxiety, and the range and severity of associated symptoms. The emergency physician should consider the anxiogenic effects of medications, including beta-adrenergic agonists, theophylline, corticosteroids, thyroid hormones, and sympathomimetics. Potential contributory medical illness (e.g., thyroid dysfunction, hypoglycemic episodes in diabetes, hyperparathyroidism, dysrhythmias, chronic obstructive pulmonary disease, seizure disorders), substance use (e.g., caffeine, amphetamines, cocaine) and withdrawal states (e.g., alcohol, sedative-hypnotics) must also be considered. If a somatic concern is the major component of the acute anxiety attack, a physical examination with particular attention to the area of complaint is important, even when there is overwhelming evidence of a functional etiology to the patient’s complaints. Anxiety attacks are stressful experiences in themselves and can cause deterioration in marginally compensated organ systems. Careful evaluation reassures the patient and avoids the problem of a premature “medical clearance.” Abnormal vital signs should immediately alert the emergency physician to an organic cause of the anxiety symptoms.9 Because of the physical nature of the symptoms, patients with anxiety and panic attacks often seek treatment in the ED rather than in a psychiatric setting. A calm manner and willingness to listen usually relieves some of the patient’s initial anxiety. An anxiety or panic reaction may be precipitated by the loss of a significant relationship, a job, a living situation, or self-esteem, as well as by physical illness or injury. Once the patient describes a trigger event, the emergency physician should restate it, as if experiencing a similar situation. This gives the patient authoritative approval for expressing embarrassing feelings. A patient who has frequent anxiety reactions is usually suggestible and will respond to reassurance. Conversely, an anxious or unsympathetic physician will only compound the problem.1 Even an apparently calm patient may communicate anxiety through worried looks, nervousness, pressured speech, or
Chapter 110 / Anxiety Disorders
Patients with somatoform disorders report a variety of somatic symptoms, including panic attacks, and 68% report a history of anxiety. Patients claim to have most of the physical symptoms they are asked about, even when evidence excluding illness is presented to the patient. Fear and anxiety initiate, facilitate, and maintain many of the symptoms encountered in the somatoform patient. Patients with “pure” anxiety disorders tend to be hypochondriacal, while those with somatization are more likely to improve transiently on active medication or placebo but rarely respond so well that they stop seeking unnecessary medical attention. Patients with panic disorders, however, seek at least as much psychiatric attention as those with somatoform disorders.31 Approximately 50% of patients with a primary panic disorder develop major depression and many others are bothered by some degree of depression in mood. Twenty percent of patients with depression have panic attacks, and the remainder have considerable anxiety. Depression with panic attacks responds less well to treatment. Agitated depression with anxiety and psychosis, sometimes called “involutional melancholia,” responds well to electroconvulsive therapy. Depression with anxiety and hostility responds well to antidepressants but benzodiazepines can exacerbate symptoms.32 Post-traumatic stress disorder is an anxiety disorder characterized by the reexperiencing of an extremely traumatic event. The symptoms are closely related to and worsened by reminders of the trauma. The “flashbacks,” in which patients reexperience the original trauma, can have the same symptoms as panic attacks. These patients often avoid crowds or social situations.33 A panic disorder is one of the easier psychiatric diseases to feign because most of the symptoms can be duplicated by intentional hyperventilation. Functional hyperventilation can be distinguished from organic hyperventilation by its irregularity and interruptions by sighs. When in doubt, formal psychiatric evaluation is indicated, particularly before prescribing a potentially dangerous or addictive drug therapy. A phobia is an irrational fear that results in avoidance and is considered normal in children. The objects of fear tend to be things that seem dangerous to a child (e.g., spiders, snakes, bats, cats, enclosed places, the dark, open spaces). Phobia becomes a disorder when it interferes with day-to-day function in an individual’s life. A social phobia is characterized by clinically significant anxiety provoked by exposure to a specific feared object or situation, often leading to avoidance behavior. Social phobias prevent a patient from doing such activities as public speaking, performing, visiting, using public showers or restrooms, or eating in public places. Agoraphobia is a fear of being alone in public places. Nearly 75% of agoraphobic patients have panic attacks.34 Those with panic attacks are more likely to seek treatment, whereas those with uncomplicated agoraphobia tend to stay at home. Agoraphobia without panic attacks may not differ fundamentally from other simple phobias. Most panic disorder patients have multiple phobias, including agoraphobia. The latter is believed to result from the panic patient’s increasing attempts to avoid places or situations in which the panic attacks would be particularly inconvenient or difficult to control. Agoraphobic patients particularly avoid places from which escape would be difficult (e.g., bridges, crowded theaters). When they do attend theaters, they favor seats on the aisle and near the door. Panic attacks in agoraphobic patients are more likely to include fear of losing control, whereas those not associated with agoraphobia are more likely to include dyspnea and dizziness.35 An obsessive-compulsive disorder (OCD) is characterized by recurrent, obtrusive, unwanted thoughts (obsessions), such as fears of contamination, and compulsive behaviors or rituals
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1450
covert assaults on the physician’s competence. In turn, the physician may empathetically respond to the patient’s hidden anxiety by also becoming anxious. This is a strong clue that the patient’s anxiety is real and significant. Without this selfawareness, physicians may focus on a patient’s physical symptoms rather than on the irrational anxiety. The pressure to see patients quickly and to move them out of the ED expeditiously may result in limited interactions with patients, the misdiagnosis of anxiety disorders, and excessive and unnecessary medical workups. A careful medical evaluation is important, but excessive focus on unlikely illness suggests to the patient a reason to worry, avoids recognition of crucial psychological factors, and may increase anxiety and the severity of symptoms. After organic illness, medications, and obvious psychiatric causes of the acute anxiety state have been ruled out, the physician should determine whether the anxiety is endogenous or exogenous. If the anxiety arises spontaneously without an identifiable stress, is unpredictable, and is accompanied by agoraphobia, an endogenous component is likely to be present. Such patients should be referred to a psychiatrist for evaluation and treatment. If the anxiety appears to be related to an identifiable external event or circumstance, the anxiety is exogenous and patients should be encouraged to discuss their feelings with a mental health worker. Talking about fears allows anxious patients some sense of mastery and control over events. These patients often require ongoing advice, support, and assistance in mobilizing necessary resources from family members, friends, and social agencies to achieve realistic expectations. Anxiety is common in elders, with prevalence rates conservatively estimated at 10%, with higher rates in patients with chronic illness. Anxiety disorders may be the most common psychiatric ailments experienced by older adults, but that age group is the least studied of all patients.36 Older patients with anxiety often have somatic complaints. These patients require a careful investigation for underlying medical illness, other psychiatric conditions, and the use of over-the-counter and prescription drugs.
Pharmacologic Treatment Before medication is prescribed, education of the patient about their illness is a key component in the treatment of anxiety disorders. Patients are often worried and confused about their illness. Reassurance that they are not alone, education about what to expect and that therapy is available, and involvement of family are all critical pieces in the treatment of anxiety. Use of intravenous medication is rare but may be necessary when an anxiety state renders a patient so helpless and out of control that there is a significant threat of safety to self or others. Intravenous medication is also appropriate for the anxious patient experiencing a significant medical illness or undergoing a medical procedure. Lorazepam in small increments every 20 minutes can be helpful in alleviating the anxiety associated with substance withdrawal states. Midazolam is frequently used to reduce anxiety and increase amnesia for ED procedures. Selective serotonin-reuptake inhibitors (SSRIs) have become the first-line treatment for most anxiety disorders because of their broad spectrum of efficacy and good tolerance by most patients. SSRIs have a lower potential for dependence and are safer than the previous classes of antidepressants and anxiolytics. This class of drugs includes fluoxetine, sertraline, fluvoxamine, paroxetine, citalopram, venlafaxine, and sertraline. Improvement is usually seen in 3 to 4 weeks, and the medication may have to be adjusted if no improvements in anxiety are seen. It is important to start the patient on low doses of SSRIs as an initial increase in anxiety may be seen.
In the past few years, emergency physicians and the public have become increasingly concerned about the growing use of benzodiazepines in the United States. Over 1 million Americans are physically dependent on tranquilizers. When tranquilizers are given in place of understanding, support, and intrapersonal therapies, patients are taught to rely on the external support of a pill rather than on inner resources.8 Benzodiazepines can be prescribed for motivated patients with acute exogenous anxiety for time-limited stress. Patients who are cooperative, employed, educated, married, and aware that their symptoms have a psychological basis are more likely to respond. Benzodiazepines are an attractive alternative to SSRIs when an immediate reduction of symptoms is desired because of the delayed response with SSRIs or a short-term treatment is needed. Benzodiazepines can be given in one or two daily doses to make use of their short half-lives; alternatively, a bedtime dose may minimize daytime sedation and still manifest a daytime anxiolytic effect. Benzodiazepines should not be prescribed for more than a week. Patients who do not improve within a week are unlikely to benefit from the drug. Patients with a history of alcoholism or drug abuse, who are excessively and emotionally dependent, or who become anxious in response to normal stress are at greater risk of drug dependency and are not good candidates for this treatment from an emergency physician. Dependence and abstinence syndromes have been reported to occur with low doses of tranquilizing drugs, especially if they are taken for more than 8 months. Short-acting benzodiazepines (e.g., lorazepam, oxazepam) should be prescribed at low dosages for patients with liver disease, organic brain syndrome, and those taking medications that either depress central nervous system function or inhibit benzodiazepine metabolism and clearance. Withdrawal rebound symptoms are more common with discontinuation of benzodiazepines than with other antianxiety treatments. Short-acting benzodiazepines produce a more severe abstinence syndrome when they are stopped abruptly, and thus many physicians prefer the longer-acting benzodiazepines.7 For some patients, switching from a short-acting agent (e.g., alprazolam) to a long-acting agent (e.g., clonazepam) can be helpful before initiating a taper. Buspirone is a nonbenzodiazepine tranquilizer used in the treatment of generalized anxiety disorder. Buspirone does not appear to cause dependency, is less sedating than benzodiazepines, and tolerance does not occur at therapeutic doses. It is the therapeutic lag in efficacy of 2 to 3 weeks that has limited the use of buspirone. It has had variable and sometimes disappointing results in clinical practice, particularly when used in patients with prior exposure to benzodiazepines.36 Monoamine oxidase inhibitors (MAOIs) demonstrate high effectiveness in the treatment of social phobia, panic, generalized anxiety disorders, OCD, and comorbid conditions (e.g., atypical depression). MAOIs (phenelzine and tranylcypromine) may be difficult to tolerate and require discipline and strict dietary restrictions and thus are rarely appropriate in the emergency setting. Tricyclic antidepressants (TCAs) are effective for panic disorders and generalized anxiety disorders but are ineffective for social phobias and, with the exception of clomipramine, are largely ineffective for OCD as well. TCAs have been used effectively for depressive and anxiety symptoms associated with post-traumatic stress disorder. TCAs include imipramine, nortriptyline, desipramine, amitriptyline, and doxepin. The TCAs have been supplanted by the SSRIs as first-line interventions for the treatment of anxiety and depressive disorders.37 Patients with endogenous anxiety (panic attacks with or without agoraphobia) should be referred to a psychiatrist to
1451
Table 110-1 Pharmacotherapy for Anxiety Disorders TCAs
MAOIs
BDZs
BUSPIRONE
CBT
+ + + − + +
+ + − − +/− −*
+ + + − + +
+ + + +/− +/− +/−†
− + − − − +/−†
+ + + + + +
*Clomipramine is effective. † Used adjunctively with serotonergic antidepressant. BDZs, benzodiazepines; CBT, cognitive-behavioral therapy; GAD, generalized anxiety disorder; MAOIs, monoamine oxidase inhibitors; OCD, obsessive-compulsive disorder; PTSD, post-traumatic stress disorder; SSRIs, selective serotonin-reuptake inhibitors; TCAs, tricyclic antidepressants.
establish a good therapeutic relationship before using anxiolytic medication. Benzodiazepines, tricyclics, SSRIs, and MAOIs are safe and effective in endogenously anxious patients who are under psychiatric care (Table 110-1). Recurrence rates of panic attacks are high when drug therapy is discontinued.8
Nonpharmacologic Therapy Psychotherapies may be helpful for individuals whose psychological makeup, coping style, interpersonal dynamics, and situational stressors contribute to their pathologic anxiety. The use of supportive, insight-oriented family is helpful when these factors appear prominently in the patient’s presentation.8 Cognitive-behavioral therapy is predicated on the theory that the distress and impairment associated with anxiety and panic are mediated by maladaptive cognitive responses that promote anxiety and avoidance. The core components of cognitivebehavioral therapy for panic disorder include correction of cognitive misperceptions and overreactions to anxiety symptoms, breathing retraining, muscle relaxation, as well as exposure and desensitization to phobic situations. Cognitivebehavioral therapy is very effective, but requires commitment from the patient.8,9 Meditation (e.g., Zen, yoga, transcendental) has been proposed by many authorities, but little clinical data support its efficacy in anxiety disorders. Biofeedback appears promising for the treatment of generalized anxiety disorder. Hypnotic suggestion may be effective because anxious patients tend to be cognitively scattered, unable to focus their attention, and highly suggestible. A hypnotic state can often be induced by certain stimuli.38 These nonpharmacologic techniques take anxious patients out of the future, about which they are frightened, and place them into the present. These techniques should be reinforced by the development of a physically and psychologically healthy lifestyle. A significant social support system not only protects against vulnerability to illness but also is highly anxiolytic. Regular exercise (e.g., dancing, swimming, bicycling, walking, jogging) also promotes tranquility. Encouraging activity that focuses on hand-eye-ear coordination (e.g., painting, playing keyboard, needlework) helps anxious patients regain and maintain control by bringing them into the present.1
■ DISPOSITION Many patients with anxiety-related symptoms can be effectively treated in the ED. The emergency physician can proceed with the following general measures:
1. Rule out organic illnesses as cause of anxiety. 2. Evaluate for substance abuse and medications associated with anxiety. 3. Determine whether anxiety is endogenous or exogenous. 4. Clarify what is currently frightening the patient. 5. Evaluate the patient’s capacity for self-awareness. 6. Assess techniques that have worked in the past. 7. Support coping skills. 8. Give the patient as much control over the care plan as feasible. 9. Select patients to start on a short course of benzodiazepines and educate patients about treatment. 10. Apply adjunctive techniques as appropriate for the patient’s personality and the physician’s preference (e.g., hypnotic suggestion, breathing exercises). Patients with a panic disorder associated with suicidal or homicidal ideation or with severe depression require urgent psychiatric attention and admission to the hospital. Other patients with suspected endogenous or severe exogenous anxiety disorders should be referred for psychiatric evaluation. The Anxiety Disorders Association of America can be contacted (240-485-1001) for a national registry of clinicians and treatment programs specializing in anxiety disorders or can be found online at www.adaa.org.
KEY CONCEPTS ■
Anxiety may accompany the onset of serious disease, may have significant metabolic demands, and may cause a marginally compensated organ system to fail. ■ As many as 42% of patients thought to have anxiety disorders are later found to have organic disease. ■ Anxiety caused by physical illness is usually suggested by the patient’s physical findings but may require adjunctive testing. ■ Anxiety affects at least 10% of elderly patients. ■ Intravenous medication may be necessary for patients who are a significant threat to themselves or others and for anxious patients with significant medical illness. ■ Limited benzodiazepine therapy may be helpful for select patients with exogenous anxiety. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 110 / Anxiety Disorders
Panic disorder GAD Social phobia Specific phobia PTSD OCD
SSRIs
Chapter 111
Somatoform Disorders
Adria O. Winter and Thomas B. Purcell
■ BACKGROUND Occasionally patients present to the emergency department (ED) with a myriad of physical symptoms but no apparent physical disease. This presents a vexing conundrum for the physicians who are wary of missing subtle presentations of physical disease. Even when a somatoform disorder is strongly suspected, emergency physicians are reluctant to attribute physical complaints to functional etiologies.1 In one study, 20% of 1500 consecutive patient visits to primary care physicians involved somatization.2 The prevalence of somatization has risen over the past 30 to 40 years, possibly because of a general decline of patients’ tolerance for mild and self-limited ailments.3,4 It is not unusual for physicians to feel that patients with somatoform disorders have nonlegitimate disturbances, because their presentation and diagnosis are more generally felt to be the province of psychiatry and not emergency medicine. Nevertheless, proper diagnosis and treatment of patients’ somatoform disorders is essential, as misidentification and mismanagement unnecessarily prolongs patients’ distress and adds to the overall burden on the health care delivery system.
■ CLINICAL FEATURES Somatization refers to a tendency to experience and communicate psychological distress as physical symptoms in the absence of identifiable pathology.5,6 Patients with somatization disorders seek medical attention because they are convinced that their symptoms reflect real physical disease.7–11 In direct contrast to malingering and factitious disorder, the symptoms are neither feigned nor under the voluntary control of the patient.12,13 Somatization is most often associated with concurrent depressive and anxiety disorders.14 The term somatoform disorders embraces all disorders that have somatization as a common factor. In some cases, a demonstrable physical disorder does exist, often iatrogenic, but the complaints are out of proportion to the physical findings.8 Almost any complaint involving any body system may occur, sometimes suggesting fashionable diseases recently popularized by the media.11 For example, the environmental somatization syndrome refers to individuals convinced that their symptoms are caused by exposure to chemical or physical components of their environment, such as poisonous substances, electromagnetic fields, or ergonomic stress attributed to repetitive movements.15 In general, it is the multiplicity of 1452
symptoms, rather than the specific symptom, that is most indicative of somatization.16,17 In women with more than five and in men with more than three unexplained somatic complaints, the likelihood of a diagnosable psychiatric disorder doubles.16,18,19 Certain categories of patients, such as women with chronic pelvic pain, have a high rate of sexual or physical abuse in childhood.20 Somatizers have difficulty describing their feelings in words, a phenomenon termed alexithymia (“without words for mood”), resulting in alternative (somatic) forms of expression.21 They steadfastly insist that their symptoms are caused by serious physical disorders even in the presence of conclusive evidence to the contrary.8 Somatization may be unconsciously motivated by a desire to assume the “sick role” and seek the privileges afforded to a sick person by society, such as release from normal obligations and absolution from blame for their condition.21,22 Individual somatic complaints such as headache, low energy levels, and recurrent abdominal pain are common in children and adolescents in the general population and generally are not associated with significant social or emotional impairment. Pronounced polysymptomatic somatization, although uncommon in this age group, may occur in late childhood or early adolescence, and these patients are at increased risk for later development of major depression, anxiety disorders, panic attacks, and drug and alcohol abuse.23,24 Interestingly, most physicians have experienced somatization, since up to 80% of all medical students become convinced they have a disease.25 The somatoform disorders were introduced in the Diagnostic and Statistical Manual of Mental Disorders, 3rd edition (DSMIII), and updated in DSM-IV as the initial classification of the “hysterical” and “hypochondriacal” neuroses and then their updated conceptualization. There is ongoing controversy whether the somatoform disorders require a reclassification since the diagnosis does not explain the etiology, nor does it predict treatment response.26 The somatoform disorders may be subdivided into four specific disorders, each different in terms of clinical presentation and management: (1) somatization disorder, (2) conversion disorder, (3) pain disorder, and (4) hypochondriasis.
Somatization Disorder Historically referred to as hysteria and neurasthenia, somatization disorder was given the eponym Briquet’s syndrome by Guze in 1975 to avoid the pejorative implications associated with the traditional terms.27,28 The malady is chronic or
1453
Criteria for the Diagnosis of
1. There must be a history of medically unexplained physical symptoms beginning before the age of 30 years. 2. A history of all of the following: a. Pain related to at least four different sites (e.g., head, abdomen, back, joints, chest) or functions (e.g., during menstruation, during urination) b. At least two gastrointestinal symptoms other than pain c. At least one sexual or reproductive symptom other than pain (e.g., sexual indifference, irregular menses) d. At least one symptom or deficit suggesting a neurologic condition not limited to pain (e.g., paralysis, lump in the throat, blindness) 3. Either the above symptoms must not be explainable by any known medical condition or, when there is a related general medical condition, the complaints or impairment must be out of proportion to what might be reasonably expected. 4. The symptoms must not be intentionally produced or feigned.
repetitive, dating from young adulthood, with numerous physical symptoms and complaints involving a variety of organ systems but few or no physical findings to explain those symptoms.29–32 The diagnosis of somatization disorder requires several criteria (Box 111-1).12 This diagnosis is rarely made in the ED, even though it may be suspected, because the proper investigation of this disorder involves time-consuming interviews and may require four to six visits before the establishment of a definitive diagnosis. Many patients do not meet the criteria for a strict diagnosis of somatization disorder yet still have considerable functional impairment, psychological distress, and excess health care utilization.1,33 A short list of symptoms provides a rapid screen for this disorder.34 The seven symptoms that best discriminate between patients with and without somatization disorder are (1) dysmenorrhea, (2) the sensation of a “lump” in the throat, (3) vomiting, (4) shortness of breath, (5) burning in the sex organs, (6) painful extremities, and (7) amnesia lasting hours to days. Among patients displaying at least two of these symptoms, the diagnosis of somatization disorder is correctly predicted with a sensitivity of 93% and a specificity of 59%. When four or more of these seven symptoms are present, the specificity rises to 100%. Although this screening may identify patients at high risk for somatization disorder, such patients should still be evaluated thoroughly to confirm the diagnosis and exclude organic disease. True somatization disorder is relatively uncommon, having a prevalence of 0.06 to 2% among the general population and up to 9% among hospitalized patients.7,12,19,34 It tends to run in families and is rarely diagnosed in men,9,10,12,35,36 although in some cultures the prevalence in men and women may be the same.19 The typical patient is a woman in her 40s who has a 25- to 30-year history of multiple vague complaints, usually headache, dizziness, nausea and vomiting, syncope, abdominal pain, bowel trouble, fatigue, palpitations, dyspareunia, and dysmenorrhea.20,32,36,37 Symptoms usually date back to the patient’s teens and 20s, with menstrual complaints being common in these age groups.7,12,32 Only 33% of patients recover during 10- to 20-year followup,38 and new symptoms requiring medical attention tend to
Conversion Disorder Also known as hysterical neurosis, conversion type, the rare conversion disorder is characterized by the sudden dramatic onset of a single symptom, typically simulating some nonpainful neurologic disorder for which there is no pathophysiologic or anatomic explanation.7,46 In contrast to somatization disorder, conversion disorder typically revolves around a single physiologically impossible condition. The symptoms, generally conforming to the patient’s own idiosyncratic ideas about illness, are not under the patient’s voluntary control. Some symptoms provide gratification for unconscious dependency needs; other symptoms provide escape from painful external emotional stimuli (e.g., hysterical paralysis in battle).40,41 Although the symptoms may have a symbolic relationship to the precipitating factors, this is often not the case.31 The most common conversion symptoms are voluntary motor or
Chapter 111 / Somatoform Disorders
BOX 111-1 Somatization Disorder
surface at least every year.7 Despite a “lifetime of suffering,” the life span of these patients is normal.39 Somatization disorder is associated with lower socioeconomic groups, alcoholism and other addictions,13,36,40 and poor education; fewer than 25% graduate from high school.19,26 Many have occupational, interpersonal, and marital problems.7,36 The health care utilization and functional impairment of these patients are astounding. Expenditures for physician services are 14 times greater and overall health care expenditures nine times greater than for unaffected patients. The typical patient spends 7 days in the hospital each year and 7 days sick in bed each month (compared with less than a half-day for a control population). More than 82% stop work because of their health.36 When the diagnosis is recognized, however, medical resource use among these patients tends to normalize.41 Patients with somatization disorder describe their symptoms in dramatic exaggerated fashion using colorful language, with great detail about how their lives have been disrupted. They usually admit to being sickly throughout life. Although extensive, their narrative suggests no clear diagnostic constellation. Patients offer detailed accounts of multiple prior medical encounters, termed “doctor shopping,” and often display multiple abdominal scars because they undergo two to three times the number of surgeries of other patients.1,36 Their medical records have numerous and exotic test results, and they faithfully consume an impressive array of medications acquired from multiple primary physicians and specialists. They report allergies to a comprehensive list of antibiotics and analgesics. These patients are often emotional and vain, exhibit limited interpersonal skills, and have few close personal relationships.32 They are typically dependent and highly manipulative. Of these patients, 68% fulfill the criteria for histrionic personality disorder.42 Somatization disorder may be closely related to some anxiety and affective disorders; more than 80% of patients with somatization disorder report a lifetime history of major depression and 68% a history of anxiety.36,43–45 These individuals may threaten or attempt suicide. Completed suicide is usually associated with psychoactive substance abuse.7 Women with this disorder tend to marry men with antisocial personalities. The husband is often overly solicitous, demanding that his wife receive many clinical studies and quick, decisive action. Predictably, he usually shows some degree of dissatisfaction with physicians in general.32 Patients who do not meet the full criteria of somatization disorder but have suggestive symptoms for 6 months or longer are classified as having undifferentiated somatoform disorder, which is treated similarly to somatization disorder.12
1454
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
BOX 111-2 Presentations of Conversion Disorder 1. Motor disturbances: a. Tremors (that worsen when attention is called to the movements) b. Seizures (wild, thrashing, writhing, often mimicking copulation) c. Paralysis or paresis (often a monoplegia, stockingglove weakness, with normal reflexes and limb circumferences) d. Aphonia (wherein the patient can whisper and cough normally and vocal cords move normally with respiration) e. Coordination disturbances 2. Sensory disturbances: a. Anesthesia (the patient may not find this symptom disturbing) b. Blindness and tunnel vision16,17 3. An occasional patient may present with other nonneurologic symptoms such as vomiting or pseudocyesis.13
sensory functions and are therefore called pseudoneurologic (Box 111-2).12 The most common ED presentations are pseudoseizures, syncope or coma, and paralysis or other movement disorders.47 Most patients are women, except for those in military service and industrial accidents.12,41 Conversion disorder typically appears in adolescence and early adulthood and is more common among lower socioeconomic groups. Symptoms tend to be of sudden onset, waxing and waning in response to environmental stresses.7,9,10,31,40 The history may show similar symptoms in the past, as well as anxiety, depression, phobias, and sexual disturbances.31 Up to 29% of patients have a history of past psychiatric illness.47 Patients describe their symptoms with a lack of appropriate concern about their profound bodily dysfunction, termed la belle indifference, although, this presentation is not necessary for diagnosis of conversion disorder, since it may be absent in over 50% of patients and is also seen in organic disease.48
Pain Disorder Also termed somatoform pain disorder,7 this condition is similar to conversion disorder in that stressful events are translated into somatic symptoms. The primary and often exclusive symptom is distressful pain that (1) is not intentionally feigned, (2) is persistent in nature, (3) limits daily function, (4) involves one or more organ systems, and (5) cannot be pathophysiologically explained.12,13,31 The pain most frequently occurs in the face, low back, neck, or pelvic area and causes significant functional impairment, ultimately becoming a major focus in the patient’s life.4,9,10,12,20 One half of all patients have some precipitating traumatic event at the outset (e.g., motor vehicle accident, industrial injury).7 Chronic pain behavior patterns are typically fixed within 3 months after the onset of symptoms, and patients who do not resume normal activities within 2 weeks deserve reevaluation and a careful psychosocial review.4 Associated features include frequent visits to physicians despite medical reassurance, excessive use of analgesics, requests for surgery, and eventually the role of permanent invalid after the pain has forced the patient to discontinue gainful employment.4,7 Onset occurs most often in the 30- to 50-year-old age group but can occur at any age. Symptoms such as headaches or
musculoskeletal pain are more likely in women.7,9,12 The pain often approximates real pain from physical disease that the patient has experienced in the past (e.g., the patient with a history of pancreatitis may develop recurrent epigastric pain when stressed). Frequent surgical intervention may produce multiple and genuine iatrogenic pain symptoms.20
Hypochondriasis The term hypochondriasis comes from regio hypochondriaca, a Latin term referring to the upper lateral regions of the abdomen inferior to the costal cartilages, especially the area of the spleen, which early physicians presumed to be the seat of this disorder. Hypochondriasis has four characteristics: (1) physical symptoms disproportionate to demonstrable organic disease; (2) a fear of disease and a conviction that one is sick, leading to “illness-claiming behavior” (a compulsive insistence on being considered a physical cripple); (3) a preoccupation with one’s own body; and (4) persistent and unsatisfying pursuit of medical care (doctor shopping) with a history of numerous procedures and surgeries and eventual return of symptoms.31 These unfortunate patients manifest both a heightened awareness and an unrealistic interpretation of normal physical signs or sensations, such as bowel habits, heartbeat, sweating, or peristalsis. These sensations are perceived as abnormal, noxious, and alarming, a phenomenon known as amplification.35 These aberrant perceptions result in a chronic morbid preoccupation with bodily functions and a lingering fear of having a disease despite medical reassurance.7,12,13,31 A distinguishing feature of hypochondriasis is that the patient’s symptoms do exist and often are confirmed by physical examination, but the patient exaggerates and misinterprets them. Hypochondriasis is relatively common. Its prevalence in general medical practice ranges from 4 to 9%.12 It has a peak incidence among men in their 30s and women in their 40s, affecting men and women equally.9,10,21 Hypochondriacs have an increased sense of responsibility for, and place high value on, their personal health and physical appearance. They have an acute sense of body vulnerability and a heightened aversion to death and aging.42 There is a strong correlation of hypochondriasis with major depression.45 A milder form of this disorder may be an exaggerated interest in bodily function and health (“health nuts”).49 The hypochondriac complains at length and in detail, using medical jargon. The complaints focus on the head, neck, and trunk, often in the form of pain. Hypochondriacs often believe they have lost control of their lives and have been described as “experts at defeating doctors in order to feel more powerful.”31 Consequently, physicians perceive hypochondriacal patients as more angry and hostile than other patients.50 The diagnosis may be suggested when the physician feels “frustration, helplessness, or anger associated with a wish to be rid of the patient.”22,31 Reactive hypochondriasis, or transient hypochondriasis, is an acute response to a psychosocial stress or life crisis, such as an acute myocardial infarction, terminal illness, or recent loss of a family member. In contrast to true hypochondriasis, this form is reversible and does respond to reassurance.31,49
■ DIAGNOSTIC STRATEGIES Physicians are generally unwilling to consider somatoform disorders in their initial differential diagnosis. The often dramatic presentation of symptoms creates a sense of urgency to take action, a fear of undiscovered medical illness, and a subsequent exhaustive evaluation of every complaint. Repetitive or extensive diagnostic testing rarely excludes organic disease
■ DIFFERENTIAL CONSIDERATIONS Distinguishing between the various somatoform disorders is less important than the diagnosis of treatable organic disease or the detection of anxiety and depression, which are both more common and more likely to respond to treatment. Coexistent depression or anxiety disorder should always be considered.44,53 Patients who have a relatively recent onset of somatization are more likely than patients with long-standing complaints to be exhibiting subtle signs of acute psychosis, organic brain syndrome, grief reaction, depression, or anxiety.
Depression Approximately 50 to 70% of depressed patients consult their physician for various somatic complaints.54 Depressed patients may not be aware of a depressed mood or may feel their depression is secondary to the somatic symptoms.53 As a result, depression is the psychiatric disorder most often mistaken for somatoform disorder.13 Although somatoform disorders often coexist with depression, the two conditions must be distinguished. Depression is worse in the morning, better at night, and often associated with a positive family history. The patient is reluctant to describe the symptoms and has vegetative signs of depression (e.g., sleep disturbances, decreased appetite with weight loss).52,55 Pain is a common symptom, particularly headache and pain involving the back, chest, or pelvic area.53,54 Somatoform disorders, on the other hand, are worse at the end of the day, and patients have a marked propensity to discuss their symptoms, usually do not have a family history, and show no vegetative signs.31
1455
In general, elderly patients do not have more physical symptoms than younger patients. Multiple somatic complaints should not be dismissed as a normal consequence of aging but rather considered a symptom of another underlying problem, usually depression or medical disease. Older patients may communicate somatic complaints as a way of expressing anger and provoking guilt among family members.21
Anxiety Patients with acute anxiety often hyperventilate and frequently exhibit physical signs of increased sympathetic activity. They may be hypervigilant and irritable and may show signs of muscular tension.31 They may offer a history of excessive worrying about their health, feeling “on edge” or irritable, having difficulty relaxing, or sleeping poorly or having trouble falling asleep and report symptoms of headache, tingling, dizzy spells, and diarrhea.55 Patients with somatoform disorders have a high prevalence of anxiety disorders, especially generalized anxiety disorder.56
Physical Illness When patients with somatization disorder develop true organic disease, they present similar to other patients, with specific complaints, clear chronology, and objective findings that should be appropriately investigated.32 Unfortunately, subjective reports of distress are often not dependable in these patients, and the physician must rely on more objective evidence, including the physical examination and routine laboratory tests.22 Multiple physical symptoms starting late in life are frequently the result of physical disease.7,12 In addition, patients who have a short duration of symptoms are more likely to have organic disease. Although any organic disease may be mistaken for a somatoform disorder, the occasionally bizarre and atypical manifestations and presentations of the disorders listed in Box 111-3 merit special consideration.13
Factitious Disease and Malingering Patients with somatoform disorders are not deliberately feigning illness; they are exhibiting the result of an unconscious behavior modification. For a subset of patients, they have unintentionally secured secondary gain from the sick role in the form of sympathy, encouragement, attention, support, and relief from responsibilities and challenges without significant loss of self-esteem.49 In contrast, factitious disorder and malingering are both characterized by the intentional and conscious
Organic Diseases That May Be Mistaken for
BOX 111-3 Somatoform Disorders
Endocrine disorders: hyperparathyroidism, thyroid disorders, Addison’s disease, insulinoma, panhypopituitarism Poisonings: botulism, carbon monoxide, heavy metals Porphyria Multiple sclerosis Systemic lupus erythematosus Wilson’s disease Myasthenia gravis Guillain-Barré syndrome Uremia
Chapter 111 / Somatoform Disorders
with absolute certainty, however, and may yield false-positive results, prompting further testing. Somatizing patients are more likely to have morbidity from repeated or invasive evaluations than from undiagnosed organic disease.1 Yielding to the temptation to institute further diagnostic procedures or interventions typically leads to a temporary improvement, closely followed by renewal of symptoms and mutual physician-patient disappointment. This gives rise to inevitable dissatisfaction of the patient with the physician and vice versa, leading to an unsatisfactory parting of ways and a perpetuation of the doctor-shopping cycle.51 Managed care and capitated reimbursement have created an additional quandary by restricting the supply of care in a time of rising demand for care from patients whose symptoms are relatively minor.3,16 The most effective diagnostic tool with somatizers is the interview. Evaluation starts with a thorough but focused history and, if available, a review of the patient’s medical record. This is followed by a careful problem-oriented physical examination, with meticulous inspection of the area of complaint, and simple or routine diagnostic testing, when appropriate, until attaining a reasonable level of diagnostic certainty.16 Further investigations or hospital admissions should be initiated solely on the basis of new objective signs of disease and only after confirming that the tests have not been performed. One rule of thumb in ordering laboratory tests is to do exactly what would be done if the patient were not a somatizer.16,22 However, the clinician must resist the impassioned entreaties of the patient when it is clear that further complex or hazardous studies are unlikely to be productive.9,10,16,52 Multiple medical and surgical consultations generally prove counterproductive. Hypochondriacs perceive this as a test of their claim to illness and respond simply by propagating and demonstrating symptoms with redoubled zeal.31
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1456
simulation or production of disease (see Chapter 112). Because such deception is difficult to uncover in the ED, these patients are often mistaken for having a somatoform disorder.
■ MANAGEMENT The symptoms of conversion disorder may provide a protective coping value for the patient, and the physician should be cautious about removing them without first providing adequate psychological support and treatment. Otherwise, new symptoms may arise to replace previous ones. The external precipitating stress or cause of anxiety should be removed if possible. These patients require psychiatric evaluation and management,9,10 and psychiatric consultation in the ED can be beneficial.47 Recurrence is common, but the prognosis associated with an individual episode of conversion disorder is good and the likelihood of recovery from symptoms exceeds that of other somatoform disorders.12,21,41 Factors associated with a good prognosis include (1) good premorbid health, (2) absence of organic illness or concomitant major psychiatric syndromes, (3) acute and recent onset, (4) definite precipitation by a stressful event, and (5) presenting symptoms of paralysis, aphonia, or blindness.12
Reassurance Young patients with no underlying medical or psychiatric illnesses who present with somatization in response to a clear psychosocial stress can often be reassured successfully with an appropriate explanation of their symptoms. Patients with chronic somatization, however, perceive this as an official denial of their sick role and are almost invariably unwilling to accept reassurance. Because they desire the acknowledgment and recognition that come with the designation of illness, which they feel is rightfully theirs, they are disappointed when no pathologic condition is discovered. Conversely, they are elated when given a diagnosis, but they resist recovery because subconsciously the “specter of cure” poses a threat to their sick role.49 Accordingly, attempts to cure the condition are countered with side effects, allergic reactions, and new symptoms. Such patients require another management strategy.
Legitimization of Symptoms Most patients with chronic somatization interpret a psychological explanation for their symptoms as an accusation of lying or feeblemindedness. It is important to convince them that the physician believes in their symptoms and will not try to “talk them out of it.” The priority is to listen and truly understand what the patient is feeling and trying to convey. Suffering is always a subjective phenomenon and, in that sense, is genuine in these patients.22 The physician should convey empathy for the patient’s physical discomfort. If the physician acknowledges the legitimacy of the claim to illness and assures the somatizer of ongoing care, limits may be set on the patient’s illness behavior.9,10,16,31 Patients should be allowed to tell their story without interruption. They should be told that they have an illness that causes them to experience many symptoms but that these symptoms will not lead to medical deterioration.1,57 The physician should offer only guarded projections regarding chances for complete “cure” of the condition. Ironically, this may be better received by these patients than overly optimistic assurances because the former serves to safeguard their sick role and shifts the physician away from an adversarial position.16,52
Diagnosis Diagnostic labels are of critical importance for somatizers, but the precise meaning of the term should be clarified for the patient to avoid misinterpretation. Explanations for symptoms that incorporate somatic responses and descriptions such as hyperventilation, tension headache, muscle tension, muscle strain, chest wall muscle spasm, or stress may be better accepted than purely psychiatric diagnoses. This reassures the patient that the physician shares the belief that the symptoms result from socially acceptable ailments while allowing more in-depth explanations that incorporate the relationship of bodily function to psychological stress. This, in turn, serves as a preparation for future psychiatric consultation or psychotropic medication.8–10,52 At times, the best approach may be to share the diagnostic uncertainty with the patient, using such terms as “atypical pain” or “multiple complaints following injury.” On a broader scale, managed care organizations must be encouraged to educate their enrollees about the process of somatization, the negative side effects of medications and other interventions, and the range of bodily symptoms in healthy people.3
Medications Patients with somatoform disorder have a high affinity for medications and are reluctant to discontinue drugs, even those with no benefit.31 Physicians should avoid drugs that produce an abstinence syndrome or dependence and those that cannot be safely continued indefinitely.32 Pain medications, if given, should be prescribed for regular intervals, not “as needed.”31 Patients with somatoform pain disorder may benefit significantly from treatment with antidepressants, including tricyclic antidepressants.58 Patients with somatization disorder with major depression may also improve with pharmacologic management of the depression.45 Therapy should be kept simple and limited to exercise, diet, physical therapy, and vitamins when possible.52 Hospitalization and narcotics should be avoided. Benign remedies, such as lotions, nutritional supplements, elastic bandages, and heating pads, may be helpful.16,22 Drug regimens should be simplified and only the most distressing symptoms addressed. Before starting any type of symptomatic drug treatment, specific target symptoms should be identified. The goal is to restore function and to make the target symptoms tolerable, not to remove them completely. If ED patients request an increase in dosage or a stronger medication, they should be told to review their medications with their regular physician before any changes are made. Insistent patients should be informed that long-term opioid use is associated with significant adverse effects, especially constipation, sedation, impaired cognition, and progressive development of tolerance and addiction.22
Mental Health Consultation Patients with somatoform disorders have difficulty confronting their own emotions, view psychiatric evaluation as threatening to their sick role, and take offense at any suggestion that their fears or beliefs may be unwarranted.7,11 They usually resist psychiatric consultation and interpret it as an attempt to be “dumped on the psychiatrist.”31 Nevertheless, psychiatric consultation may be appropriate (1) to confirm the diagnosis or discuss medications, (2) when the patient has coexistent manifestations of chronic depression or psychosis, (3) when symptoms suddenly change or become bizarre, (4) when the patient expresses suicidal ideation or severely disruptive behavior, (5)
Physician Attitudes The key to diagnosing and treating patients with somatoform disorder is effective and appropriate communication skills on the part of the physician. Somatizing patients can present a challenge because it is tempting to point out to them that there is nothing “wrong” with them and that their symptoms need no treatment. Physicians caring for these patients predictably react with feelings of uncertainty, helplessness, anger, or guilt when they can find no physiologic pathology to explain the patient’s distress. Patients with somatoform disorders can become as frustrated with their physician as the physician is with them. It is common for patients with a somatoform disorder to be quickly labeled as a “difficult patient” by the physician and staff.61,62 Unfortunately, the frustration of working with these patients quickly overwhelms the physician’s natural tendency towards compassion and can lead to a swift breakdown in communication. Despite the large number of distressing symptoms reported, physicians rarely demonstrate empathy with these patients.63,64
Treatment Goals Somatizing patients, despite lack of objective physiologic pathology, are, in fact, patients who are in need of tangible and effective help. For some, attaining invalid status enables them to be cared for and nurtured. It offers them a sense of selfimportance and respect not otherwise available to them, as well as an honorable release from noxious personal and vocational responsibilities and duties.40 To attempt a cure poses a threat to this role, and unduly positive projections by physicians are therefore understandably met with disappointment, disbelief, and even thinly veiled reproaches regarding their professional competence.4,16,52 Thus, the goal of therapy must be control of disability and appropriate referrals, rather than cure.21,59 The course of management most likely to prove successful begins with performing a sympathetic and thorough problem-oriented history and physical examination, then offering the patient the paradoxical reassurance that he or she will probably always be ill. When pain is the dominant feature, the patient should not be promised complete relief; rather, a major task of the patient should be to “learn to live with some pain.”21 Treatment goals should focus on modification of illness behavior and improvement of functional status.40 Achievable endpoints include (1) decreased frequency and urgency of medical use, in particular a reduction in ED and unscheduled office visits; (2) avoidance of expensive and hazardous procedures; (3) improved work or school performance; (4) more social activities; and (5) better personal relationships.9,13,22 These principles apply equally to pediatric patients with somatoform disorders. Unnecessary tests and procedures, in addition to placing the patient at risk, may encourage somatization. Physician acknowledgment of the patient’s suffering and family concerns, a “rehabilitative” approach emphasizing
1457
return to normal activities prior to definitive symptom relief, rewarding healthy behavior and discouraging the sick role, assumption by the patient of responsibility for coping with the symptoms, and treatment of coexistent anxiety or depression are the cornerstones of therapy.24 Patients with somatoform disorder have been described as the “least insightful, the least introspective and the least cognitively oriented patients one is likely to encounter.”32 Understanding the link between emotional and somatic distress need not be a treatment goal for these patients, and insight-oriented psychotherapy is neither productive nor cost effective.16,21,22,32 On the other hand, both the physician and the patient must accept fundamental alterations in the traditional paternalistic physician-patient relationship. Increasing responsibility for health and disease management must be incrementally turned over to the somatizing patient.21
■ DISPOSITION Appropriate psychiatric referrals should be provided for the patient. Outpatient tests or hospitalization should be avoided unless clear objective signs indicate a need for diagnostic investigation or therapeutic intervention.21,32 As a rule, management is best carried out by a single primary care physician who becomes the gatekeeper for all medical consultation and care.8–10,13,16,31,52,59 The patient should be told that no alarming findings have come to light, that further testing and additional medications are not indicated at this time, and that ongoing care and periodic reassessment are indicated and will be arranged. Patients with chronic somatization should initially be seen every 2 to 4 weeks, preferably by their primary care physician, even if their symptoms are stable. The visits should be on a time-contingent, not a needcontingent, basis. For the patient, this severs the association between medical contact and the necessity for worsening or additional symptoms and complaints. It also decreases the patient’s fear of abandonment by the physician and permits repeated evaluation for early detection of objective signs of organic disease.13,21 The patient seems to value the visit to the physician more highly than any treatment.32
KEY CONCEPTS ■
The behavior of patients with somatoform disorders is unconsciously driven. They are not “faking” the symptoms or their distress. ■ Short-term management should include the legitimization of symptoms, communication of compassion, and assurance that ongoing vigilance of the patient’s medical condition will be arranged and maintained. ■ Long-term cure of somatization disorder is unlikely. However, a steady state of symptom coping with improved function is an achievable goal. This can be done only in the primary care setting, not in the ED. ■ Laboratory tests, specialty consultations, initiation of medications, and hospitalization should be avoided (unless new objective clinical findings are found). ■ Care decisions should be deferred to the patient’s primary care physician when possible.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 111 / Somatoform Disorders
when current management is not working, or (6) when the patient requests psychotherapy.9,10,31 Favorable prognostic indicators include youth, acute onset, concurrent anxiety or depression, and limited medical comorbidity.59 Many patients accept psychological treatment under the rubric of “stress management” as long as it targets physical symptoms and somatic distress.16 Group therapy techniques presented as education rather than psychotherapy have had some limited success.60 Patients should be reassured that their relationship with the primary physician will continue to avoid the false interpretation that the referral is an abandonment.
Chapter 112
Factitious Disorders and Malingering
Jag S. Heer and Thomas B. Purcell
■ PERSPECTIVE Patients may present to the emergency department with symptoms that are intentionally produced or simulated. The inducements that generate this behavior define two distinct varieties: factitious disorders and malingering. Factitious disorders are characterized by symptoms or signs that are intentionally produced or feigned by the patient in the absence of apparent external incentives.1,2 Factitious disorders have been present throughout history. In the second century, Galen described Roman patients inducing and feigning vomiting and rectal bleeding.3 Hector Gavin sought to categorize this behavior in 1834.3 These patients constitute approximately 1% of general psychiatric referrals, but this percentage is somewhat lower than that seen in emergency medicine because these patients rarely accept psychiatric treatment.1,4 Of patients referred to infectious disease specialists for fever of unknown origin, 9.3% of the disorders are factitious.5 Between 5 and 20% of patients followed in epilepsy clinics have psychogenic seizures, and in some primary care settings the number reaches 44%.6 Among patients submitting kidney stones for analysis, up to 3.5% have been found to be fraudulent.7 Munchausen syndrome, the most dramatic and exasperating of the factitious disorders, was originally described in 1951.8 This fortunately rare syndrome takes its name from Baron Karl F. von Munchausen (1720–1797), a revered German military officer and noted raconteur who had his embellished life stories stolen and parodied in a 1785 pamphlet.3 The diagnosis is appropriate for only 10 to 20% of patients with factitious disorders.1,9 Other names applied include the “hospital hobo syndrome” (patients wander from hospital to hospital seeking admission), peregrinating (wandering) problem patients, hospital addict, polysurgical addiction, laparotomaphilia migrans, Kopenickades syndrome, Ahasuerus syndrome, and hospital vagrant.4,10,11 Munchausen syndrome by proxy (MSBP) is an especially pernicious variant that involves the simulation or production of factitious disease in children by a parent or caregiver, and was first described in 1977.12 There are approximately 1200 estimated new cases of MSBP per year in the United States.3 The condition excludes straightforward physical abuse or neglect and simple failure to thrive; mere lying to cover up physical abuse is not MSBP.13,14 The key discriminator is motive: the mother is making the child ill so that she can vicariously assume the sick role with all its benefits. The mortality rate from MSBP is 9 to 31%.14,15 Children who die are generally 1458
younger than 3 years, and the most frequent causes of death are suffocation and poisoning.14 Permanent disfigurement or permanent impairment of function resulting directly from induced disease or indirectly from invasive procedures, multiple medications, or major surgery occurs in at least 8% of these children.14,16 Other names applied include Polle’s syndrome (Polle was a child of Baron Munchausen who died mysteriously),3,10 factitious disorder by proxy,17 pediatric condition falsification,18 and Meadow’s syndrome.2 Malingering is the simulation of disease by the intentional production of false or grossly exaggerated physical or psychological symptoms, motivated by external incentives such as avoiding military conscription or duty, avoiding work, obtaining financial compensation, evading criminal prosecution, obtaining drugs, gaining hospital admission (for the purpose of obtaining free room and board), or securing better living conditions.2,19–21 The most common goal among such “patients” presenting to the emergency department is obtaining drugs, whereas in the office or clinic the gain is more commonly insurance payments or industrial injury settlements.22 Because of underreporting, the true incidence of malingering is difficult to gauge, but estimates include a 1% incidence among mental health patients in civilian clinical practice, 5% in the military, and as high as 10 to 20% among patients presenting in a litigious context.20 A survey by the American Board of Clinical Neuropsychology members revealed the most likely conditions to be feigned were mild head injury, fibromyalgia, chronic fatigue syndrome, and chronic pain.23
■ CLINICAL FEATURES Factitious Disorders With a factitious disorder, the production of symptoms and signs is compulsive in that the patient is unable to refrain from the behavior even when its risks are known. The behavior is voluntary only in the sense that it is deliberate and purposeful (intentional) but not in the sense that the acts can be fully controlled.2 The underlying motivation for producing these deceptions, securing the sick role, is primarily unconscious.9,24,25 Individuals who readily admit that they have produced their own injuries (e.g., self-mutilation) are not included in the category of factitious disorders.17 Presentations may be acute, in response to an identifiable recent psychosocial stress (termination of romantic relationship, threats to self-esteem), or a chronic life pattern, reflective of the way the person deals with
life in general.26 The symptoms involved may be either psychological or physical.
This disorder is the intentional production or feigning of psychological (often psychotic) symptoms suggestive of a mental disorder. Stimulants may be used to induce restlessness or insomnia, hallucinogens to create altered levels of consciousness, and hypnotics to produce lethargy. This psychological factitious condition is less common than factitious disorders with physical symptoms and is almost always superimposed on a severe personality disorder.2,17
Physical Symptoms The intentional production of physical symptoms may take the form of fabricating symptoms without signs (e.g., feigning abdominal pain), simulation of signs suggesting illness (e.g., fraudulent pyuria, induced anemia), self-inflicted pathology (e.g., producing abscesses by injecting contaminated material under the skin), or genuine complications from the intentional misuse of medications (e.g., diuretics, hypoglycemic agents).21 These patients are predominantly unmarried women younger than 40 years. They typically accept their illness with few complaints and are generally well-educated, responsible workers or students with moral attitudes and otherwise conscientious behavior.21,27,28 Many are in health care occupations, including nurses, aides, and physicians. These patients are willing to undergo incredible hardship, limb amputation, organ loss, and even death to perpetuate the masquerade.21 Although multiple hospitalizations often lead to iatrogenic physical conditions, such as postoperative pain syndromes and drug addictions, patients continue to crave hospitalization for its own sake. They typically have a fragile and fragmented self-image and are susceptible to psychotic, and even suicidal, episodes.27 Interactions with the health care system and relationships with caregivers provide the needed structure that stabilizes the patients’ sense of self. The hospital may be perceived as a refuge, sanctuary, or womb-like environment.4,21,24,29 Some patients are apparently driven by the conviction that they have a real, but as yet undiscovered, illness. Consequently, artificial symptoms are contrived to convince the physician to continue a search for the elusive disease process.21 Factitious illness behavior has even emerged on the Internet. “Virtual support groups” offering person-to-person communications through chat rooms and bulletin boards have been perpetrated by individuals, under the pretense of illness or personal crisis, for the purpose of extracting attention or sympathy, acting out anger, or exercising control over others.30 There has been increasing recognition of factitious illness produced by children (distinct from the MSBP described later). These children, ranging in age from 8 to 18 years, are typically “bland, flat and indifferent during their extensive medical interventions … depressed, socially isolated and often obese.”31 Among the most common presentations are fever without clear etiology, diabetic ketoacidosis, purpura, and recurrent infections. The prognosis is good if identification and psychotherapeutic intervention can be carried out at a young age.31
Munchausen Syndrome The uncommon patient with true Munchausen syndrome has a prolonged pattern of “medical imposture,” usually years in duration. The behavior usually begins before age 20 years and is diagnosed between ages 35 and 39. Twice as many men are
Munchausen Syndrome by Proxy The diagnosis of MSBP depends on specific criteria (Box 1121).14 The presenting complaints typically evade definitive diagnosis and are refractory to conventional therapy for no apparent reason.14 The symptoms are usually more than five in number, presented in a confused picture, are unusual or serious, and, by design, are unverifiable. They invariably occur when the mother is alone with the child or otherwise unobserved.36 In 72 to 95% of cases, simulation or production of illness occurs while the victim is hospitalized.14 Simulated illness, faked by the mother without producing direct harm to the child (e.g., adding blood to a urine specimen), is present in 25% of cases. Produced illness, which the
Chapter 112 / Factitious Disorders and Malingering
Psychological Symptoms
1459
affected as women.4,32 Patients’ entire adult lives may consist of trying to gain admission to hospitals and then steadfastly resisting discharge. Their career of imposture usually lasts about 9 years but has continued unabated for as long as 50 years.4 The quest for repeated hospitalizations often takes these patients to numerous and widespread cities, states, and countries.2 These individuals see themselves as important people, or at least related to such persons, and their life events are depicted as exceptional.32 They possess extensive knowledge of medical terminology. Frequently there is a history of genuine disease, and the individual may exhibit objective physical findings.27 The symptoms presented are “limited only by the person’s medical knowledge, sophistication, and imagination.”2 The alleged illnesses involved have been termed dilemma diagnoses in that investigators rarely can totally rule out the disorder, clarify the cause, or prove that it did not exist at one time.4 Common presentations are those that most reliably result in admission to the hospital, such as abdominal pain, self-injection of a foreign substance,10,11 feculent urine, bleeding disorders, hemoptysis, paroxysmal headaches, seizures, shortness of breath, asthma with respiratory failure,4,33 chronic pain,25 acute cardiovascular symptoms (e.g., chest pain, induced hypertension and syncope),32 renal colic and spurious urolithiasis,7 fever of unknown origin (hyperpyrexia figmentatica),5,11 profound hypoglycemia, and coma with anisocoria.34 Such self-induced conditions themselves may prove highly injurious or even lethal.10 The patient usually presents to the emergency department during evenings or on weekends so as to minimize accessibility to psychiatric consultants, personal physicians, and past medical records.11,27 In teaching institutions these patients typically present in July, shortly after the change in resident house officers.4 They relate their history in a precise, dramatic, even intriguing fashion, embellished with flourishes of pathologic lying and self-aggrandizement. Pseudologica fantastica, or pathologic lying, is a distinctive peculiarity of these patients. In a chronic, often lifelong behavior pattern, the patient typically takes a central and heroic role in these tales, which may function as a way to act out fantasy.35 The history quickly becomes vague and inconsistent, however, when the patient is questioned in detail about medical contacts.2,29 Attempts to manage the complaint on an outpatient basis are adamantly resisted.25 Once admitted, the patient initially appeals to the physician’s qualities of nurturance and omnipotence, lavishing praise on the caregivers. Behavior rapidly evolves, however, as the patient creates havoc on the ward by insisting on excessive attention while ignoring both hospital rules and the prescribed therapeutic regimen.2 When the hoax is uncovered and the patient confronted, fear of rejection abruptly changes into rage against the treating physician, closely followed by departure from the hospital against medical advice.10,11,25
1460
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
Criteria for the Diagnosis of Munchausen
BOX 112-1 Syndrome by Proxy
1. Apparent illness or health-related abnormality that the parent or someone who is in loco parentis has concocted or produced 2. Presentation of the child for medical treatment, usually persistently 3. Failure by the perpetrator to acknowledge the true etiology or the deception 4. Cessation of the acute symptoms and signs of illness when the child is separated from the perpetrator
mother actually inflicts on the child (e.g., injection of feces into an intravenous line), is found in 50% of cases. Both simulated and produced illnesses are found in 25% of cases.14 MSBP most commonly arises with factitious bleeding, seizures, central nervous system (CNS) depression, apnea, diarrhea, vomiting, fever, and rash.14 Reported techniques of simulation or production of disease include administration of drugs or toxins (e.g., chronic arsenic poisoning, ipecac, warfarin, phenolphthalein, hydrocarbons, salt, imipramine, laxatives, CNS depressants), caustics applied to the skin, and nasal aspiration of cooking oil.12,14–16,27,37 Techniques of asphyxiation include (1) covering the mouth or nose with one or both hands, a cloth, or plastic film and (2) inserting the fingers into the back of the mouth. In such instances, even struggling infants may sustain no cutaneous markings.38 Cases involving seizures are common and may involve third-party witnesses. On personal questioning, however, these witnesses frequently deny the occurrence of seizure activity.14 In a variant of MSBP termed serial Munchausen syndrome by proxy, there may be a history of similar strange presentations in multiple siblings, although typically only one child is involved at a time.15,16,38 In 9% of such cases there is a history of siblings who died under mysterious circumstances.14 Perpetrator Characteristics. Ninety-eight percent of perpetrators are biologic mothers from all socioeconomic groups.14 Many have a background in health professions or social work, features of Munchausen syndrome themselves, or a past history of psychiatric treatment, marital problems, or suicide attempts.14,15 Depression, anxiety, and somatization are common, but frankly psychotic behavior by the mother is atypical.14,16 Perpetrators of MSBP have an inherent skill in manipulating health workers and child protection services.39 They are pleasant, socially adept, cooperative, and appreciative of good medical care. They often display a peculiar eagerness to have invasive procedures performed on their child.3 They often choose to stay in the hospital with their child, cultivate unusually close relationships with hospital staff, and thrive on the staff’s attention.12–14,16,37 This affable relationship with the medical team rapidly changes to excessive anger and denial when confronted with suspicions.19 Most of these mothers have had an abusive experience early in life, and they use the health care system as a means to satisfy personal nurturing demands.36,40 They often cannot distinguish their needs from the child’s and satisfy their own needs first. They derive a sense of purpose from the medical and nursing attention gained when their children are in the hospital.13,14,16,40 Alternatively, the behavior may enable the mothers to escape from their own physical or psychological illnesses, marital difficulties, or social problems.36 Victim Characteristics. Victims of MSBP are equally male and female children. The mean age at diagnosis is 40 months, and
the mean duration from the onset of signs and symptoms to diagnosis is 15 months.14 A known physical illness that explains part of the symptoms is common among these children.40 Most have a history of significant failure to thrive and have been hospitalized in more than one institution. Delays in many areas of performance and learning, difficulty with family relationships, attention deficit disorder, or clinical depression may coexist.16 Some of these victims may develop factitious disorder later in life.3 Victims of MSBP are also found among the elderly population, although this is uncommon.41
Malingering Malingering is frequently found in association with antisocial personality disorder. On questioning, malingerers are vague about prior hospitalizations or treatments. The physicians who previously treated them are usually unavailable. At times, malingerers may be careless about their symptoms and abandon them when they believe no one is watching.27 In some “patients,” such as those seeking drugs, homeless persons seeking hospital admission on a cold night, or prisoners wanting a holiday from incarceration, the secondary gain may be clear. In other persons the external incentive may be obscure. In contrast to the person with factitious disorders, the malingerer prefers counterfeit mental illness because it is objectively difficult to verify or disprove. Amnesia is the most common psychological presentation, followed by paranoia, morbid depression, suicidal ideation, and psychosis.20
■ DIAGNOSTIC STRATEGIES Factitious Disorders Initial diagnosis is often delayed because the possibility of factitious disease is not considered, physicians may be unfamiliar with this problem, or the patient does not exhibit the type of personality expected with this behavior.7 Diagnosis may be confounded by genuine medical illnesses predating and coexisting with a factitious disorder. For example, patients with factitious hypoglycemia may have a history of insulindependent diabetes mellitus, or factitious skin disorders may be preceded by true dermatologic diseases.1 Identification of a factitious disorder is usually made in one of four ways: (1) the patient is accidentally discovered in the act, (2) incriminating items are found, (3) laboratory values suggest nonorganic etiology, or (4) the diagnosis is made by exclusion.17 Wallach provides a useful review of the laboratory diagnosis of factitious disorders including feigned endocrine, hematologic, genitourinary, gastrointestinal, and infectious disorders.42 Suspected MSBP requires a detailed description of the event or illness and a search for caregiver witnesses, who should be interviewed personally. Although it is essential to see the child when the symptoms are present, the parents show great ingenuity at frustrating this effort.40 Additional history of unusual illness in siblings and parents should be sought. Child victims who are verbal should be interviewed in private regarding foods, medicines, and their recollection of the symptoms or events. Prior medical records of the victim and, if possible, the siblings should be examined, although parents may impede such data gathering. The major obstacle to early discovery of MSBP is its omission from the differential diagnosis. When it is considered, the diagnosis is generally made easily and quickly.14 A suspected diagnosis may be confirmed through separation of the parent from the child (with consequent cessation of symptoms), covert video surveillance during hospitalization, or toxin screens.16,36,38 In the majority of cases, the caregiver attempts
1461
BOX 112-2 Characteristics of Malingering
to induce episodes surreptitiously while in the hospital, often during the first day of admission.14,36
Malingering Malingering should be strongly suspected with any combination of certain factors (Box 112-2).2,43 A definitive diagnosis of malingering can be established only by securing the patient’s confession, a rare circumstance.44 Because malingering constitutes criminal behavior, documentation of this diagnosis must be made with care.20 In the absence of proof of wrongdoing, it is best to assume that the patient is not a malingerer but rather a common somatizer.44 Malingerers who pursue drugs may report an unusually large number of drug allergies to persuade the physician toward prescribing their drug of choice or simply insist on a specific drug (e.g., Demerol or Dilaudid).45 One unfortunate circumstance of the Internet is the wide availability of quality medical advice on how to convincingly feign pain and disability.45
■ DIFFERENTIAL CONSIDERATIONS Patients with factitious disorders are distinguished from malingerers because their desired hospitalization or surgery seems to offer no secondary gain other than to play the sick role.2,10,22 The clinical presentation of the majority of patients with factitious disorders, unlike those with Munchausen syndrome, is relatively subtle and convincing. The complaints are generally chronic in nature rather than emergent and precipitous, and there are no obvious associated behavioral aberrations.21 The most important diagnosis to exclude when faced with a likely factitous disorder is a genuine medical condition that might account for the illness. Malingering is usually associated with less chronicity than factitious disorder, and malingerers are more reluctant to accept expensive, possibly painful, or dangerous tests or surgery.22
■ MANAGEMENT Treatment options for factitious disorders depend on the patient’s characteristics. Although it is challenging, managing common forms of factitious disorder can be more rewarding, especially with adolescents, than managing Munchausen syndrome.1,9,24,29 Cases stemming from an underlying depression have a more favorable prognosis than those associated with borderline personalities.27 The best approach to patients with factitious disorder, other than Munchausen syndrome and MSBP, remains an area of controversy. Direct nonaccusatory confrontation has been advocated as “the foundation of effective management” when coupled with the assurance that an ongoing relationship with a physician will be provided.4,21,22,27 This may be the first step in the acceptance of outpatient therapy.4
■ DISPOSITION Patients with factitious disorder should receive primary care follow-up and ongoing care. If acceptable, psychiatric referral should be arranged. Referral to other medical specialists or hospitalization should be avoided when possible. The manner of presentation and the unavailability of past medical history often allow patients with Munchausen syndrome to achieve hospital admission. If the patient is discharged from the emergency department, outpatient primary care follow-up and psychiatric referral should be offered, although both are likely to be refused.25 Because perpetrators of MSBP typically induce symptomatic episodes soon after hospitalization, admission of the victims (children or elderly persons) without taking appropri-
Chapter 112 / Factitious Disorders and Malingering
1. Medicolegal context of the presentation (e.g., the patient was referred by his or her attorney). 2. Marked discrepancy between the person’s claimed stress or disability and objective findings. 3. Poor cooperation during the diagnostic evaluation or poor compliance with previously prescribed treatment regimens. 4. The person exhibits, or has a history of, antisocial behavior.
Others point out that confrontation is ineffective in most patients and may even be counterproductive in that it threatens to undermine a needed psychological defense. Enforced recognition of external objective reality, while simultaneously disallowing the patient’s subjective experience, may generate even more dysfunction directed at legitimizing and maintaining symptoms and may even place the patient at risk for suicide.9,17,29,46,47 Some patients may relinquish this defense if they feel safe in doing so and may abandon a claim to disease if some face-saving option is offered. This approach, termed the therapeutic double bind or contingency management, involves informing the patient that a factitious disorder may exist. The patient is further told that failure to respond fully to medical care would constitute conclusive evidence that the patient’s problem is not organic but rather psychiatric. The problem is therefore reframed or redefined in such a way that (1) symptoms and their resolution are both legitimized and (2) the patient has little choice but to accept and respond to a proposed course of action or seek care elsewhere.9,47 Individuals with Munchausen syndrome typically demonstrate overt sociopathic traits or a borderline personality disorder and are demanding and manipulative, especially regarding analgesics.5,9 They have been described as “essentially untreatable,” and successful management of this condition is, in fact, considered reportable. Early confrontation or limit setting, especially regarding drug use, is advocated.9,11,21,25,27 Although Munchausen patients typically do not want to be examined extensively, a thorough physical examination should be performed to rule out physical pathology. MSBP constitutes a form of child (or elder) abuse, and appropriate action to protect the victim, including notification of welfare services, should take immediate priority.38,41 When the diagnosis has been established and the parents confronted, psychiatric care should be made immediately available to the parents because maternal suicide is a significant risk.14 Malingerers do not want to be treated. Because they are “gaming the system” for personal advantage, the last thing they want is an accurate identification of their behavior and appropriate intervention. The emergency physician should maintain clinical neutrality, offering the reassurance that the symptoms and examination are not consistent with any serious disease. Some authors have characterized patients’ use of medical resources under false pretenses as criminal behavior, and several states have enacted legislation against the fraudulent acquisition of medical services. Successful prosecution of such behavior has been reported.48 Conversely, patients with factitious disorders can and do sue. In dealing with such patients, it is advisable to involve hospital administration and risk management. Clandestine searches are inadvisable, and respect for the patient’s confidentiality should be maintained.17
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1462
ate precautions may actually place them at increased risk.14 Visits by the suspected perpetrator should be closely supervised, and no food, drink, or medicines should be brought in by the family. Protective services should be notified. Out-ofhome placement of children in established cases of MSBP is advisable, and best outcomes are seen among children taken into long-term care at an early age without access to their mother. Children allowed to return home have a high rate of
repeated abuse.39 In 20% of reported deaths, the parents had been confronted and the child sent home to them, subsequently to die.14 After courteous but assertive reassurance, suspected malingerers should be offered primary care follow-up if the symptoms do not resolve. These individuals may become threatening when they are either denied treatment or overtly confronted.19
KEY CONCEPTS ■
Emergency department patients who have consciously synthesized symptoms and signs may be divided into two broad diagnostic categories: (1) those with obvious secondary gain (malingering), who control their actions, and (2) those with a motivation of achieving the sick role (factitious disorders), who cannot control their actions. ■ Emergency department management of patients suspected of fabricating disease includes a caring
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
attitude and a search for objective clinical evidence of treatable medical or psychiatric illness. ■ Unnecessary tests, medications, and hospitalizations should be avoided in the absence of objective evidence of a medical or psychiatric disease, and patients should be referred for ongoing primary care. ■ In cases of suspected MSBP involving children or elderly persons, protection of the victim takes first priority.
Chapter 113
Suicide
Stephen A. Colucciello
■ PERSPECTIVE Although suicide has occurred in all societies since the beginning of recorded history, attitudes toward suicide have differed dramatically among various eras and cultures. Seneca viewed suicide as the ultimate expression of personal freedom, but later Judeo-Christian religions have routinely condemned it. Shakespeare portrayed suicide sympathetically and expressed either pity or admiration for the victim.1 In the United States, suicide is illegal in 49 states, and only since 1994 has assisted suicide of terminally ill patients been sanctioned in Oregon. On the Internet, suicide help groups provide active advice on methods, and numerous bulletin boards condemn them. More than 100,000 sites about suicide now appear on the Internet.2 Suicidal patients comprise almost 2% of emergency department (ED) visits.3 Two facts are especially important to remember when approaching potentially suicidal patients in the ED. First, many suicide attempts occur during an acute crisis, such as a personal loss or the exacerbation of an underlying psychiatric disorder. This acute crisis is usually time limited and may be resolvable or treatable. Second, except for the acutely psychotic patient, suicidal patients are usually ambivalent about dying. The attitude and approach of the emergency physician can help a patient choose crisis resolution rather than death.
Definitions The term suicide, from the Latin suicidum (to kill the self), refers to a continuum of thought and action that runs from ideation to completion of the act.4 Parasuicide is used by the British to describe an attempted suicide that is more of a gesture than a serious act. Statistically, there are 10 to 40 suicide attempts for every completed act.5 Chronic suicidal behavior consists of recurrent self-destructive acts, such as heavy drinking in the presence of alcoholic liver disease. Occult suicide refers to self-destructive acts disguised as accidents, such as the intoxicated, depressed driver in an apparently accidental car crash. Silent suicide describes the act of slowly killing oneself by nonviolent means, such as starvation or noncompliance with essential medical treatment. Silent suicide is most common in elders and frequently goes unrecognized. A suicide pact involves an agreement between two people who are intimately involved and accounts for 0.6% of all suicides.6 Mass suicide or group suicide involves a number of willing
and sometimes not-so-willing persons, such as members of an apocalyptic cult.
Epidemiology Suicidal ideation is common, with as many as one in three people considering suicide during their lifetime.7 Suicide is the eleventh leading cause of death in the United States, claiming more than 32,000 lives annually, with an overall rate of 11.05 per 100,000 population in 2004.8 This is the equivalent of 89 suicides per day or one suicide every 16 minutes. Over a 5-year period, there were approximately 412,000 annual ED visits for attempted suicide and self-inflicted injury. The most common method of injury was poisoning (68%), followed by cutting or piercing (20%). One third were admitted to the hospital, with 31% going to the intensive care unit. A psychiatric diagnosis was assigned for 55% of visits, with depressive disorder accounting for 34% and alcohol abuse for 16%.9 Suicide rates vary with age, gender, race, and marital status. Suicides are highest among older individuals, particularly elderly white men. The case fatality rate also increases tremendously with advancing age, from 5% in youths aged 5 to 14 years, to 34% in adults older than 64 years.10 White men commit 73% of all suicides in the United States. Whites and Native Americans are much more likely to commit suicide than African Americans, Hispanics, or Asians. Marriage decreases the likelihood of suicide, but separated or divorced people have a higher rate of suicide than those who never had a close relationship. Women attempt suicide three to four times more often than men, whereas men are three to four times more likely to succeed. In one U.S. study, only 5% of suicide attempts by females were fatal, compared to 23% of those by males.10 In general, men tend to use more lethal methods, such as firearms. Worldwide, Chinese and Indian women have higher rates of suicide than women of other nationalities.11 Pregnant women are at a significantly lower risk than women of childbearing age who are not pregnant. Motherhood seems to protect against suicide, except that postpartum depression is associated with a higher than normal suicide rate. Most people who attempt suicide have one or more known risk factors (Box 113-1). Individuals with the highest risk include those with psychiatric disorders, alcohol or substance abusers, adolescents, elders, and patients with certain chronic illnesses. In patients hospitalized for psychiatric disorders, the first month after discharge carries a high risk of suicide,12 and 1463
1464
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
BOX 113-1 Risk Factors for Suicide Demographics White men older than 65 years Women older than 60 years Males 15 to 24 years old American Indian or Native Alaskan 15 to 34 years old Psychiatric disorders Major depression Bipolar disorder Schizophrenia Borderline personality disorder Panic disorder Substance abuse Alcoholism Drug abuse (especially cocaine) Medical history Prior suicide attempts Chronic pain or illness Physical or sexual abuse Recent psychiatric hospital discharge Terminal illness (especially cancer and AIDS) Low intelligence scores (in men) Lower body mass index Short stature (in men) Cosmetic breast augmentation (in women) Family history Family violence Suicide in family Social factors Firearm in home Living alone Separated, widowed, or divorced Unemployed Homeless Recent personal loss Veterans Recent incarceration Lack of religious affiliation Emotional factors Hopelessness Chronic loneliness Fixation on death
that risk is especially great for patients who are in the first week after discharge from a psychiatric facility.13 Any prior history of suicide, even in the remote past, is an important risk factor.14 A strong association may exist between suicide risk and bisexuality or homosexuality in men.15 This association is also seen in lesbian, gay, and bisexual adolescents.16 Unemployment appears to be a risk factor for suicide among 18- to 24year-old men.17 Homeless people with mental illness are at particularly high risk for suicidal behavior, in part because of the high prevalence of traditional risk factors. Recent incarceration is also a risk factor for suicide. During the first 2 weeks after release, the risk of death among former inmates is more than 12 times that of the general population. The leading causes of death among former inmates include drug overdose, cardiovascular disease, homicide, and suicide.18 In general, the risk of suicide in recently released prisoners is approaching that seen in recently discharged psychiatric patients.19 Suicide completers and suicide attempters represent separate but overlapping populations.20 Although 10 to 15% of suicide attempters ultimately complete suicide, 60 to 70% of
suicide completers have no prior history of attempts and commit suicide on the first known attempt.21 In individuals who committed suicide while not in contact with mental health services, nearly one third of cases (32%) had no concurrent mental disorder.22
■ PATHOPHYSIOLOGY AND ETIOLOGY Societal, Psychiatric, and Biologic Factors There are many motivations for attempting suicide. It may be seen as the only escape from a terminal disease or intense chronic pain. It may be an act of revenge or political protest. Most suicide attempts occur in individuals with intense feelings of hopelessness, guilt, or self-hatred, often compounded by the exacerbation of an underlying psychiatric disorder or by the occurrence or perception of a great personal loss. The underlying causes for suicide are similar for adults and adolescents; however, adolescents tend to romanticize suicide, and “copycat” suicides are frequent after the suicide of celebrities or friends. Regardless of the motivation, most suicide attempters are ambivalent, and their attraction to death is usually counterbalanced by a desire to live. This internal conflict is reflected in the high ratio of attempted to completed suicides, and the fact that most people consult a physician shortly before their death. Psychoanalysts explain suicide in terms of psychic forces. Freud believed that suicide stems from aggression, initially directed toward another person, which ultimately turns against the self. Depression and suicide in the Freudian model represent internalized anger. Many authorities have recognized this association between aggression and suicide. In the United States, more than 1000 deaths each year result from murdersuicides. The perpetrators are usually depressed mothers, despairing elderly men or young men with intense sexual jealousy. Their victims are usually young children, blood relatives, or female sexual partners. The dual risk for suicide and violence is greatest in alcoholics.23 The impulses that lead to suicide differ between violent and nonviolent people. “Suicidality” is correlated with anger, fear, and suspiciousness in violent individuals and with feelings of sadness and despair in nonviolent persons. The psychic roots of suicide may arise from childhood trauma. Chronic loneliness during childhood is associated with subsequent suicide attempts during adolescence, and a history of sexual molestation is linked to suicide attempts in women and adolescents. Current research suggests a biologic basis for depression and suicide involving the serotonergic and dopaminergic systems. People who attempt suicide have altered serotonin receptor function and low serotonin levels.24,25 These abnormalities may be regulated through serotonergic-related genes in persons with major depression.26 The genetic basis of suicide is not clearly understood. The STin2 genetic locus might, at least in part, account for the observed familial aggregation of suicidal behavior.27 Polymorphisms in the tryptophan hydroxylase gene may affect the synthesis of serotonin. Recent data also indicate that certain genetic markers may be linked to suicidal ideation associated with medications used to treat depression.28 The genetic susceptibility to suicide, however, may affect individuals only when associated with psychiatric illness or stress.29 The rate of suicide is twice as high in families of suicide victims, and a family history of suicide predicts suicide independent of severe mental disorder.30 Relatives of suicide completers are over 10 times more likely than relatives of comparison subjects to attempt or complete suicide.31 Depressed patients who attempt suicide excrete less homovanillic acid in their urine and produce less dopamine
Methods of Attempting Suicide Most completed suicides involve firearms (70%), whereas most attempted suicides involve the ingestion of drugs or poisons (72%).38 In one large study, poisoning with drugs accounted for 74% of acts but only 14% of fatalities; firearms and hanging accounted for only 10% of acts but 67% of fatalities. Firearms were the most lethal means (91% resulted in death), followed by drowning (84%) and hanging (82%).10 Episodes involving firearms are 2.6 times more lethal than the second most lethal suicide method, suffocation.39 In 2005, 17,002 suicides involved firearms in the United States.40 Guns represent the most common method of suicide in all victim subgroups, especially among older persons and adolescents, and the use of guns has increased dramatically in the past decade, recently replacing ingestion as the major cause of suicide among women.41 The simple presence of a gun in the home represents an independent risk factor for firearm-related suicide, but not by nonfirearm means.42 This is particularly true for adolescents, whose risk for suicide increases 5 to 10 times when there is a gun in the household.43,44 In general, firearm prevalence is positively related to the suicide rate, even after controlling for rates of attempted suicide.45 Suicide by handgun is often associated with drug or alcohol use.46 The rate of gun-related suicide is 57 times higher in the first week after purchasing a handgun.47 After gun-related deaths, lethal methods chosen by men tend to include hanging, suffocation, or jumping from a height, whereas women are more likely to commit suicide by poisoning. Antidepressant overdose is the most common cause of suicide by ingestion.48 Cyclic antidepressants are associated with more deaths because of their widespread use and high potential for lethality. Most patients hospitalized for selfpoisoning have ingested drugs prescribed by their physicians for depression.49 SSRIs, including fluoxetine (Prozac), sertraline (Zoloft), and paroxetine (Paxil), are less lethal when taken in overdose and have replaced cyclic antidepressants as the first therapy in depression.
1465
The method of suicide depends on many factors, including psychic issues of self-hate, the desire for a peaceful versus violent death, and the availability of fatal means. Those who jump to their death are more likely single, unemployed, or psychotic. Those who use firearms are more likely to be male, alcoholic, to have been arrested, or to have an antisocial or borderline personality disorder.50 Communities with tall buildings and bridges have higher rates of suicide from falls, whereas suicide by gunfire occurs more often in areas where firearms are prevalent. “Suicide by cop” occurs when a suicidal individual intentionally provokes a police officer by orchestrating a lethal situation where the officer is forced to shoot in self-defense or to protect other civilians. This may account for as many as 11% of officer-involved shootings in Los Angeles.51 Certain individuals carry a suicide note; some offer an eerie postmortem apology to the police officer who ultimately kills them.
■ CLINICAL FEATURES Psychiatric Illness Although most psychiatric patients never attempt suicide, most people who commit suicide have either a diagnosable psychiatric illness or alcoholism. Exceptions include those with mental retardation, dementia, and agoraphobia.52 Patients with an affective disorder, especially major depression, are at highest risk.53,54 Approximately 15 to 20% of people with major depression commit suicide, usually while under psychiatric care.55 Individuals who experience hopelessness, anhedonia (loss of ability to experience pleasure), and mood cycling are at highest risk. Impulsive-aggressive personality disorders and alcohol abuse/dependence are independent predictors of suicide in major depression.56 Approximately 10% of schizophrenic patients will kill themselves. Psychotic patients who commit suicide are most often unmarried whites with high intelligence quotient scores.57 Patients with borderline personality disorders are also predisposed to commit suicide. Women with borderline personality disorder who attempt suicide often have a history of childhood sexual abuse and impulsive behavior. The risk is especially high when patients require hospitalization for psychiatric illness and is greatest the first month after discharge.58,59 Approximately 40% of patients with panic disorder attempt suicide at some point in their lives. These patients usually have an additional comorbid psychiatric diagnosis (e.g., borderline personality disorder, substance abuse, emotional instability). However, any preexisting anxiety disorder (including social phobia, simple phobia, generalized anxiety disorder, panic disorder, agoraphobia, obsessive-compulsive disorder) is an independent risk factor for subsequent suicidal ideation and attempts.60 Post-traumatic stress disorder, sustained by military combat personnel and disaster survivors, is also associated with suicide. Post-traumatic stress disorder frequently coexists with a major depressive episode, and this combined psychopathology enhances the risk for suicidal behavior.61 In 2005, there were at least 6256 suicides among those who served in the U.S. armed forces; a rate double that of nonveterans. Veterans between the ages of 20 and 24 years who served in Iraq or Afghanistan had the highest suicide rate among all veterans.62
Alcoholism and Substance Abuse Of successful suicides reported by the National Violent Death Reporting System, 33.3% tested positive for alcohol, 16.4% for
Chapter 113 / Suicide
than depressed patients who have not attempted suicide. Low concentrations of dopamine and serotonin metabolites in the cerebrospinal fluid also correlate with suicidal behavior. Suicide attempts in women vary with estrogen levels, with 42% of attempts occurring during the first week of the menstrual cycle.32 Neuroanatomy may also influence suicidality; suicide victims have smaller right-sided parahippocampi than control subjects.33 Although there are no currently available laboratory tests that can identify individuals at increased risk for suicide, research holds promise for biologic markers in the future. Some drugs, including reserpine, benzodiazepines, and barbiturates, are associated with depression and suicidal behavior. The Food and Drug Administration recently linked suicidal ideation among children and adolescents to selective serotonin-reuptake inhibitors (SSRIs) and added a black box warning regarding SSRI use for all age groups.34 It is ironic that in both the United States and the Netherlands, SSRI prescriptions for children and adolescents decreased after these regulatory warnings, and that these decreases were associated with increases in suicide rates in both children and adolescents.35 Patients who commit suicide shortly after the initiation of antidepressant medications are explained by the “mobilization of energy” theory.36,37 According to this theory, patients who are profoundly depressed may develop the energy to attempt suicide only as their condition improves with treatment. Such patients must be monitored very closely during their initial phase of treatment.
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1466
opiates, 9.4% for cocaine, 7.7% for marijuana, and 3.9% for amphetamines.63 Other studies show that approximately 25% of all suicides involve alcoholics, and estimates of the lifetime risk of suicide from chronic alcoholism range from 3 to 25%.64 Alcoholics who commit suicide usually have multiple risk factors, including major depressive episodes, unemployment, medical illness, and interpersonal loss. In psychiatric patients, the use of alcohol increases depression and suicidal behavior. Substance abuse among mentally ill patients is increasingly common. Psychoactive substance abuse is associated with a greater frequency, repetitiveness, and lethality in suicide attempts than are most other medications. Nearly one half of all adolescents who attempt suicide use drugs shortly before the attempt, and alcohol intoxication is strongly associated with suicide by firearms.65 Cocaine use is particularly dangerous; in New York City, 20% of all suicide victims younger than 61 years used cocaine within days of their death. Among young Hispanic men, nearly one half of the suicide victims have toxicologic screens positive for cocaine.66 In general, cocaine abusers choose violent means of selfdestruction, especially firearms.
Adolescents Suicide among adolescents has quadrupled during the past 40 years and is now the third leading cause of death (after accidents and homicides) in people between 5 and 24 years of age.67 Approximately 2 million U.S. adolescents attempt suicide each year, and 19% report serious consideration of suicide in the past year.68 Although some authorities believe that the rise in adolescent suicide simply corresponds to changing demographics in the United States, others believe that the increase is related to a growing sense of hopelessness, increased economic pressures, and access to firearms. In a survey of high school students in North Carolina, 24% had seriously considered suicide, 19% had planned suicide, and 9% had actually attempted suicide during a 1-year period.69 Adolescent girls are more likely to attempt suicide, whereas adolescent boys are more likely to complete suicide; the ratio of attempted to completed suicides is 25 : 1 for adolescent girls and 3 : 1 for boys.70 Most adolescents who complete suicide have made previous suicide threats. The majority of youths who kill themselves meet criteria for diagnosable psychiatric disorders, and both alcohol and substance abuse play a significant role in teenage suicide attempts. Adolescents with panic attacks are twice as likely to make suicide attempts as adolescents without panic attacks.71 Gay, lesbian, bisexual, or “not sure” youths may also be more prone to self-harm.72 Belonging to the “Goth” subculture is a predictor of self-harm and attempted suicide.73 Nearly 40% of youths in runaway programs report prior suicide attempts.74 Young people may also be influenced by movies or television shows that feature suicide. Teenage suicides typically increase after television broadcasts on the subject. From 1989 to 1995, suicide by firearm in young people increased dramatically.75 The firearm-related suicide rate in U.S. adolescents is 11 times higher than the combined rates of 25 other industrialized countries.76 Having a gun in the home places the troubled adolescent in great danger, and storing the gun in a locked cabinet or separating it from the ammunition does not deter suicide attempts.77 Surprisingly, up to 23% of adolescents who have attempted suicide report that their families continue to keep firearms and ammunition in the home despite their suicide attempt.
Older Adults The highest rates of completed suicide occur in elders. Suicide by firearm is the fourth leading cause of injury-related deaths among older U.S. residents.78 Older Americans use highly lethal methods when attempting suicide and, unlike adolescents, rarely stage an attempt that permits rescue. Self-inflicted gunshot wounds account for 88% of elder suicides. White men older than 65 years account for approximately 80% of suicide deaths, whereas suicide is rare among elderly people belonging to minority groups.79 Suicide among older adults is especially common in those with prior suicide attempts or major depression. Severity of depression is the strongest predictor of suicide in elders.80 Physicians often overlook signs of depression in older patients, even though most who commit suicide see their primary care physician during the month before their death. In one study, almost half of all elders who committed suicide visited a physician in the preceding week.81 Elders also have more chronic illnesses that predispose to suicide. Perceived poor health, poor sleep quality, and limited presence of a relative or friend to confide in are also associated with suicide among elders.82
Chronic Illness Patients with terminal illnesses may commit suicide to end their suffering and to reduce the emotional and financial burden on their families. Diseases more highly associated with suicide include cancer, stroke, renal failure, congestive heart failure, and chronic lung disease. In the elderly, congestive heart failure, chronic obstructive lung disease, seizure disorder, urinary incontinence, anxiety disorders, depression, bipolar disorder, and moderate to severe pain are specific illnesses associated with suicide.81 A history of cancer is an especially strong risk factor in elders. The acquired immunodeficiency syndrome (AIDS) epidemic has also increased suicide rates, and the relative risk of suicide in men with AIDS is nearly 37 times higher than in uninfected men. Patients who are positive for human immunodeficiency virus but do not have AIDS-defining con ditions are more likely to be suicidal than those with active disease.83
History Recognition of Depression and Suicide Potential Recognition of suicide potential is relatively straightforward in patients who present shortly after a suicide attempt, as well as in individuals who complain of depression or express suicidal ideation during their evaluation. The potential for suicide should also be considered in patients with any acute problem related to chronic alcoholism, substance abuse, or any psychiatric disorder. Silent suicide is possible with patients who present to the ED repeatedly because of noncompliance with treatment of their medical disorders. Occult suicide should be suspected in patients who “unintentionally” overdose or have had “accidental” gunshot wounds, lacerated wrists, automobile crashes, or falls from heights.
Patients Who Present after a Suicide Attempt or Have Suicidal Ideation Following a suicide attempt, patients with a normal mental status should be queried regarding the specifics of the act after medical evaluation and treatment are initiated. Suicidal patients may give inaccurate histories or may refuse to speak
1467
depressed or suicidal are often thankful and relieved for the intervention.
Physical Examination Patients should be examined closely for evidence of drug ingestion, trauma, or an associated medical illness. Look at the wrists for evidence of prior cutting attempts. The patient’s mental status, vital signs, pupils, skin, and nervous system are helpful in detecting organic conditions, particularly the toxidromes associated with common ingestions (see Chapter 145). Altered mental status should be assessed to determine whether the condition is caused by an organic (medical) or functional (psychiatric) cause (Table 113-2). Physical findings associated with chronic disease, alcoholism, and substance abuse should be sought. The physical examination is often overlooked or performed in a cursory manner in patients with psychiatric complaints. Up to 50% of patients with an acute psychiatric presentation harbor unrecognized medical illnesses.87
■ DIAGNOSTIC STRATEGIES Routine toxicologic screening tests are unnecessary in the evaluation of suicidal patients. Nearly all patients with dangerous overdoses and poisonings demonstrate clinical signs within several hours of ingestion. An electrolyte panel, while it need not be routine, may be useful in certain ingestions, particularly if an acid-base abnormality is suspected as in salicylate or methanol toxicity. Although the likelihood of potentially lethal acetaminophen ingestion is small in patients who deny taking acetaminophen, the emergency physician should consider measuring the acetaminophen level in patients with intentional overdose.88 An electrocardiogram should be obtained if cyclic antidepressant overdose is suspected. Patients with acute depression, particularly if newly diagnosed, may need screening tests for underlying medical disorders; however, a
Table 113-1 Modified SAD PERSONS Scale Patients Suspected of Occult or Silent Suicide Attempts Patients who are not overtly depressed or suicidal but who exhibit one or more of the high-risk presentations previously described should be assessed in a sympathetic but direct manner using a “graduated” approach. First, rapport should be established during an assessment of the presenting complaint. This should include a general medical and psychiatric history, as well as an evaluation of the patient’s home, work, and social situation, followed by specific questions regarding the signs and symptoms of depression. The emergency physician should ask direct questions regarding suicide, such as, “Have you ever had the thought that life is not worth living?”; “Do you have thoughts of killing yourself now?”; and “What plans, if any, have you made to do this?” Patients who are not depressed or suicidal are generally not offended by this approach, and it does not place the concept of suicide into the mind of someone who has not been considering it. Patients who are
FACTOR
POINTS ASSIGNED
Sex (male) Age (≤19 or ≥45 years) Depression or hopelessness Previous attempts or psychiatric care Excessive alcohol or drug use Rational thinking loss Separated, divorced, or widowed Organized or serious attempt No social supports Stated future intent
1 1 2 1 1 2 1 2 1 2
Five points or fewer, questionable outpatient treatment; 6 or more points, emergency psychiatric treatment/evaluation; more than 9 points, psychiatric hospitalization. From Hockberger RS, Rothstein RJ: Assessment of suicide potential by nonpsychiatrists using the SAD PERSONS score. J Emerg Med 6:99, 1988.
Table 113-2 Factors in Differentiating Organic from Functional Psychosis
Organic Functional
ONSET
AGE OF ONSET
HALLUCINATIONS
ORIENTATION
VITAL SIGNS
May be acute Subacute to chronic
Any age 14–40 yr
Often visual Usually auditory
Often disoriented Normal
Abnormal Normal
Chapter 113 / Suicide
to the physician. Because most people who attempt suicide communicate their intent to others at some point, an attempt should be made to interview family, friends, police, and paramedics regarding the patient’s recent actions and possible motivations. They may also provide information regarding the specifics of the current suicide attempt. Although some physicians worry that current federal laws regarding patient privacy conflict with the need to obtain information with the family, an emergency exception to the Health Insurance Portability and Privacy Act rule exists. Section 164.512(j), “Uses and Disclosures to Avert a Serious Threat to Health or Safety,” allows physicians to disclose protected health information without individual authorization “based on a reasonable belief that use or disclosure of the protected health information was necessary to prevent or lessen a serious and imminent threat to health or safety of an individual or of the public.”84 Once the patient is medically stable, the presence of risk factors for suicide should be determined. Such factors may include a history of previous suicide attempts or psychiatric care; a history of excessive alcohol or drug use, both acute and long term; family history of suicide; and signs of depression, including a sense of hopelessness. Patients who have a history of deliberate self-harm (self-poisoning, cutting, burning, or hitting oneself) have a higher risk of suicide, especially male patients.85 Over 5% of people seen at a hospital after self-harm commit suicide within 9 years.86 The patient’s marital status and social support are important factors, and the motivation for and the seriousness of the suicide attempt are assessed. Some physicians ask patients to provide their own lists of “reasons why they want to live”—a sort of reverse score for suicidality. If discharge is being considered, patients should be asked whether they would harm themselves if they were released from the ED. Additional demographic information may be helpful (see Box 113-1). The SAD PERSONS mnemonic can be used to document salient points and facilitate subsequent communications with primary care providers and psychiatrists (Table 113-1).
1468
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
primary care physician or psychiatrist can safely perform this evaluation during follow-up.
■ MANAGEMENT Out-of-Hospital Care Out-of-hospital management must focus on the patient’s injuries and potential harm from poisoning or overdose. If the patient refuses to be transported to the hospital or becomes aggressive, emergency medical personnel should involve law enforcement officers. All states give police the right to place individuals into protective custody if they are suspected of being a danger to self or others. The presence of law enforcement officers, or even the threat of calling the police for assistance, usually ensures patient cooperation during transport. Hospitalized psychiatric patients who are transported to the medical facility by police have an independent risk factor for an imminent suicide attempt.89
Emergency Department The clinical assessment of suicide potential requires an empathetic approach. Patients feel more comfortable discussing personal issues when health care personnel are friendly, nonjudgmental, and supportive. Unfortunately, ED staff may be unsympathetic toward patients who attempt suicide because of religious or philosophic beliefs, lack of formal psychiatric training, or inadequate time and personnel to provide appropriate psychiatric evaluation. They may perceive the patient’s behavior as abusive or manipulative and may become frustrated regarding ineffective disposition and follow-up options. Failure to anticipate and overcome these factors can result in inadequate patient assessment and reinforce these patients’ already low self-esteem.
Medical Clearance The first priority in managing patients is medical stabilization and treatment of injuries, poisoning, or overdose. The second priority is the identification and treatment of associated medical conditions that may underlie a patient’s altered mental status or violent behavior. Patients with significant injury, poisoning, or other medical problems should be hospitalized, sometimes in an intensive care setting, where their medical problems can be treated as they remain under constant observation. Five to 40 inpatients commit suicide for every 100,000 hospitalizations.90,91
Suicide Precautions Most suicide attempts involve minor injury or overdose that can be definitively treated in the ED. Suicidal patients who are calm and cooperative should be placed in an area where they can be safely observed by staff. Having a dedicated “sitter” to watch the patient is helpful and decreases the need for restraints. The sitter should accompany the patient when leaving the area to use the restroom or smoke. No potentially suicidal patient should be allowed to leave the ED before an evaluation is completed. The use of family members as “sitters” is discouraged because they may collude with the suicidal patient to leave or may not intervene if the patient attempts to leave. The use of “wander alert” bracelets, which set off an alarm if a patient wearing one crosses an established threshold may help in monitoring. Security personnel should search all potentially suicidal patients early in the ED stay. Having the patient change into
a hospital gown facilitates removal of weapons, medications, and other possessions that might be used to inflict injury, such as belts, neckties, and long shoelaces. The patient’s room should be cleared of all potentially harmful objects, including medications, instruments, and glass objects. Suture carts, which may contain scissors, scalpels, or other dangerous objects, must also be kept out of reach. Some rooms designed for psychiatric patients may contain a lockable “pull-down” wall, like those used on storefronts at night, that cover medical equipment and cabinets when a psychiatric patient is in the room.
Use of Restraints Mechanical and chemical restraint use is based primarily on the physician’s impression regarding the immediate risk of elopement or subsequent suicide attempt as well as personnel available to contain the patient. Some authorities believe that placing a depressed patient in mechanical restraints can impair rapport and contribute to the patient’s diminished self-esteem. Chemical restraints may calm a violent patient but may make subsequent psychiatric evaluation more difficult in the short term. Nevertheless, restraints may be essential and even lifesaving for uncooperative, violent, or psychotic patients and for those at high risk for elopement or self-harm. The Joint Commission, as well as state and federal government, have stringent requirements regarding the use of restraints. A hospital policy that conforms to these guidelines and an approved and consistently employed restraint flow sheet are important to comply with these regulations. A timed and dated physician order, as well as frequent rechecks of distal neurovascular function, are required for all patients placed in restraints.
Determination of Risk Once a patient has demonstrated suicidal behavior or ideation, the physician must determine whether the risk is imminent (i.e., within 48 hours), short term (i.e., within days to weeks), or long term.7 The likelihood of an impending repeat attempt will drive disposition: whether psychiatric hospitalization, emergency psychiatric consultation, or discharge and referral for follow-up. One must consider the potential lethality of the method chosen; for example, ingesting a handful of birth control pills is less worrisome than shooting oneself or setting oneself on fire. It is also important to consider if the patient believed the attempt would result in death. Even if the method chosen was nonlethal from a medical standpoint (such as ingestion of antibiotics), but the patient believed it was lethal, the patient may be at risk for a future, more lethal attempt. Patients who plan and hide their suicide attempt may be more desperate to die than someone who makes the attempt in front of a family member who rescues them. In depressed patients, an intense sense of desperation is an important predictor of suicide.92 Another important consideration in the evaluation of the suicidal patient involves the wish to live versus the wish to die. Not surprisingly, those patients with a wish to live are six times less likely to commit suicide than those who wish to die.93 Although one should determine an individual patient’s likelihood for committing suicide if discharged, this assessment is far from exact. No single psychological test can accurately predict suicidal attempts.94 In a group of 4800 psychiatric patients who were followed prospectively over 5 years, 44% of all suicides were not foreseen by the psychiatrist.95 In another group of individuals who completed suicides after evaluation by an emergency psychiatric service, no specific factors could be identified that predicted imminent suicide.96
1. Are you here because you tried to hurt yourself? 2. In the past week, have you been having thoughts about killing yourself? 3. Have you ever tried to hurt yourself in the past? 4. Has something very stressful happened to you in the past few weeks? Any positive answer on this rapid screen correlated with potential risk for self-harm when compared to a longer Suicidal Ideation Questionnaire, but correlation with suicidal outcomes is unknown. Another simple scoring system is the British Manchester Self-Harm Rule. It uses the following four clinical correlates to determine future risk of suicide attempts104: 1. Any history of self-harm 2. Previous psychiatric treatment 3. Benzodiazepine use in this attempt 4. Any current psychiatric treatment For 9086 patients, any positive answer on the four-question Manchester Self-Harm Rule had a sensitivity of 94% and specificity of 25% for a repeat attempt. Whether this rule
1469
would perform as well in a U.S. population with a far greater access to handguns remains unclear. Suicide assessment should be performed after the patient metabolizes any drugs or alcohol. Intoxicated patients who complain of depression or state ambivalence regarding their future intentions to commit suicide may disavow these feelings once they are sober (but may still be at risk despite the disavowal). In addition, the information obtained from a potentially suicidal patient is best confirmed with a family member or friend as patients who are determined to commit suicide may give false or misleading information. The crises that precipitate suicide attempts are often time limited, usually lasting from a few hours to a few days. If a crisis has passed or can be adequately addressed, the risk of subsequent suicide is substantially diminished. Hospitalization or emergency psychiatric evaluation should be strongly considered when a patient cannot or will not participate in an evaluation of the current crisis or when the problem is unlikely to be resolved. Ultimately, the assessment of suicide risk remains a highly individualized process. The crisis that precipitated the suicide event, the patient’s current emotional state, and the presence or absence of a supportive home environment must also be considered. When emergency physicians are uncertain regarding the need for hospitalization, they should err on the side of caution and either admit the patient for psychiatric care or request emergency psychiatric evaluation. A psychiatric social worker or other paraprofessional may assist in gathering information and in making decisions about the need for hospitalization; however, the physician still must make an independent judgment about the patient’s suicide risk.105 Despite the medical-legal threats involved, nearly a quarter of emergency physicians occasionally send patients with suicidal ideation home without any evaluation by a mental health professional.3 Some investigators recommend hospitalization for adolescents who have attempted suicide and cannot be adequately monitored at home. However, discharge to home is an option for those with suicidal thoughts if urgent psychiatric follow-up is arranged and the caregivers can adequately supervise and protect the youth. Discharged adolescents should not be actively suicidal, should not have access to lethal methods, should have a supervising adult to closely monitor them, and should have a mental health evaluation before ED discharge whenever feasible.68
Involuntary Commitment Many patients who are severely depressed or suicidal will agree to be hospitalized for further evaluation and care; however, others may resist hospitalization. Patients who refuse recommended medical treatment usually do so because of anger or fear. Patients may be angry for being brought to the ED against their wishes or for having to wait for evaluation. Alternatively, they may fear the loss of control associated with hospitalization or the perceived negative stigma associated with a psychiatric disorder. When a patient is reluctant to be hospitalized, the physician should attempt to identify and address the specific concerns. The patient’s family and friends may help convince the patient to accept voluntary hospitalization. Involuntary admission may be necessary if the physician believes the patient may inflict self-harm. Depression alone is not a criterion for involuntary commitment; the accepted legal standard requires the imminent risk of harm to self or others. In some states, even patients who voluntarily agree to hospitalization may need involuntary commitment papers filled out if they are an acute danger to self or others. This is to provide
Chapter 113 / Suicide
There are at least 31 different English-language scales devised to predict the risk of suicide, but the vast majority are not designed for or suitable for use in the ED.97 Some researchers are even using “fuzzy logic” and neural networks to achieve computerized prediction models for suicidality.98 One study evaluated six clinical suicide assessment scales to identify high-risk patients: the Modified SAD PERSONS scale, revised Beck Depression Inventory, Beck Anxiety Inventory, Beck Hopelessness Scale, Beck Scale for Suicidal Ideation, and the High-Risk Construct Scale.99 The outcome measured was psychiatric hospitalization admission for suicide risk, not completed suicide or future suicide attempts. All of the scales showed 100% sensitivity and negative predictive value, but lower specificity (38–90%) and positive predictive value (28– 71%). Although scoring systems might help in determining the need for hospitalization, they cannot predict future attempts at self-harm.100,101 Nevertheless, an attempt to determine a patient’s immediate risk of self-harm should occur and, when indicated, be communicated to other health care providers. The SAD PERSONS mnemonic provides a “suicide score” (see Table 113-1) and is well suited for use in the ED. Two points are given for each of four high-risk factors: (1) complaints of depression or hopelessness, (2) existence of an organic brain syndrome or acute psychosis, (3) presence of a well-conceived plan or life-threatening presentation, and (4) expression of determination or ambivalence regarding future suicidal behavior. One point is assigned for other important but less significant factors: male gender; age younger than 19 or older than 45 years; a history of previous suicide attempts or psychiatric care; stigmata of chronic alcoholism or substance abuse or the history of recent increased use of these substances; a patient who is separated, divorced, or widowed; and the absence of social support systems, such as close family, friends, job, or active religious affiliation. A SAD PERSONS score of 6 or more has a sensitivity of 94% and a specificity of 71% compared with formal psychiatric evaluation in identifying the need for hospitalization in patients who present immediately after a suicide attempt.102 A score of less than 6 has a negative predictive value of 95%. No deaths at 6 to 12 months occurred in patients with low scores. Another risk assessment tool used for adolescents is the Risk of Suicide Questionnaire.103 The four most useful questions on the Risk of Suicide Questionnaire were the following:
PART III ■ Medicine and Surgery / Section Eight • Psychiatric and Behavioral Disorders
1470
the legal basis for holding them against their will during transport if they change their mind about hospitalization during the trip between the ED and the receiving psychiatric facility. Civil commitment statutes differ among the 50 states and the District of Columbia. Nearly 85% of the jurisdictions require dangerousness to self to be the result of a mental illness, and only two jurisdictions mandate attempts at involuntary commitment if a person is deemed to be an imminent harm to self.106 The efficacy of commitment as a long-term preventive measure is controversial.93 Involuntary commitment is not proven to prevent future suicide and may even precipitate adverse psychiatric consequences (e.g., increased feelings of hopelessness and dependency) or cause rebellion in some patients. Many authorities believe that people who are determined to kill themselves will probably prevail despite the best efforts of family members and health care professionals.7 Still, despite the lack of hard data, involuntary commitment remains a primary intervention when patients are deemed acutely suicidal.
■ DISPOSITION Most patients at high risk for suicide will need psychiatric evaluation or hospitalization in a psychiatric facility. If patients require transfer to such a facility, they must be protected from self-harm during transport. Deaths from jumping from an ambulance during such transports have been reported.107 Many patients who attempt suicide or have symptoms of depression can be safely managed as outpatients, if the risk of subsequent suicide is judged to be acceptably low (Box 113-2). Before discharging a patient, one should address the crisis that precipitated the suicide attempt. The patient should also be considered low risk for subsequent suicide (e.g., low SAD PERSONS score). Evaluation by a mental health professional is useful, especially if the safety of outpatient management is in doubt. Some physicians ask the patient to form a verbal or written “contract” with them. This “no harm” agreement usually involves patients vowing not to hurt themselves and agreeing to return to the ED to seek help if the situation worsens before follow-up. Although the “no harm” agreement has not been validated, some believe this is a reasonable approach, whereas others believe that it provides only a false sense of security for the physician. On its own, it will certainly not absolve the clinician from malpractice liability if the patient commits suicide.108
BOX 113-2 Factors for Patients at Low Risk for Suicide 1. Few significant risk factors (e.g., low SAD PERSONS score) 2. Stable and supportive home environment 3. Patient agrees to “no harm” contract and will return to emergency department if situation worsens 4. Family member or friend staying with or available to patient 5. Phone contact with health care provider responsible for follow-up 6. Specific appointment made for follow-up within 24 to 48 hours 7. No gun in home 8. Young female who took a nonlethal ingestion or made “hesitation cuts” to wrists 9. Patient expresses a strong desire to live
If the patient is to be discharged, a family member or friend should agree to stay with the patient or to be immediately available to the patient until follow-up is provided. The patient should be discharged to a stable and supportive home environment that is free of guns and lethal medications. Adequate disposition may include a conversation with the ongoing provider. If possible, the follow-up appointment should be scheduled within a few days of discharge and should be specific regarding the location and time. This approach maximizes patient compliance with follow-up. Providing a card that indicates how to contact an available physician may be helpful.109
■ DOCUMENTATION Documentation is important when patients are committed or discharged, as well as when restraints are used. If a patient requires involuntary commitment, one should document why the patient is a danger to self or others. If the patient is to be discharged, the record should reflect that the patient is low risk and does not intend self-harm after leaving the ED. Documenting that there is no gun in the home is also useful. Preformatted charts may improve documentation.110
■ PREVENTION The incidence of suicide parallels the incidence of alcoholism, drug abuse, and psychiatric disease in society. Although most suicide prevention programs are of mixed value, legislation to control access to lethal drugs and handguns is effective. Laws requiring prescriptions for all sedative and hypnotic drugs in Japan led to a decrease in their use for suicide, with no increase in the use of other methods.111 In Canada, suicide decreased after gun control laws were tightened.112 Keeping a gun locked and unloaded and storing ammunition locked in a separate location are each associated with a protective effect against suicide by firearms in children and teens.113 Screening for suicidal ideation or depression may be a useful strategy. Physician education in depression recognition and treatment may reduce suicide rates.114 Programs that screen for depression in the elderly,115 and school-based programs that screen for suicide risk in teens,116 may decrease suicidal ideation and suicide attempts. Occult suicidality is common in ambulatory ED patients who present for nonpsychiatric problems. In one study, waiting room patients were screened for suicidality with a computerized mental health assessment tool. Eleven percent had passive thoughts of suicide while another 8% had active thoughts about killing themselves.117 While screening for occult suicide in all ED patients is an intriguing notion, the implications of referring 20% of all ED patients for psychiatric evaluation are staggering. Contacting patients by telephone 1 month after being discharged from an ED for deliberate self-poisoning may reduce the number of reattempted suicides over 1 year.118 Despite great interest in the prevention of suicide, it is unclear which, if any, interventions are effective to prevent future attempts at self-harm.119 No strong evidence suggests that antidepressants prevent self-harm in patients with prior suicide attempts; however, one small study did show an advantage to using depot flupenthixol in multiple repeaters.120 In another small study, dialectic behavior therapy was more effective than standard aftercare in preventing further episodes of self-harm.121 Suicide attempts correlate with future suicide. In one study on survivors of self-poisoning, the 5-year mortality rate from suicide was 65.5 times greater than expected in the female group (compared to a control group who did not attempt selfharm) and 41.5 times greater among males.122
1471
KEY CONCEPTS Suicide is often provoked by a treatable or reversible short-term crisis. ■ Suicidal patients frequently see a physician shortly before their death. ■ The most complete information can be elicited with an empathetic approach to the patient and communication with family members, friends, health care providers, and others. ■ Suicide precautions in the ED include appropriate use of “sitters” and, when necessary, physical and chemical restraints and involuntary commitment.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
■
The emergency physician should identify risk factors for suicide, even though determination of suicide risk is not a hard science. The SAD PERSONS score can be of help in documenting this assessment. ■ Older men and those who attempt suicide using a firearm are at highest risk for a future completed suicide. Young females, especially those who cut themselves or take a nonlethal ingestion, are generally at lower risk. ■ If patients are sent home because their risk of suicide is low, ensure a safe and supportive gun-free environment and early psychiatric follow-up.
Chapter 113 / Suicide
■
•
Section Nine Immunologic and Inflammatory Chapter 114
Arthritis
Bruce D. Adams and Douglas W. Lowery III
■ PERSPECTIVE Socioeconomics Arthritis and the over 100 related rheumatic diseases represent the leading cause of disability in the Western World.1,2 Many of the arthritides, including rheumatoid arthritis (RA), are associated with premature mortality and their treatments often manifest with major toxicity.3 Because of the tremendous pain and disability associated with joint inflammation, patients with the gamut of arthritic maladies present frequently for emergency evaluation.2 Up to 25% of all patients with rheumatologic disorders visiting the emergency department (ED) require admission, one third of these to an intensive care unit.4 There are three reasons to develop a comprehensive approach to patients with arthritis symptoms: to utilize the symptoms as a diagnostic clue to a serious systemic illness (e.g., acute rheumatic fever [ARF], scleroderma, and renal crisis), to evaluate for acute emergencies related to or masked by the disease, and to alleviate acute and chronic pain.4,5
“nodes” as early as 1802, and Bouchard described his in 1884, but the entity of osteoarthritis was not clearly defined until 1907.15 The other connective tissue diseases—systemic lupus erythematosus, systemic sclerosis, and polymyositis—were all described in the middle to late 1880s.15 Ankylosing spondylitis was first mentioned in 1831 but was not accurately described until the 1930s.15 The generic term reactive arthritis is now preferred over the eponym of the discredited Nazi war criminal Hans Reiter, who was probably not the first to describe the syndrome.17–19 Did the course of history shift because of this group of rheumatic maladies? Gout, long known as the “disease of kings” because of its predilection for diets of rich foods and alcohol, afflicted emperors throughout the ages.20–22 Despite the untreated pain of RA, Presidents Jefferson, Madison, and Franklin Roosevelt succeeded to the extent of being honored on U.S. currency and the artists Renoir and Rubens both painted masterpieces for many years.16,23 Some have even theorized that gout led to the fall of the Roman Empire (via its association with lead poisoning) and the rise of the American one (by afflicting its gluttonous European colonizers).24–26
History Rheumatic diseases (derived from the Greek rheuma, “a substance which flows”) were among the first diseases recognized. Egyptian physicians more than three millennia ago recognized gout, as did Hippocrates in the 5th century bce (who described other joint ailments, including scleroderma).6,7 Podagra, the foot-torturess and offspring of Aphrodite (Venus) and Dionysus (Bacchus), was a bad-tempered virgin who attacked victims after they overindulged to dietary or sexual excess.8 Roman surgeons under Nero first applied colchicine alkaloids from the meadow saffron (Colchicum autumnale) over 2000 years ago.8,9 Sir Thomas Sydenham in the 17th century advanced more scientifically modern descriptions of gout (from which he suffered), but it was his keen observations of ARF, to include his eponymous chorea, that are probably better remembered from the man that first advised modern physicians “primum non nocere.”10,11 Dundas proposed the term acute rheumatic fever in 1808 and Trousseau linked the rash of Scarlatina with it in 1873.12,13 Its causal relationship with streptococcal infection was finally confirmed in 1900.14 Similarly, Swediaur noted a relationship between urethritis and arthritis in 1784, but the discovery of gonococcal arthritis waited until 1883.15 Garrod in 1858 coined the term rheumatoid arthritis, although Landre-Beauvais in 1800 and Brodie in 1819 are credited with its early scientific descriptions.16 Heberden described his 1472
■ PRINCIPLES OF DISEASE Anatomy and Physiology Joints are designed to bear weight and allow motion with as little wear as possible.27 Three classes of joints are identified: synarthroses (suture lines of the skull), amphiarthroses (fibrocartilaginous unions of the pubic symphysis and the lower third of the sacroiliac joint), and diarthroses (moving joints). The most common type is the diarthrosis or synovial joint, which consists of two ends of subchondral bone (one convex, one concave) almost completely covered by articular cartilage.27 The cartilage consists of a matrix of collagen fibers and proteoglycans, which are synthesized by the chondrocytes within it. The cartilaginous surfaces are well lubricated and slide against each other.28 The joint is surrounded by a capsule that is supported by ligaments, tendons, and muscle and is lined with a synovial membrane (Fig. 114-1). Cartilage is deformable, compressible, and lubricated by synovial fluid secreted by cells of the synovial membrane lining the joint space. The synovium is up to three cells thick and consists of two cell types: type A cells, which contain lysosomes, and type B cells, which synthesize the fluid. Both types multiply in synovitis and interact with the vasculature to produce arthritis.28 Joint fluid has a high viscosity because
1473 Sites and types of rheumatic disease (site: pathophysiology [typical disease])
Figure 114-1. Anatomic structures of the joint with location of selected arthritis diseases. (Redrawn from Goldman: Cecil Medicine, 23rd ed. Copyright © 2007 Saunders, An Imprint of Elsevier.)
Joint Synovial space membrane Tendon
Muscle
Bone
Periosteum Bony end plate
Bone
Joint capsule
Enthesis Cartilage
A
of its major component, a polysaccharide, hyaluronic acid. The fluid also contains water, glucose, electrolytes, and proteins of low molecular weight.
Pathophysiology The disease-specific pathologies are further detailed in the following sections, but the final common pathways of arthritis are initially triggered by trauma, infection, or endogenous cell and humoral inflammatory components.29 The joint’s metabolic balance then shifts towards catabolic mediators and tissue destruction.30 The trigger for this inflammatory reaction is different with different diseases. In nongonococcal bacterial arthritis, the cells of the synovial lining phagocytize bacteria. In gout and pseudogout, crystals are released from cells lining the synovium by conditions that precipitate an acute attack.31 Joint inflammation of the traditional rheumatic diseases has a complex immunologic basis.32,33
■ CLINICAL FINDINGS
Bony end plate: (osteonecrosis)
B
Cartilage: Cartilage degeneration (osteoarthritis)
Enthesis: Enthesopathy (ankylosing spondylitis)
Table 114-1 Causes of Joint Pain MONARTICULAR
POLYARTICULAR
Acute (6 wk) Adult-onset Still’s Osteoarthritis disease Relapsing polychondritis Rheumatoid arthritis Seronegative spondyloarthropathies
Symptoms and Signs Pain is the hallmark complaint of patients with joint problems who come to the ED (Table 114-1). The pain may be acute or chronic (with the typical duration cutoff being 6 weeks), or an acute episode in a person with chronic disease. The patient may have had similar pain before, so it is important to know whether a diagnosis was previously made and what treatment, if any, was instituted. One must first determine whether the source of the inflammation or pain is articular or periarticular (outside the joint capsule). True arthritis produces generalized joint pain, warmth, swelling, and tenderness. Discomfort increases with both passive and active motion of the joint because the inflamed synovium is exquisitely sensitive to stretching and because all parts of the joint are involved in the inflammatory process. By contrast, periarticular inflammation (bursitis, tendinitis, or localized cellulitis) tends to be more focal. Tenderness and swelling do not occur uniformly across the joint, and pain is produced only with certain movements, with the most common being resisted active contraction or passive stretching of the affected muscles or tendons.
Patterns If the site of the patient’s pain is articular, one must next determine whether the arthritis is monarticular or polyarticular. The arthritis may be symmetrical (e.g., rheumatoid or drug induced) or asymmetrical (e.g., rubella, ARF, or gonococcal). In addition, it may also be migratory (e.g., gonococcal or rubella), subsiding in one area before presenting in another, or additive, remaining in the first joint and progressing to additional joints (Box 114-1).
Distribution The distribution of joint involvement may give some clues to the disease: the first metatarsophalangeal (MTP) joint is classically affected in gout; the metacarpophalangeal (MP) joints and proximal interphalangeal (PIP) joints in RA; and the distal interphalangeal (DIP) joints and the first carpometacarpal joint in osteoarthritis.34 Patients with inflammatory arthritis may have low-grade fever, but high fever with chills is more likely
Chapter 114 / Arthritis
Bursa
Joint space: Joint infection Muscle inflammation microcrystalline arthritis (gout) (myositis) Synovial Bursa: membrane: Localized site Synovitis (bursitis) (rheumatoid arthritis) Muscle
1474
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
Differential Diagnosis of Arthritis in the
BOX 114-1 Emergency Department
Monarticular Septic arthritis Gout Pseudogout Osteoarthritis Trauma/hemarthrosis Charcot’s joint Polyarticular Symmetrical Gonococcal arthritis Viral arthritis Lyme disease Drug-induced arthritis Reiter’s syndrome Rheumatic fever Seronegative spondyloarthropathies Asymmetrical Gonococcal Acute rheumatic fever Lyme disease Systemic lupus erythematosus Immune complex diseases (viral) Reactive arthritis
to be caused by septic arthritis. Morning stiffness (gel phenomenon) and improvement of symptoms with activity suggest inflammatory arthritis, while improvement with rest suggests mechanical disorders.34 Concomitant renal stones suggest gout, genital ulcerations occur in Behçet disease and reactive arthritis, and a purulent urethral discharge suggests gonococcal arthritis or reactive arthritis. The use of isoniazid, procainamide, and hydralazine can precipitate lupus, and thiazides can increase the serum uric acid level, leading to gouty arthritis. Many of the newer RA medicines convey serious potential toxicity.35,36
Physical Examination General Examination A thorough physical examination should be performed that specifically searches for evidence of particular rheumatic diseases. The skin, eyes, and cardiac, pulmonary, and neurologic systems should be examined carefully, searching for systemic manifestations of rheumatic disorders (Table 114-2).
Joint Examination The most important diagnostic tool for evaluating acute arthritis is the examiner’s own hands. Each joint in question should be specifically examined for the following attributes: warmth (the dorsum of the hand can detect a 0.5° C difference); effusion; synovial thickening; deformity; range of motion; pain on actively loaded motion; and tenderness (generalized or localized, articular or periarticular).34,37 Spine. When evaluating the spine, the patient should stand, and the vertebral column should be assessed for abnormal curvature or asymmetry. One should perform the Schober maneuver to assess for the limitation of the lumbar spine motion that occurs in ankylosing spondylitis while examining the sacrum and anterior iliac crests to elicit pain in the sacroiliac joints (Fig. 114-2).37
Table 114-2 Clinical Signs of Arthritis Diseases ORGAN SYSTEM
FINDINGS
DISEASES
Airway Cardiac
Airway obstruction Pericarditis Murmurs Iritis, uveitis Conjunctivitis IBD Dysentery Lesions, urethral discharge Aplastic anemia Anemia of chronic disease Cauda equina syndrome Cervical spine instability Ulcerations Pleuritis, nodules Renal crisis, ARF Plaques on elbows, knees Sclerodactyly, calcinosis Erythema chronicum migrans Tophi Erythema marginatum Subcutaneous nodules
RA, RP RA, ARF ARF, RP, AS Spondyloarthropathies Reactive arthritis Spondyloarthropathies Reactive arthritis Reactive arthritis, gonococcemia Parvovirus Any chronic arthritis disease AS
Eyes Gastrointestinal Genitalia Hematologic Neurologic
Oral mucosa Pulmonary Renal Skin
AS, RA, OA Reactive arthritis RA Scleroderma Psoriasis Scleroderma Lyme disease Gout Rheumatic fever Rheumatoid arthritis
Excludes rheumatic vasculitis diseases. ARF, acute rheumatic fever; AS, ankylosing spondylitis; IBD, inflammatory bowel disease; OA, osteoarthritis; RA, rheumatoid arthritis; RP, relapsing polychondritis.
Upper Extremities. Localized tenderness and pain associated with active movement are more likely to be periarticular in origin. A shoulder affected by chronic arthritis or bursitis will have atrophy of the deltoid muscle. Generalized tenderness and pain, both at rest and with active and passive motion, suggest joint involvement. With the patient’s hands placed behind the head, one may then perform the back tests for external and internal rotation, respectively. Early signs of joint inflammation in the elbow are limitation of extension and an increase in the normal angle at which the patient holds the elbow at the side. The hand and wrist provide many clues to the presence of long-standing rheumatic diseases: MP and PIP joints are affected in RA; and the first carpometacarpal, PIP, and DIP joints are affected in osteoarthritis. The fingers may be swollen or sausage-like in appearance, an indication of psoriasis or reactive arthritis. Subluxation at the MP joints, ulnar deviation, and swan neck deformities occur in RA. The nails may have pitting characteristic of psoriatic arthritis. Although the wrist may not be obviously swollen, discomfort and decreased range of motion, particularly on extension, may indicate synovial involvement. Hip. Inflammation affecting the hip joint can be reported by the patient as pain in the anterior thigh, knee, or groin. A hip joint effusion will cause the patient to hold the hip partially flexed. An externally rotated and abducted leg in a neonate strongly suggests infection.38 Range of motion of the hip is most easily tested by flexing the hip, bending the knee at a
1475 *
10
*
0
*
15
*
15 20
* *
A
Internal rotation 45°
B
Figure 114-2. A, Schober’s maneuver. Limited lumbar mobility secondary to spondylitis is documented with an abnormal Schober’s maneuver. To perform this maneuver, the clinician makes a mark at the level of the posterior superior iliac spines (dimples of the pelvis) and another mark 10 cm above the first mark with the patient standing. The patient is then asked to bend forward, attempting to touch his or her toes. The distance between the two marks is remeasured. With normal lumbar motion, the distance between the two points increases by at least 5 cm. (Redrawn from Frontera: Essentials of Physical Medicine and Rehabilitation, 1st ed. Copyright © 2002 Hanley and Belfus.) B, Loss of lumbodorsal spine mobility in ankylosing spondylitis. The lower spine remains straight when the patient bends forward. (From Behrman: Nelson Textbook of Pediatrics, 17th ed. Copyright © 2004 Saunders, An Imprint of Elsevier.)
right angle, and rotating the heel medially and laterally to test for external and internal rotation, respectively (Fig. 114-3).39 Knee. An effusion of the knee joint is relatively easy to detect when it appears as a ballotable fullness medially and laterally. Small effusions can be detected by examining for a transmitted fluid wave. Fullness of the popliteal fossa may indicate a Baker cyst. Passive range of motion may elicit crepitus or clicking. Tibiotalar joint effusions produce swelling under the medial malleolus and make it difficult to palpate the extensor hallucis longus tendon. Tenderness, warmth, and swelling of the great toe MTP joint occur in cases of gout but can also occur with osteoarthritis and RA. Sausage-like swellings of the toes are seen in reactive arthritis.
External rotation 45°
Figure 114-3. Testing for intrinsic disease of the hip joint. (From Branch WT: Office Practice of Medicine, 2nd ed. Saunders, 1987.)
The Radiograph of a Joint Can Be Surveyed
BOX 114-2 Using This Systematic Approach38
A—Alignment: Rheumatoid arthritis manifests with ulnar deviation. B—Bone mineralization: Osteoporosis is more typical of the inflammatory arthritides, while periostitis (reactive bone formation) indicates seronegative spondylarthropathies or osteoarthritis. C—Calcification: Gouty tophi and typical linear cartilage calcification of CPPD. D—Distribution: RA is typically symmetrical. Specific target sites suggest diagnoses. E—Erosions: Overhanging edges and sclerosis characteristic of gout. S—Soft tissue: Assess for soft tissue swelling, nail hypertrophy (psoriatic arthritis), and sclerodactyly (scleroderma).
Radiology ■ DIAGNOSTIC STRATEGIES
Plain Radiographs
Laboratory Tests
Plain radiographs are more helpful in patients with chronic disease than in those with acute arthritis (Box 114-2).43–45 Common findings that help distinguish the different forms of arthritis are set out in Table 114-3.43
Laboratory tests other than synovial fluid analysis generally convey only modest diagnostic value in evaluating acute arthritis in the ED.38–40 Three widely used screening tests are a white blood cell (WBC) count, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP). These values are modestly sensitive but quite nonspecific.39,41 An elevated ESR can be used to screen for inflammatory arthritis when taken into careful clinical context.40 Likewise for CRP with the notable exceptions of scleroderma, polymyositis, and dermatomyositis.40 The ESR is particularly useful when suspecting polymyalgia rheumatica or giant cell arteritis.40 Antistreptolysin O titers or a throat culture provides evidence of antecedent streptococcal infection in ARF. Rheumatoid factor, antinuclear antibody, HLA B27, and other disease specific serologies are generally best deferred to the follow-up physician.41 The serum uric acid level is not helpful in diagnosing acute gouty arthritis because it can actually normalize during the acute phase.42
Computed Tomography, Magnetic Resonance Imaging, and Sonography Other radiologic modalities are occasionally performed as part of the workup in an emergency setting. Ultrasonography is useful in evaluating joint effusions and synovitis.46 Computed tomography (CT) detects early sacroiliac joint disease in difficult cases and is the preferred method for evaluating the sternoclavicular joint.47 Sonography, CT, and magnetic resonance imaging (MRI) have been used to evaluate acute hip pain in children to distinguish septic arthritis and transient synovitis.44 MRI is excellent for (1) imaging cruciate ligaments of the knee, (2) detecting early edema in periarticular structures and fluid collection in tendon sheaths, and (3) determining the extent of cartilage destruction.44
Chapter 114 / Arthritis
Patient lying on examining table
0
1476
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
Arthrocentesis Arthrocentesis is the most important diagnostic modality for evaluating the acutely inflamed joint, especially when considering septic arthritis.45,48
Indications and Contraindications The emergency indications for arthrocentesis in evaluating joint pain include obtaining joint fluid for analysis, draining tense hemarthroses in patients with hemophilia (of the elbows, knees, or ankles and after the appropriate clotting
Table 114-3 Common Radiologic Findings in Arthritis ARTHRITIS
FINDINGS
Acute arthritis (gout, pseudogout, septic arthritis) Late septic arthritis (need at least 8–10 days to see changes)
Soft tissue swelling
Late pseudogout (knee, hip, radiocarpal, midcarpal, all MP)
Degenerative arthritis (acromioclavicular, first carpometacarpal, first MTP, DIP, knee, hip, cervical spine, lumbosacral spine) Tuberculous arthritis (knee, hip, shoulder)
Late rheumatoid arthritis (wrist, MP, PIP, MTP, first IP, foot, atlantoaxial, glenohumeral)
Subchondral bone destruction Periosteal new bone Loss of joint space Osteoporosis Late joint space narrowing Linear calcification in cartilage Asymmetrical joint space narrowing MP “hook spurs” in HHC Osteophyte formation Subchondral cyst formation Lack of osteoporosis Asymmetrical joint space narrowing Sclerosis of juxta-articular bone Bone spurs and cysts—adjacent to severe cartilage degeneration No osteoporosis Soft tissue swelling Marked demineralization Bony rarefaction Little reactive sclerosis Late bony destruction Joint space preserved Symmetrical joint space narrowing Osteoporosis of periarticular bone Marginal erosions (no overhanging margins as in gout) Little reactive bone formation
DIP, distal interphalangeal; HHC, hereditary hemochromatosis; IP, interphalangeal; MP, metacarpophalangeal; MTP, metatarsophalangeal; PIP, proximal interphalangeal.
factor replacement), and instilling analgesics and antiinflammatory agents for the treatment of acute and chronic arthritis.49–51 Emergency arthrocentesis is relatively contraindicated with an overlying cellulitis unless the infected area can be avoided during the puncture. Coagulopathy is the other major relative contraindication, but arthrocentesis can be safely performed in the presence of therapeutic levels of warfarin, ideally with a smaller needle.52,53 Arthrocentesis of prosthetic joints should only be performed to rule out infection and are best done in consultation with an orthopedic surgeon.54
Complications The primary complications of arthrocentesis are bleeding or infection in the joint space, allergic reaction to anesthetic agents, and long-term corticosteroid-related complications. Dry taps (when no fluid is aspirated after joint puncture) are more common in patients with chronic arthritis owing to obstructing tophi or anatomic abnormalities in the synovium and periarticular tissues.55,56 Using a smaller syringe or a larger needle may help in such cases.49 Even if no synovial fluid is apparent in the syringe barrel after multiple attempts, the capped syringe can be sent for culture analysis and just a single drop from the needle bevel can be used for crystal analysis on a slide mount.49 An average on the slide of fewer than two WBCs per high-power field predicts a noninflammatory effusion.50
Technique The patient should be positioned comfortably, with adequate exposure and cushioned support for the joint. Muscle tension during the procedure can reduce the joint volume, making the procedure more difficult; thus, every opportunity to provide for the patient’s comfort should be employed. Bony landmarks should be carefully palpated. Under aseptic technique, the skin should be prepared with an appropriate surgical scrub. Adequate local anesthesia can be achieved either by use of a vapor coolant or by local infiltration with anesthetic solution such as 1 or 2% lidocaine.51 Using an 18- to 22-gauge needle attached to a syringe, the joint space is punctured and aspirated while carefully avoiding abrasion of the articular cartilage. After aspiration, a long-acting anesthetic can be instilled to alleviate pain, but only if septic arthritis has been excluded can therapeutic steroids be injected.49
Synovial Fluid Examination Analysis of synovial fluid is essential for identifying crystalline and suppurative causes of acute arthritis (Table 114-4).
Table 114-4 Synovial Fluid Interpretation PROPERTY
CLASS I (NONINFLAMMATORY)
CLASS II (INFLAMMATORY)
CLASS III (SEPTIC)
CLASS IV (HEMORRHAGIC)
Color Viscosity WBC count (/mm3) Differential Culture Leading diagnosis
Clear/yellow High 100,000 >95% PMNs Positive Bacterial arthritis
Opaque, may contain fat droplets (lipohemarthrosis)
Variable Trauma, coagulopathy
PMNs, polymorphonuclear neutrophils. From Fye KH, Morehead K: Joint aspiration & injection. In Imboden J, Hellman D, Stone J (eds): Current Rheumatology Diagnosis & Treatment. New York, McGrawHill, 2007, p 12.
1477
Chapter 114 / Arthritis
Figure 114-4. Pseudogout (A, blue crystal with a more plump structure) and gout (B, yellow crystal in parallel refraction with a longer, more needle-like structure) crystals under polarizing light. (From Goldman L: Cecil Textbook of Medicine, 23rd ed. Copyright © 2007 Saunders. An Imprint of Elsevier.)
A
B
General Appearance
Electrocardiography
Bedside inspection of synovial fluid for color, clarity, and viscosity can provide diagnostic clues. Viscosity is judged using the “string test”; normal viscosity allows stretching of a “string” of fluid between the thumb and forefinger, whereas inflamed fluid acts more like water.57,58 Hemarthrosis presents after acute trauma or in the presence of coagulopathy, particularly with hemophilia.55 Lipohemarthrosis indicates ligamentous injury or an intra-articular fracture.49,56 Brownish synovial fluid may suggest the rare diagnosis of pigmented villonodular synovitis.59
Electrocardiography is indicated for evaluating a patient with acute arthritis for whom a diagnosis of ARF is being considered.66 Prolongation of the P-R interval is a nonspecific minor Jones criterion, but any number of electrocardiographic findings can manifest.66,67
Synovial Fluid Studies Routine laboratory analysis should be sent for cell count, differential, Gram’s stain, and crystal analysis. A positive Gram’s stain is diagnostic, but a negative result for bacteria does not rule out septic arthritis; cultures should also be obtained.48 Synovial glucose or protein is of little assistance.48,57 White Blood Cell Count. The WBC count dictates the inflammatory class of the fluid, but there is unfortunately significant overlap between inflammatory and septic causes of acute arthritis.48,58,60,61 This test can be analogously compared to that of cerebrospinal fluid and meningitis. A very high fluid WBC or polymorphonuclear cell (PMN) pleocytosis indicates infection, but a more modest WBC or differential does not exclude it.48,58 The likelihood ratio (LR) for septic arthritis increases as the joint WBC count rises: a WBC count of less than 25,000/ mm3 = LR of 0.32; a WBC count of greater than 25,000/mm3 = LR of 2.9; a WBC count of greater than 50,000/mm3 = LR of 7.7; and a WBC count of greater than 100,000/mm3 = LR of 28.0.48 However, low WBC counts do occur early in infectious arthritis and in partially treated infections; high WBC counts (>50,000/mm3) can occur in RA, gout, and pseudogout. Most of the cells in both septic and severe inflammatory arthritis are PMNs. Cell counts should not be used to absolutely rule out a septic cause; rather, bacterial cultures (in blood culture bottles) should be obtained if there is any suspicion of infection.58,62 Crystal Analysis. Crystal analysis is best performed using polarizing microscopy of a drop of synovial fluid or postcentrifugation sediment placed on a slide with cover slip (Fig. 114-4).45,49,63 Monosodium urate crystals are needle shaped and negatively birefringent (yellow when parallel to the compensator and blue when perpendicular), ranging in size from 2 to 10 µm. Calcium pyrophosphate crystals, in contrast, are polymorphic, rhomboid, and positively birefringent (yellow when perpendicular to the compensator, blue when parallel).49,64 Crystal and septic arthritis can coexist because the infection could actually represent the precipitant of the former.65
■ DIFFERENTIAL CONSIDERATIONS Although certain disease entities (i.e., RA, gonococcal arthritis, Lyme arthritis [LA]) can be placed in several anatomic or temporal categories, the most useful classification of arthritis to guide the differential diagnosis is by chronicity and number of joints involved (see Table 114-1).32,56,68–70 Some rheumatology textbooks include an intermediate category of pauciarthritis or oligoarthritis, but the simpler category system appears to better facilitate care and decision making in the ED.70,71
■ MANAGEMENT Descriptive guidelines are provided for the diagnostic strategies for patients with monarticular (Fig. 114-5) and polyarticular (Fig. 114-6) presentations. After a specific diagnosis is made, treatment varies by underlying pathology.
■ MONARTICULAR ARTHRITIS—ACUTE Inflammatory conditions of the periarticular soft tissues such as olecranon bursitis, rotator cuff shoulder tendonitis, and prepatellar bursitis can mimic monarticular arthritis, and classically polyarticular conditions such as RA and the seronegative spondyloarthropathies may present initially in only one joint. The emergency patient with monarticular arthritis should be considered to have septic arthritis until proven otherwise.72,73
Septic Arthritis—Nongonococcal Bacterial Epidemiology The incidence of septic arthritis is approximately 2 to 10 cases per 100,000 per year, with an age distribution curve for septic arthritis revealing bimodal peaks for young children and adults over age 55.73–75 Additional risk factors include low socioeconomic status, intravenous drug abuse, alcoholism, diabetes, skin infections, HIV and other immunocompromised states, chronic arthritis (particularly rheumatoid, crystalline, and degenerative osteoarthritis), and following intra-articular corticosteroid injections or prosthetic implants.39,64,66,76–78 Importantly, septic arthritis can occur simultaneously with other forms of arthritis, especially RA and gout.65,79 The diagnosis of
1478
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
Arthralgia limited to one joint
Complete history and physical examination
Significant trauma (+) or focal bone pain
X-ray
Fracture, tumor, or metabolic bone disease
Abnormal
(−) (+) Effusion or signs of inflamation?
Normal
(−) Probable inflammatory process
Unsuccessful
Joint aspiration
Point tenderness or trigger points
(+)
Bursitis, tendinitis, or fibromyalgia
(−)
Successful Re-evaluate
Bloody
>2,000 WBCs? (−) >75% PMNs?
Bone marrow elements present
(+) Coagulopathy, pseudogout, tumor, trauma, or Charcot joint Check: PT/PTT Platelet count Bleeding time
Intraarticular fracture Crystals identified
Positive culture
Monosodium urate (gout) Calcium pyrophosphate dihydrate (pseudogout)
Infectious arthritis *Synoviral fluid culture as well as cervical, urethral, pharyngeal, and/or rectal evaluations for Gonococcus and Chlamydia when suspected.
Osteoarthritis, internal derangement, soft tissue injury, or viral infection
Sterile inflammatory pint fluid
Suspect: RA, JRA, viral, SLE, Lyme, sarcoidosis, or spondylarthropathy Check: CBC, ESR, RF Consider: LFTs, HLA-B27, ANA, Lyme serologies, and pelvis radiographs
Figure 114-5. An initial approach to the patient with monarticular arthritis. The majority of diagnoses will be determined by the history and physical
examination findings. ANA, antinuclear antibodies; CBC, complete blood cell count; ESR, erythrocyte sedimentation rate; JRA, juvenile rheumatoid arthritis; LFTs, liver function tests; PMNs, polymorphonuclear neutrophils; PT, prothrombin time; PTT, partial thromboplastin time; RA, rheumatoid arthritis; RF, rheumatoid factor; SLE, systemic lupus erythematosus; WBCs, white blood cells. (Adapted from American College of Rheumatology Ad Hoc Committee on Clinical Guidelines: Arthritis Rheum 39:1, 1996.)
infectious arthritis in a patient with known crystal arthritis can be thorny because acute flare-ups of pseudogout or gout can cause fever, while crystals can emerge from an infected joint.42,80 Gram’s stain and culture must always be done on joint fluid, even when a diagnosis of crystal arthritis is evident by microscopy.54 Joint infection in injection drug users usually involves the axial skeleton but can involve the extremities.54
Pathophysiology Any joint, whether native, chronically diseased, or prosthetic, can develop septic arthritis.68 Bacterial pathogens infect the joint space most commonly by hematogenous spread, but direct inoculation and contiguous spread from bony or soft tissue
infections occurs also.73,74 Once in the joint, bacteria proliferate essentially unchecked in the highly vascular synovium, which has no limiting basement membrane.68 This induces an inflammatory cascade, synovial proliferation with neovascularization, and subsequent enzymatic, cellular, and cytokine degradation of articular cartilage.61,69 Septic arthritis, unless rapidly recognized and treated, results in serious disabling morbidity with mortality ranging from 5 to 15% or higher.48,68
Microbiology The microbiology of nongonococcal arthritis has remained fairly constant over time, except for the notable decline of Haemophilus and pneumococcal species in the postimmuniza-
1479 Polyarthralgia
Synovitis?
(−)
(+)
Figure 114-6. An initial approach
to the patient with polyarticular joint symptoms. ANA, antinuclear antibodies; CBC, complete blood cell count; ESR, erythrocyte sedimentation rate; RF, rheumatic factor. (Adapted from American College of Rheumatology Ad Hoc Committee on Clinical Guidelines: Arthritis Rheum 39:1, 1996.)
Tender points? (−)
(+)
Fibromyalgia or multiple sites of bursitis or tendinitis or relapsing polychondritis and PMR
Symptoms >6 weeks (+)
Systemic rheumatic disease
(−)
Viral arthritis Early systemic rheumatic disease
Viral arthralgia Osteoarthritis Soft tissue abnormalities Hypothyroidism Neuropathic pain Metabolic bone disease Depression
Careful follow-up
Check: CBC, ESR, RF, and/or ANA; creatine, urinalysis, joint aspiration
Check: Blood count Liver function tests Consider: Hepatitis B and C serology Parvovirus serology
tion era.68,73 Acute nongonococcal septic arthritis in adults is caused most often by gram-positive organisms (75–90%) followed by gram-negative bacilli (10–20%) and then anaerobes, mycobacterium, fungal and other unusual organisms.63,74,75,81 Overall, Staphylococcus aureus is still the most common cause of septic arthritis, with the rapid emergence of methicillinresistant strains.82,83 Neisseria gonorrhoeae accounts for only 20% of cases of all monarticular septic arthritis and more commonly presents with polyarthritis; it is discussed separately. Select populations have higher propensities for specific infecting organisms (Table 114-5).
Prosthetic Joint Infection Prosthetic joint infections are classified as early (50 mm/hr Abnormal findings on temporal artery biopsy From Hunder GG, et al: The American College of Rheumatology 1990 criteria for the classification of giant cell arteritis. Arthritis Rheum 33:1122, 1990.
for any patient with a high clinical suspicion of temporal arteritis. Steroids do not significantly change the results of the biopsy and may prevent progression to visual loss.58 Prednisone should be started at a dosage of 1 mg/kg/day until biopsy can be performed. Patients with severe disease or impending visual loss should be hospitalized and given high-dose steroids until the diagnosis is obtained. The likelihood of visual symptom resolution correlates with the time of corticosteroid initiation: More than 50% of patients treated in the first 24 hours can expect to recover compared to only 6% of patients treated after the first 24 hours.59 Although there is no consensus regarding treatment duration or tapering schedules, most patients are given the initial dose of steroids for 2 to 4 weeks and then a slow tapering of the steroid dosage over several weeks, although relapse rates are lower in those who use them longer (1 or 2 years).60,61
Takayasu’s Arteritis Takayasu’s arteritis (pulseless disease), named for the Japanese ophthalmologist who first described the ocular manifestations in 1908, is a chronic, recurrent, inflammatory vascular disease that primarily affects the aorta and its major branches.62 It is characterized by lymphocytic infiltration and fibrosis of the vessels, resulting in marked thickening of the intima and adventitia, eventually leading to obstruction of the arteries and ischemic complications. Adolescent girls and women in their second and third decades are predominantly affected, and women are affected eight times more frequently than men. The syndrome is most commonly seen in Japan, Southeast Asia, India, and Mexico.63 In the United States, the annual incidence is approximately 1 to 3 cases per 1 million people.56 The cause of Takayasu’s arteritis is unknown. The clinical features can be broadly categorized into two groups: those caused by systemic inflammation and those caused by vascular damage. In the prepulseless or early phase, the diagnosis is difficult. Almost half of patients experience constitutional symptoms including fatigue, weight loss, and low-grade fever caused by systemic inflammation. Hypertension is frequently seen secondary to aortic or renal artery involvement. With progression of the disease, ischemic symptoms caused by the vascular manifestations appear, including diminished or unequal pulses, claudication, retinopathy, lightheadedness (associated with vertebral or carotid artery dissection), and visual loss. Strokes, syncope, subclavian steal syndrome, abdominal pain, and coronary ischemia are also reported. Early diagnosis is difficult because symptoms are nonspecific and there is no specific laboratory test. Disease activity may be subclinical and not recognized by laboratory markers in up to half of cases.64 Later, pulses may be unequal, bruits may be auscultated, and there may be blood pressure differ-
Medium Vessel Vasculitides Polyarteritis Nodosa Polyareteritis nodosa (PAN) is a necrotizing vasculitis of predominantly medium-sized arteries. Immune deposits are minimal or absent, and test results for anti-neutrophil cytoplasmic antibodies (ANCAs) are typically negative. A distinction has been made between PAN and microscopic polyangiitis (MPA). PAN includes vasculitis associated with nervous system and gastrointestinal tract involvement, whereas MPA is positive for ANCAs directed at myeloperoxidase and is associated with nerve, glomerular, and lung tissue. The etiology of PAN is unknown. Viral hepatitis B or C is associated with a vasculitis identical to PAN but is treated differently, and PAN is excluded if chronic hepatitis B or C is found.66 PAN is also linked with drug reactions, serum sickness, and HIV. Patients with classic PAN can be any age, including children, but the peak age at onset is in the fifth or sixth decade, and there is an approximately 2 : 1 male-to-female ratio. Reported annual incidence rate of PAN ranges from 2 to 9 cases per 1 million people.67 The early clinical picture consists of constitutional symptoms including fever, malaise, arthralgias, and myalgias. Cutaneous manifestations occur in up to one third of patients, usually seen as areas of palpable purpura, sometimes with ulceration. Frequent sites include the fingers, the ankles around the malleoli, and the pretibial areas. In severe cases, there may be widespread digital cyanosis secondary to ischemia. Splinter hemorrhages and livedo reticularis are also commonly observed. PAN can progress to peripheral neuropathy and bowel ischemia, and it can cause hypertension related to renal artery inflammation. The diagnosis is made by the clinical pattern and histopathology seen on biopsy. After endocarditis and concomitant infections are ruled out as a cause of the vasculitis, biopsy of the involved segment may reveal the diagnosis, although the irregular distribution of the disease can make obtaining a diagnostic biopsy difficult. Probably the most common biopsy site is the sural nerve. In patients with a neuropathy, especially if the sural nerve conduction is abnormal, the biopsy finding is positive in more than 80% of cases.68 Mesenteric angiography may be useful to demonstrate widespread aneurysms, but these may not be present early in the course of the illness. PAN is treated with corticosteroids, especially in cases without organ involvement. Patients with organ involvement
1505
may need additional therapy with immunosuppressive agents such as cyclophosphamide.69 Active viral hepatitis, if present, should be treated with antiviral therapy.
Buerger’s Disease Buerger’s disease, also known as thromboangiitis obliterans, is an inflammatory vaso-occlusive disease that primarily affects the lower extremities in young adult male cigarette smokers, although women may also be affected. The role of tobacco, especially cigarette smoking, is clear, but the pathogenesis is unknown. In most cases, Buerger’s disease is limited to medium to small arteries and veins of the distal extremities. The disease typically begins with bilateral pain and ischemia of both lower extremities. At onset, the symptoms may be mild, such as paresthesias or pain only with exposure to cold, but most cases rapidly progress to a painful condition with digital cyanosis and severe claudication. Digital ulcers often occur with minor trauma.70 Characteristic angiographic changes include multiple bilateral areas of narrowing or occlusion in the digital, palmer, planter, ulnar, radial, tibial, and peroneal arteries. The disease must be differentiated from premature atherosclerosis and other rheumatic diseases. Treatment consists of abstinence from all forms of tobacco. Affected limbs must be protected from trauma and cold. Calcium channel blockers and pentoxyfylline have reportedly been beneficial in some patients. Therapy using transfer of the gene for vascular endothelial growth factor is under investigation.71 Approximately half of individuals who continue to smoke will require amputation, often multiple times as more proximal vessels become involved.
Small Vessel Vasculitides Behçet’s Disease Behçet’s disease is a chronic relapsing vasculitis characterized by oral and genital ulceration; cutaneous lesions; and ophthalmologic, neurologic, or gastrointestinal manifestations. The prevalence of Behçet’s disease ranges from 1 in 1000 in Japan to 1 in 150,000 in the United States and Europe.72,73 It affects men more often than women, mainly young adults, with age of onset of 25 to 35 years. In Japan, it has been linked to histocompatibility antigen HLA-B5. Behçet’s disease is one of the rare forms of vasculitis also capable of affecting large vessels. Oral ulceration is the hallmark of the disease, tends to be the earliest manifestation, and is required for the diagnosis of Behçet’s disease. Genital aphthae occur slightly less often than oral ulceration. The frequency of eye involvement is approximately 70% and includes bilateral or unilateral iritis, uveitis, and optic neuritis, all of which can lead to blindness.74 An additional hallmark of Behçet’s disease, a hypopyon uveitis, is seen rarely. CNS vasculitis, resulting in meningoencephalitis, intracranial hypertension, or a multiple sclerosis–like syndrome, may develop in 10 to 20% of patients. Gastrointestinal ulceration, including ileocecal perforation, has been reported in Japanese patients. Skin lesions, including erythema nodosum and cutaneous vasculitis, may occur. Cardiac and renal involvement is rare.72,75 Nondeforming arthritis involving the knees and ankles has been described. Laboratory examination is nonspecific. The diagnosis of Behçet’s disease is made when a consistent clinical syndrome is associated with a nonnecrotizing perivascular infiltrate of lymphocytes and monocytes on biopsy of affected tissue. Therapeutic options should be based on the
Chapter 116 / Systemic Lupus Erythematosus and the Vasculitides
ences in the extremities. Definitive diagnosis is made with arteriography demonstrating stenotic lesions, poststenotic dilation, aneurysms, and increased collateral circulation. Treatment with prednisone 1 mg/kg/day for 1 to 3 months induces remission in up to 60% of patients.65 Other cytotoxic agents, such as methotrexate, cyclophosphamide, or azathioprine, or tumor necrosis factor modulatory therapy may be added to achieve remission if relapses occur. Infections may complicate therapy. Complications of Takayasu’s arteritis, such as hypertension, congestive heart failure, angina, or aortic regurgitation, may benefit from other forms of medical therapy. Hypertension may be treated with calcium channel blockers and angiotensin-converting enzyme inhibitors, although this may be especially difficult in patients with extensive disease who may have two or more arterial beds with substantially different blood pressures. Antiplatelet agents may be of benefit. Bypass grafting and endarterectomy are useful in patients with significant disease. Mortality has resulted mainly from renal failure, cardiac failure, or infectious complications of immunosuppressive treatment.64
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1506
degree of involvement.76 Patients with mucocutaneous lesions only may be treated with intralesional topical or aerosolized (not inhaled) corticosteroids. Topical tacrolimus can also be used, often in combination with topical corticosteroids. Patients with more severe mucocutaneous involvement may be treated with thalidomide. Systemic involvement is generally well controlled with glucocorticoids at 1 mg/kg/day. Gastrointestinal disease can be controlled with sulfasalazine 2 to 6 g/day. Patients with eye involvement should be referred to an ophthalmologist because the most common cause of morbidity is ocular involvement, which can cause blindness. Serious manifestations of uveitis and CNS involvement warrant use of azathioprine or cyclophosphamide, and patients should be admitted to the hospital. Deep vein thrombosis associated with Behçet’s disease rarely results in pulmonary emboli but should be treated with systemic anticoagulation.
Wegener’s Granulomatosis Wegener’s granulomatosis (WG) is a necrotizing granulomatous vasculitis involving the respiratory tract, kidneys, and, to varying degrees, the medium to small vessels in other organs. Based on the National Hospital Discharge Survey, the 1986 to 1990 period prevalence of WG was estimated to be approximately 3 per 100,000 people. Both sexes were equally represented. The mean age at onset was 45 years, and most patients were white (80.9%).77 Patients first complain of upper respiratory tract symptoms with sinusitis, otitis, and nasal ulceration. Destruction of the sinus walls may also occur. Lower respiratory tract symptoms include cough, dyspnea, hemoptysis, and asymptomatic pulmonary infiltrates, occasionally with cavitation. Tracheal stenosis occurs in 13% of patients.78 Renal involvement is a later finding with glomerulonephritis, which may be aggressive in 85% of patients.79,80 Eye involvement includes conjunctivitis and scleritis caused by granulomatous deposition in the sclera. Skin lesions include ulcers, nodules, and granuloma formation. Neurologic involvement is rarely a presenting feature of WG, but it may develop during the course of disease in 22% to more than 50% of cases81,82 and includes cerebral vasculitis, granulomatous deposition in cranial nerves, and peripheral nerve vasculitis resulting in neuropathies. Coronary vasculitis, pericarditis, and conduction defects are rare.78,83 Laboratory examination includes findings of a markedly elevated ESR, normochromic normocytic anemia, and, occasionally, thrombocytopenia. Urinalysis may show hematuria, active sediment excretion, proteinuria, and red blood cell casts. Antibodies against cytoplasmic components of polymorphonuclear cells (c-ANCA) have been found to be sensitive and specific for a diagnosis of WG.79 ANAs are usually absent. The chest x-ray study shows multiple sharply demarcated nodular densities, predominantly in the lower lung fields, with pleural effusions in 25% of patients.78 Lymphadenopathy is rarely seen on radiography. Diagnosis of WG is confirmed by lung biopsy. Transbronchial biopsies are rarely diagnostic (70%), mononeuritis multiplex (60%), fevers (55%), and a variety of cutaneous findings (>60%). In contrast, alveolar hemorrhage occurs in a comparative minority (12%).90 MPA usually requires both glucocorticoids and a cytotoxic agent to control disease. The usual regimen to induce remission in patients with severe organ involvement includes high doses of prednisone (often preceded by a 3-day “pulse” of methylprednisolone, 1 g/day) plus cyclophosphamide. Following the induction of remission, patients may be switched to either azathioprine (up to 2 mg/kg/day) or methotrexate (up to 25 mg/week, assuming that residual renal dysfunction does not preclude this medication). Plasma exchange has also been shown to be useful. If MPA is diagnosed early and treated promptly, patients have a high likelihood (>90%) of achieving disease remission, although approximately one third of patients suffer disease flares after the achievement of remission.91
Goodpasture’s Syndrome Goodpasture’s syndrome, or anti-glomerular basement membrane disease (anti-GBM), is characterized by the linear deposition of anti-glomerular basement membrane antibodies in the glomerular and alveolar basement membrane resulting in alveolar hemorrhage and progressive glomerulonephritis. Environmental factors and genetic susceptibility are thought to play a role, but no clear etiology has been identified. All age groups are affected. In adults, the mean age of onset is 20 to 30 years, with a peak incidence in young men age 20 to 30 years. A second peak occurs in those age 50 to 70 years, with men and women equally affected. The annual incidence is estimated to be 1 case per 2 million people among people of white European descent.92 General malaise, weight loss, fever, or arthralgia may be the initial features of anti-GBM disease. Patients may also have cough, dyspnea, and hemoptysis. The principal clinical features of the disease are development of renal failure due to progressive glomerulonephritis or pulmonary hemorrhage. Historically, hemoptysis has been the most common presenting feature, occurring in approximately 70% of reported cases.92
1507
Initially, the pulmonary hemorrhage may be mild, or it may be severe and life threatening. Hypoxia is common. The renal manifestations are varied; some patients have normal renal function, whereas others have a rapidly progressing glomerulonephritis. Patients may also have skin involvement with palpable purpura. Laboratory examination is notable for elevated ESR and urinalysis with red blood cell casts. On blood testing, anti-GBM antibodies can be measured, but the level of circulating antibodies does not correlate with the severity of the disease. Complement levels are normal, and in contrast to findings in Wegener’s granulomatosis, c-ANCA tests are negative. A chest radiograph shows hilar pulmonary infiltrates. The differential diagnosis includes SLE and Wegener’s granulomatosis. Diagnosis is made by renal biopsy. Lung tissue shows pulmonary alveolar hemorrhage, with similar linear deposition of antibodies along the alveolar basement membrane. Management of the airway is the first priority in patients with severe pulmonary hemorrhage. Treatment with methylprednisolone 10 to 15 mg/kg IV is necessary if rapidly progressive glomerulonephritis or severe pulmonary hemorrhage complicates the patient’s course. The use of cytotoxic agents such as cyclophosphamide, as well as plasmapheresis (2–4 L/ day of plasma), has been associated with improvement in pulmonary hemorrhage and glomerular lesions if extensive renal damage has not already occurred. Patients who progress to end-stage renal disease are candidates for transplantation if anti-GBM antibodies return to undetectable levels with treatment; otherwise, the disease may recur in the transplanted kidney.93 Some patients present with the occasional flare of pulmonary hemorrhage. These patients should be admitted to the hospital to be observed for airway complications and for development of renal disease. The prognosis is varied but has greatly improved in the past 15 years because plasma exchange has been used more aggressively.
Henoch-Schönlein Purpura Henoch-Schönlein purpura (HSP) is a small vessel vasculitis involving the skin, gut, and kidney and is characterized by immunoglobulin A (IgA)-dominant immune deposits in target organs. This vasculitis is particularly common in children, although patients of any age can be affected. The annual incidence in children is approximately 15 cases per 100,000 children.94 The male-to-female ratio is 2 : 1. The syndrome occurs most often in the winter and early spring; two thirds of the cases follow an upper respiratory tract infection, with onset an average of 10 days after the start of respiratory symptoms.95 Other inciting agents associated with HSP include insect stings and drugs. The classic full presentation includes the acute onset of fever, palpable purpura on the lower extremities and buttocks, abdominal pain, arthritis, and hematuria. The rash is accompanied by arthralgias of the lower extremities, most commonly the ankles, with swollen, tender joints. The rash can be urticarial or purpuric and is usually in the lower extremities (Fig. 116-2). Frank arthritis is usually absent. Gastrointestinal complaints, seen in 70% of patients, include abdominal pain, nausea, vomiting, and diarrhea with blood and mucus per rectum.79 Renal involvement occurs in 50% of patients with hematuria and red blood cell casts; however, it rarely progresses to renal failure.73 Nervous system involvement is rare, especially in children. The syndrome is relapsing and remitting over several weeks. Antigens to drugs, infectious agents, food, insect bites, and immunizations have been implicated in the pathogenesis of the IgA-dominant immune deposition. Treatment of the
Chapter 116 / Systemic Lupus Erythematosus and the Vasculitides
vasculitic neuropathy, cardiac involvement, or glomerulonephritis, should trigger the use of cyclophosphamide as part of the remission induction strategy.89
1508
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
Drugs Associated with
BOX 116-5 Hypersensitivity Vasculitis Nonsteroidal anti-inflammatory drugs Ibuprofen, diclofenac, etodolac, piroxicam, naproxen Antibiotics β-Lactams, macrolides, sulfonamides, quinolones, antivirals Growth factors G-CSF, GM-CSF Selective serotonin reuptake inhibitors Paroxetine, fluoxetine Antithyroid drugs Propylthiouracil Antihypertensive medications Thiazides, beta-blockers Anticonvulsants Phenytoin Vaccinations Allopurinol Aspirin Acetaminophen Insulin Methotrexate Anti–tumor necrosis factor agents Etanercept, infliximab Figure 116-2. Henoch-Schönlein purpura is a small vessel vasculitis
involving the skin, gut, and kidney. (From Habif TP: Clinical Dermatology, 4th ed. New York, Mosby, 2004.)
precipitating infection or discontinuation of the inciting drug is necessary. The treatment of HSP is an area with little evidence but much ongoing investigation.96 NSAIDs may alleviate arthralgias but can aggravate gastrointestinal symptoms and should be avoided in any patient with renal disease. Dapsone may be effective in cases of HSP, possibly through interference with the interactions of IgA and neutrophils. Children with more severe arthralgias and abdominal pain benefit from prednisone at 1 mg/kg/day orally. Adults who have symptoms may be given prednisone 60 mg/day. Glucocorticoids do not appear to improve the rash, and their effectiveness in renal disease is controversial. Uncontrolled trials suggest that high-dose methylprednisolone followed by oral prednisone or high-dose prednisone combined with azathioprine or cyclophosphamide may help patients with severe nephritis.94 Most patients do well with supportive care because HSP follows a self-limited course, resolving without substantial morbidity. The vast majority of cases resolve within 6 to 8 weeks. The disease may recur in up to 33% of patients. A small percentage of patients have prolonged renal impairment.95
Hypersensitivity Vasculitis Hypersensitivity vasculitis, also known as cutaneous leukocytoclastic angiitis, is a small-vessel vasculitis that is restricted to the skin and not associated with any other form of primary or secondary vasculitis. It can be seen at any age and has no gender predominance. The etiology is often unknown, although drugs seem the most commonly identifiable offender.97 Vaccinations have more rarely been associated with hypersensitivity vasculitis. Onset can be abrupt after exposure to the etiologic agent, but it usually ranges from 7 to 21 days after initial exposure. Although almost any medication can be associated with a hypersensitivity vasculitis, certain classes of drugs are more commonly associated with a hypersensitivity
G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocytemacrophage colony-stimulating factor. Adapted from Harris E, et al (eds): Kelley’s Textbook of Rheumatology, 7th ed. Philadelphia, Saunders, 2005.
vasculitis (Box 116-5).98 In up to 40% of cases, however, no specific cause is identified.99 The skin lesions of hypersensitivity vasculitis include purpura (either palpable or nonpalpable), papules, urticaria/ angioedema, erythema multiforme, vesicles, pustules, ulcers, and necrosis. The lesions typically occur first and most prominently in dependent regions (i.e., the lower extremities or buttocks). The lesions tend to occur in cohorts or “crops” that are the same age. The occurrence of the lesions may be asymptomatic but is usually accompanied by a burning or tingling sensation. Arthralgias and even frank arthritis, with a predominance for large joints, are sometimes present. Laboratory examination is nonspecific, with a mildly elevated ESR and mild leukocytosis. The pathologic finding in the vessels, primarily the postcapillary venules, is immune complex deposition with infiltration by polymorphonuclear leukocytes with or without destruction of vessel walls. Hypersensitivity vasculitis must be distinguished primarily from other small vessel vasculitides, from autoimmune inflammatory conditions associated with joint disease and rashes, and from other cutaneous reactions to medications. Treatment includes discontinuation of the offending agent. In mild cases, symptomatic treatment may be needed. For more severe involvement, corticosteroids (40–60 mg of prednisone per day gradually tapering the dose) are indicated.100
Mixed Cryoglobulinemia Cryoglobulins are immunoglobulins and immune complexes that precipitate in the cold (4° C) and dissolve on rewarming. Mixed cryoglobulinemia syndrome is a consequence of immune complex-mediated vasculitis and is characterized by a typical clinical triad of purpura, weakness, and arthralgias. Middle-aged women are most commonly affected. Precipitat-
Serum Sickness Serum sickness is a type III hypersensitivity reaction that results from injection of foreign proteins or serum. Reactions secondary to the administration of nonprotein drugs are clinically similar to serum sickness reactions. Not all substances that are recognized as foreign will elicit an immune response. The antigen must be of characteristic size or have specific antigenic determinants and physiological properties to be an effective stimulator of the immune system. After an appropriate antigen is introduced, an individual’s immune system responds by synthesizing antibodies. The antibody reacts with the antigen, forming soluble circulating immune complexes that may diffuse into the vascular walls, where they may initiate fixation and activation of complement. Complementcontaining immune complexes generate an influx of polymorphonuclear leukocytes into the vessel wall, where proteolytic enzymes that can mediate tissue damage are released. Immune complex deposition and the subsequent inflammatory response are responsible for the widespread vasculitic lesions seen in serum sickness. Serum sickness has distinctive skin findings. Typically, erythema first occurs on the sides of the fingers, toes, and hands, before a more widespread eruption that is most often morbilliform (in two thirds of patients), sometimes with urticaria. Urticaria is seldom seen alone. Approximately half the cases of serum sickness have visceral involvement. Rash, fever, constitutional symptoms, arthralgia, and arthritis are the most frequent clinical findings.104 Usually, symptoms start 12 to 36 hours after ingestion if there is a previously immunizing exposure, but they may occur up to 7 to 21 days after antigen exposure. The manifestations seen in serum sickness are due to immune complex deposition but not to systemic vasculitis as described in this chapter. In serum sickness, serum C3 and C4 complement levels are markedly decreased.105 Withdrawing the drug usually leads to rapid resolution. Treatment is
1509
supportive with antihistamines and antipyretics. Systemic corticosteroids may benefit some patients. Symptoms usually last 1 or 2 weeks before spontaneously subsiding. Long-lasting sequelae generally do not occur.
Erythema Nodosum Erythema nodosum is a vasculitis of the venules in the subcutaneous layers of the skin. The cause is unclear, but it is usually the result of a hypersensitivity vasculitis from infections, drugs, or a systemic disease. The disease is seen most commonly in spring and fall. The incidence of erythema nodosum is approximately 1 to 5 per 100,000 people. Women are more commonly affected than men, with a male-to-female ratio of 1 : 6. The peak incidence is in the third decade of life.106,107 Erythema nodosum is thought to be caused by a circulating immune complex-mediated process, although the exact pathogenic process is not clear. The hallmark lesions of erythema nodosum are tender erythematous subcutaneous nodules that have a blue hue as they resolve (Fig. 116-3). Symmetrical pretibial involvement is most common, although the extensor surfaces of the forearm, thigh, and trunk may also be affected. Patients may have just the nodules or may have systemic symptoms, including fever and malaise. Arthralgias are seen in 90% of patients at some time during the disease course and have been known to persist for up to 2 years after resolution.107 Erythema nodosum is usually idiopathic, although patients who have erythema nodosum should be considered for underlying diseases such as viral upper respiratory tract infection, streptococcal infection, sarcoidosis, tuberculosis, and drug exposure. Much rarer causes include inflammatory bowel diseases, histoplasmosis, Yersinia, Salmonella, Chlamydia,
Figure 116-3. The hallmark lesions of erythema nodosum are tender
erythematous subcutaneous nodules that have a blue hue as they resolve. (From Habif TP: Clinical Dermatology, 4th ed. New York, Mosby, 2004.)
Chapter 116 / Systemic Lupus Erythematosus and the Vasculitides
ing antigens include hepatitis A, B, and C; cytomegalovirus; or Epstein-Barr virus.101 The clinical features of mixed cryoglobulinemia include leukocytoclastic vasculitis (manifested as palpable purpura), polyarthralgias and arthritis, lymphadenopathy, hepatosplenomegaly, peripheral neuropathy, and hypocomplementemia, especially decreased C4 serum levels. Recurrent palpable purpura occurs in virtually all patients. The most serious involvement is renal deposition of the cryoglobulins, resulting in glomerulonephritis, and it occurs in 20 to 60% of patients.102 Patients may have fulminant or slowly progressive chronic renal disease. Laboratory examination demonstrates an elevated ESR, decreased serum complement levels, and the presence of cryoglobulins. Diagnosis is made clinically in the presence of cryoglobulins; however, it may be difficult to distinguish from SLE or HSP. Treatment depends on the extent of involvement. Patients with disease limited to the skin may try low-dose steroids, and patients with systemic manifestations are usually started on prednisone 60 mg/day orally, although low-dose steroids should be avoided during initial antiviral therapy. Cyclophosphamide has been helpful in controlling systemic disease and allows decreases in steroid dosages. Approximately 75% of patients with mild to moderate hepatitis C-associated cryoglobulinemic vasculitis respond to antiviral therapy initially, but sustained responses are low.103 Plasmapheresis has also been used in life-threatening cases but often requires concomitant immunosuppressive treatment. Rituximab, an antiCD20 monoclonal antibody that depletes B cells, is also being explored as a therapeutic agent.
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1510
coccidioidomycosis, psittacosis, and autoimmune diseases such as SLE. Drugs implicated include penicillins, sulfa drugs, aspartame, phenytoin (Dilantin), and oral contraceptives. The eruption can last up to 6 weeks. Although the lesions are exquisitely tender, erythema nodosum tends to be self-limited. The most common approach is treatment of any underlying disorder and supportive therapy. NSAIDs may be useful in controlling the arthralgias. Disposition depends on suspicion of underlying disease but is usually outpatient follow-up.
hypercoagulable state should be investigated, most of the time venous insufficiency of the lower extremities is usually the only precipitating factor. Treatment of thrombophlebitis is conservative, with application of a support stocking on the involved leg. In chronic and recurrent cases, especially those associated with malignancy, heparin, and fibrinolytic drugs may be used.110
KEY CONCEPTS Panniculitis The panniculitides are a group of heterogeneous inflammatory diseases involving the subcutaneous fat. Biopsy of the involved subcutaneous tissue shows fat cell necrosis, infiltration of inflammatory cells with macrophages, and vasculitis. Several forms of panniculitis exist. The inflammatory infiltrate can involve the septa or the lobule, and vasculitis may or may not be present. Diseases associated with panniculitis are erythema nodosum, erythema induratum, lupus profundus, pancreatitis, superficial thrombophlebitis, α1-antitrypsin deficiency, lightchain paraproteinemia, and C1 inhibitor deficiency.108 Erythema nodosum can exist as a manifestation of systemic disease or as a hypersensitivity to drugs (as discussed previously). Erythema induratum (Bazin’s disease) and nodular vasculitis are synonymous terms used to describe a vasculitis of the skin of the calf associated with tuberculosis. Typical erythema induratum is a disease of middle-aged women in whom erythematous subcutaneous nodules and plaques appear on the posterior aspects of the lower leg. The bilateral lesions are usually tender and begin as nodules but then ulcerate and scar. The course is protracted and recurrent episodes over years or decades are common. Mycobacteria are rarely found in the lesions. Therapy is supportive with dressing changes and elevation, unless evidence of active tuberculosis is found elsewhere.109 Lupus profundus is a chronic recurrent panniculitis that appears in approximately 1 to 3% of patients with cutaneous lupus erythematous. Patients have subcutaneous nodules in the scalp, face, breasts, thighs, and buttocks. The lesions ulcerate and then heal. The differential includes erythema nodosum, but in lupus profundus the lesions are usually more chronic and nontender.109 Superficial thrombophlebitis presents with erythematous, tender subcutaneous nodules with a linear arrangement along a tender cordlike thickening of the involved vein. Usually, these cords are located on the lower limbs and are associated with venous insufficiency. Although a primary or secondary
Systemic Lupus Erythematosus ■ Patients with SLE can have multiple and varied symptom complexes. The diagnosis should be considered in patients with fever, rash, or unexplained systemic complaints. ■ Patients with deep vein thrombosis without risk factors should be considered for APS. Patients with evidence of thrombosis should receive anticoagulation and be admitted for further workup. ■ Febrile patients with SLE receiving immunosuppressive therapy should be hospitalized and treated aggressively because they have a high risk for gramnegative or streptococcal sepsis. ■ Patients with worsening renal function or with involvement of the heart, lungs, or CNS should be hospitalized for aggressive treatment to prevent progression of the disease and symptoms. ■ Patients with symptoms of coronary ischemia should be aggressively treated. Even young patients with risk factors for CAD should be evaluated for coronary ischemia. The Vasculitides ■ The diagnosis of systemic vasculitis is difficult and should be considered in patients with rash and pulmonary or renal complaints. ■ Consultation with a rheumatologist is helpful for determining management when patients have flares of a known vasculitis. ■ Febrile patients receiving immunosuppressive therapy or corticosteroids have a high risk for sepsis or disseminated viral infections and should be treated aggressively. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 117
Allergy, Hypersensitivity, and Anaphylaxis
T. Paul Tran and Robert L. Muelleman
■ PERSPECTIVE Over millions of years, the human immune system has evolved to become a highly complex, elegant, and efficient organ whose chief function is to protect the human host (self ) from harmful offenders (nonself ). Antigens are foreign (or self) molecules that will elicit an immune response. Immunologic responses to antigens in humans are coordinated by two immune systems: the ancient innate immune system, which humans inherited from invertebrates, and the recently evolved adaptive immune system, which is present in humans and vertebrates. The innate system is considered the first line of defense. Its effector components include mast cells, macrophages, dendritic cells, natural killer cells, granulocytes, antimicrobial peptides, complements, and cytokines. Encoded in the germline, receptors in the innate immune system can recognize foreign molecular patterns that are highly conserved in microbes but not in humans, and they immediately begin the process of clearing the antigens. The adaptive system, on the other hand, must allow time for the antigenspecific cells (B and T cells) to amplify through a process known as clonal expansion to mount an effective immune response. Its effector components include B and T lymphocytes and cytokines. The adaptive immune system is characterized by having an immune memory and enormous diversity, capable of recognizing the myriad antigens through a vast library of antibodies and receptors (up to 1015).1 This diversity of antigen receptors is achieved by somatic rearrangement of fewer than 400 genes. Despite the complexity, the two immune systems work in concert and with great fidelity to provide the human host immunity. However, they can overreact, causing allergic disease. For practical purposes, the term allergy is used herein to refer to an IgE-mediated (also known as type I or immediate) hypersensitivity reaction. Antigens that elicit an allergic reaction are referred to as allergens. Cases of allergy and anaphylaxis have historically been documented to the days of antiquity. Pharaoh Menes died of anaphylaxis in 2641 bce. In 1902, Portier and Richet discovered that although a dog tolerated an injection of sea anemone toxin the first time, it died within minutes when injected again several weeks later.2 They coined the term anaphylaxis from Greek (ana, against; phylax, guard or protect), meaning “against protection.” For this and subsequent work in anaphylaxis, Richet was awarded the Nobel Prize in Medicine and Physiology in 1913. Today, anaphylaxis refers to a life-threatening allergic syndrome (i.e., IgE mediated) characterized by multiorgan involvement and rapid
onset. Patients having an anaphylactic reaction would prototypically present in distress with prominent pruritic urticaria, orolaryngeal edema, bronchospasm, hypotension, and central nervous system and gastrointestinal (GI) symptoms. Common allergens that can induce an anaphylactic reaction include drugs, foods, insect stings, and latex. Deaths from anaphylaxis usually result from acute respiratory failure or cardiovascular collapse. Prompt recognition and aggressive treatment in the emergency department (ED) can usually stave off this potentially life-threatening allergic reaction. The term anaphylactoid reaction refers to a syndrome clinically similar to anaphylaxis that is not mediated by IgE. Its clinical presentation and treatment are identical to those of anaphylaxis. Anaphylactoid reactions seem to result from direct degranulation of mast cells (and basophils) and may follow a single, first-time exposure to certain inciting agents. In this chapter, the term anaphylaxis is used to refer to both IgE- and non-IgE-mediated reactions, obviating the need for the term anaphylactoid reaction.
Definition, Epidemiology, and Risk Factors for Anaphylaxis Although anaphylaxis has traditionally been considered a clinical syndrome, there has been no universal agreement on a clinical definition of an anaphylactic reaction; anaphylaxis is usually defined in mechanistic terms as an IgE-mediated event. In an attempt to provide a working definition that clinicians can apply in the clinical setting, the National Institute of Allergy and Infectious Disease, in collaboration with the Food Allergy and Anaphylaxis Network, advanced a set of clinical criteria for identifying and diagnosing anaphylaxis (Box 117-1).3 In the most common scenario encountered in the ED, the diagnosis of anaphylaxis is considered “highly likely” if a patient presents with acute onset (within minutes after an exposure) of rash or mucosal swelling and respiratory insufficiency or hypotension. The epidemiology of anaphylaxis is not known with certainty, and estimates in medical literature vary widely. As a measure of risk, the annual incidence rate for all-cause anaphylaxis is generally believed to be 21 per 100,000 personyears4; the case mortality in this series is 0.65%. As a measure of burden of disease, the prevalence rate of anaphylaxis ranges from 1.21 to 15.04% for the U.S. population.5 Given the U.S. population of 272 million in 1999, these prevalence figures imply that between 3.3 million and 43 million Americans are at risk for an anaphylactic reaction. It is commonly believed that there are probably 100,000 cases of anaphylaxis per year 1511
1512
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
BOX 117-1 Clinical Criteria for Diagnosing Anaphylaxis Anaphylaxis is highly likely when any one of the following three criteria is fulfilled: 1. Acute onset of an illness (minutes to several hours) with involvement of the skin, mucosal tissue, or both (e.g., generalized hives, pruritus or flushing, swollen lips-tongue-uvula) AND AT LEAST ONE OF THE FOLLOWING: a. Respiratory compromise (e.g., dyspnea, wheezebronchospasm, stridor, reduced PEF, hypoxemia) b. Reduced BP or associated symptoms of end-organ dysfunction (e.g., hypotonia [collapse], syncope, incontinence) 2. Two or more of the following occurring rapidly (minutes to several hours) after exposure to a likely allergen for that patient: a. Involvement of the skin-mucosal tissue (e.g., generalized hives, itch-flush, swollen lips-tongue-uvula) b. Respiratory compromise (e.g., dyspnea, wheezebronchospasm, stridor, reduced PEF, hypoxemia) c. Reduced BP or associated symptoms (e.g., hypotonia [collapse], syncope, incontinence) d. Persistent gastrointestinal symptoms (e.g., crampy abdominal pain, vomiting) 3. Reduced BP after exposure to known allergen for that patient (minutes to several hours): a. Infants and children: low systolic BP (age specific) or greater than 30% decrease in systolic BP* b. Adults: systolic BP of less than 90 mm Hg or greater than 30% decrease from that person’s baseline *
Low systolic blood pressure for children is defined as less than 70 mm Hg from 1 month to 1 year, less than 70 mm Hg + 2 × age from 1 to 10 years, and less than 90 mm Hg from 11 to 17 years. BP, blood pressure; PEF, peak expiratory flow. From Sampson HA, Munoz-Furlong A, Campbell RA, et al: Second symposium on the definition and management of anaphylaxis: Summary report—Second National Institute of Allergy and Infectious Disease/Food Allergy and Anaphylaxis Network symposium. J Allergy Clin Immunol 117:391, 1996.
in the United States; of these, approximately two thirds are new cases and 1% are fatal.6 The cause of anaphylaxis remains unidentified in more than one third of cases.7,8 When a cause can be determined, food, drugs, insect stings, and latex are the most common etiologic agents. Food is the cause in one third of known cases, most commonly peanuts and crustaceans.7 Insect stings, particularly bee and wasp stings, are an important cause of anaphylaxis, occurring in 0.5 to 5% of the population and accounting for 50 to 100 deaths per year in the United States.5 Penicillin allergy occurs at a frequency of one to five reactions per 10,000 treatments, with 1 per 50,000 to 100,000 treatments resulting in a fatality.9 Radiocontrast medium (RCM) reactions occur in 0.22% of patients given the hyperosmolar contrast agents but in only 0.04% of patients given the newer nonionic agents.10 Latex, the natural rubber widely used in the manufacture of medical products (latex gloves and latex medical devices), is a common source of allergy but an uncommon cause of anaphylactic reactions.5 Although race, geographic location, and occupation do not seem to predispose individuals to the risk of anaphylaxis, other risk factors tend to show a multifactorial pattern. For example, atopy is a risk factor for anaphylaxis in an all-cause anaphylactic series.11 This is particularly true when the allergen is
administered by the mucosal route (e.g., food). Atopy does not, however, seem to be a risk factor when the allergen is administered parenterally (e.g., penicillin). Limited data tend to suggest that economic status, age, sex, and season of the year seem to affect the risks of having an anaphylactic reaction. Anaphylactic reactions seem to be more common in summer and early fall (the outdoor season), in people of higher socioeconomic status, in women older than 30 years, and in adults. The dose, frequency, duration, and route of administration of a drug also affect the tendency to develop an anaphylactic reaction, with the parenteral and topical routes more likely to lead to an anaphylactic reaction. One interesting aspect of drug-related anaphylaxis is the constancy of administration. An anaphylactic reaction may not occur in a susceptible patient as long as a drug is administered at regular intervals. The same patient may, however, experience an anaphylactic reaction if the drug is resumed after an interruption of therapy. Lastly, the more distant the last exposure, the lower the risks of anaphylaxis upon reexposure, presumably because of some forgetfulness of the immune memory.
■ PRINCIPLES OF DISEASE Development of the Immune System and Mechanism of Immune-Mediated Injury The adaptive and innate immune systems originate from the common pluripotential hematopoietic stem cells, which are derived from the yolk sac and later reside in the bone marrow. These stem cells differentiate and develop into the lymphoid precursor cells and the colony-forming unit for granulocyte, erythroid, myeloid, and megakaryocyte (CFU-GEMM) stem cells. The lymphoid precursor cells in turn differentiate into bursa-equivalent lymphocytes (B cells), thymus-derived lymphocytes (T cells), and natural killer (NK) cells; the CFUGEMM cells, in the meantime, develop into mast cells, basophils, and others (Fig. 117-1). When the body encounters an allergen, the cellular components of the adaptive immune system interact with the cellular and protein components of the innate immune system to mount a concerted defense aimed at neutralizing and removing the harmful allergen.
T Cell Development Lymphoid precursor cells migrate from the bone marrow into the thymus, where they progress through ontogeny. Under regulation by cytokines and cell-cell interaction, these precursors undergo gene rearrangement and positive and negative selection. In the process, T cells acquire the T cell antigen receptors and various surface markers and eventuate into two main T cell lineages. Using the cluster of differentiation (CD) classification, there are principally two types of mature T cells that eventuate out of the thymus: CD4+, also called helper T cells, and CD8+, also called suppressor T cells. Depending on the type of cytokine produced, T helper cells are subdivided into type 1 helper cells (Th1) and type 2 helper cells (Th2), with opposing activities. Whereas Th1 cells inhibit IgE production and IgE isotype switching, Th2 cells stimulate IgE production and IgE isotype switching. The balance of these stimulatory and inhibitory activities of the Th1 and Th2 cells is believed to determine an individual’s propensity to develop allergic disease or atopy and may help explain the increased prevalence of allergy in urbanized and Western societies in the past three decades. Early in utero and soon after birth, naive T lymphocytes in the infant’s immune system are dominated by the allergy-prone Th2 cells and their associated cytokines (interleukins [ILs] 4, 5, and 13). These cytokines are
1513
B Cell
Lymphoid Precursor
NK Cell
Pluripotent Stem Cell Mast Cell
CFU-GEMM
Basophil
Monocyte Erythrocyte Neutrophil Eosinophil Dendritic Cell Platelet
Figure 117-1. Developmental pathways of the immune and
hematopoietic systems. CFU-GEMM, colony-forming unit for granulocyte, erythroid, myeloid, and megakaryocyte; NK, natural killer. (Redrawn from Shearer WT, Fleisher TA: The immune system. In Middleton E, et al [eds]: Allergy: Clinical and Practice. St. Louis, Mosby, 1998, pp 1–13.)
important inducers for production of IgE antibodies. Later, during infancy through early childhood and adolescence, the nonatopic infant’s immune system gradually shifts from this allergy-prone Th2 environment to an allergy-protective Th1 environment. The cytokines associated with this Th1 environment include IL-2 and interferon-γ. This shift is thought to be caused by the continual exposure of the young individual’s immune system to allergenic stimuli from the surrounding environment, mainly microbes. Features of Western lifestyles (“hygiene hypothesis”) such as changes in infant diets, widespread use of antibiotics, smaller family size, and cleaner childcare are believed to reduce this stimulatory antigenic exposure in an individual’s early years, leading to an environment in which the immune system is dominated by a persistent allergyprone Th2 system. This lack of balance between the two immune systems supposedly leads to atopy and thus an allergyprone population.12,13
B Cell Development and Immunoglobulins B cell ontogeny can be divided into antigen-independent and antigen-dependent stages. During the antigen-independent stage, B cells mature in primary lymphoid organs (bone marrow
Classification of Reactions There are four types of hypersensitivity reactions in the Coombs and Gell classification system. Type I (immediate hypersensitivity) is IgE (and IgG4) mediated and accounts for most allergic and anaphylactic reactions observed in humans. Type II (cytotoxic) denotes antibody-mediated cytotoxic reaction. In this scheme, complement-fixing IgG (or IgM) engages cell-bound antigen, activating the classic complement pathway, leading to the fixation of membrane attack complexes and cell lysis. In the process, anaphylatoxins C3a and C5a may also cause mast cell mediators to be released, producing the same action as the classic mediators of anaphylaxis. Type III (immune complex) is IgG or IgM complex mediated. Circulating soluble antigen-antibody immune complexes migrate from the circulation to deposit in the perivascular interstitial space, thereby activating the complement system. Anaphylactic reactions to blood transfusions or blood component therapy, including serotherapy (immunoglobulin administration), are clinical examples of the overlap of type II and type III reactivity; thus, they have been classified as complement-mediated or immune complex-mediated anaphylaxis. Type IV (delayed hypersensitivity) is T cell mediated and has no documented relationship to the pathogenesis of anaphylaxis.
■ PATHOPHYSIOLOGY Mast cells (and basophils) and their contents are the central effector cells and mediators in allergy and anaphylaxis. Exposure of a genetically predisposed individual to a novel allergen leads to the synthesis and release of allergen-specific IgE by plasma cells into the circulation. Fixation of this allergenspecific IgE to surface receptors on mast cells (FcεRI) com-
Chapter 117 / Allergy, Hypersensitivity, and Anaphylaxis
T Cell
and fetal liver), where they undergo gene rearrangement in a stochastic manner and acquire various surface markers. Later during the antigen-dependent stage in the secondary lymphoid organs (lymph nodes and spleen), B cells differentiate into memory B cells and plasma cells and are ready to secrete immunoglobulins. Throughout B cell ontogeny, B cell maturation, isotype switching, and immunoglobulin production are driven by activated T cells, cytokines, and interaction with antigen and bone marrow stromal cells. Immunoglobulins are protein molecules composed of two identical polypeptide heavy chains and two identical polypeptide light chains, covalently linked by disulfide bonds (Fig. 117-2). The heavy (H) chains have one variable domain, VH, and three or four constant domains, CH. The light (L) chains have one variable domain, VL, and one constant domain, CL. The variable domains of the heavy and light chains together form a pair of identical antigen binding sites and, together with the adjacent constant heavy domain pair, make up the Fab (antibody-binding fragment) region of the immunoglobulin molecule. The remaining constant domains of the heavy chains together form the Fc (crystallizable fragment) region of the immunoglobulin molecule. The Fc binds to the surface receptors of effector cells such as mast cells, B cells, or macrophages. There are five isotypes or classes of immunoglobulins, IgG, IgA, IgM, IgD, and IgE, with isotype IgG having four subclasses (IgG1, IgG2, IgG3, and IgG4) and IgA two subclasses (IgA1 and IgA2). The body usually produces IgM antibodies when it first encounters an antigen. Repeated antigenic exposure, however, may cause the constant region of the IgM to switch to another class (IgA, IgG, or IgE), a process also known as isotype switching. Isotype IgE (and IgG4) is the most important antibody in the pathogenesis of allergic disease and anaphylaxis.
1514
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
NH2
VL
VH
VL
NH2
CL CL
VH
CH1
CH1
CH2
CH2
CH3
CH3
Fab
Fc
Allergen COOH Immunoglobulin molecule Ig is composed of a pair of heavy chains and a pair of light chains with variable (V) and constant (C) domains.
IgE
Mast cell
Figure 117-2. Activation of
mast cells with degranulation of mast cell mediators by antigen (Ag) cross-linking adjacent immunoglobulin E (IgE) on cell surface. PAF, platelet-activating factor.
Preformed mediators (histamine) Lipid mediators (leukotrienes, PAF, prostaglandins, thromboxane) Cytokines
pletes the process known as sensitization. These IgE-bearing mast cells usually reside in the mucosal surfaces, submucosal tissue (around venules), and cutaneous surfaces, and they are capable of becoming activated upon reexposure to a specific allergen. Cross-linking of the FcεRI on the mast cells by a specific multivalent allergen sets off a cascade of conformational and biochemical events, leading eventually to the degranulation of preformed mediators and the generation and release of arachidonic acid metabolites and cytokines from mast cells (and basophils). At the target tissue level, these mediators cause enhanced capillary permeability, vasodilation, smooth muscle contraction, sensory nerve stimulation, myocardial depression, and activation of secondary inflammatory pathways. These pathologic events result in clinical manifestations that include flushing syndrome, urticaria and angioedema, pruritus, nausea, vomiting, diarrhea, abdominal pain, chest pain, dyspnea, wheezing, respiratory insufficiency and failure, dizziness, syncope, hypotension, and shock.
Immunoglobulin E–Mediated Signal Transduction System Following cross-linking of FcεRI, signal transduction is initiated by a tyrosine kinase, termed Lyn, which is constitutively associated with the FcεRI receptor (Fig. 117-3).14,15 Subsequently, Lyn phosphorylates the immunoreceptor tyrosinebased activation motif (ITAM) of the subunits of the FcεRI, leading to the activation and binding of protein tyrosine kinases (PTK) with SRc homology domain 2 (SH2). One such PTK is spleen tyrosine kinase (Syk). An activated Syk subsequently leads to further tyrosine phosphorylations of several “adapter” proteins. Principally among these are (1) the phos-
phorylation of linker for activation of T cells (LAT), which serves as the attachment site for several proteins including leukocyte-specific phosphoprotein of 76 kd (SLP-76), and (2) the phosphorylation of phospholipase C γ (PLC-γ). Phosphorylated PLC-γ generates diacylglycerol (DAG) and inositol1,4,5-triphosphate (IP3) from membrane phospholipids. DAG in turn activates protein kinase C, which promotes exocytosis of preformed granules and cytokine transcription factors. IP3 binds to receptors in the endoplasmic reticulum, causing a rise in intracellular Ca2+ and resulting in a depletion of Ca2+ store, which leads to more influx of calcium from extracellular space via the activation of the Ca2+ release-activated Ca2+ channel (ICRAC). Spikes in intracellular calcium activate a number of calcium-gated kinases, including phospholipases, which cleave membrane phospholipids to generate lysophospholipids, which in turn facilitate the fusion of secretory granules with the cell membrane, leading to the exocytosis of the secretory granules. Through the LAT/SLP-76 multimolecular complex, tyrosine phosphorylation and activation of a number of enzymes and adapters lead to the activation of JNK and ERK (among others) and ultimately the synthesis and release of cytokines and arachidonic acid (AA) metabolism. The end result is the secretion of preformed mediators, AA metabolites, and cytokines (Box 117-2) into the circulation, which act on target organs to cause the clinical syndrome of anaphylaxis.
Mediators of Anaphylaxis The mediators released by the mast cells and basophils can be categorized into three main groups: preformed mediators, lipid-derived metabolites (AA metabolism), and cytokines (see Box 117-2). Of the preformed mediators, histamine is the most
n
E
A ll e
PL
F
ERK
c εR
cPLA2
I
F
c εR
Lyn
Sy
k
I
LAT SLP76
Arachidonic Acid Metabolism
PLC-γ
PL IP3
Cyclooxygenase
DAG
PGD2
PKC
5-Lipoxygenase 5-HETE
JNK
Degranulation
↑Ca
2+
Extracellular Influx
Cytokine Gene Transcription
LTA4
LTC4
LTB4
LTD4 LTE4
Figure 117-3. Signal transduction sequence following the cross-linking of adjacent immunoglobulin E (IgE) on the surface of mast cell. cPLA2, cytoplasmic phospholipase A2; DAG, diacylglycerol; ERK, extracellular signal-regulated kinase; HETE, hydroxyeicosatetraenoic (acid); IP3, inositol triphosphate; JNK, S-Jun N-terminal kinase; LAT, linker for activation of T cell; LTA4, leukotriene A4; Lyn, Src-like protein-tyrosine kinase; PGD2, prostaglandin D2; PKC, protein kinase C; PTK, protein tyrosine kinase; SLP76, leukocyte-specific phosphoprotein of 76 kd; Syk, spleen tyrosine kinase.
BOX 117-2
Mediators of Activated Mast Cells and Basophils
Preformed Mediators (Granule) Histamine Tryptase Carboxypeptides Chymase Cathepsin G Heparin Proteoglycans Arachidonic Acid Metabolism Products (Membrane) LTB4 LTC4 PGD2 Platelet-activating factor Cytokines (Nucleus) TNF CCL2, CCL3, CCL5 GM-CSF IL-3, -4, -5, -6, -8, -10, -13 CCL, CC chemokine ligand; GM-CSF, granulocyte-macrophage colonystimulating factor; IL, interleukin; LTB4, leukotriene B4; PGD2, prostaglandin D2.
Physiologic Effects The chemical mediators released from mast cells exert their effect on various target organs to produce the clinical syndrome of anaphylaxis. Increased vascular permeability can lead to urticaria, angioedema, laryngeal edema, nasal congestion, or gastrointestinal swelling with abdominal cramping and
Chapter 117 / Allergy, Hypersensitivity, and Anaphylaxis
rg
e
Ig
LAT
1515
clinically apparent and is responsible for the immediate symptoms. Histamine is an essential mediator in immediate hypersensitivity and inflammation, and infusion of histamine has been shown to produce the majority of the clinical features of anaphylaxis syndrome.16 Histamine is produced and stored in preformed granules (pg) in mast cells and basophils, at approximately 1 or 2 pg/cell. There are three classes of histamine receptors—H1, H2, and H3—that mediate the activity of histamine in the body. H1 receptor stimulation produces bronchial, intestinal, and uterine smooth muscle contraction; increased vascular permeability; nasal mucus production; coronary artery spasm; and increased eosinophil and neutrophil chemokinesis and chemotaxis. H2 receptor stimulation increases the rate and force of ventricular and atrial contraction, gastric acid secretion, airway mucus production, and vascular permeability while also causing bronchodilation and inhibition of basophil histamine release. H3 receptors, found in neurons (in the central nervous system) and peripheral tissues, control the synthesis and release of histamine. The roles of the other preformed mediators in mast cell and basophil degranulation syndrome are not well delineated. In contrast to the preformed mediators, cytokines and lipid metabolites are elaborated following the activation of mast cells and basophils (see Fig. 117-3). Prostaglandin D2 (PGD2) is the main AA metabolite released by activated mast cells (but not basophils). PGD2 (and thromboxanes) is synthesized from AA by the cyclooxygenase pathway (via both COX-1 and COX-2). PGD2 is responsible for causing hypotension, inhibition of platelet aggregation, and bronchospasm; PGD2 is approximately 30 times more potent than histamine in causing bronchoconstriction. The leukotrienes—LTB4, LTC4, LTD4, and LTE4—also referred to as cysteinyl leukotrienes or slowreacting substances of anaphylaxis, are synthesized from AA via the lipoxygenase pathway. LTB4 and LTC4 are first synthesized intracellularly in mast cells and basophils and then secreted; LTC4 is subsequently converted to LTD4, and LTE4 in the extracellular space (by gamma-glutamyl transpeptidase and dipeptidase). They are involved in cholinergicindependent bronchospasm, increased vascular permeability, and increased mucous gland production. These cysteinyl leukotrienes have a slow onset but are 10 to 1000 times more potent than histamine in causing bronchoconstriction when administered by aerosol.17 They also have a longer duration of action and potentiate the effects of other bronchoconstrictors such as histamine. Platelet-activating factor (PAF) is an unstored phospholipid and the most potent compound known to cause aggregation of human platelets with subsequent release of platelet-derived vasoactive mediators. Its other actions include neutrophil activation and chemotaxis and ileal and parenchymal lung strip smooth muscle contraction. PAF has been demonstrated to produce many of the important clinical manifestations of anaphylaxis, including decreased myocardial contractile force, coronary vasoconstriction, pulmonary edema, and a prolonged increase in total pulmonary resistance with a decrease in dynamic compliance. Indeed, blockage of PAF with experimental antagonists leads to improved cardiac function, suggesting that PAF may be involved in the late cardiac dysfunction and lethality associated with anaphylaxis.18
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1516
vomiting. Vasodilation can lead to flushing, headaches, reduced peripheral vascular resistance, hypotension, and syncope. Contraction of smooth muscle can lead to bronchospasm, abdominal cramping, or diarrhea. Pulmonary vessel vasoconstriction can lead to pulmonary hypertension, pulmonary edema, and decreased cardiac filling pressures. Coronary vasoconstriction can lead to myocardial ischemia and decreased myocardial contractile force. Changes in atrial chronotropy and ventricular and atrial isotropy can lead to cardiac dysrhythmias. In addition to the direct actions on the target tissues, these preformed mediators, lipid-derived mediators, and cytokines activate a number of inflammatory pathways, including the complement system, clotting and clot lysis systems, and the kallikrein-kinin (contact) system, to contribute to the clinical manifestations of allergy and anaphylaxis. Cardiovascular collapse in anaphylaxis has classically been described as a result of peripheral vasodilation, enhanced vascular permeability, leakage of plasma, and intravascular volume depletion (the “empty ventricle” syndrome). However, hemodynamic reports for humans experiencing anaphylactic shock indicate that the explanation may be more complicated. In a variety of clinical settings, hypotension in anaphylaxis has been associated with an initial increase in cardiac index, which may become depressed, and altered peripheral and pulmonary vascular resistance (increased/decreased). In the setting of decreased cardiac index and elevated peripheral vascular resistance, organ perfusion is compromised, resulting in metabolic acidosis. The use of pressors alone in this situation may not improve hemodynamics as peripheral blood vessels are already maximally vasoconstricted. Aggressive volume expansion with crystalloid (or colloid) is needed in this scenario. Pathologic features identified at autopsy in fatal cases of anaphylaxis are most commonly observed in the respiratory and cardiac systems. These include orolaryngeal edema, pulmonary hyperinflation, peribronchial vascular congestion, intra-alveolar hemorrhage, pulmonary edema, increased tracheobronchial secretions, and eosinophilic infiltration of the bronchial walls. Death from asphyxia is usually caused by angioedema of the epiglottis, larynx, hypopharynx, and, to some extent, the trachea. Patients who die of vascular collapse show varying degrees of myocardial damage, visceral congestion, and other findings suggestive of a loss of intravascular blood volume. Other autopsy findings include urticarial eruptions, angioedema, visceral congestion, submucosal edema, and hemorrhagic gastritis. Notably, autopsy findings may also be normal after an anaphylactic death.19 A summary of the physiologic effects and clinical signs and symptoms is given in Table 117-1.
■ ETIOLOGY Numerous agents are known to cause anaphylaxis in humans. They are organized by immunopathogenetic mechanism: IgE mediated, immune complex mediated, nonimmunologic activators, or AA modulators (Box 117-3). Reactions without identifiable causative agents are classified as physically induced or iatrogenic anaphylaxis (IA).
Immunoglobulin E–Mediated Agents This diverse group of agents includes foods, antibiotics, latex, drugs, and Hymenoptera stings.
Foods Foods are the major identifiable causative agents, accounting for approximately one third of the cases of anaphylaxis.7 A
variety of foods ranging from the well-known, such as nuts, shellfish, and eggs, to the obscure, such as chamomile tea (which may have cross-reactivity with ragweed), have been identified. Cow’s milk, egg, peanut, soy, wheat, fish, shellfish, and tree nuts are foods that most commonly cause anaphylaxis. Even for a person with a known history of food allergy, it may be difficult to avoid foods that may cause allergic reactions because their identity may be obscured in processing (e.g., wine contaminated with Hymenoptera venom).20 Because allergenic foods are first absorbed transmucosally, symptoms of food anaphylaxis may first appear localized to the upper airway of the respiratory tract. When anaphylactic allergens are administered parenterally, symptoms of anaphylaxis tend to be more cardiovascular and systemic. Allergic reactions to foodstuffs are more common in children, ranging from 0.3 to 7.5%.21 Therapeutic and prophylactic use of large quantities of antibiotics is common in the production of beef cattle, swine, fish, poultry, and sometimes vegetables and fruits. Along with antibiotics, sodium and potassium bisulfites and metabisulfites are used as preservatives in foods. Sulfites have been used as antioxidants in the food and restaurant industry to prevent discoloration of vegetables (e.g., salad bars and avocado dips), fruits, and potatoes and to preserve fruit and vegetable juices. They are also used to prevent bacterial contamination and oxidation of wines, beers, and distilled beverages. Sensitivity to ingested sulfites has been well documented, especially among the asthmatic population.22 Establishing a particular foodstuff or preservative as the causative agent of anaphylaxis can be difficult.
Antibiotics Benzylpenicillin, semisynthetic penicillin, and cephalosporins are the most commonly used antibiotics, with penicillin perhaps the most commonly reported medication allergy. The first penicillin-induced anaphylactic fatality was reported in 1949.23 Because of their low molecular weights, these antimicrobials do not themselves possess antigenic properties. Immunologically, they are haptens, simple chemicals that are not antigenic in themselves but become antigenic after the chemicals or their metabolites form a stable bond with the host proteins. Certain binding properties of particular drugs make them more likely to induce sensitization. Although patients often report a history of penicillin allergy, this usually does not stand up to close scrutiny.24 Studies have shown that up to 9 of 10 individuals with a reported history of penicillin allergy can safely use penicillin; these individuals usually are either mislabeled as penicillin allergic or lose their allergy after years of avoidance.25 Depending on the studies, the frequency of allergic reactions to penicillin varies from 0.01 to 0.05% of administrations of penicillin (1 to 5 reactions per 10,000), with an anaphylactic reaction rate less than 0.01% and a fatality rate less than 0.002% (less than 1 fatality per 50,000 penicillin administrations).9,26 Parenterally administered penicillin is responsible for most of the anaphylactic reactions. The extensive use of this drug in unsuspected sources such as foods, in which it is used as a bacteriostatic agent, may make it difficult to ascertain historically that penicillin is not the causative agent. Cephalosporins share the β-lactam ring structure and side chains of the penicillins and have been incriminated for allergic cross-sensitivity in 1 to 8% of patients.27 It is unclear which epitope is responsible for the cross-reactivity. Patients who have urticaria or anaphylactic reactions after taking penicillin are approximately four times more likely to have an adverse reaction to cephalosporins. Even in this setting, the risk of an anaphylactic reaction to cephalosporins is still less than 0.1%.
1517
Table 117-1 Clinical Manifestations of Anaphylaxis and Related Pathophysiology REACTION
SYMPTOMS
SIGNS
PATHOPHYSIOLOGY
Respiratory tract Upper
Rhinitis
Nasal congestion Nasal itching Sneezing
Nasal mucosal edema Rhinorrhea
Laryngeal edema
Dyspnea Hoarseness Throat tightness Hypersalivation Cough Wheezing Retrosternal tightness Dyspnea
Laryngeal stridor Supraglottic and glottic edema
Increased vascular permeability Vasodilation Stimulation of nerve endings As above, plus increased exocrine gland secretions
Lower
Cardiovascular system
Bronchospasm
Circulatory collapse
Light-headedness Generalized weakness Syncope Ischemic chest pain
Dysrhythmias
As above, plus palpitations
Cardiac arrest
Skin
Urticaria
Angioedema Eye
Conjunctivitis
Gastrointestinal tract
Miscellaneous central nervous system Hematologic
Genitourinary
ECG, electrocardiographic.
Fibrinolysis and disseminated intravascular coagulation
Pruritus Tingling and warmth Flushing Hives Nonpruritic extremity, periorbital and perioral swelling Ocular itching Increased larcrimation Red eye Dysphagia Cramping, abdominal pain Nausea and vomiting Diarrhea (rarely bloody) Tenesmus Apprehension Sense of impending doom Headache Confusion Abnormal bleeding and bruising
Pelvic pain Vaginal bleeding Urinary incontinence
Cough Wheeze, rhonchi Tachypnea Respiratory distress Cyanosis Tachycardia Hypotension Shock ECG changes: Tachycardia Nonspecific and ischemic ST-T wave changes Right ventricular strain Premature atrial and ventricular contractions Nodal rhythm Atrial fibrillation Pulseless ECG changes: Ventricular fibrillation Asystole Urticaria Diffuse erythema Nonpitting edema, frequently asymmetric Conjunctival inflammation
As above, plus bronchiole smooth muscle contraction
Increased vascular permeability Vasodilation Loss of vasomotor tone Increased venous capacitance Decreased cardiac output Decreased mediator-induced myocardial suppression Decreased effective plasma volume Decreased preload Decreased afterload Hypoxia and ischemia Dysrhythmias Iatrogenic effects of drugs used in treatment Preexisting heart disease
Increased vascular permeability Vasodilation Increased vascular permeability Stimulation of nerve endings
Nonspecific
Increased mucus secretions Gastrointestinal smooth muscle contraction
Anxiety Seizures (rarely) Coma (late)
Secondary to cerebral hypoxia and hypoperfusion Vasodilation
Mucous membrane bleeding, disseminated intravascular coagulation Increased uterine tone Vaginal bleeding Urinary incontinence
Mediator recruitment and activation Uterine smooth muscle contraction Bladder smooth muscle contraction
Chapter 117 / Allergy, Hypersensitivity, and Anaphylaxis
ORGAN SYSTEM
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1518
Etiologic Agents and Pathogenetic Mechanisms BOX 117-3 of Anaphylaxis IgE-Mediated Agents Foods (eggs, peanuts, tree nuts, cow’s milk, fruits, shellfish, shrimp, other crustaceans, many others) Antibiotics (penicillins, cephalosporins, sulfonamides, nitrofurantoin, tetracycline, streptomycin) Other therapeutics (methylparaben, human diploid cell rabies vaccine, egg-based vaccines: measles, mumps, rubella, vasopressin, antilymphocyte globulin) Insect stings (Hymenoptera venoms, fire ant stings) Latex Allergens used in immunotherapy Heterologous and human sera Hormones (insulin, methylprednisolone, parathormone, estradiol, progesterone, corticotropin) Enzymes (trypsin, streptokinase, chymotrypsin, chymopapain, l-asparaginase) Polysaccharides (dextran, iron dextran) Local anesthetics (mostly ester family, procaine, tetracaine, benzocaine) Direct Mast Cell Degranulation Radiocontrast media Opiates Curare, d-tubocurarine Protamine Polysaccharides (some are IgE mediated) ACE inhibitor used during hemodialysis with certain HD membranes Ethylene oxide gas on dialysis tubing Immune Complex Mediated Whole blood (transfusion reaction to formed elements) Immunoglobulin administration Arachidonic Acid Metabolism Aspirin and NSAIDs (presumed) Benzoates (presumed) Food colorants (tartrazine [possibly]) Physical Factors Exercise Temperature (heat or cold) Idiopathic Factitious Undifferentiated somatoform idiopathic anaphylaxis Idiopathic ACE, angiotensin-converting enzyme; HD, hemodialysis; IgE, immunoglobulin E; NSAID, nonsteroidal anti-inflammatory drug. Modified from Kemp SF, Lockey RF: Anaphylaxis: A review of causes and mechanisms. J Allergy Clin Immunol 110:341, 2002.
Although it may be prudent to administer a class of antibiotics other than the cephalosporins when a well-documented significant history of penicillin allergy is obtained, if no other antibiotic choices are available, the first dose of cephalosporin can be administered orally under medical supervision and observation.28
Latex Allergy to natural rubber latex in gloves and other medical products has become a health issue for health care providers,
patients, and rubber industry workers.29 Latex is derived from the commercial rubber tree Hevea brasiliensis, which is native to the southern Amazon and harvested commercially from plantations in Southeast Asia and Africa. The functional unit is a rubber particle coated with a layer of proteins, lipids, and phospholipids to provide structural integrity. Latex allergy refers to sensitivity to either the proteins or the chemical products contained in the latex products. The sensitivity reaction can be delayed (type IV) contact dermatitis or an immediate hypersensitivity (type I) reaction (asthma, urticaria, and anaphylaxis). Although latex allergy is common, affecting 0.7 to 1.1% of the population,30 in the United States, only 220 people per year are estimated to be at risk for an anaphylactic reaction from latex.5 The most common symptoms of latex allergy include allergic urticaria, rhinitis, conjunctivitis, and occupational asthma. There is evidence that patients with specific food allergies are predisposed to latex allergy.31 The true risks of latex-induced anaphylaxis are not known, but in a review of anaphylactic incidents in 50 children, 27% were due to latex allergy.32 Diagnostic tools include serologic assays and skin prick testing. There is no effective prophylaxis for latex allergy. In fact, routine use of the H2 blocker ranitidine for gastroesophageal reflux was reported to increase the risk of a heart conduction block in a case of anaphylaxis caused by latex.33 Avoidance of latex-containing products is the recommended approach.
Insect Stings Hymenoptera venoms and fire ant stings are responsible for significant anaphylactic morbidity and mortality.34 The first recorded fatality from anaphylaxis was probably the hieroglyphic-documented death of King Menes of Egypt in 2641 bce, when he succumbed to the sting of a wasp or hornet. Stinging Hymenoptera insects affect up to 13.6 million Americans annually (circa 1999), accounting for approxi mately 50 to 100 deaths annually.5 Allergic sensitization to Hymenoptera has been reported in 0.4 to 4% of the general population. The principal offenders (in decreasing order of frequency) are yellow jackets, honeybees, wasps, and yellow and bald-faced hornets. The imported fire ant has become a significant pest responsible for anaphylaxis, spreading from the Atlantic and Gulf coasts inland.35 The introduction of killer (Africanized) bees in Brazil and their subsequent northern migration make them a significant cause of sting-induced anaphylaxis in areas of Texas, Arizona, and the southwestern United States.36 The Hymenoptera venoms are complex mixtures of pharmacologically and biochemically active substances. Honeybee venom has been subjected to the greatest amount of research and contains two major enzymes—hyaluronidase and phospholipase A (PLA)—and other peptides, including a mast celldegranulating peptide. Yellow jacket venom contains not only phospholipases A and B and hyaluronidase but also kinins. Hornet venom has, in addition, acetylcholine. Wasp venom has not been extensively studied. Fire ant venom is mostly a nonproteinaceous alkaloid suspension containing PLA and hyaluronidase.
Other Therapeutic Agents Heterologous sera that were used in diphtheria and tetanus equine antitoxins in the past can act as whole antigenic markers. In fact, until the advent of penicillin, these two therapeutic agents were the most common iatrogenic causes of
Immune Complex–Mediated Agents Anaphylactic-type reactions are uncommon complications after the administration of whole blood and immunoglobulins. The fixation of antibodies to formed elements such as red blood cells, platelets, and leukocytes and soluble components activates the complement system. This is particularly relevant in IgA-deficient patients exposed to multiple transfusions, who may have produced antibodies to IgA present in previous transfusions. With subsequent transfusions, an antigen (IgA)–
1519
anti-IgA antibody (IgG) immune complex forms, and subsequent activation of the complement cascade may occur.
Nonimmunologic Activators Many of the opioid analgesics can cause anaphylactic reaction through a direct histamine release mechanism, although there is evidence that, rarely, some reactions are true IgE-mediated reactions. It is unclear how much cross-sensitivity is present among these agents.42 Radiocontrast media (RCM) represents one important class of drugs that can cause an anaphylactic reaction.10,11 RCM is available as high-osmolar RCM (HRCM, >1400 mOsm/kg), iso-osmolar RCM (same as serum, 290 mOsm/kg), and lowosmolar RCM (LRCM, 500–900 mOsm/kg). In addition, depending on the charge of the iodine molecule used, RCM is classified as ionic or nonionic. The majority of RCM is ionic HRCM or nonionic LRCM. Approximately 10 million radiologic studies using RCM are performed in the United States annually, most of which use nonionic LRCM. Of these studies, approximately 35 per 100,000 exposed patients have a serious reaction, with an estimated fatality rate of 1 to 3.9 per 1 million injections.9 The older hyperosmolar agents can cause a reaction in up to 5.6% of patients and fatalities in up to 0.01%.43 Subsequent studies put the risk of a serious reaction to high-osmolar contrast media at 0.22%, with 11.7 fatalities per 1 million injections.44 Protocols have been developed to minimize risks of a serious allergic reaction in patients who have had a previous adverse reaction to RCM but who still require additional radiographic studies with contrast. One common protocol is shown in Box 117-4. Anaphylactic reactions to RCM are largely idiosyncratic, occur within minutes of infusion, and are independent of the dose. Risk factors for an anaphylactic reaction include a previous adverse reaction to RCM, a history of atopy or allergic disease, asthma, and certain medications. Of note, it is a misconception that a history of allergy to fish or shellfish is a contraindication to the use of RCM or increases the risk of an adverse reaction to RCM.45 The pathophysiology of anaphylactic reactions to RCM is unknown, but it is believed to be nonimmunologic. Suggested mechanisms include direct histamine release, alternative complement pathway activation, and activation of the contact system.
Modulators of Arachidonic Acid Metabolism Interruption of AA metabolism by aspirin (ASA) and other nonsteroidal anti-inflammatory drugs (NSAIDs) has been
A Standard Treatment Protocol for
BOX 117-4 Patients with a History of
Radiocontrast-Induced Anaphylaxis
Prednisone, 50 mg by mouth given 13, 7, and 1 hour before the procedure. Diphenhydramine, 50 mg intramuscularly given 1 hour before the procedure. Consider ephedrine, 25 mg by mouth given 1 hour before the procedure. Consider an H2 antagonist, such as ranitidine, 300 mg by mouth given 3 hours before the procedure. From Lieberman P: Anaphylaxis. Med Clin North Am 90:77, 2006
Chapter 117 / Allergy, Hypersensitivity, and Anaphylaxis
anaphylaxis in humans. The use of human antisera was associated with a marked reduction in the incidence of seruminduced anaphylactic reactions. In fact, no adverse reactions were noted on repeated immunization using human tetanus antisera in approximately 250 patients who had previous anaphylactic reactions to equine tetanus antisera.37 Equine antisera are still used in the administration of antilymphocyte serum and in the management of venomous snake bites. Although anaphylactic reactions are rare (1 : 500,000), the equine antiserum should still be diluted and pretested. Since the development of heterologous insulin hormone therapy for the management of diabetes mellitus, local and systemic allergic complications have been recognized. A large percentage of local reactions were eliminated with the introduction of purified, single-peak pork insulin. With the introduction of Humulin, an insulin preparation prepared from recombinant DNA, the incidence of anaphylaxis and insulin resistance has declined dramatically. Allergen extracts are used diagnostically in skin testing and therapeutically in immunotherapy (also known as hyposensitization or desensitization).38 Exposure to therapeutic pollens, by injection or inhalation, can result in local allergic or systemic anaphylactic reactions. High-dose therapy, too frequent administration, or inadvertent intravascular injection increases the risk of anaphylaxis with immunotherapy. Although corticosteroids are used in the management of acute allergic syndromes and anaphylaxis, adverse reactions to these medications have been observed after parenteral administration.39 Skin testing may demonstrate the specific class of steroids responsible for hypersensitivity, and substitution of a different class should be considered. Local anesthetics occasionally produce adverse reactions.40 Most of these reactions are not allergic in nature but are related to a direct effect of the medication. True allergic reactions are rare and are most commonly seen with local anesthetics from the ester family (e.g., procaine, tetracaine, and benzocaine). Allergic reactions to local anesthetics belonging to the amide family (e.g., lidocaine, bupivacaine, mepivacaine, and dibucaine) are extremely rare if they occur at all. Multidose vials of lidocaine contain the preservative methylparaben, which belongs structurally to the ester family. This preservative has been implicated in allergic reactions in patients with a history of previous lidocaine hypersensitivity.41 Pure lidocaine (without the methylparaben preservative) should be used intravenously (as in Bier’s block). Anaphylactic reactions have occurred after the administration of egg embryo-grown vaccines, including the combined measles, mumps, and rubella (MMR), yellow fever, and influenza vaccines. A patient who is able to tolerate eggs orally, even if he or she has previously experienced anaphylaxis and shows a positive skin test to eggs, is likely to tolerate the vaccines. Anaphylactic reactions have also occurred against ethylene oxide (ETO), which is used to sterilize hemodialyzers. ETO can bind with human proteins such as human serum albumin (HSA), thus rendering the ETO-HSA complex allergenic.
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1520
postulated as the mechanism responsible for anaphylactoid reactions resulting from these agents, although AA modulation, anaphylatoxin generation, and direct histamine release may all be partially responsible.46 The incidence of anaphylactoid reactions to aspirin and NSAIDs varies widely, depending on the population (healthy, atopic, or those with nasal polyps). One study estimates the incidence as 2.1 anaphylactoid reactions per 100,000 exposed patients.9 Aspirin allergy refers to a clinical spectrum ranging from ASA GI intolerance to ASAinduced asthma exacerbation, which is a true ASA-anaphylactoid reaction. Desensitization protocols have been suggested for cardiovascular patients with a history of aspirin allergy.47 For ASA-induced cutaneous disease, an ASA desensitization protocol that can be used for cardiovascular patients in the ED is to administer ASA every 15 minutes, starting with 0.1 mg, up to 325 mg at 135 minutes.48 ASA should be avoided in patients with true anaphylactic reaction to ASA. For noncardiovascular applications, most aspirin-sensitive patients can tolerate sodium salicylate or acetaminophen as an aspirin substitute. Of note, one of many food additives, tartrazine (foods, drugs, and cosmetics [FD&C] yellow dye number 5), is a stable azo coloring agent present in thousands of foods and drugs in the United States.49 The exact mechanism of tartrazine sensitivity is unknown, although modulation of AA metabolism and several other theories have been proposed.
Physically Induced Anaphylaxis Thermomechanical and physical factors (heat and cold), especially exercise, have increasingly been recognized as etiologic agents in certain anaphylactic-like incidents.50 The mechanism is unclear, but release of mast cell and basophil mediators has been implicated. Patients with exercise-induced anaphylaxis are generally dedicated athletes who may have a personal or family atopic history. Exercise-induced anaphylaxis has been demonstrated in some cases to depend on previous ingestion of food to which the patient may be subclinically sensitive. Provocative foods, if identified, should be avoided. Patients should discontinue the exercise when they experience pruritus. When exercise is continued beyond this point, clinical deterioration is likely in susceptible individuals. Prophylactic treatment with an antihistamine as a single agent or in combination with other agents may be helpful. Avoidance of precipitating factors, modification of exercise, and use of a self-injectable epinephrine kit have been recommended for patients with exercise-induced anaphylaxis.
Idiopathic Anaphylaxis Prednisone-responsive IA refers to an anaphylactic condition in which no eliciting factors can be detected.11,51 In the United States, it is estimated that 20,000 to 47,000 patients annually see allergists for signs and symptoms of IA. A specific causative agent cannot be found historically. Laboratory studies including a complete blood count with differential leukocyte count, erythrocyte sedimentation rate, blood chemistries, complement levels, C1 esterase inhibitor levels, serum and urinary histamine levels, urinalysis, skin testing, and occasionally more specialized tests when clinically indicated are all nondiagnostic. Food diaries and efforts to find systemic disease are often fruitless. Although IA may be life-threatening, it is usually responsive to conventional therapies, including antihistamines, sympathomimetics, and especially prednisone.51 Some cases of IA may be caused by kissing or conversion disorders.52 The overall prognosis for IA is good, but certain patients may experience recurrent IA despite intensive prophylactic administration of antihistamines, sympathomimetics, or steroids.
Sometimes, IA can represent “progesterone” anaphylaxis. Women suffering from this disorder may present with recurrent episodes of anaphylaxis that are temporally related to the menstrual cycle. Other patients may have anaphylactic reactions to injection of medroxyprogesterone or luteinizing hormone-releasing hormone.
■ CLINICAL FEATURES Anaphylaxis in humans primarily affects organs that are rich in mast cells—the cutaneous, upper and lower respiratory, cardiovascular, neurologic, and gastrointestinal systems. Anaphylactic reactions range from mild to fatal with variable durations of attack. The clinical expression depends on the degree of hypersensitivity; the quantity, route, and rate of antigen exposure; the pattern of mediator release; and the target organ sensitivity and responsiveness. One cardinal feature of a serious anaphylactic reaction is its rapid onset. Most anaphylactic reactions become clinically evident within minutes (average 5 to 30 minutes) after a parenteral exposure; onset can be slower (average 2 hours) after ingestion. Most fatalities occur within the first 30 minutes after antigenic exposure. Symptoms can sometimes resolve and recur hours later in what has been termed biphasic anaphylaxis.53 In general, the sooner the clinical syndrome is manifest after antigenic exposure, the more severe the reaction. Anaphylactic reactions after parenteral antigenic exposure are usually more immediate in onset, progress more rapidly, and are more severe in quality than those occurring after topical or oral exposures. Fatal cases of anaphylaxis are usually caused by cardiovascular collapse or respiratory failure. Although urticaria and angioedema are the most common presenting symptoms in anaphylaxis (88%), fatal anaphylaxis syndromes with laryngeal edema and circulatory collapse can occur even in the absence of any premonitory warning symptoms or signs or cutaneous manifestations. The first clinical manifestation of anaphylaxis usually involves the skin; the patient experiences generalized warmth and tingling of the face, mouth, upper chest, palms, soles, or the site of antigenic exposure. Pruritus is a nearly universal feature and may be accompanied by generalized flushing and urticaria. Patients presenting with angioedema may complain of swelling and a sensation of burning under the skin but no itchy rash. This may be followed by mild to severe respiratory distress. The patient may describe a cough; a sense of chest tightness, dyspnea, and wheeze from bronchospasm; or throat tightness, dyspnea, odynophagia, or hoarseness associated with laryngeal edema or oropharyngeal angioedema. Hypotension or dysrhythmias may manifest as light-headedness or syncope. Seizure activity caused by decreased cerebral perfusion may rarely be observed. Any of these clinical patterns may occur independently or in association with nasal congestion and sneezing; ocular itching and tearing; cramping abdominal pain with nausea, vomiting, diarrhea, and tenesmus; incontinence; pelvic pain and uterine cramping; headache; or a sense of impending doom. The physical examination may reveal tachypnea, tachycardia, and hypotension. Laryngeal stridor, hypersalivation, hoarseness, and angioedema indicate upper airway obstruction, whereas coughing, wheezing, rhonchi, and diminished air flow suggest lower respiratory tract bronchoconstriction. Tachycardia and hypotension suggest cardiac insufficiency. Commonly observed dysrhythmias include sinus tachycardia, premature atrial and ventricular contractions, nodal rhythm, and atrial fibrillation. Other electrocardiographic changes include nonspecific and ischemic ST-T wave changes, right ventricular strain, and intraventricular conduction defects. The patient may have a depressed level of consciousness due
■ DIAGNOSTIC STRATEGIES A good history and physical examination, coupled with a high index of suspicion, are the best diagnostic tools in approaching patients with possible anaphylaxis. A new set of clinical parameters has been advanced to aid in the clinical diagnosis of an anaphylactic reaction. The diagnosis of anaphylaxis is considered to be “highly likely” when patients present with cutaneous symptoms (itchy urticaria, flushing, and swollen lips, tongue, and throat) and either respiratory difficulty (dyspnea, wheezing, and stridor) or reduced blood pressure or symptoms of end-organ dysfunction (see Box 117-1). Anaphylaxis can be confirmed by testing for allergen-specific IgE and serum tryptase—tests that are not normally available in the ED. As clinically indicated, other diagnostic modalities may be employed concurrently to rule out other emergencies. The initial screening studies can include a complete blood count, complete metabolic panel (hypoglycemia), coagulation panel (prothrombin time, partial thromboplastin time, and international normalized ratio), cardiac enzymes, an electrocardiogram to rule out an acute coronary syndrome, urine analysis, erythrocyte sedimentation rate, and a portable chest radiograph. Serum level of serotonin and urinary 5-hydroxyindole acetic acid, catecholamines, and vanillylmandelic acid are useful to rule out carcinoid syndrome. Serum and urinary histamine and serum tryptase levels are helpful to confirm the diagnosis of anaphylaxis after the fact. The optimal time to obtain the serum histamine level is within 1 hour and serum tryptase within 1 or 2 hours (but no longer than 6 hours) of the onset of symptoms. Samples of vomitus may be collected for the allergist to create a custom radioallergosorbent test panel for later desensitization therapy. Serial arterial blood gases may help monitor clinical response. Blood culture, urine culture, computed tomography of the head and lateral soft tissue of the neck, and indirect and direct laryngoscopy can be considered depending on clinical suspicion.
1521
exclude epiglottitis and supraglottitis, retropharyngeal or peritonsillar abscess, laryngeal spasm, foreign body aspiration, or tumor.
Bronchospasm Obstructive lung diseases such as acute asthma and status asthmaticus are usually not associated with other symptoms and signs of anaphylaxis. Patients with acute pulmonary embolism may present with shock, respiratory distress, and bronchospasm. Exercise-induced anaphylaxis should be differentiated from exercise-induced asthma because the former is usually accompanied by pruritus and other systemic manifestations.
Syncope Vasovagal syncope is the most common differential diagnosis in the patient arriving with collapse as a result of parenteral administration of an antigen. Classically, the patient has bradycardia, hypotension, and pallor as opposed to the tachycardia, hypotension, and diaphoresis usually associated with anaphylaxis. The absence of any other clinical manifestations of anaphylaxis, along with history of stress, pain, and previous episodes of simple faints, helps point toward the diagnosis of vasovagal syncope. Other causes of syncope, such as seizure, stroke, hypoglycemia, acute coronary syndrome, or cardiac dysrhythmia, also need to be considered. Ordinary allergic reactions and especially anaphylaxis can precipitate an acute coronary syndrome.54
Shock Clinically, anaphylactic, septic, and spinal shock may appear similarly with signs and symptoms of shock including endorgan hypoperfusion and vasodilation. Skin is usually moist and warm, suggesting a state of decreased peripheral vascular resistance. Cardiogenic, restrictive, hypovolemic, or hemorrhagic shock would more likely be seen with cold, clammy skin, suggesting a state of heightened peripheral vascular resistance. Because anaphylactic shock can progress to cardiogenic shock, measurement of central venous pressures (CVPs) may be necessary.
■ DIFFERENTIAL CONSIDERATIONS
■ MANAGEMENT
The diagnosis of an anaphylactic reaction depends largely on recognizing the key symptoms and signs (see Box 117-1) that occur abruptly after an exposure to a suspected inciting agent. Considerations for other diseases with overlapping presentations should include those shown in Table 117-2.
Out-of-Hospital
Flush Syndromes and Rash Flushing disorders range from the benign, such as alcoholinduced flushing, to the more pathologic, such as mastocytosis or carcinoid syndrome. Scombroidosis refers to histamine poisoning from the ingestion of fish improperly stored at an elevated temperature. Histamine and cis-urocanic acid are produced by various bacteria that multiply in the spoiled fish. Patients usually present with a frightening flush but no urticaria, palpitations, syncope, nausea, vomiting, or diarrhea.
Stridor In the absence of oropharyngeal angioedema or other clinical manifestations of anaphylaxis, the diagnosis of laryngeal edema should be confirmed by direct or indirect laryngoscopy to
When a susceptible patient is reexposed to an antigen to which there has been a previous reaction, 50 mg of oral diphenhy dramine should be taken if available. At the first signs of any clinical manifestations of anaphylaxis, the patient should self-administer epinephrine if available (adult dose, 0.3 mL of 1 : 1000 intramuscular; pediatric dose, 0.01 mL/kg of 1 : 1000 intramuscular). Susceptible patients may even use aerosolized epinephrine from a metered-dose inhaler to counteract the effects of laryngeal edema, bronchoconstriction, and other manifestations of anaphylaxis.55 Multiple inhalations (e.g., 10–20 doses, resulting in the inhalation of 1.5–3 mg of epinephrine) produce therapeutic plasma levels, with the advantages of ease of administration, rapid absorption, and locally high epinephrine levels in the upper and lower airways. Epinephrine must be used with caution in elderly people and in those with a history of cardiac or hypertensive problems. Out-of-hospital personnel may be required to resuscitate a moribund patient using basic life support. Their first priority should be to establish and maintain ventilation, intravenous access, cardiac monitoring, and administration of supplemental
Chapter 117 / Allergy, Hypersensitivity, and Anaphylaxis
to hypotension; rarely, this may be caused by a postictal state due to seizure activity. Urticaria, angioedema, rhinitis, and conjunctivitis may be evident. A summary of the observed clinical manifestations of anaphylaxis along with their related pathophysiology is presented in Table 117-1.
1522
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
Table 117-2 Differential Diagnosis of Anaphylactic Reaction and Anaphylactic Shock FINDING
DISORDERS
COMMENTS
Flush syndromes/rash
Carcinoid syndrome Medullary carcinoma of the thyroid Vasointestinal peptide secreting tumor (VIPoma) Systemic mastocytosis Urticaria pigmentosa Pheochromocytoma Scombroidosis Occult infections Alcohol flush syndrome Sulfite and monosodium glutamate toxicity Leukemia (basophilic/acute promyelocytic) Hydatid cyst
Syncope/altered mental status
Vasovagal reaction Seizure/epilepsy Stroke Neuroleptic malignant syndrome (NMS) Serotonin syndrome Hypoglycemia Acute coronary syndrome Cardiac dysrhythmia
Stridor
Epiglottitis Supraglottitis Retropharyngeal Peritonsillar abscess Laryngeal spasm Foreign body aspiration Tumor Asthma/status asthmaticus Obstructive airway diseases Pulmonary embolism Spontaneous pneumothorax Cardiogenic Hemorrhagic/hypovolemic Septic Pulmonary embolism Urticarial vasculitis Hereditary angioedema Progesterone Redman’s syndrome Progesterone anaphylaxis Capillary leak syndrome Postmenopausal state
Urticaria and hypotension not typical in carcinoid syndrome. Serum serotonin and urinary 5-hydroxyindole acetic acid are elevated. There is episodic hypertension in pheochromocytoma. Check levels of metanephrine and urinary vanillylmandelic acid. Mastocytosis is a risk factor for anaphylaxis, and anaphylaxis may be the presenting manifestation of systemic mastocytosis. Scombroidosis is caused by histamine poisoning, which in turn is caused by ingestion of spoiled fish. The flushing is prominent and prolonged but no urticaria; tryptase level is low. Flush syndrome induced by alcohol is more common in Asian populations. Bradycardia, hypotension, nausea, diaphoresis, and pallor favor the diagnosis of a vasovagal reaction, whereas tachycardia, hypotension, and diaphoresis favor the diagnosis of anaphylaxis. Anaphylaxis can occur concurrently with or precipitate an acute coronary syndrome. Fever, encephalopathy, muscle rigidity, and hemodynamic instability characterize NMS. Fever, encephalopathy, hypertension, clonus, hyperreflexia, and other autonomic instability characterize serotonin syndrome. After stabilization and arrangement for good backup for airway support, investigations can be started to locate source of stridor.
Acute respiratory insufficiency Shock
Miscellaneous
oxygen to keep the oxygen saturation level greater than 90%. Local measures to decrease antigen absorption from an extremity include dependent positioning of the extremity, ice to vasoconstrict locally, and application of a loose tourniquet to obstruct the venous and lymphatic circulation. The tourniquet should be released for 1 of every 10 minutes. If an insect stinger remains, the wound should not be squeezed because it may inject more venom into the patient. The stinger should be removed gently with instruments, avoiding disturbance of the venom apparatus.
Emergency Department Since most of the morbidity and mortality associated with anaphylaxis originates from acute respiratory failure or cardio-
Exercise-induced asthma does not have the stigmata of exercise-induced anaphylaxis. Moist and warm skin suggests decreased peripheral vascular resistance; cold, clammy skin suggests increased peripheral vascular resistance. Pruritic urticaria is usually absent in hereditary and acquired angioedema. Rapid infusion of vancomycin occurs in Redman’s syndrome.
vascular collapse, the immediate goal in the ED is to stabilize any cardiorespiratory insufficiencies while confirming the diagnosis of anaphylaxis and anaphylactic shock along with other diagnostic alternatives. Box 117-5 summarizes the treatment options for acute anaphylaxis. Epinephrine and antihistamines (H1 and H2) should be administered early in most cases. Patients should have supplemental oxygen administered, large-bore (e.g., 16-gauge) intravenous lines inserted to infuse crystalloid or colloid solutions, and continuous cardiac monitoring. A large volume of crystalloid fluid may be required to reverse the hypotension associated with anaphylaxis. Upper airway obstruction from laryngeal edema or angioedema can progress rapidly. While preparing for more definitive airway management, a chin lift or jaw thrust may help obtain a patent airway. Suctioning the oropharynx of excess
1523
BOX 117-5 Treatment Options for Anaphylaxis (c) Aerosolized beta-agonists and others (i) Adult (1) Albuterol: 2.5 mg, diluted to 3 mL of NS, may be given continuously (2) Levalbuterol: 0.625–1.25 mg, diluted to 3 mL of NS, may be given continuously (3) Ipratropium: 0.5 mg in 3 mL of NS, repeat as necessary (ii) Pediatric (1) Albuterol: 2.5 mg, diluted to 3 mL of NS, may be given continuously (2) Levalbuterol: 0.31–0.625 mg, diluted to 3 mL of NS, may be given continuously (3) Ipratropium: 0.25 mg in 3 mL of NS, repeat as necessary (d) Methylprednisolone (i) Adult: 125–250 mg IV (ii) Pediatric: 1–2 mg/kg/IV 3. Special situations (a) Refractory hypotension (i) Glucagon: 1–5 mg IV over 5 min, followed by 5–15 µg/min continuous infusion (ii) Consider: (1) Discontinue epinephrine (2) Dopamine, 5–20 µg/kg/min continuous infusion and/or dobutamine 5–20 µg/kg/min continuous infusion (3) Norepinephrine: 8–12 µg/min (2–3 mL/min; 4 mg added to 1000 mL of D5W provides a concentration of 4 µg/mL) (b) Patients on beta-blockade (i) Glucagon: 1–5 mg IV over 5 min, followed by 5–15 µg/min continuous infusion (ii) Transcutaneous pacing for bradycardia (iii) Atropine for bradycardia (1) Adult: 0.3–0.5 mg IV/subcutaneous, to a maximum of 3 mg (2) Pediatric: 0.02 mg/kg IV/subcutaneous, to a maximum of 2 mg (iv) Isoproterenol: 0.05–0.2 µg/kg/min (1–2 mg in 500 mL of D5W, infused at a rate of 0.5–2 mL/min) (c) Refractory bronchospasm (i) Aminophylline: 5.6 mg/kg loading dose IV over 20 min, followed by 0.1–1.1 mg/kg/hr continuous infusion (d) Hypertensive crisis due to unopposed alpha-blockade (i) Nitroprusside: 0.3–10 µg/kg/min (6 µg/kg/min, neonates) continuous infusion (ii) Phentolamine: 5–20 mg IV (e) Dysrhythmia (i) Lidocaine 1–2 mg/kg IV bolus, followed by 2 mg/min continuous infusion
CXR, chest x-ray; D5W, 5% dextrose in water; ECG, electrocardiogram; LMA, laryngeal mask airway; NS, normal saline.
secretions may be necessary. A nasopharyngeal or oropharyngeal airway may aid in maintaining a patent airway at this stage. Racemic epinephrine, delivered as a 2.25% solution (0.5 mL placed in a nebulizer in 2.5 mL of normal saline), may be used as a temporizing measure. A laryngeal airway mask,
jet ventilation, or surgical airway may be needed for difficult airways. The success rate of intubation is improved when it is performed early and before soft tissue swelling progresses. Oral endotracheal intubation is the route of choice because
Chapter 117 / Allergy, Hypersensitivity, and Anaphylaxis
1. Immediate general interventions (a) Remove any triggering agent (b) Place patient in the Trendelenburg position if hypotensive (c) Assessment of airway patency, breathing, and circulation (i) Hyperextension of neck, jaw thrust, chin lift (ii) Administer supplemental oxygen by nasal cannula or non-rebreather mask (iii) Racemic epinephrine 0.5 mL of 2.25% in 2.5 mL of NS by nebulizer while awaiting definitive airway management (iv) Establish airway (1) Endotracheal intubation with or without RSI (rapid sequence intubation) (2) Adjunct airway technique (jet ventilation, LMA, surgical airway) as per local institution (v) Establish large-bore IVs (1) Administer colloid/crystalloid, titrate to blood pressure (vi) Pulse oxymetry (vii) Cardiac monitoring/ECG (viii) Portable CXR (ix) Blood draw (x) Place a loose tourniquet proximal to the reaction site; if reaction site on extremity, place extremity in dependent position (xi) Inject 0.1–0.2 mL 1 : 1000 epinephrine locally to the reaction site 2. Specific measures (a) Epinephrine (i) Intramuscular (subcutaneous route acceptable) 1 : 1000 (1) Adult: 0.3–0.5 mL every 5 min as necessary, titrated to effects (2) Pediatric: 0.01 mL/kg, every 5 min as necessary, titrated to effects (3) Alternatively, epinephrine (EpiPen) (0.3 mL) or EpiPen Jr (0.15 mL) can be administered into anterolateral thigh. Removal of clothing is unnecessary (ii) Intravenous 1 : 100,000 (0.1 mL of 1 : 1000 in 10 mL of NS) (1) Continuous hemodynamic monitoring required (2) 10 mL of 1 : 100,000 over 10 min, titrated to effects, repeat as necessary (b) Antihistamines (i) Diphenhydramine: intravenous (or oral) (1) Adult: 50 mg, up to 400 mg/24 hr, titrated to effects (2) Pediatric: 1 mg/kg, up to 300 mg/24 hr, titrated to effects (ii) Ranitidine: intravenous (or oral) (1) Adult: 50 mg IV (150 mg oral) (2) Pediatric: 1 mg/kg IV or oral
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1524
significant anatomic distortion may be present as a result of edema. Sedation and paralysis should be used with caution because a distorted airway may preclude intubation after paralysis. Patients in acute respiratory distress should be given definitive airway treatment without waiting for results of an arterial blood gas. Once a patent airway has been obtained and supplemental oxygen delivered, therapy should focus on relieving the patient’s bronchospasm.
Continuous cardiac monitoring should be done at all times. If a percutaneous intravenous line cannot be established, alternative routes are available. In addition to the subcutaneous and intramuscular routes of administration, intraosseous or sublingual injection or endotracheal nebulization should be considered. The dosage and concentration guidelines for these routes of administration of epinephrine are the same as those for intravenous administration.
Epinephrine
Antihistamines
Drugs used in the treatment of anaphylaxis either inhibit the release of chemical mediators or reverse the effects of mediators on target tissues. Epinephrine, with its combined alphaand beta-adrenergic agonist actions, is the first drug of choice in the treatment of anaphylaxis.56 The alpha-agonist effects of epinephrine increase peripheral vascular resistance and reverse peripheral vasodilation, vascular permeability, and systemic hypotension. The beta-agonist effects of epinephrine produce bronchodilation, cause positive inotropic and chronotropic cardiac activity, and result in increased production of intracellular cyclic adenosine monophosphate (cAMP). Epinephrine therefore reverses bronchospasm, stimulates increased cardiac output, and inhibits further mediator release. The alpha- and beta-agonist actions of epinephrine can also be potentially dangerous. Excessive alpha-agonist activity can result in a hypertensive crisis. Excessive beta-agonist activity can increase myocardial oxygen consumption through increased wall tension, contractility, and chronotropism and can result in myocardial ischemia or infarction. Increased automaticity and chronotropism can produce hemodynamically significant supraventricular and ventricular tachydysrhythmia. Epinephrine should be used with caution in elders and those with known coronary artery disease and should be avoided in patients with life-threatening tachydysrhythmias. The route of epinephrine administration depends on the severity of the clinical presentation. Subcutaneous epinephrine is usually effective in situations in which the clinical manifestations are mild and the patient is normotensive. In the patient with diffuse, generalized urticaria, subcutaneous absorption of epinephrine may be slow and unpredictable and the intramuscular route may be more efficacious. For subcutaneous and intramuscular injections, the initial dose of epinephrine is 0.01 mL/kg of a 1 : 1000 solution to a maximum of 0.5 mL of 1 : 1000 solution (0.5 mg). A fraction of the total dose (0.1 or 0.2 mL) should be administered at the site of antigenic exposure if accessible (e.g., a bee sting or antigen injection in an extremity). If the patient demonstrates severe upper airway obstruction, acute respiratory failure, or shock (systolic blood pressure 0.5 cm in diameter and depth Nodule filled with expressible material Blisters 0.5 cm in diameter filled with clear fluid Vesicle filled with cloudy or purulent fluid Liquid debris that has dried on the skin surface; usually moist and yellowish brown Visibly thickened stratum corneum; usually white Epidermal thickening characterized by visible and palpable skin thickening and accentuated skin markings Dermal thickening that feels thick and firm Papule or plaque of dermal edema; often with central pallor and irregular borders Red appearance of skin caused by vasodilation of dermal blood vessels; blanchable Red appearance of skin caused by blood extravasated from disrupted dermal blood vessels; nonblanchable Flat, nonpalpable Elevated, palpable
Papule Plaque Nodule Cyst Vesicle Bullae Pustule Crust Scale Lichenification Induration Wheal Erythema Purpura Macular purpura Papular purpura
Modified from Lookingbill DP, Marks JG: Principles of Dermatology, 3rd ed. Philadelphia, Saunders, 1993.
evaluation. Higher dosages may be needed. Alternative therapy includes fluconazole 200 mg/day (adults) or 3 to 5 mg/kg/day (children), itraconazole 200 mg daily (adults) and 3 to 5 mg/kg/ day (children) for 4 to 6 weeks, oral terbinafine at 3 to 6 mg/ kg/day for 4 to 6 weeks, or terbinafine cream once a day for 8 weeks.5,9,13,14 Selenium sulfide shampoo 250 mg twice weekly decreases shedding of spores.9 Family members should be evaluated.
Kerion A kerion is a dermatophytic infection, usually of the scalp, that appears as an indurated, boggy inflammatory plaque studded with pustules.3 It is commonly confused with bacterial infections. Kerions should be treated as tinea capitis, with the addition of prednisone 1 mg/kg/day for 1 or 2 weeks to help decrease the inflammatory reaction and subsequent scarring.6,15,16 If bacterial superinfection exists, oral cephalexin or dicloxacillin can be added for the first week of treatment.9
Tinea Corporis Clinical Features. Tinea corporis is the classic “ringworm” infection. It affects the arms, legs, and trunk and is classically a
Figure 118-1. Tinea corporis. (Courtesy of David Effron, MD.)
sharply marginated, annular lesion with raised or vesicular margins and central clearing (Fig. 118-1). Lesions may be single or multiple, the latter occasionally being concentric. Tinea cruris, which involves the groin, is similar in appearance and may also include the perineum, thighs, and buttocks, but the scrotum is characteristically spared. Differential Considerations. The differential diagnosis of tinea cruris includes granuloma annular psoriasis, intertrigo with secondary candidiasis, and erythrasma.17 Management. Infections of the body, groin, and extremities usually respond to topical measures alone.17 A number of effective topical antifungal agents are available, including clotrimazole (Lotrimin), haloprogin (Halotex), miconazole (Micatin), tolnaftate (Tinactin), terbinafine, naftifine, and griseofulvin 1%. Two or three daily applications of the cream form of any of these preparations result in healing of most superficial lesions in 1 to 3 weeks.5,9,18-20 Acute inflammatory lesions displaying oozing or blisters should be treated additionally (four times a day) with open, wet compresses of Burow’s solution—an aluminum acetate solution that is useful as a soothing wet dressing for inflammatory skin conditions. There is often involvement of the feet and toenails.5
Tinea Pedis Tinea pedis, or athlete’s foot, appears with scaling, maceration, vesiculation, and fissuring between the toes and on the plantar surface of the foot. In extensive cases, the entire sole may be involved. A secondary bacterial infection may occur. The vesicular pustular form of tinea pedis should be considered when vesicles and pustules on the instep are noted. The differential diagnosis includes contact dermatitis and dyshidrotic eczema. A KOH preparation should help differentiate between these processes. Treatment is similar to that of tinea corporis.21
1531
Candidiasis Perspective Infection by Candida albicans can occur in infancy and old age; in people with acquired immunodeficiency syndrome (AIDS), pregnancy, obesity, malnutrition, diabetes and other endocrine imbalances, and malignancy; and those with other debilitating illnesses. Patients treated with corticosteroids, immunosuppressive agents, and antibiotics are also prone to cutaneous fungal infections.
Oral Thrush
Figure 118-2. Tinea versicolor. (Courtesy of David Effron, MD.)
Tinea Versicolor Clinical Features. Tinea versicolor is a superficial yeast infection caused by Pityrosporum ovale.22 Superficial scaling patches occur mainly on the chest and trunk but may extend to the head and limbs. As the name implies, lesions can be a variety of colors, including pink, tan, or white.3 The disease may be associated with pruritus, but medical care is often sought because the spots do not tan. On physical examination, a fine subtle scale is noted that may appear hypopigmented (Fig. 118-2). Pale yellow or orange fluorescence under Wood’s light is sometimes present. The differential diagnosis includes vitiligo and seborrheic dermatitis. A KOH preparation reveals short hyphae mixed with spores (“chopped spaghetti and meatballs”). Management. Tinea versicolor is treated with 2.5% selenium sulfide shampoo, imidazole creams, or oral ketoconazole as a single 400-mg dose or 200 mg daily for 3 to 5 days.5,22-24 Recurrence rates vary from 15 to 50%, and recurrence is considered the rule rather than the exception.22 Monthly prophylaxis with propylene glycol and water, selenium shampoo, or azole creams can help prevent recurrences.9,22 Pigmentation may not return to normal for months.
Tinea Unguium Clinical Features. Tinea unguium results in nails that are opaque, thickened, cracked, and crumbled. Subungual debris is present, and the nail may contain yellowish longitudinal streaks. The nail of the great toe is most commonly involved. Involvement of all of the nails of the hands and feet is rare. Management. Topical therapy of the nails alone rarely results in a cure because penetration into the nail keratin is poor. Fingernails typically respond more rapidly to therapy than toenails. Oral griseofulvin and ketoconazole require prolonged courses with high relapse rates and numerous side effects.24
Clinical Features. Oral thrush is the most common clinical expression of Candida infection.25 Thrush is most common in newborns, with one third being affected by the first week of life. It appears as patches of white or gray friable material covering an erythematous base on the buccal mucosa, gingiva, tongue, palate, or tonsils. Fissures or crust at the corners of the mouth may be present. The differential diagnosis of oral thrush includes lichen planus, which is not easily scraped off like C. albicans. Oral mucous membrane infection with C. albicans is an AIDS-defining illness.9 If the patient does not use dentures and has not taken antibiotics, underlying immunosuppression should be considered. Management. Treatment of oral thrush involves painting the mouth with 1 mL of oral nystatin suspension (100,000 U/mL) four times a day for infants or 4 to 6 mL four times a day swish and swallow for older children and adults. Treatment should be continued for 5 to 7 days after the lesions disappear. Clotrimazole troches dissolved in the mouth two to five times daily is a preferable treatment option for adults.3 If topical therapy is not effective or in cases of chronic candidosis, oral ketoconazole, itraconazole, or fluconazole may be prescribed.25 Patients with oral candidiasis because of dentures should soak their dentures overnight in dilute (1 : 10) sodium hypochlorite solution.3
Cutaneous Candidiasis Clinical Features. Cutaneous candidiasis favors the moisture and maceration of the intertriginous areas—the interdigital web spaces, groin, axilla, and intergluteal and inframammary folds. Lesions appear as moist, bright red macules rimmed with a collarette of scale, which represents the pustule roof with scalloped borders. Small satellite papules or pustules are just peripheral to the main body of the rash. These satellite lesions are the most typical indicators of a Candida infection. Intertriginous lesions are prone to bacterial superinfection. Candidal onychia and paronychia are occupational conditions in those whose hands are frequently immersed in hot water. These infections also occur with thumb sucking by children who have thrush. The paronychial area becomes red and swollen and the nails thick and brittle, with transverse ridging. Destruction of the nail plate may occur. Differential Considerations and Diagnostic Strategies. The differential diagnosis of cutaneous candidiasis includes contact dermatitis,
Chapter 118 / Dermatologic Presentations
Newer agents such as itraconazole, fluconazole, and terbinafine are safer and more effective. They also offer shorter treatment periods, thus improving compliance.24 The infection may be resistant to this regimen as well, however, and surgical removal of the nail is occasionally required.17 Recurrence is common.
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1532
tinea cruris, intertrigo, malaria, or folliculitis. Candidiasis, however, is less sharply demarcated than tinea cruris and brighter red than intertrigo. A KOH preparation taken from a pustule and roof of the lesion will reveal hyphae and pseudohyphae. Management. Treatment of intertriginous lesions requires the removal of excessive moisture and maceration. Lesions should be exposed to circulating air from a fan several times a day. Inflammatory lesions should be soaked in or covered with compresses of cool water or Burow’s solution. Topical imidazole creams, such as clotrimazole and miconazole, should be applied sparingly to affected areas. Prescription creams, such as econazole, ketoconazole, or sulconazole, are also effective. Protecting the hands from water is an integral part of the treatment of candidal paronychia. Prolonged immersion should be avoided and contact with water prevented by gloves with cotton liners. Nystatin or clotrimazole cream should be applied frequently to the nail folds for 6 to 8 weeks. A search for underlying immunocompromise should begin in patients with chronic, recurrent candidiasis.
Diaper Dermatitis Clinical Features Diaper dermatitis is a common disorder that is exacerbated by heat, moisture, friction, and the presence of urine and fecal material. Occlusive clothing in infants tends to foster all of these. Lesions begin as erythematous plaques in the genital, perianal, gluteal, and inguinal areas. More severe involvement results in moist, eroded lesions that may extend beyond the primary areas of appearance. Infection with C. albicans and fecal bacterial flora is an important contributory factor to the development of diaper dermatitis. Lesions infected with Candida are moist, red patches with well-demarcated borders. Papular or pustular satellite lesions are also present. Diaper dermatitis may reflect the presence of atopic or seborrheic dermatitis in the infant. The presence of lesions elsewhere on the body—particularly on the face in cases of atopic dermatitis or the scalp in cases of seborrhea—alerts the physician to these possibilities. Ammonia and bacterially produced putrefactive enzymes produce dermatitis as contact irritants. Such rashes are accompanied by characteristic odors. The existence of diaper dermatitis as a true allergic contact dermatitis is rare.
Management Treatment consists primarily of altering the physical environment in which diaper dermatitis thrives. Excess clothing should be removed, and occlusive plastic or rubber diaper covers should not be used. Diapers should be changed frequently and left off for prolonged periods if possible. Sterilized cloth diapers are preferred. If exudative lesions are present, treatment with topical cool wet compresses of saline or Burow’s solution is indicated for 2 or 3 days. Continuous air exposure of the area should be attempted.26 Zinc oxide (Desitin) may dry the area. Severe contact or seborrheic dermatitis may require short-term treatment with topical corticosteroids, such as 1% hydrocortisone in a cream base.26 Ointment-based topical medications for treatment of diaper dermatitis should be avoided because their occlusive nature enhances moisture retention. Nystatin cream or powder should be applied to lesions infected with Candida.
■ SCALY PAPULES Fungal lesions are typically scaly, as are lesions of secondary syphilis. Additional scaly diseases are discussed next.
Pityriasis Rosea Pityriasis rosea is a mild skin eruption predominantly found in children and young adults. The lesions are multiple pink or pigmented oval papules or plaques 1 to 2 cm in diameter on the trunk and proximal extremities. Mild scaling may be present. The lesions are parallel to the ribs, forming a Christmas tree–like distribution on the trunk. Oral lesions are rare. In children, papular or vesicular variants of the disease may occur.3 In half the cases, the generalized eruption is preceded by 1 week by the appearance of a “herald patch.” This is a larger lesion, 2 to 6 cm in diameter, that resembles the smaller lesions in other respects. The eruption is usually asymptomatic, although pruritus may be present. Pityriasis rosea is self-limited, resolving in 8 to 12 weeks. Its cause is unknown, although a virus is suspected. The differential diagnosis includes tinea corporis, guttate psoriasis, lichen planus, drug eruption, and secondary syphilis. Recurrences are rare. Treatment is usually unnecessary, except for symptomatic alleviation of bothersome pruritus.
Atopic Dermatitis Principles of Disease Atopic dermatitis (AD) is a common dermatologic condition encountered in the ED and commonly referred to as “eczema” or “chronic dermatitis.” AD is the cutaneous manifestation of an atopic state, and although it is not an allergic disorder, it is associated with allergic diseases such as asthma and allergic rhinitis. Patients with AD are known to have abnormalities of both humoral and cell-mediated immunity.25 The exact mechanism is unclear, but eosinophil, mast cell, and lymphocyte activation triggered by increased production of interleukin-4 by specific T helper cells seems to be involved. Increased IgE levels are found in most but not all patients with AD, but there is a poor correlation between the severity of the dermatitis and the serum IgE level.25 The course of AD involves remissions and exacerbations. More than 90% of patients have the onset of AD before 5 years of age. New-onset AD in older children or adults should raise suspicion for other diagnoses.
Clinical Features Atopic dermatitis has no pathognomonic skin lesions or unique laboratory parameters. The United Kingdom’s Working Party revised diagnostic criteria include itchy skin plus three or more of the following: history of flexural involvement, generalized dry skin, history of asthma or hay fever, onset of rash before 2 years of age, and flexural dermatitis.27 These criteria are quite sensitive (85%) and specific (96%). Skin lesions generally appear as inflammatory thickened, papular, or papulovesicular lichenification and hyperpigmentation.28 The skin is typically dry and may be scaly, but in the acute phase, it may also be vesicular, weeping, or oozing. The distribution of lesions varies with the age of the patient. In infants, inflammatory exudative plaques are seen on the cheeks, extensor surfaces, and in the diaper area. Older children and adults have lesions in the antecubital and popliteal flexion areas, neck, face, and upper chest. Infantile AD usually
Differential Considerations The differential diagnosis of infantile AD includes histiocytosis X, Wiskott-Aldrich syndrome, chronic seborrheic dermatitis, phenylketonuria, Bruton’s X-linked agammaglobulinemia, psoriasis, and scabies. Fixed-drug eruptions and contact dermatitis round out the differential diagnosis regardless of age.25,27 Complications of AD include pyogenic skin infections, otitis externa, cataracts, keratoconus, retinal detachment, and cutaneous viral infections.
Management The optimal protocol for management in children has not been established. Treatment should be aimed at controlling inflammation, dryness, and itching. The use of sedating antihistamines at bedtime can be beneficial in patients with AD who have comorbid allergic conditions and sleep disturbances. Daily skin care should be reviewed with patients or caregivers. General recommendations for all patients include avoidance of nonspecific skin irritants, wool, nonessential toiletries and detergents, and using cotton clothing as much as possible. Patients should take daily warm baths or showers for approximately 10 to 15 minutes to hydrate the skin. Baths are followed by gentle pat drying and immediate application of a topical anti-inflammatory medication on the affected areas and a moisturizer such as Cetaphil cream on the asymptomatic areas. Medium-potency topical corticosteroids may be sufficient to treat moderate flares. Ointment-based medications are usually better tolerated by most patients during an acute flare. Skin dryness may be treated by the application of lubricating ointments such as Vaseline or 10% urea in Eucerin cream (not lotion). Treatment of exudative areas includes the application of wet dressings. Such dressings are useful for their moisturizing, anti-inflammatory, and antipruritic actions. Two or three layers of gauze soaked in Burow’s solution should be applied for 15 to 20 minutes four times a day. Antihistamines may be helpful in reducing the pruritus and are also useful for their sedative and soporific effects, although there is no convincing evidence that H1 antihistamines decrease itching in patients with atopic eczema.2 Topical corticosteroids are the cornerstone of therapy and should be prescribed in ointment form. When the dermatitis is severe, a fluorinated corticosteroid ointment such as halfstrength betamethasone valerate should be applied to affected areas of the body three times a day. Fluorinated corticosteroids should not be used on the face because they can produce permanent cutaneous atrophy. Milder corticosteroid prepara-
1533
tions such as 0.025% triamcinolone ointment may be used on the face and intertriginous areas. Patients with extremely severe disease may require systemic steroids. Ultraviolet B treatment is moderately effective, although its mechanism of action is not well understood.28 Cyclosporine and other immunosuppressant agents are being used with some promising benefit. Further studies are needed to determine ideal dosing and safety profiles for these agents.28 Topical calcineurin inhibitors, including tacrolimus ointment and pimecrolimus cream, are nonsteroidal topical immunosuppressants approved in the United States for use on children 2 years or older and are useful for treating lesions on the thinner skin areas (face, groin, and axillae) where repeated applications of topical corticosteroids may result in skin atrophy or striae.29 A burning sensation at the site of application may occur. Note that the Food and Drug Administration has issued a “black box” warning concerning longterm continuous treatment with topical calcineurin inhibitors and cancer, although there is currently no evidence for a causal link.30,31 Inpatient admission should be strongly considered for patients with generalized erythema and exfoliation (erythroderma), and intractable itching as skin breakdown and severe secondary bacterial or viral skin infections may occur.
Skin Infections in Patients with Atopic Dermatitis Patients with AD are susceptible to infection and colonization by a variety of organisms because of their defective skin barrier functions and local skin immunodeficiency. Widespread disseminated viral infections, such as eczema molluscatum, eczema vaccinatum, or eczema herpeticum, and recurrent staphylococcal pustulosis are especially concerning.29 Eczema molluscatum is self-limited. Eczema vaccinatum results from exposure of patients to vaccinia virus either via intentional inoculation or via contact with someone recently immunized against smallpox. Therapy of eczema vaccinatum requires prompt administration of intravenous immunoglobulin, which can be obtained from the Centers for Disease Control and Prevention.32 Eczema herpeticum constitutes a medical emergency. Patients present with disseminated eruptions of dome-shaped vesicles that may or may not be superimposed on areas of eczematous rashes, with the head, neck, and trunk commonly affected. Fever, malaise, and local lymphadenopathy are variable depending on the timing of presentation and host characteristics. Complications include keratoconjunctivitis, viremia, multiorgan involvement with meningitis, and encephalitis.32 Clinical suspicion of eczema herpeticum mandates initiation of intravenous acyclovir in conjunction with antistaphylococcal antibiotics for possible bacterial superinfection. Lumbar puncture should not be attempted if infected lesions are present over the lumbar area. Ophthalmology consultation is needed for patients with periocular or suspected eye involvement.
■ PUSTULES Impetigo Principles of Disease Impetigo is a slowly evolving pustular eruption, most common in preschool children. Currently, Staphylococcus aureus is the most common pathogen, with group A streptococcus a distant second.33 Poor health and hygiene, malnutrition, and various antecedent dermatoses, especially atopic dermatitis, predispose individuals to impetigo.
Chapter 118 / Dermatologic Presentations
begins in the fourth to sixth month of life and improves by the third to fifth year of life. The childhood form occurs between 3 and 6 years of age and resolves spontaneously or continues into the adult form.28 Intense pruritus is a hallmark of AD. During flares, patients may present with complaints of intense itching and failure of routine treatments to control their symptoms. Patients may also present with secondary infections. The itching may be focal or generalized, is worse during the winter, and is triggered by increased body temperature and emotional stress. It may be particularly annoying at night. Excoriations may be prominent, and secondary bacterial infection of excoriated lesions is common. Repeated scratching and rubbing produce lichenification, a condition of hyperpigmentation, thickening of the skin, and accentuation of skin furrows. Lichenification is a common feature of chronic AD.
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1534
Clinical Features
Management
Streptococcal impetigo is found most often on the face and other exposed areas. The eruption often begins as a single pustule but develops into multiple lesions. It begins as 1to 2-mm vesicles with erythematous margins. When these break, they leave red erosions covered with a golden yellow crust. Lesions may be pruritic but usually are not painful. Regional lymphadenopathy is commonly present. Lesions are contagious among infants and young children and less so in older children and adults. Postpyodermal acute glomerulonephritis is a recognized complication of streptococcal impetigo. Staphylococcal impetigo may be differentiated from streptococcal impetigo (ecthyma) by little surrounding erythema in the staphylococcal infection that is more superficial.3 Other diagnostic considerations are herpes simplex virus (HSV) and inflammatory fungal infections. A Gram’s stain obtained from the weepy erosion after removing the crust will reveal grampositive cocci. Bullous impetigo is caused by staphylococci infected by phage group 2. This form is seen primarily in infants and young children. The initial skin lesions are thin-walled, 1- to 2-cm bullae. When these rupture, they leave a thin serous crust and collarette-like remnant of the blister roof at the rim of the crust. The face, neck, and extremities are most often affected. The differential diagnosis is contact dermatitis, HSV infection, superficial fungal infections, and pemphigus vulgaris. A Gram’s stain of the fluid from a bulla reveals gram-positive cocci. Cultures are positive in 95% of cases.
Treatment with an antiseptic cleanser such as povidone-iodine or chlorhexidine every day or every other day for several weeks is usually adequate. For patients with extensive involvement, a 10-day course of erythromycin, 250 mg four times a day, or dicloxacillin, 250 mg four times a day, may be added.3,35,37
Management Systemic and topical therapies are equally successful in treating impetigo.33-35 For more extensive lesions, systemic treatment should be used. There is no evidence, however, that systemic antibiotics prevent the development of acute glomerulonephritis.31,36 The efficacy of topical mupirocin 2% ointment three times a day and oral erythromycin, 250 mg four times a day for 10 days in adults or 30 mg/kg/day in children, or cephalexin, 30 to 40 mg/kg/day three times for 7 to 10 days, is similar.9,33-35,37 Mupiricin should be avoided if there is concern about methicillin-resistant strains. Therapy for bullous impetigo consists of an oral penicillinase-resistant semisynthetic penicillin such as dicloxacillin, 250 mg four times a day for 5 to 7 days for adults, or erythromycin, 250 mg four times a day in adults or 30 to 50 mg/kg/day in children. If the infection is limited to a small area, mupirocin 2% ointment three times a day may be applied. Without treatment, impetigo heals within 3 to 6 weeks.33-35,37
Folliculitis Clinical Features Folliculitis is an inflammation in the hair follicle, usually caused by S. aureus. It appears as a pustule with a central hair. The lesions are usually on the buttocks and thighs, occasionally in the beard or scalp, and may cause mild discomfort. Differential diagnosis includes acne, keratosis pilaris, and fungal infection. Gram-negative folliculitis with Pseudomonas aeruginosa occurs with infected hot tubs and swimming pools or in individuals taking antibiotics for acne, and it can be differentiated from staphylococcal folliculitis by a Gram’s stain of the lesion.
Hidradenitis Suppurativa Hidradenitis suppurativa affects the apocrine sweat glands. Recurrent abscess formation in the axillae and groin resembles localized furunculosis. The condition tends to be recurrent and may be extremely resistant to therapy. Hidradenitis suppurativa may be treated with drainage of abscesses. Antistaphylococcal antibiotics are useful if administered early and for a prolonged period.9 Many cases do not respond, however, and eventually require local excision and skin grafting of the involved area. Antiandrogen therapy may be considered if antibiotics do not produce improvement.9
Carbuncle A carbuncle is a large abscess that develops in the thick, inelastic skin of the back of the neck, back, or thighs. Carbuncles produce severe pain and fever. Septicemia may accompany the lesions. The diagnosis of skin abscess, furuncle, or carbuncle is usually made clinically. Local heat should be applied to furuncles and carbuncles, which should be incised and drained when fluctuant. Antibiotics are unnecessary with incision and drainage unless cellulitis or septicemia is present.
Community-Associated Methicillin-Resistant Staphylococcus Aureus The incidence of community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) has soared since the first report in 1993.38 In many major U.S. cities, CA-MRSA is now the most common pathogen cultured from ED patients presenting with skin and soft tissue infections.39 Concern exists that CA-MRSA may be more virulent than methicillin-sensitive strains and colonization with CA-MRSA may produce more overt infections.39
Epidemiology Hospital-acquired MRSA isolates can survive on a variety of inanimate surfaces, sometimes for weeks. It is unclear whether this is also true for CA-MRSA isolates; if it is true, their presence on such items as clothing, towels, and athletic equipment might contribute to outbreaks. Pets (including dogs and cats), livestock, and birds have been identified as MRSA carriers40; their role in MRSA transmission to humans requires further evaluation.
Clinical Features CA-MRSA infections most often present as skin and soft tissue suppuration such as an abscess, furuncle, or cellulites. Lesions frequently exhibit central necrosis and are often confused with spider bites by patients. No clinical features distinguish with certainty skin and soft tissue infections caused by MRSA from those caused by methicillin-susceptible S. aureus.41 Although rare, CA-MRSA infection can also present as necrotizing fasciitis.42 Recurrences of CA-MRSA cellulitis are common. Contagion among the close household contacts of patients, as well
as correctional facility, school, and sports-team contacts, is well recognized.
Several studies have demonstrated excellent outcomes for abscesses caused by CA-MRSA that are treated with incision and drainage alone.43 If antibiotics are needed, information on local antibiotic-resistance patterns can help clinicians assess the likelihood of CA-MRSA infection and guide decisions regarding empirical treatment. Obtaining a specimen for culture and susceptibility testing, which was considered to be unnecessary in the pre-CA-MRSA era, may be useful in guiding therapy. Specimens are obtained at the time of incision and drainage of purulent lesions. In patients with larger abscesses, systemic signs of infection, or both, antimicrobial therapy is needed in addition to incision and drainage. The optimal oral antimicrobial regimen for the treatment of skin and soft tissue infections is not known. The type and route of therapy should be guided by the severity of the clinical syndrome. Clindamycin combines MRSA activity with effectiveness against the majority of other gram-positive organisms. Side effects include diarrhea, Clostridium difficile colitis, and increasing rates of clindamycin resistance.44 Rifamycin has anti-MRSA activity, but resistance readily develops, so it should not be used alone. Its long half-life allows once-a-day administration. It penetrates well into all tissues and body fluids. It has a high potential for drug-drug interactions.45 Linezolid, a newer antimicrobial agent, is active against almost all CA-MRSA isolates and group A streptococci. Disadvantages of its use include high cost, lack of routine availability, hematologic side effects, and potential for resistance among S. aureus strains. Prolonged linezolid administration increases the likelihood of resistance.46 Trimethoprim-sulfamethoxazole or tetracycline is not recommended as sole empirical therapy for a nonpurulent cellulitis of unknown cause because of group A streptococci resistance to these agents.44 A β-lactam antibiotic may augment treatment. Cephalosporins and macrolides, including newer ones, are ineffective against CA-MRSA.10 Fluoroquinolones should be avoided because S. aureus resistance develops readily and is already widely prevalent.39 Patients with large abscesses, abscesses in high-risk locations, fever, signs of systemic infection, young age, or immunodeficiency should prompt consideration of hospitalization. The detailed management of invasive disease due to CAMRSA is discussed elsewhere. Vancomycin is still considered the parenteral drug of choice for patients with invasive S. aureus infection, although clinical failures have been reported. It seems reasonable to combine vancomycin with another effective antistaphylococcal agent because many antibiotics have better bactericidal activity. In severely ill patients, carbapenems such as meropenem, panipenem, and ertapenem, which are active against CA-MRSA and synergistic with vancomycin, should be used. 47 Use of parenteral clindamycin (not recommended as monotherapy), bactrim, and linezolid has been described. In addition, daptomycin and tigecycline are now approved for the treatment of skin and soft tissue infections caused by MRSA.48,49 A fixed combination of the streptogramins quinupristin and dalfopristin (Synercid) can be used to treat CA-MRSA skin and soft tissue infections. Its use has been limited by the potential for drug-drug interactions and by side effects.50 Recurrent infections are generally treated like initial episodes. Some providers recommend “decolonization” strategies, although neither the indications for their use nor their effectiveness in reducing the risk of recurrences is established.
Prevention Common antiseptics appear to retain reasonable activity against CA-MRSA, although the results of recent studies are somewhat conflicting. Good personal hygiene including appropriate hand-washing techniques, separation of infected patients from other types of patients, and routine cleaning of shared equipment are essential to limiting CA-MRSA spread.50,51
Gonococcal Dermatitis Clinical Features The arthritis-dermatitis syndrome is the most common presentation of disseminated gonococcal disease.52,53 It occurs in 1 or 2% of patients with gonorrhea, affecting women primarily.52 Fever and migratory polyarthralgias commonly accompany the skin lesions. The lesions are often multiple and have a predilection for periarticular regions of the distal extremities.52 The lesions begin as erythematous or hemorrhagic papules that evolve into pustules and vesicles with an erythematous halo (Fig. 118-3). They closely resemble the lesions of meningococcemia at this stage. They are tender and may have a gray necrotic or hemorrhagic center. Healing with crust formation usually occurs within 4 or 5 days, although recurrent crops of lesions may appear even after antibiotics have been started.52
Diagnostic Strategies The lesions usually have a negative culture for gonococci, and the Gram’s stain only occasionally reveals the organisms. A more reliable diagnostic technique is immunofluorescent antibody staining of direct smears from pustules.52 This method indicates that the lesions may be the result of hematogenous dissemination of nonviable gonococci.52
Management Current treatment of disseminated gonococcal infection is ceftriaxone, 1 g intramuscularly (IM) or intravenously (IV) every 24 hours, or ceftizoxime or cefotaxime, 1 g IV every
Figure 118-3. Typical skin lesions of disseminated gonococcal disease. (Courtesy of David Effron, MD.)
Chapter 118 / Dermatologic Presentations
Management
1535
Decolonization strategies include the use of intranasal mupirocin to reduce nasal carriage of MRSA; however, eradication of nasal colonization appears to be transient. The efficacy of attempts to eradicate CA-MRSA among household members has not been studied.
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1536
Figure 118-4. Facial cellulitis. (Courtesy of David Effron, MD.) 8 hours. Patients allergic to β-lactam antibiotics may be treated with spectinomycin 2 g IM every 12 hours. A total of 7 days of antibiotic therapy is required, with the remaining course of cefixime, 400 mg twice a day, cefuroxime or ciprofloxacin, 500 mg twice a day, or ofloxacin, 400 mg twice a day. Ciprofloxacin and ofloxacin are not recommended due to increasing resistance patterns or for pregnant women or children younger than 17 years.52,53 Hospitalization is recommended for patients in whom the diagnosis is uncertain and for those who have septic arthritis, meningitis, or endocarditis.
Figure 118-5. Urticaria (hives). (Courtesy of David Effron, MD.)
■ ERYTHEMA Cellulitis is an infection of the skin tissue denoted by erythema, swelling, and local tenderness (Fig. 118-4).54-58 Erysipelas is a streptococcal infection of the skin and subcutaneous tissue. The involved area is red, indurated, and edematous.59 These disorders are discussed in Chapter 135.
■ RED MACULES Drug Eruption Principles of Disease A given drug can produce a skin eruption of a different appearance in different patients or a different appearance in the same patient on different occasions. The most common eruptions are urticaria (hives) (Fig. 118-5) and, more commonly, morbilliform rashes (Fig. 118-6). Drug reactions tend to appear within a week after the drug is taken, with the exception of reactions to semisynthetic penicillins, which commonly occur later. Skin lesions may appear after a drug has been discontinued and may worsen if the drug or its metabolites persist in the system. Special note should be made of penicillin because it is the most common cause of drug reaction. Serum sickness and urticaria are the most common manifestations of penicillin allergy. Atopic patients and those with a history of hay fever, asthma, or eczema are at special risk. On the other hand, a number of drugs in common use rarely produce eruptions. Among these are acetaminophen, aluminum hydroxide (Maalox), codeine, digoxin, erythromycin, ferrous sulfate, meperidine (Demerol), morphine, and prednisone.
Clinical Features Some of the more common skin reactions produced by commonly used drugs are listed in Table 118-2. Exanthematous drug
Figure 118-6. Morbilliform drug eruption. (Courtesy of David Effron, MD.)
eruptions resemble the skin manifestations of various viral or bacterial infections and are usually widespread symmetric maculopapular eruptions. Severe cases may progress to exfoliative dermatitis. Eczematous drug rashes resemble those of contact dermatitis but are generally more extensive. They begin as erythematous or papular eruptions that may become vesicular. Prior sensitization to a topical medication is common in cases of this type of eruption. Vasculitic lesions begin as erythematous papules or nodules but may ulcerate and become gangrenous. Urticarial vasculitis is characterized by persistent urticarial lesions with histologic evidence of leukocytoclastic vasculitis. Wheel-and-flare–like lesions that hurt or burn more than itch, lesions lasting more than 24 hours, and urticarial lesions that leave prolonged pigmentary changes or inflammatory lesions should prompt suspicion for urticarial vasculitis.60 Purpuric drug eruptions
× × × × × × × × Chloral hydrate
× × ×
×
×
EXANTHEMATOUS
Opiates
× × × × × × × ×
× × × × × × × × ×
URTICARIAL*
× Tolbutamide, phenytoin
× ×
×
× ×
×
×
× ×
ERYTHEMA MULTIFORME†
*The most common causes of drug-induced urticaria are aspirin and penicillins. † The long-acting sulfonamides have been linked to Stevens-Johnson syndrome.
Aminophylline Anovulatory drugs Barbiturates Bromides Chloramphenicol Insulin Iodides Isoniazid Meprobamate Penicillin Phenacetin Phenolphthalein Phenothiazines Phenylbutazone Quinidine Quinine Salicylates Sulfonamides Tetracycline Thiazides Others
THERAPEUTIC AGENTS
Tolbutamide
× ×
×
×
×
×
TOXIC EPIDERMAL NECROLYSIS
× Diphenhydramine ephedrine, thiamine, methyldopa
×
×
× ×
×
ECZEMATOUS
× ×
×
×
×
×
ERYTHEMA NODOSUM
TYPE OF ERUPTION
Antimalarial drugs, guanethidine
× ×
× ×
×
×
×
VASCULITIS
× ×
× ×
×
×
PURPURA
× × × Antimalarial drugs, chlordiazepoxide, reserpine
×
×
PHOTOSENSITIVE
Diazepam, indomethacin
× × ×
× × × × ×
×
× ×
FIXED
Chapter 118 / Dermatologic Presentations
Table 118-2 Types of Lesions Characteristically Caused by Commonly Used Drugs
1537
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1538
water baths with colloidal oatmeal (Aveeno) emollient or cornstarch, and diphenhydramine (Benadryl), 50 mg (5 mg/kg/24 hr in children) every 6 hours, are likely to be beneficial.
Staphylococcal Scalded Skin Clinical Features
Figure 118-7. Purpuric lesions. (Courtesy of David Effron, MD.)
may be the result of bone marrow suppression, platelet destruction, or vasculitis (Fig. 118-7). Ultimately, a skin biopsy is needed to confirm the diagnosis of vasculitis. Photosensitive drug reactions require the presence of sunlight and are seen most commonly on sun-exposed areas of skin. This class of reactions is commonly divided into phototoxic and photoallergic. Phototoxic reactions are more common. Sulfonamides, sulfonylureas, thiazide diuretics, and tetracyclines are common causes (see Fig. 118-7). This type of reaction does not primarily involve immunologic mechanisms and occurs in any person taking an adequate quantity of the drug and exposed to sunlight. The lesions usually have the appearance of a severe sunburn but may be bullous or papular. Pruritus is typically minimal or absent.61 Photoallergic reactions are the result of antigen formation that results in the formation of sensitized lymphocytes. These reactions therefore represent a delayed immunologic response. A photoallergic reaction occurs only in sensitized individuals, usually 2 weeks or longer after exposure to the drug and sunlight. Its occurrence is not dose related, and the eruption usually appears eczematous and intensely pruritic. Chlorpromazine, promethazine, and chlordiazepoxide are common sensitizers of photoallergic reactions.61 Patients who develop photoallergic reactions should be withdrawn from inciting drugs. Patients who are subject to photosensitive drug eruptions may be required to avoid prolonged sunlight exposure. Sunscreen containing 5% aminobenzoic acid should be used during any such exposure. Fixed-drug eruptions appear and recur at the same anatomic site after repeated exposure to the same drug. The lesions are usually sharply marginated and round or oval. They may be pigmented, erythematous, or violaceous. Pruritus may be prominent.
Differential Considerations The differential diagnosis of drug eruptions includes viral exanthem, chronic exfoliative erythroderma caused by psoriasis or atopic dermatitis, malignancy, scarlet fever, staphylococcal scarlatiniform eruptions, and Kawasaki disease.9,61
Management Treatment of drug eruptions should begin with discontinuation of the inciting agent. Patients should be warned that drug eruptions clear slowly after discontinuation of the offending agent. Itching may be treated with the application of a drying antipruritic lotion such as calamine. Cool compresses, tepid
Staphylococcal scalded skin syndrome generally occurs in children 6 years of age or younger. It is caused by an infection with phage group 2 exotoxin-producing staphylococci. The illness begins with erythema and crusting around the mouth. The erythema then spreads down the body, followed by bulla formation and desquamation. Mucous membranes are usually not involved, but minimal involvement is occasionally seen. After desquamation occurs, the lesions dry up quickly, with clinical resolution in 3 to 7 days.
Management Most group 2 toxin-producing organisms are penicillin resistant. Although most patients will recover without antibiotic treatment, IV therapy with 50 to 100 mg/kg of nafcillin daily or oral cloxacillin 50 mg/kg/day or dicloxacillin is recommended.9,62,63
Toxic Epidermal Necrolysis Principles of Disease Many consider Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) as a continuous spectrum of the same disease. Both are true dermatologic emergencies. The main feature of non–staphylococcal-induced TEN, or Lyell’s disease, is the separation of large sheets of epidermis from underlying dermis. Drugs, including the long-acting sulfa drugs, penicillin, aspirin, barbiturates, phenytoin, carbamazepine, allopurinol, and nonsteroidal anti-inflammatory drugs, are an important cause of TEN. TEN has occurred after vaccination and immunization against poliomyelitis, measles, smallpox, diphtheria, and tetanus. It has also been found in association with lymphoma.
Clinical Features Toxic epidermal necrolysis commonly begins with prodromal symptoms, such as malaise, rhinitis, sore throat, body aches, and fever. These are followed by the abrupt development of a macular rash that may or may not appear as target lesions.64 Mucous membrane involvement commonly precedes the rash in TEN. The macular exanthem usually starts centrally and then spreads to the extremities. The exanthem becomes confluent and dermal-epidermal dissociation ensues, resulting in a positive Nikolsky’s sign, denudation with shear stress, and the skin is commonly painful to the touch. Mucous membrane involvement becomes more apparent during the progression phase.64 Involvement of the conjunctivae and cornea may lead to permanent scarring and blindness. The full thickness of epidermis is involved. The two conditions are easily histologically distinguishable with a skin biopsy (Fig. 118-8). A mortality rate of 15 to 20% is expected with this condition.61
Management The treatment of TEN includes discontinuation of the offending agent, fluid replacement, and aggressive infection control.9,61
1539
Management Initial treatment of TSS consists of IV fluid replacement, ventilatory support, pressor agents, penicillinase-resistant antibiotics, and drainage of infected sites.63,66
Urticaria Principles of Disease
Figure 118-8. Toxic epidermal necrolysis. (Courtesy of David Effron, MD.) Administration of systemic corticosteroids is controversial.61 They have little effect on the disease and may mask signs of impending sepsis. Plasmapheresis is considered experimental.45 The mainstay of treatment is excellent supportive care, prevention of secondary infection, and expert wound management. This is usually best accomplished in a center with burn expertise.
Toxic Shock Syndrome Principles of Disease Toxic shock syndrome (TSS) is an acute febrile illness characterized by a diffuse desquamating erythroderma. Classically composed of high fever, hypotension, constitutional symptoms, multiorgan involvement, and rash, the syndrome gained notoriety in the early 1980s because of association with tampon use. However, it is also well-known in men and children. Its appearance has often been linked to exotoxin-producing S. aureus. Most cases of nonmenstrual TSS occur in the postoperative setting. TSS has also been associated with various staphylococcal and streptococcal infections, including empyema, osteomyelitis, fasciitis, septic abortion, peritonsillar abscess, sinusitis, burns, and subcutaneous abscess.63 TSS is associated with severe group A beta-hemolytic streptococcal infections. It has been reported in previously healthy patients, immunocompromised patients, and elderly patients. Fatigue, localized pain, and nonspecific symptoms herald the onset of this disease, followed by septic shock and multisystem organ failure.63,65
Clinical Features Diagnosis of TSS requires the presence of (1) fever of at least 38.9° C; (2) hypotension, with a systolic blood pressure of 90 mm Hg or less; (3) skin rash; and (4) involvement of at least three organ systems.9,63 Systemic involvement may include the gastrointestinal (GI) tract, muscular system, or central
Urticaria may occur in isolation or as part of a systemic anaphylactic reaction. The following discussion pertains to urticaria occurring in the absence of systemic symptoms. Anaphylactic reactions are discussed in Chapter 117. Approximately 15 to 20% of the population experience urticaria during their lifetime. Acute urticaria is seen in both sexes and is more likely to have an allergic cause. Chronic urticaria is more common in women in their 40s and 50s. Half of all patients with chronic urticaria have the disease for 5 years and one fourth for 20 years.67 Various mediators, including histamine, bradykinin, kallikrein, and acetylcholine, are thought to play a role in urticaria production. Urticaria may be initiated by immunologic or nonimmunologic mechanisms. Hives found in anaphylaxis and serum sickness represent an immunologic reaction. Nonimmunologic urticaria may be produced by degranulation of mast cells, which may be caused by a number of foods and drugs, including aspirin and narcotics. Substances that can cause urticaria by contact with the skin include foods, textiles, animal dander and saliva, plants, topical medications, chemicals, and cosmetics.68 The role of drugs in the production of urticaria is discussed in the section on drug eruption. Almost any drug may produce urticaria, although penicillin and aspirin are the most common. Traces of penicillin may be present in dairy products as well as in medications. The mechanism of production of urticaria by aspirin is unknown but is probably nonimmunologic, and the effects of aspirin may persist for a number of weeks after ingestion.68 A variety of food allergies, such as fish, eggs, or nuts, may result in urticaria. In addition, foods such as lobster and strawberries can release histamine through a nonimmunologic mechanism. Hereditary forms of urticaria include familial cold urticaria and hereditary angioneurotic edema. Infections are an uncommon cause of urticaria, except in children in whom viral infections often cause hives. Occult infections with Candida, the dermatophytes, bacteria, viruses, and parasites may trigger hives. Viral infections that produce urticaria include hepatitis, mononucleosis, and coxsackievirus infections. The inhalation of pollens, mold, animal dander, dust, plant products, and aerosols may produce urticaria. Respiratory symptoms may accompany the dermatosis, and a seasonal pattern of occurrence may be present. Stings and bites of insects, arthropods, and various marine animals may also produce an urticarial eruption.
Chapter 118 / Dermatologic Presentations
nervous system (CNS) and laboratory evidence of renal, hepatic, or hematologic dysfunction. Headache, myalgias, arthralgia, alteration of consciousness, nausea, vomiting, and diarrhea may be present. The rash is typically a diffuse, blanching, macular erythroderma. Accompanying nonexudative mucous membrane inflammation is common. Pharyngitis, sometimes accompanied by a “strawberry tongue,” conjunctivitis, or vaginitis may be seen. As a rule, the rash fades within 3 days of its appearance. This is followed by a full-thickness desquamation, most commonly involving the hands and feet.
1540
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
Differential Considerations The differential diagnosis of urticaria includes erythema multiforme, erythema marginatum, and juvenile rheumatoid arthritis.
Management
Figure 118-9. Dermatographism. (Courtesy of David Effron, MD.)
Occasionally, patients with systemic lupus erythematosus, lymphoma, carcinoma, hyperthyroidism, rheumatic fever, and juvenile rheumatoid arthritis develop an urticarial eruption. The association is uncommon enough that it is not necessary for a urticaria workup to include a search for malignancy in most cases. A number of physical agents produce urticaria. Dermatographism is present when firm stroking of the skin produces an urticarial wheal within 30 minutes (Fig. 118-9) and is the most common form of physical urticaria. Pressure urticaria is distinct from dermatographism in that the onset of urticaria is delayed by 4 to 8 hours after the application of physical pressure. There is no other particular significance to this form of urticaria. Cold urticaria may be either familial or, more commonly, acquired. Cold urticaria may also be associated with underlying illness, such as cryoglobulinemia, cryofibrinogenemia, syphilis, and connective tissue disease.67,68 Cyproheptadine, 2 to 4 mg two or three times a day, is useful in the suppression of primary cold urticaria.67 Side effects of this drug include drowsiness and an increased appetite.67 Antihistamines taken 30 to 60 minutes before cold exposure may be helpful. Doxepin is also useful; begin at 10 mg at bedtime and gradually increase to 10 to 25 mg three times a day.67 Cholinergic urticaria is induced by exercise, heat, or emotional stress. It may be associated with pruritus, nausea, abdominal pain, and headache.67The lesions of cholinergic urticaria are wheals 1 to 3 mm in diameter surrounded by extensive erythematous flares and, occasionally, satellite wheals. Cholinergic urticaria responds better to hydroxyzine than do other physical urticarias.67 Heat is a rare cause of hives. Solar urticaria, also uncommon, is confined to sun-exposed areas of skin and clears rapidly when the light stimulus is removed. Extensive sun exposure may cause wheezing, dizziness, and syncope in a susceptible individual.67 Sunscreens have not been proven to be effective for the prevention of solar urticaria.67 The cause of chronic urticaria in adults is often not determined, although the etiologic factors responsible for urticaria in children are more readily identifiable.69
Clinical Features Urticaria appears as edematous plaques with pale centers and red borders and is easily recognizable (see Fig. 118-5). Individual hives are transient, lasting less than 24 hours, although new hives may continuously develop, which represent localized dermal edema produced by transvascular fluid extravasation.
Treatment of urticaria involves the removal of the inciting factor, when applicable, and the administration of antihistamines or other antipruritics. Hydroxyzine (Atarax and Vistaril) in a dose of 10 to 25 mg (2 mg/kg/24 hr in children) is usually effective in providing symptomatic relief. Alternatives are nonsedating antihistamines, such as terfenadine 60 mg twice a day, astemizole 10 mg daily, or fexofenadine 60 mg twice a day.70 Prednisone is also effective, but the urticaria can rebound, making cessation of prednisone sometimes difficult. For chronic urticaria, long-term therapy with antihistamines may be needed.
Serum Sickness Serum sickness is a clinical syndrome most commonly caused by drugs and characterized by fever, lymphadenopathy, arthralgias, cutaneous eruptions, gastrointestinal disturbances, and malaise. It is often associated with proteinuria, without evidence of glomerulonephritis.71 A widespread morbilliform or urticarial rash or erythema multiforme–like eruption develops, sometimes involving the palms and soles. 72 The most common cause of serum sickness and serum sickness–like reactions is a hypersensitivity reaction to drugs.73 Cefaclor is a common culprit in causing serum sickness–like reactions. Serum sickness usually begins 1 to 3 weeks after the start of administration of the medication, although it can occur within 12 to 36 hours in individuals who have been sensitized during a previous exposure. Serum sickness is mediated by the tissue deposition of circulating immune complexes, the activation of complement, and the ensuing inflammatory response. This is a type III (immune complex) reaction, or Arthus reaction.
Management Discontinuation of the culprit drug and symptomatic treatment with antihistamines and topical corticosteroids are recommended. A short course of oral corticosteroids may be required in patients with more severe symptoms.73 The drug causing the reaction should be avoided in the future. For cefaclor and cefprozil, the risk of cross-reaction with other β-lactam antibiotics is small, and the further administration of another cephalosporin is usually well tolerated.73 However, some clinicians recommend that patients who experience serum sickness–like reactions from cefaclor avoid all β-lactam drugs.
■ EXANTHEMS Principles of Disease An exanthem is defined as a skin eruption that occurs as a symptom of a general disease. Approximately 30 enteroviruses, predominantly the coxsackievirus and echovirus groups, and four types of adenoviruses are known to produce exanthems. Other viruses may do so as well. The exanthems of the coxsackievirus and echovirus are most thoroughly documented. Most viral exanthems are maculopapular, although scarlatiniform, erythematous, vesicular, and petechial rashes
1541
person within 6 days of exposure. The recommended dose of ISG is 0.25 mL/kg IM in children. Live measles virus vaccine given within 72 hours of exposure may be effective in preventing measles.74 Some authors suggest vitamin A soon after exposure. The incidence of measles has decreased since the resurgence seen in 1989 to 1991.74 The patterns observed during outbreaks include a shift from preschool-aged children to older adults and among groups who do not routinely obtain vaccination, such as immigrants.
Measles
Clinical Features
Clinical Features
The onset of the illness is usually abrupt, with headache, nausea and vomiting, myalgias, chills, and a fever spiking to 40° C. Occasionally, the onset is more gradual, with progressive anorexia, malaise, and fever. The disease may last 3 weeks and may be severe with prominent CNS, cardiac, pulmonary, GI, renal, and other organ involvement; disseminated intravascular coagulation; or shock. The rash develops on the second to fourth day or, occasionally, as late as the sixth day of the illness. It begins with erythematous macules that blanch on pressure, appearing first on the wrists and ankles. These macules spread up the extremities and to the trunk and face in a matter of hours. They may become petechial or hemorrhagic. Lesions on the palms and soles are particularly characteristic. Increased capillary fragility and splenomegaly may be present.
Measles is a highly contagious viral illness spread by contact with infectious droplets, with an incubation period of 10 to 14 days. Patients are contagious from 1 or 2 days before onset of symptoms up to 4 days after the appearance of the rash.74 Symptoms begin with fever and malaise. The fever usually increases daily in a stepwise manner until it reaches approximately 40.5° C on the fifth or sixth day of the illness. Cough, coryza, and conjunctivitis begin within 24 hours of the onset of symptoms. On the second day of the illness, Koplik’s spots, which are pathognomonic of the disease, appear on the buccal mucosa as small, irregular, bright red spots with bluish-white centers. Beginning opposite the molars, Koplik’s spots spread to involve a variable extent of the oropharynx. The cutaneous eruption of measles begins on the third to fifth day of the illness. Maculopapular erythematous lesions involve the forehead and upper neck and spread to involve the face, trunk, arms, and finally the legs and feet. Koplik’s spots begin to disappear coincident with the appearance of the rash. By the third day of its presence, the rash begins to fade, doing so in the order of its appearance, and the fever subsides. Complications include otitis media, encephalitis, and pneumonitis. Otitis media is the most common complication. Encephalitis occurs in approximately 1 in 1000 cases of measles and carries a 15% mortality. Measles pneumonia may also be life threatening.
Management If bacterial invasion occurs with otitis or pneumonia, the use of antibiotics is indicated. Otherwise, treatment is supportive. Isolation of infected children is of limited value because exposure usually occurs before the appearance of the rash and the presence of Koplik’s spots renders the disease diagnosable. Measles is not contagious after the fifth day of the presence of the rash. Infection confers lifelong immunity. The illness can be modified or prevented by the administration of human immune serum globulin (ISG) in a susceptible
Rocky Mountain Spotted Fever Principles of Disease Rocky Mountain spotted fever is caused by Rickettsia rickettsii, an organism harbored by a variety of ticks. The organism is transmitted to humans through tick saliva at the time of a tick bite or when the tick is crushed while in contact with the host. Although originally described in the Rocky Mountain region, this disease occurs in other areas of North, South, and Central America. Most reported cases are from the southeastern United States.
Diagnostic Strategies The Weil-Felix reaction is the best known serologic diagnostic test, but the development of Weil-Felix agglutinins in cases of Rocky Mountain spotted fever is not constant, and more specific immunofluorescent procedures have been developed.75 Treatment should not await the result of such tests, however, but should begin as soon as the disease is suspected on clinical grounds.
Management Tetracycline (25–30 mg/kg/day in divided doses) is the antibiotic of choice. If the patient is unable to take oral medications, tetracycline may be administered IV, with a 15 mg/kg loading dose followed by a maintenance dosage of 15 mg/kg/day. Doxycycline may be used as well in a dosage of 4.4 mg/kg/day divided every 6 hours followed by 1.1 mg/kg twice a day, up to 30 mg/day. Chloramphenicol may be used for patients allergic to tetracycline and in children younger than 9 years. A usual course is 6 to 10 days and should continue for 72 hours after defervescence.75 Sulfa drugs should be avoided because they can exacerbate the illness. Rickettsiae are routinely resistant to penicillins, cephalosporins, aminoglycosides, and erythromycin.75
Chapter 118 / Dermatologic Presentations
are occasionally seen. The eruptions are variable in their extent, nonpruritic, and do not desquamate. Oropharyngeal lesions may be present. Infection with echovirus type 9 may be accompanied by meningitis and a petechial exanthem resembling meningococcemia, although the exanthem also occurs without meningeal involvement. Infections caused by echovirus type 16 (Boston exanthem) and coxsackievirus group B, type 5, may resemble roseola infantum but are more likely to occur in adults. Infections caused by coxsackievirus group A, type 16, cause a distinctive syndrome of vesicular stomatitis and 1- to 4-mm oral vesicles involving the dorsa of the hands and lateral borders of the feet. Disease caused by coxsackievirus group A, type 9, has been the most extensively studied. It may be associated with meningoencephalitis or interstitial pneumonia. The rash is usually maculopapular, begins on the face or trunk, and spreads to the extremities. A vesicular eruption resembling varicella may occur. The classic viral exanthems are rubeola (measles), rubella (German measles), herpesvirus 6 (roseola), parvovirus B19 (erythema infectiosum, or fifth disease), and the enteroviruses (echovirus and coxsackievirus).3,9 Widespread immunization programs have reduced the incidence of rubeola and rubella.
1542
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
Roseola Infantum Roseola infantum, otherwise known as exanthem subitum or sixth disease, is a benign illness caused by human herpesvirus 6 and characterized by fever and a skin eruption. A roseola-like illness has occasionally been associated with other illnesses.74 Ninety-five percent of cases are seen in children 6 months to 3 years of age, and most of these are in infants younger than 2 years. A febrile seizure may occur. The fever typically has an abrupt onset, rising rapidly to 39° C to 41° C, and is present consistently or intermittently for 3 or 4 days, at which time the temperature drops precipitously to normal. The rash appears with defervescence. The lesions are discrete pink or rose-colored macules or maculopapules 2 or 3 mm in diameter, which blanch on pressure and rarely coalesce. The trunk is involved initially, with the eruption typically spreading to the neck and extremities. Occasionally, the eruptions are limited to the trunk. The rash clears over 1 or 2 days without desquamation. Despite the presence of a high fever, the infant usually appears well. Encephalitis is a very rare complication.74 The prognosis is excellent, and no treatment is necessary.
Rubella Rubella, or German measles, is a viral illness characterized by fever, skin eruption, and generalized lymphadenopathy. It is spread by droplet contact, and peak incidence is in the winter and early spring. The incubation period is typically 14 to 21 days, and the rash heralds the onset of the illness in children. The maximum time of communicability is in the few days before and 5 to 7 days after the onset of the rash.74 Infants with congenital rubella can shed virus for more than 1 year.74 In adults, a 1- to 6-day prodrome of headache, malaise, sore throat, coryza, and low-grade fever precedes the rash. These symptoms generally disappear within 24 hours after the appearance of the skin eruption. The rash of pink to red maculopapules appears first on the face and spreads rapidly to the neck, trunk, and extremities. Those on the trunk may coalesce, but lesions on the extremities do not. The rash remains for 1 to 5 days, classically dis appearing at the end of 3 days. Although clearing may be accompanied by fine desquamation, this sign is usually absent. Lymphadenopathy may begin as early as 1 week before the rash. Although this is generalized, the nodes most apparent are the suboccipital, postauricular, and posterior cervical groups. Palpable adenopathy may be apparent several weeks after other signs and symptoms have subsided. The major complications of rubella include encephalitis, arthritis, and thrombocytopenia. The most severe complication is fetal damage. A total of 24% of infected fetuses have a congenital defect. A maternal infection may be determined by obtaining serum for hemagglutination inhibition antibody determinations, acutely and in 2 weeks. A fourfold rise in the titer is diagnostic of rubella infection. The routine use of postexposure prophylaxis of rubella in an unvaccinated woman in early pregnancy is not recommended. No treatment is required in many cases of rubella. Antipyretics are usually adequate for the treatment of headache, arthralgias, and painful lymphadenopathy.
Arthralgia and arthritis occur commonly in adults but rarely in children. The rash is intensely red on the face and gives a “slapped-cheek” appearance with circumoral pallor. A maculopapular lacelike rash, which may be noted on the arms, moves caudally to the trunk, buttocks, and thighs. The rash may recur with changes in temperature and exposure to sunlight. The incubation period is usually between 4 and 14 days.74 Parvovirus B19 infection may also result in asymptomatic infection, upper respiratory infection, atypical rash, and arthritis without rash. Rarely, it has been reported to cause hepatitis.74 Infected immunodeficient patients may experience chronic anemia as a result of this disease. Patients with sickle cell disease or other hemolytic anemias may develop an aplastic crisis lasting 7 to 10 days.74 Parvovirus B19 infection during pregnancy can cause fetal hydrops and death.74 No congenital anomalies have been reported. No treatment is required.
Scarlet Fever Clinical Features The incidence of scarlet fever has declined in recent years. The illness has an abrupt onset with fever, chills, malaise, and sore throat followed within 12 to 48 hours by a distinctive rash that begins on the chest and spreads rapidly, usually within 24 hours. Circumoral pallor may be noted. The skin has a rough sandpaper-like texture because of the multitude of pinheadsized lesions. The pharynx is injected, and there may be erythematous lesions or petechiae on the palate. After the resolution of symptoms, desquamation of the involved areas occurs and is characteristic of the disease. Complications include the development of a streptococcal infection of lymph nodes, tonsils, the middle ear, and the respiratory tract. Late complications include rheumatic fever and acute glomerulonephritis (Fig. 118-10).
Management Treatment is aimed at providing adequate antistreptococcal blood antibiotic levels for at least 10 days. Oral penicillin VK 50 mg/kg/day (40,000–80,000 units) in four divided doses in children or 250 mg four times a day in adults is administered. Benzathine penicillin (given as Bicillin CR) is administered IM. In patients weighing less than 30 pounds, 300,000 units of benzathine penicillin is used; in patients weighing 31 to 60
Erythema Infectiosum Erythema infectiosum, or fifth disease, is caused by parvovirus B19 infection. It is characterized by mild systemic symptoms, fever in 10 to 15% of patients, and a characteristic rash.
Figure 118-10. Erythema marginatum associated with rheumatic fever. (Courtesy of David Effron, MD.)
■ PAPULAR LESIONS Contact Dermatitis Principles of Disease Contact dermatitis is an inflammatory reaction of the skin to a chemical, physical, or biologic agent. The inducing agent acts as an irritant or allergic sensitizer. Allergic contact dermatitis is a form of delayed hypersensitivity mediated by lymphocytes sensitized by the contact of the allergen to the skin. It is less common than irritant contact dermatitis.26 Caustics, industrial solvents, and detergents are common causes of irritant dermatitis. Dermatitis may result from brief contact with a potent caustic or from repeated or prolonged contact with milder irritants. Clothing, jewelry, soaps, cosmetics, plants, and medications contain allergens that commonly cause allergic contact dermatitis. The most common allergens include rubber compounds, plants of the Rhus genus (poison ivy, oak, and sumac), nickel (often used in jewelry alloys), paraphenyldenediamine (an ingredient in hair dyes and industrial chemicals), and ethylenediamine (a stabilizer in topical medications).26 Sensitization to poison ivy results in sensitization to other plants in this family, such as cashew, mango, lacquer, and ginkgo trees.76
Clinical Features The primary lesions of contact dermatitis are papules, vesicles, or bullae on an erythematous bed. Of the allergens, Rhus species are the most likely to cause bullous eruptions. Oozing, crusting, scaling, and fissuring may be found, along with lichenification in chronic lesions. The distribution of the eruption depends on the specific contactant and may be localized, asymmetric linear, or unilateral (Figs. 118-11 and 118-12). Mucous membranes are usually spared unless directly exposed to the inciting agent. A history of exposure is the most significant factor favoring the diagnosis. If doubt exists about the diagnosis, the patient should be referred for allergic patch testing.
Figure 118-11. Contact dermatitis secondary to nickel. (Courtesy of David Effron, MD.)
1543
Management Treatment of contact dermatitis includes avoidance of the irritant or allergen and treatment of secondary bacterial infection. Oozing or vesiculated lesions should be treated with cool wet compresses of Burow’s solution applied for 15 minutes three or four times a day. Topical baths, available over the counter, may also be comforting. A course of systemic corticosteroids is often necessary.26 Prednisone in a dosage of 30 to 80 mg/day (depending on the severity of involvement) should be prescribed initially. This should be tapered over at least 10 to 14 days, and 21 days for poison ivy. The long, slow taper is needed to prevent rebound of the disease. The treatment may be discontinued when a daily dose of 10 mg is reached. Systemic antihistamines, such as hydroxyzine or diphenhydramine, may help control pruritus.9,26,76 The patient should also be counseled to wash all clothes that might have contacted the plant because the irritant plant oil can persist. Once the offending agent is reliably removed from the skin and clothes, ongoing outbreak is attributable to the initial contact, not spread from the serous fluid from the bullae. The patient is not contagious to others unless there is direct contact with the plant oil in people who are sensitized.
Erythema Multiforme Principles of Disease The most common precipitating factors in erythema multiforme are exposure to drugs and HSV infection. Other causes include other viral infections, especially hepatitis and influenza A. Less common causes include fungal diseases, such as dermatophytosis, histoplasmosis, and coccidioidomycosis, and bacterial infections, especially streptococcal infections and tuberculosis. Various collagen vascular disorders have been known to precipitate erythema multiforme, particularly rheumatoid arthritis, systemic lupus erythematosus, dermatomyositis, and periarteritis nodosa. Pregnancy and various malignancies have also been associated with erythema multiforme. No provocative factor can be identified in approximately half of all cases. Differential diagnosis includes urticaria, scalded skin syndrome, pemphigus, and pemphigoid and viral exanthems.
Clinical Features Erythema multiforme is an acute, usually self-limited disease precipitated by a variety of factors. It is characterized by the
Figure 118-12. Typical linear lesions of contact dermatitis secondary to poison ivy. (Courtesy of David Effron, MD.)
Chapter 118 / Dermatologic Presentations
pounds, 600,000 units of benzathine is used; in patients weighing 61 to 90 pounds, 900,000 units of benzathine is used; and in those weighing more than 90 pounds, 1.2 million units of benzathine is used. In patients allergic to penicillin, 250 mg of erythromycin four times a day or 40 mg/kg/day should be given orally for 10 days. Other macrolides and certain other cephalosporins may also be used.
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1544
sudden appearance of skin lesions that are erythematous or violaceous macules, papules, vesicles, or bullae. Their distribution is often symmetrical, most commonly involving the soles and palms, the backs of the hands or feet, and the extensor surfaces of the extremities. The presence of lesions of the palms and soles is particularly characteristic.61 The target lesion with three zones of color is the hallmark of erythema multiforme. It is a central, dark papule or vesicle that is surrounded by a pale zone, a halo of erythema (Fig. 118-13), and is commonly found on the hands or wrist. Stevens-Johnson syndrome, a severe form of erythema multiforme, is occasionally fatal. It is characterized by bullae, mucous membrane lesions, and multisystem involvement (Fig. 118-14). The patient may be toxic; complaining of chills, headache, and malaise; and displaying fever, tachycardia, and tachypnea. Systemic involvement may occur, with renal, GI, or respiratory tract lesions, resulting in hematuria, diarrhea, bronchitis, or pneumonia. Purulent conjunctivitis may be severe enough to cause the eyes to swell shut. Death results from infection and dehydration.
Management Treatment should begin with a search for the underlying cause. Mild forms resolve spontaneously in 2 or 3 weeks. Severe cases may last up to 6 weeks and may require hospital admission for IV hydration, local skin care, systemic analgesia, and systemic corticosteroid therapy, which should consist of 80 to 120 mg of prednisone daily in divided doses. Bullous
lesions should be treated with the application of wet compresses soaked in a 1 : 16,000 solution of potassium permanganate or a 0.05% silver nitrate solution several times a day. The major complications of Stevens-Johnson syndrome are infection and fluid loss. Renal involvement and pneumonia are rare. Severe conjunctivitis may result in corneal scarring and blindness. Reported mortality rates for Stevens-Johnson syndrome range from 0 to 15%.3,15
Pediculosis Clinical Features The diagnosis is made by identification of nits or adult lice on microscopic examination of plucked hairs from the symptomatic area. Nits are relatively more common than the adult louse form. Nits attach to the bases of hair shafts and appear as white dots (Fig. 118-15). Adult forms look like blue or black grains. The patient complains of intense itching and scratching. A secondary infection may result from the latter. The organisms causing pediculosis corporis reside in the seams of clothing and bedding materials while they feed on the human host. Except for heavily infested individuals, the parasites are absent from the body. Erythematous macules or wheals may be present, along with intense pruritus. The treatment consists of laundering or boiling clothing and bed linen. If nits are found in the body hair, a treatment with lindane lotion may be instituted, but this is not necessary is most cases (Figs. 118-16 and 118-17). Pediculosis capitis is seen more commonly in small children than in adults. Pruritus is the major symptom and may be confined to the occipital or postauricular scalp. Excoriations commonly result in secondary bacterial infections and regional lymphadenopathy.
Diagnostic Strategies The diagnosis is made by the identification of nits cemented to hairs at the hair-scalp junction (see Fig. 118-15).
Management
Figure 118-13. Erythema multiforme. (Courtesy of David Effron, MD.)
Figure 118-14. Stevens-Johnson syndrome. (Courtesy of David Effron, MD.)
Lindane (Kwell) lotion or cream is no longer the preferred prescription topical treatment.77 Permethrin (Nix) is the recommended treatment. It remains active for 2 weeks. Creme rinses and conditioning shampoos should not be used during this period because they coat the hairs and protect the lice from the insecticide. Permethrin is applied to the scalp after
Figure 118-15. Nits as seen in head lice. (Courtesy of David Effron, MD.)
1545
Figure 118-16. Body lice. (Courtesy of David Effron, MD.)
household members need not undergo a course of therapy. Underclothing, pajamas, and sheets and pillowcases should be machine washed (hot water) and dried, laundered and ironed, or boiled. Pruritus that persists after the course of therapy may result from an irritation of the skin by the pediculicide, sensitization, or patient anxiety. Permethrin is used to treat pediculosis capitis. A single dose of oral ivermectin, 200 µg/kg repeated in 10 days, has been shown to eradicate head lice.77 Lindane should be reserved for treatment failures. Household contacts should be examined for involvement, but uninfected persons need not be treated.
Scabies Clinical Features
Figure 118-17. Body lice. (Courtesy of David Effron, MD.)
the hair is shampooed and dried. It is rinsed out with water after 10 minutes. It must be applied when the hair is dry because lice can close down their respiratory airways for up to 30 minutes when immersed in water.77 Higher cure rates are achieved if the dose is repeated 1 week after the initial usage. Because the condition may be spread by sexual contact, sexual partners should also be treated. Other uninfested
Scabies is a mite infestation characterized by severe itching, which usually worsens at night. The areas of the body most commonly involved are the interdigital web spaces, flexion areas of the wrists, axillae, buttocks, lower back, penis, scrotum, and breasts (Fig. 118-18). The infestation tends to be more generalized in infants and children than in adults. The typical lesions are reddish papules or vesicles surrounded by an erythematous border and scratch marks. Scabies in infants and young children often have generalized skin involvement, including the face, scalp, palms, and soles. In infants, the most common presenting lesions are papules and vesiculopustules.78 Nodular scabies is a clinical variant in which extremely pruritic nodules are present on the male genitalia, buttocks, groin, and axillary regions. The nodules are reddish to brown, do not contain mites, and are thought to represent hypersensitivity reactions. They can persist for weeks despite adequate scabicidal treatment. Immunosuppressed patients may develop Norwegian scabies, which is manifested by extensive hyperkeratosis and crusting of the hands, feet, and scalp. It is highly contagious because of excessive mite proliferation.79,80 Secondary infections of these lesions are common. Close personal contact is involved in transmission of scabies. Multiple family members are likely to become infested. The infestation is also transmitted with sexual contact.
Management Treatment options include crotamiton (Eurax) lotion and cream or permethrin 5% cream (Elimite) and ivermectin. Lindane is no longer the preferred treatment. Patients in
Chapter 118 / Dermatologic Presentations
Figure 118-18. Scabies. (Courtesy of David Effron, MD.)
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1546
whom the former treatment fails may respond to the latter. Permethrin 5% cream (Elimite) applied overnight once weekly for 2 weeks over the entire body is the treatment of choice for infants and small children. It is more effective than crotamiton (Eurax) in eliminating the mite, in reducing secondary bacterial infection, and in reducing pruritus. Postscabietic nodules and pruritus may persist for months, even after successful treatment.62,63,66,67 Treatment of Norwegian scabies may require repeated treatment with scabicides and sometimes sequential use of several agents. A single dose of oral ivermectin, 200 µg/kg, may also be used.81 A second dose given 1 week later has been demonstrated to substantially improve the cure rate. Patients with crusted scabies may require repeat doses of ivermectin (200 µg/kg) along with topical scabicides (full-body application, repeated initially every few days) and keratolytics. The full benefit of ivermectin becomes evident when eradication of scabies in epidemic or endemic situations is needed since ivermectin leads to reliable disease control.82 The safety of ivermectin has been documented in millions of people with microfilarial diseases. Although ivermectin does not normally penetrate the blood-brain barrier and there should be no risk of seizures, neurotoxicity has been reported in the elderly.83 Because of limited safety data, ivermectin should not be used in children younger than 5 years or during pregnancy or lactation. The essential oil of the tea tree (Melaleuca alternifolia) and the essential oil of the Lippia multiflora Moldenke have also been noted to have scabicidal and antibacterial activity, although the dosing schedule of tea tree oil has not been established.84 All family members and sexual contacts should also be treated. Intimate articles of clothing and sheets and pillowcases should be washed and dried by machine (hot water), laundered and ironed, or boiled. It may take several weeks after therapy for the signs and symptoms to abate. A hypersensitivity state or anxiety may prolong symptoms long after the mites have been destroyed.
Figure 118-19. Secondary syphilis. (Courtesy of David Effron, MD.)
Figure 118-20. Cutaneous manifestation of secondary syphilis on the soles of the feet. (Courtesy of David Effron, MD.)
Syphilis Clinical Features Syphilis is transmitted only by direct contact with an infectious lesion. The causative organism is the spirochete Treponema pallidum. After an incubation period of 10 to 90 days, the primary lesion appears, which lasts from 3 to 12 weeks and heals spontaneously. In 6 weeks to 6 months after exposure, the disease enters the secondary stage, which involves a variety of mucocutaneous lesions. These lesions also heal spontaneously in 2 to 6 weeks as the disease enters the latent phase. Either a prolonged latent phase or tertiary syphilis follows. Of untreated patients, 25% display at least one relapse of mucocutaneous lesions of the oral cavity or anogenital region. The chancre is the dermatologic manifestation of primary syphilis. Chancres usually appear as single lesions but may be multiple. They appear at the site of spirochete inoculation, usually the mucous membranes of the mouth or genitalia. The chancre begins as a papule and characteristically develops into an ulcer approximately 1 cm in diameter with a clean base and raised borders. The chancre is painless unless secondarily infected, and it may be accompanied by painless lymphadenopathy. The secondary stage usually follows the primary stage by 6 weeks or more but rarely overlaps primary syphilis. There are a number of cutaneous manifestations of secondary syphilis. Lesions may be erythematous or pink macules or papules,
Figure 118-21. Cutaneous manifestation of secondary syphilis on the palms of the hands. (Courtesy of David Effron, MD.)
usually with a generalized symmetrical distribution (Fig. 11819). Pigmented macules and papules classically appear on the palms and soles (Figs. 118-20 and 118-21). The lesions may be scaly but are rarely pruritic. Papular, annular, and circinate lesions are more common in people of color. Generalized lymphadenopathy and malaise accompany the skin lesions. Irregular, patchy alopecia may be seen. Moist, flat, verrucous condyloma latum may appear in the genital area. These lesions are highly contagious.
1547
Diagnostic Strategies
Management Incubating syphilis, the stage before the appearance of primary lesions, may be treated with 4.8 million units of procaine penicillin IM after 1 g of probenecid orally. Primary and secondary syphilis is treated with benzathine penicillin G in a dose of 2.4 million units IM. Patients allergic to penicillin should be treated for 14 days with doxycycline, 100 mg twice a day, tetracycline, 500 mg four times a day, or erythromycin, 500 mg four times a day.52 HIV-infected patients require more intensive therapy. Treatment may be administered in the ED if the diagnosis can be made on clinical, microscopic, or serologic grounds. If this cannot be done, a serologic sample should be drawn and the patient referred for treatment. The VDRL test may be expected to return to nonreactive 6 to 12 months after the treatment of primary disease or 1 to 1 1 2 years after the treatment of secondary disease. Patients with tertiary syphilis who are adequately treated may nevertheless retain a positive serologic result. Within 12 hours of receiving therapy, patients may experience a febrile reaction and diffuse rash called the JarischHerxheimer reaction. The reaction resolves spontaneously, usually within 24 hours.
Figure 118-22. Erythema nodosum. (Courtesy of David Effron, MD.)
the lesions evolve, they may turn yellow-purple and resemble bruises (Fig. 118-22). Women are affected three times more often than men, with the highest incidence in the third to fifth decades of life.68 A number of underlying conditions produce erythema nodosum: tuberculosis, sarcoidosis, coccidioidomycosis, histoplasmosis, ulcerative colitis, regional enteritis, pregnancy, infections with streptococci, Yersinia enterocolitica, and Chlamydia. As with erythema multiforme, many cases of erythema nodosum are idiopathic. The relationship of drugs to erythema nodosum was noted in the section on drug eruption. Oral contraceptive agents are a leading cause of drug-induced cases. The differential diagnosis includes traumatic bruises and subcutaneous fat necrosis.
Management When an underlying condition can be determined, this should be treated as indicated. Chest radiograph may be considered to rule out sarcoidosis, tuberculosis, or deep fungal infection. Bed rest, elevating the legs, and wearing elastic stockings reduce pain and edema. Aspirin in a dosage of 600 mg every 4 hours or nonsteroidal anti-inflammatory agents may also afford some relief.9,85 Erythema nodosum is a self-limited process that usually resolves in 3 to 8 weeks.9 Patients with severe pain may be treated with 360 to 900 mg of potassium iodide daily for 3 or 4 weeks. Stopping therapy before this time may result in a relapse. Potassium iodide may act through an immunosuppressive mechanism mediated via heparin release from mast cells.9,85
■ VESICULAR LESIONS Perspective
■ NODULAR LESIONS Erythema Nodosum Clinical Features Erythema nodosum is an inflammatory reaction of the dermis and adipose tissue that is seen with painful red to violet nodules. Nodules are elevated lesions located deep in the skin, and the skin over the nodules can be moved by palpation. These painful nodules occur most commonly over the anterior tibia but may also be seen on the arms or body. Fever and arthralgia of the ankles and knees may precede the rash.3,9 As
Vesicles are elevated lesions that contain clear fluid. Vesicles greater than 1 cm are known as bullae. Vesicles may sometimes be associated with red papular lesions, as in contact dermatitis or erythema multiforme.
Pemphigus Vulgaris Clinical Features Pemphigus vulgaris is an uncommon, but important, dermatologic disorder. The mortality rate before the use of steroids was approximately 95%. The current mortality rate is 10 to
Chapter 118 / Dermatologic Presentations
The diagnosis of primary syphilis is made primarily by the identification of spirochetes with darkfield microscopy. Because a darkfield microscope is often not available to the emergency physician, the diagnosis of primary syphilis must be suspected on clinical grounds and the patient referred to a dermatologist or appropriate public agency for diagnosis and treatment. The Venereal Disease Research Laboratory (VDRL) test, the most commonly used diagnostic serologic test, is positive in approximately three fourths of patients with primary syphilis, but the test tends to be negative early in the course of the disease.53 The VDRL test is invariably positive in cases of secondary syphilis, usually in titers of 1 : 16 or greater. The darkfield examination of moist lesions may also be positive, but the diagnosis in this stage is based on a positive serologic test. The most specific and sensitive serologic test is the fluorescent treponemal antibody absorption (FTA-ABS) test.53 A biologic false-positive serologic test for syphilis is defined as a positive VDRL test with a negative FTA-ABS test. This situation is seen acutely after vaccination or infections, especially mycoplasmal pneumonia, mononucleosis, hepatitis, measles, varicella, and malaria, and in pregnancy. Chronic biologic false-positive reactions (i.e., those lasting longer than 6 months) may occur with systemic lupus erythematosus, thyroiditis, lymphoma, narcotic addiction, or in elderly patients. Most false-positive reactions are in low titer ranges of 1 : 1 to 1 : 4.
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1548
15%, related more to steroid-induced complications than to the disease. Pemphigus is a bullous disease, affecting both sexes equally, and is most common in patients 40 to 60 years old.86 The disease is mostly prevalent in people of Jewish, Mediterranean, or south Asian descent.86 The typical skin lesions are small, flaccid bullae that break easily, forming superficial erosions and crusted ulcerations. Any area of the body may be involved. Nikolsky’s sign is present and characteristic of the disease. Blisters may be extended or new bullae may be formed by firm tangential pressure of a finger on the intact epidermis. Before the appearance of the skin involvement, mucous membrane lesions occur; 50 to 60% of patients have oral lesions. The oral lesions typically antedate the cutaneous lesions by several months.9,86 The most common site is in the mouth, especially the gums and vermilion borders of the lips. Oral lesions are bullous but commonly break, leaving painful, denuded areas of superficial ulceration. The cause of pemphigus is unknown, although studies suggest an autoimmune mechanism. The development of pemphigus has been associated in a few instances with the use of medications, most notably penicillamine and captopril.9 A positive Tzanck cytologic test suggests the diagnosis (i.e., finding acantholytic cells or degenerated, rounded epithelial cells with amorphous nuclei). Acantholytic cells are not specific for pemphigus, however, and the diagnosis must be confirmed by serum immunofluorescence. The differential diagnosis includes bullous pemphigoid, epidermolysis, dermatitis herpetiformis, toxic epidermal necrolysis, bullous scabies, and bullous systemic lupus erythematosus (Fig. 118-23).9,86
steroids, and the mortality rate continues to be substantial.9 Deaths are related to an uncontrolled spread of the disease, secondary infection, dehydration, and thromboembolism. Other medical illnesses, as well as the side effects of highdosage corticosteroids, also contribute to mortality.
Herpes Simplex Perspective Two known variants of HSV cause human infection: HSV-1 and HSV-2. The former primarily affects nongenital sites, whereas lesions caused by the latter are found predominantly in the genital area and are transmitted primarily by venereal contact.
Clinical Features The hallmark of skin infection with HSV is painful, grouped vesicles on an erythematous base. Those above the waist are usually caused by HSV-1, whereas those below the waist generally result from HSV-2 (Figs. 118-24 and 118-25). The lesions are usually localized in a nondermatomal distribution. The skin distribution may become more generalized in patients with atopic eczema and other dermatoses. Adults with HSV infection should avoid contact with children with atopic dermatitis, especially in the first 3 to 5 days of infection.
Management Pain control and local wound care are essential components of therapy. Once the diagnosis is made, treatment with oral glucocorticoids in initial doses of 100 to 300 mg of prednisone, or an equivalent drug, should be instituted in conjunction with a dermatologist. Other immunosuppressant drugs may also be used. Despite the condition’s localization to the skin and mucous membranes, death was the rule before treatment with
Figure 118-24. RSV-1 infection. (Courtesy of David Effron, MD.)
Figure 118-23. Bullous pemphigus. (Courtesy of David Effron, MD.)
Figure 118-25. Herpetic whitlow. (Courtesy of David Effron, MD.)
1549
Figure 118-26. Chickenpox. (Courtesy of David Effron, MD.)
Management Recommended treatment for a first clinical episode of genital herpes is with acyclovir (Zovirax), 200 mg orally five times a day for 7 to 10 days, famciclovir, 125 mg twice a day, or valacyclovir, 500 mg three times a day or until clinical resolution occurs. These agents reduce the duration of viral shedding, accelerate healing, and shorten the duration of symptoms, but they have not succeeded in preventing recurrent episodes.9 Prophylactic administration of acyclovir may be effective in ameliorating the severity of recurrent genital herpes, but the effects of long-term administration are unknown.9 Although many episodes of recurrent herpes infection do not benefit from acyclovir therapy, 200 mg five times a day may be given orally for recurrences at the beginning of the prodrome. Famciclovir, 125 mg twice a day for 5 days, and valacyclovir, 500 mg three times a day for the same duration, are equally effective.9 Severe initial attacks of genital herpes have been successfully treated with the IV infusion of acyclovir. Admission to the hospital is required, however, because such treatment is necessary for several days, especially for the immunocompromised patient. A mucocutaneous herpes infection in such patients is potentially fatal because it has a propensity for generalization and dissemination to the internal organs. Supportive care is important and pain control is a major concern. Systemic analgesics and topical anesthetic agents may be useful. Patient education regarding the prevention or spread of the disease during sexual contact and the birth process is imperative.
Varicella Clinical Features Varicella, or chickenpox, is an infection caused by the varicella-zoster virus. After an incubation period of 14 to 21 days, the illness begins with a low-grade fever, headache, and malaise. The exanthem coincides with these symptoms in children and follows them by 1 or 2 days in adults. The skin lesions rapidly progress from macules to papules to vesicles to crusting, sometimes within 6 to 8 hours. The vesicle of varicella is 2 or 3 mm in diameter and surrounded
Figure 118-27. Bullous chickenpox. (Courtesy of David Effron, MD.) by an erythematous border (Fig. 118-26). An unusual form of varicella presents with larger bullae (Fig. 118-27). The drying of the vesicle begins centrally, producing umbilication. The dried scabs fall off in 5 to 20 days. Lesions appear in crops on the trunk, where they are seen in the highest concentration, and on the scalp, face, and extremities. The hallmark of varicella is the appearance of lesions in all stages of development in one region of the body. Extensive eruptions are often associated with a high and prolonged fever. Complications of chickenpox include encephalitis or meningitis, pneumonia, staphylococcal or streptococcal cellulitis, thrombocytopenia, arthritis, hepatitis, and glomerulonephritis.73 Varicella pneumonia occurs more commonly in adults than in children.
Management The illness is self-limited, and treatment is symptomatic only. Salicylates should be avoided in patients with chickenpox to minimize the risk of subsequent Reye’s syndrome. Oral acyclovir may be effective if it can be started within 24 hours of development of rash for patients with chronic respiratory or skin disease. Some studies report a diminution in duration and magnitude of fever and number and duration of lesions with the early use of acyclovir.74 Isolation of infected patients is often futile because the disease may be transmitted before the diagnosis is clinically evident. Because the disease has the potential to be contagious until all vesicles are crusted and dried, infected persons should be kept at home until this stage is reached.
Chapter 118 / Dermatologic Presentations
The mouth is the most common site of HSV-1 infections. Children are affected more commonly than adults.9 Small clusters of vesicles appear but are soon broken, leaving irregularly shaped, crusted erosions. The severity of gingivostomatitis varies from the presence of small ulcers to extensive ulceration of the mouth, tongue, and gums accompanied by fever and cervical lymphadenopathy. The infection may be so severe that oral fluid intake is difficult, and dehydration may result. Healing typically occurs in 7 to 14 days, unless a secondary infection with streptococci or staphylococci occurs. HSV-2 infections in men are seen with either single or multiple vesicles on the shaft or glans penis. Fever, malaise, and regional adenopathy may be present.53 A prodrome of local pain and hyperesthesia may precede the appearance of the cutaneous lesions. The vesicles erode after several days, become crusted, and heal in 10 to 14 days. Infections in women involve the introitus, cervix, or vagina. Vesicles may be grouped or confluent. Herpetic cervicitis or vaginitis may be the cause of severe pelvic pain, dysuria, or vaginal discharge.9,53 Recurrence is common, but recurrent episodes tend to be less severe. A correlation based on serologic and epidemiologic data has been discovered between HSV-2 reproductive tract infections and carcinoma of the cervix.9,53
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
1550
Varicella-zoster and varicella titers should be checked in pregnant women and immunocompromised patients who are exposed to chickenpox, and if negative, varicella-zoster immune globulin should be administered within 96 hours of exposure.87 Fetal infection after maternal varicella in the first or early second trimester of pregnancy may result in varicella embryopathy, a condition characterized by limb atrophy, scarring on extremities, and CNS and ocular manifestations.34 Maternal varicella that occurs between 5 days before delivery and 2 days after delivery may result in disseminated herpes in the newborn.34 The varicella vaccine is a live attenuated virus; it is highly efficacious and very safe.88 A single dose is effective in children between the ages of 1 and 13 years. For older children, two doses separated by 4 to 8 weeks is recommended.88 In addition, the incidence of zoster occurring after vaccination appears to be lower than after naturally acquired disease.88
Herpes Zoster Clinical Features
latent period between the two illnesses, the virus is thought to reside in dorsal root ganglion cells.9,89 Herpes zoster has a very low mortality rate and is rarely life threatening, even when dissemination to the visceral organs occurs. Complications include CNS involvement, ocular infection, and neuralgia. Meningoencephalitis, myelitis, and peripheral neuropathy have been reported. Ocular complications occur in 20 to 70% of cases involving the ophthalmic division of the trigeminal nerve. The severity varies from mild conjunctivitis to panophthalmitis, which threatens the eye.89 Eye involvement produces anterior uveitis, secondary glaucoma, and corneal scarring. There is a close correlation between eye involvement and vesicles located at the tip of the nose. Postherpetic neuralgia, pain that persists after the lesions have healed, occurs more commonly in elderly and immunosuppressed patients.89 It may last a number of months and is often resistant to treatment with standard analgesic medications. Herpes zoster generally tends to be more severe in immunosuppressed patients, especially those with AIDS, Hodgkin’s disease, or other lymphomas.89 Cutaneous dissemination occurs more commonly in these patients than in the general population. Visceral and CNS dissemination is also more likely to occur in these patients; therefore, they should be considered for hospitalization.
Herpes zoster, or shingles, is an infection caused by the varicella zoster virus. It occurs exclusively in individuals who have previously had chickenpox. Before the rash appears, the patient typically develops pain in a dermatomal distribution. This pain precedes the eruption by 1 to 10 days; is of variable intensity; and is described as sharp, dull, or burning in quality. The rash consists of grouped vesicles on an erythematous base involving one or several dermatomes. The thorax is involved in most cases, and the trigeminal distribution is the next most commonly involved region.89 The vesicles initially appear clear and then become cloudy and progress to scab and crust formation. This process takes 10 to 12 days, and the crusts fall off in 2 or 3 weeks (Figs. 118-28 and 118-29). Herpes zoster has a peak incidence in patients 50 to 70 years old and is unusual in children. Although the association with leukemia, Hodgkin’s lymphoma, and other malignancies is well known, rarely does the appearance antedate the diagnosis of such diseases. Most cases of herpes zoster occur in healthy individuals.89 Herpes zoster may be transmitted from patients with chickenpox to susceptible individuals. Chickenpox may also be acquired by contact with shingles, although this is less common.89 It is generally believed, however, that herpes zoster is caused by a reactivation of latent varicella-zoster virus present since the initial infection with chickenpox. During the
Treatment other than analgesia is rarely necessary. Burow’s solution compresses diluted 1 : 20 to 1 : 40 in water may be applied to hasten drying. Early systemic corticosteroid therapy may shorten the duration of postherpetic neuralgia but does not lessen the severity of pain or the rate of the healing of the lesions.90 Antiviral chemotherapy, with acyclovir, famciclovir, vidarabine, foscarnet, valacyclovir, and interferon-α, has been shown to be effective for immunocompromised patients.89 Postherpetic neuralgia is a complicated problem with few satisfactory solutions. Some success has been achieved using capsaicin cream, but this cannot be applied to inflamed or eroded skin.9 Intravenous acyclovir may be of some benefit in the treatment of severe ocular herpes zoster. Treatment includes mydriasis and the application of topical corticosteroids. Unlike the situation with herpes simplex conjunctivitis, eye involvement caused by herpes zoster does not appear to be exacerbated by corticosteroids.
Figure 118-28. Herpes zoster. (Courtesy of David Effron, MD.)
Figure 118-29. Herpes zoster infection. (Courtesy of David Effron, MD.)
Management
1551
Table 118-3 Differentiation of Chickenpox from Smallpox SMALLPOX
Prodromal signs/symptoms
Prodromal signs/ symptoms absent or mild
Illness severity
Illness usually not severe unless complications/ immunosuppressed Superficial vesicles developing rapidly (1 day) and in multiple stages in each affected area
1–4 days of systemic signs/ symptoms before onset of rash Very ill from onset, may be toxic
Lesion development
Lesion locations
Commonly on face and trunk, not palms and soles
Contagiousness
Contagious until all lesions crusted over
Hard, circumscribed pustules developing slowly (over days); lesions in same stage in every affected area Commonly on face and extremities, including palms and soles Contagious until all scabs have fallen off
Chapter 118 / Dermatologic Presentations
CHICKENPOX
Figure 118-30. Smallpox. (From the Centers for Disease Control and Prevention Public Health Image Library [http://phil.cdc.gov].)
Smallpox The last naturally occurring case of smallpox was in Somalia in 1977. Subsequently, the routine vaccination of the general public was stopped. Except for laboratory stockpiles, the variola virus had been eliminated.82 Due to recent concerns regarding biological agents as weapons, it is important that smallpox be differentiated from chickenpox (Table 118-3; Fig. 118-30).91
Cutaneous Anthrax Cutaneous anthrax begins as a pruritic pustule or vesicle that enlarges and erodes over 1 or 2 days. Subsequently, a necrotic ulcer with central black eschar is formed.92 The lesion may be painless and may be surrounded by significant edema (Fig. 118-31).
Figure 118-31. Cutaneous anthrax. (From the Centers for Disease Control and Prevention Public Health Image Library [http://phil.cdc.gov].)
■ SKIN LESIONS ASSOCIATED WITH SYSTEMIC DISEASE Numerous systemic illnesses have cutaneous manifestations (Table 118-4; Figs. 118-32 to 118-39). Some of the most common illnesses include AIDS, diabetes mellitus, connective tissue diseases, and endocrine disorders.
■ CLINICAL FEATURES OF LESIONS ASSOCIATED WITH INTERNAL MALIGNANCY Cutaneous lesions most directly indicative of an internal malignancy arise from the extension of the tumor to the skin or by hematogenous or lymphatic metastasis. The neoplasms that most commonly produce such a cutaneous extension are lymphomas, leukemias, and carcinomas of the breast, GI tract, lung, ovary, prostate, uterus, and bladder. Skin metastases generally signify a poor prognosis.93
Figure 118-32. Kaposi’s sarcoma associated with AIDS. (Courtesy of David Effron, MD.)
1552
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
Table 118-4 Skin Lesions Associated with Systemic Disease DISEASE
LESIONS
AIDS3,9,87
Chronic ulcerative herpes simplex Kaposi’s sarcoma (Figs. 118-32 and 118-33) Severe herpes zoster Oral hairy leukoplakia Genital warts Molluscum contagiosum (Fig. 118-34) Seborrheic dermatitis 2 Pityrosporum Recurrent staphylococcal abscesses Mycobacterial papules, nodules, abscesses Oral and rectal squamous cell carcinoma Lymphoma Severe psoriasis Acquired ichthyosis Folliculitis Human papillomavirus infection Lichenoid photoeruptions Diabetic dermopathy Necrobiosis lipoidica diabeticorum Cellulitis (Fig. 118-35) Vascular ulceration (Fig. 118-36) Acanthosis nigricans Bullosis diabeticorum Diabetic thick skin Scleroderma Heliotrope discoloration and edema of eyelids Scaly erythema of malar prominences
Diabetes mellitus85
Dermatomyositis
Systemic lupus erythematosus
Rheumatoid arthritis
Hyperthyroidism85
Hypothyroidism85
Ulcerative colitis93
Erythematous dermatitis over joint extensor surfaces, especially hands (Fig. 118-37) Raynaud’s phenomenon Discoid lesions Malar erythema (Fig. 118-38) Hypertrophic or verrucous palm and sole lesions Lupus panniculitis Oral ulcers Raynaud’s phenomenon Rheumatoid nodules and necrobiosis Vasculitic lesions Pyoderma gangrenosum Urticaria Fine, velvety, smooth skin Increased sweating Hyperpigmentation or hypopigmentation Pretibial edema Alopecia Onychosis Urticaria Dry, coarse skin Myxedema (Fig. 118-39) Carotene color Pruritus Atopic dermatitis Ichthyosis Erythema nodosum Easy bruising Alopecia (lateral third of eyebrows) Pyoderma gangrenosum Erythema nodosum Aphthous stomatitis
COMMENTS
Diagnostic for AIDS
Most common Most characteristic Control of diabetes does not affect presence
Skin lesions may precede muscle disease Symmetrical proximal weakness, remissions, exacerbations Increased creatine phosphokinase aldolase with active disease Patients with cutaneous discoid lupus generally have benign diseases
Still’s disease
Associated with state of disease
1553
Effron, MD.)
Figure 118-36. Vascular ulceration secondary to diabetes. (Courtesy of David Effron, MD.)
Figure 118-34. Molluscum contagiosum caused by a virus is more
Figure 118-37. Erythematous dermatitis over the joint extensor surfaces,
prevalent with AIDS. (Courtesy of David Effron, MD.)
dermatomyositis. (Courtesy of David Effron, MD.)
Figure 118-35. Gangrene of the toe with cellulitis in a diabetic patient.
Figure 118-38. Malar erythema in a patient with systemic lupus
(Courtesy of David Effron, MD.)
erythematosus. (Courtesy of David Effron, MD.)
Acanthosis Nigricans
The lesion appears as a hyperpigmented verrucous, velvetlike hyperplasia and hypertrophy of the skin accompanied with accentuation of the skin markings. The chief sites of involvement are the body folds, especially the axillae, antecubital fossae, neck, and groin. More than 90% of cases of “malignant” acanthosis nigricans are associated with intra-abdominal malignancies, of which two thirds are adenocarcinomas of the stomach.93 Carcinomas of the colon, ovary, pancreas, rectum, and uterus make up the majority of the rest.93 Regardless of the tumor type, acanthosis
Acanthosis nigricans is associated with internal malignancy, despite the fact that most patients do not have tumors.93 Benign cases may be familial or related to endocrine disease or obesity. The term malignant acanthosis nigricans is used to designate the form associated with neoplastic disease. This phrasing is misleading because acanthosis nigricans is only a marker of the underlying disease and is never infiltrated with malignant cells.
Chapter 118 / Dermatologic Presentations
Figure 118-33. Kaposi’s sarcoma in an AIDS patient. (Courtesy of David
1554
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
BOX 118-1 Causes of Purpura Thrombocytopenic Aplastic anemia Drug induced Idiopathic Malignant disease Sarcoidosis Splenomegaly Systemic lupus erythematosus Thrombotic Tuberculosis
Figure 118-39. Severe myxedema in a hypothyroid patient. (Courtesy of David Effron, MD.)
nigricans is associated with tumors that are usually highly malignant and metastasize early.93 The mechanism of this dermatosis in cases of internal malignant disease is postulated to be a result of tumor products that bind to and stimulate insulinlike growth factors in the skin.77
Dermatomyositis The incidence of dermatomyositis with malignant disease ranges from 6 to 55% and is generally higher in older patients. In younger individuals, the appearance of dermatomyositis does not necessarily call for a tumor workup. Tumors commonly associated with dermatomyositis are carcinomas of the breast, ovary, and GI and female genital tracts. Polymyositis occurring alone without the accompanying skin findings is rarely associated with malignancies.85
Erythema Multiforme Erythema multiforme may be associated with acute forms of leukemia. It is seen with acute monocytic, lymphocytic, and granulocytic forms and is also found in chronic leukemias and Hodgkin’s disease.9,94
Erythema Nodosum Erythema nodosum is another reaction found in association with leukemia and Hodgkin’s lymphoma, as well as with metastatic carcinoma and inflammatory bowel disease.94
Erythroderma Generalized erythroderma is almost pathognomonic for Hodgkin’s disease; however, it is also a common skin manifestation of lymphocytic leukemia. Although less common, it is also seen with other forms of leukemia, carcinoma, and mycosis fungoides. The appearance of erythroderma may precede the diagnosis of internal malignant disease by many years. The skin eruption is invariably accompanied by intractable pruritus.94
Acquired Ichthyosis Acquired ichthyosis is a skin condition manifested as generalized dryness of the skin, scaling, and superficial cracking or as hyperkeratosis of the palms and soles. Hodgkin’s disease is the most common malignant disease associated with the
Nonthrombocytopenic Drugs Infection (meningococcemia, Rocky Mountain spotted fever) Qualitative platelet defect Vasculitis
nonfamilial form of ichthyosis. Non-Hodgkin’s lymphoma and carcinomas of the breast, lung, colon, and cervix have also been associated with acquired ichthyosis.85
Pruritus Itching may be an important indicator of Hodgkin’s disease, leukemia, adenocarcinoma or squamous cell carcinoma of various organs, carcinoid syndrome, multiple myeloma, and polycythemia vera. It may appear years before the underlying malignancy is identified.85 In cases of Hodgkin’s disease, the itching is usually continuous and may be accompanied by a severe burning sensation. Although usually generalized, pruritus commonly begins in the feet and may be limited to the lower extremities. It may be intractable and associated with urticaria, erythroderma, excoriation, or lichenification. The pruritus of leukemia and systemic carcinoma is generally less severe than that found with Hodgkin’s disease. Nevertheless, itching associated with internal malignant disease may be difficult to control. Conventional anti-H1 antihistamines, cimetidine, cholestyramine, and cyproheptadine have each been used with variable results.94 Occasionally, only the suppression of the tumor is beneficial.
Purpura Purpura is the most common manifestation of acute granulocytic and monocytic leukemia. It may also be associated with myeloma, lymphoma, and polycythemia vera. Although the most common cause of purpura in these conditions is thrombocytopenia secondary to bone marrow infiltration, in some instances the platelet count is normal and the causative mechanism obscure.9 Purpura is caused by vascular abnormalities, thrombocytopenia, or other coagulation defects. A variety of diseases and conditions may be the underlying cause, and the treatment should be directed toward this cause whenever possible (Boxes 118-1 and 118-2).94,95 Thrombocytopenic and nonthrombocytopenic forms are differentiated by the results of the patient’s platelet count. Serious bleeding seldom occurs if the platelet count is greater than 50,000/mm3. If the platelet count is less than 10,000/mm3 or serious bleeding is encountered, platelet transfusion should be initiated. Because of the short circulating half-life of infused platelets, transfusion should be used as a short-term measure only.
1555
BOX 118-2 Commonly Used Drugs Associated with Purpura
CAUSE
COMMON RESPONSIBLE FACTORS
Bacterial infection
Streptococcus Staphylococcus Yersinia Mycobacterium Herpes simplex virus Epstein-Barr virus Cytomegalovirus Hepatitis viruses (especially B) Many acute viral syndromes (adenovirus, enterovirus) Parasites Coccidioidomycosis Histoplasmosis Rickettsia Spirochete (Lyme disease) Bees Wasps Scorpions Spiders Jellyfish Fleas Mites Penicillin Sulfa Cephalosporins Salicylates Morphine, codeine, other opioids Nonsteroidal anti-inflammatory drugs Barbiturates Amphetamines Blood and blood products Nuts Shellfish Eggs Strawberries Tomatoes Milk, cheese Chocolate Chemicals Cosmetics Topical medications Plants Textiles Foods Dust Pollen Animal dander Chemicals/aerosols Mold spores Heat Cold Light Pressure (dermatographism) Water Collagen vascular disease Lupus, juvenile rheumatoid arthritis, polyarteritis nodosa, dermatomyositis, Sjögren’s syndrome, rheumatic fever Inflammatory bowel disease Crohn’s disease, ulcerative colitis Malignancy Carcinoma, leukemia, lymphoma Miscellaneous Serum sickness, thyroiditis, aphthous stomatitis, Behçet’s disease
Viral infection
Other infections
Envenomation
Drugs
Foods
Contacts
Figure 118-40. Tracks secondary to intravenous heroin abuse. (Courtesy of David Effron, MD.)
Urticaria Urticaria is occasionally found in Hodgkin’s disease and more rarely in leukemia and internal carcinoma. Cold urticaria may occur with multiple myeloma (Table 118-5).
Inhalants
Physical agents
■ CLINICAL FEATURES OF LESIONS ASSOCIATED WITH NARCOTIC ADDICTION Individuals who inject opiates and other drugs parenterally develop characteristic skin lesions secondary to such use. Skin lesions have been most extensively described in heroin addicts. Skin tracks, or indurated linear hyperpigmented streaks, are produced by repeated IV injection (Fig. 118-40). They follow the course of the superficial veins used in the injection, most commonly in the antecubital fossae and the dorsa of the hands. Subcutaneous injection results in round or oval hyperpigmented atrophic depressed scars 1 to 3 cm in diameter (Fig. 118-41). Abscesses, which often require drainage, commonly precede the development of such scars. Hypertrophic scarring
Diseases
Data from Edwards L: Dermatology in Emergency Medicine. New York, Churchill Livingstone, 1997; and Westo WL, Badgett JT: Urticaria. Pediatr Rev 19:240, 1998.
Chapter 118 / Dermatologic Presentations
Amitriptyline Aspirin Cephalothin Chloramphenicol Chlorpromazine Chlorpropamide Diazoxide Digitoxin Furosemide Hydrochlorothiazide Indomethacin Isoniazid Meprobamate Methyldopa Penicillin Phenacetin Phenobarbital Phenylbutazone Quinidine Rifampin Sulfonamides Tolbutamide
Table 118-5 Common Causes of Urticaria
1556
PART III ■ Medicine and Surgery / Section Nine • Immunologic and Inflammatory
KEY CONCEPTS ■
Figure 118-41. Scars from subcutaneous illicit drug injection. (Courtesy of David Effron, MD.)
and keloid formation may also occur. Increased pigmentation may occur in sun-exposed areas and at the site of tourniquet applications. In addition to the characteristic skin lesions associated with drug injection, people who inject intravenous drugs are prone to sharp foreign body retention, pseudoaneurysm, gram-negative local and systemic infections, wound botulism (associated with the use of black tar heroin), and numerous other illnesses.
Infection with C. albicans can occur normally in infancy, in obese people, during pregnancy, and in old age. In other patients, the following underlying problems should be considered: AIDS and other immunodeficiency states, diabetes and other endocrine imbalances, malignancy, malnutrition, and other debilitating illnesses. ■ Rashes that are associated with mucosal lesions, blisters, or desquamating skin are often caused by significant soft tissue infections, drug eruptions, or immune disorders. ■ Purpura result from blood leaking from vessels into the skin and do not blanch when pressure is applied. Purpura less than 3 mm in diameter are called petechiae. Nonpalpable purpura are often caused by coagulation defects (usually platelet abnormalities), whereas palpable purpura are usually a sign of vasculitis. ■ Diffuse pruritus in the absence of a skin rash may be a sign of underlying malignancy.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
•
Section Ten Hematology and Oncology Chapter 119
Anemia, Polycythemia, and White Blood Cell Disorders
Timothy G. Janz and Glenn C. Hamilton
■ ANEMIA Definition Anemia is an absolute decrease in the number of circulating red blood cells (RBCs). The diagnosis is made when laboratory measurements fall below accepted normal values (Table 119-1). In emergency medicine, anemia may be divided into two broad categories: emergent, having immediate life-threatening complications, and nonemergent, with less imminent patient danger. Factors other than the absolute number of circulating RBCs may place the patient in one category or another (e.g., rate of onset and underlying hemodynamic reserve of the patient).1,2 Both groups necessitate a sound diagnostic approach, but emergent anemia may require supportive therapy concomitant with or in advance of the definitive diagnosis. Although patients with nonemergent anemia are usually referred to a specialist, they are seen in the hospital often enough to make an understanding of anemia necessary for emergency physicians. The urgency of consultation depends predominantly on the patient’s hemodynamic tolerance of the anemia.1,3
Pathophysiology The major function of the RBC is oxygen transport from the lung to the tissue and carbon dioxide transport in the reverse direction. Oxygen transport is influenced by the amount of hemoglobin, its oxygen affinity, and blood flow. An alteration in any of the major components usually results in compensatory changes in the other two. For example, a decrease in hemoglobin from anemia is compensated by both inotropic and chronotropic cardiac changes that result in increased blood flow and decreased hemoglobin affinity at the tissue level, thereby allowing more oxygen release. These compensatory responses may collapse because of disease severity or underlying pathologic conditions. The result is tissue hypoxia and eventual cell death.1,4 Anemia often stimulates the compensatory mechanism of erythropoiesis controlled by the hormone erythropoietin. Erythropoietin is a glycoprotein produced in the kidney (90%) and the liver (10%). It regulates the production of RBCs by controlling differentiation of the committed erythroid stem cell. It is stimulated by tissue hypoxia and products of RBC destruction during hemolysis. Erythropoietin levels are elevated in many types of anemia.5,6
Bone marrow contains pluripotent stem cells that can differentiate into erythroid, myeloid, megakaryocytic, and lymphoid progenitors. Erythropoietin enhances the growth and differentiation of erythroid progenitors. When the late normoblast extrudes its nucleus, it still contains a ribosomal network, which identifies the reticulocyte (Fig. 119-1). The reticulocyte retains its ribosomal network for approximately 4 days, 3 of which are spent in bone marrow and 1 in the peripheral circulation. The RBC matures as the reticulocyte loses its ribosomal network and circulates for 110 to 120 days. The erythrocyte is then removed by macrophages that detect senescent signals. Under steady-state conditions, the rate of RBC production equals the rate of destruction. RBC mass remains constant because an equal number of reticulocytes replace the destroyed, senescent erythrocytes during the same period.5 Common sites of blood loss in trauma include the pleural, peritoneal, pelvic, and retroperitoneal spaces. In nontraumatic circumstances, especially in patients on anticoagulants, the gastrointestinal tract, retroperitoneal space, uterus, and adnexa must be considered. Causes other than blood loss may be responsible for severe anemia of rapid onset. Certain rare hemolytic conditions can cause rapid intravascular destruction of RBCs (Box 119-1). More common are patients with chronic compensated hemolytic anemia (e.g., sickle cell disease), who decompensate with an acute-onset anemia as a result of decreased erythrocyte production triggered by a viral infection. Beyond red cell destruction, the status of hemoglobin function must be considered. Impaired hemoglobin transport of oxygen is seen in cases of carbon monoxide poisoning. Methemoglobinemia from nitrates, cyanhemoglobin from cyanide, and sulfhemoglobinemia resulting from hydrogen sulfide may severely decrease functional hemoglobin. These patients often have fatigue, altered mental status, shortness of breath, and other manifestations of hypoxia without signs of RBC loss or volume depletion.7,8
Diagnostic Findings in Emergent Anemia Clinical Features The most common cause of clinically severe anemia is blood loss. The clinical manifestation of anemia depends on how rapidly the hematocrit falls and also on the patient’s ability to compensate. 1557
1558
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
Table 119-1 Hemogram Normal Values
History and Physical Examination for Clinically
AGE
HEMOGLOBIN (G/DL)
HEMATOCRIT (ML/DL)
RED BLOOD CELL COUNT (×106)
3 mo 3–7 yr Adult man Adult woman
10.4–12.2 11.7–13.5 14.0–18.0 12.0–16.0
30–36 34–40 40–52 35–47
3.4–4.0 4.4–5.0 4.4–5.9 3.8–5.2
BOX 119-2 Severe Anemia
History General Out-of-hospital status, therapy, response to therapy Bleeding diathesis Previous blood transfusion Underlying diseases, including allergies Current medications, especially those causing platelet inhibition Trauma Nature and time of injury Blood loss at scene Nontrauma Skin: petechiae, ecchymoses Gastrointestinal: hematemesis, hematochezia, melena, peptic ulcer Genitourinary: last menstruation, menorrhagia, metrorrhagia, hematuria Physical Examination Vital Signs Measured Serially Blood pressure, pulse, respiratory rate, oxygen saturation Orthostatic blood pressure and pulse (contraindicated with severe hypotension) Level and content of consciousness
Figure 119-1. Reticulocytes with reticular material after methylene blue staining. (From Hoffbrand AV, Pettite JE: Color Atlas of Clinical Hematology, 3rd ed. London, Mosby, 2000, p 18.)
Causes of Rapid Intravascular Red Blood
BOX 119-1 Cell Destruction
Mechanical hemolysis associated with disseminated intravascular coagulation Massive burns Toxins (e.g., some poisonous venoms—brown recluse spider, cobra) Infections such as malaria or Clostridium sepsis Severe glucose-6-phosphate dehydrogenase deficiency with exposure to oxidant stress ABO incompatibility transfusion reaction Cold agglutinin hemolysis (e.g., Mycoplasma organisms, infectious mononucleosis) Paroxysmal nocturnal hemoglobinuria exacerbated by transfusion Immune complex hemolysis (e.g., quinidine) Clinical signs and symptoms include tachycardia, decreased blood pressure, postural hypotension, light-headedness, increased heart rate, and increased respiratory rate. Complaints of thirst, altered mental status, and decreased urine output may also be present. The patient’s age, concomitant illness, and underlying hematologic, cerebral, and cardiovascular status tremendously influence the clinical findings. Children and young adults may tolerate significant blood loss with unaltered vital signs until a precipitant hypotensive episode occurs. Pediatric patients may become markedly tachycardic, physiologically attempting to maintain cardiac output, since their ability to increase stroke volume is limited. Elderly patients commonly have underlying disease states that compromise their ability to compensate for blood loss.9
Skin Pallor Diaphoresis Jaundice Cyanosis Purpura, ecchymoses, petechiae Evidence of penetrating wounds Cardiovascular Murmurs, S3, S4 Quality of femoral and carotid pulses Abdomen Hepatosplenomegaly Pain, guarding, rebound on palpation Rectal and pelvic examination Masses Stool hemoglobin testing
Pertinent elements of the history and physical examination of patients with acute anemia are listed in Box 119-2.10
Ancillary Evaluation Stabilization of emergent anemia commonly runs parallel to assessment. If the signs and symptoms suggest potential life-threatening conditions, intravenous lines are placed and samples for the following initial laboratory tests are drawn: 1. Complete blood count and peripheral smear 2. Blood sample for type and crossmatch 3. Prothrombin time 4. Partial thromboplastin time 5. Electrolyte levels 6. Glucose level (especially if the patient has altered consciousness) 7. Creatinine level
1559
BOX 119-3 History and Physical Examination for Nonemergent Anemia Previous treatment of anemia Weight loss Back pain Physical Examination Skin Pallor Purpura, petechiae, angiomas Ulcerations Eye Conjunctival jaundice, pallor Funduscopic hemorrhage, petechiae Oral Tongue atrophy, papillary soreness Cardiopulmonary Heart size, murmurs, extra cardiac sounds Rales, other signs of pulmonary edema Abdomen Hepatomegaly, splenomegaly Ascites Masses Lymph nodes Neurologic Altered positions or vibratory sense Peripheral neuritis Rectal and pelvic
8. Urinalysis for free hemoglobin 9. Clotting and unclotted blood samples for later testing
Most of these patients do not need immediate stabilization and can be further evaluated as outpatients.
If possible, a blood sample is obtained for measurement of hematocrit in the emergency department. Although it may take hours before the hematocrit correctly reflects the degree of blood loss, the initial value is useful in determining an initial baseline. Occasionally, this value reveals an underlying anemia with the acute blood loss superimposed. Depending on severity, a blood sample is sent for type and crossmatch. Peripheral smear interpretation is done on pretreatment blood samples. Measurements of coagulation status, electrolytes, glucose, blood urea nitrogen, and creatinine are useful in the diagnosis of underlying disease processes that may relate to the patient’s anemia. Values of folate, vitamin B12, iron, total iron-binding capacity, reticulocytes, and direct antiglobulin (Coombs’ test) are altered by transfusion. Therefore, pretreatment samples are best saved.11,12
Ancillary Evaluation The initial laboratory evaluation includes a complete blood count with leukocyte differential, reticulocyte count, peripheral smear (Fig. 119-2), and RBC indices, including mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin concentration (MCHC).
Diagnostic Findings in Nonemergent Anemia Clinical Features Nonemergent anemias are usually seen in ambulatory patients complaining of fatigue and feeling “washed out.” Other voiced complaints include irritability, headache, postural dizziness, angina, decreased exercise tolerance, shortness of breath, and decreased libido. When the anemia is of slow onset, the patient may adapt until the hemoglobin is very low. Alternatively, patients with rapid blood loss may experience lightheadedness or syncope even when the measured hemoglobin is not critically low. For patients without evidence of acute bleeding or emergent condition, elements of history and physical examination may help identify the cause (Box 119-3).
Figure 119-2. Normal smear. (From Hoffbrand AV, Pettite JE: Color Atlas of Clinical Hematology, 3rd ed. London, Mosby, 2000, p 22.)
Chapter 119 / Anemia, Polycythemia, and White Blood Cell Disorders
History Symptoms of anemia Chest pain, decreased exercise tolerance, dyspnea Weakness, fatigue, dizziness, syncope Bleeding diathesis Bleeding after trauma, injections, tooth extractions Spontaneous bleeding, such as epistaxis, menorrhagia Spontaneous purpura and petechiae Sites of blood loss Respiratory: epistaxis, hemoptysis Gastrointestinal: hematemesis, hematochezia, melena Genitourinary: abnormal menses, pregnancies, hematuria Skin: petechiae, ecchymoses Intermittent jaundice, dark urine Dietary history Vegetarianism Poor nutrition Drug use and toxin exposure, including alcohol Racial background, family history Underlying disease Uremia, liver disease, hypothyroidism Chronic disease states such as cancer, rheumatic or renal disease Previous surgery Miscellaneous
1560
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
BOX 119-4 Admission Criteria for Nonemergent Anemia Developing cardiac symptoms, such as shortness of breath or chest pain, or neurologic symptoms Initial unexplained hemoglobin value less than 8–10 g/dL or hematocrit less than 25–30% Major difficulty in obtaining outpatient care for patients whose hemoglobin levels are significantly low or when comorbidity is present
BOX 119-5
Differential Diagnosis of Anemias Caused by Decreased Red Blood Cell Production: Subclassification by Red Blood Cell Indices
Hypochromic Microcytic Anemias (Decreased MCV and Hemoglobin Concentration) Iron deficiency Thalassemia Sideroblastic anemia or lead poisoning Chronic disease (e.g., cancer, renal or inflammatory disease); normochromic and normocytic indices often found Macrocytic (Elevated MCV) Vitamin B12 deficiency Folate deficiency Liver disease Hypothyroidism Normocytic (Normal MCV and Hemoglobin Concentration) Primary bone marrow involvement: aplastic anemia, myeloid metaplasia with myelofibrosis, myelophthisic anemia Resulting from underlying disease: hypoendocrine state (thyroid, adrenal, pituitary), uremia, chronic inflammation, liver disease MCV, mean corpuscular volume.
Disposition Reasonable criteria for the admission of patients with non emergent anemia are shown in Box 119-4.12
Differential Diagnosis The differential diagnosis of anemia is facilitated by classifying the anemia into one of three groups: decreased RBC production, increased RBC destruction, and blood loss.12 A complementary approach uses RBC morphology and indices.13
Decreased Red Blood Cell Production Anemias caused by decreased RBC production have a natural history of insidious onset and an associated decreased reticulocyte count. A subclassification by indices of anemias caused by decreased RBC production is listed in Box 119-5. The RBC indices and morphology manifested in a peripheral smear are useful in securing the diagnosis. The definitive diagnosis is usually made outside the emergency department and may require bone marrow examination. The emergency physician rarely initiates replacement therapy, except in circumstances that require transfusion. Appropriate diagnostic tests may be initiated, but replacement of iron, vitamin B12, or folate without proof of cause is unnecessary and unwise.
of Red Blood Cell Indices and Table 119-2 Calculation Normal Values INDEX
Mean corpuscular volume Mean corpuscular hemoglobin Mean corpuscular hemoglobin concentration
FORMULA FOR CALCULATION
NORMAL RANGE
Hematocrit (%) divided by red blood cell count (106/µL) Hemoglobin (g/dL) divided by red blood cell count (106/µL) Hemoglobin (g/dL) divided by hematocrit (%)
81–100 fL 26–34 pg 31–36%
fL, femtoliter.
RBC indices are useful in classifying anemias caused by a production deficit. Their calculation and normal ranges are provided in Table 119-2. MCV is a measure of RBC size. Decreases and increases reflect microcytosis and macrocytosis, respectively. MCH incorporates both RBC size and hemoglobin concentration. It is influenced by both and is the least helpful of the indices. The MCHC index is a measure of the concentration of hemoglobin. Low values represent hypochromia, whereas high values are noted only in patients with decreased cell membrane relative to cell volume, such as in the case of spherocytosis. An additional index is the RBC distribution width (RDW), which is a measure of the homogenicity of the RBCs measured. RDW is automatically calculated as the standard deviation of MCV divided by MCV multiplied by 100. A normal RDW is 13.5 ± 1.5%. It is useful in differentiating iron deficiency from thalassemia.14 Microcytic Anemias. Hypochromic microcytic anemias can be subdivided into deficiencies of the three building blocks of hemoglobin: iron (iron-deficiency anemia [Fig. 119-3]), globin (thalassemia), and porphyrin (sideroblastic anemia and lead poisoning). Anemia of chronic disease, a secondary iron abnormality, rounds out the differential diagnosis. Not all microcytic anemias are the result of iron deficiency, and routine iron therapy for a patient with a low MCV and MCHC is inappropriate. Iron-Deficiency Anemia. Iron deficiency is a frequent cause of chronic anemia seen in the emergency department. It is the most common anemia in women of childbearing age. In older patients, occult blood loss, especially gastrointestinal, may initially appear as iron-deficiency anemia. Because changes in RBC size and hemoglobin content occur only after bone marrow and cytochrome iron stores are depleted, a patient may have early symptoms of iron deficiency (e.g., fatigue) without manifesting changes in RBC structure. Actually, a low MCV is relatively rare in iron-deficiency anemia. The diagnosis is made by laboratory evaluation of the fasting level of serum iron, serum ferritin, and total iron-binding capacity. The laboratory interpretation and pitfalls are outlined in Table 119-3. A concentrated search for occult blood loss is vital. Therapy consists of oral iron replacement. A cost-effective form is ferrous sulfate. The dosage is 300 mg for adults (60 mg of elemental iron) or 3 mg/kg/day for children. This medication is generally well tolerated, although it may cause nausea, vomiting, or constipation. Patients should be warned that their stools will be blackened. In rare patients with poor oral tolerance or absorption, parenteral iron therapy may be necessary.
1561
microcytic cells and poikilocytes (abnormally shaped cells). (From Hoffbrand AV, Pettite JE: Color Atlas of Clinical Hematology, 3rd ed. London, Mosby, 2000, p 44.)
Table 119-3 Diagnostic Tests for Iron-Deficiency Anemia TEST
NORMAL RESULT
IRON-DEFICIENCY LEVEL
INTERPRETATION
Fasting serum iron
60–180 µg/dL
400 µg/dL
Diurnal variation (draw in morning); increased by hepatitis, hemochromatosis, hemolytic anemia, and aplastic anemia; decreased in infection Increased in late pregnancy or hepatitis; decreased in infection
Percentage of saturation (serum iron) of total iron-binding capacity Serum ferritin
15–45%
900 pg/mL, unbound vitamin B12–binding capacity 18.5 g/dL In women: Hgb > 16.5 g/dL Normal arterial oxygen saturation (>92%) Splenomegaly
Disposition
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
1574
becoming more popular, however. Normal values for the WBC count are listed in Table 119-8. The “normal” count is age dependent until childhood and may be shifted upward by exercise, gender (women), smoking, and pregnancy. Decreases in the total WBC count range of 1000 to 1200 cells/mm3 have been noted in the African American population. Laboratory errors may be due to improper sample preparation, nucleated RBCs, or platelet clumping. The blood smear differential count may also be influenced by small sample size, improper cell identification, and age group (children). Differential ranges are listed in Table 119-9. One common but easily corrected error in laboratory reporting is giving the results in terms of the percentage of cell types. Absolute counts for each cell type are more accurate and useful in assessing the risk for infection.75,76
Abnormal Values Because of the wide range of normal values, all abnormal WBC counts should be interpreted in the context of the patient’s condition. A careful history and physical examination, absolute cell counts, and review of the peripheral smear differential count are the starting points for determining the origins of quantitative WBC disorders.
Leukocytosis Most cases of leukocytosis are caused by increases in the neutrophil or lymphocyte cell lines. Neutrophil leukocytosis (neutrophilia) is an absolute neutrophil count greater than 7500 cells/mm3 and is commonly associated with infection or inflammation (Box 119-16). Because increased neutrophil destruction is associated with both these pathologic processes, bone marrow stores are drawn on, and the usual ratio of 1 band to 10 neutrophils increases. This increase is manifested as a “left shift” in the differential count and represents immature
Ranges for the Blood Leukocyte Table 119-8 Normal Count (Cells/mm3) AGE
AVERAGE
95% RANGE (AVERAGE VALUE ± 2 SD)
1 wk 6 mo 12 mo 4 yr 8 yr Adults
12,200 11,900 11,400 9,100 8,300 7,400
5,000–21,000 6,000–17,500 6,000–17,500 5,500–15,500 4,500–13,500 4,500–11,000
Modified from Miale JB: Laboratory Medicine: Hematology, 6th ed. St. Louis, Mosby, 1982.
neutrophils from the postmitotic pool moving into the circulation. WBC counts can increase without a left shift or an increase in band forms by demarginating neutrophils from the vessel walls. It is often seen as a response to stress, exercise, or epinephrine. Severe stress can raise the WBC count to 18,000 to 20,000 cells/mm3.75,76
Chronic Myeloid Leukemia One of the myeloproliferative causes of neutrophilic leukocytosis is chronic myeloid leukemia (CML). Although it is the least common of the major leukemias (60% acute, 31% chronic lymphocytic leukemia, and 15% CML), it must be considered in neutrophilia. Patients with CML are usually older than 40 years and have WBC counts greater than 50,000 cells/mm3. The differential count shows elevated polymorphonuclear neutrophils and metamyelocytes. Less often, the basophil and eosinophil counts are increased. CML is a stem cell disorder in which the WBC count is elevated and the differential is normal. Mature and intermediate granulocytes are overproduced. Platelets may also be increased, but RBC production is down, thereby resulting in anemia. The patient often complains of fatigue, anorexia, sweating, and weight loss. Physical findings include pallor, sternal tenderness, and splenomegaly (90% of patients; Fig. 119-11). In the laboratory, decreased leukocyte alkaline phosphatase and increased vitamin B12 levels are found, which helps differentiate CML from other causes of neutrophilia. The Philadelphia chromosome (Ph1) is almost constantly associated with the disease. The chronic phase of CML is treated with an alkylating agent (e.g., busulfan) or an antimetabolite (e.g., hydroxyurea). Selected patients may benefit from bone marrow transplantation.77,78 The need for urgent therapy in CML is usually related to hyperuricemia and renal injury or severe anemia and subsequent angina or heart failure. Rarely, hyperleukocytosis occurs, but the more mature, “less sticky” cells in CML do not usually cause problems unless the count exceeds 500,000 cells/mm3. A higher cell count may cause leukostasis and result in deafness, visual impairment, pulmonary ventilationperfusion abnormalities, and priapism. Treatment involves hydration, leukapheresis, transfusion as necessary, allopurinol to prevent severe hyperuricemia, and specific chemotherapy (hydroxyurea). Late problems in the natural history of CML involve progressive loss of cell differentiation and response to therapy. The term blastic crisis represents the sudden appearance of an acute form of leukemia, which is a rare substage of the evolving deterioration.77,78 The condition may occur in lymphoid or myeloid forms. Blast counts greater than 50,000 cells/mm3 may predispose the patient to the complications of leukostasis.
Table 119-9 Normal Percentage Ranges for the Leukocyte Differential Count in Blood* AGE
SEGMENTED NEUTROPHILS
BAND NEUTROPHILS
1 wk 6 mo 12 mo 4 yr 8 yr Adult
34 ± 15 (4100) 23 ± 10 (2710) 23 ± 10 (2680) 34 ± 11 (3040) 45 ± 11 (3700) 51 ± 15 (3800)
11.8 8.8 8.1 8.0 8.0 8.0
± ± ± ± ± ±
4 3 3 3 3 3
(1420) (1000) (990) (730) (660) (620)
LYMPHOCYTES
41 61 61 50 39 34
± ± ± ± ± ±
*Numbers in parentheses indicate the average number of cells per cubic millimeter. Modified from Miale JB: Laboratory Medicine: Hematology, 6th ed. St. Louis, Mosby, 1982.
5 (5000) 15 (7300) 15 (7000) 15 (4500) 15 (3300) 10 (2500)
MONOCYTES
EOSINOPHILS
BASOPHILS
9.1 4.8 4.8 5.0 4.2 4.0
4.1 2.5 2.6 2.8 2.4 2.7
0–4 0–4 0–4 0–6 0–6 0–5
(1100) (480) (550) (450) (350) (300)
(500) (300) (300) (250) (200) (200)
(50) (50) (50) (50) (50) (40)
1575
BOX 119-16 Causes of Leukocytosis
Lymphocytosis (Absolute Count >9000/mm3, Ages 1–6; 7000/mm3, Ages 7–16; 4000/mm3, Adults) Viral infection (primary cause): mononucleosis, rubeola, rubella, varicella, toxoplasmosis Bacterial infection: pertussis, tuberculosis, hepatitis, cytomegalovirus Lymphoproliferative: acute or chronic lymphocytic leukemia Immunologic response: immunization, autoimmune diseases, graft rejection Endocrine: hypothyroidism Relative lymphocytosis associated with granulocytopenia Modified from Miale JB: Laboratory Medicine: Hematology, 6th ed. St. Louis, Mosby, 1982.
A leukemoid reaction is a nonleukemic reactive granulocytic leukocytosis that resembles CML but has no associated Ph1 chromosome, no absolute increase in basophils and eosinophils, and an increase in leukocyte alkaline phosphatase. It is difficult to distinguish from CML in the emergency department, and both must be considered as a potential diagnosis in granulocytic leukocytosis. WBC counts are usually greater than 50,000 cells/mm3. A leukemoid reaction may be seen in tuberculosis, Hodgkin’s disease, sepsis, and metastatic tumor, particularly bronchogenic, gastric, and renal carcinoma.76
Lymphocytic Leukocytosis Lymphocytic leukocytosis (lymphocytosis) is an agedependent definition: 9000 cells/mm3, ages 1 to 6 years; 7000 cells/mm3, ages 6 to 16 years; and 4000 cells/mm3, adults. It is seen in a variety of disorders, primarily infections and lymphoproliferative disease.79 In the past, acute and chronic were descriptive terms applied to lymphocytic neoplasms with respect to patient survival time before present therapy was available. The terms acute and chronic are currently used to describe the cell maturity, rapidity of onset, and aggressiveness of therapy. Chronic Lymphocytic Leukemia. Chronic lymphocytic leukemia is primarily a B cell disorder and is the most common type of leukemia in the population 50 years or older. Patients initially complain of fatigue, weight loss, increased susceptibility to infection, rashes, and easy bruising. The lymph nodes are nontender and smooth, and they may appear in only one or two areas. Splenic and hepatic enlargement occurs in more than 50% of patients. Laboratory support of the diagnosis is an absolute lymphocyte count greater than 5000 cells/mm3 in adults. Anemia, thrombocytopenia, and neutropenia are often found. Autoimmune hemolytic anemia, a positive direct antiglobulin test, and other altered immune system problems are seen. Early therapy may be directed toward complications of anemia, thrombocytopenia, an impaired or accentuated immune response, or enlarged lymph nodes or spleen. Leukostasis is seldom seen in chronic lymphocytic leukemia, but therapy is considered when the total count rises to higher than 200,000/mm3.79 Acute Lymphocytic Leukemia. Acute lymphocytic leukemia is most commonly diagnosed in children younger than 10 years. It is the most frequent malignancy in children younger than 15 years. The potential for leukostasis increases in acute lymphocytic leukemia when the blast count rises above 50,000 cells/mm3. Oncologic therapy is based on clinical staging and includes chemotherapy or radiation therapy. Aggressive therapy has improved childhood survival from 1 to 15 years or more. This response to treatment has not been found to the same degree in adults.80-82
Leukopenia
Figure 119-11. Chronic myeloid leukemia. (From Hoffbrand AV, Pettite JE: Color Atlas of Clinical Hematology, 3rd ed. London, Mosby, 2000, p 169.)
In adults, leukopenia is defined as an absolute blood cell count less than 4000 cells/mm3. Leukopenia is commonly associated with a reduction in one cell type, the neutrophil, and this decrease has the greatest clinical significance. The absolute neutrophil count is calculated by multiplying the WBC count by the combined percentage of band and segmented neutrophils. The absolute neutrophil count can be classified as mild (1000–1500 cells/mm3), moderate (500–1000 cells/mm3), or severe (7500 Cells/mm3) Inflammation: rheumatoid arthritis, gout Infection: bacterial most common Tissue necrosis: cancer, burns, infarctions Metabolic disorders: diabetic ketoacidosis, thyrotoxicosis, uremia Rapid RBC turnover: hemorrhage, hemolysis Myeloproliferative disorders: chronic myeloid leukemia, polycythemia vera Malignancy (e.g., gastrointestinal cancers) Stress: exercise, pain, surgery, hypoxia, seizures, trauma Drugs: epinephrine, corticosteroids, lithium, cocaine Pregnancy Heredity or idiopathic disease Laboratory error: automated counters, platelet clumping, precipitated cryoglobulin
Leukemoid Reaction
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
1576
Linkage of Leukopenia to Phases of Table 119-10 Neutrophil Maturation MECHANISM
EXAMPLE
Proliferation in bone marrow
Aplastic anemia, leukemia, cancer chemotherapy (cyclophosphamide, azathioprine, methotrexate, chlorambucil) Drugs: phenothiazines, phenylbutazone, indomethacin, propylthiouracil, phenytoin, cimetidine, semisynthetic penicillins, sulfonamides Infection: viral, tuberculosis, sepsis Folate or vitamin B12 deficiency, chronic idiopathic neutropenia Starvation Hypersplenism: sarcoidosis, portal hypertension, malaria Infection: viral most common (mononucleosis, rubella, rubeola), Rickettsia organisms, overwhelming bacterial infection Autoimmune disease: systemic lupus erythematosus, AIDS, Felty’s syndrome Leukocyte clumping, long delay in performing test
Maturation in bone marrow Distribution Increased use
Laboratory error
The physical signs of infection may be minimal in severe neutropenia because there are too few cells to generate a substantial inflammatory or purulent response. Neutropenia may be caused by decreased production, increased destruction, or movement of circulating neutrophils into marginal or tissue pools. Until recently, it was most often caused by a decrease in bone marrow production (Table 119-10). Autoimmune neutropenia is also becoming more commonly diagnosed because it is thought to have a role in acquired immunodeficiency syndrome.76,83 A thorough medication history must be taken in all patients found to have neutropenia. A previous history of neutropenia, a review of recent infection, and a family history are obtained. The review of systems focuses on bleeding problems, fatigue, sweats, weight loss, and autoimmune symptoms. The physical examination is directed toward sites of infection, lymphadenopathy, hepatosplenomegaly, and underlying disease. In patients with severe neutropenia and fever, a full radiologic and direct examination of commonly involved areas, such as the chest and urine, should be performed and sputum, urine, and blood cultures obtained. Basic isolation techniques, early admission, and consultation with another specialist are recommended. Specific therapies may be started after cultures and consultation are completed. A number of empirical antibiotic regimens are recommended for febrile patients with neutropenia.84,85 Human granulocyte colony-stimulating factor is often used in the setting of neutropenia, but it is best done in consultation with a hematologist.86,87 Patients with a clear reversible source or without significant clinical findings and mild to moderate levels of neutropenia may have outpatient follow-up arranged, preferably after discussion with their physician.
■ RATIONALE FOR SELECTION OF WHITE BLOOD CELL AND DIFFERENTIAL COUNTS The WBC count has not proved to be a highly sensitive or specific test for the diagnosis of a variety of disease entities.
BOX 119-17
Agents and Conditions That Elevate the WBC Count
Acetylcholine Acidosis Adrenergic drugs Alcohol Allergic reactions Bacterial infection Blood donation Burns Competitive running Crying in infants Fever Gout Hemolysis Heparin Histamine Hypoxia Iron overdose Juvenile rheumatoid arthritis Lead and other toxins Lithium Menstruation Myocardial infarction Neonatal asphyxia Neoplasm Normal pregnancy Pain Polyarteritis nodosa Prednisone Pulmonary infarction Seizures Snake bite Supraventricular tachycardia Surgery Trauma Uremia Viral infection Vomiting
However, in certain clinical situations, the WBC count may have utility. For example, studies evaluating the WBC count for the diagnosis of abdominal pain have found it to be a useful confirming test or helpful in selecting patients for observation. In evaluating the bacterial versus viral infectious potential in febrile children, the WBC and differential counts have demonstrated limited usefulness, except in children younger than 2 years, in whom counts greater than 15,000 cells/mm3 have an increased correlation with bacteremia.88-90 In addition to the nonspecificity of the WBC count, the differential count provides additional helpful information in less than 1% of cases. The absolute leukocyte count or the differential cannot reliably distinguish between viral and bacterial infection. The test should be viewed as having limited screening value in the acute care setting. Multiple agents and conditions increase the WBC count (Box 119-17), thus making the test less specific for infection than previously assumed.88-90
1577
KEY CONCEPTS Anemia in the elderly often occurs as an exacerbation of preexisting comorbid diseases. ■ Anemia of uncertain etiology should be thoroughly evaluated. If the patient has no adverse hemodynamic consequences, the evaluation can proceed on an outpatient basis. ■ One of the most important but often overlooked studies in the evaluation of suspected hemolytic anemia is the peripheral blood smear.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
■
Patients with sickle cell disease who come to the emergency department are most commonly having a true crisis and are not simply exhibiting drug-seeking behavior. ■ The white blood cell determination in the emergency department has poor sensitivity and specificity for disease.
Chapter 119 / Anemia, Polycythemia, and White Blood Cell Disorders
■
Chapter 120
Disorders of Hemostasis
Timothy G. Janz and Glenn C. Hamilton
■ PERSPECTIVE Hemostasis is the process of blood clot formation and represents a coordinated response to vessel injury. It requires an orchestrated response from platelets, the clotting cascade, blood vessel endothelium, and fibrinolysis. Thrombinstimulated clot formation and plasmin-induced clot lysis are closely related and regulated. This dynamic process is often viewed in phases: formation of a platelet plug, propagation of the coagulation cascade, formation of a clot, and fibrinolysis of the clot. Most hemostatic abnormalities are acquired and result from drugs (e.g., aspirin or warfarin [Coumadin]), from associated disease (e.g., hepatic insufficiency), or from iatrogenic causes (e.g., multiple transfusions).
■ PATHOPHYSIOLOGY Hemostasis depends on normal function and integration of the vasculature, platelets, and the coagulation pathway.
Vasculature Vascular integrity is maintained by a lining of nonreactive overlapping endothelial cells supported by a basement membrane, connective tissue, and smooth muscle. These cells are important in maintaining a barrier to macromolecules and, when injured, in contributing to the metabolic response and local vasoconstriction. The vascular wall is an important contributor to hemostasis.1 The endothelium contributes to both clot formation and regulation by producing substances such as von Willebrand’s factor (vWF), antithrombin III, heparin sulfate, prostacyclin, nitric oxide, and tissue factor pathway inhibitor.
Platelets Platelets have multiple and ever-expanding roles in our understanding of hemostasis. They are complex cytoplasmic fragments released from bone marrow megakaryocytes under the control of thrombopoietin. Platelets contain lysosomes, granules, a trilaminar plasma membrane, microtubules, and a canalicular system. Granules are an important component of hemostasis and contain platelet factor 4, adhesive and aggregation glycoproteins, coagulation factors, and fibrinolytic inhibi1578
tors. Each participates in the process of coagulation. The platelet’s role is termed primary hemostasis, and it serves as the initial defense against blood loss. A fibrin clot that incorporates coagulation factors usually reinforces a platelet clot. Platelet activity is summarized in Box 120-1. Any of the steps listed may be absent, altered, or inhibited by inherited or acquired disorders.2-6
Coagulation Pathway The coagulation pathway is a complex system of checks and balances that results in controlled formation of a fibrin clot. Coagulation factors have been given standard Roman numerals matching their order of discovery (Box 120-2).7 A simplified version of the coagulation pathway is presented in Figure 120-1. The clotting cascade is traditionally depicted as consisting of intrinsic and extrinsic pathways. The intrinsic pathway is initiated by exposure of blood to a negatively charged surface, such as a glass surface in the activated partial thromboplastin clotting time. The extrinsic pathway is activated by tissue factor exposed at the site of vessel injury or thromboplastin. Both pathways converge to activate factor X, which then activates prothrombin to thrombin. The primary physiologic event that initiates clotting is exposure of tissue factor at the injured vessel site. Tissue factor is a critical cofactor that is required for activation of factor VII. Activated factor VII activates factor X directly, as well as indirectly by activating factor IX. Because of limited amounts of tissue factor and rapid inactivation by tissue factor pathway inhibitor, the extrinsic pathway initiates the clot process. Sustained generation of thrombin and clot formation depend on the intrinsic pathway through activation of factor IX by activated factor VII, which helps explain the bleeding problems associated with hemophilia.7,8 Intrinsic, extrinsic, and common pathways must function normally for hemostasis to occur, and each may be evaluated with laboratory tests.1,7 The clinically important groups of coagulation factors are as follows: 1. Thrombin-sensitive factors contributing to the metabolic response and local vasoconstriction: I, V, VIII, XIII 2. Vitamin K–sensitive factors: II, VII, IX, X 3. Sites of heparin activity: IIa, IXa, Xa (major site), XIa, platelet factor 3
BOX 120-1 Role of Platelets in Hemostasis Adhesion to subendothelial connective tissue: collagen, basement membrane, and noncollagenous microfibrils; serum factor VIII (von Willebrand’s) permits this function; adhesion creates the initial bleeding arrest plug Release of adenosine diphosphate, the primary mediator and amplifier of aggregation; release of thromboxane A, another aggregator and potent vasoconstrictor; release of calcium, serotonin, epinephrine, and trace thrombin Platelet aggregation over the area of endothelial injury Stabilization of the hemostatic plug by interaction with the coagulation system: Platelet factor 3, a phospholipid that helps accelerate certain steps in the coagulation system Platelet factor 4, a protein that neutralizes heparin Pathway initiation and acceleration by thrombin production Possible secretion of active forms of coagulation proteins Stimulation of limiting reactions of platelet activity Contact
1579
Coagulation Control All the components of the coagulation reaction are necessary to prevent excessive bleeding. Hemostasis is a balance between the excessive bleeding state and thrombosis. Once coagulation is initiated, controls are necessary to prevent local or generalized thrombosis. These controls include the following:11,13-18 1. Removal and dilution of activated clotting factors via blood flow, which also mechanically opposes growth of the hemostatic plug
BOX 120-2 Coagulation Factors I. Fibrinogens II. Prothrombin III. Tissue thromboplastin IV. Calcium V. Labile factor (proaccelerin) VI. Not assigned VII. Proconvertin VIII. Antihemophilic A factor IX. Antihemophilic B factor (plasma thromboplastin component, Christmas factor) X. Stuart-Prower factor XI. Plasma thromboplastin antecedent XII. Hageman factor (contact factor) XIII. Fibrin-stabilizing factor
Intrinsic
Extrinsic
Surface contact
Tissue thromboplastin (III), released from damaged tissue
XII XIIa
VIIa Ca2+ Tissue lipid
XIa
XI
2+
Prekallikrein
Ca IX
Kallikrein
VII
IXa
VIIIAHF Ca2+ Platelet lipid
Kinins
Inflammation Common pathway
Xa
X
V
Va Ca2+ Lipid II
XIII IIa XIIIa
I (fibrinogen)
Figure 120-1. Coagulation pathway.
Ia (fibrin)
Insoluble fibrin
Chapter 120 / Disorders of Hemostasis
Thrombin-sensitive factors are activated by thrombin and may give rise to a bleeding disorder if defective synthesis occurs. Vitamin K–sensitive factors may also cause bleeding from defective synthesis, as occurs with liver disease and warfarin anticoagulants. Heparin in combination with antithrombin III affects the coagulation pathway at multiple sites.9-12
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
1580
2. Modulation of platelet activity by endothelial-generated nitric oxide and prostacyclin 3. Removal of activated coagulation components by the reticuloendothelial system 4. Regulation of the clotting cascade by antithrombin III, protein C, protein S, and tissue factor pathway inhibitor 5. Activation of the fibrinolytic system
■ CLINICAL FEATURES Out-of-Hospital Treatment The out-of-hospital treatment of bleeding problems presents no special concerns. Local pressure and volume repletion are the mainstays of therapy for blood loss. The out-of-hospital team must be aware that inherited coagulopathies may complicate any medical or traumatic problems and that coagulopathy can rapidly develop in critically ill patients. Patients who do not respond quickly to the usual measures of hemostasis either in the field or in the emergency department (ED) should be considered to have a potential bleeding disorder.
History and Physical Examination An outline of the history and physical examination is presented in Box 120-3. The history alone may be useful in differentiating between platelet and coagulation factor abnormalities. Platelet disorders are usually manifested as acquired petechiae, purpura, or mucosal bleeding and are more common in women. Coagulation problems are commonly congenital, are characterized by delayed deep muscle or joint bleeding, and are seen more often in men.
Ancillary Evaluation A definitive diagnosis depends on laboratory evaluation. Tests pertinent to the ED are discussed in the following sections and listed in Box 120-4.
BOX 120-3 Clinical Evaluation of a Bleeding Patient History Nature of bleeding Petechiae Purpura Ecchymosis Significant bleeding episodes Sites of bleeding Skin Mucosa: oral or nasal Muscle Gastrointestinal Genitourinary Joints Patterns of bleeding Recent onset or lifelong Frequency and severity Spontaneous or after injury Challenges to hemostasis Tooth extraction Operative procedures Association with medication, particularly aspirin Medications Associated diseases Uremia Liver disease Infection Malignancy Previous transfusion Family history Physical Examination Vital signs Skin: nature of bleeding, signs of liver disease Mucosa: oral or nasal Lymphadenopathy Abdomen: liver size and shape, splenomegaly Joints: signs of previous bleeding Other sites of blood loss: pelvic, rectal, urinary tract
Complete Blood Count and Blood Smear The complete blood count (CBC) assesses the degree of anemia associated with the bleeding episode. Reductions in hemoglobin and hematocrit often lag behind the actual loss of red blood cells (RBCs) in acute hemorrhage because of a slow equilibration time. The peripheral blood smear may demonstrate schistocytes or fragmented RBCs in disseminated intravascular coagulation (DIC). Teardrop-shaped or nucleated RBCs may reflect myelophthisic disease. A characteristic white blood cell morphologic condition is seen with thrombocytopenia associated with infectious mononucleosis, folate or vitamin B12 deficiency, or leukemia.19
Platelet Count The platelet count may be estimated from the smear. Normally, one platelet is present per 10 to 20 RBCs. Often, the count is automated, the normal range being 150,000 to 400,000/mm3. Platelet counts less than 100,000/mm3 define thrombocytopenia. With normal platelet function, the bleeding time increases in direct relation to a decrease in the platelet count below 100,000/mm3. Levels below 20,000/mm3 may be associated with serious spontaneous hemorrhage. However, the count gives no information about the functional capability of platelets.20
BOX 120-4 Coagulation Studies CBC and smear (EDTA—purple top) Platelet count (EDTA—purple top) Bleeding time Prothrombin time (citrate—blue top) Partial thromboplastin time (citrate—blue top) Other coagulation studies: fibrinogen level, thrombin time, clot solubility, factor levels, inhibitor screens As necessary: electrolytes, glucose, BUN, creatinine, type and crossmatch BUN, blood urea nitrogen; CBC, complete blood count; EDTA, ethylenediaminetetraacetic acid.
Bleeding Time Bleeding time is the best test for determining both vascular integrity and platelet function that can be performed in the ED. The test is performed after making two standard incisions 1 mm deep and 1 cm long on the volar aspect of the forearm under 40 mm Hg pressure via a blood pressure cuff utilizing a template to ensure appropriate incisions. The time is mea-
Prothrombin Time The prothrombin time (PT) tests the factors of the extrinsic and common pathways. The patient’s anticoagulated plasma is combined with calcium and tissue factor prepared from rabbit or human brain tissue. Sensitivity to factor deficiencies depends on the source of the tissue factor. The PT detects deficiencies in fibrinogen, prothrombin, factor V, factor VII, and factor X. It is used to test the extrinsic pathway. A normal control sample is simultaneously run, and the clotting times of both are recorded. The time in seconds is usually given over the normal control time, for example, 12.5/11.5. A PT 2 seconds or more over the control time can be considered significant. Results are usually reported as the international normalized ratio (INR), which compensates for differences in sensitivity of various thromboplastin reagents to the effects of warfarin. The test is helpful in monitoring the use of coumarin anticoagulants, and the time may be prolonged in patients with liver disease and other abnormalities of vitamin K–sensitive factors.20
Partial Thromboplastin Time The partial thromboplastin time (PTT) tests the components of the intrinsic and common pathways, that is, essentially all factors but VII and XIII in the entire clotting cascade. In this test a phospholipid source and a contact-activating agent (kaolin) are added to anticoagulated citrate plasma. After an incubation period that allows factor XII to become activated, calcium is added and the clotting time is recorded. A normal control sample is run simultaneously. Normal ranges may vary, and each hospital laboratory should be checked. The average time is 25 to 29 seconds. The sensitivity of the test varies from factor to factor, but factor levels must usually be less than 40% before the PTT is prolonged. The test may be altered by clotting factor inhibitors of external origin (e.g., heparin) or internal origin (e.g., anti-VIII antibody). Inappropriately high values may occur if the plasma is too turbid or icteric. The activated PTT is most sensitive to abnormalities in the sequence of the coagulation cascade that precedes activation of factor X.20,23,24
Fibrinogen Fibrinogen is present in sufficient concentration to be measured directly. Because it is the final coagulation substrate, its level reflects the balance between production and consumption. It may be decreased by hypoproduction, as in severe liver disease, or by overconsumption, as in DIC. Low levels or altered function increase the PT, PTT, and thrombin clotting time. Because fibrinogen is an acute phase reactant, certain
1581
conditions, including malignancy, sepsis, inflammation, and pregnancy, may alter the test result.
Thrombin Time Measurement of the thrombin clotting time bypasses the intrinsic and extrinsic pathways by directly converting fibrinogen to fibrin. It is a useful screening test for both qualitative and quantitative abnormalities of fibrinogen and inhibitors such as heparin and fibrin split products.20
Clot Solubility The result of clot solubility testing may be the only abnormality in disorders involving factor XIII deficiency and some abnormal fibrinogen. A washed clot is incubated in acetic acid or urea. If the clot is not properly cross-linked, it dissolves.12
Factor Level Assays Factor levels are determined either by bioassay, in which the ability of the sample of plasma to normalize controlled substrate-deficient plasma is evaluated, or by immunologic assay. Inhibitor screening tests reveal antibodies in plasma that prolong the normal plasma clotting time when mixed.9,18,24
■ DIFFERENTIAL DIAGNOSIS AND MANAGEMENT When a bleeding disorder is diagnosed or suggested, the assessment initially includes stabilization, which may necessitate volume, RBC, and coagulation factor replacement. If the disorder is known, clinical complications associated with its underlying pathophysiologic condition must be considered. If the disorder is unknown, a rapid differential diagnosis must be made. A clinically useful scheme approaches bleeding disorders in terms of three constituents: vascular integrity, platelets, and coagulation factors. This differential diagnostic approach can be further divided into inherited and acquired disorders.
Vascular Disorders Vascular disorders have signs and symptoms similar to those of thrombocytopenic states. The inherited forms are rare. Acquired forms are usually associated with connective tissue changes or endothelial damage. The differential diagnosis of vascular disorders is listed in Box 120-5.25
Platelet Disorders General Approach Most platelet abnormalities occur in women and are acquired. The bleeding source is usually capillary, with resultant cutaneous and mucosal petechiae or ecchymosis. Epistaxis, menorrhagia, and gastrointestinal bleeding are common initial symptoms. The bleeding is generally mild and occurs immediately after surgery or dental extractions. Petechiae and purpura may be noted on physical examination, and superficial ecchymoses may be found around a venipuncture site. Deep muscle hematomas and hemarthroses are not aspects of the clinical picture. The bleeding time is prolonged, and the platelet count may be low, normal, or high. The differential diagnosis of platelet disorders is listed in Box 120-6.
Chapter 120 / Disorders of Hemostasis
sured from the incision to the moment when the blood oozing from the wound is no longer absorbed by filter paper. Some institutions have replaced the traditional bleeding time with a platelet function analyzer instrument, which is just as accurate and more convenient. A normal time is 8 minutes, a time of 8 to 10 minutes is borderline, and a time longer than 10 minutes is typically abnormal. Because of the high incidence of druginduced platelet dysfunction, it is important to ask the patient about medications, particularly aspirin and other antiplatelet medications (e.g., clopidogrel). The test is independent of the coagulation pathways.20-22 As mentioned previously, the bleeding time is prolonged with platelet counts below 100,000/mm3, but such prolongation does not represent platelet dysfunction. However, a prolonged bleeding time associated with platelet counts greater than 100,000/mm3 suggests impaired function.
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
1582
BOX 120-5 Differential Diagnosis of Vascular Disorders
BOX 120-6 Differential Diagnosis of Platelet Disorders
Inherited Disorders of connective tissue Pseudoxanthoma elasticum Ehlers-Danlos syndrome Osteogenesis imperfecta Disorders of blood vessels Hemorrhagic telangiectasia
Thrombocytopenia Decreased production Decreased megakaryocytes secondary to drugs, toxins, or infection Normal megakaryocytes with megaloblastic hematopoiesis or hereditary origin Platelet pooling and splenic sequestration Increased destruction Immunologic Related to collagen vascular disease, lymphoma, leukemia Drug related Infection Post-transfusion Idiopathic (autoimmune) thrombocytopenic purpura Mechanical Disseminated intravascular coagulation Thrombotic thrombocytopenic purpura Hemolytic-uremic syndrome Vasculitis Dilutional secondary to massive blood transfusion
Acquired Scurvy (vitamin C deficiency) Simple or senile purpura Purpura secondary to steroid use Vascular damage Infection (meningococcemia) Hemolytic-uremic syndrome Hypoxemia Thrombotic thrombocytopenic purpura Snakebite Dysproteinemic purpura
Thrombocytopenia Decreased Production. Thrombocytopenia from decreased bone marrow production is usually caused by the effects of chemotherapeutic drugs, myelophthisic disease, or direct bone marrow effects of alcohol or thiazides. Splenic Sequestration. Splenic sequestration is rare and seen primarily with hypersplenism resulting from hematologic malignancy, portal hypertension, or disorders involving increased splenic RBC destruction, such as hereditary spherocytosis or autoimmune hemolytic anemia.26 Increased Destruction Immune Thrombocytopenia. Thrombocytopenia associated with increased peripheral destruction of platelets and shortened platelet survival caused by an antiplatelet antibody is seen in a number of diseases. In most cases a cause is identifiable. Collagen vascular diseases, particularly systemic lupus erythematosus, may cause an antiplatelet antibody–related platelet decrease. Similar associations have been noted with leukemia and lymphoma, particularly lymphocytic lymphoma. All evaluations of suggested immune thrombocytopenia should include a CBC, peripheral smear, antinuclear antibody test, and bone marrow examination.27 A number of drugs have been associated with thrombocytopenia of immunologic origin. Quinine and quinidine are common offenders that affect platelets through an “innocent bystander” mechanism. The platelet is coated with a drug-antibody complex, complement is fixed, and intravascular platelet lysis occurs. Because of its relatively high frequency, heparin is an important cause of drug-induced thrombocytopenia in hospitalized patients. Platelets are activated by the formation of an IgG-heparin complex. Low-molecular-weight heparin may be associated with less thrombocytopenia than standard, unfractionated heparin; however, both forms of heparin demonstrate cross-reactivity.28 Heparin-induced thrombocytopenia (HIT) is a serious immune-mediated side effect associated with heparin. HIT occurs in 0.76 to 2.6% of patients receiving unfractionated heparin and in less than 1% of those receiving low-molecularweight heparin. It usually occurs within 5 to 7 days of heparin treatment. Thrombus develops in approximately half the patients with HIT. The thrombotic complications can lead to loss of a limb in up to 20% and death in as many as 30%. The diagnosis should be suggested in the presence of absolute thrombocytopenia or a greater than 50% reduction in platelets after the initiation of heparin. The most specific diagnostic
Thrombocytopathy Adhesion defects such as von Willebrand’s disease Release defects: acquired and drug related Aggregation defects such as in thrombasthenia Thrombocytosis Autonomous (primary thrombocythemia) Reactive (secondary thrombocythemia) Iron deficiency Infection/inflammatory Trauma Nonhematologic malignancy Postsplenectomy Rebound from alcohol, cytotoxic drug therapy, folate/vitamin B12 deficiency
tests for HIT are serotonin release assays, heparin-induced platelet aggregation assays, and solid-phase immunoassays. Platelet-associated IgG levels are commonly elevated, but this finding is less specific or sensitive than the other diagnostic tests. More concerning to the emergency physician is delayedonset HIT. This form of HIT occurs a median of 14 days after the initiation of heparin, but it has been reported to occur up to 40 days after starting heparin. Arterial or venous thrombosis typically develops in patients with HIT after receiving heparin. Treatment of thrombotic complications in these patients involves the use of direct thrombin inhibitors (lepirudin, ar gatroban), factor Xa inhibitors (fondaparinux), or heparinoids (danaproid).29-32 Digitoxin, sulfonamides, phenytoin, and aspirin are other problem drugs. The patient has usually ingested the medication within 24 hours. An idiopathic thrombocytopenic purpura (ITP) type of syndrome has been reported in intravenous cocaine users.33 Clinical trials with platelet glycoprotein IIbIIIa antagonists suggest that intravenous glycoprotein IIb-IIIa inhibitors may confer an increased risk for associated thrombocytopenia, independent of heparin therapy.34 The platelet count may fall below 10,000/mm3 and be complicated by serious bleeding. Laboratory testing may confirm the presence of antibody, especially with the use of quinine and quinidine. After stopping administration of the drug, the platelet count improves slowly over a period of 3 to 7 days. A short course of
1583
static effect. Though rare, life-threatening bleeding should be treated with platelet transfusions, intravenous immune globulin (1 g/kg), and methylprednisolone (30 mg/kg IV). Otherwise, care is supportive. The use of all nonessential drugs should be stopped, particularly those that might inhibit platelet function, such as aspirin.35,38,39,42,43 A similar pattern of thrombocytopenic purpura has been reported in sexually active homosexual men. Although the clinical findings and response to therapy mimic ITP, the mechanism is believed to be nonspecific deposition of immune complexes and complement rather than antiplatelet IgG.44 Nonimmune Thrombocytopenia. Nonimmune platelet destruction is usually consumptive or mechanical. Consumption occurs as part of the process of intravascular coagulation, although it may be seen at sites of significant endothelial loss. Thrombotic thrombocytopenic purpura (TTP), hemolytic uremic syndrome, and vasculitis all initiate platelet destruction through endothelial damage.45,46 The most striking difference between the first two is the age at onset and the prognosis. Thrombotic Thrombocytopenic Purpura. The pathologic state of TTP is the result of subendothelial and intraluminal deposits of fibrin and platelet aggregates in capillaries and arterioles. Hemolytic uremic syndrome is considered to be very similar to TTP; however, the former is associated with less central nervous system and more renal involvement than TTP is. Although the initiating event is unclear, prostacyclin and abnormal platelet aggregation are believed to play a central role in pathogenesis of the disease. The disease may affect patients of any age or sex, but the majority are 10 to 40 years of age and 60% of cases occur in women. Most cases of TTP are idiopathic. However, TTP can be associated with medications. Quinine is the most common drug associated with the disease. The antiplatelet drugs ticlopidine and clopidogrel, which are used to treat a variety of cardiovascular and cerebrovascular disorders, have also been associated with TTP. It is classically seen as the constellation of thrombocytopenic purpura, microangiopathic hemolytic anemia, fluctuating neurologic symptoms, renal disease, and fever, but only 40% of cases have the classic pentad. The platelet count ranges from 10,000/mm3 to 50,000/mm3, and generalized purpura and bleeding complaints are common. Anemia is universal, with hematocrit levels commonly less than 20%. The hemolysis may cause jaundice or pallor, and the blood smear characteristically contains numerous schistocytes and fragmented RBCs. Neurologic symptoms include stroke, seizures, paresthesias, altered levels of consciousness, and coma, all of which characteristically fluctuate in severity. The renal component varies from hematuria and proteinuria to acute renal failure. Fever is present in 90% of patients. Untreated, the disease follows a progressive and fatal course, with 80% mortality rate 1 to 3 months after diagnosis. Therapy has included corticosteroids, splenectomy, anticoagulation, exchange transfusion, and dextran. However, plasma exchange with fresh frozen plasma (plasmapheresis) is the current treatment of choice. Over the last several years, the aggressive use of plasma exchange has reduced the mortality rate from 90 to 17%. In addition to plasma exchange, initial therapy may also include steroids such as prednisone and antiplatelet agents such as aspirin and dipyridamole (Persantine). Splenectomy, immune globulin, vincristine, and other therapies may have a role in resistant cases. With the exception of life-threatening bleeding, platelet transfusion should be avoided because platelets may cause additional thrombi in the microcirculation.45-50 Dilutional Thrombocytopenia. Dilutional thrombocytopenia occurs in cases of massive transfusion, exchange transfusion, or extracorporeal circulation. Volume replacement with stored bank blood is platelet-poor because platelets have a life span
Chapter 120 / Disorders of Hemostasis
corticosteroid therapy such as prednisone in a dose of 1 mg/kg with rapid tapering may facilitate recovery.35,36 Postinfectious immune thrombocytopenia is usually associated with viral diseases such as rubella, rubeola, and varicella. Although many cases associated with sepsis have a mechanical origin, some immune mechanisms have been demonstrated.35 Post-transfusion thrombocytopenia is a rare disorder that causes a precipitous fall in platelets approximately 1 week after the transfusion. In 90% of cases, its origin is linked to the 98% of the population carrying a PLA1 antigen on platelets. Despite the fact that 2% of blood recipients are mismatched with respect to this antigen, it is fortunately a rare occurrence. When transfused into a PLA1 antigen-negative patient, the platelets with attached PLA1 antibodies provoke an anamnestic response, but the actual mechanism of platelet destruction remains unknown. The platelet count often falls precipitously below 10,000/mm3, with a significant risk for major bleeding. Intracranial hemorrhage occurs in approximately 10% of such cases. Patients are usually middle-aged women with a history of pregnancy who may have been previously sensitized to the PLA1 antigen during pregnancy. Plasma exchange therapy is an effective antidote.35,37 Idiopathic Thrombocytopenic Purpura. Autoimmune idiopathic thrombocytopenic purpura (ITP) should be considered after other causes have been excluded. ITP is associated with an IgG antiplatelet antibody that has proved difficult to detect. The two clinically important forms are acute and chronic.27,38,39 The acute form of ITP is seen most often in children 2 to 6 years of age. A viral prodrome commonly occurs within 3 weeks of its onset. The platelet count falls, usually to less than 20,000/mm3. The course is self-limited, with a greater than 90% rate of spontaneous remission. Morbidity and mortality rates are low, although full recovery may take several weeks. Treatment is supportive, and steroid therapy does not alter the disease course.35,39 The more chronic form of ITP is primarily an adult disease found three times more often in women than men. The onset of chronic ITP is insidious, without a prodrome, and it is manifested as easy bruising, prolonged menses, and mucosal bleeding. The patient may have petechiae or purpura, and platelet counts between 30,000/mm3 and 100,000/mm3 are common. Bleeding complications are of unpredictable frequency and severity, although the long-term mortality rate is approximately 1%.38,39 Splenomegaly is unusual in either acute or chronic ITP. Recently a thrombopoietin-receptor agonist, eltrombopag, has been shown to increase platelet counts in patient with relapsed or refractory ITP.40 Eltrombopag has also been used to increase platelet counts in patients with cirrhosis secondary to hepatitis C virus.41 The role that this and other thrombopoietic agents have in the standard treatment of thrombocytopenia remains to be defined. The course is one of waxing and waning severity, and spontaneous remission is rare. Associated diseases, such as lymphoma and systemic lupus erythematosus, must be ruled out before the diagnosis can be made. Quantitative laboratory tests of antiplatelet antibody may differentiate between patients who will favorably respond to therapy and those who will not. Hospitalization is recommended during the initial evaluation because the differential diagnosis is complex and the risk of bleeding is significant. Treatment usually includes corticosteroids, splenectomy, and in refractory cases, immunosuppressive therapy such as with cyclophosphamide, azathioprine, or vincristine. Plasmapheresis, androgens, immune globulin, anti-Rh(D), danazol, and colchicine have all met with varied success. Platelet transfusions are used only to control life-threatening bleeding because of increased antiplatelet antibody titers and short-lived hemo-
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
1584
of only 9 days. The number of transfusions directly correlates with the degree of thrombocytopenia. Current transfusion practice is to monitor platelet counts for every 10 U of RBCs and transfuse once the platelet count approaches 50,000/mm3.46
Thrombocytopathy Knowledge of abnormal platelet function as a clinical disorder has grown rapidly in recent years. The drug-induced form may be one of the most commonly seen causes of abnormal bleeding.36 Defects may occur at any level of platelet function, including adhesion, release, and aggregation. Adhesion Defects. The representative adhesion disorder is von Willebrand’s disease, which is more a factor VIII problem than a platelet deficiency. Platelets are normal in terms of their morphologic condition, number, release, and aggregation. The abnormal adhesion results not from the platelet but from an endothelium-based plasma deficiency of a factor VIII component (vWF) that permits platelet adhesion.51,52 Release Defects. Release defects include “storage pool” syndromes in which release is normal but amounts of adenosine diphosphate, calcium, and serotonin are decreased. Release defects may be congenital or acquired, as in systemic lupus erythematosus, alcoholism, or lymphoma. Drugs induce the most common release problem. Aspirin and related drugs block the enzyme cyclooxygenase, which participates in thromboxane A2 formation. Decreased release of thromboxane A2 results in decreased aggregation and less local vasoconstriction. Both may contribute to an increased risk of bleeding. Testing for this risk has been suggested by development of the postaspirin bleeding time as a screening test for hemostatic disorders. Aspirin is unique in that it permanently poisons this reaction for the life of the platelet in dosages of only 300 to 600 mg. Phenylbutazone and indomethacin affect function only while measurably circulating. A similar problem may occur in patients with uremia or dysproteinemia and as a rare inherited form.5,6,53 Aggregation Defects. Primary aggregation defects are associated with the rare recessive trait thrombasthenia. This platelet membrane abnormality may be detected by the lack of clot retraction during a 2-hour clot retraction test.20 Platelet Transfusions. Most platelet function disorders are not treated by platelet transfusion because its efficacy is questionable and alloimmunization may occur. Platelet transfusions are commonly indicated for primary bone marrow disorders (e.g., aplastic anemia or acute leukemia). Assessing the risk for spontaneous bleeding by using platelet counts is an imprecise science. Less mature platelets associated with peripheral consumption or sequestration are less likely to allow spontaneous hemorrhage than are those associated with primary bone marrow involvement. An estimate of functionality is combined with the platelet count for a better predictor of primary hemostasis potential. At counts below 50,000/mm3, a variable degree of risk exists, especially that associated with trauma, ulcers, or invasive procedures. At counts higher than 50,000/mm3, hemorrhage caused by platelet deficiency is unlikely. The transfusion threshold for platelets in trauma is not well defined and may be as high as 75,000/mm3 to 80,000/mm3. Spontaneous bleeding in the absence of surgery, trauma, or other risk factors may occur in patients with platelet counts less than 10,000/mm3.54
Thrombocytosis Thrombocytosis may be discovered in the ED. The reactive form is considered benign. The differential diagnosis (see
Box 120-6) should be considered when confronted with a platelet count higher than 600,000 to 1,000,000/mm3. The primary or autonomous state may be associated with bleeding or thrombosis. It is often an associated finding in patients with polycythemia vera, myelofibrosis, or chronic myelogenous leukemia. Suggested autonomous thrombocytosis requires a full hematologic evaluation.1,55
Disorders of the Coagulation Pathway The coagulation system accomplishes secondary hemostasis through a complex enzymatic cascade. The clinically significant disorders have a number of characteristic features that help differentiate them from platelet disorders, including the following:18 1. The bleeding source is often an intramuscular or deep soft tissue hematoma from small arterioles. 2. The congenital form of the disease occurs predominantly in men, often as a sex-linked inheritance. 3. Bleeding may occur after surgery or trauma but is delayed in onset up to 72 hours. 4. Epistaxis, menorrhagia, and gastrointestinal sources of bleeding are rare, whereas hematuria and hemarthrosis are common in severe cases. 5. The bleeding time is normal except in patients with von Willebrand’s disease. The PT and PTT are the basic laboratory diagnostic tools for the evaluation of coagulation disorders and can be used to organize the approach to their diagnosis.18
Abnormal Prothrombin Time and Other Tests Normal An elevated PT reflects an extrinsic pathway abnormality mediated through deficiency of factor VII. The hereditary form is caused by a rare autosomal recessive gene. The acquired form is commonly seen and may be a result of vitamin K deficiency, coumarin use, or liver disease. Because factor VII has the shortest half-life (3–5 hours) of the coagulation factors, it is the first to manifest a deficiency when its active form is underproduced. The PT is a sensitive gauge of hepatic function and the efficacy of warfarin administration. INRs calculate the prothrombin ratio raised to the power of an international sensitivity index for specific thromboplastin reagents. It is recommended with most warfarin therapy that the INR be maintained between 2.0 and 3.0.56-58
Abnormal Partial Thromboplastin Time and Other Tests Normal Two groups of inherited disorders manifest an isolated elevation in the PTT. The first group consists of the contact factors (e.g., XII [Hageman factor]), prekallikrein (Fletcher factor), and high-molecular-weight kinogen. They cause a benign disorder in which the PTT is elevated but the patient has no bleeding diathesis. These deficiencies exist as isolated laboratory abnormalities, and thus they should not be invoked as a cause of the patient’s bleeding problem. They may be specifically assayed when a precise diagnosis is necessary.11,17 The second group causes significant bleeding problems resulting from deficiencies of factors within the intrinsic coagulation system. They are the most common inherited abnormalities of the entire clotting system. Deficiencies of factors VIII, IX, and XI account for 99% of inherited bleeding disorders. Patients with active life-threatening bleeding who are thought to have a congenital bleeding disorder can be sup-
1585
and anticipation of potential life threats and admission of known bleeders for observation in selected circumstances. At one time, treatment of hemophilia-associated bleeding was a relatively common emergency medicine activity, but since 1975, hemophilia home therapy has increasingly been instituted. Therefore, many hemophiliacs now come to the ED only with complicated problems or trauma-related difficulties, and most are knowledgeable about their disease.60-63 Preparation. In preparing for the problem, the emergency physician should have updated information covering disease processes and current therapy. A cooperative effort should be made between the ED and the hematology service to generate a file of known hemophiliacs in the area who are monitored at the hospital. The file should include the primary physician, diagnosis, factor VIII activity level, blood type, presence of antihemophilic factor antibodies, and time of last hospitalization. A protocol should be developed for ordering and administering factor VIII. Replacement Therapy. The accepted therapy for hemophilia A is factor VIII replacement with cryoprecipitate or factor VIII:C concentrates. These concentrates are exposed to heat treatment or solvent-detergent mixtures to decrease transmission of hepatitis B, hepatitis C, and human immunodeficiency virus. In the past, the concentrate was made from fractionated freeze-dried antihemophilic factor and contained 250 to 1500 IU of factor VIII:C in a reconstituted volume. Factor VIII is also produced by recombinant DNA techniques and is considered to be the replacement product of choice. Recombinant-derived factor VIII is comparable to plasma-derived factor VIII in terms of characteristics and control of bleeding, but it has no discernible side effects. Factor VIII:C concentrates are commonly used in severe hemophilia and for home use. Cryoprecipitate is the cold precipitable protein fraction derived from fresh frozen plasma thawed at 1° C to 6° C. It was once the mainstay of hemophilia A therapy and may be used when noninfectious factor VIII concentrates are not available.63-67 Plasma-derived replacement therapies pose some risk for hepatitis C and hepatitis B. Persistent hepatitis B surface antigen occurs in the blood of 5% of hemophiliacs, whereas the anti-B surface antigen is found in 80%. This problem has been overshadowed by the association of acquired immunodeficiency syndrome with hemophilia. The association is related to blood product use, and although the total number is low, the incidence is high—3.6 per 1000 hemophilia A patients.64,65 Therapy for a bleeding episode includes a number of considerations: the circumstances in which factor VIII is given, the dosage, the timing of maintenance, the duration of the dosage, the presence of antibodies, and the means of gauging effectiveness. Tables 120-1 and 120-2 include guidelines for the recommended treatment in a variety of circumstances. Most importantly, the emergency physician should believe patients who say that they are bleeding and institute early therapy.61,63 The response to therapy can be monitored by clinical improvement, a decreasing PTT, and, optimally, serial factor VIII:C activity levels. The infusion of 1 U of factor VIII per kilogram increases factor VIII levels by 2%. The lack of a response to factor VIII administration should raise the possibility of circulating antibodies. All hemophiliacs should be screened for the development of these antihemophilic factor antibodies when they are given in-hospital therapy or if their condition becomes refractory to home therapy. The 7 to 20% of patients in whom these IgG antibodies develop usually have a severe deficiency necessitating multiple factor VIII transfusions. The treatment may be complex, and hospitalization is
Chapter 120 / Disorders of Hemostasis
ported with fresh frozen plasma, 15 mL/kg, while diagnostic studies are being performed. The risk of viral transmission of hepatitis B or C or human immunodeficiency virus must be considered. In a patient with a prolonged PTT and a lifelong history of bleeding, the most important tests in initiating the differential diagnosis are factor VIII and factor IX assays. This test measures the ability of the patient’s plasma to correct the prolonged PTT of plasma deficient in factor VIII. This ability is compared with that of normal plasma and the result is given as a percentage of normal. The test measures the procoagulant activity of factor VIII but does not discriminate between abnormal activity resulting from abnormal factor VIII or low levels of normal factor VIII. The two forms of this deficiency are hemophilia A and von Willebrand’s disease.9 Hemophilia A. Hemophilia A is caused by a variant form of factor VIII that is present in normal levels but lacks a clotpromoting property. The incidence is 60 to 80 persons per million population. Of known cases, 70% have been found to have a sex-linked recessive nature; that is, the disease is carried on the X chromosome at location Xq28. Factor VIII circulates in plasma in very low concentration and is normally bound to vWF. The source of factor VIII production is uncertain, but the liver is thought to be a significant source because hemophilia A can be corrected by liver transplantation. A female carrier mating with a normal man would be predicted to pass the disease to half her sons. Likewise, a male hemophiliac would have all normal sons and all carrier daughters. The remaining 25 to 30% of cases of the disease are believed to result from a spontaneous genetic abnormality. The familial form has a remarkable consistency of severity from generation to generation, although the degree of severity has considerable variation. This severity may be directly related to the level of factor VIII coagulant (factor VIII:C) activity. Cases with less than 1% activity are severe, with a tendency toward spontaneous bleeding. Cases with 1 to 5% activity are moderate, with rare spontaneous bleeding but increased problems with surgery or trauma. Cases with 5 to 10% activity and above are considered mild, with little risk of spontaneous bleeding but still with hazards after trauma and surgery. A number of hemophiliacs may have activity above 10% but have few problems under normal conditions. The PTT may lack sensitivity for this group because it is significantly prolonged only at factor VIII: C levels less than 35 to 40%.9,59-62 The disease is seen as a disorder of secondary hemostasis with a characteristic pattern of bleeding. Bleeding can occur anywhere, but deep muscles, joints, the urinary tract, and intracranial sites are the most common. Recurrent hemarthrosis and progressive joint destruction are major causes of morbidity in hemophilia. Intracranial bleeding is the major cause of death in all age groups of hemophiliacs. Mucosal bleeding such as epistaxis and oral bleeding or menorrhagia is rare unless the disease is associated with von Willebrand’s disease or platelet inhibition, such as with aspirin use. Gastrointestinal bleeding is rare unless peptic ulcer disease is also present. Trauma is a common initiator of bleeding in all stages of severity. This potential hazard must be viewed expectantly in all hemophiliacs because late bleeding may occur, usually by 8 hours but potentially up to 1 to 5 days, and, rarely, even longer after traumatic injury.60-62 Management of Hemophilia A. Comprehensive management of hemophilia involves a team effort of physicians, specialized nurses, physical therapists, social workers, the patient, and the patient’s family. The therapeutic responsibility of the emergency physician consists of three areas: preparation for and identification of the problem, initial evaluation, and admission of new bleeders; replacement therapy for bleeding episodes;
1586
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
Table 120-1 Recommended Factor VIII Therapy for Specific Problems in Hemophilia TYPE OF BLEEDING
Skin Abrasion Laceration Superficial Deep Nasal epistaxis Spontaneous Traumatic Oral Mucosa or tongue bites Traumatic (laceration) or dental extraction
Soft tissue/muscle hematomas Hemarthrosis Early
INITIAL DOSAGE
DURATION
COMMENT
None Usually none; if necessary, treat as minor
None None
Treat with local pressure and topical thrombin Local pressure and anesthetic with epinephrine may benefit; watch 4 hours after suturing; reexamine in 24 hours
Minor bleeding (12.5 mg/kg)
Single-dose coverage
May need hospitalization for observation; repeat may be necessary for suture removal
Usually none; may need to be treated as mild bleeding Moderate bleeding (25 mg/kg)
None
Uncommon; consider platelet inhibition; treat in usual manner
Up to 5–7 days
Trauma-related bleeding can be significant
Usually none; treat as minor if persists Moderate (25 U/kg) to severe (50 U/kg)
Single dose
Commonly seen
Single dose; may need more
Moderate (25 U/kg) to severe (50 U/kg)
2–5 days
Saliva rich in fibrin lytic activity; oral ε-aminocaproic acid (Amicar) may be given at 100 mg every 6 hr for 7 days to block fibrinolysis; check contraindications; hospitalize patients with severe bleeding May be complicated by local pressure on nerves or vessels (e.g., iliopsoas, forearm, calf)
Mild (12.5 U/kg)
Single dose
Late or unresponsive cases of early hemarthrosis Hematuria
Mild to moderate (25 U/kg)
3–4 days
Mild (12.5 U/kg)
2–3 days
Major bleeding
Major bleeding (50 U/kg)
7–10 days or 3–5 days after bleeding ceases
Treat as earliest symptom (pain); knee, elbow, ankle more common Arthrocentesis rarely necessary and only with 50% level coverage; immobilization is critical point of therapy Urokinase, the fibrinolytic enzyme, is in urine; with persistent hematuria an organic cause should be ruled out In head trauma, therapy should be given prophylactically; early CT scan of head recommended for all
Gastrointestinal severe bleeding Neck/sublingual Retroperitoneal Intra-abdominal Major trauma Head injury (see text) Central nervous system bleeding (see text) Surgical procedure CT, computed tomography
necessary. A variety of therapies have been considered, including “overwhelming” factor VIII doses, exchange plasmaphe resis, immunosuppressive therapy, and the infusion of prothrombin complexes containing activated clotting factors. Other recommended therapies include porcine factor VIII, which has less cross-reactivity with the human product, and probably in the future, recombinant activated factor VIIa. Recombinant factor VIIa has been used in the treatment of some nonhemophiliac patients with serious or intractable bleeding but has not proved more effective than placebo. Until more data are available, the efficacy of recombinant factor VIIa in settings outside of congenital coagulation disorders remains to be determined.63,68-71 Acquired IgG antihemo-
philic factor antibodies may exist in nonhemophiliac patients. They can occur in the postpartum period, as immunologic reactions to penicillin or phenytoin, and in association with systemic lupus erythematosus, rheumatoid arthritis, or inflammatory bowel disease. The diagnosis is made by the occurrence of an acquired hemophilia-like syndrome with positive antibody titers in the appropriate setting. The “lupus anticoagulant” is unique in that it may be associated with an increased risk for thrombosis, as well as a hemorrhagic diathesis.61,72 Desmopressin acetate has been shown to increase levels of factors VIII:C and VIII:Ag in patients with hemophilia A and in some with von Willebrand’s disease. It is given intrave-
1587
Table 120-2 Dosage of Factor VIII (Antihemophilic Factor) DESIRED FACTOR VIII LEVEL (%)
INITIAL DOSE (U/KG)
Mild Moderate Severe
5–10 20–30 50 or greater
12.5 25 50
Standard Calculation 1. Patient’s plasma volume (50 mL/kg × weight in kg) × (Desired level of factor VIII [percent]) − (Present level of factor VIII [percent]) = Number of units for initial dose. 2. In emergency therapy, the present level of factor VIII is assumed to be zero. 3. One unit is the activity of the coagulation factor present in 1 mL of normal human plasma. 4. Because the half-life of factor VIII is 8–12 hr, the desired level is maintained by giving half the initial dose every 8–12 hr. 5. Cryoprecipitate is assumed to have 80–100 U of factor VIII:C per bag; factor VIII:C concentrates list the units per bottle on the label.
nously at 0.3 µg/kg per dose. Benefits are primarily noted in patients with mild to moderate disease and last for 4 to 6 hours.73,74 Prophylaxis. The anticipation of delayed bleeding in patients with hemophilia may necessitate admission and observation for a variety of trauma-related injuries. Candidates for prophylactic admission are patients with deep lacerations; those with soft tissue injuries in areas where the pressure from a developing hematoma could be destructive, such as in the eye, mouth, neck, back, and spinal column; and patients with a history of major trauma forces without injury. Head trauma is potentially life-threatening to hemophiliacs, and central nervous system bleeding is the major cause of death for patients in all age groups. Studies find a 3 to 13% risk of intracranial hemorrhage, yet no patient given replacement therapy within 6 hours had intracranial bleeding. It is recommended that patients who sustain head trauma but who have normal CT scans have factor VIII therapy initiated to greater than 50% activity level.75,76 All hemophiliacs with head trauma should be considered for admission and their primary physician and hematologist consulted early. Gene therapy represents a potential development in the treatment of hemophilia. With cloning of the genes encoding factor VIII, the possibility exists for either a partial or complete cure of hemophilia. The goal of gene therapy is not to restore factor levels to normal but rather to convert from a severe to a mild phenotype and dramatically improve clinical outcomes. Early studies are encouraging. Although genetic testing and counseling are currently available, no genetic therapies for hemophilia A are available at present.61,77-80 von Willebrand’s Disease. To understand von Willebrand’s disease, it is helpful to review the nomenclature used to refer to factor VIII in some centers. Factor VIII has at least three activities. First is its antihemophilic, or coagulant, activity, VIII:C. All references to factor VIII in this chapter thus far have been to this activity. A second activity supports platelet adhesion and in vitro aggregation with the antibiotic ristocetin; it is called von Willebrand’s factor activity, or VIII/vWF. A third component reacts with rabbit antibodies to factor VIII. It is termed the factor VIII antigen, or VIII:Ag, and relates to the measured plasma level rather than the activity of factor VIII. The antigen and cofactor activity for platelet function are structurally related.52,81 von Willebrand’s disease has both decreased factor VIII:Ag levels and decreased VIII:C activity secondary to underproduction. The patient’s platelets are normal in number, morphologic condition, and other functions, but in the absence of circulating factor VIII/vWF, their
Miscellaneous Coagulation Disorders A number of other disorders may be caused by a deficiency in the common coagulation pathway. An altered fibrinogen level or abnormal function is a relatively common cause. Patients with this deficiency also have an abnormal thrombin time. The inherited forms are rare. The acquired forms have been related to fibrin-blocking substances and hypofibrinogenemia, which are found most often in cases of DIC and dysfibrinogenemia associated with macroglobulinemia, multiple myeloma, and hepatoma. In the context of emergency medicine, fibrinogen’s most important role relates to its activity in DIC.
Chapter 120 / Disorders of Hemostasis
BLEEDING RISK
adhering properties are diminished. Von Willebrand’s disease is the most common hereditary bleeding disorder, with an estimated prevalence of 1%. The disease occurs in 5 to 10 persons per million population as an autosomal dominant trait with a variable penetrance pattern. A rare X-linked inheritance has been described.52,81,82 Manifestations of von Willebrand’s disease are usually milder and less crippling than those of hemophilia. The factor VIII:C level is in the 6 to 50% range. Bleeding sites are predominantly mucosal (e.g., epistaxis) and cutaneous. Hemarthroses are rare, but menorrhagia and gastrointestinal bleeding are common. Laboratory differentiation from hemophilia A includes an abnormal bleeding time, a decreased level of factor VIII:Ag, and abnormal platelet aggregation with ristocetin.83 In patients with severe disease, replacement therapy with factor VIII in the form of intermediate purity factor VIII concentrate is the method of choice. The initial dose is 20 to 30 IU/kg every 12 hours to keep vWF levels at 50% or to control bleeding. A unique response to the transfusion of plasma components in patients with von Willebrand’s disease is the stimulation of a progressive increase in VIII:C activity that lasts 12 to 40 hours. After the initial dose, fewer units are necessary, and longer dosage schedules may be followed by a clinical response and a combination of factor VIII:C activity and serial bleeding times. In extreme circumstances without alternatives, fresh frozen plasma may be used. A factor VIII concentrate (Humate-P) has also demonstrated sufficient VIII/vWF to treat the disease.54,84 Drug therapy with desmopressin is of benefit in patients with mild to moderately severe von Willebrand’s disease. It is most useful in the common form of the disease and ideally would be given in consultation with a hematologist.35,85,86 Hemophilia B (Christmas Disease). Hemophilia B is a deficiency of factor IX activity. Its genetic pattern and clinical findings are indistinguishable from those of hemophilia A, but its incidence is only a fifth that of hemophilia A. Factor IX is a vitamin K–dependent glycoprotein. Its deficiency is diagnosed by a factor IX assay, usually after the factor VIII:C assay is found to be normal. The replacement schedule for factor IX is similar to that for hemophilia A, but a purified factor IX concentrate or recombinant factor IX preparation is used. The plasma prothrombin complex (factors II, VII, IX, and X) and fresh frozen plasma are also useful, but they pose a higher risk of viral transmission and venous or arterial thrombosis. The maintenance dosage schedule is increased to every 24 hours because of the longer half-life of factor IX. Clinical concerns and treatment strategies associated with hemophilia A also apply to hemophilia B.63,87,88 Similar to hemophilia A, gene testing and counseling are available. Gene therapy in animals has demonstrated promising results, and preliminary results from a human study suggest that the severity of hemophilia B can be altered and improved by gene manipulation.77,79,89,90
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
1588
The other components of the common pathway (factors II, V, and X) have rare inherited deficiencies. The acquired forms are far more common and relate to vitamin K deficiency (decreased factor II, VII, IX, and X activity), warfarin use (same factors as with vitamin K deficiency), hepatic insufficiency (potentially all factors except VIII), and massive transfusion of stored blood (low in factors V and VIII and platelets). Disseminated Intravascular Coagulation. DIC is a relatively common acquired coagulopathy. Its ubiquitous nature, multiple origins, and potentially devastating sequelae, balanced by an effective mode of therapy, make early diagnosis of this hematologic process critical. It is most often encountered in the critical care setting. Hemostasis is achieved by a fine balance between procoagulants and inhibitors and thrombus formation and lysis. The balance may be disturbed by pathologic processes that result in an out-of-control coagulation and fibrinolytic cascade within the systemic circulation. The following occurs in this abnormal clotting sequence: 1. Platelets and coagulation factors are consumed, especially fibrinogen and factors V, VIII, and XIII. 2. Thrombin is formed, and it overwhelms its inhibitor system and acts to accelerate the coagulation process and directly activate fibrinogen. 3. Fibrin is deposited in small vessels in multiple organs. 4. The fibrinolytic system by means of plasmin may lyse fibrin and impair thrombin formation. 5. Fibrin degradation products are released and affect platelet function and inhibit fibrin polymerization. 6. Coagulation inhibition levels (e.g., antithrombin III, protein C, and tissue factor pathway inhibitor) are decreased. The clinical consequence of these processes is the lifethreatening combination of a bleeding diathesis from loss of platelets and clotting factors, fibrinolysis, and fibrin degradation product interference; small vessel obstruction and tissue ischemia from fibrin deposition; and RBC injury and anemia from fibrin deposition. The condition must be considered in any patient in whom purpura, a bleeding tendency, and signs of organ injury, particularly the central nervous system and kidney, develop. This broad description is further confused clinically by the variable acuteness and intensity of intravascular clotting, the effectiveness of fibrinolysis, and other systemic manifestations of the initiating disease.91-93 The clinical diagnosis is necessarily supported by laboratory tests. The tests recommended in Table 120-3 usually confirm the presence of DIC. Other tests (e.g., specific degradation products of fibrin, particularly D-dimer, and fibrinogen) can confirm the diagnosis. These tests are rarely available in the ED. Two conditions that may simulate DIC are severe liver disease and primary fibrinolysis. Liver disease of this severity is usually manifested by clinical jaundice and splenomegaly. Primary fibrinolysis is a rare disorder that affects fibrinogen and fibrin but generally leaves the coagulation components (platelets, factor V, and factor VIII) in the low normal range. The paracoagulation test is negative, and the euglobulin lysis time is rapid.92,93 When planning therapy, the emergency physician must remember that defibrination is always secondary to a serious underlying pathologic process. Once the diagnosis is confirmed, the initial treatment is focused on reversing the triggering mechanism. Many episodes of DIC are self-limited, such as in a transfusion reaction, or compensated, such as associated with a tumor mass, and do not require intervention other than support.91,92
Diagnosis of Disseminated Table 120-3 Laboratory Intravascular Coagulation TEST
FINDING
PATHOPHYSIOLOGY
Peripheral smear
Low platelets, schistocytes, RBC fragments Low (usually 1500 U/L) Advanced disease with abdominal involvement Preexisting renal dysfunction Post-treatment renal failure Acidic urine Concentrated urine Preexisting volume depletion Young age
calcemia. Hyperuricemia with resultant urate nephropathy is the most commonly recognized metabolic cause of renal insufficiency.24 The kidney provides the primary mechanism for excretion of uric acid, potassium, and phosphate. Rapid proliferation of tumor cells may exceed the removal rate of the respective substances, resulting in increased levels. In fact, increased quantities of these substances have been observed in patients undergoing rapid lysis of chemosensitive tumors. The integrity of renal function is a critical factor in determining the degree of metabolic derangements. In patients with preexisting renal insufficiency, the metabolic derangements of acute tumor lysis are more likely to be severe. However, even when renal function appears normal at the start of treatment, the rapid lysis of certain tumors may overwhelm the excretory capacity of the kidney. Similar to hyperuricemia, hyperphosphatemia may also cause renal failure. A possible mechanism is precipitation of calcium phosphate within the kidney.24 Hyperkalemia, along with a contributing hypocalcemia, may result in life-threatening ventricular dysrhythmias.22 Hypocalcemia may also cause neuromuscular instability with muscle cramps and occasionally tetany. Confusion and convulsions also have been described in case reports.24
Management In approaching a patient with potential TLS, it is “easier to stay out of trouble than get out of trouble.”24 The main principles of TLS management are (1) identification of high-risk patients with initiation of preventive therapy and (2) early recognition of metabolic and renal complications with prompt supportive care, including hemodialysis. Most of the complications can be readily managed when they are recognized early; however, delay in recognition and initiation of treatment of TLS can be life-threatening.25 Chemotherapy should be delayed, if possible, until metabolic disturbances, especially prerenal azotemia and hyperuricemia are corrected. Initial management is aimed at the control of preexisting hyperuricemia with hydration, allopurinol, and alkalinization of the urine to a pH greater than 7. Diuretics are added if necessary, and frequent monitoring of electrolytes, calcium, and phosphorus is essential.
Hydration Volume depletion is a major risk factor for TLS and must be corrected vigorously. Rapid intravenous (IV) hydration is the single most important intervention. Hydration not only helps correct electrolyte disturbances by diluting extracellular fluid, but it also increases intravascular volume. Increased volume enhances renal blood flow, glomerular filtration rate, and urine volume, which consequently decreases the concentration of solutes in the distal nephron and medullary microcirculation. Continuous infusion rates as high as 4 to 5 L/day yielding urine volumes of at least 2 to 3 L/day should be given unless the patient’s cardiovascular status indicates impending volume overload. Ideally, IV hydration in high risk patients should begin 24 to 48 hours prior to initiation of cancer therapy and continue for 48 to 72 hours after completion of chemotherapy.25
Hyperuricemia Allopurinol is a xanthine oxidase inhibitor and is given to reduce the conversion of nucleic acid by-products to uric acid in order to prevent urate nephropathy and subsequent oliguric
1595
BOX 121-5 Criteria for Instituting Hemodialysis
Pathophysiology renal failure. Because allopurinol inhibits the synthesis of uric acid but has no effect on preexisting uric acid, uric acid levels usually do not fall until after 48 to 72 hours of treatment. Allopurinol usually is given orally between 300 and 600 mg/ day for prophylaxis and 600 to 900 mg/day for treatment of TLS.25 Rasburicase (recombinant urate oxidase) is a newer therapy that can be used when the uric acid levels cannot be lowered sufficiently by standard approaches. This modality of treatment is rarely initiated by emergency physicians. Rasburicase is useful in cases of hyperuricemia. Humans don’t express urate oxidase; urate oxidase catalyses the conversion of poorly soluble uric acid to soluble allantoin. By converting uric acid to water-soluble metabolites, it effectively and rapidly decreases plasma and urinary uric acid levels. Unlike allopurinol, rasburicase does not increase the excretion of xanthine and other purine metabolites; therefore, it does not increase tubule crystallization of these compounds, thereby decreasing the risk of urate nephropathy.25 Most articles agree that it is wise to alkalinize the urine as a prophylactic measure against hyperuricemia, but caution is advised should hyperphosphatemia and hypocalcemia develop. In patients with concomitant hyperphosphatemia, alkalinization favors precipitation of calcium/phosphate complexes in the renal tubules. Furthermore, alkali therapy may aggravate manifestations of hypocalcemia such as tetany.22,26 Although alkalinization increases the solubility of uric acid, the primary means of uric acid control is hydration and diuresis to maintain adequate urinary flow.24,26 The use of furosemide or mannitol for osmotic diuresis has not proven to be beneficial as front-line therapy. In fact, these modalities may contribute to uric acid or calcium phosphate precipitation in renal tubules in a volume-contracted patient. Instead, diuretics should be reserved for well-hydrated patients with insufficient diuresis, and furosemide alone should be considered for the normovolemic patient with hyperkalemia or for the patient with evidence of fluid overload. If TLS develops and it is refractory to the previously mentioned treatments, hemodialysis should be considered as a potentially lifesaving measure. This therapy is effective in lowering uric acid, potassium, and phosphate levels, as well as in controlling uremic symptoms. See the suggested criteria for instituting hemodialysis in Box 121-5. The prognosis is good in the absence of renal failure. If renal failure exists and hemodialysis of 5 to 7 days is necessary, the prognosis is grave. With aggressive management, the incidence of renal and metabolic complications of cytoreductive therapy may be decreased.
■ HYPERVISCOSITY SYNDROME Hyperviscosity syndrome (HVS) refers to the clinical sequelae of increased blood viscosity. Viscosity is the resistance that a liquid exhibits to the flow of one layer over another. Excessive elevations in certain paraproteins (circulating immunoglobu-
HVS is most commonly associated with plasma cell dyscrasias (the paraproteinemias) and is due to the large size of the excess immunoglobulin M (IgM) paraproteins in these disorders. Waldenström’s macroglobulinemia is the most common cause and accounts for about 85 to 90% of cases of HVS. Less frequently, the disease can occur in multiple myeloma (especially with myeloma proteins of the IgA and IgG3 types). Other causes include cryoglobulinemia, a benign hyperglobulinemia of the IgM-IgG type, and leukemias.27-29 The blastic phase of chronic myelogenous leukemia, chronic granulocytic leukemia, and the blast cell crisis of acute lymphoblastic and nonlymphoblastic leukemias also commonly cause HVS.27,28 Other more benign causes include leukemoid reaction, polycythemia vera, and the accumulation of abnormal hemoglobins in sickle cell disease. The incidence of HVS in Waldenström’s macroglobulinemia is found to be approximately 20%, in IgG myeloma approximately 4.2%, and in IgA myeloma as high as 25%.28 The inherent physiochemical properties of the dysproteinemias along with extremely high concentrations of these proteins seem to predispose to the development of hyperviscosity. Paradoxically, HSV also has been reported in κ-light-chain disease owing to a greater tendency to form unstable, highly polymerized circulating aggregates. The etiologic factor most responsible for HVS in the leukemias appears to be leukocytosis with WBC counts in excess of 100,000, usually accompanied by blast forms exceeding 100,000 in the peripheral smear. The clinical manifestations of HVS become most apparent when the serum viscosity relative to water is greater than 4 to 5, normal serum viscosity relative to water being 1.4 to 1.8.27-29
Clinical Features A symptomatic triad of mucosal bleeding, visual disturbances, and neurologic manifestations is a classic presentation of HSV. Visual disturbances and, on occasion, visual loss may occur with retinopathy characterized by venous engorgement (e.g., “sausage-link” or “boxcar” segmentation), which is also seen in the bulbar conjunctiva, microaneurysms, hemorrhages, exudates, and occasionally papilledema. Persistent bleeding di atheses from mucosal surfaces, especially nasal mucosa, the gastrointestinal (GI) tract, and sites of minor surgery or trauma, even in the face of a normal platelet count, are common. Other clinical findings encompass myriad neurologic disturbances, including headache, dizziness, jacksonian and generalized seizures, somnolence, lethargy, coma, auditory disturbances (including hearing loss), and hypotension. Constitutional symptoms of fatigue, anorexia, and weight loss that are nonspecific early on are commonly associated with the underlying malignancy or with numerous electrolyte disturbances related to the underlying malignant process. Cardiopulmonary findings, including acute respiratory failure and hypoxemia, congestive heart failure, myocardial infarction, and valvular abnormalities have all been reported. Renal insufficiency and
Chapter 121 / Selected Oncologic Emergencies
Serum potassium 6 mEq (6 mmol/L) Serum uric acid 10 mg/dL (590 mol/L) Serum creatinine 10 mg/dL (880 mol/L) Serum phosphorus 10 mg/dL (phosphate 3.2 mmol/L) or rapidly rising To reduce volume overload Symptomatic hypocalcemia
lins) or cellular blood components (leukocytosis, erythrocytosis, and thrombocytosis) can result in elevated serum viscosity and the development of significant sludging, decreased perfusion of the microcirculation, and vascular stasis. The outcome of these pathophysiologic events leads to the development of HVS, which requires urgent medical therapy to forestall or reverse the effects of sludging in the microcirculation of the CNS, visual system, and cardiopulmonary system.27
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
1596
failure may be a complication of the syndrome and will exacerbate existing clinical findings secondary to the expanded plasma volume.27,28 The laboratory evaluation of the patient with suggested HVS should include a coagulation study and renal, electrolyte, and differential white count profiles. Serum and urine protein electrophoresis should be done in all cases of suggested dysproteinemias; the diagnosis is supported by a large spike on the serum electrophoresis. A clue to the presence of hyperviscosity may be the inability of the laboratory to perform chemical tests on the blood because of the serum stasis and increased viscosity that jams analyzers. In multiple myeloma significant hypercalcemia may also occur, and with high M-protein fractions a factitious hyponatremia may be present. The diagnosis may be also entertained when a patient is brought to the emergency room in a stupor or coma and anemia and rouleaux formation are found on the peripheral smear.30 Because HVS is often a presenting characteristic of dysproteinemias and leukemias with blastic transformation and because a history of previously documented disease is often absent, this syndrome should be considered in patients with unexplained somnolence and coma.
Management Emergency leukapheresis or plasmapheresis is the definitive treatment. Temporizing measures provided by the emergency physician should focus on adequate rehydration and diuresis. An immediate temporizing measure in a patient with frank coma and an established paraproteinemia is a two-unit phlebotomy with replacement of the patient’s RBCs with physiologic saline.27-30 After plasmapheresis or leukapheresis has adequately alleviated the clinical findings, chemotherapeutic modalities can be used.
■ HYPERURICEMIA Hyperuricemia is a serious and well-known consequence of certain malignant disorders, which, if recognized early, can result in a significant decrease in morbidity for the cancer patient. The major source is cell breakdown, and its major excretory pathway is via the kidneys.
Pathophysiology The pathogenesis of hyperuricemia results from either the increased production or decreased excretion of uric acid, or both. Increased production of uric acid commonly results from accelerated generation of uric acid through purine metabolism as a result of rapid dissolution of neoplastic tissues (cell death) following chemotherapy or radiation therapy of undifferentiated lymphomas or lymphoblastic lymphomas and rapid cell proliferation and turnover with acute lymphoblastic leukemias. In addition, hyperuricemia may be seen with multiple myeloma and occasionally with disseminated metastatic carcinoma. With massive release of precursors, uric acid levels rise precipitously and may reach levels as high as 15 to 20 mg/dL. As a result, uric acid crystals form in the highly concentrated and acidified urine of the distal tubules, intrarenal obstruction follows, and acute renal failure ensues.24,31 Chronic, moderately elevated levels of the serum uric acid may result in renal colic, obstructive uropathy, or chronic renal failure. Either uric acid renal calculi or interstitial deposits of sodium urate may develop. This situation is associated with neoplastic overproduction of uric acid precursors. Polycythemia vera, myeloid metaplasia, mast cell disease, and chronic
granulocytic leukemia are often associated with this type of hyperuricemia. Decreased excretion may be a result of underlying renal insufficiency or as a consequence of precipitation of urates in the renal tubules, parenchyma, or ureters with subsequent development of renal insufficiency and further reduction in excretion of uric acid. Three types of renal diseases are attributable to hyperuricemia: acute hyperuricemic nephropathy, uric acid nephrolithiasis, and gouty nephropathy.
Clinical Features Hyperuricemia can occur with or without symptoms. Symptoms may be associated with the underlying malignancy. Hyperuricemia precipitated or aggravated by therapy of these diseases may occur as an isolated metabolic disturbance or may be accompanied by other manifestations of the TLS (see previous discussion on TLS). If an underlying neoplastic disease has been diagnosed, the possibility of hyperuricemia should be investigated before, during, and after treatment with chemotherapy or radiation. The hyperuricemia should be treated to prevent renal damage. In patients with urate stones and hyperuricemia, examination of the peripheral blood may provide evidence of an underlying myeloproliferative disorder. Acute oliguria following chemotherapy or radiation therapy suggests the diagnosis of hyperuricemia, and the uric acid level in the blood often far exceeds that associated with acute renal failure. A number of benign diseases are associated with hyperuricemia that may coexist with neoplasia. These include hereditary gout, hyperparathyroidism, psoriasis, sarcoidosis, and renal failure of any cause. The long-term administration of certain drugs may lead to elevation of the serum uric acid level. Various diuretics, including thiazides and furosemide, are important examples.24,32 From a therapeutic standpoint, however, the finding of hyperuricemia obviates the importance of the primary cause; the therapy is the same.
Management When possible, hyperuricemia should be treated before chemotherapy or radiation therapy, especially with bulky tumors or if the serum uric acid level is borderline or increased. If a uric acid elevation of more than 9 mg/dL is found, allopurinol, fluids, and alkalinization of the urine should be initiated. If possible, this regimen should be started a day or two before the initiation of chemotherapy or radiation treatment. Patients with histories of gouty arthritis should also receive colchicine (0.6 mg orally twice a day) to avoid the acute attacks that can be associated with allopurinol administration. Patients should be kept well hydrated. In patients with acute distal tubular uric acid obstruction, treatment includes the administration of allopurinol, together with the fluid and electrolyte management used in other forms of acute renal failure. If hyperuricemia is secondary to malignancy, cytolytic therapy should be stopped. Allopurinol in dosages of 300 to 600 mg/day usually causes a decrease in the serum uric acid level in approximately 3 days, so its administration should be started 2 or 3 days before cytolytic therapy, if time permits. Hydration is vital in maintaining a urine output above 2 L/day. Again, rasburicase (recombinant urate oxidase) is a newer therapy that can be used when the uric acid levels cannot be lowered sufficiently by standard approaches (see previous discussion on TLS). Alkalinization to keep the urine pH above 7 can be accomplished by administering sodium bicarbonate (9–12 g/day). Diuretics are to be used as needed. Acetazolamide (Diamox)
■ HYPERCALCEMIA Hypercalcemia occurs in approximately 20 to 40% of cancer patients and is the most common life-threatening metabolic disorder associated with cancer.33 It affects multiple organ systems and induces a variety of pathophysiologic events that may be more immediate threats to life than the cancer itself. For the purpose of this discussion, we discuss nonparathyroid hormone-mediated hypercalemia, which is associated with malignancy.34
Pathophysiology Two mechanisms have been proposed to explain the development of hypercalcemia associated with malignancy. The first mechanism involves patients with metastatic bone involvement. This hypercalcemia is most likely associated with the release of calcium and phosphate caused by associated increased osteoclastic activity within the bone. The second mechanism involves those patients with no bone disease. A variety of tumor-produced hormone-like substances have been associated with the development of hypercalcemia, including parathyroid hormone, prostaglandins, and peptides, all of which affect bone turnover. Hypercalcemia is a common feature of many malignancies but most often complicates cancer of the breast, lung, head, and neck, as well as multiple myeloma and leukemia. Bony metastases are not a prerequisite for hypercalcemia and when present do not necessarily cause hypercalcemia. In patients who are hypercalcemic from squamous cell lung cancer, only one in six has bone metastases. In small-cell lung carcinoma; hypercalcemia is almost never seen, despite the presence of bone marrow metastases in 20 to 50% of cases. A complex interaction of various substances (parathyroid hormone, prostaglandins, peptides, steroids, osteoclastic factors) appears to result in both increased bone synthesis and degradation. The exception is multiple myeloma, in which bone destruction is accompanied by minimal bone synthesis. Other entities that cause hypercalcemia are listed in Box 121-6.35,36 Rarer still are factitious hypercalcemia, idiopathic hypercalcemia of infancy (with elfin facies), familial hypocalciuric hypocalcemia, and hypercalcemia from pheochromocytoma or periostitis
Clinical Features The development of symptoms of hypercalcemia is nonspecific. There is little correlation between serum calcium levels and the presence and severity of symptoms. Acute hypercalcemia results in marked CNS effects ranging from personality changes (depression, paranoia, lethargy, somnolence) to coma. With chronic hypercalcemia, symptoms include a history of anorexia, nausea, vomiting, constipation, polyuria, polydipsia,
1597
BOX 121-6 Non-neoplastic Causes of Hypercalcemia Hyperparathyroidism Hyperthyroidism Renal insufficiency (diuretic phase of acute renal failure, after transplantation, secondary hyperparathyroidism) Drugs (thiazide diuretics, lithium, and calcium carbonate) Hypervitaminosis (A and D) Acute adrenal insufficiency Immobilization (Paget’s disease, fracture, paraplegia) Acromegaly Myxedema Milk-alkali syndrome Sarcoidosis Benign monoclonal gammopathy Rarer still are factitious hypercalcemia, idiopathic hypercalcemia of infancy (with elfin facies), familial hypocalciuric hypercalcemia, and hypercalcemia from pheochromocytoma or periostitis
BOX 121-7
Common Signs and Symptoms of Hypercalcemia in Malignancy
General Itching Neurologic Fatigue, muscle weakness, hyporeflexia, lethargy, apathy, disturbances of perception and behavior, stupor, coma Renal Polyuria, polydipsia, renal insufficiency Gastrointestinal Anorexia, nausea, vomiting, constipation, abdominal pain Cardiovascular Hypertension, dysrhythmias, digitalis sensitivity and memory loss. The signs, symptoms, and complications of hypercalcemia are summarized in Box 121-7. In patients with carcinoma, any of these symptoms should suggest the diagnosis of hypercalcemia, but the emergency physician should be particularly alert to the possibility of hypercalcemia in any cancer patient with lethargy or a change in mental status. Many may also have electrolyte abnormalities such as hypokalemia and dehydration. Thus evaluation of serum electrolytes should accompany the measurement of serum calcium, phosphorus, albumin, and alkaline phosphate. In general, a serum calcium level above 14 mg/dL constitutes a medical emergency. In chronic hypercalcemia, one may see patients with blood calcium levels as high as 15 mg/dL with only mild symptoms. With an acute onset, one can see patients comatose at a level of only 12 to 13 mg/dL.24,31-34 Many benign conditions can result in hypercalcemia. The most common are hyperparathyroidism and Paget’s disease of bone. Clinical features include a long history of hypercalcemia symptoms, particularly renal stones. Chronic changes on bone films, such as subperiosteal reaction and cysts or a “groundglass” appearance of the skull, suggest hyperparathyroidism. Diagnosis of Paget’s disease rests in biopsy results. Vitamin D excess, milk-alkali syndrome, and adrenal insufficiency are other common causes in the differential diagnosis of hypercalcemia.24,31 The acute onset of severe hypercalcemia or chronic exposure of the renal tubules to elevated calcium levels may reduce
Chapter 121 / Selected Oncologic Emergencies
in doses of 1 g/day usually alkalinizes the urine temporarily until allopurinol becomes effective. If oliguria occurs, IV mannitol may be started with 12.5 g of a 20% solution given over 3 minutes to keep urine output more than 250 mL/hr. The dose of mannitol is limited to 100 g/24 hours to avoid clinical features resembling water intoxication. If these measures fail, peritoneal dialysis or hemodialysis or flushing the ureters via retrograde catheters may be considered. Clearly, prevention of this complication is far better than treatment. The cancer patient who comes to the ED with renal colic warrants careful evaluation for hyperuricemia. The prognosis depends on the underlying malignancy and degree of renal failure.24,31
1598
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
the glomerular filtration rate and renal blood flow, resulting in acute renal failure.24
Management The therapeutic modalities used in the treatment of hypercalcemia are numerous, but they should always be used in conjunction with therapy of the underlying malignant disease. The exception to this is breast cancer, when hormone therapy should be stopped until hypercalcemia is regulated. The treatment depends on the clinical status of the patient and on the calcium level in the blood, but the general principles of treatment include treating the cancer when possible, encouraging ambulation, correcting dehydration, increasing urinary calcium excretion; inhibiting osteoclastic activity (calcium removal from bone) and reducing calcium intake. If serum calcium levels are below 14 mg/dL, oral rehydration and ambulation may suffice. Normal saline solution can be administered if the oral intake is not sufficient. If the serum phosphate level is not elevated, oral phosphates may be used cautiously. Phosphosoda (5 mL by mouth, two or three times daily) is usually tolerated with mild to no diarrhea. IV phosphates are able to effectively lower the serum calcium level through precipitation of inorganic calcium phosphate salts in bone. This modality of treatment is usually not recommended, however, and if needed, it should only be done in consultation with a nephrologist or oncologist in view of their serious complications, which include widespread visceral calcifications, shock, and renal failure. This agent is usually reserved for hypercaecemia unresponsive to other agents. Mithramycin (given as 25 µg/kg IM once every 4 to 5 days) is not generally part of the initial emergency management of hypercalcemia and has been supplanted in most cases by the bisphosphonates. Prednisone (60–80 mg) or other corticosteroids may be effective within a few days to a week. This drug is more useful for long-term treatment than for acute control. Corticosteroids are particularly valuable in breast carcinoma, myeloma, and lymphoma. They should not be initiated without oncologic consultation because they are chemotherapeutic agents for these malignancies. If the serum calcium level is greater than 14 mg/dL or significant symptoms are present, a more vigorous management should be undertaken. Continuous cardiac monitoring in the ED is necessary and central venous or pulmonary artery pressure monitoring may be required. Saline rehydration and diuresis stimulates renal tubular excretion of calcium and is the most important initial component of the emergency management of hypercalcemia. Dehydration should be corrected within 1 to 2 hours with normal saline solution. When urine flow is adequate, furosemide (40– 60 mg IV) may be given to increase excretion of calcium. Although the calciuric effect of furosemide is modest, it is also useful in preventing fluid overload in patients predisposed to cardiac failure. Careful attention to fluid input and output to ensure that the patient remains euvolemic is necessary. Calcitonin is a naturally occurring hormone that inhibits bone resorption and increased excretion of calcium. Calcitonin may be effective in doses of 4 to 8 IU/kg IM/SC. This treatment, although relatively safe when renal function is normal, is not generally part of the initial emergency management of hypercalcemia. Fifty percent of hypercalcemic cancer patients also have hypokalemia. Serum potassium levels should be monitored every 4 hours and potassium chloride (20–40 mEq, IV or PO) supplemented as necessary to prevent severe hypokalemia.24,31,33,35
In the past 5 years following approval by the Food and Drug Administration, bisphosphonates have become the treatment of choice for management of cancer-induced hypercalcemia supplanting all other pharmacologic approaches except corticosteroids. Bisphosphonates act by binding to hydroxyapatite in bone and thereby inhibiting the dissolution of crystals. These agents prevent osteoclast attachment to bone matrix and interfere with osteoclast recruitment without inhibiting bone formation and mineralization. Several agents are now available, including clodronate, pamidronate, and ibandronate, with other more potent bisphosphonates in development. Pamidronate (90 mg, given as an infusion over 4–24 hr) effectively and safely achieves normocalcemia within a few days (mean 4 days) in over 90 to 95% of patients.32-35
■ NEOPLASTIC CARDIAC TAMPONADE Although cardiac tamponade resulting from neoplasm is uncommon, it can occur abruptly and result in death if not treated quickly. In most cases neoplastic cardiac tamponade is observed in patients with a previous diagnosis of cancer, typically at late stages of the disease. It is rarely seen as the initial manifestation of an extracardiac malignancy. The decompensated state of cardiac function comes from a marked rise in intrapericardial pressure caused by accumulation of fluid within the pericardial sac resulting from malignancy or from pericardial thickening with scar formation, which results in a thick constrictive neoplastic encasement. This condition, if not recognized and decompressed promptly, can lead to circulatory compromise and death. Signs and symptoms are partially affected by the rapidity of development. In the era prior to diagnostic ultrasound this medical/oncologic emergency was often unrecognized. In one early series prior to the advent of ultrasound, the diagnosis was missed by the first physician in 11 of 17 patients and a number of times was missed by more than a single examiner.37 In most instances, pericardial effusion is accompanied by signs and symptoms that presage the development of the clinical picture of tamponade, including dyspnea, apprehension, anxiety, and chest pain. In rare instances, tamponade may be the first manifestation of the malignancy, solid tumor, or leukemia. Any patient in the ED with a history of cancer, shortness of breath, and hypotension should be suspected of having pericardial tamponade. The diagnoses of pulmonary embolism, congestive heart failure, and anxiety can be mistakenly made in this setting.
Etiology The most common cause of neoplastic pericardial tamponade is malignant pericardial effusion, often associated with postirradiation pericarditis, fibrosis, and effusion. Only rarely does a tumor or radiation fibrosis cause a neoplastic constrictive pericarditis with resultant tamponade. In most reported cases, cardiac tamponade represents a clinical progression of neoplastic or postirradiation pericarditis. Neoplastic pericarditis can result from any number of benign, malignant, primary, or secondary tumors of the pericardium or mediastinum.38-40 The most common benign tumors of the pericardium or mediastinum are fibromas, angiomas, and teratomas. Pericardial mesothelioma can have a clinical course characterized by rapid accumulation of massive quantities of bloody pericardial fluid, eventually leading to tamponade. Secondary involvement of the pericardium may result from either direct invasion from structures or metastases from a distant primary tumor. These metastases are usually multiple rather
Pathophysiology The severity of cardiac tamponade and eventual cardiopulmonary decompensation depend on the rate of pericardial fluid accumulation, the fluid volume, and the underlying cardiac function. Clinically the progressive elevation of intracardial pressure interferes with ventricular expansion and results in a decrease in the cardiac volume. Intracardial chamber pressures rise rapidly with subsequent transmission of this pressure peripherally in pulmonary and vena caval beds. In an effort to maintain cardiac output, various compensatory mechanisms come into play (tachycardia, peripheral vasoconstriction, decrease in renal flow with resultant increase in blood volume by sodium and water retention), all to maintain arterial pressure and venous return. When these compensatory mechanisms fail to maintain cardiac output, ventricular end-diastolic pressure increases and subsequent circulatory collapse is impending. The signs and symptoms parallel these pathophysiologic changes. The most common symptoms include extreme anxiety and apprehension, a precordial oppressive feeling, or actual retrosternal chest pain with dyspnea of varying degrees. True orthopnea and paroxysmal nocturnal dyspnea are uncommon, but when they occur the patient assumes a variety of positions to get relief from the chest pain and the dyspnea. Other prominent symptoms include cough, hoarseness, hiccups, and occasional GI manifestations such as dysphagia, nausea, vomiting, and epigastric or right upper quadrant abdominal pain that is probably the result of visceral congestion.12,40-43
Clinical Features Patients with severe tamponade are acutely ill and may appear ashen, pale, or markedly diaphoretic with an impaired consciousness ranging from mildly confused to unresponsive.
1599
Evaluation of Neoplastic Table 121-1 Physical Cardiac Tamponade Beck Triad or Acute Compression Triad Described in 1935, this complex of physical findings refers to increased jugular venous pressure, hypotension, and diminished heart sounds. These findings result from a rapid accumulation of pericardial fluid. However, this classic triad is usually observed in patients with acute cardiac tamponade. Pulsus Paradoxus or Paradoxical Pulse This is an exaggeration (>12 mm Hg, or 9%) of the normal inspiratory decrease in systemic blood pressure. To measure the pulsus paradoxus, patients are often placed in a semirecumbent position; respirations should be normal. The blood pressure cuff is inflated to at least 20 mm Hg above the systolic pressure and slowly deflated until the first Korotkoff sounds are heard only during expiration. At this pressure reading, if the cuff is not further deflated and a pulsus paradoxus is present, the first Korotkoff sound is not audible during inspiration. As the cuff is further deflated, the point at which the first Korotkoff sound is audible during both inspiration and expiration is recorded. If the difference between the first and second measurement is greater than 12 mm Hg, an abnormal pulsus paradoxus is present. The paradox is that while listening to the heart sounds during inspiration, the pulse weakens or may not be palpated with certain heartbeats, while S1 is heard with all heartbeats. A pulsus paradoxus can be observed in patients with other conditions, such as constrictive pericarditis, severe obstructive pulmonary disease, restrictive cardiomyopathy, pulmonary embolism, rapid and labored breathing, and right ventricular infarction with shock. A pulsus paradoxus may be absent in patients with markedly elevated left ventricular diastolic pressures, atrial septal defect, pulmonary hypertension, and aortic regurgitation. Kussmaul’s Sign This was described by Adolph Kussmaul as a paradoxical increase in venous distention and pressure during inspiration. This sign is usually observed in patients with constrictive pericarditis but occasionally is observed in patients with effusive-constrictive pericarditis and cardiac tamponade.
Rapid, shallow, and occasionally labored breathing may be present along with peripheral cyanosis and distended jugular veins. Seizures have been reported. Striking facial plethora and a full neck secondary to edema (Stoke’s collar) can also be seen in SVCS. Pulses are soft and easily compressible. The systolic blood pressure is usually low, with a decreased pulse pressure, although normal systolic, diastolic, and pulse pressures have been reported with moderate degrees of tamponade. Kussmaul’s signs (muffled heart sounds, an enlarged cardiomediastinal silhouette, tachycardia, and, most notably, pulsus paradoxus) are extremely useful findings in the physical evaluation of tamponade (Table 121-1).44 Ascites, hepatomegaly, peripheral edema, and mottling are other findings that reflect the elevation in venous pressure and decrease in cardiac output.12,40,41,45
Ancillary Evaluation Low-voltage and the nonspecific findings of pericardial effusion, sinus tachycardia, ST elevation, and nonspecific ST-T wave changes may occur. Electrical alternans with 1 : 1 total atrial-ventricular complexes has been considered almost pathognomonic of cardiac tamponade. Electrical alternans is the alternation of electrocardiographic QRS complexes, usually
Chapter 121 / Selected Oncologic Emergencies
than solitary lesions. The tumors most commonly associated with pericardial involvement include those of the lung and breast, leukemia, Hodgkin’s and non-Hodgkin’s lymphomas, melanomas, GI primary tumors, and sarcomas.40,41 Clinically recognizable symptoms or signs of pericardial disease are difficult to appreciate before death. Less than 30% of patients with autopsy-proven malignant pericardial disease were diagnosed antemortem.40,41 Radiation pericarditis has been a well-known complication of radiotherapy since the introduction of modern megavoltage techniques. The cardiac effects of radiotherapy may manifest themselves immediately with acute pericarditis or be delayed for months to years, although the majority develop effusion within the first year. The acute forms are inflammatory or effusive, usually self-limited, and subside without residual constriction; the chronic effusive and constrictive types may lead to tamponade and death.42 Neoplastic constrictive pericarditis, although rare, may be caused by the invasion of the pericardium by metastatic lesions or indirectly from the complication of radiation therapy with resultant fibrous thickening of the pericardium. Each of these entities can progress to cardiac tamponade because of thickening by tumor or radiation fibrosis, resulting in a decrease in the distensibility of the pericardium, thus reaching the critical point of cardiopulmonary decompensation earlier, despite smaller volumes of slowly accumulating effusion. The symptoms and signs of neoplastic and radiation pericarditis mimic pericarditis from other causes, and because of the usual insidious onset of the effusion of fibrous pericardial thickening, the condition might be attributed to the underlying malignancy and not considered until the full-blown picture of cardiac tamponade develops.
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
1600
in a 2 : 1 ratio. This is due to movement of the heart in the pericardial space. Electrical alternans is also observed in patients with myocardial ischemia, acute pulmonary embolism, and tachyarrhythmias.44 Approximately two thirds of the reported cases of pulsus alternans occur in patients with tamponade caused by massive pericardial effusion in neoplastic pericarditis. The alternation customarily disappears soon after removal of a small volume of fluid, but it can also disappear spontaneously or be observed in attendance with a fluid increase.41 Radiographic signs of tamponade suggestive of pericardial effusion include an enlarged cardiac silhouette with clear lung fields and normal vascular pattern, although a normal chest radiograph does not exclude tamponade. The typical “waterbottle” appearance of the heart on a plain radiograph is often present. Echocardiography is the simplest and most sensitive of diagnostic tests and can be done at the bedside for confirmation of pericardial effusion. Thoracic CT has also become an important diagnostic tool in diagnosing pericardial effusions.41,46 Cardiac tamponade should be considered in any cancer patient with dyspnea. Highly suggestive symptoms include clouded sensorium, thready pulse, pulsus paradoxus exceeding 50% of the pulse pressure, low systolic pressure, engorged neck veins with a rising peripheral venous pressure above 130 mm H2O, a falling pulse pressure below 20 mm Hg, and electrical alternans. This is an uncommon yet pathognomonic sinusoidal variation in QRS size secondary to the pendular effect of the heart swinging in the fluid medium of the pericardial sac.45 In this setting, sudden death may occur and pericardiocentesis should be performed as soon as possible.
Management In the ED, the only lifesaving treatment for tamponade that is effective is immediate removal of the pericardial effusion via pericardiocentesis. The procedure carries some risk, including induction of cardiac dysrhythmias and hemorrhage from an injured coronary vessel. Aspiration of as little as 50 to 100 mL of fluid has been shown to temporarily alleviate the pathologic process.12,40,41 Emergency subxiphoid percutaneous drainage is a lifesaving bedside procedure. The subxiphoid approach is extrapleural; hence, it is the safest for blind pericardiocentesis. A 16- or 18-gauge needle is inserted at an angle of 30 to 45 degrees to the skin, near the left xiphocostal angle, aiming toward the left shoulder. When performed emergently, this procedure is associated with a reported mortality rate of approximately 4% and a complication rate of 17%.44 Echocardiographically guided pericardiocentesis can also be performed in the ED, but the cardiac catheterization laboratory is a more controlled setting and is preferable: this is usually performed from the left intercostal space. First, mark the site of entry based on the area of maximal fluid accumulation closest to the transducer. Then, measure the distance from the skin to the pericardial space. The angle of the transducer should be the trajectory of the needle during the procedure. Avoid the inferior rib margin while advancing the needle to prevent neurovascular injury. Leave a 16-gauge catheter in place for continuous drainage.44 Removal of the maximal amount of fluid is advisable, along with inserting of an indwelling catheter, during the first pericardiocentesis since fluid may reaccumulate during the first 24 hours. Once the pericardial fluid has been obtained, it must be sent for biochemical and cytologic analysis. Neoplastic cardiac tamponade accounts for at least 50% of all reported cases of pericardial fluid collection. Other types of supportive therapy may be needed during the evaluation process while
preparing for pericardiocentesis, such as IV hydration with normal saline and oxygen therapy. Once the patient has been stabilized, additional therapeutic interventions should be planned and initiated by the appropriate admitting services because reaccumulation of effusion in neoplastic tamponade is not easily managed on a short-term basis. Pericardial windows, radiotherapy, intrapericardial chemotherapy, and pericardiectomy may be justified.12,40,41 The prognosis after neoplastic cardiac tamponade depends on the underlying type and extent of cancer. The presence of total electrical alternans is an adverse prognostic sign, even when the alternans disappears with pericardiocentesis. Despite a poor prognosis for patients with cancers such as melanoma or non-small-cell lung cancer, some patients with treatmentresponsive lymphomas have survived long-term after neoplastic cardiac tamponade.
■ NEUROLOGIC EMERGENCIES Of all patients with cancer, 15 to 20% have neurologic complications.47 Neurologic symptoms are occasionally the presenting complaint in patients with systemic cancer, but more often symptoms develop in patients known to have cancer. Neurologic emergencies in cancer patients include cerebral herniation, seizures, epidural spinal cord compression, CNS infections, and reversible toxic or metabolic encephalopathies. Treatment is needed urgently after the patient arrives at the ED to prevent permanent neurologic dysfunction or death.
■ CEREBRAL HERNIATION Pathophysiology Cerebral herniation occurs when the ICP increases locally within the skull from an expanding mass lesion. The increase produces a shift of brain substance in the direction of least resistance caudally through the tentorial opening and the foramen magnum. Causes of cerebral herniation in cancer patients commonly include primary or metastatic brain tumors and intracerebral hemorrhage. Less common causes include subdural hematoma, brain abscess, acute hydrocephalus, and radiation-induced brain necrosis.46 Primary brain tumors account for approximately one half of intracranial tumors. Metastatic brain tumors are seen most commonly in lung, breast, colon, kidney, and testicular cancer and in patients with choriocarcinoma and malignant melanoma.47,48
Clinical Features Three distinct herniation syndromes have been described: uncal, central, and tonsillar herniation. In uncal herniation a lateral mass displaces the temporal lobe, which compresses the upper brainstem. A rapid loss of consciousness is seen in conjunction with unilateral pupillary dilatation and ipsilateral hemiparesis. Central herniation usually results from slowly expanding, multifocal lesions that cause a downward and lateral shift of the diencephalon and upper pons. A slowly decreasing level of consciousness, small reactive pupils, and Cheyne-Stokes respirations, without focal signs, are seen clinically. Central herniation is sometimes mistaken for toxic or metabolic encephalopathy because of the lack of focal signs. A history of headache or focal neurologic complaints or any lateralizing findings indicates the need for prompt CT of the head to rule out a herniating mass lesion before lumbar puncture. Tonsillar herniation is produced by a large posterior fossa mass that pushes the cerebellar tonsils through the foramen magnum, compressing the medulla and resulting in a rapidly decreasing level of consciousness, occipital headache, vomit-
ing, hiccups, hypertension, meningismus, and abrupt changes in the respiratory pattern.47-50
When the clinical diagnosis of cerebral herniation is made, emergency management is necessary before the cause can be established. Intubation with hyperventilation to a carbon dioxide partial pressure (Pco2) of 25 to 30 mm Hg temporarily lowers the ICP by producing cerebral vasoconstriction. This should be avoided if possible but may be necessary for brief periods in response to reversible, acute neurologic deterioration. Excessive or prolonged hyperventilation may cause paradoxical vasodilation and should be avoided. Mannitol (1 g/kg IV) should be given and may be repeated in 4 to 6 hours. Dexamethasone (12–24 mg IV) has not been shown to improve outcome or reduce ICP acutely in severe head injury,51,52 but is often administered in patients with raised ICP or impending herniation caused by CNS malignancy because of the effect of corticosteroids on reducing cerebral edema associated with the neoplastic process. CT of the brain should be obtained as soon as emergency stabilization is accomplished. Epidural or subdural hematoma and hydrocephalus usually require surgery, whereas abscess and metastases are usually managed with antibiotics and antineoplastics or radiation, or both, respectively. When stabilization and an initial diagnosis have been made, neurologic or neurosurgical consultation and prompt admission to an intensive care unit are mandatory.47,49
■ SEIZURES Seizures are common in patients with cancer. Their immediate management is necessary to prevent physical injury, increased ICP, and risk of aspiration. Seizures increase the brain’s metabolic requirements and lead to increased cerebral blood flow. This may precipitate increased ICP in susceptible patients. Seizures may be due to brain metastases, toxic or metabolic disturbances (usually hyponatremia or uremia), vascular problems (especially intracerebral hemorrhage or subdural hematomas), and infections. Diagnostic laboratory studies should include a CBC, electrolytes, glucose level, blood urea nitrogen (BUN), measurement of calcium and magnesium levels, liver function tests, coagulation studies, and appropriate cultures. CT of the head should be done and followed by a lumbar puncture, when indicated.47,49 The therapy for seizures depends on the specific cause and the patient’s clinical status. For example, a single hypoglycemic or hypoxic seizure usually requires only correction of the underlying metabolic defect. Patients with a single seizure whose workup reveals a chronic problem (e.g., a cerebral metastasis) require anticonvulsants and therapy specific for the malignancy. A loading dose of phenytoin (15–18 mg/kg IV) may be given followed by oral maintenance. Prolonged single seizures or repetitive seizures require more vigorous treatment, including diazepam (5–10 mg IV) or lorazepam (1–2 mg IV) followed by IV phenytoin. Active airway and ventilatory management is essential. A bedside fingerstick glucose level should be obtained promptly. Thiamine and naloxone are not routinely indicated. In addition, when repetitive seizures have occurred, management of the underlying cause should be initiated rapidly and the patient admitted to an intensive care unit.47,49
■ EPIDURAL SPINAL CORD COMPRESSION Principles of Disease Epidural spinal cord compression from metastatic cancer is common, serious, and potentially treatable. It is most often
Clinical Features Back pain, either local or radicular, is the initial symptom in 95% of patients with epidural metastasis. It may be acute in onset or develop insidiously over weeks to months and usually predates other symptoms. The pain may increase during physical examination with spinal percussion, neck flexion, Valsalva maneuver, or straight leg raising and is usually located at the level of the tumor.50,54,55 Other symptoms are usually present at the time of diagnosis and may include weakness (75% of patients) and autonomic or sensory symptoms (50% of patients). Fifty percent of patients are not ambulatory at the time of diagnosis. The neurologic examination usually reveals symmetrical weakness with either flaccidity and hyporeflexia (if the diagnosis is made very early) or spasticity and hyperreflexia (if the diagnosis is made later).
Diagnostic Strategies Plain films show evidence of tumor in the vertebral body in 70 to 90% of patients with vertebral metastases.47,51 Immediate myelography or MRI is indicated if the plain films are abnormal, regardless of whether the neurologic examination is abnormal or is consistent with spinal cord compression or the plain film findings. In cases with questionable findings on plain films of the spine, tomograms, coned-down views, or CT may reveal bone metastases not otherwise appreciated. Myelography can demonstrate a complete or near-complete obstruction of contrast dye flow at the level of vertebral body involvement. MRI has emerged as the procedure of choice for intramedullary metastases and has also replaced myelography, which is associated with significant morbidity related to lumbar puncture and dye insertion at multiple levels (including cisternal puncture), to demonstrate the length of the compression or skip lesions along the spinal cord.50,55
Management Because minimal weakness at the time of presentation may progress to profound, irreversible weakness over several hours, treatment should be started rapidly. In the ED, a loading dose of dexamethasone (10–100 mg IV), followed by 4 to 24 mg every 6 hours for 3 days to reduce cord edema, is initiated at the time of diagnosis. Immediate oncology and radiation oncology consultations should be obtained. Although corticosteroids are routinely administered to patients with suggested spinal cord compression, high-dose corticosteroids, such as dexamethasone (100 mg), have been associated with complications and their use is controversial.55 Radiation treatment is the usual therapy and can be initiated after steroid treatment. The prognosis depends on the radiosensitivity of the tumor, the location of the compression, the pretreatment performance status, and the rate of decompensation. Surgery is indicated
Chapter 121 / Selected Oncologic Emergencies
Management
1601
caused by lymphoma or lung, breast, or prostate carcinoma. With the exception of lymphoma, which extends through the intervertebral foramina from paravertebral lymph nodes, these tumors metastasize to the vertebral body and extend into the spinal canal to compress the spinal cord. Less common causes of spinal cord compression in patients with cancer include melanoma, myeloma, renal cell carcinoma, vertebral subluxation, spinal epidural hematomas, and intramedullary metastasis. Acute myelopathy in patients with cancer may also be caused by radiation, paraneoplastic necrotizing myelitis, a ruptured intervertebral disk, and meningeal carcinomatosis with spinal cord involvement. Most cases (68%) of epidural cord compression occur in the thoracic spine, 15% occur in the cervical spine, and 19% in the lumbosacral spine.53
PART III ■ Medicine and Surgery / Section Ten • Hematology and Oncology
1602
only if the diagnosis is in doubt, if a tissue diagnosis is required, if the spine is unstable, or when maximal doses of radiation have already been given to the involved area.47,48,56 Intramedullary metastases are similar in presentation and treatment to epidural cord compression but are associated with a very poor prognosis. Epidural hematomas have been described in patients with thrombocytopenia or a coagulopathy as a complication of lumbar puncture. A rapidly progressive paraparesis and back pain are seen. MRI or myelography can establish the diagnosis; the treatment is surgical decompression. Platelet transfusions may limit progression in the ED.47,56
■ CENTRAL NERVOUS SYSTEM INFECTIONS Principles of Disease Patients with cancer are susceptible to a variety of CNS infections. These patients may have impaired immune responses secondary to their underlying disease or treatment with steroids, chemotherapy, splenectomy, or irradiation. Most CNS infections occur in patients with leukemia, lymphoma, or head and neck cancer. Patients with head and neck cancer are susceptible (in addition to the reasons discussed) because of fistula formation and tumor invasion, which allows organisms access to the CNS. Important CNS infections include meningitis, brain abscess, and encephalitis. These often have similar presentations, making their differentiation in the ED difficult.
the course of an abscess, with the classic well-defined mass with a low-density center and a contrast-enhancing ring seen later. Edema and mass effect are common. A lumbar puncture is not helpful in making the diagnosis and may precipitate cerebral herniation. Organisms that cause abscess include gram-negative rods, Aspergillus and Phycomycetes species, and Toxoplasma gondii. Emergency management includes highdosage antibiotics. If herniation develops, immediate steps to reduce the ICP, followed by emergency surgery, are indicated. Encephalitis is rare in patients with cancer and is most often caused by herpes zoster or T. gondii. The presenting complaints are usually headache, fever, and altered mental status. The CT scan is commonly normal but may show diffuse edema, whereas the lumbar puncture may show pleocytosis with an elevated protein level but no demonstrable organism. It is difficult to distinguish encephalitis from meningitis in the ED, but the overall clinical picture in both diseases mandates hospital admission for further evaluation.
Management
Meningitis is characterized by fever, headache, and altered mental status. Meningismus is often absent. The diagnosis of meningitis in patients with cancer is often delayed because the manifestations of the disease are attributed to other processes: fever to systemic infection, headache to cerebral metastases, and altered mental status to a toxic or metabolic encephalopathy.
Until further evaluation is able to distinguish between meningitis and encephalitis, empirical broad-spectrum antibiotic coverage with a third-generation cephalosporin (ceftriaxone or ceftazidime) and vancomycin should be initiated for all patients. Ampicillin may be added when there is suggestion of Listeria. Ceftazidime with or without an aminoglycoside is generally selected when the likelihood of infection with Pseudomonas is high. Neutropenic patients (polymorphonuclear WBC count < 1000/mm3) with either leukemia or lymphoma usually have a gram-negative infection (often with P. aeruginosa). Patients with lymphoma and a normal WBC count are commonly infected with Listeria monocytogenes, Streptococcus pneumoniae, or Cryptococcus neoformans. Infections with Haemophilus influenzae and Neisseria meningitidis are uncommon. Patients with head and neck tumors may develop staphylococcal infection.47-50
Diagnostic Strategies
Encephalopathy
All cancer patients with fever and an altered mental status require a lumbar puncture, which should be preceded by head CT if cerebral metastases are suggested.49,52 In addition, thrombocytopenia and coagulopathy should be considered and either ruled out or treated appropriately with platelet transfusions or fresh frozen plasma, respectively, before a lumbar puncture is done. Platelet transfusion is usually reserved for patients with platelet count less than 10,000/µL. The fluid obtained should be sent for a cell count and differential cell count, Gram’s stain, India ink stain, protein and glucose levels, bacterial and fungal cultures, cryptococcal antigen level, and cytologic examination. The absence of WBCs in the cerebrospinal fluid does not rule out meningitis, especially in neutropenic patients. The likely organisms responsible for meningitis vary with the underlying disease and the peripheral WBC count.
Toxic and metabolic encephalopathy should be actively considered when patients with cancer have an acute or subacute altered mental status in the absence of fever or headache. Toxic and metabolic causes should be routinely excluded even when infection or a metastatic complication is suggested. Signs of encephalopathy include confusion, aberrant behavior, and a decreased level of consciousness. These may develop acutely or insidiously over days to weeks. Patients with cancer are particularly susceptible to toxic and metabolic encephalopathy because their disease can have multiple organ system involvement, can cause electrolyte and nutritional abnormalities, and the drugs used to treat the disease (especially chemotherapeutic agents and narcotics) can cause encephalopathy even when used in therapeutic doses.47,48 In the ED, encephalopathic patients should first be evaluated carefully for a possible infection or mass lesion. The metabolic workup should include electrolytes; BUN, creatinine, glucose, and calcium levels; arterial blood gases; and liver function tests. Toxicology screens should be considered in possible ingestions and in patients who are unable to give a history. Naloxone and 50% dextrose should be given while the workup is proceeding. Specific treatment is indicated for any abnormalities found during the workup. Hospital admission is usually required unless the cause is easily and rapidly reversible and is unlikely to recur.
Clinical Features
Differential Considerations Brain abscess is usually seen in patients with leukemia or head and neck tumors and accounts for 30% of CNS infections in cancer patients.49 Patients have symptoms of elevated ICP (headache, vomiting, and papilledema), lateralizing findings, and a source of infection.48,52 Fever is usually present. Head CT characteristically demonstrates an ill-defined mass early in
1603
KEY CONCEPTS Hypercalcemia due to malignancy is unrelated to bone metastases in 20% of patients and is associated with a poor prognosis independent of therapeutic response. Hydration and use of bisphosphonates (e.g., pamidronate) have become the mainstays of initial treatment. ■ Spinal cord compression arises as back pain in more than 95% of patients. If ambulatory at the time of diagnosis, 80% of patients maintain the ability to ambulate. MRI has become the diagnostic modality of choice, and high-dose dexamethasone is given to all patients, followed by radiation therapy in most cases. ■ Superior vena cava obstruction is rarely life-threatening and requires tissue diagnosis. Although caused by malignancy in 70 to 80% of cases, thrombosis secondary to indwelling central lines is increasing as a cause. Stenting of the SVC has become the approach to SVC obstruction unresponsive to chemotherapy or radiation therapy, or both. ■ Fever and neutropenia in the cancer patient are a true medical emergency requiring rapid diagnosis, cultures,
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
and treatment with broad-spectrum, bactericidal, synergistic antimicrobials. An aminoglycoside plus an extended-spectrum penicillin and third-generation cephalosporin with or without vancomycin remain the standard combinations in patients without penicillin allergy. ■ Neoplastic pericardial effusion can arise insidiously with symptoms such as apprehension, anxiety, dyspnea, and weakness. Bedside ultrasonography has become a rapid, safe imaging modality for establishing the diagnosis prior to the development of clinically apparent tamponade in a critically ill patient. ■ Acute TLS, previously limited to hematologic malignancies, is now being described in patients receiving chemotherapy for solid tumors. It can arise with dyspnea, mental status changes, cardiac dysrhythmia, or seizures. Treatment includes urinary alkalinization and emergency hemodialysis in cases complicated by acute renal failure.
Chapter 121 / Selected Oncologic Emergencies
■
•
Section Eleven Metabolism and Endocrinology Chapter 122
Acid-Base Disorders
Jamie L. Collings
The human body must be maintained in a precise acid-base balance to maintain healthy cellular function. This balance is controlled by the lungs, kidneys, and serum buffers, interacting and responding to physiologic changes. Physiologic insults such as vomiting, diarrhea, respiratory failure, kidney dysfunction, diabetes, toxic ingestions, among others can result in life-threatening acid-base crises. Identifying and optimally treating the underlying condition is often achieved only through the diagnostic insights gained from acid-base measurements and calculations. This chapter presents essential acid-base physiology, beginning with an overview of the principles of acid-base function. This is followed by a discussion of primary respiratory acidosis and alkalosis and then metabolic causes of acidosis and alkalosis. Finally, mixed acid-base disorders are analyzed. Clinical implications are included along with the mathematical knowledge that the emergency physician requires to expertly manage these complex and potentially life-threatening conditions.
water (pH 7.0). Blood pH must be maintained within relatively narrow limits because protein and enzyme systems function properly only within a narrow pH spectrum. A pH outside the range of 6.8 to 7.8 is generally associated with serious disease processes and the potential for considerable morbidity or mortality. Acidemia is defined as a serum pH of less than 7.36. Conversely, alkalemia is defined as a pH of greater than 7.44. Acidosis is defined as a pathologic process that lowers the [HCO3−] (metabolic acidosis) or raises the Paco2 (respiratory acidosis); alkalosis is defined as a pathologic process that raises the [HCO3−] (metabolic alkalosis) or lowers the Paco2 (respiratory alkalosis). A simple acid-base disorder is a single acid-base disturbance with its accompanying compensatory response. Mixed acid-base disorders are the result of two or more primary disturbances.
■ PRINCIPLES OF DISEASE
Physiologic buffers, defined as a weak acid and its salt, oppose marked changes in pH after the addition of an organic acid or a base, as follows:
The kidneys, lungs, and physiologic buffers normally maintain the serum pH within a narrow spectrum, between 7.36 and 7.44. Each of these three systems dynamically responds to small changes in acid-base balance. Such precise physiologic control is required for normal cellular function. Consequently, disorders of kidneys, lungs, and physiologic buffers result in acid-base abnormalities. Blood pH is determined by the ratio of the serum bicarbonate concentration and Paco2 (partial pressure of CO2 in arterial blood). Primary metabolic acid-base disorders and the secondary metabolic compensation for primary respiratory disturbances alter the serum bicarbonate concentration [HCO3−]. Primary respiratory acid-base disorders and the secondary respiratory compensation for primary metabolic disturbances alter the Paco2. The Henderson-Hasselbalch equation relates the concentrations of the acid-base pair to the pH. As the pH changes, so does the concentration. Because the equation produces a logarithmic result, subtle changes in the serum pH can cause large and often significant alterations in the concentration of the acid-base pair. Clinically, this equation dictates how drugs disperse, enzymes react, and medications bind at a given serum pH. In humans, hydrogen ion concentration [H+] is extremely low (approximately 4 × 10−12 mEq/L) and strictly regulated. Normally, blood is slightly alkalemic relative to 1604
Physiologic Buffers
H+ + buffer − Na + buffer − H+ + Na + The human body uses three important physiologic buffers to minimize surges in pH: (1) the bicarbonate/carbonic acid system (primarily located in red blood cells), (2) intracellular protein buffers, and (3) phosphate buffers located within bone. Patients with malnutrition or chronic disease, and thus low albumin and bone density, and anemic patients have an ineffective buffering capability.
Bicarbonate/Carbonic Acid Buffer System The bicarbonate/carbonic acid buffer system is unique among physiologic buffering systems. The system is open-ended; continuous removal of organic acid is made possible by the exhalation of carbon dioxide (CO2). In equilibrium, the equation is as follows: H+ + HCO3− H2CO3 H2O + CO2 Bicarbonate is present in large quantities and can be controlled by the lungs and kidneys; thus, it serves as the major contributor to the maintenance of acid-base balance. Clinically, its importance is in the transient buffering of serum and interstitial fluid.
1605
Intracellular Protein Buffers
Bone as Buffer Bone contains a large reservoir of bicarbonate and phosphate and can buffer a significant acute acid load. Bone is probably involved in providing some buffering (mostly ionic exchange) in most acute acid-base disorders, but there is very little research in this area. In terms of duration, only two types of metabolic acidosis are long-lasting enough to be associated with the loss of bone mineral (through release of calcium carbonate): renal tubular acidosis and uremic acidosis. In uremic acidosis, loss of bone crystal is multifactorial (changes in vitamin D metabolism, phosphate metabolism and secondary hyperparathyroidism) and acidosis is only a minor factor.
Pulmonary Compensation The second compensatory system for pH changes involves a relationship between the peripheral chemoreceptors, located in the carotid bodies, and central chemoreceptors, located in the medulla oblongata. Both these receptors influence respiratory drive and can initiate changes in minute ventilation. A drop in pH stimulates the respiratory center, resulting in increased minute ventilation. This in turn lowers the Paco2, driving the pH toward the normal range. Conversely, an increase in pH decreases ventilatory effort, which increases Paco2 and lowers the pH back toward normal. A diabetic patient in ketoacidosis hyperventilates to compensate for the organic acidemia and would be expected to have a low Paco2. This compensatory response is the expected reaction to a fall in serum pH. In general, compensatory processes return the pH toward normal over a period of 4 to 12 hours, but do not fully normalize it. Respiratory alkalosis is the only primary acid-base disorder in which the pH does often normalize with time.
Renal Compensation The kidneys play little role in the acute compensation of acidbase disorders because they do not immediately respond to changes in pH. More than 6 to 12 hours of sustained acidosis results in active excretion of H+ (predominantly in the form of ammonium, NH4+, with retention of bicarbonate, HCO3−). Conversely, more than 6 hours of alkalemia stimulates renal excretion of bicarbonate with retention of H+ in the form of organic acids, resulting in near-normalization of pH. In metabolic acidosis, there is either an excess production or an infusion of H+ (e.g., lactic acid production, ketoacid production) or an excessive loss of anion (HCO3−) and accompanying sodium and potassium cations (Na+, K+; e.g., diarrhea). In general, the kidneys attempt to preserve Na+ by exchanging it for excreted H+ or K+. The quantity of potassium excreted depends on the level of acidosis and the serum K+ level. In the presence of an H+ load, hydrogen ions move from the extracellular fluid (ECF) into the intracellular fluid. For this to occur, K+ moves outside the cell into the ECF to maintain electroneutrality. In cases of severe acidosis, significant overall depletion of total body K+ stores can occur despite serum
■ DIAGNOSTIC STRATEGIES A stepwise clinical approach to acid-base disorders starts with a well-conducted history and physical examination. Particular attention should be paid to the patient’s past medical history, current medications, chance of toxic ingestion, occurrence of vomiting or diarrhea, level of consciousness on admission, respiratory rate, skin turgor, and urine output. Evaluation progresses with analysis of serum electrolytes and pH, and calculation of any anion gap, and calculation of the delta gap. These calculations assist in determining the type of acidosis or alkalosis present and whether it is part of a mixed condition. The anion gap (AG) can be calculated as follows: AG = Na + − (Cl− + HCO3− ) Traditionally, a normal AG has been considered 12 ± 3 mEq/ L. This number can vary from one laboratory to another (mostly based on whether potassium is included in the calculation) and the clinician should take this possibility into consideration. The “gap” provides an estimate of unmeasured anions in plasma, primarily albumin plus small amounts of sulfate, phosphate, and organic anions (e.g., citrate). If there are excess organic acids in the circulation, the organic acids dissociate and the resulting H+ is titrated by HCO3−, which increases the AG. If the AG is increased, especially when it is more than 10 mEq/L above the upper limit of the reference range, the clinician should consider an excess in organic acids or acidic substances. With smaller gaps up to one third of patients will not have a metabolic acidosis. The concept of a “low” AG (200 mg/dL), ketonemia (>1 : 2 dilutions), and acidemia (pH < 7.3). DKA can be caused by any condition that reduces insulin availability or activity or that increases glucagon. DKA occurs most often in type 1 diabetic patients with little or no endogenous insulin; however, its occurrence in patients with type 2 diabetes, particularly obese African Americans, is not as rare as once thought. DKA in these patients results from increased lipolysis, and the breakdown of free fatty acids leads to production of ketoacids. Precipitating events usually include infections, surgery, and emotional or physical stressors. Treatment is aimed at fluid replacement over the first 24 to 48 hours, insulin replacement, and potassium replacement. Isoniazid and Iron Toxicity. Isoniazid is a common, important, but potentially lethal medication used for the treatment of tuberculosis. Clinicians must be aware that ingestions of greater than 40 to 60 mg/kg pose a danger of not only recurrent seizures but also life-threatening metabolic acidosis (as a result of the lactate-producing seizure activity). Treatment involves pyridoxine administration to control seizures and hemodialysis to reduce both intravascular drug concentration and acidemia.
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1610
is supportive, and gastrointestinal decontamination with charcoal may be indicated. Fluid and electrolyte replacement will likely be necessary, and both urinary and serum alkalinization may be of benefit in enhancing elimination and minimizing toxicity.
Normal Anion Gap Metabolic Acidosis Metabolic acidosis with a normal AG is caused by either an excessive loss of HCO3− or an inability to excrete H+ (see Box 122-4). Any condition that causes excessive loss of intestinal fluid distal to the stomach can cause a normal AG metabolic acidosis. Normal AG metabolic acidosis is primarily a HCO3−wasting condition and in 95% of cases results from diarrhea. Other possible, although less common, causes include tube drainage and skin fistulae, with loss of HCO3−-rich intestinal, biliary, or pancreatic fluids. Ureterosigmoidostomy (surgical insertion of ureters into the sigmoid colon) produces a hyperchloremic acidosis because of loss of HCO3− in exchange for the reabsorption of Cl−. Patients with renal failure develop an inability to excrete their dietary H+ load; the severity is proportional to the degree of reduction in the GFR. Patients with renal tubular acidosis type 1 are unable to secrete H+ at the distal tubule, whereas impairment of HCO3− reabsorption at the proximal tubule is the defect in renal tubular acidosis type 2. Calculation of the urinary anion gap (UAG = [Na+ + K+] − Cl−) may be helpful; a negative urinary AG suggests GI loss of HCO3−, whereas a positive urinary AG suggests altered urinary acidification, indicating a renal tubule abnormality. Other causes of normal AG metabolic acidosis include hyperparathyroidism, medications such as carbonic anhydrase inhibitors (e.g., acetazolamide [Diamox], mafenide acetate [Sulfamylon]), spironolactone and cholestyramine, chloridecontaining acids ingestion (e.g., NH4 Cl−, arginine HCl, lycine), renal tubular acidosis, sulfur, CaCl2 and MgCl2 ingestions, and hyperalimentation with excess arginine, lysine, or Cl−.
Physiologic Compensation The body responds to acidemia by utilizing four buffering systems: (1) extracellular bicarbonate/carbonic acid (HCO3−/ H2CO3) system, (2) intracellular blood protein system, and (3) renal and (4) respiratory compensation systems (see Fig. 122-2). The first two processes minimize the initial [H+], while the kidneys eliminate excessive H+ in the urine, reabsorb HCO3−, and restore acid-base homeostasis. The CNS responds to increased [H+], through direct stimulation of the chemoreceptors in the medulla oblongata, by stimulating the respiratory center. This results in an increase in alveolar ventilation, producing a compensatory elimination of Paco2 and elimination of excess H+. It may take 12 to 24 hours to achieve a maximal respiratory response to a sustained metabolic acidosis. When the arterial pH is 7.1 or less, the minute ventilation can reach 30 L/min, and at this level of pH, Kussmaul’s respiration and its prominent hyperventilation can be seen. In response to metabolic acidosis, H+ is excreted by the kidney while HCO3− is reabsorbed. The rate-limiting reaction (the synthesis of H2CO3 from CO2 and H2O) is catalyzed by carbonic anhydrase. Therefore, inhibitors of this enzyme can create a metabolic acidosis by preventing the renal excretion of H+. The excretion of H+ requires buffering with HPO4− or NH3, with NH4+ playing the largest role. This buffering is called titratable acidity. The kidney responds to an increased
H+ load by the augmentation of cellular NH3 production and consequently NH4+ excretion. In summary, H+ is acutely buffered by extracellular and intracellular mechanisms. However, these mechanisms are not potent enough to correct acidosis sufficiently. Acidemia stimulates the CNS ventilatory center, and the Paco2 is reduced secondary to Kussmaul’s respiration. With continued and chronic acidemia, the kidneys secrete H+ (as NH4+ and H2PO4−) and reabsorb HCO3− in an attempt to neutralize the acidosis.
Management In treating patients with metabolic acidosis, primary efforts should be directed at restoring their homeostatic mechanisms. The clinician must treat the patient, using laboratory markers only as a guide. Individual therapies are directed toward the particular cause of the acidosis. Active correction of the pH depends on the severity of the acid-base imbalance, the cause, the patient’s compensatory capabilities, and the potential harm caused by therapy. Most patients with metabolic acidosis do not require aggressive attempts at pH manipulation. For many, the causality is easily discernible, and treatment involves stabilization of homeostatic mechanisms. For example, metabolic acidosis (average pH 7.1) after a seizure resolves within approximately 15 minutes, and the bicarbonate normalizes within 45 to 60 minutes. Rather than administration of sodium bicarbonate (NaHCO3), immediate treatment would involve termination of the seizure activity, maintenance of the airway, and provision for acid-base normalization by ventilatory loss of CO2. Therapy with NaHCO3 has some inherent complications, and rapid NaHCO3 replacement can result in paradoxical CNS intracellular acidosis, impaired oxygen delivery, hypokalemia, hypocalcemia, “overshoot” alkalosis, hypernatremia, volume overload, and hyperosmolality. Bicarbonate penetration into the CNS across the blood-brain barrier is very slow; consequently, intravenous HCO3− therapy alkalinizes the plasma much faster than the CNS. As the serum pH increases, the peripheral chemoreceptors decrease minute ventilation, raising Paco2 in an attempt to normalize the serum pH. CO2, which rapidly diffuses across the blood-brain barrier, rises intracerebrally, and the CNS becomes more acidemic despite alkalinization of the plasma. This inverse reaction is referred to as paradoxical CNS acidosis. Much discussion surrounds this phenomenon and intravenous HCO3− use. Buffer therapy during out-of-hospital cardiac arrest had little to no benefit in one study, regardless of the arterial pH.12 The only prospective, randomized, controlled study was done on hypovolemic rats and failed to demonstrate any difference between the HCO3− and control groups.13 Furthermore, alkali therapy can lead to ECF volume overload (especially in patients with congestive heart failure) and hypokalemia, which may lead to respiratory muscle weakness and inability to hyperventilate if it is severe. Administration of loop diuretics may prevent or treat this complication, but if adequate diuresis cannot be established, emergent dialysis may be necessary. Because NaHCO3 imparts a significant sodium load on the patient, several low-sodium buffers have been developed. Unfortunately, none has proven to be clinically more efficacious than NaHCO3.14 Because of the inherent complications associated with HCO3− replacement, the rule of thumb is to consider treatment in patients who have pH less than 7.1 with NaHCO3 1 mEq/kg unless the condition of acidemia is expected to be self-limited. For example, many experts do not recommend administration of HCO3− in patients with DKA and pHs as low
as 6.9. Another formula available to assist in determining the adequate dose is the following: Half should be replaced initially, and further NaHCO3 therapy should be determined by patient response and laboratory parameters. Patients with normal AG metabolic acidosis have a greater loss of HCO3− than those with an increased AG, and therefore the clinician may have a lower threshold for replacement (i.e., treat serum HCO3− less than 8 mEq/L and correct to 12–15 mEq/L versus AG acidosis where it should be corrected to 10 mEq/L).
■ METABOLIC ALKALOSIS Metabolic alkalosis is produced by conditions that increase HCO3− or reduce H+. This usually requires either the loss of H+ or the retention of HCO3−. The diagnosis requires knowledge of the Paco2, because elevation of the plasma HCO3− may be secondary to renal compensation of a chronic respiratory acidosis.
Etiology Metabolic alkalosis is usually caused by an increase in HCO3− reabsorption secondary to volume, potassium, or Cl− loss (Box 122-5). Loss of H+ and Cl− from protracted vomiting and nasogastric suctioning can also lead to HCO3− retention. Renal impairment of HCO3− excretion, especially in the setting of alkali therapy, can lead to a significant metabolic alkalosis. An ECF volume reduction can increase the plasma HCO3− concentration when combined salt and water losses occur, typically in patients using diuretics. This state forces a contraction of the ECF around a constant plasma HCO3−, creating a relative excess in HCO3− concentration; this is known as contraction alkalosis. Metabolic alkalosis can be caused by hypokalemia as H+ is shifted intracellularly in exchange for the osmotic movement of K+ extracellularly. There is also an increase in renal H+
BOX 122-5 Causes of Metabolic Alkalosis Volume-Contracted (Saline-Responsive) Vomiting/gastric suction Diuretics Ion-deficient baby formula Colonic adenomas Postrespiratory acidosis Normal Volume/Volume-Expanded (Saline-Resistant) Hyperaldosteronism (primary, secondary, or exogenous mineralic corticoids, e.g., licorice, tobacco) Cushing’s syndrome Severe potassium depletion Adenocarcinoma Bartter’s syndrome Ectopic adrenocorticotropic hormone Unclassified Milk-alkali syndrome Carbenicillin therapy Metabolism of organic acid anion (bicarbonate, lactate, citrate) Massive transfusion with citrate anticoagulant or plasmanate (acetate) if renal impairment Nonparathyroid hypercalcemia
Physiologic Compensation Although somewhat less predictable, acute compensation of metabolic alkalosis involves the respiratory center, and chronic compensation involves the renal system. During acute compensation, chemoreceptors controlling ventilation respond to an increased pH by inducing hypoventilation, thus increasing Paco2 and forming H+, which lowers the pH back to normal. A Paco2 of greater than 55 mm Hg is unlikely to be caused by simple respiratory compensation of metabolic alkalosis, and this value should alert the clinician to a ventilation disorder complicating the picture. Chronic compensation for metabolic acidosis results from the kidneys excreting excess HCO3− in the urine. In patients with renal failure, impairment in renal HCO3− excretion results in sustained metabolic alkalosis.
Management Clinicians can easily treat the simple loss of H+ from protracted vomiting or nasogastric suction. For more complicated causes, however, management can be directed by measurement of the urinary Cl−, which helps classify metabolic alkalosis into saline-responsive or saline-resistant.
Saline-Responsive Alkalosis Patients with saline-responsive alkalosis have a urinary Cl− level less than 10 mEq/L. Treatment is directed toward correcting the urinary excretion of HCO3−. Administration of NaCl and KCl suppresses both renal acid excretion and renal HCO3− excretion. Administration of NaCl and KCl should be considered for patients with mild to moderate salineresponsive alkalosis. In patients who are severely volumedepleted, consultation for admission and administration of intravenous mineral acids (e.g., arginine monohydrochloride) may be necessary. In edematous states for which saline therapy may be contraindicated, acetazolamide increases the excretion of NaHCO3, treating both the alkalosis and the edema. In renal failure patients, severe metabolic alkalosis should be treated with dialysis.
Saline-Resistant Alkalosis Patients with saline-resistant alkalosis have a urinary Cl− level greater than 10 mEq/L. In mineralocorticoid excess, hypokalemia and increased secretion of aldosterone lead to excessive renal excretion of H+ and a reabsorption of HCO3−. Treatment can be successful with K+ replacement by reversing the intracellular shift of H+. This reduction of cellular H+ also enhances HCO3− excretion. Additional therapy can be directed toward reducing mineralocorticoid activity (e.g., administering spir onolactone, an aldosterone antagonist).
■ MIXED ACID-BASE DISORDERS Double and triple primary acid-base disturbances are common. Traditionally, mixed disorders have been difficult to evaluate in the emergency department. However, recent literature pro-
Chapter 122 / Acid-Base Disorders
NaHCO3 (mEq ) = 25 − (measured HCO3− ) × ( weight [kg ] 2)
1611
secretion and HCO3− reabsorption. The net effect is ECF alkalosis with paradoxical intracellular acidosis, which is easily reversed with K+ therapy. Primary hyperaldosteronism, hyper-reninism, licorice ingestion, Cushing’s syndrome, and congenital adrenal hyperplasia are associated with mineralocorticoid excess. This leads to an increased Na+ reabsorption in the distal tubule with its accompanying H+ and K+ secretion to maintain electroneutrality.
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1612
vides some guidelines for ascertaining the mixed disorder and its causes. Clues to the presence of a mixed acid-base disturbance can either be historical (e.g., polydrug ingestion) or clinical, with varied chemistry and arterial blood gas findings that differ from those anticipated. We can use a six-step approach to analyzing acid-base disturbances as outline below and in Figure 122-3. Step 1 involves measuring the pH. It is necessary to first assess whether the patient has an acidemia (pH < 7.36) or alkalemia (pH > 7.44). The human body almost never fully compensates for any primary acid-base disturbance except for chronic respiratory alkalosis. Step 2: Is the primary disturbance respiratory or metabolic? Step 3 requires the clinician to calculate the AG. Box 122-3 lists possible causes of an AG greater than 15 mEq/L, and Box 122-4 lists possible causes for a case in which the AG is normal but the patient has a metabolic acidosis. Step 4 involves calculating the delta gap (ΔG = deviation of AG from normal − deviation of HCO3− from normal) to help resolve the possibility of a mixed acid-base disorder or further differentiate an elevated AG metabolic acidosis.
Values for the ΔG are all gaussian, and therefore the mean value should be near zero.15 An expected normal range for the ΔG would be 0 ± 6. A positive ΔG (+6 or greater) is almost always caused by high AG acidosis and a primary metabolic alkalosis. DKA or AKA with severe vomiting, lactic acidosis in the setting of chronic diuretic use, and renal disease with vomiting are clinical examples. A negative ΔG (−6 or less), on the other hand, can be of varied clinical representation. Most often there is either a mixed high AG and normal AG acidosis, or a high AG acidosis with chronic respiratory alkalosis and a compensating hyperchloremic acidosis. Clinically, these patients often have severe underlying metabolic disease with ongoing toxic ingestion (e.g., profound hypermagnesemia, hyponatremia, or hypercalcemia in patients with lithium toxicity) or chronic lung disease, acute lactic acidosis, and furosemide use. Other relationships in these disorders can also assist in rapid interpretation of mixed acid-base disturbances (Box 122-6). Step 5 asks whether the respiratory disturbance (if there is one) is acute or chronic. If acute, for each change in Pco2 of 10 mm Hg, the pH changes by 0.08 in the opposite direction. If chronic, for each change in Pco2 of 10, the pH changes by 0.03 in the opposite direction. Step 6 involves determining if the respiratory system has compensated fully when the primary disturbance is a metabolic acidosis. Using Winter’s formula (Pco2 = 1.5 (HCO3−) + 8 ± 2), you can calculate the degree of compensation. Mixed acid-base disturbances typically result from compensation failure, excessive compensation, or more than one disease process. Examples of compensation failures resulting in metabolic acidosis and respiratory acidosis include: cardiac arrest patients, chronic obstructive pulmonary disease patients with respiratory failure and hypoxemia, and hypoventilation and acidosis-causing toxins. Metabolic alkalosis and respiratory alkalosis may result from compensation failure in patients who are pregnant with hyperemesis and in patients with postoperative pain and vomiting. Patients with salicylate overdose,
BOX 122-6 Relationships in Acid-Base Disturbances Respiratory Acidosis Acute HCO3− increases 1 mEq/L (range, 0.25–1.75) for every 10-mm Hg increase in Pco2. pH drops 0.08 for every 10-mEq/L rise in HCO3−. Chronic (>5 days of hypercapnia) HCO3− increases 4 mEq/L for every 10-mm Hg increase in Pco2 (±4). Limit of compensation: bicarbonate will rarely exceed 38–45 mEq/L. Metabolic Acidosis Note: It may take 12–24 hours for maximal respiratory response to develop. Paco2 = (1.5 × HCO3) + 8 ± 2. Paco2 is equivalent to last two digits of pH (e.g., if Pco2 is 20, pH should be 7.20). ΔPco2 − 1 [1.3 × (ΔHCO3−)] For pure AG acidosis, the rise in anion gap should be equal to the fall in [HCO3−] (i.e., ΔG should equal 0). For pure non-AG (hyperchloremic) acidosis, the fall in HCO3− should be equal to the rise in [Cl−] (i.e., ΔHCO3− = −ΔCl−). Limit of compensation: Paco2 will not fall below 10–15 mm Hg. Respiratory Alkalosis Acute HCO3− drops 1 to 3.5 mEq/L for every 10-mm Hg drop in Pco2. Limit of compensation: bicarbonate is rarely below 18 mEq/L. Chronic (renal compensation starts within 6 hours and is usually at a steady state by 1 1 2–2 days) HCO3− drops 2–5 mEq/L for every 10-mm Hg drop in Pco2. Limit of compensation: bicarbonate is rarely below 12–14 mEq/L. Metabolic Alkalosis Pco2 = 0.9 (HCO3−) + 9. Pco2 increases 0.6 mm Hg for each 1-mEq/L increase in HCO3−. Limit of compensation: Pco2 rarely exceeds 55 mm Hg but has been reported as high as 75.
pulmonary edema, sepsis, and hepatic failure may exhibit excessive compensation and a combined metabolic acidosis and respiratory alkalosis that results in a near normal pH. Superimposed metabolic acidosis and alkalosis may also be present in patients with excessive compensation due to vomiting in association with DKA or AKA. In the alcoholic patient with AKA you may find a triple acid-base disturbance as a result of vomiting (metabolic alkalosis), withdrawal (respiratory alkalosis), and the AKA (metabolic acidosis). A thorough history and physical examination can be especially important in prompting the clinician to consider a complicated acid-base disturbance.
1613 What is the pH?
40 then metabolic alkalosis
PCO2 > 40 then respiratory acidosis
PCO2 < 40 then metabolic acidosis
Calculate anion gap Na+ − (Cl– + CO2)
Anion gap >15, then patient also has metabolic acidosis or patient has anion gap metabolic acidosis if metabolic acidosis already known
If no gap anion gap acidosis less likely
Very low or negative anion gap; consider bromism, lithium toxicity, multiple myeloma, and hypoalbuminemia
Calculate delta gap (∆ G = deviation of AG from normal – deviation of HCO3– from normal); ∆G = (calculated AG – 12) – (24 – measured HCO3–)
Values less than –6 most often result from either a mixed high AG and normal AG acidosis or a high AG acidosis with chronic respiratory alkalosis and a compensating hyperchloremic acidosis
Values greater than +6 almost always result from high AG acidosis and primary metabolic acidosis
If respiratory, is this acute or chronic? If acute, for each change PCO2 of 10 mm Hg, the pH changes by 0.08 in the opposite direction If chronic, for each change in PCO2 of 10, the pH changes by 0.03 in the opposite direction
If the primary disturbance is a metabolic acidosis, has the respiratory system compensated sufficiently? Winter’s formula: PCO2 = 1.5 (HCO3–) + 8 +/−2 • Only applicable if metabolic acidosis is primary disturbance • Answer is a range
Figure 122-3. Algorithm for acid-base calculation.
Chapter 122 / Acid-Base Disorders
>7.44
1614
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
KEY CONCEPTS ■
Changes in serum pH are dealt with by three compensatory systems: (1) the physiologic buffers, (2) the lungs, and (3) the kidneys. ■ HCO3− is present in large quantities and can be controlled by the lungs and kidneys, making it the major contributor to the maintenance of acid-base balance and the primary system to handle the acute load of organic acidemia. ■ Respiratory acidosis is defined as decreased pH that results from pulmonary CO2 retention. This CO2 retention leads to excess H2CO3 production and acidemia. ■ Increased minute ventilation is the primary cause of respiratory alkalosis, characterized by decreased Paco2 and increased pH. ■ Metabolic acidosis can be caused by one of three mechanisms: (1) increased production of acids, (2)
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
decreased renal excretion of acids, or (3) loss of alkali. The causes of metabolic acidosis can be divided into those that create an elevation in the AG and those that do not. ■ Metabolic alkalosis is usually caused by an increase in HCO3− reabsorption secondary to volume, potassium, or Cl− loss. ■ Contraction alkalosis can result from extracellular volume reduction, with a consequent increase in the plasma HCO3− concentration, when combined salt and water losses occur. This typically occurs in patients using diuretics. ■ Determination of a mixed acid-base disorder requires knowledge of the pH, calculation of the AG, and calculation of the ΔG.
Chapter 123
Electrolyte Disturbances
Michael A. Gibbs and Vivek S. Tayal
■ PERSPECTIVE An electrolyte is any substance that has free ions and therefore can conduct an electrical charge when in solution. The principle electrolytes in human physiology are sodium (Na+), potassium (K+), calcium (Ca2+), magnesium (Mg2+), chloride (Cl−), and hydrogen phosphate (HPO42−). Gradients of electrolytes between the intracellular and extracellular spaces are carefully maintained and are responsible for the electrical conduction required for muscles and nerves to function. The electrolyte concentrations of the body are maintained principally through kidney function, but also through action of hormones such as antidiuretic hormone, aldosterone, and parathyroid hormone. Dysfunction of any of these mechanisms or even severe physiologic stress, can disrupt electrolyte balance and result in a life-threatening emergency.
■ SODIUM Normal Physiology Since electrolytes exist in solution in the human body, fluid balance and electrolyte balance are linked, and under hormonal control. Water makes up approximately 60% of body weight and is distributed in three compartments: the intracellular space, the interstitial space, and the intravascular space. The intracellular space makes up approximately two thirds of total body water, with the remaining one third in the interstitial and intravascular spaces. The concentration of sodium [Na+], the predominant extracellular cation, governs the movement of water among these three compartments. When the extracellular [Na+] decreases, water shifts to the intracellular space to restore osmotic equilibrium. When the extracellular [Na+] rises, water shifts out of the intracellular space. Under normal conditions, Na+ leaks passively into cells down a concentration gradient and is transported back out of the cell by the sodiumpotassium adenosine triphosphatase (Na+-K+ ATPase) pump. Na+ homeostasis and water balance are under the hormonal regulation of the renin-angiotensin system and antidiuretic hormone, respectively. Renin, an enzyme produced by the kidney, is released in response to decreases in circulating intravascular volume. Renin catalyzes the production of angiotensin I, which is then converted to angiotensin II in the lung. Angiotensin II stimulates the production of aldosterone, a mineralocorticoid hormone produced by the zona glomerulosa of the adrenal glands. Aldosterone enhances Na+ reabsorption and K+ excretion in the distal nephron.
Antidiuretic hormone (ADH, vasopressin, arginine vasopressin) is synthesized in the hypothalamus and secreted from the posterior pituitary. ADH is released primarily in response to rises in serum osmolality but also to decreases in intravascular volume or arterial pressure. Volume depletion is the most potent stimulus for ADH production, and with decreases in plasma volume, ADH may be secreted even in the face of hypotonicity. ADH enhances renal water reabsorption by increasing tubular water permeability. Other factors that may stimulate ADH release include angiotensin, catecholamines, opiates, caffeine, hypoglycemia, hypoxia, and stress.
Hyponatremia Principles of Disease Hyponatremia is defined as a serum [Na+] of less than 135 mEq/ L. Hyponatremia can be classified into three categories based on the patient’s clinical volume status: (1) hypovolemic hyponatremia, (2) euvolemic hyponatremia, and (3) hypervolemic hyponatremia (Box 123-1). When assessing the patient with a low serum Na+ level, it is also important to consider the possibility of sampling errors (e.g., phlebotomy from a venous site proximal to an infusion of hypotonic solution), as well as pseudohyponatremia and redistributive hyponatremia. Pseudohyponatremia. Pseudohyponatremia refers to a falsely low serum Na+ measurement in patients whose plasma contains excessive protein or lipid. The relative percentage of water in plasma is reduced. Flame photometry, which determines Na+ content per unit of plasma, shows an artifactually low Na+ level, although both the total Na+ content and the serum osmolarity remain within the normal range. Measurement of the serum Na+ by direct potentiometry avoids this problem.1 Redistributive Hyponatremia. Redistributive hyponatremia is caused by osmotically active solutes in the extracellular space that draw water from the cell, diluting the serum [Na+]. Common situations causing such hyperosmolar states include hyperglycemia (e.g., diabetic ketoacidosis [DKA]) and parenteral administration of mannitol or glycerol for the management of intracranial hypertension or glaucoma. The measured serum Na+ in patients with hyperglycemia can be corrected by adding approximately 1.6 mEq/L for every 100-mg/dL rise in the serum glucose over 100 mg/dL. Hypovolemic Hyponatremia. Hypovolemic hyponatremia results from the loss of water and Na+ with a greater relative loss of Na+. Typical causes include vomiting, diarrhea, gastrointesti1615
1616
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
BOX 123-1 Causes of Hyponatremia Sampling error Pseudohyponatremia Hyperlipidemia Hyperproteinemia Redistributive type Hyperglycemia Mannitol Hypovolemic type Renal losses Gastrointestinal Third-space losses Excessive sweating Addison’s disease Euvolemic type SIADH Psychogenic polydipsia Hypervolemic type Congestive heart failure Hepatic cirrhosis Nephrotic syndrome
BOX 123-2
Causes of Syndrome of Inappropriate Secretion of Antidiuretic Hormone
CNS Disease Brain tumor, infarction, injury, or abscess Meningitis Encephalitis Pulmonary Disease Pneumonia Tuberculosis Lung abscess Pulmonary aspergillosis Drugs Exogenous vasopressin Diuretics Chlorpropamide Vincristine Thioridazine Cyclophosphamide CNS, central nervous system.
SIADH, syndrome of inappropriate secretion of antidiuretic hormone.
Clinical Features nal suction or drainage tubes, fistulas, and “third spacing” of fluids (e.g., burns, intra-abdominal sepsis, bowel obstruction, pancreatitis). Causes specifically attributable to renal losses include diuretic use, mineralocorticoid deficiency, renal tubular acidosis, and salt-wasting nephropathy. When Na+ losses are sufficient to decrease the glomerular filtration rate (GFR) significantly, the amount of filtrate delivered to the loop of Henle (where free water is generated) is decreased, and little free water appears in the urine. Also, because ADH is released in response to intravascular volume deficits despite hypotonicity, hyponatremia may be maintained even in patients whose GFR would otherwise be adequate to excrete excess free water. Hypovolemic hyponatremia can also be worsened when fluid losses are replaced with hypotonic fluids. Euvolemic Hyponatremia. The many causes of euvolemic hyponatremia include the syndrome of inappropriate secretion of ADH (SIADH), defined as the secretion of ADH in the absence of an appropriate physiologic stimulus. Its hallmark is an inappropriately concentrated urine despite the presence of a low serum osmolality and a normal circulating blood volume. Causes of SIADH include central nervous system (CNS) disorders, pulmonary disease, drugs, stress, pain, and surgery (Box 123-2). Before the diagnosis of SIADH can be confirmed, other potential causes of euvolemic hyponatremia (e.g., hypoadrenalism, hypothyroidism, renal failure) should be ruled out. Psychogenic polydipsia is a rare cause of euvolemic hyponatremia. This is most often seen in patients with psychiatric disorders who consume large volumes of water, usually in excess of 1 L/hr, overwhelming the capacity of the kidneys to excrete free water in the urine.2 In contrast to SIADH, the urine in patients with psychogenic polydipsia is maximally dilute. Hypervolemic Hyponatremia. Hypervolemic hyponatremia results when Na+ is retained but retention of water exceeds that of Na+. This is seen in edematous states such as congestive heart failure, hepatic cirrhosis, and renal failure. In these conditions, decreased effective renal perfusion causes the secretion of both ADH and aldosterone. This leads to increased tubular reabsorption of both Na+ and water, decreased delivery of water to the distal nephron, and inability to produce hypotonic urine.
The primary symptoms of hyponatremia are CNS symptoms, including lethargy, apathy, confusion, disorientation, agitation, depression, and psychosis. Focal neurologic deficits, ataxia, and seizures have been reported.3 Other nonspecific signs and symptoms include muscle cramps, anorexia, nausea, and weakness. The signs and symptoms of hyponatremia depend on the rapidity with which the serum [Na+] declines, as well as on its absolute level. The acutely hyponatremic patient is almost always symptomatic when the serum Na+ level falls below 120 mEq/L, whereas patients with chronic hyponatremia may tolerate much lower levels. Very young and very old patients typically develop symptoms with lesser decreases in the serum Na+ level.
Diagnostic Strategies The urinary [Na+] can be a useful tool in the assessment of the patient with hyponatremia. Patients with hypovolemic hyponatremia caused by renal Na+ wasting typically have an inappropriately high urinary [Na+] (>20 mEq/L); those with extrarenal Na+ wasting and intact renal Na+-conserving mechanisms have a low urinary [Na+] (14 mg/dL) require rapid treatment regardless of symptoms. The four basic goals of therapy are (1) restoration of intravascular volume, (2) enhancement of renal calcium elimination, (3) reduction of osteoclastic activity, and (4) treatment of the primary disorder (Box 123-11). Although it may not be realistic to expect to achieve these goals in the emergency department, it is important for the emergency physician to initiate therapy and involve the appropriate consultants as early as possible. Fluid Administration. The administration of isotonic saline is the first step in the management of severe hypercalcemia. Once the intravascular volume has been restored to normal, the serum calcium level will usually have decreased by 1.6 to 2.4 mg/dL, although hydration alone rarely leads to complete normalization. The expansion of intravascular volume increases renal calcium clearance by increasing GFR and Na+ delivery to the distal tubules. The rate of fluid administration should be based on the severity of hypercalcemia, the degree of dehy-
BOX 123-11 Management of Hypercalcemia Restoration of intravascular volume Correct dehydration with isotonic solution Correct associated electrolyte abnormalities Enhancement of renal calcium elimination Saline diuresis Loop diuretics (e.g., furosemide) Avoid thiazide diuretics Reduction of osteoclastic activity (Consult specialist for agent selection dosing.) Bisphosphonates Etidronate Pamidronate Zoledronic acid Calcitonin Hydrocortisone Treatment of primary disorder Parathyroidectomy for hyperparathyroidism Withdrawal of causative medications Treatment of nonparathyroid endocrine disorders
dration, and the patient’s cardiovascular tolerance of acute volume expansion. In elderly patients and those with poor left ventricular function, central venous pressure monitoring can be used to adjust fluid administration rates. Two to 5 L per day is often required. Coexisting electrolyte deficiencies should also be corrected. Furosemide. Loop diuretics such as furosemide inhibit the resorption of calcium in the thick ascending loop of Henle, increasing the calciuric effect of hydration. Volume expansion must precede the administration of furosemide, however, because the drug’s effect depends on the delivery of calcium to the distal nephron. IV doses of 10 to 40 mg every 6 to 8 hours are usually sufficient. Thiazide diuretics should not be used because they enhance distal absorption of calcium and may worsen hypercalcemia. Osteoclast Inhibitors. Therapy for severe hypercalcemia should also include agents that reduce the mobilization of calcium from bone. Drugs that inhibit osteoclast-mediated bone resorption include the bisphosphonates, calcitonin, glucocorticoids, and gallium nitrate. Because these drugs are used very infrequently in the emergency department, consultation with a specialist and/or pharmacist to select the best agent and dosing strategy is advised. The bisphosphonates act by inhibiting osteoclastic bone resorption and decreasing the viability of osteoclasts.43 Etidronate, palmidronate, and zoledronic acid have similar efficacy and a reasonable adverse effect profile.44-47 Calcitonin is a naturally occurring hormone that lowers serum calcium by inhibiting osteoclastic activity. Among the anticalcemic agents available, calcitonin has the most rapid onset of action, although it causes only a modest reduction in the serum calcium level.48 When hypercalcemia is severe and the need to lower the serum calcium is urgent, it is reasonable to administer a dose of calcitonin in combination with a more potent agent such as a bisphosphonate. The glucocorticoids act by inhibiting the action of vitamin D. They may be effective calcium-lowering agents in patients with hypercalcemia caused by hematologic malignancies, granulomatous disorders, or vitamin D intoxication. Underlying Cause. Pharmacologic therapy does not permanently normalize the serum calcium concentration. The underlying cause of the hypercalcemia needs to be treated as well. Primary hyperparathyroidism is definitively managed by parathyroidectomy. In the hands of experienced surgeons, more than 90% of patients are cured. When hypercalcemia is caused by malignancy, treatment must be directed at the underlying tumor because normocalcemia is difficult to sustain without successful treatment of the underlying cause. Hypercalcemia caused by medication responds to discontinuation of the offending agent. Hypercalcemia caused by nonparathyroid endocrine disease responds to treatment of the underlying disorder.
■ MAGNESIUM Normal Physiology Magnesium (Mg2+) is the second most abundant intracellular cation. It is a cofactor in hundreds of enzymatic reactions, including all those involving adenosine triphosphate (ATP). Magnesium is essential for the production and use of energy, DNA, and protein synthesis, ion channel gating, hormone receptor binding, neurotransmission, cardiac excitability, and muscle contraction.49 The adult human body contains approximately 2000 mEq of magnesium. One half of total magnesium is in the mineral
Hypomagnesemia Principles of Disease Hypomagnesemia is one of the most common electrolyte deficiencies in clinical practice.49 Approximately 10 to 20% of hospitalized patients and 50 to 60% of patients admitted to the intensive care unit are hypomagnesemic.51 Despite the high prevalence of hypomagnesemia, several factors can make the diagnosis a challenge. First, the clinical manifestations of hypomagnesemia are nonspecific, so the disorder is often overlooked. Second, the serum magnesium level is not measured as part of the “routine” electrolyte panel.52 Third, the serum magnesium level is an insensitive indicator of magnesium deficiency. Although a low serum magnesium level is indicative of a magnesium deficit, patients with a normal magnesium level may still have a severe deficiency. Fourth, hypomagnesemia often coexists with and may be masked by other electrolyte deficiency states. Numerous studies have demonstrated the high prevalence of hypomagnesemia in patients with hypokalemia.53 Because magnesium is required for the normal functioning of the Na+K+ ATPase pump, hypomagnesemia can result in refractory hypokalemia that is not correctable by the administration of potassium alone. Magnesium replacement enhances potassium retention and decreases the amount of supplemental potassium required to achieve a net positive balance.54 Magnesium is also required for the normal synthesis and release of PTH. Patients with hypomagnesemic hypocalcemia typically have inappropriately low levels of PTH and target organ resistance to the hormone, which are corrected by magnesium administration. A high prevalence of hypophosphatemia in patients who are hypomagnesemic has also been described. Because the kidneys normally conserve magnesium efficiently, significant hypomagnesemia usually occurs only when there is renal magnesium wasting or when intestinal losses exceed dietary intake and absorption (Box 123-12). In the emergency department, hypomagnesemia is most often associated with the use of diuretics and with alcohol abuse. Diuretics. Patients taking diuretics for the treatment of hypertension, congestive heart failure, or both are at significant risk for hypomagnesemia. Both the thiazide and the loop diuretics promote renal magnesium loss and may cause severe magnesium deficiency.55 In one study, typical diuretic doses increased urinary magnesium excretion by 25 to 50%. Some authors recommend that all patients receiving diuretics be considered candidates for magnesium supplementation. The use of a
1627
BOX 123-12 Causes of Hypomagnesemia Alcohol abuse Diuretic use Renal losses Acute and chronic renal failure Postobstructive diuresis Acute tubular necrosis Chronic glomerulonephritis Chronic pyelonephritis Interstitial nephropathy Renal transplantation Gastrointestinal losses Chronic diarrhea Nasogastric suctioning Short-bowel syndrome Protein-calorie malnutrition Bowel fistula Total parenteral nutrition Acute pancreatitis Endocrine disorders Diabetes mellitus Hyperaldosteronism Hyperthyroidism Hyperparathyroidism Acute intermittent porphyria Pregnancy Drugs Aminoglycosides Amphotericin Beta-agonists Cisplatin Cyclosporine Diuretics Foscarnet Pentamidine Theophylline Congenital disorders Familial hypomagnesemia Maternal diabetes Maternal hypothyroidism Maternal hyperparathyroidism
potassium-sparing diuretic in conjunction with a conventional diuretic is less likely to cause hypomagnesemia because these agents also have a magnesium-sparing effect. Alcoholism. The reported prevalence of hypomagnesemia in alcoholic patients varies widely, from 30 to 80%.56,57 Hypomagnesemia in the alcoholic patient is multifactorial; potential causes include poor nutrition, increased urinary excretion, gastrointestinal losses from vomiting and diarrhea, and pancreatic insufficiency. Renal, Gastrointestinal, and Endocrine Disorders. Hypomagnesemia can also result from renal magnesium wasting or from decreased production of (or end-organ responsiveness to) PTH.58 Magnesium wasting may be seen in some patients with post obstructive diuresis, acute tubular necrosis, chronic glomerulonephritis, chronic pyelonephritis, or interstitial nephropathy, as well as after renal transplantation. The decreased magnesium excretion typically found with acute and chronic renal failure, however, generally results in these patients tending to be hypermagnesemic.
Chapter 123 / Electrolyte Disturbances
component of bone, and 40 to 50% is found in the intracellular compartment. Only 1 to 2% of the body’s magnesium is present in the extracellular fluid, so the serum magnesium level is often a poor reflection of the total magnesium content. One third of the serum magnesium is bound to albumin, with the rest in the biologically active ionized form. The normal range for serum magnesium is 1.8 to 3.0 mg/dL. A balance between gastrointestinal absorption and renal excretion maintains magnesium homeostasis. Dietary sources of magnesium include green vegetables, meats, fish, beans, nuts, and grains. Absorption of ingested magnesium occurs in the small intestine through both active and passive transport mechanisms. In the kidney, 95% of the filtered load of magnesium is reabsorbed in the proximal tubule and loop of Henle.50 In deficiency states, magnesium resorption is enhanced in the distal convoluted tubule under the influence of PTH. In hypermagnesemic states, renal excretion of magnesium increases.
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1628
Gastrointestinal causes of hypomagnesemia include shortbowel syndrome, protein-calorie malnutrition, bowel fistula, continuous nasogastric suctioning, chronic diarrhea, and administration of total parenteral nutrition.58 Patients with acute pancreatitis typically have an intracellular magnesium deficiency despite normal serum concentrations. This is most likely in patients who are also hypocalcemic. Hypomagnesemia is the most common electrolyte abnormality in ambulatory diabetic patients and is also a common finding in patients with DKA. Excessive urinary loss associated with glycosuria and transcellular shifts of the cation are the proposed mechanisms. The clinical consequences of magnesium deficiency include impairment of insulin secretion and peripheral insulin resistance. Hypomagnesemia also may play a role in the development of retinopathy, hypertension, and the abnormal platelet function often observed in diabetic patients. Other endocrine and metabolic causes of hypo magnesemia include primary and secondary aldosteronism, hyperthyroidism, primary hyperparathyroidism, and acute intermittent porphyria. Pregnancy. Pregnancy is marked by a state of hypomagnesemia. Serum levels usually decline in the third trimester. Patients with preterm labor are more likely to have a significantly depressed serum magnesium level.59 Drugs. Hypomagnesemia has also been associated with a number of drugs.60 It can result from renal magnesium wasting (e.g., aminoglycosides, amphotericin B, cisplatin, diuretics, foscarnet, pentamidine) or from transcellular magnesium shifts (e.g., beta-agonists, cyclosporine, theophylline).61 Congenital Disorders. Congenital disorders causing hypomagnesemia include primary infantile hypomagnesemia and familial hypomagnesemia. Maternal diabetes, maternal hyperparathyroidism, and maternal hypothyroidism are also associated with hypomagnesemia in the newborn.
Clinical Features The clinical manifestations of hypomagnesemia are nonspecific and can easily be confused with those caused by other metabolic abnormalities. Symptoms are inconsistent, variable in severity, and not well correlated with a specific serum magnesium level. However, patients are usually symptomatic at serum levels of 1.2 mg/dL or less. The clinical manifestations of hypomagnesemia most likely to be prominent in the emergency setting involve the neuromuscular and cardiovascular systems. Neuromuscular manifestations include muscle weakness, tremor, hyper-reflexia, tetany, and a positive Chvostek’s or Trousseau’s sign. CNS findings range from apathy, irritability, and dizziness to seizures, papilledema, and coma. Focal neurologic findings have also been described. Dysrhythmia is the most common cardiovascular manifestation of hypomagnesemia. A number of studies demonstrate an increased incidence of supraventricular dysrhythmias (atrial fibrillation, multifocal atrial tachycardia, paroxysmal supraventricular tachycardia) and ventricular dysrhythmias (premature ventricular contractions, ventricular tachycardia, torsades de pointes, ventricular fibrillation) in patients who are magnesium deficient.62 Patients taking diuretics for the treatment of congestive heart failure are particularly vulnerable. Digitalisinduced dysrhythmias are also more likely in the presence of hypomagnesemia. Because magnesium is an essential cofactor for the Na+-K+ ATPase pump that is inhibited by digitalis, hypomagnesemia typically worsens the manifestations of digitalis toxicity. Hypomagnesemia has been associated with a wide range of ECG findings, including prolongation of the PR, QRS, and
QT intervals; ST-T segment abnormalities; flattening and widening of the T wave; and presence of U waves. These findings, however, are nonspecific and may be at least partly caused by associated hypokalemia. Thus, the ECG should not be used to rule out magnesium disturbances. The relationship between hypomagnesemia and ischemic heart disease is controversial. Hypomagnesemia is common in emergency department patients with chest pain and in those admitted to the coronary care unit.62 Patients who have a myocardial infarction are more likely to be hypomagnesemic than those who do not. This finding has been shown to be independent of concomitant diuretic use. Serum magnesium levels decline transiently after acute myocardial infarction, increasing the risk of dysrhythmia.63,64 Proposed mechanisms include transcellular shifts of the cation and chelation with free fatty acids released after acute myocardial infarction. Although several studies demonstrate a benefit of empiric magnesium administration after acute myocardial infarction, the largest trial to date, the International Study of Infarct Survival, failed to confirm a significant benefit.65
Management Because it is often an inaccurate reflection of total magnesium stores, the serum magnesium level should not be used alone to guide therapy. However, magnesium administration is appropriate in patients with a low serum level (350 low 130s ≈ 4.5–6.0 700 140s ≈5 >15 >50 Absent
BUN, blood urea nitrogen.
of all episodes of DKA occur in patients whose diabetes was previously undiagnosed.6,38
Diagnostic Strategies History Clinically, most patients with DKA complain of a recent history of polydipsia, polyuria, polyphagia, visual blurring, weakness, weight loss, nausea, vomiting, and abdominal pain. Approximately one half of these patients, especially children, report abdominal pain. In children this pain is usually idiopathic and probably caused by gastric distention or stretching of the liver capsule; it resolves as the metabolic abnormalities are corrected. In adults, however, abdominal pain more often signifies true abdominal disease.
Physical Examination Physical examination may or may not demonstrate a depressed sensorium. Typical findings include tachypnea with Kussmaul’s respiration, tachycardia, frank hypotension or orthostatic blood pressure changes, the odor of acetone on the breath, and signs of dehydration.39 An elevated temperature is rarely caused by DKA itself and suggests the presence of infection.
Laboratory Tests Initial tests allow preliminary confirmation of the diagnosis and immediate initiation of therapy (Table 124-2). Subsequent tests are made to determine more specifically the degree of dehydration, acidosis, and electrolyte imbalance and to reveal the precipitant of DKA. On the patient’s arrival to the ED, serum and urine glucose and ketones, electrolytes, and arterial blood gases (ABGs) should be checked. Glucose is usually elevated above 350 mg/ dL; however, euglycemic DKA (blood glucose < 300 mg/dL) has been reported in up to 18% of patients.4 ABGs demonstrate a low pH. Venous pH is not significantly different from arterial pH in patients with DKA, and some researchers con-
1641
lipase determination helps to distinguish pancreatitis from elevated salivary amylase levels.
Differential Considerations Alcoholics, especially those who have recently abstained from drinking, with Kussmaul’s respiration, a fruity odor to the breath, and acidemic ABG values may have alcoholic ketoacidosis. These patients may be euglycemic or hypoglycemic, and a large part of their acidosis is often caused by the unmeasured β-hydroxybutyric acid. Alcoholic ketoacidosis accounts for approximately 20% of all cases of ketoacidosis. Ketoacidosis can also develop with fasting in the third trimester of pregnancy and in nursing mothers who do not eat.41 Other entities that may manifest with various combinations of altered mental status, acidosis, and abdominal pain include hypoglycemia, cerebrovascular accident (stroke), trauma, sepsis, hyperglycemic hyperosmolar nonketotic coma, postictal states, lactic acidosis, uremic acidosis, and abdominal emergencies. Intoxications by ethanol, salicylates, methanol, isopropyl alcohol, chloral hydrate, paraldehyde, ethylene glycol, and cyanide all share some features of DKA.
Management General Measures The approach to the patient with severe DKA is the same as that to any patient in extremis. The comatose patient, especially if vomiting, requires intubation. Once the patient is intubated, hyperventilation should be maintained to prevent worsening acidosis. The patient in hypovolemic shock requires aggressive fluid resuscitation with 0.9% saline solution rather than pressors. Other possible causes of shock (e.g., sepsis or myocardial dysfunction secondary to MI) should be considered. Close monitoring of vital signs is essential. In the patient whose therapy may precipitate fluid overload caused by cardiac compromise or renal failure, a central venous pressure line or Swan-Ganz catheter should be inserted. The diagnosis of DKA is generally simple. When hyperglycemia, ketosis, and acidosis have been established, fluid, electrolyte, and insulin therapy should begin (Box 124-7).
Insulin DKA cannot be reversed without insulin, and insulin therapy should be initiated as soon as the diagnosis is certain. In the past, very high dosages of insulin were administered to diabetic patients in DKA because they were thought to be extremely insulin-resistant. However, low-dosage insulin therapy has proved as effective as high-dosage therapy.6 The rate of decrease in blood sugar is equal or only slightly more gradual. The overall potassium requirement is less. High dosages of insulin have potentially harmful effects, including a greater incidence of iatrogenic hypoglycemia and hypokalemia.36 The exact amount of insulin administered varies. Many start therapy with a bolus of 10 U of regular insulin intravenously. This initial bolus may produce certain problems, however, and makes no significant difference in therapy.42 The current therapy of choice is regular insulin infused at 0.1 U/kg/hr up to 5 to 10 U/kg/hr, mixed with the IV fluids. Regular insulin, 10 to 20 U/hr, administered intramuscularly accomplishes similar effects but subjects the patient to repeated painful injections. In theory, intramuscular insulin may accumulate at a poorly perfused administration site, failing to enter the systemic circulation in a timely manner.
Chapter 124 / Diabetes Mellitus and Disorders of Glucose Homeostasis
sider the use of venous blood superior to repeated arterial puncture. Metabolic acidosis with an anion gap is primarily the result of elevated plasma levels of acetoacetate and β-hydroxybutyrate, although lactate, FFAs, phosphates, volume depletion, and several medications also contribute to this condition.4 Rarely, a well-hydrated patient with DKA may have a pure hyperchloremic acidosis with no anion gap. If an immediate potassium level is not available through ABG, an electrocardiogram may indicate potassium levels. Despite initial potassium levels that are normal to high, a total potassium deficit of several hundred milliequivalents results from potassium and hydrogen shifts. Other tests may include complete blood count with differential, magnesium, calcium, amylase, blood urea nitrogen (BUN), creatinine, phosphorus, ketone, and lactate level determinations. A complete urinalysis helps in the determination of possible infection or renal disease. Elevations of urine specific gravity, BUN, and hematocrit suggest dehydration. Appropriate cultures should be dictated by clinical findings. The serum sodium value is often misleading in DKA. Sodium is often low in the presence of significant dehydration because it is strongly affected by hyperglycemia; hypertriglyceridemia; salt-poor fluid intake; and increased GI, renal, and insensible losses. When hyperglycemia is marked, water flows from the cells into the vessels to decrease the osmolar gradient, thereby creating dilutional hyponatremia. Lipids also dilute the blood, thereby further lowering the value of sodium. Newer autoanalyzers remove triglycerides before assay, thus eliminating this artifact. Hypertriglyceridemia is common in DKA because of impaired lipoprotein lipase activity and hepatic overproduction of very-low-density lipoproteins.4 In the absence of marked lipidemia, the true value of sodium may be approximated by adding 1.3 to 1.6 mEq/L to the sodium value on the laboratory report for every 100 mg/dL glucose over the norm. Thus, if the laboratory reports a serum sodium value of 130 mEq/L and a blood glucose value of 700 mEq/L, the total serum sodium value is more accurately assessed to be between 137.8 and 139.6 mEq/L. Acidosis and the hyperosmolarity induced by hyperglycemia shift potassium, magnesium, and phosphorus from the intracellular to the extracellular space. Dehydration produces hemoconcentration, which contributes to normal or high initial serum potassium, magnesium, and phosphorus readings in DKA, even with profound total body deficits. The effect of acidosis on the serum potassium determination can be corrected by subtracting 0.6 mEq/L from the laboratory potassium value for every 0.1 decrease in pH noted in the ABG analysis.40 Thus, if the potassium is reported as 5 mEq/L and the pH is 6.94, the corrected potassium value would be only 2 mEq/L, representing severe hypokalemia. While insulin is administered and the hydrogen ion concentration decreases, the patient needs considerable potassium replacement. Finally, hyperglycemia and the anion gap have significant effects on the plasma potassium concentration, independent of acidosis.37 No conversion factor has been developed for estimating true magnesium levels, although initial values may be high. All laboratory determinations must be interpreted with caution. Serum creatinine determinations made by autoanalyzer may be falsely elevated.36 Leukocytosis more closely reflects the degree of ketosis than the presence of infection. Only the elevation of band neutrophils has been demonstrated to indicate the presence of infection, with a sensitivity of 100% and a specificity of 80%. The diagnosis of pancreatitis is confounded by the usually elevated urine and serum amylase levels in DKA. Typically, this is salivary amylase, but most laboratories are not equipped to make this distinction. A serum
1642
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
BOX 124-7 Summary of Treatment for Diabetic Ketoacidosis Identify DKA: serum glucose, electrolytes, ketones, and ABG; also draw CBC with differential; urinalysis; chest radiograph film and ECG, if indicated. 1. Supplement insulin. ± Bolus: 0.1 U/kg regular insulin IV Maintenance: 0.1 U/kg regular insulin IV Change IV solution to D5W 0.45% NS when glucose ≤ 300 mg/dL. 2. Rehydrate. 1–2 L NS IV over 1–3 hours Children: 20 mL/kg NS over first hour Follow with 0.45% NS 3. Correct electrolyte abnormalities. Sodium Correct with administration of NS and 0.45% NS. Potassium Ensure adequate renal function. Add 20–40 mEq KCl to each liter of fluid. Phosphorus Usually unnecessary to replenish Magnesium Correct with 1–2 g MgSO4 (in first 2 L if magnesium is low). 4. Correct acidosis. Add 44–88 mEq/L to first liter of IV fluids only if pH ≤ 7.0. Correct to pH 7.1. 5. Search for and correct underlying precipitant. 6. Monitor progress and keep meticulous flow sheets. Vital signs Fluid intake and urine output Serum glucose, K+, Cl−, HCO3−, CO2, pH Amount of insulin administered 7. Admit to hospital or intensive care unit. Consider outpatient therapy in children with reliable caretaker and Initial pH > 7.35 Initial HCO3− ≥ 20 mEq/L Can tolerate PO fluids Resolution of symptoms after treatment in emergency department No underlying precipitant requiring hospitalization ABG, arterial blood gas; CBC, complete blood count; DKA, diabetic ketoacidosis; ECG, electrocardiogram; NS, normal saline; PO, by mouth.
In children, the IV dosage of regular insulin may be calculated at 0.1 U/kg. Children are more likely than adults to develop cerebral edema in response to a rapid lowering of plasma osmolarity. Thus, reduction of glucose levels in children should be gradual. Because the half-life of regular insulin is 3 to 10 minutes, IV insulin should be administered by constant infusion rather than by repeated bolus. When the blood glucose has dropped to 250 to 300 mg/dL, dextrose should be added to the IV fluids to prevent iatrogenic hypoglycemia and cerebral edema. In patients with euglycemic DKA, dextrose should be added to the IV fluids at the start of insulin therapy. Insulin adheres to the walls of glass and polyvinyl bottles and tubing, making the exact amount of insulin being administered uncertain.43 Running approximately 10 U of the insulin infusion through the tubing accomplishes adherence without altering the delivered concentration of the remainder of the infusate.
Insulin resistance occurs rarely in diabetic patients and requires an increase in dosage to obtain a satisfactory response. Resistance may be caused by obesity or accelerated insulin degradation.4 Two general categories of insulin resistance are described.44 Postreceptor-binding resistance is the more common type. It is a mild to moderate resistance probably caused by defects in intracellular metabolism. It takes only a small amount of insulin to fill the receptors. When they are filled, administration of additional insulin produces no further effect. The second type, prereceptor-receptor resistance, is rare. This severe resistance may be caused by insulin antibodies, high concentrations of stress hormones, antireceptor antibodies, or any combination of these.45
Dehydration The severely dehydrated patient is likely to have a fluid deficit of 3 to 5 L. No uniformly accepted formula exists for the administration of fluid in this disorder. If the patient is in hypovolemic shock, normal saline (NS) should be administered as rapidly as possible in the adult, or in 20 mL/kg boluses in the child, until a systolic pressure of 80 mm Hg is obtained. In the adult who has marked dehydration in the absence of clinical shock or heart failure, 1 L of NS may be administered in the first hour. In general, 2 L of NS over the first 1 to 3 hours is followed by a slower infusion of half-NS solution. Patients with DKA without extreme volume depletion may be successfully treated with a lower volume of IV fluid replacement.38 NS solution at 20 mL/kg over the first hour is the usual fluid resuscitation therapy for a child. Thereafter, fluid rate should be adjusted according to age, cardiac status, and degree of dehydration to achieve a urine output of 1 to 2 mL/kg/hr. Whereas some authors advocate half-NS or colloid solution, most evidence and practice favor initial resuscitation with 0.9% NS solution. Fluid resuscitation alone may help to lower hyperglycemia. Because even in DKA a low level of circulating insulin may be present, increased perfusion may transport insulin to previously unreached receptor sites. In addition, a large volume of glucose may be cleared by the kidneys in response to improved renal perfusion. The mean plasma glucose concentration has been noted to drop 18% after administration of saline solution without insulin.6 Acidosis also decreases after fluid infusion alone. Increased perfusion improves tissue oxygenation, thus diminishing the formation of lactate. Increased renal perfusion promotes renal hydrogen ion loss, and the improved action of insulin in the better-hydrated patient inhibits ketogenesis. Some authors believe that the rapid decrease of the hyperosmolarity of DKA caused by the administration of 0.45% NS solution may precipitate cerebral edema, one of the most dangerous complications associated with the patient in DKA, especially children.46-48
Potassium Potassium replacement is invariably needed in DKA. The initial potassium level is often normal or high despite a large deficit because of severe acidosis. Potassium levels often plummet with correction of acidosis and administration of insulin. Potassium should be administered with the fluids while the laboratory value is in the upper half of the normal range. Renal function should be monitored. In patients with low serum potassium at presentation, hypokalemia may become life-threatening when insulin therapy is administered.
Magnesium Magnesium deficiency is a common problem in patients with DKA without renal disease. Both the initial pathophysiology and the therapy for DKA induce profound magnesium diuresis. Magnesium deficiency may exacerbate vomiting and mental changes, promote hypokalemia and hypocalcemia, or induce fatal cardiac dysrhythmia. The normal person requires 0.30 to 0.35 mEq/kg/day. Thus, it is reasonable to include 0.35 mEq/kg of magnesium in the fluids of the first 3 to 4 hours, with further replacement dependent on blood levels and the clinical picture. This amounts to 1.0 to 3 g of magnesium sulfate in the 70-kg patient.
Acidosis Bicarbonate therapy may be indicated in severely acidemic patients (pH ≤ 7.0).6 The use of bicarbonate is not warranted in less ill patients for several reasons.49,50 1. Bicarbonate worsens the inhibition of oxygen release from red blood cells caused by the 2,3-DPG deficiency seen in phosphorus-depleted patients with DKA. 2. Overly rapid correction of acidosis is contraindicated because the blood-brain barrier (BBB) is much more permeable to carbon dioxide than to bicarbonate. Thus, the correction of intravascular acidosis terminates Kussmaul’s respiration, further augmenting the blood carbon dioxide available to cross the BBB. Slowly, sufficient bicarbonate crosses the BBB to provide adequate buffering. In the short term, however, as the blood acidosis is corrected, the acidity of the fluid surrounding the brain increases, causing paradoxical cerebrospinal fluid acidosis. The clinical significance of an acid cerebrospinal fluid pH is controversial. 3. The administration of bicarbonate increases the potassium requirement, both immediately by driving potassium into the cell and more gradually by affecting the kidney, making iatrogenic hypokalemia more likely. When bicarbonate is used, serum potassium levels need to be followed even more closely. 4. The overaggressive use of bicarbonate may produce alkalosis, which induces dysrhythmias largely through its effect on the distribution of electrolytes. Alkalemia occurring late in the course of therapy is more common in patients who have received bicarbonate because ketones are metabolized to carbon dioxide, water, and bicarbonate. 5. Evidence suggests that lowered pH produces a feedback mechanism that directly inhibits ketogenesis. Bicarbonate
1643
can increase ketonuria and delay the fall in serum ketones compared with saline infusion alone. 6. Patients treated with bicarbonate fare no better and possibly fare worse than patients treated without bicarbonate. Studies indicate that bicarbonate worsens the prognosis even in patients with severe acidosis and pH values in the range 6.9 to 7.1. It is possible to manage severe DKA with fluids and insulin alone. When this is done, pH normalization is similar to that in a bicarbonate control group.
When bicarbonate therapy is deemed necessary, the pH should not be corrected above 7.1. Response to therapy should be followed initially with hourly vital signs; fluids should be administered and urine output measured; insulin should be given; and glucose, pH, and anion gap measurements taken. Plasma bicarbonate may remain low even while pH increases and anion gap narrows because of the hyperchloremia that develops from rapid saline infusion, loss of bicarbonate in the urine as ketones, and exchanges with intracellular buffers.
Complications The precipitating causes of DKA may have associated morbidity and mortality rates equal to or worse than those for DKA itself. These include iatrogenic causes as well as infection and MI. Morbidity in DKA is largely iatrogenic: (1) hypokalemia from inadequate potassium replacement, (2) hypoglycemia from inadequate glucose monitoring and failure to replenish glucose in IV solutions when serum glucose drops below 250 to 300 mg/dL, (3) alkalosis from overaggressive bicarbonate replacement, (4) congestive heart failure from overaggressive hydration, and (5) cerebral edema probably caused by too rapid osmolal shifts. DKA is responsible for 70% of diabetesrelated deaths in children. Poor prognostic signs include hypotension, azotemia, coma, and underlying illness.51 The mortality rate in treated DKA decreased from approximately 38% between 1930 and 1959 to about 5 to 7% in the 1980s. The primary causes of death remain infection (especially pneumonia), arterial thromboses, and shock. The decrease in mortality rate demonstrates that appropriate therapy can make a difference. Cerebral edema should be considered when the patient remains comatose or lapses into coma after the reversal of acidosis. It generally occurs 6 to 10 hours after the initiation of therapy. There are no warning signs, and the associated mortality rate is currently 90%. Cerebral edema has been associated with low partial pressures of arterial carbon dioxide, high BUN concentration, and the use of bicarbonate. Subclinical cerebral edema in children is probably very common. Furthermore, subclinical cerebral edema may either precede or follow the onset of therapy, raising the question of whether this entity is caused by therapy or is simply a manifestation of the basic pathophysiology of DKA. Because clinically evident cerebral edema does not usually occur unless the blood sugar level is below 250 mg/dL and insulin is being used, insulin may directly antagonize the brain’s defenses against fluid shifts while the plasma glucose level approaches normal values. Other theories attribute the formation of cerebral edema to (1) “idiogenic osmols” developed in the brain as a result of insulin therapy, (2) the rate of fluid administration, and (3) the rate of correction of the acidosis. Other less common causes have been suggested. Several authors recommend the administration of mannitol, 0.25 to 2 mg/kg, at the first sign of altered mental status in children being treated for DKA.47 Steroids are ineffective treatment for cerebral edema secondary to DKA and may worsen DKA.
Chapter 124 / Diabetes Mellitus and Disorders of Glucose Homeostasis
IV potassium should be vigorously administered in concentrations of 20 to 40 mEq/L as required. Some clinicians administer a portion of the potassium as the phosphate salt. In DKA, phosphate falls from a mean value of 9.2 to 2.8 mg/dL within 12 hours of therapy, reflecting an average total deficit of 0.5 to 1.5 mmol/kg.44 This may result in a decreased level of 2,3-diphosphoglycerate (2,3-DPG) and subsequent poor oxygen delivery to red blood cells. Other problems associated with the hypophosphatemia are depressed myocardial and respiratory muscle performance, hemolysis, impaired phagocytosis, thrombocytopenia, platelet dysfunction, confusion, and disorientation.6 The only caveat with phosphorus administration is that its magnesium- and calciumlowering properties may induce symptomatic hypomagnesemia and hypocalcemia. Despite theoretical benefits, no clinical benefit has been shown from the routine administration of phosphorus in DKA.6
1644
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
Disposition Most patients with DKA require hospital admission, often to the intensive care unit. All pregnant diabetic patients in DKA require admission and consultation with an endocrinologist and obstetrician specializing in the care of highrisk pregnancies. Some children (initial pH > 7.35, bicarbonate ≥ 20 mEq/L) with resolution of findings who can tolerate oral fluids after 3 or 4 hours of treatment may be discharged home with a reliable caregiver.52 Patients who have mild DKA may be treated on an outpatient basis if (1) the patient or parent is reliable, (2) the underlying causes do not require inpatient therapy, and (3) close follow-up is pursued.
■ HYPERGLYCEMIC HYPEROSMOLAR NONKETOTIC COMA Hyperglycemic hyperosmolar nonketotic coma (HHNC) represents a syndrome of acute diabetic decompensation characterized by marked hyperglycemia, hyperosmolarity and dehydration, and decreased mental functioning that may progress to frank coma. Ketosis and acidosis are generally minimal or absent. Focal neurologic signs are common. DKA and HHNC may occur together; some even consider HHNC and DKA to be at two ends of a spectrum, with many patients in the middle.6,53
Principles of Disease Pathophysiology As with DKA, the pathophysiology of HHNC varies with the particular patient. Because most patients with HHNC are elderly, decreased renal clearance of glucose produced by the decline of renal function with age often contributes to the illness. As with DKA, decreased insulin action results in glycogenolysis, gluconeogenesis, and decreased peripheral uptake of glucose. The hyperglycemia pulls fluid from the intracellular space into the extracellular space, transiently maintaining adequate perfusion. Soon, however, this fluid is lost in a profound osmotic diuresis, limited finally by hypotension and a subsequent drop in the glomerular filtration rate (GFR). The urine is extremely hypotonic, with a urine sodium concentration between 50 and 70 mEq/L, compared with 140 mEq/L in extracellular fluid. This hypotonic diuresis produces profound dehydration, leading to hyperglycemia, hypernatremia, and associated hypertonicity. Often the patient is prevented from taking in adequate fluids by stroke, Alzheimer’s disease, or other diseases, greatly exacerbating the dehydration of renal origin. The reason for the absence of ketoacidosis in HHNC is unknown. FFA levels are lower than in DKA, thus limiting substrates needed to form ketones. The most likely reason for the blunted counter-regulatory hormone release and lack of ketosis seems to be that these patients continue to secrete the tiny amount of insulin required to block ketogenesis.54
Etiology HHNC is a syndrome of severe dehydration that results from a sustained hyperglycemic diuresis under circumstances in which the patient is unable to drink sufficient fluids to offset the urinary losses. The full-blown syndrome does not usually occur until volume depletion has progressed to the point of decreased urine output. HHNC is most common in elders with type 2 diabetes but has been reported in children with type 1 diabetes.55
Precipitants of Hyperglycemic Hyperosmolar
BOX 124-8 Nonketotic Coma External Insult Trauma Burns Dialysis Hyperalimentation
Disease Process Cushing’s syndrome and other endocrinopathies Hemorrhage Myocardial infarction Renal disease Subdural hematoma Cerebrovascular accident Infection Down syndrome Drugs Antimetabolites l-Asparaginase Chlorpromazine Chlorpropamide Cimetidine Diazoxide Didanosine Ethacrynic acid Furosemide Glucocorticoids Immunosuppressants Olanzapine and other atypical antipsychotics Phenytoin Propranolol Thiazides
Box 124-8 lists the broad range of predisposing factors.5 HHNC may occur in patients who are not diabetic, especially after burns, parenteral hyperalimentation, peritoneal dialysis, or hemodialysis.56
Clinical Features The prodrome of HHNC is significantly longer than that of DKA. Clinically, extreme dehydration, hyperosmolarity, volume depletion, and CNS findings predominate.54 If awake, patients may complain of fever, thirst, polyuria, or oliguria. Approximately 20% of patients have no known history of type 2 diabetes. The most common associated diseases are chronic renal insufficiency, gram-negative pneumonia, GI bleeding, and gram-negative sepsis. Of these patients, 85% have underlying renal or cardiac impairment as a predisposing factor. Arterial and venous thromboses often complicate the picture. The patient often exhibits orthostatic hypotension or frank hypotension, tachycardia, and fever with signs of marked dehydration. On average, the HHNC patient has a 24% fluid deficit, or 9 L in the 70-kg patient. The depression of the sensorium correlates directly with the degree and rate of development of hyperosmolarity. Some patients have normal mental status. Seizures are usually associated with neurologic findings, especially epilepsia partialis continua (continuous focal seizures) and intermittent focal motor seizures. Stroke and hemiplegia are also common. Less common neurologic findings include choreoathetosis, ballismus, dysphagia,
Diagnostic Strategies Laboratory findings usually reveal a blood glucose level greater than 600 mg/dL and serum osmolarity greater than 350 mOsm/L. The BUN concentration is invariably elevated (see Table 124-2). Although patients with HHNC do not have a ketoacidosis caused by diabetes, they may have a metabolic acidosis secondary to some combination of lactic acidosis, starvation ketosis, and retention of inorganic acids attributable to renal hypoperfusion. The patient with HHNC typically manifests more profound electrolyte imbalance than the patient with DKA. Levels of potassium, magnesium, and phosphorus may seem initially high, even in the presence of marked total deficit. In the absence of acidemia, however, the discrepancy between the initial electrolyte reading and body stores is less than that of DKA. Initial serum sodium readings are inaccurate because of hyperglycemia.
Differential Considerations The differential diagnosis of HHNC is identical to that for DKA. In addition, diabetic patients receiving chlorpropamide are subject to water intoxication with dilutional hyponatremia, which may manifest as coma without acidosis that is clinically indistinguishable from HHNC. The patient with HHNC who has a sharply depressed sensorium may not be initially distinguishable from the patient with profound hypoglycemia. When blood glucose cannot be rapidly checked, the immediate administration of one ampule of D50W minimally worsens HHNC and may be lifesaving for patients with hypoglycemia.
Management The fluid, electrolyte, and insulin regimens for the initial resuscitation in HHNC are subject to the same controversies as the therapies for DKA (Box 124-9). Whereas some physicians use half-NS solution rapidly infused, most use NS solution, switching to half-NS later in the resuscitation. Just as in DKA, overly rapid correction of serum osmolarity may predispose to the development of cerebral edema in children. There are few reports of cerebral edema complicating HHNC in adults.
Dehydration Under central venous or pulmonary artery pressure monitoring, rapid administration of NS in a similar fashion to initial therapy for DKA is generally safe. For patients in coma or hypovolemic shock, initial IV fluid infusion should be given as rapidly as possible. If the patient does not have central monitoring and is not in coma or shock, vigorous IV fluid administration, such as 1 L/hr, is prudent and provides sufficient rehydration. In any circumstance, after 2 to 3 L of NS, 0.45% NS solution should be substituted. As the fluid deficits and hyperosmolarity are corrected, the infusion rate must be slowed and electrolytes managed. In patients with concomitant congestive heart failure, sterile water has been successfully administered by central venous line at a rate of 500 mL/hr, with no detectable hemolysis or other complications. Glucose should be added to resuscitation fluids when the blood glucose level drops below 300 mg/dL.
1645
Summary of Treatment for Hyperglycemic
BOX 124-9 Hyperosmolar Nonketotic Coma
Identify HHNC, then treatment is the same as initial DKA treatment. 1. Supplement insulin. ± Bolus: 0.05–0.1 U/kg regular insulin IV Maintenance: 0.05–0.1 U/kg regular insulin IV Caution: Serum glucose rapidly corrects with fluid administration alone; monitor glucose to avoid hypoglycemia. Change IV solution to D5W 0.45% NS when glucose ≤ 300 mg/dL. 2. Rehydrate. Rapid administration of 2–3 L NS over first several hours CVP monitoring may be necessary in patients with history of heart disease. Correct one half of fluid deficit in first 8 hours, remainder over 24 hours. 3. Correct electrolyte abnormalities. Sodium Correct with administration of NS and 0.45% NS. Potassium First ensure adequate renal function. Add 20–40 mEq KCl to each liter of fluid. Phosphorus Usually unnecessary to replenish Magnesium Correct with 1–2 g MgSO4 (in first 2 L if magnesium is low). 4. Correct acidosis. Add 44–88 mEq/L to first liter of IV fluids only if pH ≤ 7.0. Correct to pH 7.1. 5. Search for and correct underlying precipitant. 6. Monitor progress and keep meticulous flow sheets. Vital signs Fluid intake and urine output Serum glucose, K+, Cl−, HCO3−, CO2, pH, ketones Amount of insulin administered 7. Admit to hospital or intensive care unit. CVP, central venous pressure; DKA, diabetic ketoacidosis; HHNC, hyperglycemic hyperosmolar nonketotic coma; NS, normal saline.
Electrolytes The guidelines for the administration of potassium, magnesium, and phosphorus are similar to those for DKA.
Insulin Low-dosage insulin, such as that administered in the patient with DKA, is generally effective and safe when the restoration of volume has been instituted.
Other Considerations A vigorous search for the underlying precipitant for HHNC must be pursued. Response to therapy should be followed in the manner described for patients in DKA. Phenytoin (Dilantin) is contraindicated for the seizures of HHNC because it is often ineffective and may impair endogenous insulin release. Phenytoin-induced HHNC even occurs in nondiabetic patients. Low-dosage subcutaneous heparin may be
Chapter 124 / Diabetes Mellitus and Disorders of Glucose Homeostasis
segmental myoclonus, hemiparesis, hemianopsia, central hyperpyrexia, nystagmus, visual hallucinations, and acute quadriplegia.
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1646
indicated to lessen the risk of thrombosis, which is increased by the volume depletion, hyperviscosity, hypotension, and inactivity associated with HHNC.
Complications Reasons for high morbidity and mortality rates are not always clear, but many patients with HHNC are elders who have underlying cardiac and renal disease. Pediatric HHNC differs from adult HHNC in that children have a much higher incidence of fatal cerebral edema.57 Other causes for morbidity and mortality are similar to those described for DKA. The mortality rate of treated HHNC patients has been 40 to 70% in the past, but now ranges from 8 to 25%.58
Disposition Patients with HHNC should be hospitalized for IV hydration, glucose control, and evaluation of precipitating and complicating conditions.
■ LATE COMPLICATIONS OF DIABETES Late complications of diabetes cause significant morbidity and mortality. They develop approximately 15 to 20 years after the onset of overt hyperglycemia. The Diabetes Control and Complications Trial showed that tight glycemic control significantly reduces the risk of microvascular disease, such as microalbuminuria (the earliest sign of nephropathy), neuropathy, and retinopathy, but at the expense of greatly increasing the risk of recurrent hypoglycemia.59-61
Vascular Complications Diabetes is associated with an increased risk for atherosclerosis and thromboembolic complications, which are a major cause of morbidity and premature death.62 The cause of accelerated atherosclerosis is unknown, although it is probably related to oxidated low-density lipoprotein and increased platelet activity. Atherosclerotic lesions are widespread, causing symptoms in many organ systems. Coronary artery disease and stroke are common. Diabetic patients have an increased incidence of “silent” MI, complicated MIs, and congestive heart failure.7,63,64 Peripheral vascular disease is noted clinically by claudication, nonhealing ulcers, gangrene, and impotence.
Diabetic Nephropathy Renal disease is a leading cause of death and disability in diabetic patients. Approximately one half of end-stage renal disease in the United States is caused by diabetic nephropathy.65 Diabetic nephropathy involves two pathologic patterns: diffuse and nodular. Clinical renal dysfunction does not correlate well with the histologic abnormalities. Disease usually progresses from enlarged kidneys with elevated GFR to the appearance of microalbuminuria, to macroproteinuria with hypertension, reduced GFR, and renal failure.66 The appearance of microalbuminuria correlates with the presence of coronary artery disease and retinopathy.67 Azotemia generally does not begin until 10 to 15 years after the diagnosis of diabetes. Progression of renal disease is accelerated by hypertension. Meticulous control of diabetes can reverse microalbuminuria and may slow the progression of nephropathy.2,18,59-61 Hypertension should be aggressively managed. Angiotensin-converting enzyme inhibitors are effective in controlling hypertension and lowering microalbuminuria.67,68 Chronic hemodialysis and renal transplantation are
unfortunate endpoints for many diabetic patients with renal disease.
Retinopathy Diabetes is a leading cause of adult blindness in the United States. Approximately 11 to 18% of all diabetic patients have treatable diabetic retinopathy ranging from mild to severe and manifesting in many forms. The severity of diabetic retinopathy is clearly related to the quality of glycemic control.64 Background (simple) retinopathy is found in most diabetic patients who have prolonged disease. Background retinopathy is characterized by microaneurysms, small vessel obstruction, cotton-wool spots or soft exudates (microinfarcts), hard exudates, and macular ischemia.66 The characteristics of proliferative retinopathy are new vessel formation and scarring. Complications of proliferative retinopathy are vitreal hemorrhage and retinal detachment, which may ultimately cause unilateral vision loss. Treatment for diabetic retinopathy is photocoagulation. Maculopathy is background retinopathy with macular involvement. It results primarily in a deficit of central vision. As with proliferative retinopathy, it is vital that the patient be under the care of an ophthalmologist. Laser therapy in the early stages can dramatically alter the course of this disabling disease. The diabetic patient may present with complaints ranging from acute blurring of vision to sudden unilateral or even bilateral blindness. Less often, diabetic patients have more gradual vision loss caused by the common senile cataract or the “snowflake” cataract, which may disappear as hyperglycemia is corrected. The associated hyperlipidemia of diabetes may lighten the color of retinal vessels, producing lipidemia retinalis. Anterior ischemic optic neuropathy has been reported. Diabetic patients with retinopathy should be referred to an ophthalmologist. Even in those with normal vision, ophthalmologic procedures may limit visual loss or prevent crises such as neovascular glaucoma.
Neuropathy Both autonomic and peripheral neuropathies are well-known complications of diabetes. The prevalence of peripheral neuropathy ranges from 15 to 60%.4 The cause of the neuropathy is not clearly understood, but evidence suggests several factors in its development. Neuropathy may result from the effects of diabetic vascular disease on the vasa nervorum. Myoinositol, the polyol pathway, and nonenzymatic glycosylation of protein may have roles. All these factors are related to an elevated blood glucose level. Neurologic manifestations of diabetes may regress with improved glycemic control. Pathologically, segmental demyelinization occurs with loss of both myelinated and unmyelinated axons, particularly those affecting the distal part of the peripheral nerve. Several distinct types of neuropathy have been recognized in diabetes. Peripheral symmetrical neuropathy is a slowly progressive, primary sensory disorder manifesting bilaterally with anesthesia, hyperesthesia, or pain. The pain is often severe and worse at night. It affects upper and lower extremities, although lower extremities and the most distal sections of the involved nerves are most often affected. There may be a motor deficiency as well. Pain is very difficult to control. Simple pain medications, amitriptyline, and fluphenazine are effective for some patients.66 Mononeuropathy, or mononeuropathy multiplex, affects both motor and sensory nerves, generally one nerve at a time.
Infections Diabetic patients are more susceptible to complications of infections because of their inability to limit microbial invasion with effective polymorphonuclear leukocytes and lymphocytes.4 They have an increased incidence of extremity infections and pyelonephritis compared with the general population. In addition, they are particularly susceptible to certain other
1647
infections such as tuberculosis, mucocutaneous candidiasis, intertrigo, mucormycosis, soft tissue infections, nonclostridial gas gangrene, osteomyelitis, and malignant Pseudomonas otitis externa.4 Treatment for diabetic patients with infection includes rapid culture and antibiotics, glycemic control, and generally hospitalization.
Cutaneous Manifestations Dermal hypersensitivity is manifested by pruritic, erythematous indurations that occur at insulin injection sites. The declining prevalence of this condition has paralleled the improved purification of insulin. Insulin lipoatrophy likewise seems to be a result of insulin impurities and is manifested as subcutaneous depressions at injection sites. Although lipoatrophy is now more common than dermal hypersensitivity, its prevalence has also declined sharply because insulin preparations have improved. Insulin lipohypertrophy is manifested by raised areas of subcutaneous fat deposits at insulin injection sites. These lesions generally reflect the failure of the patient to rotate injections sites adequately. They resolve spontaneously over months if insulin injection is avoided in the affected areas and sites are properly rotated. Insulin pumps are often associated with localized skin problems, usually a reaction to the tape securing the tubing and needles. Occasionally, sensitivity to the catheters is seen. Skin infections at the site of injection are the most common complication of insulin pumps. Changing the patient to buffered pure-pork from unbuffered beef-pork insulin is the only intervention that seems to reduce the rate of infection. A few patients have been noted to develop hard nodules at the injection site. The cause of these nodules is uncertain. Diabetic patients who use oral hypoglycemic agents may develop rashes associated with these medications. After consuming ethanol, approximately 38% of type 2 patients taking chlorpropamide exhibit a “flush” consisting of redness of the face and neck and a sense of warmness or burning. Patients may demonstrate urticaria in response to both insulin and oral hypoglycemics. Diabetic skin conditions include fungal infections, acanthosis nigricans, necrobiosis lipoidica diabeticorum, xanthoma diabeticorum, bullosis diabeticorum, and diabetic dermopathy. Acanthosis nigricans is characterized by a velvety brownblack thickening of the keratin layer, most often in the flexor surfaces. It is the cutaneous marker for a group of endocrine disorders with insulin resistance.8 Necrobiosis lipoidica diabeticorum begins as erythematous papular or nodular lesions, usually in the pretibial area, but in other areas as well. The early lesions may contain telangiectasias. These lesions spread and frequently form a single pigmented area of atrophic skin, often with a yellow and sometimes ulcerated center and an erythematous margin. A history of previous trauma is sometimes found. The three forms of diabetic thick skin are (1) sclerodermalike skin changes of the fingers and dorsum of the hand associated with stiff joints and limited mobility, (2) clinically inapparent but measurable thick skin, and (3) “scleroderma adultorum,” or increased dermal thickness on the back and posterior upper neck in middle-aged, overweight patients with type 2 diabetes. Xanthoma diabeticorum is evidence of the hyperlipidemia associated with diabetes. It is similar to the xanthoma found in nondiabetic hyperlipidemic patients. Xanthomas have an erythematous base and a yellowish hue. Bullosis diabeticorum is a rare occurrence. Bullae are usually filled with a clear fluid and are most often found on the extremities, especially the feet. The fluid is occasionally
Chapter 124 / Diabetes Mellitus and Disorders of Glucose Homeostasis
The onset is rapid, with wasting and tenderness of the involved muscles. Clinically, sudden onset of wristdrop, footdrop, or paralysis of cranial nerves III, IV, and VI is noted. Diabetic truncal mononeuropathy occurs rapidly in a radicular distribution. In contrast to other mononeuropathies, it is primarily, if not exclusively, sensory. If it causes pain, it may mimic that of an MI or acute abdominal inflammation. Like diabetic mononeuropathy, it may be most bothersome at night and generally resolves in a few months. Whereas diabetic mononeuropathy is often the first clue of diabetes, truncal mononeuropathy is more often found in known diabetic patients. Autonomic neuropathy occurs in many forms. Neuropathy of the GI tract is manifested by difficulty swallowing, delayed gastric emptying, constipation, or nocturnal diarrhea. Impotence and bladder dysfunction or paralysis may occur. Orthostatic hypotension, syncope, and even cardiac arrest have resulted from autonomic neuropathy. Diabetic diarrhea responds to diphenoxylate and atropine, loperamide, or clonidine. Orthostatic hypotension is treated by sleeping with the head of the bed elevated, avoidance of sudden standing or sitting, and the use of full-length elastic stockings. The Diabetic Foot. Approximately 20% of hospitalizations in diabetic patients are related to foot problems. Sensory neuropathy, ischemia, and infection are the principal contributors to diabetic foot disease. Loss of sensation leads to pressure necrosis from poorly fitting footwear and small wounds going unnoticed. The most common cause of injury is pressure on plantar bony prominences. All neuropathic foot ulcers should be assessed for infection and débrided of devitalized tissue, and radiographs should be obtained to examine for the presence of foreign bodies, soft tissue gas, or bone abnormalities. Weightbearing must be eliminated by total-contact casting.69 Not all ulcers are infected. Infection is suggested by local inflammation or crepitation. Conversely, some uninflamed ulcers are associated with underlying osteomyelitis. Most mild infections are caused by gram-positive cocci, such as Staphylococcus aureus or streptococci, and may be treated with oral antibiotics, a strict non-weight-bearing regimen, meticulous wound care, and daily follow-up. This approach may not be possible when patients are deemed unreliable, do not have good home support, or do not have ready access to follow-up care. Deeper, limb-threatening infections—as evidenced by fullthickness ulceration, cellulitis greater than 2 cm in diameter with or without lymphangitis, bone or joint involvement, or systemic toxicity—are usually polymicrobial in origin and caused by aerobic gram-positive cocci, gram-negative bacilli, and anaerobes. These patients require hospitalization and, after culture, IV empirical antimicrobial therapy with ampicillin-sulbactam, ticarcillin-sulbactam, cefoxitin, imipenem, or a fluoroquinolone and clindamycin; strict non-weight-bearing status; tight glycemic control; early surgical intervention for débridement; drainage; and meticulous wound care.69,70 Occult osteomyelitis should be considered in all cases of neuropathic ulceration. Up to one third of patients must undergo amputation.
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1648
slightly hemorrhagic. The bullae usually heal spontaneously without scarring. Diabetic dermopathy, or “skin spots,” is the most common finding in diabetes. It arises as discrete, depressed, and brownish lesions generally less than 15 mm in diameter and found in the pretibial area. Resistant, aggressive impetigo or intertrigo should suggest diabetes.
evaluate treatment. Initial therapy with sulfonylureas is appropriate; glyburide (2.5–5 mg once daily) or glipizide (5 mg once daily) is recommended. In obese patients or those in whom sulfonylureas are contraindicated, metformin may be an alternative. Follow-up should be stressed and warning signs of hypoglycemia discussed.
Insulin Allergy
The widespread availability of a variety of oral medications for hyperglycemia, some with serious side effects, requires the emergency physician to be familiar with these drugs. Sulfonylureas, developed in the 1940s, continue to be the mainstay of oral diabetes treatment. These drugs increase insulin secretion by binding to specific beta-cell receptors.75 This class of drugs works best in patients with early onset of type 2 diabetes and fasting glucose levels less than 300 mg/dL.76 This class of drugs is contraindicated in patients with a known allergy to sulfa agents. Patients with renal failure may be predisposed to hypoglycemia. Metformin is an agent in the biguanide class. It works by decreasing hepatic glucose output and increasing peripheral uptake of glucose, leading to decreased insulin resistance and lower blood glucose. Used alone, metformin does not cause hypoglycemia, but it is contraindicated in patients with renal insufficiency and metabolic acidosis. Metformin is renally excreted. Metformin should be withheld for 48 hours before or after administration of iodinated contrast media because of the risk of acidosis from transient decrease in renal function. Metformin must be used with caution in patients with hypoxemia, liver compromise, and alcohol abuse. These patients are at increased risk for developing lactic acidosis, which has a 50% mortality rate.76,77 The thiazolidinediones reduce insulin resistance and are especially useful in patients who require large amounts of insulin and still lack adequate glucose control. Because of hepatotoxicity, troglitazone has been removed from the market.77 Pioglitazone and rosiglitazone are approved for monotherapy. Liver function should be monitored for at least 1 year after the initiation of therapy with thiazolidinediones. α-Glucosidase inhibitors delay intestinal monosaccharide absorption and prevent complex carbohydrate breakdown.76 They must be titrated to minimize GI side effects and should not be used in patients with certain GI disorders. Liver function monitoring is required because of dose-dependent hepatotoxicity. Repaglinide is similar to the sulfonylureas in action and mechanism. It has a more rapid onset of action, involves less risk of hypoglycemia, and is suitable for patients allergic to sulfa.77 Care should be used in patients with renal or hepatic dysfunction.76 Exanatide is a GLP (glucagon-like peptide) agonist. GLP stimulates release of insulin from pancreatic cells. GLP itself has a half-life of only a few minutes, but the GLP agonists bind to the GLP receptor on the pancreas and have a much longer half-life. Vildagliptin and sitagliptin are classified as DPP-4 (dipeptidyl peptidase-4) inhibitors. DPP-4 degrades endogenous GLP; the DPP-4 inhibitors, by preventing this degradation, prolong the half-life of GLP and increases insulin secretion.78
Insulin allergy is mediated by immunoglobulin E and is manifested by local itching or pain and delayed brawny edema, urticaria, or anaphylaxis. Systemic reactions are usually seen in patients who have previously discontinued insulin and then resumed therapy. Mild reactions may be treated with antihistamines, whereas anaphylaxis must be treated with epinephrine. Patients with significant reactions must be admitted for desensitization.4
■ DIABETES IN PREGNANCY Before the discovery of insulin in 1922, diabetes in pregnancy was associated with a fetal death rate of 60 to 72% and maternal morbidity of approximately 30%. In 1977, a linear relationship between glycemic control and perinatal mortality was discovered. Strict metabolic control is now a goal in all diabetic pregnancies.41,57 Pregnant patients should be watched extremely closely and aggressively treated for impending or actual DKA. For a variety of reasons, pregnant women have a special predisposition to both glucose intolerance and excess ketone production. Although uncommon, DKA may cause perinatal asphyxia and reduce fetal oxygen delivery.41,57 Intellectual deficits in offspring have been associated with maternal ketonuria from any cause. Pregnancy is associated with progression of retinopathy for unknown reasons.71 Whether pregnancy worsens diabetic nephropathy or hastens the progression to end-stage renal disease is controversial.72 Although nephrotic syndrome develops in 71% of pregnancies, blood pressure and proteinuria eventually return to first-trimester values. Diabetic nephropathy is associated with an increased risk of preterm labor, stillbirth, neonatal death, fetal distress, and intrauterine growth retardation; otherwise, the literature is sparse on the effect of pregnancy on diabetic neuropathy. Autonomic neuropathy, particularly gastroparesis, makes adequate nutrition difficult for both mother and fetus. Pregnant women should be referred for parenteral feedings if conservative therapy fails to control vomiting.41 Hypoglycemia is common in pregnancy in part because of intensive insulin treatment to maintain euglycemia.73 Hypoglycemic unawareness is not uncommon. The effects of hypoglycemia on the fetus are unclear. Ketoacidosis is associated with a 50 to 90% fetal mortality rate.41,74
■ NEW-ONSET HYPERGLYCEMIA Patients often present to the ED with typical diabetic symptoms such as polyuria, polydipsia, and polyphagia. Many have serum glucose greater than 200 mg/dL but are not ketotic. These patients with normal electrolytes may be treated with IV hydration alone or with insulin, often reducing the glucose to 150 mg/dL. In reliable patients whose initial glucose is greater than 400 mg/dL, initiation of oral hypoglycemic therapy may be appropriate, with lifestyle modification. An HbA1c value should be obtained before initiation of therapy to help
■ ORAL HYPOGLYCEMIC AGENTS
■ NEW TRENDS Changes in the therapy of diabetes include greater use of human insulin, which has prevented some of the adverse reactions to beef and pork products. Unfortunately, some patients demonstrate sensitivity reactions even to subcutaneously
1649
glycemic control. The mostly beneficial but occasionally deleterious effects of exercise have also been elaborated.75 Newer therapies include pancreatic and pancreatic beta-cell transplants. Solid-organ pancreatic transplantation remains controversial among both diabetologists and transplant surgeons. Transplantation ameliorates many secondary complications of diabetes, such as nephropathy, neuropathy, gastroparesis, retinopathy, and microvascular changes. The percentage of grafts functioning after 1 year and the 1-year survival rate of patients are greater than 75% in selected medical centers. Rejection, post-transplantation pancreatitis, and graft thrombosis, as well as other vascular and immunosuppression problems, continue to plague transplant recipients. Pramlintide is part of a new class of drugs called amylin analogues. These decrease gastric emptying and decrease glucagon secretion. Currently, pramlintide must be administered by subcutaneous injection after a meal.78 Other new areas for research have included agents that increase urinary excretion of glucose or increase hepatic gluconeogenesis.
KEY CONCEPTS ■
Hypoglycemia may be associated with significant morbidity and mortality. When the diagnosis is suggested and, if possible, confirmed by laboratory evaluation, treatment should be initiated immediately. ■ Hypoglycemia caused by oral hypoglycemic agents may be prolonged. Patients should be observed for an extended period or hospitalized. ■ The essential treatment of DKA includes restoration of insulin, correction of dehydration, correction of potassium level, correction of acidosis, and treatment of the underlying cause. ■ Hyperglycemic hyperosmolar nonketotic coma is often associated with focal neurologic signs that resolve with treatment. The essentials of treatment are correction of profound dehydration, correction of electrolytes, and treatment of the underlying cause. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 124 / Diabetes Mellitus and Disorders of Glucose Homeostasis
injected human insulin.79 More physicians are teaching their type 1 patients and families how to administer glucagon to treat severe hypoglycemia. Initiation of immunosuppressive therapy at the initial diagnosis of type 1 diabetes can prolong the patient’s ability to secrete insulin. However, this beneficial effect, whether achieved by azathioprine or cyclosporine, is not usually sustainable.80 The potential side effects of immunosuppressive agents have precluded large trials in patients early in their disease.81 Prophylactic insulin therapy, nicotinamide, oral insulin, or glutamate decarboxylase and avoidance of cow’s milk may prevent or delay the onset of type 1 diabetes in patients at risk.5 Glycemic control now involves improved technology and more widespread individual monitoring. More patients alter their insulin dosages daily in response to their findings. Diabetic patients with tight glycemic control benefit by limiting the progression of microvascular disease: neuropathy, renal disease, and certain types of retinopathy. However, they are more likely than other diabetic patients to experience hypoglycemic episodes. Emergency physicians and out-of-hospital care providers are encountering patients with insulin pumps. Many insulin pumps are available, each having a pump mechanism, a reservoir for insulin, tubing, and indwelling subcutaneous needles. They are attached, usually with tapes, to the patient’s body and administer insulin at a regular adjustable rate. Most pumps also allow the patient to administer additional boluses of insulin as necessary. These pumps support tight glycemic control and are acceptable to some patients. However, motivated patients can achieve equivalent control by adjusting daily injections. Insulin pumps are associated with a variety of complications (e.g., iatrogenic hypoglycemia). Because glucose rotates the polarization of light waves, new fiberoptic technology has been developed to determine blood glucose noninvasively. This technology may be applied to the insulin pumps in the future. Inhaled insulins have been studied, and one was even marketed briefly; further research into this area is likely. A variety of available insulins, including ultra-short-acting (Humalog) and long acting (Lantus) have come out in recent years, which have provided a wider array of possibilities in diabetes management. The basic concepts of the diabetic diet remain unchanged, although many studies emphasize foods and medications that alter glucose absorption. Various high-fiber diets have improved
Chapter 125
Rhabdomyolysis
Laura J. Bontempo and Amy H. Kaji
■ PERSPECTIVE Rhabdomyolysis, defined as the dissolution or disintegration of striated muscle,1 describes a clinical and biochemical syndrome resulting from the release of intracellular contents into the extracellular fluid and circulation. The diagnosis of rhabdomyolysis rests on measurement of these released substances in either plasma or urine. The classic presentation includes symptoms of myalgias, weakness, red to brown urine due to myoglobinuria, and elevated serum muscle enzymes, such as creatine kinase (CK).2 Approximately 26,000 incident cases of rhabdomyolysis are reported annually.3 The spectrum of disease severity ranges from an asymptomatic elevation of muscle enzymes to life-threatening electrolyte imbalances, acute renal failure, multiorgan failure syndrome, and death. Historically, the earliest reference to rhabdomyolysis occurs in the Old Testament, from the Book of Numbers.4 During the Exodus, the Israelites consumed large amounts of quail, which fed on hemlock seeds, and many became ill and died from an illness described by intense muscle pain and weakness. In the late 19th century a clinical syndrome of muscle pain, weakness, and brown urine was called “Meyer-Betz disease” in the German literature.5 In 1941, Bywaters and Beall6 described the clinical course of four victims with crush injuries to the limbs after air raids during World War II. They noted the link between muscle injury and renal dysfunction in their classic monograph, as follows: “The patient has been buried for several hours with pressure on a limb. On admission he looks in good condition except for swelling of the limb, some local anesthesia and whealing. The hemoglobin, however, is raised and a few hours later despite vasoconstriction made manifest by pallor, coldness and sweating, the blood pressure falls. This is restored to the preshock level by (often multiple) transfusions of serum, plasma, or occasionally, blood. Anxiety may now arise concerning the circulation in the injured limb, which may show diminution of arterial pulsation distally, accompanied by all the changes of incipient gangrene. Signs of renal damage soon appear and progress, even though the crushed limb be amputated. The urinary output, initially small, owing perhaps to the severity of the shock, diminishes further. The urine contains albumin and many dark brown or black granular casts, which later decrease in number. The patient is alternately drowsy and anxiously aware of the severity of his illness. Slight generalized edema, thirst, and incessant vomiting develop, and the blood pressure 1650
often remains slightly raised. The blood urea and potassium, raised at an early stage, become progressively higher, and death occurs comparatively suddenly, frequently within a week. Necropsy reveals necrosis of muscle and in the renal tubules, degenerative changes and casts containing brown pigment.” Acute renal failure (ARF) is one of the most serious complications of rhabdomyolysis, and the presence of ARF is associated with multisystem organ failure and a higher mortality rate.7 Well-designed prospective studies of rhabdomyolysis and its complications are lacking, so the true incidence of ARF in this setting is unknown but is estimated at 4 to 33%.8 Approximately 5 to 15% of patients hospitalized with ARF in the United States have rhabdomyolysis as the cause.9
■ PRINCIPLES OF DISEASE Anatomy and Physiology Skeletal muscle is the largest organ in the human body. The functioning of muscle cells is critically dependent on a healthy cell membrane, the sarcolemma, which maintains ionic gradients and ensures proper metabolic functioning. The sarcolemma contains sodium-potassium pumps, calcium proteincarrier pumps, and other channels and structures.10 The sodium-potassium pump moves sodium out and potassium into the sarcoplasm. More sodium is transported out than potassium is transported in, creating a net negative intracellular charge. A sodium concentration gradient is thus created. Normally, the concentration of intracellular sodium ions is very low, approximately 10 mEq/L, when compared with the extracellular fluid.10 Located in the sarcolemma, the calcium pumps work to maintain a low intracellular fluid calcium concentration by moving calcium out of the sarcoplasm to the extracellular space. Additional pumps help move calcium into the muscle cell’s internal structures, the sarcoplasmic reticulum and the mitochondria. As sodium ions move down the electrochemical gradient (i.e., return into the cell), calcium ions are able to move out of the cell into the extracellular fluid (Fig. 125-1). With the exception of the sodium-calcium exchange, all of these pumps depend on active transport, with adenosine triphosphate (ATP) as the energy source.10 Also found within a cell’s sarcoplasm, myoglobin is the major heme protein supplying oxygen to skeletal and cardiac
1651 External medium K+
Sarcoplasm
K+ 1 K+ K+
ATP + K 2+
Ca2+ 3 Na+
Na+1
Ca1 K+
2 ATP
K+ 3
Ca2+ SR K+ Sarcoplasmic reticulum
0
Diagnostic Parameters in Acute Renal Failure and Acute Intrinsic Renal Failure
Odorless urine Specific gravity 4 hours), presence of diabetes, and body mass index (BMI) greater than 40.36
Exertion Rhabdomyolysis can result from prolonged or strenuous exercise and is seen in both trained and untrained athletes.37-39 Eccentric exercise (work done by a muscle during lengthening) is more damaging to muscle fibers, as evidenced by higher CK levels, than concentric exercise (work done by a muscle during shortening).14 Hot conditions contribute to the incidence of exertional rhabdomyolysis because of increased dehydration and increased activity of heat-sensitive degradative enzymes.39,40 Hypokalemia increases the risk of exertional rhabdomyolysis, since hypokalemia limits vasodilation and perfusion of the muscle microvasculature.41 With prolonged exercise, the sarcolemmic ion pumps may also fail because of a depletion of cellular energy sources, specifically ATP.11 The failure of these pumps leads to elevated intracellular calcium and subsequently rhabdomyolysis, which, coupled with dehydration and acidosis from lactic acid production, can cause ARF.13,39,42 Exertional rhabdomyolysis is not always the result of voluntary muscle exertion, since the same pathophysiology is seen in patients with status epilepticus, myoclonus, dystonia, chorea, tetanus, psychotic agitation, and mania.11,43-45
Electrical Current Rhabdomyolysis occurs in approximately 10% of patients who initially survive a high-voltage electrical injury or lightning strike.18 Note that the severity of rhabdomyolysis is not related to the size of the wound or the site of entry.18 Rhabdomyolysis from electrical current appears to be a result of both the heat generated by the electrical current and the direct effects of the current on the sarcolemma (electroporation).46
Heat and Cold Injury Multiple disorders can raise the core body temperature and result in sarcolemma disruption. Neuroleptic malignant syndrome (fever in patients treated with phenothiazines or haloperidol), malignant hyperthermia (rapid rise in body temperature after anesthesia with halogenated hydrocarbons or succinylcholine), and both classic and exertional heatstroke47 are some of the most common causes.19,48,49 In hyperpyrexic syndromes, cellular energy demands outstrip available energy supplies, causing membrane dysfunction and cellular injury.50 Hypothermia may also cause rhabdomyolysis, most likely through cold-induced ischemia and direct injury to components of the sarcolemma, which cannot maintain structural integrity below certain temperature levels.51
Drugs and Toxins Drugs in almost every class of medication have been implicated as a cause of rhabdomyolysis.11 Common offenders include ethanol, cocaine and other licit and illicit drugs, lipidlowering agents, carbon monoxide, and biologic toxins. Ethanol. Ethanol is directly toxic to the skeletal muscle cell membrane, and this toxicity appears to be potentiated by starvation.13 For this reason, ethanol-induced rhabdomyolysis is often seen in patients who are “binge drinkers.” Electrolyte abnormalities also play a role, since chronic alcohol abusers often have hypokalemia, hypophosphatemia, and hypomagnesemia.19 These deficiencies, coupled with ethanol’s direct sarcolemmic toxic effects, make the ethanol abuser more susceptible to rhabdomyolysis.13 Ethanol is also a sedative-hypnotic, which can induce obtundation and lead to immobilization of a body part with
Infections Bacterial, viral, and parasitic infections have been associated with rhabdomyolysis.70 Many viruses have been implicated, including influenza, coxsackievirus, parainfluenza, adenovirus, herpes simplex, Epstein-Barr, cytomegalovirus, and human immunodeficiency virus (HIV).71 Patients classically report a history of a viral illness 1 to 2 weeks prior to the onset of myalgias and myoglobinuria. Influenza viruses A and B72 are the most frequently cited viral causes.73,74 Although viruses may be directly toxic to myocytes, this has not been proven.13,75 Whereas rhabdomyolysis has been reported in many patients with HIV, the independent role of this virus is unclear because
1653
many of the patients in these studies were taking multiple medications or had concurrent infections.76,77 Bacterial infections may cause muscle damage through multiple mechanisms, including direct muscle infection in pyomyositis, as well as the release of exotoxins and cytokines, and the induction of fevers and rigors. Legionella is the most common known bacterial cause of rhabdomyolysis. Its myotoxic effects are mediated through an endotoxin.13,78,79 Salmonella and Streptococcus also can induce rhabdomyolysis through both direct myocyte invasion and inhibition of glycolytic enzymes.13,78,80 Of parasitic infections, falciparum malaria (Plasmodium falciparum) is the most notorious for causing rhabdomyolysis, as patients have high fevers, rigors, vomiting, and ARF.81
Electrolyte Abnormalities A variety of electrolyte disorders, particularly hypophosphatemia and hypokalemia, have been associated with rhabdomyolysis. Hypophosphatemia is believed to cause membrane injury by severe depletion of ATP, and most cases have been described in alcoholic patients82 or those receiving treatment for diabetic ketoacidosis.83 Since potassium is a vasodilator of the microcirculation for metabolically active muscle cells, hypokalemia may prevent local vasodilation and lead to focal muscle ischemia.9,13 Hypocalcemia has also been associated with rhabdomyolysis in the setting of hypoparathyroidism.84 Both hyponatremia and hypernatremia have also been associated with rhabdomyolysis,11 with case reports of the former primarily involving hyponatremia induced by psychogenic polydipsia.85
Hypoxia and Ischemia Intrinsic vascular injury or obstruction, hypotension, and external compression of the blood supply to a muscle group may all cause tissue hypoxia and rhabdomyolysis. For example, it may follow orthopedic or vascular reconstruction procedures in which a tourniquet is used. When circulation is reestablished, the damaged cell is reperfused and the extruded intracellular contents, including myoglobin, are brought into the general circulation.13,86 Reperfusion is associated with an influx of neutrophils, which release proteolytic enzymes that can directly occlude the microcirculation and cause further muscle ischemia.11,13 Certain blood disorders (e.g., sickle cell anemia)87 may cause vascular thrombosis, resulting in tissue hypoxia and subsequent muscle injury, as well.17,88
Miscellaneous Causes Endocrine disorders, such as diabetic ketoacidosis and nonketotic hyperglycemia, have been associated with rhabdomyolysis, possibly due to hypophosphatemia and hypokalemia and the hyperosmolar state.89 Rhabdomyolysis has also been infrequently described in association with hypothyroidism, hyperthyroidism,90 and pheochromocytoma.91 Case reports have also been documented in the literature of rhabdomyolysis occurring in patients with inflammatory myopathies, such as dermatomyositis and polymyositis.92
■ CLINICAL FEATURES Patients with rhabdomyolysis classically present with the complaints of muscle weakness, pain, and tea-colored urine. The myalgias may be focal or diffuse, depending on the underlying cause of the disease. However, a high clinical suggestion for rhabdomyolysis must be maintained in patients at risk, because
Chapter 125 / Rhabdomyolysis
external compression of its blood supply. In addition, excessive motor activity from seizures or delirium tremens can induce rhabdomyolysis. Cocaine. The incidence of rhabdomyolysis in patients who use cocaine varies from 5 to 30% in published reports.48 Cocaine may produce rhabdomyolysis by several mechanisms; hypotheses include cocaine-induced vasospasm with resultant muscle ischemia, excessive energy demands placed on the sarcolemma, and direct toxic effects on myocytes.52-54 Seizures, agitation, trauma, and hyperpyrexia may also play a role.11 In general, the severity of rhabdomyolysis parallels the severity of intoxication.52,55 Intravenous cocaine may be associated with a higher incidence of rhabdomyolysis-induced ARF than for smoking cocaine.52 Other Illicit Drugs. Agents such as d-lysergic acid diethylamide (LSD); phencyclidine hydrochloride (PCP); sympathomimetics, such as amphetamines; and “ecstasy” (MDMA; 3,4methylenedioxymethamphetamine)56 may also cause rhabdomyolysis.11,57 Delirium and agitation, resulting in involuntary and voluntary muscle contraction, may raise energy demands of muscle cells. The increased demand may outstrip the supply of normally available ATP.58 Lipid-Lowering Agents and Other Licit Drugs. The 3-hydroxy-3methylglutaryl coenzyme A (HMG-CoA) reductase inhibitor lipid-lowering agents (e.g., lovastatin, simvastatin) have been associated with rhabdomyolysis,59 as have the branched-chain fatty esters that inhibit liver triglyceride synthesis (e.g., clofibrate, gemfibrozil).11,60,61 The mechanism of action is unclear, although statins are considered to be direct myotoxins that may interfere with ATP production through the electron transport chain. Statins may cause rhabdomyolysis when used alone or with other drugs, particularly gemfibrozil.60,62,63 Patients with preexisting renal dysfunction, hypothyroidism, and inflammatory myopathies may be at greater risk.64 Other direct myotoxins include colchicine and immunosuppressants such as cyclosporine. Overuse of diuretics or cathartics, leading to severe hypokalemia and dehydration, may also result in rhabdomyolysis. Carbon Monoxide. Rhabdomyolysis is a known complication of carbon monoxide poisoning.65,66 The pathophysiology is unknown, but hypoxia, muscle compression from coma, and direct myocyte toxic effects may play a role.65 Biologic Toxins. Some snake envenomations cause rhabdomyolysis through direct myocyte injury resulting in the release of intracellular contents to the extracellular circulation. Species known to do this include the European adder, Australian tiger snake, Australian king brown snake, sea snakes, North and South American rattlesnakes, and the death adder. Multiple myocyte toxins may be present in a single venom.11 Stings from Africanized bees (“killer bees”) and honeybees can also cause rhabdomyolysis. This is also mediated through direct myotoxins.11,67,68 Mushroom poisoning has also been associated with rhabdomyolysis.69
up to 50% of those with serologically proven rhabdomyolysis do not complain of myalgias or muscle weakness.93 In fact, in the United States, rhabdomyolysis is most often due to prolonged muscle compression in the intoxicated patient who lays motionless or in the elderly patient with dementia following a fall.93
History The history can be extremely helpful in making the diagnosis. The history should include any recent trauma or compression, excessive exertion, envenomations, infections, electrical shock, or temperature extremes. Other areas of interest are the use of prescription medications, over-the-counter drugs, alcohol or illicit drugs, known medical conditions, and a family history of muscle dysfunction or disease. Unfortunately, many patients with rhabdomyolysis are unable to provide an adequate history because of an altered sensorium.
Physical Examination Physical examination may reveal weakness with tenderness to palpation of the affected muscle groups. Skin may be discolored, and there may be evidence of pressure necrosis. With trauma or compression, the affected area may have sensory and motor deficits that do not follow a single nerve distribution.16 Some patients with severe rhabdomyolysis show bradypnea, presumably caused by diaphragmatic muscle involvement. Patients may also appear clinically dehydrated from the reduced extracellular fluid volume. Compartment syndrome is a relatively common complication of rhabdomyolysis, and thus, the exam may reveal firm muscle compartments, pain with passive extension, or neurovascular compromise of the affected extremity. Characteristic physical signs are present in only 4 to 15% of patients, however.93 Therefore the absence of these exam findings does not rule out the diagnosis.
■ DIAGNOSTIC STRATEGIES Myoglobin In the past, the diagnosis of rhabdomyolysis rested on the demonstration of myoglobin in the serum. Serum myoglobin, however, is an insensitive marker for rhabdomyolysis. The half-life of myoglobin in plasma is 1 to 3 hours and can be cleared completely from plasma within 6 hours after injury.93 Similarly, urine myoglobin, which is rapidly excreted, may also be an inaccurate measure. Myoglobinuria may be absent in patients who present late in the course of their illness, and the amount of myoglobinuria depends on the plasma concentration of myoglobin, renal function, the glomerular filtration rate, the extent of myoglobin binding in plasma, and urine flow rate.93 Methods used to measure urine myoglobin include immunodiffusion, radioimmunoassay, and specific dipstick tests. The dipstick tests involve use of reagents (e.g., guaiac, or o-toluidine) and are only slightly less sensitive than radioimmunoassay, which is the best method available.94 Urinalysis typically shows brown urine with a large amount of blood on dipstick evaluation but few, if any, red blood cells (RBCs) on microscopic evaluation. This occurs because most dipstick tests cannot distinguish myoglobinuria from hematuria or hemoglobinuria.94 Protein, brown casts, and renal tubular epithelial cells may also be found.11
Creatine Kinase Measurement of CK levels (formerly CPK, creatine/creatinine phosphokinase) is a more sensitive method than myoglobin
10 9 Units of measurement
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1654
8 7 6 5 4 3 2 1 0 24
48
72
96
120
Hours from injury
Figure 125-2. Typical creatine kinase elimination curve. testing. CK is an excellent marker for rhabdomyolysis because it is easily measured, is present in the serum immediately after muscle injury, and is not rapidly cleared from serum (halflife is 1.5 days).95 In general, peak CK levels occur within 24 to 36 hours of muscle injury and diminish approximately 39% per day (Fig. 125-2).49 Failure of levels to decrease in this manner suggests ongoing muscle injury, possibly from an undetected compartment syndrome. Rhabdomyolysis is not defined by a specific CK level. In general, however, in the absence of cerebral or myocardial infarction, a CK above 5000 indicates serious muscle injury, and a CK level greater than five times normal is thought to be diagnostic, although levels as high as several hundred thousand have been reported.14,42,53,76 Higher CK levels, especially levels greater than 16,000, are correlated with the development of ARF.96 However, since patients may have significant morbidity with only moderately elevated CK levels,53,96,97 even modest elevations of CK must be taken seriously. The CK subtype present in skeletal muscle is MM, but when considerable skeletal muscle injury occurs, a small amount of CK-MB is also released. The CK-MB fraction rarely exceeds 3 to 5% and is indicative of the CK-MB released from skeletal muscle rather than from concomitant myocardial damage.
Other Tests Other laboratory evaluation may show hyperkalemia, hyperphosphatemia, hypocalcemia, hyperuricemia, and hypoalbuminemia. An elevated anion gap is characteristically present.98 Hyperkalemia (>5.5 mEq/L) has been reported on initial laboratory studies in 20 to 40% of patients.93 It is caused by a combination of intracellular potassium release from muscle necrosis and decreased renal excretion. Hyperphosphatemia results from a leakage of phosphorus from injured muscle. Levels usually do not exceed 7 mg/dL, but a normal serum phosphate level in the setting of significant rhabdomyolysis raises the possibility that hypophosphatemia was the underlying cause of the rhabdomyolysis. The most common electrolyte abnormality—hypocalcemia—occurs early, and it can be exacerbated by hyperphosphatemia. Hypocalcemia results from the deposition of calcium into damaged muscle and decreased bone responsiveness to parathyroid hormone.99 In one series of 76 patients, hypocalcemia was present in 63%.100 Hypercalcemia often develops later. Although the exact cause of hypercalcemia is unknown, it is hypothesized that calcium is mobilized from damaged muscle, and parathyroid hormone
■ DIFFERENTIAL CONSIDERATIONS Pigmenturia has a variety of causes (Box 125-3). Hematuria can be distinguished from myoglobinuria through microscopic identification of RBCs in the urine with hematuria. Note that hematuria can be present with rhabdomyolysis if there is concomitant renal trauma. Similar to myoglobinura, hemoglobinuria demonstrates a positive dipstick test for blood but no, or few, RBCs on microscopic analysis. With hemoglobinuria, however, the plasma appears discolored as brown or red.19 Pigmenturia can be associated with acute intermittent porphyria, as well, although these patients generally have a very different clinical presentation, and their urine contains porphobilinogen.11 Bilirubin, a degradation product of heme, also causes pigmenturia when present in the urine. In this case, the urine tests positive for urobilinogen. Pigmenturia can also be a direct effect from certain drugs or foods, though the urine should test negative for blood, and the microscopic evaluation should demonstrate no RBCs. In crush injury, the motor weakness and possible paralysis may mimic spinal injury.102 All trauma patients must be treated with spinal precautions, and laboratory-proven rhabdomyolysis does not rule out concurrent spinal injury. With rhabdomyolysis, however, motor function often improves as the disease is treated.
BOX 125-3 Differential Diagnosis of Pigmenturia Hemoglobinuria Hemolysis Hematuria Renal causes Trauma Acute Intermittent Porphyria Bilirubinuria Food Beets Drugs Vitamin B12 Rifampin Phenytoin Laxatives
1655
Myocardial infarction must be considered in patients with an elevated CK level and pain, especially if the pain is localized to the chest. Measuring serum troponin, obtaining an electrocardiogram, and eliciting a history of ischemic chest pain help discern a cardiac cause of the CK elevation.
■ COMPLICATIONS Complications of rhabdomyolysis may be categorized as early, if they occur within the first 24 hours after injury. Such early complications of rhabdomyolysis include electrolyte abnormalities, such as hyperkalemia, hypocalcemia or hypercalcemia, hyperphosphatemia, and hyperuricemia. The hyperkalemia may lead to cardiac arrhythmia and death. Hepatic dysfunction, another early complication, manifests as elevation in liver enzymes and occurs in 25% of patients. Proteases released from injured muscle are implicated in the hepatic inflammation.101 Renal failure and disseminated intravascular coagulopathy are later complications, more commonly developing after 24 to 48 hours. In contrast, compartment syndrome, usually developing in muscles whose expansion is restricted by tight fascia, such as the tibialis anterior, may be an early or late complication of fluid resuscitation with worsening edema.
■ MANAGEMENT After initial stabilization and resuscitation, the primary objective in managing rhabdomyolysis should be to identify and treat the underlying cause and to mitigate the associated complications: electrolyte derangements, renal failure, coagulopathy, and compartment syndrome.
Saline Infusion The mainstay of therapy for rhabdomyolysis is the administration of large volumes of saline very early in the course of the disease. Fluid is sequestered in necrotic muscle and contributes to intravascular hypovolemia and prerenal renal failure. In patients with trauma or compression, saline resuscitation should begin in the field.103 Delays in initiating rehydration increase the risk of oliguric and anuric renal failure.104 In one study, no patients who underwent aggressive saline rehydration within the first 6 hours of admission developed ARF.13,105 Initial resuscitation should be undertaken with normal saline. Potassium-containing fluids should be avoided due to the risk of rhabdomyolysis-associated hyperkalemia. Highvolume infusions should be started as soon as possible and infusion rates titrated for a urine output of 200 to 300 mL/hr.18 Patients may require up to 20 L of fluid in the first 24 hours to achieve adequate urine flow rates. Ideally, the amount of fluid to be administered should be determined on the basis of the clinical course or central venous pressure measurements.106
Mannitol Mannitol use is somewhat controversial in the treatment of rhabdomyolysis, since its use is mostly supported by animal studies and retrospective clinical studies.107,108 In one study, mannitol did not confer any benefit compared with normal saline alone.109 Mannitol is an osmotic diuretic, an intravascular volume expander, a renal vasodilator, and possibly a free radical scavenger. As a diuretic, mannitol increases urine flow, which may help prevent obstruction from myoglobin casts. Renal vasodilation increases renal blood flow and GFR, and
Chapter 125 / Rhabdomyolysis
and 1,25-dihydroxycholecalciferol levels are increased during the recovery period.13 The combination of elevated serum phosphate and calcium may result in precipitation of calcium phosphate in soft tissue, blood vessels, and eyes.11 Hyperuricemia results from the release of purines from damaged muscles. Hyperuricemia is more likely to occur in well-trained athletes with exertional rhabdomyolysis due to their increased muscle mass. Hypoalbuminemia may result from leakage of protein from injured vessels coupled with proteinuria. Many patients with acute rhabdomyolysis demonstrate evidence of disseminated intravascular coagulopathy as a late complication. Thrombocytopenia, hypofibrinogenemia, and an elevated D-dimer with prolongation of prothrombin time may be seen. The coagulopathy is a result of muscle necrosis and liberation of activating substances (e.g., thromboplastin) from injured cells. Some patients may have elevated levels of aspartate transaminase, alanine transaminase, and lactate dehydrogenase due primarily to muscle necrosis and less to hepatic injury from proteases released from injured muscle.101
may also decrease tubular obstruction. As a volume expander, mannitol draws fluid from the interstitial space, decreasing intravascular dehydration and potentially reducing muscular swelling.11,19,104 In cases of early ARF, mannitol may convert oliguric renal failure to nonoliguric renal failure, which has a somewhat better prognosis. Since mannitol is a diuretic, however, adequate volume resuscitation and urine flow should be established prior to its administration. Loop diuretics (e.g., furosemide) can acidify the urine and should not be used.18
Urine Alkalinization Myoglobin precipitation is enhanced in acidic conditions, and myoglobin alone may not be nephrotoxic unless accompanied by intravascular volume depletion and acidosis.20 Thus, urine alkalinization theoretically facilitates renal myoglobin clearance by increasing its solubility. The goal is to keep the urine pH greater than 6.5, which can be accomplished by adding bicarbonate to intravenous fluids. Two ampules of bicarbonate in 1 L of half normal saline produce a slightly hypertonic solution. Thus, it may be prudent to add 1.5 ampules of bicarbonate to 1 L of half normal saline or 2 ampules of bicarbonate in 5% dextrose in water, especially when mannitol, a hyperosmolar agent, is also administered. Potential adverse effects of sodium bicarbonate therapy include hypernatremia, aggravation of fluid overload in patients with congestive heart failure, and the exacerbation of hypocalcemia. Similar to mannitol administration, alkalinization as a therapeutic modality for rhabdomyolysis is being questioned, since no randomized studies have demonstrated a benefit. In fact, Hosmi and colleagues109 found no benefit of bicarbonate therapy over vigorous fluid hydration. More recently, Brown and associates reviewed the case records of 1771 critically ill trauma patients with increased CK levels and demonstrated no difference in the incidence of renal failure (22% vs. 18%), dialysis (7% vs. 6%), or mortality (15% vs. 18%) between the group that received mannitol and bicarbonate versus the group that only received saline hydration.110 Thus, further studies are needed to demonstrate if any treatment regimen is truly superior to early, vigorous saline hydration.111
given to a patient with hyperphosphatemia, since the calcium and phosphate may precipitate.11 Moreover, correction of initial hypocalcemia with intravenous calcium can exacerbate the normal hypercalcemia that occurs in the recovery phase of rhabdomyolysis when calcium deposited in the injured muscle cells is mobilized back into the extracellular space. Thus, the use of calcium for asymptomatic hypocalcemic patients should be avoided because it may raise intracellular calcium levels, promoting further muscle injury.84 In addition to administering potassium resin binders, as well as insulin, glucose, and bicarbonate, dialysis may be required to treat hyperkalemia. Symptomatic hypercalcemia generally requires only volume expansion and diuretic therapy. For patients with a rising or elevated potassium level, persistent acidosis, uremia, or oliguric or anuric renal failure with fluid overload, dialysis may be necessary.11,13 Dialysis with supportive care should effectively limit the morbidity and mortality from ARF associated with rhabdomyolysis.
Coagulopathy/Disseminated Intravascular Coagulopathy Therapy for coagulopathy is directed at treatment of the underlying disease process. However, if hemorrhagic complications occur, therapy with platelets, vitamin K, and fresh frozen plasma may be necessary.
Compartment Syndrome Clinicians should monitor compartment pressures in the patient with suggested or existing compartment syndrome. When compartmental pressure exceeds 30 to 35 mm Hg, fasciotomy should be strongly considered, although the decision to perform a fasciotomy must be decided on a case-by-case basis.115 The failure of CK levels to decline appropriately suggests ongoing muscle injury from a compartment syndrome (Fig. 125-3).
■ DISPOSITION No good prospective studies support a standardized approach to disposition of the patient with rhabdomyolysis. The high
Experimental Therapies A potential role for iron chelators, such as desferrioxamine, is under investigation. Chelation therapy reduces renal injury in animal models.112 The theory is that iron chelation diminishes exposure to free iron, thereby decreasing lipid peroxidation and myocyte breakdown. One case report demonstrated that hyperbaric oxygen was beneficial when used as adjunctive therapy for a patient with compartment syndrome, rhabdomyolysis, and ARF after a heroin overdose.113 Other experimental models demonstrate a reduction in muscle necrosis due to ischemia-reperfusion injury with the administration of free-radical scavengers and antioxidants, such as vitamin E and vitamin C, and minerals, such as zinc, manganese, and selenium. These, therefore, are potential therapeutic agents in the management of rhabdomyolysis.114
General Measures Electrolyte Abnormalities Hyperkalemia is a potentially life-threatening complication of rhabdomyolysis and must be treated. However, intravenous calcium may be ineffective as a treatment for hyperkalemia if
40,000 Decreased pulses
30,000 CP lv/L
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1656
Paresthesias 20,000 “Normal” CP T/2 = 48 hr
10,000 0 0
12
24
36
48
60
72
Hours Rhabdomyolysis
Fasciotomy
Figure 125-3. Rhabdomyolysis: second-wave phenomenon. Serum
creatine kinase (CK) activity resulting from a single bout of muscle injury usually peaks at about 24 hours. Its half-life (T/2) is about 48 hours. A second rise may occur if necrosis has involved a muscle in a tight fascial compartment in which sufficient edema accumulates to produce ischemia and a second wave of necrosis. (Redrawn from McGoldrick MD: Diagnosis and management of acute renal failure: Part I. Cardiovasc Rev Rep 5:1031, 1984.)
1657
assess for ongoing muscle injury. If the patient is not a victim of trauma or compression injury, the underlying etiology of the rhabdomyolysis requires investigation to prevent recurrences.
KEY CONCEPTS ■
Classic clinical manifestations of rhabdomyolysis include myalgias, weakness, and tea-colored urine. Because only half of patients present classically, the emergency physician should consider rhabdomyolysis in patients at risk, particularly when they present with an altered sensorium. ■ Alcohol abuse, illicit drug use, certain medications, muscle overexertion, and traumatic muscle compression are the most common causes of rhabdomyolysis. ■ In rhabdomyolysis, urine dipstick testing is strongly positive for blood, with few, or no, RBCs on microscopic examination. The diagnosis is confirmed by an elevated serum CK level. ■ Early fluid resuscitation resulting in a urine output of 200 to 300 mL/hr reduces the risk of rhabdomyolysis-induced
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
renal failure. Intravenous fluids are the primary treatment, with mannitol and urine alkalinization as adjuncts of unproven benefit. Loop diuretics should not be used since they can acidify the urine. ■ Hyperkalemia can precipitate malignant cardiac dysrhythmias and must be treated promptly. ■ To avoid exacerbating the hypercalcemia, which normally occurs during the recovery period when excess intracellular calcium shifts into the extracellular space, initial asymptomatic hypocalcemia should not be corrected. ■ CK levels that do not decrease or continue to rise beyond 48 hours may indicate continued muscle injury and warrant a careful evaluation for compartment syndrome.
Chapter 125 / Rhabdomyolysis
risk for renal failure, however, mandates close monitoring of renal function, electrolytes, and hydration status, which usually requires admission to the hospital. Also, if the patient is a victim of trauma, serial CK levels should be followed to
Chapter 126
Thyroid and Adrenal Disorders
David Zull
Hyperthyroidism, hypothyroidism, and adrenal insufficiency are endocrine disorders that often manifest with chronic, nonspecific symptoms such as fatigue, weakness, and depression, and as a result are difficult to recognize in a typical medical encounter. With increased severity, each disorder has classic clinical manifestations that are more easily recognizable. Most importantly for the emergency physician, acute stresses can precipitate life-threatening illnesses in these patients, requiring vigorous medical management based on clinical judgment and suggestive laboratory data alone.
■ HYPERTHYROIDISM Perspective Background and Epidemiology Although the terms hyperthyroidism and thyrotoxicosis are often used interchangeably, hyperthyroidism refers to conditions in which the production of thyroid hormone is increased, whereas thyrotoxicosis is defined as any state in which thyroid hormone levels are increased in the blood, whether it be from overproduction (Graves’ disease, toxic multinodular goiter [TMG]), thyroid hormone release from an injured gland (thyroiditis), or exogenous thyroid hormone. The clinical spectrum of hyperthyroidism is a continuum from asymptomatic or subclinical disease to life-threatening thyroid storm. On large random population screenings, the prevalence of hyperthyroidism is 0.5 to 2.2%, with more than half of these patients considered subclinical (prehyperthyroid state with depressed thyroid-stimulating hormone [TSH] and normal tetraiodothyronine, thyroxine [T4]).1-3 The prevalence of hyperthyroidism in women is tenfold greater than in men. Graves’ disease is the predominant cause of thyrotoxicosis, with TMG becoming more common with increasing age, exceeding Graves’ disease by 2 to 1 in patients older than 55 years.4,5 Hyperthyroidism is rare in childhood, but when seen is related to Graves’ disease. It is estimated that 1 to 2% of patients with thyrotoxicosis will progress on to thyroid storm when an acute intercurrent stress supervenes.6
Principles of Disease The follicular cells of the thyroid gland produce T4 and triiodothyronine (T3), which are regulated by a feedback loop with the anterior pituitary gland, which produces TSH. If levels of 1658
T4 drop, TSH production is stimulated, whereas if the T4 level is high, TSH is suppressed. TSH is in turn regulated by the hypothalamus’s production of thyrotropin-releasing hormone (TRH). Thyroid hormone synthesis by the follicular cells starts with the production of thyroglobulin, a large hormonal precursor protein with numerous tyrosines in its structure. Iodine is then actively transported into follicular cells where it is oxidized and then bound to tyrosine residues. Linking of iodotyrosines within thyroglobulin produces T4 and T3, which are released into the circulation by proteolyisis. All of T4 is produced in the thyroid gland, whereas only 15 to 20% of T3 is synthesized directly; the remainder is formed by deiodonation of T4 in peripheral tissues. During systemic illness, deiodonation occurs, but at an inner ring of T4, rather than the outer ring, and reverse T3 is produced. T4 is a prohormone with only mild intrinsic activity, whereas T3 is the biologically active hormone, and reverse T3 is inactive. Over 99.5% of thyroid hormones are protein-bound in the serum to thyronine-binding globulin (TBG) and other proteins, rendering them metabolically inactive. As a result, only free T4 and free T3 are clinically relevant.4,6,7 Although iodide is a substrate for thyroid hormone production, excess iodide inhibits iodide trapping and thyroglobulin iodination (the Wolff-Chaikoff effect) and most importantly blocks the release of thyroid hormone from the gland. Iodide’s inhibition of thyroid hormone production and release is transient, with the gland escaping inhibition after 10 to 14 days. In contrast, an iodide load can induce hyperthyroidism (Jod-Basedow effect) in some patients with multinodular goiter and latent Graves’ disease, especially if the patient is iodine-deficient to begin with.4,6,7 Thyroid hormone has effects on the metabolism of all tissues, exerting these at several levels. Thyroid hormone regulates gene activity by interaction at nuclear receptors. It has direct effects on metabolism by interaction with cellular enzymes, like adenosine triphosphatase. Most importantly, T3 and T4 increase the number and sensitivity of beta-adrenergic receptors, dramatically increasing response to endogenous catecholamines.4,6 Graves’ disease is the most common cause of thyrotoxicosis and consists of the syndrome of hyperthyroidism, a diffuse symmetrical goiter, ophthalmopathy, and dermopathy. Graves’ disease primarily affects females between the ages of 20 and 40 years, often those with a family history of thyroid disease. It is an autoimmune disorder in which B lymphocytes produce
1659
deficiency). Most of these cases result from a destructive thyroiditis, but a minority are due to amiodarone’s iodine load (400 times the daily requirement), which may unmask hyperthyroidism in patients with multinodular goiter and subclinical Graves’ disease. An exacerbation of the tachyarrhythmia that the patient is being treated for or heart failure is the typical presentation of a patient with thyrotoxicosis related to amiodarone. Other drugs that may induce thyroiditis include interferon, interleukin-2, granulocyte-macrophage colonystimulating factor, and lithium.12,15,16 Chronic excess ingestion of thyroid hormone can result in thyrotoxicosis, referred to as thyrotoxicosis factitia. Although iatrogenic or patient errors may be responsible, medical personnel with psychiatric disease account for the majority of reported cases. Inadvertent ingestion of thyroid hormone in herbal products for weight reduction or in contaminated ground beef has been reported as well. Surprisingly, acute ingestions of thyroid hormone usually manifest only minor toxicity. The reasons for this are multiple: the 7-day half-life of T4, the suppression of T4-to-T3 conversion and inhibition of endogenous hormone production by negative feedback loops, and the down-regulation of thyroid hormone receptors.17,18 A variety of rare forms of thyrotoxicosis have been described, including struma ovarii, thyroid carcinoma, hydatidiform mole, choriocarcinoma, and TSH-secreting pituitary adenomas4,6,19 (Box 126-1).
Clinical Features The symptoms and signs of thyrotoxicosis are caused by a hypermetabolic state and increased beta-adrenergic activity. Clinical manifestations vary from minimal (apathetic hyperthyroidism) to life-threatening (thyroid storm) and depend on the patient’s age, duration of disease, the level and rate of rise of hormone levels, drug interactions, and the stress of intercurrent illness. Hyperadrenergic manifestations are often masked in the elderly. Thyrotoxicosis of long duration and gradual course may go unnoticed by many patients, or symptoms may be attributed to other causes like emotional stress, dieting, or physical deconditioning.4,6
BOX 126-1 Causes of Thyrotoxicosis Graves’ disease (toxic diffuse goiter) Toxic multinodular goiter Toxic adenoma (single hot nodule) Factitious thyrotoxicosis Thyrotoxicosis associated with thyroiditis Hashimoto’s thyroiditis Subacute (de Quervain’s) thyroiditis Postpartum thyroiditis Sporadic thyroiditis Amiodarone thyroiditis Iodine-induced hyperthyroidism (areas of iodine deficiency) Amiodarone Radiocontrast media Metastatic follicular thyroid carcinoma hCG-mediated thyrotoxicosis Hydatidiform mole Metastatic choriocarcinoma Hyperemesis gravidarum TSH-producing pituitary tumors Struma ovarii hCG, human chorionic gonadotropin; TSH, thyroid-stimulating hormone.
Chapter 126 / Thyroid and Adrenal Disorders
immunoglobulins that stimulate the TSH receptor (thyroidstimulating immunoglobulin [TSI]). The eye disease that accompanies the disease is thought to result from thyroid antibodies sensitized to common antigens in orbital fibroblasts and muscle.4,6 TMG is the second leading cause of hyperthyroidism, characterized by multiple autonomously functioning nodules typically developing in women older than 50 years of age. It is unusual in youth unless the patient has a preexisting nontoxic multinodular goiter or lives in a region of iodine deficiency. The population of the United States is generally iodinesufficient, but areas of the world with populations that are deficient include Central America, South America, the Himalayas, Eastern Europe, and Central Africa. The hyperthyroidism in TMG is milder than Graves’ disease and is gradual in onset, but acute presentations can occur when iodine replacement is given to an iodine-deficient individual. Because of the age of the patients, cardiovascular manifestations like atrial fibrillation and heart failure predominate,8 whereas tremors and hypermetabolic features are less pronounced than Graves’ disease. Muscle wasting and weakness is common, and the patient is often described as apathetic.5 As multinodular goiters often extend retrosternally, obstructive symptoms may occur.9,10 A single hyperfunctioning (hot) nodule referred to as a toxic adenoma may occur in this same population, but it is less common than the multinodular form. In thyroiditis, acute thyrotoxicosis can result from thyroid gland inflammation and cell breakdown with release of preformed thyroid hormone. Thyroiditis may be autoimmune in origin, infectious, or drug-induced. Hashimoto’s thyrioditis is the most common type of thyroiditis. It is an autoimmune disorder characterized by thyroid antibodies and lymphocytic infiltration of the thyroid gland. Patients present with painless goiter and hypothyroidism, but thyrotoxicosis is rarely evident early in the disease (hashitoxicosis).11 Related autoimmune disorders of the thyroid include postpartum thyroiditis and sporadic thyroiditis. They are also referred to as painless, or silent, thyroiditis due to their small nontender goiter and mild symptoms. Five to 10% of pregnant women develop transient thyrotoxicosis 1 to 6 months postpartum, followed by a hypothyroid state for up to 6 months, then a return to baseline. There is a 70% chance of recurrence in subsequent pregnancies, and some women develop permanent hypothyroidism. Sporadic thyroiditis, which may account for up to 1% of thyrotoxicosis, is a similar entity, except for its lack of association with pregnancy.12 Subacute thyroiditis (de Quervain’s thyroiditis) appears to be a viral or postviral disease that presents with a prodrome of fatigue, myalgias, and pharyngitis, followed by fever and severe anterior neck pain. Pain often radiates to the jaw and ears, and the gland is exquisitely tender. Symptoms of hyperthyroidism with sweats, palpitations, and tremor develop during this acute painful phase and may last several weeks, transitioning to a hypothyroid state for several months, then a return to a euthyroid state. Subacute thyroiditis may account for 2% of thyrotoxic patients and as is the case for other thyroid diseases, women predominate.13 Suppurative thyroiditis is a rare disorder that also presents with fever and anterior neck pain, but is marked by neck swelling, induration, and erythema and the presence of dysphonia and dysphagia. The cause is usually bacterial infection with abscess formation, but parasites, mycobacteria, and fungi may be responsible. Most patients have preexisting thyroid disease and are immunocompromised (AIDS).14 In North America, about 2% of patients treated with amiodarone develop thyrotoxicosis (higher in areas of iodine
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1660
Constitutional symptoms such as fatigue and generalized weakness are very common in thyrotoxicosis. Despite increased calorie intake, weight loss is seen in most patients, averaging about a 15% drop from baseline. To confound things, elderly patients often have a decreased appetite, leading to a suggestion of occult cancer. Hypermetabolic symptoms like heat intolerance, excessive sweating, and preference for the cold are most pronounced in younger patients.20 Neuropsychiatric complaints include anxiety, restlessness, tremor, feeling jittery or unable to sit still, insomnia, memory loss, and poor attention span. Family members often report emotional lability and agitation, which may progress to altered mental status and coma in thyroid storm. Weakness and fatigue of proximal muscle groups may result from thyroid myopathy, with patients often complaining of difficulty combing their hair, climbing stairs, or rising from a chair.21 A sudden and profound muscle weakness progressing to flaccid paralysis is described in a thyrotoxic variant of hypokalemic periodic paralysis.22 Cardiopulmonary symptoms are very common and include palpitations, dyspnea on exertion, and reduced exercise tolerance. Older patients may present with new-onset angina, atrial fibrillation, or congestive heart failure as the presenting and only symptoms of thyroid disease.8,16,23 Gastrointestinal complaints often include more frequent bowel movements, but not diarrhea. Dysphagia may result from enlargement of the thyroid gland in Graves’ disease or retrosternal extension of the gland in TMG. Nausea and vomiting may be seen with severe thyrotoxicosis.4,6,9 Reproductive endocrine function can be affected. Women often complain of a change in their menses, anywhere from amenorrhea to menometrorrhagia, and infertility is very common. Men may complain of a decrease in libido and breast swelling.4,6 Ocular complaints may result from ophthalmopathy in Graves’ disease, but not in other causes of thyrotoxicosis. A sense of irritation and excessive tearing are early symptoms, with diplopia, retrobulbar discomfort, blurring of vision, and foreign body sensation occurring late in the disease24 (Box 126-2). Physical examination of the thyrotoxic patient may reveal distinctive findings, especially in younger individuals. The patient often appears anxious and fidgety, with a fine tremor of the hands and tongue and lightly closed eyelids. The skin feels warm, smooth, and velvety, likened to a baby’s skin, especially over the elbows. The face is rosy and blushes
BOX 126-2 Patient Complaints in Thyrotoxicosis Constitutional: Weight loss despite hyperphagia, fatigue, generalized weakness Hypermetabolic: Heat intolerance, cold preference, excessive perspiration Cardiorespiratory: Heart pounding and racing, dyspnea on exertion, chest pains Psychiatric: Anxiety, restlessness, hyperkinesis, emotional lability, confusion Muscular: Tremor, difficulty getting out of a chair or combing hair Ophthalmologic: Tearing, irritation, wind sensitivity, diplopia, foreign body sensation Thyroid Gland: Neck fullness, dysphagia, dysphonia Dermatologic: Flushed feeling, hair loss, pretibial swelling Reproductive: Oligomenorrhea, decreased libido, gynecomastia
readily. The hands may reveal palmar erythema and the distal part of the nails may separate from the nail bed (onycholysis, or Plummer’s nails). Scalp hair is fine and brittle, and diffuse alopecia may occur.4,6 Specifically in Graves’ disease, about 5% of patients develop marked thickening of the pretibial skin by mucopolysaccharide infiltration of the dermis (pretibial myxedema). These lesions are painless, raised nodules and plaques that become confluent over the pretibial area and dorsum of the feet. Hyperpigmentation and induration are present, but pitting is absent, and pretibial myxedema is always associated with Graves’ eye disease.25 Tachycardia is seen in virtually all patients, and there is widening of the pulse pressure with bounding pulses. The apical impulse is prominent and the heart sounds are enhanced. A systolic flow murmur is usually present, and rarely a friction rublike sound along the left sternal border (Means-Lerman scratch) may also be heard. Atrial fibrillation can be seen at any age in hyperthyroidism, with an overall prevalence of 2%, but the frequency is age-dependant, rising to 15% in patients older than 70 years.26 Even subclinical hyperthyroidism appears to increase the prevalence of atrial fibrillation threefold over that in the average population.27 The ventricular response in thyrotoxic atrial fibrillation may be unusually fast, and the patient may be resistant to attempts to slow the rate as well as to convert to sinus rhythm. The chronic tachycardia and high cardiac output state in hyperthyroidism may lead to dilated cardiomyopathy, especially in elders and those with atrial fibrillation, resulting in an S3 gallop and basilar crackles.8,16,23,28 Primary pulmonary hypertension, sometimes associated with tricuspid regurgitation and right heart failure, may also be seen.16,29,30 The characteristic stare of thyrotoxicosis results from retraction of the upper and lower eyelids revealing a rim of sclera beyond the limbus. As the eyelids are sympathetically innervated, the increased sensitivity to adrenergic stimuli in thyrotoxicosis leads to the widening of the palpebral fissures. Other hyperadrenergic eye findings in thyrotoxic patients include lid lag and globe lag. In lid lag, the upper lid lags behind the globe when the patient is asked to look slowly downward. In globe lag, the globe lags behind the upper lid with slow upward gaze.4,24 Although stare is frequent in any form of thyrotoxicosis, proptosis of the globe is unique to Graves’ disease, resulting from mucopolysaccharide infiltration and inflammation of the ocular muscles and soft tissue leading to exophthalmos. Imaging with ultrasound, computed tomography, or magnetic resonance imaging reveals orbital swelling in virtually all patients with Graves’ disease, but only about 50% have clinical findings. Conjunctival injection, periorbital edema, and chemosis are early findings. Proptosis is defined by the anteroposterior distance from the lateral orbital ridge to the anterior cornea as greater than 20 mm. Progressive orbital involvement may lead to infiltration of the inferior rectus muscle with limitation of upward gaze. Very late findings can include keratitis from inability to close the eyes completely, and visual loss from optic nerve compression.24 Treatment of hyperthyroidism, especially with radioactive iodine, may paradoxically aggravate Graves’ eye disease.31 Chronic proximal muscle wasting and weakness may result from thyrotoxic myopathy.21 More acute weakness with flaccidity may occur in Asian and Latino males from the thyrotoxic form of hypokalemic periodic paralysis.22 Most thyrotoxic patients have a palpable abnormality on examination of the thyroid gland. In Graves’ disease, the gland is often two to three times normal size, but may be massively enlarged. A normal-size gland is unusual in younger patients, but more than 20% of older patients lack a goiter.4 The thyroid
Thyroid Storm Thyroid storm is a life-threatening decompensation of poorly controlled, untreated, or unrecognized thyrotoxicosis. It occurs in 1 to 2% of thyrotoxic patients and in about 10% of patients hospitalized for hyperthyroidism. Thyroid storm occurs predominately in Graves’ disease, but occasionally is seen in TMG and toxic adenoma.32,33 Occurrences in thyroiditis,34 factitious thyrotoxicosis,35 struma ovarii, hydatidiform mole,19 and other causes of thyrotoxicosis are limited to rare case reports. Thyroid storm is an exaggeration of the clinical manifestations of thyrotoxicosis, further distinguished by the presence of fever, marked tachycardia, central nervous system dysfunction, and gastrointestinal symptoms. Decompensation of one or more organ systems, such as shock or heart failure, also defines thyroid storm. If untreated, thyroid storm is uniformly fatal, and even with aggressive management it still carries a 20% mortality rate.6,32,36,37
BOX 126-3 Physical Findings in Thyrotoxicosis Vital Signs: Tachycardia, widened pulse pressure, bounding pulses, fever Cardiac: Hyperdynamic precordium, systolic flow murmur, prominent heart sounds, systolic rub (Means-Lerman scratch), tricuspid regurgitation, atrial fibrillation, evidence of heart failure Ophthalmologic: Widened palpebral fissures (stare), lid lag, globe lag, conjunctival injection, periorbital edema, proptosis, limitation of superior gaze Neurologic: Fine tremor, hyper-reflexia, proximal muscle weakness Psychiatric: Fidgety, emotionally labile, poor concentration Dermatologic: Warm, moist, smooth skin; fine, brittle hair; alopecia, flushed facies; palmer erythema; hyperpigmented pretibial plaques, nodules, or induration that is nonpitting; onycholysis Neck: Diffuse symmetrical thyroid enlargement, sometimes with a bruit and palpable thrill; thyroid with multiple irregular nodules or a prominent single nodule; tracheal deviation, venous prominence with arm elevation (Pemberton’s sign)
1661
Central to the pathophysiology of thyroid storm is an increase in catecholamine-binding sites and hence a heightened response to adrenergic stimuli. Superimposed on this vulnerable state is an acute stress that causes an outpouring of catecholamines that in conjunction with high levels of free T4 and T3 precipitates the exaggerated response we call thyroid storm.4,6,37 It was once thought that thyroid storm resulted from a sudden dumping of thyroid hormone into the circulation, but except for sudden cessation of antithyroid therapy in a hyperthyroid patient and blunt or penetrating trauma to the thyroid gland (in both hyperthyroid and euthyroid patients) in which hormone leaks from injured acini, rapid rises in T4 and T3 are not responsible for storm.38-41 In fact, hormone levels in thyroid storm are not generally distinguishable from poorly controlled thyrotoxicosis.7 Infection and sepsis are the most common precipitants of thyroid storm, but as fever is a prominent feature of storm, the thyrotoxic state may be overlooked. Historically, thyroid and nonthyroid surgery have been the leading triggers of thyroid storm, but identification and treatment of thyrotoxicosis preoperatively has dramatically decreased this as a precipitant. Other common precipitating events include myocardial infarction, stroke, pulmonary embolism, diabetic ketoacidosis, parturition, trauma, and administration of iodinated contrast media and amiodarone (Box 126-4). The clinical presentation of thyroid storm is often dramatic. Although many of the findings of thyrotoxicosis are evident on
BOX 126-4 Precipitants of Thyroid Storm Medical Infection/sepsis Cerebral vascular accident Myocardial infarction Congestive heart failure Pulmonary embolism Visceral infarction Emotional stress Acute manic crisis Trauma Thyroid surgery Nonthyroid surgery Blunt and penetrating trauma to the thyroid gland Vigorous palpation of the thyroid gland Burns Endocrine Hypoglycemia Diabetic ketoacidosis Hyperosmolar nonketotic coma Drug-Related Iodine-131 therapy Premature withdrawal of antithyroid therapy Ingestion of thyroid hormone Iodinated contrast agents Amiodarone therapy Iodine ingestion Anesthesia induction Miscellaneous drugs (chemotherapy, pseudoephedrine, organophosphates, aspirin) Pregnancy-Related Toxemia of pregnancy Hyperemesis gravidarum Parturition and the immediate postpartum period
Chapter 126 / Thyroid and Adrenal Disorders
gland in Graves’ disease is symmetrical, smooth, soft to rubbery consistency, and has no evident nodules. In severe disease a palpable thrill and audible bruit are present and are usually continuous, rather than the systolic bruit seen with vascular disease. In TMG, the gland is variably enlarged and although multiple irregular nodules are often palpable, a single dominant nodule is not unusual, making distinguishing TMG from toxic adenoma difficult. The multinodular goiter may extend retrosternally, hiding it on examination unless the supraclavicular area is palpated on swallowing. By mass effect, a multinodular goiter may cause tracheal deviation and hoarseness as well as facial and neck vein engorgement, the latter becoming evident when the arms are elevated above the head (Pemberton’s sign)9,10 (Box 126-3). In subacute thyroiditis, the thyroid gland is exquisitely tender, but redness or warmth of the overlying skin is only seen in suppurative thyroiditis. In the more typical entities of sporadic, postpartum, and Hashimoto’s thyroiditis, the gland is nontender with modest to no enlargement.11-14 In factitious thyrotoxicosis, the gland may be atrophic but this may be difficult to appreciate.
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1662
exam, certain features distinguish thyroid storm. Fever out of proportion to the physician’s expectations is characteristic of storm, sometimes mimicking heatstroke with temperatures exceeding 106° F. Inappropriately excessive diaphoresis is frequently observed, and sinus tachycardia over 140 beats per minute is common.4,6,36,37 Heart rates exceeding 150 beats per minute may be indicative of atrial fibrillation or other supraventricular tachycardias. In addition, symptoms and signs of congestive heart failure often accompany such rapid rates.8,16 Altered mental status from metabolic encephalopathy is a hallmark of thyroid storm, ranging from restlessness and agitation to delirium, seizures, and coma.21,42 Gastrointestinal symptoms are often pronounced, with nausea, vomiting, and diarrhea leading to volume depletion and hypotension. Abdominal pain mimicking bowel obstruction may be present. An unusual complication of severe thyrotoxicosis is cholestatic jaundice, which carries a bad prognosis if hepatic failure ensues.43 Although hyperthermia, exaggerated tachycardia, and altered mental status can quickly identify possible thyroid storm, the clinical presentation can be difficult to differentiate from uncomplicated thyrotoxicosis. Taking into account the severity of fever, tachycardia, central nervous system dysfunction, congestive heart failure, and gastrointestinal symptoms, Burch and Wartofsky developed a scoring system to help distinguish uncomplicated thyrotoxicosis from impending thyroid storm and true thyroid storm (Table 126-1). Although this scoring system has not been rigorously tested, it may prove useful in decisions to treat in borderline cases.44
Diagnostic Strategies The best screening tool for the diagnosis of thyrotoxicosis is the ultrasensitive TSH, which is depressed or undetectable in thyrotoxicosis. A normal TSH excludes hyperthyroidism and an elevated TSH is diagnostic for hypothyroidism, except in the rare circumstance of secondary hyperthyroidism from overproduction of TSH by a pituitary adenoma. Although a nondetectable TSH is specific for thyrotoxicosis, a modest depression of the TSH measurement is not always the result of mild or subclinical hyperthyroidism. Severe systemic illness may depress TSH production, leading to low levels of TSH, free T3, and free T4. This nonthyroidal illness pattern is often referred to as the euthyroid sick syndrome, and it appears to be a transient form of central hypothyroidism, an adaptive response to slow metabolism during systemic stress. Chronic conditions in which TSH may be suppressed include anorexia nervosa, depression, and renal failure. Medications, including dopamine, glucocorticoids, somatostatin, and octreotide, may also depress TSH levels.6,7 Although screening for thyroid disease with TSH is a reasonable strategy, measurement of thyroid hormone levels is required for a definitive diagnosis. Total T3 and total T4 assays may be misleading as they are influenced by changes in TBG. Increases in TBG with resultant false elevations in total T3 and T4 are seen in pregnancy, infectious hepatitis, and drug therapy with estrogens, tamoxifen, methadone, or heroin. In contrast, decreases in TBG with subsequent low total T3 and T4 are seen in cirrhosis, malnutrition, and nephrotic syndrome, as well as treatment with androgens or glucocorticoids. Lastly, many drugs inhibit the binding of T3 and T4 to TBG, thus resulting in higher levels of free T3 and free T4 levels, which will not be reflected in the total hormone measurements. Such drugs include salicylates, nonsteroidal anti-inflammatories, heparin, furosemide, diphenylhydantoin, carbamazepine, and sulfonylureas. Because of the many limitations in the measure-
Table 126-1 Diagnostic Criteria for Thyroid Storm SCORE
Fever (°F) 99–99.9 100–100.9 101–101.9 102–102.9 103–103.9 ≥104
5 10 15 20 25 30
Tachycardia (beats/min) 90–109 110–119 120–129 130–139 ≥140
5 10 15 20 25
Mental Status Normal Mild agitation Delirium, psychosis Extreme lethargy Coma/seizures Congestive Heart Failure Absent Mild (edema) Moderate (rales) Pulmonary edema Atrial fibrillation Gastrointestinal and Hepatic Symptoms None Nausea, vomiting Diarrhea, abdominal pain Unexplained jaundice Precipitating Event None Present
0 10 20 30 0 5 10 15 10 0 10 20 0 10
Adapted from Burch HB, Wartofsky L: Life-threatening thyrotoxicosis: Thyroid storm. Endocrinol Metab Clin North Am 1993; 22: 263–277. Tally the maximum score from each category. A score of 45 or greater suggests thyroid storm; a score of 25–44 suggests impending storm, and a score below 25 is unlikely to represent thyroid storm.
ment of total hormone levels, only free T3 and free T4 assays should be relied on.6,7 The combination of both free T4 and free T3 elevation with TSH suppression is diagnostic of thyrotoxicosis. If TSH is suppressed and free T4 is normal, subclinical hyperthyroidism is likely; however, about 5% of patients with thyrotoxicosis have an elevated free T3 and normal free T4—referred to as T3 toxicosis, an entity more common in TMG. The reverse situation, in which free T3 is normal and free T4 is elevated may be seen in thyroiditis, exogenous levothyroxine ingestion, and hyperthyroidism in the elderly, often with suppressed T4 to T3 conversion due to comorbid illness4,7 (Table 126-2). Differentiating Graves’ disease from other forms of thyrotoxicosis is usually straightforward clinically, but the measurement of thyroid antibodies to thyroglobulin and thyroid peroxidase may be helpful in questionable cases. In addition to thyroid function tests, multiple laboratory abnormalities may be seen in thyrotoxicosis and thyroid storm. Hyperglycemia is the most common abnormality, seen in up to half the patients, likely related to glycogenolysis and catecholamine-mediated antagonism of insulin. Mild hypercalcemia is seen in 10% of patients and is related to hormone-mediated bone resorption, osteoporosis, and increased fracture risk.
1663
Table 126-2 Thyroid Function Test Interpretation FREE T4
FREE T3
DISEASE
Normal Low Low Low Low
Normal High Normal Normal High
Normal High Normal High Normal
Low
Low
Low
High
Normal
Normal
High High
Low High
Low High
None Hyperthyroidism Subclinical hyperthyroidism T3 toxicosis Thyroiditis, T4 ingestion, hyperthyroidism in the elderly or with comorbid illness Euthyroid sick syndrome; central hypothyroidism Subclinical hypothyroidism; recovery from euthyroid sick syndrome Primary hypothyroidism TSH producing pituitary adenoma
T3, triiodothyronine; T4, thyroxine; TSH, thyroid-stimulating hormone.
Abnormal liver function tests are frequent in hyperthyroidism. The abnormalities observed include mild increases in serum aspartate transaminase, alanine transaminase, lactate dehydrogenase, bilirubin, and, most commonly, alkaline phosphatase. Although elevated serum bilirubin occurs in hyperthyroidism, clinical jaundice develops infrequently. Other abnormalities may include a leukocytosis with a left shift, a mild normocytic normochromic anemia, and low serum cholesterol levels.7 The diagnostic evaluation for thyroiditis is more difficult. If there is exquisite gland tenderness and a sedimentation rate greater than 100, the diagnosis of subacute thyroiditis is likely. Other forms of thyroiditis, however, lack these findings.12,13 Doppler ultrasound of the thyroid may be helpful in differentiating among the hypervascular enlarged gland of Graves’ disease, the nodules of TMG, and thyroiditis or factitious thyrotoxicosis (decreased Doppler flow).45 Another option is a radioactive iodine uptake, which is depressed in thyroiditis and factitious thyrotoxicosis but increased in hyperthyroidism. If exogenous thyroid hormone abuse is suggested, measurement of thyroglobulin levels may confirm the diagnosis, being very low in factitious thyrotoxicosis but elevated in all other forms of thyrotoxicosis.35
Differential Considerations The overtly thyrotoxic patient is often thought to be very anxious, manic, or in the midst of a panic attack. In addition, hyperadreneric signs may suggest sympathomimetic (cocaine, amphetamine) or anticholinergic intoxication or a withdrawal syndrome (alcohol, narcotics, sedative-hypnotics). The high fever and altered mental status seen in thyroid storm may mimic heatstroke, neuroleptic malignant syndrome, serotonin syndrome, bacterial meningitis, and sepsis. In elders, the hyperadrenergic features of thyrotoxicosis may be masked, facial muscles may lack expression, and mental status may be depressed leading to the syndrome of apathetic hyperthyroidism. Patients with multinodular goiters and those older than 70 are most likely to present in this manner. New-onset atrial fibrillation and congestive heart failure exacerbations are often the presenting symptoms of apathetic hyperthyroidism. In addition, elders with thyrotoxicosis may have significant weight loss without increased appetite, suggestive of occult cancer.
Patients with mild thyrotoxicosis with minor symptoms can often await outpatient follow-up for initiation of treatment. Of more importance to the emergency physician is the avoidance of interventions that may increase thyroid hormone levels or accentuate adrenergic stimuli. Thyrotoxic patients should not receive iodinated contrast media or amiodarone, both of which present an iodine load that may enhance thyroid hormone production.46 Caution is advised with the use of aspirin and nonsteroidal anti-inflammatories as they may interfere with protein binding of thyroid hormone, leading to increases in free T4 and T3.47 Drugs such as pseudoephedrine, ketamine, and albuterol that increase sympathomimetic tone should also be used with caution.48 Patients with thyrotoxicosis who are symptomatic may require initiation of beta-blocker therapy in the emergency department, but the initiation of thionamides like propylthiouracil (PTU) and methimazole are rarely indicated. The prompt recognition and treatment of thyroid storm is crucial for patient survival. Therapeutic interventions have several aims: (1) reducing production of thyroid hormone, (2) inhibiting thyroid hormone release, (3) blocking peripheral conversion of T4 to T3, (4) initiating beta-adrenergic blockade, (5) instituting general supportive measures, and (6) identifying and treating the precipitating event.
Reducing Production of Thyroid Hormone The first-line treatment of thyroid storm is the use of thionamides, which inhibit oxidation and organic binding of iodine to thyroglobulin, thus blocking synthesis of thyroid hormone. PTU and methimazole are available, but PTU is preferred due to its additional effect of impairing conversion of T4 to T3. PTU is given as an initial loading dose of 600 to 1000 mg by mouth, followed by 200 to 250 mg every 4 hours. The recommended dose for methimazole is 20 to 25 mg initially with the same dose repeated every 4 hours.4,6,32,36,37 If a patient cannot take medication orally, the same dose can be given by nasogastric tube or by retention enema. Such solutions require pharmacy preparation on a case-by-case basis.49 Due to solubility limitations, there is no IV form of PTU or methimazole, yet methimazole has been tried intravenously. Such solutions also require pharmacy preparation, and can be administered 30 mg every 6 hours.50 The intravenous (IV) route should be considered only in a dire situation where oral or rectal administration is not feasible or ineffective (Box 126-5).
Inhibiting Thyroid Hormone Release Although thyroid hormone synthesis can be stopped by thionamides, preformed hormone in the gland is still available for release. Inorganic iodine blocks the release of thyroid hormone stored in the gland, but administration should be delayed at least 1 hour after PTU or methimazole is started. The reason for this delay is that an iodine load presented to an actively synthesizing gland provides further substrate for hormone production and release. Iodine is given as saturated solution of potassium iodide (SSKI) 5 gtt every 6 hours or Lugol’s solution 8 gtt every 6 hours. Published dosage recommendations range from one-half to double these doses, yet the effective iodine dose appears to be just a fraction of these numbers. As there is no apparent harm in administering larger iodine doses, this middle figure is most commonly recommended.4,6,32,36,37 Because no IV form of iodide is available, the rectal route can be used if the oral or nasogastric route cannot be used.49 If allergy to iodine is encountered, lithium is an alternative agent
Chapter 126 / Thyroid and Adrenal Disorders
TSH
Management
1664
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
BOX 126-5 Management of Thyroid Storm Inhibition of Thyroid Hormone Synthesis Propylthiouracil 600–1000 mg loading dose, then 200–250 mg every 4 hr OR Methimazole 20–25 mg initially, then 20–25 mg every 4 hr (Preferred route: PO or NG. Alternative route: PR. Enema prepared by pharmacy. Same dose for all routes. No IV preparation is available, but IV methimazole can be prepared with the use of a Millipore filter and given 30 mg every 6 hr) Inhibition of Thyroid Hormone Release Saturated solution of potassium iodide (SSKI) 5 gtt by mouth, NG, or PR every 6 hr OR Lugol’s solution 8 gtt by mouth, NG, or PR every 6 hr OR Sodium Iodide 500 mg in solution prepared by pharmacy IV every 12 hr OR If allergic to iodine, lithium carbonate 300 mg by mouth or NG every 6 hr Beta-adrenergic Blockade Propanolol 60−80 mg PO every 6 hr OR Metoprolol 50 mg PO every 6 to 12 hr If IV route required, propanolol 0.5–1.0 mg IV slow push test dose, then repeat every 15 min to desired effect, then 2–3 mg every 3 hr
OR Esmolol 250–500 µg/kg bolus, then 50–100 µg/kg/min infusion Strict contraindication to beta-blocker: reserpine 0.5 mg PO every 6 hr Administration of Corticosteroids (inhibit T4 to T3 conversion, treat relative adrenal insufficiency) Hydrocortisone 300 mg IV, followed by 100 mg every 6 hr OR Dexamethasone 2–4 mg IV every 6 hr Diagnosis and Treatment of Underlying Precipitant Consider empirical antibiotics if critical Supportive Measures Volume resuscitation and replacement of glycogen stores D5/0.9NS 125–1000 mL/hr depending on volume status and CHF Tylenol with caution Cooling blanket, fans, ice packs, ice lavage Miscellaneous Lorazepam or diazepam as anxiolytic and to decrease central sympathetic outflow l-Carnitine (block entry of thyroid hormone into cells), 1 g PO every 12 hr Cholestyramine (block enterohepatic recirculation of thyroid hormone), 4 g PO every 6 hr
CHF, congestive heart failure; D5/0.9NS, 5% dextrose in 0.9% normal saline; IV, intravenous; NG, nasogastric; PO, by mouth; PR, in rectum; T3, triiodothyronine; T4, thyroxine.
that impairs thyroid hormone release. The lithium dose is 300 mg every 6 hours by mouth or nasogastric tube, but lithium levels should be monitored to maintain a level of about 1 mg/ L. Iodine should not be used in amiodarone-induced thyrotoxicosis as iodine overload may contribute to amiodarone’s toxicity. In hyperthyroidism unmasked by iodine excess (iodinated contrast agents), lithium should be used to inhibit hormone release and further iodine administration avoided. Iodine’s effects cease after 2 to 3 weeks of therapy, thus a delayed exacerbation of hyperthyroidism may ensue unless adequate thionamide therapy has been maintained.51
Blocking Peripheral Conversion of T4 to T3 and Initiating Beta-adrenergic Blockade Blockade of peripheral hyperadrenergic activity by betablockers is a cornerstone of therapy in thyroid storm and symptomatic thyrotoxicosis. Propanolol has traditionally been the beta-blocker of choice because it blocks conversion of T4 to T3 and its nonselective effects also improve tremor, hyperpyrexia, and restlessness. Dosage recommendations vary from 20 to 120 mg by mouth at 6-hour intervals, with most authors suggesting 60 to 80 mg per dose.4,6,32,36,37 If the patient cannot take anything by mouth or rapid beta-blockade is desired while waiting for the oral dose to be effective, IV propanolol can be administered as a test dose of 0.5 to 1 mg over 10 minutes. A cautious start is recommended if there is evidence of severe heart failure or hypotension as there are rare case reports of cardiovascular collapse after propanolol administration in thyroid storm.52 If the patient tolerates the initial dose of propanolol, it can be repeated every 15 minutes until
the desired effect is achieved, then transitioning to 3-hour intervals with 1 to 3 mg boluses. If there are contraindications or concerns about beta-blocker therapy, a short-acting agent such as esmolol may be prudent. Esmolol is usually started as a loading dose of 250 to 500 µg/kg, then continued as an infusion of 50 to 100 µg/kg/min.53 Beta1-selective drugs like esmolol or metoprolol (50 mg every 6–12 hours) may be preferable in asthma patients, but if not tolerated, reserpine 0.5 mg orally every 6 hours could be considered, while monitoring for hypotension.54 Corticosteroids are recommended in thyroid storm, because they inhibit peripheral conversion of T4 to T3, as well as block the release of hormone from the gland. The synergistic effect of PTU, iodide, and steroids in thyrotoxicosis can restore the concentration of T3 to normal within 24 to 48 hours.36 Corticosteroids are also suggested due to an absolute or relative adrenal insufficiency that can occur in thyroid storm. Addison’s disease can occur concomitantly with Graves’ disease in polyglandular autoimmune syndrome type 2, but more importantly, the increased clearance of cortisol in thyrotoxicosis coupled with the high demand for cortisol in such critically ill patients leads to a relative adrenal insufficiency in most. Hydrocortisone can be given as an initial bolus of 100 to 300 mg IV, followed by 100 mg every 8 hours for several days. Dexamethasone has also been used in this setting as well, in doses of 2 to 4 mg every 6 hours or 8 mg every 24 hours.55
Instituting General Supportive Measures Supportive measures are equally important in the management of thyroid storm. Fluid resuscitation should be vigorous
1665
BOX 126-6 Thyrotoxicosis and Thyroid Storm: Special Situations
Atrial Fibrillation Beta-blocker preferred for rate control (doses as in Box 126-5) Calcium channel blockers prone to hypotension; diltiazem 10-mg test dose. Avoid verapamil Digoxin less effective but may be tried
Amiodarone should be avoided due to iodine load Refractory to conversion to sinus unless euthyroid first Thyroiditis (Subacute) NSAIDs for inflammation and pain control Prednisone 40 mg/day if refractory to NSAIDs Beta-blockade to control thyrotoxic symptoms No role for PTU, methimazole, or iodides Factitious Thyrotoxicosis Beta-blockade for thyrotoxic symptoms Cholestyramine to block absorption of ingested thyroid hormone No role for PTU, methimazole, or iodides
ACEI, angiotensin-converting enzyme inhibitor; BP, blood pressure; EF, ejection fraction; NSAIDs, nonsteroidal anti-inflammatory drugs; PTU, propylthiouracil.
unless clear signs of congestive heart failure are evident. Due to the depletion of glycogen stores in thyrotoxicosis, a 5% dextrose solution is recommended; hence for volume replacement, D5/0.9NS is a rational choice. As high fever to the point of heatstroke is not unusual in thyroid storm, therapy to dissipate heat is a priority, typically by means of cool mists, fans, ice packs, cooling blankets, and ice water lavage. Acetaminophen could be used for moderate fever, but as hepatic dysfunction may occur in storm, it should be used with caution. Aspirin is contraindicated in thyroid storm because it increases levels of free thyroid hormone.4,6,32,36,37 Treatment to hasten elimination of thyroid hormone has been described. Cholestyramine, an anion exchange resin, binds thyroid hormone in the bowel lumen, thus interrupting enterohepatic recirculation. Cholestyramine in a dose of 4 g every 6 hours has been shown to result in a more rapid decline in hormone levels than thionamides alone.56 Colestipol has been shown to have a similar effect, but not ezetimibe.57 Progressive deterioration in a patient with thyroid storm, despite aggressive multidrug therapy, may lead to the consideration of plasmapharesis, plasma exchange, or dialysis to attempt rapid reduction in thyroid hormone levels.58 Benzodiazepines for agitation and hypomania of thyrotoxicosis may be considered as the hyperadrenergic state resembles cocaine intoxication. l-Carnitine has been described in thyroid storm, its suggested mechanism being the inhibition of thyroid hormone entry into cell nuclei. The dose used is 1 g every 12 hours by mouth59 (Box 126-5). Radioactive iodine or surgery has no role in the management of thyroid storm or thyrotoxicosis until a sustained euthyroid state has been achieved, as these interventions can precipitate storm themselves.60 Beta-blockers are a mainstay in the treatment of high-output heart failure as defined by a normal or exaggerated ejection fraction on echocardiography.8,16 Preexisting heart disease aggravated by thyrotoxicosis may be associated with lowoutput congestive heart failure (low ejection fraction on echocardiography), in which circumstance caution should be exercised with the use of beta-blockers, because cases of cardiovascular collapse have been described with their use in this setting.52 Routine management with angiotensin-
converting enzyme inhibitors, diuretics, and digoxin are appropriate in both groups.8,16 The management of atrial fibrillation in thyrotoxicosis also has unique features. The rapid ventricular response generally requires high doses of beta-blocker for control. If thyroid storm is present, calcium channel blockers should be avoided as hypotension is a potential complication. Digoxin tends to be ineffective in this setting, but could be tried. Amiodarone should not be used due to its iodine load and potential to induce thyroiditis. Attempts to convert to sinus rhythm are usually fruitless while the patient remains thyrotoxic and thus should be postponed until the patient is euthyroid.16,26 Pain and tenderness as seen in subacute thyroiditis is treated with nonsteroidal anti-inflammatories. If refractory or recurrent, prednisone may be required.12,13 Thyrotoxicosis in thyroiditis is usually mild, and beta-blockers alone are recommended. In fact, thionamides and iodine have no effect in thyroiditis. If drug-related, the offending agent (amiodarone, interferon) should be stopped immediately. Thyrotoxicosis from exogenous thyroid hormone ingestion should be treated with beta-blockade alone because the thyroid gland is shut down, thereby rendering thionamides and iodine ineffective. Cholestyramine could be used to bind hormone in the gut in both the acute and chronic intoxication, but evidence to its efficacy is limited17,35 (Box 126-6). Special mention should be made regarding potential toxicity of thionamide therapy, as it may be a presenting manifestation of a thyrotoxic patient. The minor adverse reactions that occur in up to 5% of patients include drug fever, alteration in sense of taste, skin eruptions, arthralgias, and sialoadentitis. Such reactions should not lead to discontinuation of therapy in the patient with thyroid storm, but should be reason to stop in the mildly thyrotoxic individual. The most feared and life-threatening adverse reaction to PTU and methimazole is agranulocytosis, which is often heralded by onset of fever and severe sore throat. Any patient who develops fever while on thionamide therapy should have his or her white blood cell count determined and if at all depressed, therapy should be stopped immediately. Fortunately, such reactions occur only in 3 or 4 patients per 1000. Other infrequent yet serious reactions to thionamides include hepatitis, vasculitis, and polyarthritis.61
Chapter 126 / Thyroid and Adrenal Disorders
Congestive Heart Failure If rate-related, high-output failure Beta-blockade is first-line therapy (dose as in Box 126-5) ACEI, digoxin, diuretics as needed If depressed EF Avoid beta-blocker or 14 dose ACEI if BP adequate Digoxin and furosemide as needed If pulmonary hypertension Oxygen Sildenifil
1666
The clinical presentation of the hypothyroid patient can vary from asymptomatic or subclinical cases to life-threatening myxedema coma. In random population sampling, the prevalence of TSH elevation has ranged from 3.7 to 9.5%, with the majority of these having a normal free T4, which by definition is subclinical hypothyroidism. Overt hypothyroidism (elevated TSH and depressed free T4) is seen in a minority of these patients, about 0.3% of the population overall, with the prevalence rising with age, such that patients older than 80 years have a fivefold greater likelihood of developing hypothyroidism than do 12- to 49-year-olds.1-3 Some surveys estimate the incidence of hypothyroidism as high as 20% in elders.5 The female to male ratio is about 4 : 3 in hypothyroidism, in contrast to 8 : 1 for hyperthyroidism. Race differences are notable in that hypothyroidism is seen in 5.1% of whites, 4.1% of Hispanic-Americans, and 1.7% of African-Americans.1-3
ism to overt hypothyroidism and sometimes precipitating hypothyroidism de novo. In contrast, iodine-deficient patients administered an iodine load increase production of thyroid hormone and may develop hyperthyroidism. Interferon-alfa can result in hypothyroidism by precipitating Hashimoto’s thyroiditis. Overtreatment with PTU or methimazole may lead to hypothyroidism due to their inhibition of hormone synthesis. Phenytoin, carbamazepine, phenobarbital, and rifampin may aggravate hypothyroidism by enhancing metabolism of thyroid hormone.62-64 Patients on thyroid replacement therapy can develop hypothyroidism when drugs are introduced that interfere with hormone absorption, including iron, calcium, phosphate binders, sucralfate, aluminum hydroxide, cholestyramine, colestipol, and even coffee.65 Iatrogenic causes of hypothyroidism include neck irradiation for cancer or lymphoma and thyroidectomy for nodular goiter or thyroid cancer. Hypothyroidism may be seen as a late phase of thyroiditis.12 In Hashimoto’s disease the initial hyperthyroid phase is rarely identified, and hypothyroidism predominates.11 In subacute, silent, and postpartum thyroiditis the hyperthyroid phase is usually clinically evident, and the hypothyroidism is often very mild and transient, but can become chronic.13 Rare causes of hypothyroidism include inherited disorders of hormone biosynthesis and central hypothyroidism. Central causes are usually due to pituitary destruction by an adenoma, hemorrhage (Sheehan’s syndrome), or infiltration (sarcoid, amyloid), but can also result from hypothalamic dysfunction.66 A form of central hypothyroidism that appears to be an adaptive response to significant nonthyroidal illness is the euthyroid sick syndrome. Mild suppression of TSH release leads to a decrease in free T4 and T3. Impairment of T4 to T3 conversion also develops leading to elevation of reverse T3 levels. Euthyroid sick syndrome remits spontaneously with resolution of the acute illness, and treatment with thyroid replacement is not indicated. Drugs that may contribute to TSH suppression in nonthyroidal illness include glucocorticoids, dopamine, and octreotide62,64,67 (Box 126-7).
Principles of Disease
Clinical Features
The etiology of hypothyroidism includes primary thyroid failure, thyroiditis, pituitary/hypothalamic causes, drug-related and iatrogenic. The vast majority of hypothyroidism encountered in the United States is due to thyroid gland failure, and the majority of these are caused by autoimmune destruction of the gland in Hashimoto’s thyroiditis. In younger patients, the disease is associated with a goiter and elevated titers of antithyroid antibodies, specifically to thyroid peroxidase, thyroglobulin, and TSH. The TSH receptor antibody in Hashimoto’s disease blocks the receptor, in contrast to the stimulating antibody in Graves’ disease. In older patients, the thyroid gland is typically atrophic, and evidence of autoimmunity is often lacking.62-64 End-stage Graves’ disease can also result in autoimmune destruction of the thyroid gland, occurring spontaneously following several exacerbations of hyperthyroidism. More commonly, hypothyroidism follows treatment of Graves’ disease with radioactive iodine or thyroidectomy. Drug-induced hypothyroidism is often encountered with lithium carbonate because it inhibits hormone release. Iodine excess, as seen with amiodarone, iodinated contrast media, kelp supplements, and iodine-containing cough medicines, can impair thyroid hormone release and synthesis (WolffChaikoff effect), thereby converting subclinical hypothyroid-
Symptoms and signs of hypothyroidism are often very subtle and difficult to recognize in their milder presentation. Patients often ignore or tolerate symptoms when their development is very gradual, as is the case in Hashimoto’s thyroiditis, where the delay from symptom appearance and diagnosis may be several years. Acute hypothyroidism presenting over weeks to months may be seen in thyroiditis or withdrawal of exogenous thyroid hormone. Chronic disease may present acutely due to drug toxicity or when an intercurrent illness is superimposed.62-64 The clinical manifestations of hypothyroidism result from changes induced by lack of thyroid hormone, most notably a generalized slowing of metabolic processes (due to altered gene expression and decreased catecholamine sensitivity) and an accumulation of glycosaminoglycans (decreased metabolism) in interstitial fluids. Patients with hypothyroidism typically have pale, cool skin from decreased blood flow and fluid accumulation. Epidermal and sweat gland changes result in dry, scaly, rough skin. The skin is firm to the touch and appears swollen, but does not pit. In severe, chronic disease, the patient has a typical facies characterized by puffy eyelids, broad nose, swollen lips, and macroglossia. The hair in hypothyroidism becomes coarse and brittle, and alopecia is common. Thinning of the lateral third
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
Identifying and Treating the Precipitating Event Simultaneous with the previously detailed treatment measures, an aggressive search for an underlying precipitant of thyroid storm should be undertaken. As infection is the most common culprit, a chest radiograph, urinalysis, and blood cultures are routine. Silent myocardial ischemia should be assessed by an electrocardiogram and troponin, and the possibility of stroke or pulmonary embolism considered. Empirical use of antibiotics are tempting in the setting of thyroid storm, but restraint is prudent unless there is strong clinical evidence. Aggressive management of thyroid storm with PTU followed by iodine, beta-blockers, corticosteroids, fluid resuscitation, rapid cooling, and treatment of the precipitating illness can resolve fever, tachycardia, and altered mental status within a 24-hour period. All patients with storm should be admitted to an intensive care setting, and any interruption in therapy should be avoided as it can lead to a sudden recrudescence of symptoms and death.
■ HYPOTHYROIDISM Perspective
1667
BOX 126-7 Causes of Hypothyroidism
Central Hypothyroidism Euthyroid sick syndrome Pituitary disease Pituitary adenoma Hemorrhage Infiltrative (amyloid, sarcoid) Hypothalamic disease
of the eyebrows may occur. The nails also become brittle and thin. The skin may take on a yellowish tinge from carotene, which accumulates because of impaired conversion to vitamin A. Carotenemia is distinguished from jaundice by the sparing of the conjunctiva. Vitiligo may occur in association with polyglandular syndrome, whereas hyperpigmentation may be seen if the patient has concomitant Addison’s disease (Schmidt’s syndrome).62,68 Generalized edema of the face and extremities may develop—nonpitting from accumulation of glycosaminoglycans and pitting from a capillary leak phenomenon seen in hypothyroidism. A localized pretibial myxedema and exophthalmos may still be seen in patients with Graves’ disease rendered hypothyroid by surgery or radioactive iodine.68 The hypothyroid patient is usually normothermic, but complaints of cold intolerance and cool extremities are common. Blood pressure is usually normal, but 20 to 40% of patients have diastolic hypertension and narrowing of the pulse pressure. Bradycardia is common, but asymptomatic.62,63 In contrast to hyperthyroidism, in which atrial arrhythmias are frequently seen, hypothyroidism may be associated with QT prolongation
Chapter 126 / Thyroid and Adrenal Disorders
Primary Hypothyroidism Autoimmune hypothyroidism Hashimoto’s thyroiditis (chronic—atrophic thyroid, acute with goiter) Graves’ disease (end stage) Iatrogenic Radioactive iodine therapy for Graves’ disease Thyroidectomy for Graves’ disease, nodular goiter, or thyroid cancer External neck irradiation for lymphoma or head and neck cancer Iodine-related Iodine deficiency (common worldwide, but rare in North America) Iodine excess (inhibition of hormone release can unmask autoimmune thyroid disease) (see under Drug-related) Drug-related Lithium (inhibit hormone release) Amiodarone (destructive thyroiditis or iodine excess) Interferon-alfa (precipitate Hashimoto’s thyroiditis) Iodine excess (iodinated contrast media, kelp, amiodarone) Propylthiouracil, methimazole Interference with thyroid hormone absorption in patients on replacement therapy (iron, calcium, chromium, phosphate binders, cholestyramine, colestipol) Thyroiditis Subacute Silent (sporadic) Postpartum Amiodarone Congenital defect in thyroid hormone synthesis
and ventricular irritability.69,70 Patients with hypothyroidism often complain of dyspnea on exertion and decreased exercise capacity, and although decreased cardiac contractility and diastolic dysfunction is present in chronic hypothyroidism, signs of congestive heart failure are usually absent. Angina and coronary artery disease may be masked by slowed metabolism and decreased ischemic stress, but coronary disease is accelerated by elevations in cholesterol and blood pressure.62,71 Pericardial effusions may be seen in chronic hypothyroidism, but are usually small and asymptomatic. Larger effusions may result in diminished heart sounds and decreased apical impulse, but cardiac tamponade is rarely seen due to slow chronic build up.72,73 Complaints of fatigue, dyspnea, and decreased exercise capacity are most likely to be from a respiratory origin, rather than cardiac. Hypothyroidism is characterized by impaired ventilator responses to hypercapnea and hypoxia, as well as myopathy of respiratory musculature, with resultant slow, shallow respirations. Macroglossia may contribute to the respiratory distress and lead to obstructive sleep apnea. Mucopolysaccharide infiltration or edema of the vocal cords leads to a deep, husky voice in the hypothyroid patient. Primary pulmonary hypertension is reported with increased prevalence in hypothyroidism and may contribute to complaints of dyspnea or chest pain.74,75 A modest weight gain is characteristic of hypothyroidism, but massive obesity is unusual. Limiting the anticipated weight gain in this hypometabolic state is a concomitant decrease in appetite.62 Neurocognitive impairment may be a presenting feature of hypothyroidism, especially in elders. Slowness of comprehension, lethargy, decreased attention span, poor short-term memory, and impaired abstract thinking may all be present. The patient generally appears placid or depressed, moves slowly and deliberately, and speaks hesitantly.76,77 Although unusual in hypothyroidism, extreme agitation, psychosis, and even seizures have been described—the last is referred to as myxedema madness or Hashimoto’s encephalopathy.42,78 Paresthesias are common in hypothyroidism and although peripheral polyneuropathy may occur, mononeuropathies are much more prevalent. Edema of perineural and synovial tissue within the carpal tunnel leads to carpal tunnel syndrome, which is reported in about 25% of hypothyroid patients. Another common mononeuropathy in hypothyroidism involves the eighth cranial nerve, resulting in sensorineural hearing loss and tinnitus.77 Muscle-related symptoms are frequent in hypothyroid patients, often manifesting with proximal muscle weakness, myalgias, stiffness, and fatigue. Hypothyroid myopathy leads to slowing of the relaxation phase of deep tendon reflexes, referred to as hung-up or pseudomyotonic reflexes. This reflex phenomenon is best demonstrated with the Achilles tendon reflex performed while the patient is kneeling on a chair. Hung-up reflexes are not unique to hypothyroidism, as they can be seen with aging, diabetes, and pregnancy. Muscle palpation in most patients is normal, but some with hypothyroid myopathy develop firm, enlarged muscles referred to as pseudohypertrophy. Prolonged mounding of muscle mass may sometimes be evident when reflexes are elicited (myoedema). Ataxia and dysmetria reversible with thyroid replacement have been described, which may by myopathic or cerebellar in origin.77,79 Elevation of the serum creatine kinase (CK) concentration is seen in 70 to 90% of patients with hypothyroidism; however, the magnitude of the CK elevation does not correlate well with the severity of the patient’s myopathy or hypothyroidism. Rare cases of acute rhabdomyolysis have been reported in severe
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1668
hypothyroidism, precipitated by exercise, statin therapy, or renal failure.80 One of the most common complaints of the hypothyroid patient is constipation, which results from decreased bowel motility. Rarely, ileus or megacolon may occur and be confused with intestinal obstruction. Although oligo- or amenorrhea may suggest a pituitary/ hypothalamic origin, primary hypothyroidism can result in menstrual abnormalities due to altered metabolism of estrogen. Menorrhagia is also described. Decreased fertility and early abortions are commonly associated complaints. Hypothyroid men often report decreased libido, erectile dysfunction, and delayed ejaculation.62,63 Initial symptoms of hypothyroidism may be rheumatic in character. Arthralgias and stiffness, sometimes associated with noninflammatory joint effusions, may be seen. Acute monoarthritis may be seen due to an increased prevalence of hyperuricemia and gout in hypothyroid patients, and perhaps as a result of a minor link with pseudogout and chondrocalcinosis as well81,82 (Box 126-8).
Myxedema coma is a life-threatening decompensation of severe long-standing hypothyroidism, often precipitated by an acute illness or stress. The hallmarks of myxedema coma are altered mental status and hypothermia, but hypotension, bradycardia, and hypoventilation are often present as well. The typical patient is an elderly woman with chronic hypothyroidism that is untreated or unrecognized. Hashimoto’s thyroiditis is the most common underlying thyroid pathology due to its insidious nature, but any cause of hypothyroidism is possible. The history is one of progressive weakness, lethargy, and immobility that may progress to shock and death. Virtually any acute illness may precipitate myxedema coma, but the most common factors include infection, cold exposure, trauma, cerebrovascular accident, congestive heart failure, gastrointestinal bleeding, and drug effects. Sedatives and narcotics are commonly implicated drug classes, but general anesthesia, thyroid hormone noncompliance, amiodarone, lithium, iodides, phenytoin, and rifampin may be factors33,63,75,83-85 (Box 126-9).
BOX 126-8 Symptoms and Signs of Hypothyroidism Vital Signs Systolic BP—normal or low Diastolic BP—normal or elevated Slow pulse to sinus bradycardia Respirations—normal or slow, shallow Temperature—normal, but prone to hypothermia with stress
Sinus bradycardia Long QT with increased ventricular arrhythmia Chest pain—accelerated coronary disease Diastolic heart failure (delayed ventricular relaxation) Pericardial effusion (asymptomatic) Peripheral edema
Hypometabolic Complaints Cold intolerance Fatigue Weight gain, but decreased appetite
Respiratory Dyspnea on exertion Obstructive sleep apnea Primary pulmonary hypertension
Cutaneous Coarse, brittle hair Alopecia Dry skin, decreased perspiration Pallor, cool hands and feet Coarse, rough skin Yellow tinge from carotenemia Thin, brittle nails Lateral thinning of the eyebrows
Gastrointestinal Constipation Ileus Gastric atrophy
Neurologic Slow mentation and speech Impaired concentrating ability and attention span Lethargy Decreased short-term memory Agitation, psychosis Seizures Ataxia, dysmetria Mononeuropathy Carpal tunnel syndrome Sensorineural hearing loss Peripheral neuropathy Muscular Proximal myopathy Pseudohypertrophy Delayed relaxation of reflexes (hung-up or pseudomyotonic) Cardiac Decreased exercise capacity Dyspnea on exertion BP, blood pressure; HEENT, head, ear, eyes, nose, and throat.
Reproductive Oligo- and amenorrhea Menorrhagia Decreased fertility Early abortions Decreased libido Erectile dysfunction Rheumatic Polyarthralgias Joint effusions Acute gout or pseudogout HEENT Hoarseness Deep, husky voice Macroglossia Hearing loss Periorbital swelling Broad nose Swollen lips Goiter
1669
Myxedema Coma: Aggravating or
Infection/sepsis (especially pneumonia) Exposure to cold Cerebrovascular accident Drug effect Altered sensorium: Sedative-hypnotics, narcotics, anesthesia, neuroleptics Decrease T4 and T3 release: amiodarone, lithium, iodides Enhance elimination of T4 and T3: phenytion, rifampin Inadequate thyroid hormone replacement: Noncompliance; interference with absorption (iron, calcium, cholestyramine) Myocardial infarction Gastrointestinal bleeding Trauma/burns Congestive heart failure Hypoxia Hypercapnia Hyponatremia Hypoglycemia Hypercalcemia Diabetic ketoacidosis T3, triiodothyronine; T4, thyroxine.
It is important to appreciate that although infection is the most common precipitant of myxedema coma, the patient does not usually mount a febrile response due to the profound hypometabolic state. Peripheral vasoconstriction helps maintain core temperature in severe hypothyroidism, but this compensation may be tenuous. Hypothermia is characteristic of myxedema coma and it is usually marked, with body temperatures as low as 24° C (75° F) described. Temperatures below 90° F are common and carry a grave prognosis. In contrast to environmental hypothermia, shivering is often absent.63,75,85 Whereas diastolic hypertension is common in hypothyroidism, the blood pressure in myxedema coma is usually low and may be refractory to fluid resuscitation and pressors unless thyroid hormone is administered.75 Sinus bradycardia is routinely seen and may be unresponsive to atropine, but heart block is unusual. Prolongation of the QT interval with cases of torsades de pointes is described.70 Despite cardiac enlargement and decreased myocardial contractility seen in most patients, evidence of congestive heart failure is not common.62 Signs of severe hypothyroidism are usually evident in the myxedema coma patient. Dry, coarse skin; sparse, brittle hair; cool extremities; puffy eyelids and face; large tongue; hoarse voice; and slow, delayed speech and movement are typical. Pitting and nonpitting edema are prominent in the extremities.68 Pleural and pericaridial effusions are common, and ascites may be present.86 Despite the name myxedema coma, most patients present with confusion, lethargy, or stupor and are not comatose, but progression to coma is inexorable if therapy is not instituted. The cause of altered mental status is multifactorial, including thyroid hormone deficiency, hypothermia, hypercapnea, hyponatremia, hypotension, and hypoglycemia.75,83 Paradoxically, a more agitated, psychotic state may sometimes occur, referred to as myxedema madness.77 Focal or generalized seizures may occur in up to 25% of patients, and status epilepticus has been reported.78 Respiratory depression with carbon dioxide retention is common in myxedema coma, contributing further to the
BOX 126-10 Recognition of Myxedema Coma Patient profile: Elderly female in the winter Known hypothyroidism; thyroidectomy scar Hypothermia: Usually below 95.9° F; below 90° F is bad prognostic sign; as low as 75° F reported. Near normal in presence of infection Altered mental status: Lethargy and confusion to stupor and coma, agitation, psychosis and seizures (myxedema madness) Hypotension: Refractory to volume resuscitation and pressors unless thyroid hormone administered Slow, shallow respirations with hypercapnea and hypoxia; high risk of respiratory failure Bradycardia (sinus)/long QT and ventricular arrhythmias Myxedema facies: Puffy eyelids and lips, large tongue, broad nose Evidence of severe chronic hypothyroidism: Skin, hair, reflexes, bradykinesis, voice Acute precipitating illness (e.g., pneumonia) Drug toxicity (e.g., sedative, narcotic, neuroleptic) Hyponatremia
altered mental status. An enlarged tongue, supraglottic edema, and obesity further aggravate hypercapnea and hypoxia. Pneumonia, a common precipitant of myxedema coma, accelerates the downward spiral toward respiratory failure. Mechanically assisted ventilation is required in most patients, often leading to extended hospitalization74,85 (Box 126-10).
Diagnostic Strategies The diagnosis of primary hypothyroidism requires the measurement of an elevated serum TSH and a depressed free T4 level. If the index of suspicion is low, the serum TSH alone can be used as a screening test, and the free T4 test only performed if the TSH is abnormal. The total T4 level is not recommended in evaluating thyroid disease because of numerous confounding factors on protein binding that alter its measurement, thus the free T4 is preferred. The serum T3, whether total or free, should not be relied on in the diagnosis hypothyroidism due to its great variability. Almost any acute or chronic illness or physiologic stress can lead to depression of T4 5′deiodinase activity, thereby leading to decreased peripheral conversion of T4 to T3 and an increase in reverse T3 levels.7 The elevation of TSH accompanied by a normal free T4 is referred to as subclinical hypothyroidism (SCH). The prevalence of SCH is 4 to 9% in large general population screening surveys and in 7 to 26% in geriatric series.1,2,62,87 Although by definition SCH is asymptomatic, this laboratory abnormality is associated with the development of depression, cognitive impairment, subtle systolic and diastolic dysfunction, and hyperlipidemia.88 The distinction of myxedema coma from moderate to severe hypothyroidism cannot be made by thyroid functions tests alone. Elevation of the serum TSH may be blunted by any concomitant systemic illness (hypothyroid sick syndrome), resulting in a misleadingly minor elevation in TSH in a severely hypothyroid patient. Central hypothyroidism is characterized by a low serum TSH together with a low free T4 level. The euthyroid sick syndrome can have very similar results but the TSH is only mildly suppressed and the free T4 is normal or slightly low. Because central hypothyroidism is usually not a chronic condition, many of the clinical findings of hypothyroidism are absent.67,75,88
Chapter 126 / Thyroid and Adrenal Disorders
BOX 126-9 Precipitating Factors
PART III ■ Medicine and Surgery / Section Eleven • Metabolism and Endocrinology
1670
Hyponatremia is a common electrolyte abnormality in severe hypothyroidism and myxedema coma, but it is not seen in milder forms of the disease. Hyponatremia results from decreased free water clearance due to diminished renal blood flow from volume depletion and depressed cardiac output as well as excess excretion of anti-diuretic hormone. A reversible elevation in serum creatinine is also seen in this setting.62,75,84 Hypoglycemia may also occur in severe hypothyroidism and myxedema coma. Decreased gluconeogenesis and reduced insulin clearance are the likely mechanisms, but it is important to recognize that a low blood sugar may be a clue of concomitant adrenal insufficiency, which is present in up to 10% of myxedema coma patients. CK from a muscular origin is often elevated in severe hypothyroidism, but acute rhabdomyolysis is uncommon.80 Serum transaminases and lactate dehydrogenase are frequently elevated as well. Lipid clearance is decreased in hypothyroidism, resulting in elevations of total cholesterol, low density lipoprotein, and triglyceride. In the Colorado Thyroid Disease Prevalence Study of 25,862 patients, euthyroid patients averaged a total cholesterol of 214 mg/dL, whereas subclinical and overt hypothyroid subjects averaged 224 and 251 mg/dL, respectively.1-3 Severe hypothyroidism may be associated with a normocytic, normochromic anemia from decreased red cell production and a depressed white blood count that does not rise appropriately in response to infection.75 Increased bleeding times may result from an acquired von Willebrand’s syndrome. A chest radiograph may reveal an enlarged cardiac silhouette, but evidence of heart failure is unusual. Although myocardial hypertrophy may be present in hypothyroidism, cardiomegaly usually represents the presence of pericardial effusion, seen in 30 to 78% of severe, chronic disease, but less than 5% of mild hypothyroidism. Pleural effusions may be seen as well.62,72,73,75 Electrocardiographic findings include sinus bradycardia, nonspecific ST and T wave abnormalities, and decreased voltage or electrical alternans if a pericardial effusion is present. Prolongation of the QT interval and ventricular arrhythmias may be seen as well.69,70,75 If a lumbar puncture is performed to evaluate altered mental status, findings of increased opening pressure and an elevated cerebrospinal fluid protein level may be seen in myxedema coma.
Differential Considerations The overtly hypothyroid patient is often thought to be severely depressed, and the diagnosis of hypothyroidism may be overlooked. In fact, about 10 to 15% of patients hospitalized for depression are found to be hypothyroid. Roughly 25% of bipolar patients with a rapid cycling pattern are found to be hypothyroid. Even subclinical hypothyroid patients have more than twice the incidence of depression than euthyroid patients.76 The profound fatigue and weakness that may occur in hypothyroidism may be diagnosed as depression, chronic fatigue syndrome, Addison’s disease, or anemia. Respiratory failure in patients with substrates of obstructive sleep apnea such as obesity, hypoventilation, and macroglossia, should be evaluated for hypothyroidism. In addition, any patient with decreasing exercise capacity and dyspnea without any clear cardiopulmonary cause should be evaluated for hypothyroidism. The hypothermia of myxedema coma may be attributed to environmental stress, sepsis, or hypoglycemia. The altered mental status, as well, may be attributed to concomitant condi-
tions in myxedema coma, such as drug toxicity, hypothermia, hypercapnea, hypoxia, hypoglycemia, or hyponatremia, and the hypothyroidism overlooked.
Management The patient with a new diagnosis of overt or subclinical hypothyroidism generally does not require the initiation of treatment from the emergency department, yet it is important to be familiar with the principles of diagnosis and treatment. Before commitment to lifelong therapy, the serum TSH and free T4 should be repeated for confirmation. About 5% of patients with subclinical hypothyroidism normalize within 1 year and, of the remainder, approximately 5% per year develop overt hypothyroidism. Treatment with thyroid hormone replacement may be considered in SCH if symptomatic or the TSH is greater than 10.62,64,88 Levothyroxine (T4) is the mainstay of treatment of hypothyroidism. Treatment is generally started at a dose of 1.6 µg/kg/day in younger patients, whereas elders and patients with underlying coronary artery disease are often started at less than half that dosage due to their susceptibility to angina and arrhythmias. The long half-life of T4 (7 days) and its gradual conversion to T3, leads to dosage adjustments of 12.5- to 25-µg increments at no less than 6-week intervals.62,63 The management of myxedema coma requires immediate attention to airway management, fluid resuscitation, thyroid hormone replacement, general supportive measures, and treatment of the precipitating illness (Box 126-11).
BOX 126-11 Treatment of Myxedema Coma Protect the airway/ventilatory support; monitor for alkalosis Fluid resuscitation: 0.9NS or D5/0.9NS if hypoglycemia Watch for unmasking of CHF Thyroid hormone replacement: T4 alone (elderly and patients with cardiac comorbidity): T4 300–500 µg IV as initial bolus Or split bolus 200–300 µg IV day 1 and 2 Then 50–100 µg IV daily until able to take PO T3 alone (younger patient, no cardiac risks; rapid correction desired): T3 10–20 µg IV initially, then 10 µg IV every 4 hr for 1 day, then 10 µg IV every 6 hr for 1–2 days Combination T4 and T3 therapy (intermediate approach): T4 200–250 µg IV as initial bolus T3 10 µg IV initial dose, then 10 µg IV every 8–12 hr T4 100 µg IV in 24 hr, followed by 50 µg/day Hydrocortisone 50−100 mg IV every 6−8 hr Hyponatremia Avoid hypotonic fluids, use only 0.9NS or D5/0.9NS If less than 120 mEq/L, consider 3% saline, 50–100 mL boluses Passive rewarming Regular blankets, prevent heat loss If heating blankets considered, pretreat with IV fluids and monitor BP closely Avoid mechanical stimulation Treatment of any precipitating illness, with special attention to infectious causes BP, blood pressure; CHF, congestive heart failure; D5/0.9NS, 5% dextrose in 0.9% normal saline; IV, intravenous; T3, triiodothyronine; T4, thyroxine.
1671
Airway Management
Fluid Resuscitation Intravascular volume depletion is prominent in myxedema coma, even in the presence of normal vital signs. Fluid resuscitation should be started immediately, but the aggressiveness of administration should be tempered by the risk of unmasking congestive heart failure. The initial fluid of choice is D5/ 0.9NS because the myxedema coma patient is at high risk for both hyponatremia and hypoglycemia.75,83-85
Thyroid Hormone Replacement Prompt thyroid hormone replacement is critical for patient survival from myxedema coma, although the most effective regimen is unclear. Determination of the form of thyroid hormone (T4 or T3 or both) and the dosage must balance the high mortality of untreated myxedema coma against the risk of myocardial infarction or cardiac arrhythmias induced by therapy. T4 has lower risk of toxicity as its action depends on peripheral conversion to T3, which is a slow, delayed process. Proponents of T3 administration suggest that the quicker onset of action and the increased biologic activity of T3, as well as the impaired conversion of T4 to T3 in the critically ill, make T3 the logical choice. High doses of T4 or T3 appear to increase mortality, hence there is a limit on how fast hormone replacement can be given.33,63,75,83-85 The most widely published approach to myxedema coma involves the IV administration of T4 in a dose of 300 to 500 µg that depends on patient weight and cardiac risks. It is suggested that this dose replaces total body stores of T4, and the pool can be maintained by 50 to 100 µg/day. Some authors suggest splitting the loading dose over 2 days.75,83,84 For critically ill younger patients without cardiac disease, where a more rapid correction of hormone levels is desired, the use of T3 alone should be considered. An IV loading dose of 10 to 20 µg, followed by 10 µg IV every 4 hours for 24 hours, followed by 10 µg every 6 hours for 1 to 2 days is suggested.75 An intermediate approach using both T4 and T3 is suggested by some authors to speed clinical response while minimizing cardiac toxicity. T4 is administered IV in half the loading dose (200–250 µg) with 10 µg of T3, followed by T3, 10 µg every 8 to 12 hours, and maintenance T4, 50 µg every 24 hours. Due to impaired oral absorption and transit of medications in myxedema coma, the IV route is recommended until the patient is alert and able to tolerate oral intake, and then maintenance T4 alone is continued.75,83
General Supportive Measures The management of hyponatremia in myxedema coma requires water restriction, but in the face of the volume depletion and hypotension seen in myxedema coma, normal saline solutions should be used. If hyponatremia is severe ( 2 mg/dL (177 µmol/L) for adults or more than twice the upper limit of normal for age or more than twofold elevation over baseline for patients with preexisting renal disease Hematologic: Platelets ≤ 100,000/mm3 or DIC defined as prolonged clotting times, low fibrinogen level, and the presence of FDPs Hepatic: Total bilirubin, AST, and ALT at least twice the upper limit of normal for laboratory, or a twofold increase in patients with preexisting liver disease Acute respiratory distress syndrome: Acute onset of pulmonary infiltrates and hypoxemia in the absence of cardiac failure or by evidence of diffuse capillary leak manifested by acute onset of generalized edema, or pleural or peritoneal effusions with hypoalbuminemia Generalized erythematous maculopapular rash that may desquamate Soft tissue necrosis, including necrotizing fasciitis, myositis, or gangrene Laboratory criteria for diagnosis: Isolation of GAS
ALT, serum alanine aminotransferase; AST, serum aspartate aminotransferase; BUN, blood urea nitrogen; CNS, central nervous system; CSF, cerebrospinal fluid.
childbirth. Nonmenstrual TSS occurs in people of all ages and in both sexes. The CDC reported an average of about 200 cases a year (≈ 1 case per 100,000 population) from 1994 to 2001 with a steady increase in the incidence of streptococcal TSS and a slight decrease in the incidence of staphylococcal TSS.91 The age and sex distribution reflects the association with menses. In 2003, 294 cases were reported, but a steady decline continues to occur, with 182 cases reported in 2007. Streptococcal TSS accounts for a little more than half of the cases.92 Nonmenstrual staphylococcal TSS is associated with superinfection of various skin lesions, including burns, surgical sites, dialysis catheters, and lung (influenza associated). It may also occur in association with staphylococcal respiratory infections or even with colonization by a toxigenic strain of the organism, without an obvious infectious source. Streptococcal TSS is classically associated with more severe soft tissue infections such as necrotizing fasciitis and myositis as well as pneumonia, peritonitis, myometritis, and osteomyelitis.84–87 The mortality rate from staphylococcal TSS has declined since the disease was first described. The case fatality rate in 1980 was 10%, and it was less than 3% in the past several years. Streptococcal TSS remains a highly fatal disease, with a mortality rate of 30 to 70%.84–87
Case Classification Probable: A case meets the clinical case definition in the absence of another identified cause for the illness and with isolation of GAS from a nonsterile site. Confirmed: A case meets the clinical case definition with isolation of GAS from a normally sterile site (e.g., CSF or joint, pleural, or pericardial fluid). ALT, serum alanine aminotransferase; AST, serum aspartate aminotransferase; CSF, cerebrospinal fluid; DIC, disseminated intravascular coagulation; GAS, group A Streptococcus.
Principles of Disease Etiology Staphylococcal TSS is caused by colonization or infection with toxigenic strains of Staphylococcus aureus. This strain produces toxic shock syndrome toxin-1 (TSST-1). S. aureus has been detected in virtually all cases of both forms of the illness. S. aureus has been isolated from the vagina or cervix in 98% of women with menstrual TSS, compared with a colonization rate of less than 10% of unaffected women. Because the organism is often not invasive, the blood cultures are often negative. Streptococcal TSS is caused by invasive infection with toxigenic strains of GAS.84–87
Pathophysiology The shock and multiorgan dysfunction associated with TSS are caused by the effects of various exotoxins produced by S. aureus and GAS. S. aureus produces TSST-1 and enterotoxin B. TSST-1 is identified in more than 90% of menstrual cases and 60% of nonmenstrual cases. Other toxins may play a role in nonmenstrual TSS. Antibodies to these toxins are protective against disease. GAS produces streptococcal pyrogenic exotoxins A (SPEA) and B (SPEB). These exotoxins are
Chapter 127 / Bacteria
BOX 127-6 Syndrome (Revised)
PART III ■ Medicine and Surgery / Section Eleven • Infectious Diseases
1698
absorbed into the bloodstream through inflamed or traumatized mucous membranes or from areas of focal infection. Absorbed toxins act as superantigens, inducing mononuclear cells to synthesize and release cytokines, tumor necrosis factor α, and interleukins, which begin the cascade of systemic vasculitis and the multisystem manifestations of the disease. Host immune factors are important in the pathogenesis of TSS. GAS is an invasive organism, and circulating GAS organisms induce the production of tumor necrosis factor α and other cytokines by mononuclear cells.84–87,93
eral edema. Vaginal mucosal erythema and purulent vaginal discharge may be present in menstrual TSS but are not required to make the diagnosis. As multiple organ systems become involved, a wide constellation of signs and symptoms may be seen. Gastrointestinal involvement is manifested by vomiting, diarrhea, and severe abdominal pain. Hepatomegaly may be present. Patients may become hypoxic and develop rales on pulmonary examination. Comparisons between staphylococcal and streptococcal TSS are presented in Table 127-4.
Complications Clinical Features Symptoms and Sings The clinical presentations of streptococcal TSS and staphylococcal TSS are similar. The primary difference is that an identifiable infectious source is virtually always present with streptococcal TSS, and colonization alone may be the source in staphylococcal TSS. Patients may have fever, chills, nausea, vomiting, watery diarrhea, headache, myalgias, and pharyngitis. This prodromal illness may last for 2 to 3 days before progression to frank sepsis and organ dysfunction. Others patients may become abruptly symptomatic within hours. Rapid progression is more typical of streptococcal TSS. Patients may complain of pain at a site of infection more often with streptococcal TSS. Risk factors for TSS are listed in Box 127-8. The fever is usually high and abrupt in onset, although septic patients may have hypothermia on presentation. The classic rash is a nonpruritic, diffuse, blanching, macular erythroderma. It develops over the first few days of the illness and initially may be faint, evanescent, and mistaken for the flush associated with a fever. The rash is usually diffuse but may be localized to the trunk, extremities, or perineum. After about a week, a fine flaking desquamation occurs on the face, trunk, and extremities, followed by full-thickness peeling of the palms, soles, and fingers. This classic rash progression is much more common in staphylococcal TSS and is present in less than 10% of patients with streptococcal TSS.93 The patient’s mental status is frequently abnormal, out of proportion to the degree of hypotension. Confusion, somnolence, agitation, and combativeness are present in 55% of patients with streptococcal TSS and in even more patients with staphylococcal TSS.84,85,93 Other findings on physical examination may include pharyngeal and conjunctival erythema, strawberry tongue, and periph-
Complications of TSS include acute respiratory distress syndrome, shock, gangrene, DIC, renal failure, and a constellation of neuropsychiatric symptoms. Less common findings in staphylococcal TSS include rhabdomyolysis, seizures, pancreatitis, pericarditis, and cardiomyopathy. Women with the menstrual form of TSS may experience one or more recurrent episodes; recurrences of the nonmenstrual form are rare. Complication rates are higher with streptococcal TSS. Rhabdomyolysis occurs in up to 63% of patients with streptococcal TSS and is usually related to the underlying soft tissue infections.84–87,93
Diagnostic Strategies The case definition for TSS does not require a positive culture for S. aureus but does for Streptococcus organisms. These case definitions (see Boxes 127-6 and 127-7) are useful to the clinician, but they are neither specific nor foolproof. Specific tests are not required to exclude other diseases, but if such tests are obtained, the results of these studies must be negative. No specific laboratory changes are associated with TSS, but many abnormalities are common. Either leukocytosis or leukopenia can occur, but a marked bandemia is very common. Elevated creatinine levels and hemoglobinuria occur in most patients. Laboratory evidence of renal dysfunction occurs prior to hypotension in half of the patients. Hypoalbuminemia and life-threatening hypocalcemia are prominent initially and persist throughout the course of the disease. Other abnormalities include anemia, thrombocytopenia, hyperbilirubinemia, elevated transaminase levels, and sterile pyuria.84–87 Chest radiography may reveal evidence of acute respiratory distress syndrome or a pulmonary source of the organism. Plain radiographs of any infected skin or soft tissue site typically of Staphylococcal and Table 127-4 Comparison Streptococcal Toxic Shock Syndrome
BOX 127-8 Risk Factors for Toxic Shock Syndrome Use of superabsorbent tampons Postoperative wound infections Postpartum period Nasal packing Cancer Common bacterial infections Ethanol abuse Infection with influenza A Infection with varicella Diabetes mellitus Human immunodeficiency virus infection Chronic cardiac disease Chronic pulmonary disease Nonsteroidal anti-inflammatory use (may mask symptoms rather than be a risk factor)
FEATURE
STAPHYLOCOCCAL
STREPTOCOCCAL
Age Sex Severe pain Hypotension Erythroderma rash Renal failure Bacteremia Tissue necrosis Predisposing factors
Primarily 15–35 yr Greatest in women Rare 100% Very common Common Low Rare Tampons, packing, NSAID use?
Thrombocytopenia Mortality rate
Common 1 month) Histoplasmosis, disseminated or extrapulmonary Isosporiasis, chronic intestinal (>1 month’s duration) Kaposi’s sarcoma Lymphoid interstitial pneumonia or pulmonary lymphoid hyperplasia complex Lymphoma, Burkitt (or equivalent term) Lymphoma, primary, of brain Mycobacterium avium complex or Mycobacterium kansasii, disseminated or extrapulmonary Mycobacterium tuberculosis of any site, pulmonary, disseminated, or extrapulmonary Mycobacterium, other species of unidentified species, disseminated or extrapulmonary Pneumocystis jiroveci pneumonia Pneumonia, recurrent Progressive multifocal leukoencephalopathy Salmonella septicemia, recurrent Toxoplasmosis of brain, onset at age >1 month
200
Year
Figure 130-1. Proportions of AIDS cases among adults and adolescents by exposure category and year of diagnosis, 1985 to 2001, United States. (From Centers for Disease Control and Prevention: AIDS Surveillance— Trends [1985–2006]. Available at: http://www.cdc.gov/hiv/graphics/ trends.htm.)
■ PRINCIPLES OF DISEASE Pathophysiology HIV is a cytopathic human retrovirus that belongs to the lentivirus subfamily. The two major subtypes of HIV are HIV-1 and HIV-2. HIV-1 is the predominant subtype worldwide and is the cause of AIDS. HIV-2 causes a similar immune syndrome but is rarely seen in the United States being restricted primarily to western Africa. The HIV virion is composed of a central single-stranded RNA molecule and the enzyme reverse transcriptase. These are surrounded by a core protein and a lipid bilayer envelope that contain virally encoded transmembrane proteins critical for recognition and attachment to target host lymphocytes (predominantly CD4+ cells). HIV-1 has been isolated from a variety of body fluids including blood, serum, semen, vaginal secretions, urine, cerebrospinal fluid (CSF), tears, breast milk, bone marrow, alveolar fluid, synovial fluid, amniotic fluid, and saliva. Only a few modes of transmission have been proved: in semen, vaginal secretions, blood or blood products, and breast milk and transplacental transmission in utero. There have been no instances of transmission by casual contact, although one case report described possible salivary transmission. The HIV virion is extremely labile and easily neutralized by heat and common disinfecting agents such as 50% ethanol, 35% isopropyl alcohol, 0.3% hydrogen peroxide, disinfectant (Lysol), or a 1 : 10 solution of household bleach (sodium hypochlorite). HIV selectively attacks cells within the immune system (primarily TH4 helper cells, but macrophages and monocytes also may be involved), a characteristic that accounts for much of the immunodeficiency it produces in affected persons. HIV1 transmembrane proteins gp41 and gp120 play a critical role in recognition and attachment of HIV virions to receptors on host lymphocytes. After infection, viral RNA is reversetranscribed into deoxyribonucleic acid (DNA) by reverse transcriptase, one of the critical enzymes required for HIV replication. The viral genome thus becomes permanently integrated into the host’s genome. Once integrated, retroviral DNA may lie dormant, or it may be actively transcribed and translated to produce virally encoded proteins and new HIV virions. HIV protease is another critical retroviral enzyme in the life cycle of the virus, responsible for activation of viral protein precursors into the functional enzymes required for virion infectivity.
Chapter 130 / AIDS and HIV Infection
Laboratory-confirmed evidence of HIV infection and
225
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1734
Primary HIV exposure is characterized by a transient viremia and a decrease in CD4+ cell counts, followed by establishment of equilibrium between virus and host immunity. A persistent latent period, during which the virus lies dormant in the host genome, can last for years. The “set point” or steady-state viral load level in the blood of the patient allows prediction of longterm clinical outcomes. Lower levels of viremia correlate with longer clinical latency periods. In the later stages of HIV infection, a sudden increase of viremia correlates with a dramatic decrease in CD4+ T lymphocytes. These hematologic changes are followed by the appearance of opportunistic infections or malignancies and ultimately death. HIV-1 is highly heterogeneous. Multiple genetic subtypes exist in a variety of geographic and sociologic settings. Further genetic diversity exists within individual hosts owing to the highly mutable character of the virus. High error rates, which are associated with reverse transcription, ensure extensive viral diversity, a critical factor in the pathogenesis of infection and ongoing emergence of drug-resistant phenotypes.6
Tests for Human Immunodeficiency Virus Infection General Approach HIV infection most commonly is established by HIV serologic studies or by detection of antibodies to the virus. Testing involves sequential use of an enzyme immunoassay (EIA) and a Western blot assay. Criteria for positive results are positive results on EIA followed by Western blot assay. EIA detects the binding of specific serum antibodies to HIV antigens that are adherent to a microtiter plate. Western blot assay detects electrophoretically separated viral antigens in the patient’s serum. A positive Western blot result requires detection of two of the following: p24, gp41, or gp120/160. Final HIV serology results are reported as positive, negative, or indeterminate. Overall sensitivity and specificity rates for HIV serologic testing are greater than 99.9%. False-negative HIV test results are accounted for primarily by testing too early, during the “window period” (usually the first several months) of acute infection, after viral transmission but before the appearance of antibodies. Rates of falsenegative testing range from 0.3% in high-prevalence populations to less than 0.001% in low-prevalence populations. Ninety-five percent of false-negative test results become positive by 3 months and 98% by 6 months. A less common explanation for false-negative results is seroreversion, which may occur in late-stage disease or in patients on HAART regimens, or in those harboring atypical strains of HIV-1 or with HIV-2 infection. False-positive test results are exceedingly rare (with a frequency of less than 0.0004%); they may occur in several clinical populations, including (1) recipients of transfused blood containing the HIV antibody; (2) children younger than 6 months of age, in whom a positive test result may be caused by transplacentally acquired antibodies; (3) patients with cross-reacting antibodies (e.g., antihepatitis A immunoglobulin M [IgM], antihepatitis B core IgM, antinuclear, anti-smooth muscle, anti-parietal cell, and antimitochondrial antibodies); and (4) patients with cross-reactive human lymphocyte antigens (HLAs) from the H9 cell line or other human retroviruses. In populations with a low prevalence of truepositive results (e.g., heterosexual men or women in lowseroprevalence areas), the frequency of false-positive results (on both EIA and Western blot assay) is increased. This finding has been cited as one of the reasons for not offering indiscriminate HIV screening in the ED.
Indeterminate results are most common with a positive EIA result and a single band (rather than two or three) on Western blot assay. In evaluating an indeterminate result on the latter, the patient’s risk profile should be assessed. Low-risk patients with indeterminate results are rarely infected with HIV-1 or HIV-2, and repeat testing usually shows persistence of one band, with the cause rarely established. Referral to an infectious disease specialist and follow-up serologic testing at 3 months are indicated. Patients in higher-risk groups with indeterminate results usually are found to have definitively positive results on Western blot assay 3 or 6 months later and therefore should be counseled to follow appropriate risk-reduction behavior until follow-up definitive testing is completed. Other methods for detection of HIV infection include detection of virus-specific antigens and assays for HIV nucleic acid. Neither of these techniques is considered superior to routine serologic tests in terms of accuracy, and they should be used only in patients with confusing serologic results requiring clarification; in the ED, the most common situation in which viral detection testing may be considered is in cases of suspected acute retroviral infection. Quantitative plasma HIV RNA assay is most commonly employed and is routinely used for HIV staging and monitoring of response to retroviral therapy. Test results are reported as copies per milliliter, with survival time directly correlated with viral burden. Sensitivity of the various viral detection methods varies with stages of diseases but generally is greater than 99% for DNA polymerase chain reaction (PCR) assay, 90 to 95% for quantitative HIV RNA assay, and 95 to 100% for viral culture of peripheral blood mononuclear cells. The U.S. Food and Drug Administration (FDA) has approved a number of rapid HIV tests. These tests are easy to use and can be performed at the bedside or in a nearby satellite laboratory using saliva or fingerstick or whole blood specimens. Delivery of results is possible in as little as 10 to 20 minutes. Sensitivity and specificity is approximately 99%. Negative test results can be reported immediately as negative; reactive test results require confirmatory follow-up testing with a Western blot assay and should be reported as reactive. The first test to receive FDA approval is the OraQuick Rapid HIV-1 Antibody Test (OraSure Technologies, Inc., Bethlehem, Pennsylvania). Advantages of rapid test technologies include ease of specimen collection, reduced costs, rapid availability of results, and improved compliance with testing.
Human Immunodeficiency Virus Testing in the Emergency Department Traditionally it has been thought that serologic testing of patients for HIV infection in the ED is not indicated. With the advent of rapid testing methodologies, the concept of ED testing for HIV infection is being reexamined. This is being driven by a number of factors, including the ease of the testing process, the recognized value of knowing a patient’s HIV status if the acute clinical presentation raises suspicion for disease, and the now-widespread recognition that early detection of HIV infection (and early therapeutic intervention) provides a significant health benefit both for individual patients and for the community. Benefits to the infected patient include delaying progression of disease and reducing the risk of opportunistic infections; an important advantage for the community includes decreased disease transmission associated with reduction in high-risk behavior in persons who are aware of their HIV serostatus. The most recent CDC guidelines for HIV testing, released in 2007, recommend that such testing be performed in health care settings.7 These guidelines give special emphasis to the
■ CLINICAL FEATURES The broad spectrum of disease presentation for HIV-related disorders ranges widely from asymptomatic seropositive status to severe, life-threatening complications of AIDS. Included are a wide variety of opportunistic infections, malignancies, and other HIV-related diseases. Nearly every organ system may be affected by HIV infection and related conditions. Because the differential diagnosis is so broad in scope for many ED presentations, this chapter addresses clinical symptoms, signs, and focused information on some of the more common disorders.
Initial Evaluation of the Human Immunodeficiency Virus–Infected Patient The initial evaluation and management of the HIV-infected patient consist of rapid and early assessment of stability. Any problems with airway, breathing, and circulation must be promptly identified and appropriate interventions performed. For unstable patients, intravenous access, cardiac monitoring, and administration of oxygen typically are indicated. After initial stabilization, the remainder of the history and physical examination may be conducted. Relevant elements of the history include information pertinent to the chief complaint, including duration, location, quali-
1735
ties, characteristics, level of distress, and relieving or inciting factors. The past medical history should identify a previous history of similar problems, the time of diagnosis of HIV infection, previous AIDS-defining conditions, recent hospitalizations, past surgical history, current medications, and allergies. The existence of an advance directive may be important historical information, because many HIV-infected patients have expressed opinions about the level of intervention desired in various clinical settings, particularly in critical care settings and at the end of life. Information regarding potential risk factors for HIV infection may be appropriate to gather in the ED evaluation in patients not known to be HIV-seropositive, particularly in endemic areas. The infection rate may be surprisingly high, even for those patients with presenting complaints not associated with HIV infection. Furthermore, inquiries about risk factors help direct the medical evaluation, remind ED personnel of the potential for occupational exposure to the virus, and afford the opportunity to offer referral for testing and counseling to persons who engage in high-risk behavior. Many cases of early HIV infection may not be detected during ED evaluation because of a low clinical suspicion for the disease, particularly in areas with a low prevalence of AIDS. Although inquiries regarding risk factors may be offensive to some patients, any difficulty usually can be averted by beginning with tactful inquiries about previous HIV testing or risk factors and indicating that these questions are routinely asked in the ED. After initial stabilization and gathering of historical information, a focused physical examination should be conducted. In elements of the examination relevant to the chief complaint, special attention should be paid to the identification of potentially treatable disorders. The universal goals of ED management are to rapidly and effectively assess the patient, identify potentially lifethreatening disorders, administer urgent interventions, generate an appropriate differential diagnosis, and provide or arrange for appropriate initial therapy, consultation, and disposition.
Stages of Human Immunodeficiency Virus Infection Several methods of classification and staging of HIV infection have been developed. The Walter Reed classification system is based on clinical and immunologic features. Other classifications are based on CD4+ counts.18 In 1993, the CDC case definition of AIDS incorporated CD4+ counts of less than 200 cells/µL as an AIDS-defining condition. The median survival time for untreated patients with AIDS is 3.7 years for those with CD4+ cell counts less than 200 cells/µL and 1.3 years for those with their first AIDS-defining complication (see Box 130-1).
Primary Human Immunodeficiency Virus Infection Acute HIV syndrome (acute seroconversion syndrome) commonly follows primary exposure by 2 to 4 weeks and may be associated with nonspecific flulike symptoms and signs such as fever, adenopathy, fatigue, pharyngitis, diarrhea, weight loss, and rash. Additional signs and symptoms such as myopathy, peripheral neuropathy, or other neurologic or immunologic manifestations are less common.19 These relatively nonspecific problems are seen in approximately 40 to 90% of patients and usually last 1 to 3 weeks. The differential diagnosis of acute HIV infection is broad in scope; considerations mainly include a wide variety of viral illnesses, such as Epstein-Barr virus (EBV) infection and viral hepatitis. Pres-
Chapter 130 / AIDS and HIV Infection
role of the ED, driven largely by multiple studies showing that the ED is the most frequent site of encounter with the health care system for persons with unrecognized HIV infection (representing approximately 30% of all patients infected). Shortly after the CDC guidelines were published, the American College of Emergency Physicians (ACEP) released a corresponding policy statement supporting the availability of HIV testing for evaluation of related acute care conditions, indicating that testing and results should be available in an expeditious and efficient fashion, as with the management of other conditions. With regard to HIV screening, the policy suggests that individual institutions need to consider the appropriateness and feasibility of screening based on the particular characteristics of their ED and available resources. A list of important considerations that must be attended to before initiation of such a program is provided in the ACEP statement.8 For institutions considering establishing ED-based HIV testing and screening programs, the American Hospital Association provides an online guide.9 A number of EDs around the country have implemented HIV testing using both rapid and traditional testing approaches.10-12 Such HIV testing programs have included both routine, broad-based screening and focused testing based on clinical suspicion. Funding for implementation has been supported in part by the CDC as well as by state and local health departments. In spite of some success, barriers to implementing screening (more so than clinically based testing) remain. Most important among these barriers are time constraints, financial burdens, impact of testing on ED crowding, and concerns regarding the responsibility for ensuring followup care.11,13,14 Emergency physicians need to be aware of state laws and local regulations governing testing. Although separate informed consent or formal pretest counseling is no longer considered mandatory by the CDC,15 many states have laws requiring informed consent before testing.16 AIDS is a reportable disease in all 50 states, and HIV infection is reportable in most states. As of 2007, 47 states are conducting confidential name-based HIV infection reporting, based on the 2005 CDC recommendations.17
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1736
ence of a rash or mucocutaneous ulcers should raise suspicion for acute HIV seroconversion. During the acute phase of HIV infection, results of standard HIV testing (enzyme-linked immunosorbent assay [ELISA] antibody testing) usually are negative, because the median time for seroconversion is approximately 2 months. If acute HIV infection is strongly suspected (on the basis of presentation and history of recent exposure), RNA viral load testing can be performed, either in the ED or by referral. Identification of acute HIV is important because HIV viral load is significantly higher during this phase of the illness and the risk of transmission is elevated. No current consensus has emerged among HIV experts, however, regarding the optimal timing and treatment regimen for acute HIV infection. Accordingly, any patient identified as having acute infection should be referred for urgent evaluation by an HIV specialist.
Predictors of Disease Progression Although the rate of disease progression from initial HIV infection to development of AIDS-defining illnesses varies widely, the average time is 10 to 12 years. Some long-term nonprogressors have remained free of AIDS-defining conditions for more than 20 years. Clinical predictors of more rapid development of clinically significant immunodeficiency include oral candidiasis, oral hairy leukoplakia, dermatomal varicella, lymphadenopathy, and constitutional symptoms.20,21 The best predictor of immunologic susceptibility to opportunistic infection is the CD4+ cell count.22 Other laboratory markers of disease progression include neutropenia and plasma HIV-1 RNA determinations.
Complications Systemic Symptoms and Signs of Human Immunodeficiency Virus Infection Systemic symptoms and signs such as fever, weight loss, and malaise are common among patients presenting to the ED. The differential diagnosis is lengthy and includes a variety of infectious causes, malignancy, and drug reactions (Box 130-2). Fever is a common presenting complaint in patients with AIDS. Evidence of an infectious cause or other reason for fever should be sought by careful history and physical examination. Initial workup for the cause of fever in an immunocompromised patient may include a complete blood count, electrolytes, comprehensive metabolic panel (CMP), chest radiograph, urinalysis and culture, and blood cultures (aerobic, anaerobic, mycobacterial, and fungal). Additional testing, based on current and past medical history and physical exam, may include stool (for culture, examination for ova and parasites, and Gram’s stain), urine (for histoplasmosis and fungal and mycobacterial culture), and induced sputum (for smear fungal and mycobacterial culture) studies; erythrocyte sedimentation rate determination; liver function tests; serum cryptococcal antigen assay; and serologic tests for syphilis, Toxoplasma, and Coccidioides. In the absence of neurologic signs or symptoms or if no other source of fever is identified, lumbar puncture should be considered after a cranial computed tomography (CT) scan. Two of the most common causes of febrile illness in patients with later-stage HIV are disseminated atypical mycobacterial infections and cytomegalovirus (CMV) infection. Atypical mycobacteral infections, caused by Mycobacterium avium complex or M. kansasii, cause disseminated disease in up to 50% of patients with AIDS and usually are associated with CD4+ counts less than 100 cells/µL. Presentation typically includes severe weight loss, diarrhea, and various consti-
Etiology for Systemic Symptoms in Human
BOX 130-2 Immunodeficiency Virus (HIV)–Infected Patients Infections Primary HIV infection (e.g., acute retroviral syndrome, HIV wasting syndrome) Protozoal Infections Pneumocystis jiroveci pneumonia Toxoplasmosis Cryptosporidiosis Bacterial Infections Streptococcus pneumoniae infection Haemophilus influenzae infection Pseudomonas aeruginosa infection Salmonellosis Bacteremia (any organism) Atypical Bacterial Infections Mycobacterium avium-intracellulare (MAI) infection Mycobacterium tuberculosis (MTB) infection Fungal Infections Histoplasmosis Cryptococcosis Coccidioidomycosis Viral Infections Herpes simplex virus infection Herpes zoster virus infection Cytomegalovirus infection Hepatitis virus infections Noninfectious Processes Adverse drug reactions Neoplasms Kaposi’s sarcoma Lymphoma Hodgkin’s disease
tutional symptoms, such as fever, malaise, and anorexia. Anemia is common. Ziehl-Neelsen (acid-fast) stain of stool or other body fluids commonly yields positive findings, and the organism also can be cultured from blood. Treatment for M. avium complex infection consists of clarithromyin, 500 mg twice a day, and ethambutol, 15 mg/kg daily. Such regimens often reduce the degree of bacteremia and symptomatology but typically do not eradicate the organism. Clarithromycin or azithromycin should be used for prophylaxis in patients with CD4+ counts below 50 cells/µL. Disseminated CMV infections typically occur in patients with CD4+ counts below 50 cells/µL. In addition to fever, patients often present with odonophagia, abdominal pain, and diarrhea secondary to esophagitis and colitis. Diagnosis usually requires endoscopy or colonoscopy for biopsy, because culture has poor sensitivity. Complications include gastrointestinal bleeding and perforation. Treatment includes immune restoration with antiretrovirals and a regimen of ganciclovir or foscarnet. Oral ganciclovir is used for prophylaxis. Some patients with HIV infection and fever or other systemic symptoms may be managed on an outpatient basis if they are not severely immunosuppressed and not systemically ill. Requirements for discharge from the ED include ability to take oral fluids, assurance of timely follow-up including obtaining results of ED-initiated cultures, and capability of providing adequate self-care. Indications for hospital admission include
Pulmonary Involvement Pulmonary manifestations of HIV infection are among the most common reasons for ED visits among patients with AIDS. Careful consideration is mandated to establish the diagnosis and initiate early treatment. The differential diagnosis of respiratory involvement is broad in scope; considerations include bacterial infections (e.g., Streptococcus pneumoniae, Haemophilus influenzae, Chlamydia pneumoniae, Pseudomonas aeruginosa, Staphylococcus aureus, Mycobacterium tuberculosis [MTB], Mycobacterium avium-intracellulare [MAI] complex), fungal infections (e.g., Pneumocystis jiroveci [formerly Pneumocystis carinii], Cryptococcus neoformans, Histoplasma capsulatum, Aspergillus fumigatus, Blastomyces dermatitides), viral infections (e.g., cytomegalovirus, adenoviruses), protozoal infections (e.g., Toxoplasma gondii), malignancies (e.g., Kaposi’s sarcoma, carcinoma, lymphoma), and others (e.g., lymphocytic interstitial pneumonitis, pulmonary hypertension, pulmonary embolism). Occurrence of specific pulmonary infections often is related to CD4+ counts. In patients with pulmonary involvement and CD4+ counts greater than 500 cells/µL, encapsulated bacteria, tuberculosis, and malignancies are common. With lower CD4+ counts, PCP, infections due to atypical mycobacteria, fungal infections, cytomegalovirus infection, lymphoma, lymphoproliferative disorders, and Kaposi’s sarcoma are seen with increasing frequency. Patients with fever and a productive cough are likely to have a bacterial pneumonia, whereas a nonproductive cough is more likely to accompany PCP, other fungal infections, or neoplasm. Hemoptysis often is associated with pneumococcal pneumonia and tuberculosis. Fulminant respiratory failure is most likely to be caused by Pneumocystis jiroveci (the agent of PCP) or CMV. Diagnostic evaluation of patients with HIV infection and suspected pneumonia should routinely include pulse oximetry, chest radiography, and complete blood count. Additional testing based on the stage of disease and clinical presentation may include arterial blood gas (ABG) analysis, serum lactate dehydrogenase determination, assays for serum cryptococcal antigen and urine Histoplasma antigen, and induced sputum specimen studies including Gram stain, acid-fast bacillus (AFB) smear, and Gomori, Giemsa, or immunofluorescent antibody (IFA) staining for Pneumocystis jiroveci. Blood culture specimens should be obtained in the ED from all HIV-infected patients with suspected pneumonia; this component of the evaluation becomes increasingly important in patients with later-stage disease. Blood culture collection, however, should not delay the initiation of antimicrobial therapy. Although radiographic findings in many pulmonary complications may be nondiagnostic, certain patterns may be suggestive of specific disorders. A focal infiltrate on the plain chest film often suggests bacterial pneumonia, whereas a diffuse interstitial or perihilar, granular pattern is associated with PCP. PCP is suggested by increased serum lactate dehydrogenase and hypoxia (especially exercise-induced), which may be more severe than expected from radiographic findings. Hilar adenopathy with diffuse pulmonary infiltrates suggests cryptococcosis, histoplasmosis, mycobacterial infection, or neoplasm. Kaposi’s sarcoma (KS) can manifest with cough, fever, and dyspnea and may mimic PCP on the chest radiograph. Table 130-1 lists common radiographic findings and associated conditions in the HIV-infected patient.
1737
Radiographic Abnormalities: Table 130-1 Chest Differential Diagnosis in the AIDS Patient FINDING
POTENTIAL ETIOLOGIES
Diffuse interstitial infiltration
Pneumocystis jiroveci Cytomegalovirus Mycobacterium tuberculosis Mycobacterium avium complex Histoplasmosis Coccidioidomycosis Lymphoid interstitial pneumonitis Mycoplasma pneumoniae Bacterial pneumonia Mycoplasma pneumoniae Pneumocystis jiroveci Mycobacterium tuberculosis Mycobacterium avium complex Kaposi’s sarcoma Mycobacterium tuberculosis Mycobacterium avium complex Fungal lesions Toxoplasmosis Pneumocystis jiroveci Mycobacterium tuberculosis Bacterial infection Fungal infection Kaposi’s sarcoma (small effusion may be associated with any infection) Kaposi’s sarcoma Lymphoma Mycobacterium tuberculosis Cryptococcus Kaposi’s sarcoma Histoplasmosis (40%) Pneumocystis jiroveci (20%) Mycobacterium tuberculosis Cryptococcosis Many other disease entities
Focal consolidation
Nodular lesions
Cavitary lesions
Pleural effusion Adenopathy
Pneumothorax Normal radiograph
As with all disease processes, ED management of pulmonary complications must first include stabilization with appropriate support of airway, breathing, and circulation. Definitive airway management may be indicated in severe cases. Volume repletion or pressors, or both, may be indicated for hypotension. Other treatment measures should include administration of supplemental oxygen and volume repletion if indicated. If the diagnosis can be ascertained or is strongly suspected, specific treatment should be instituted while the patient is in the ED, particularly if PCP is suspected. If the symptoms are of new onset or there has been a change from previous status, hospitalization should be considered. Decisions regarding patients with known pulmonary involvement are based on comparison with baseline status, the effectiveness of ongoing or previous treatment, and the individual’s ability to obtain outpatient follow-up observation (see “Disposition”). The Pneumonia Outcomes Research Team (PORT) study did not include HIV-infected patients, and most experts suggest that hospital admission should be more readily considered for patients with HIV infection and pneumonia. Staging systems for predicting death from HIV-associated pneumonia found that clinical factors associated with increased mortality include the presence of neurologic symptoms, respiratory rate of 25 breaths per minute or less, and creatinine level greater than 1.2 mg/dL.22 Bacterial infections are the most frequent type of pulmonary infection among patients with AIDS and commonly are caused
Chapter 130 / AIDS and HIV Infection
toxic appearance, neutropenia with fever, active bleeding, or other need for urgent diagnosis and treatment. For patients with persistent fevers, in whom one or more of the discharge criteria are not met, hospitalization is warranted.
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1738
by Streptococcus pneumoniae, Haemophilus influenzae, Pseudomonas aeruginosa, and numerous other organisms.23 Infections with Mycoplasma pneumoniae, Chlamydia pneumoniae, and Legionella are relatively less common. P. aeruginosa is a more common pathogen in later stages of AIDS. Presentations of bacterial pneumonia may be typical or atypical in symptoms, duration, and severity. Severely ill patients with pneumonia being managed in an ICU should be treated with linezolid or vancomycin, an antipseudomonal agent, in addition to a macrolide or respiratory fluoroquinolone.
Pneumocystis Pneumonia PCP is one of the most common opportunistic infections in AIDS. More than 80% of patients with AIDS acquire PCP at some time during their illness, and it is the initial opportunistic infection in many cases. As noted earlier, PCP is caused by the organism Pneumocystis jiroveci (formerly Pneumocystis carinii24— the original basis for the “PC” in PCP, which remains in use to refer to Pneumocystis pneumonia). Although P. jiroveci traditionally is classified as a protozoan, its morphology has been suggested to closely resemble that of a fungus.25 The incidence of PCP has declined since the widespread use of HAART.26 Patients typically present with an insidious cough (often nonproductive), dyspnea, unexplained fever for longer than 2 weeks, chest pain, and fatigue. The chest radiograph commonly shows a diffuse interstitial infiltrate but also may be normal in appearance or reveals asymmetry, nodules, cavitation, or bullae.27 Considerations in the differential diagnosis include viral, bacterial, mycobacterial, fungal, and protozoal pneumonias, as well as malignancies. The more common causes to be considered in the differential diagnosis are shown in Box 130-2. Negative findings on the chest radiograph are reported in up to 20% of patients ultimately found to have PCP.28 In situations in which a high clinical suspicion exists for PCP, chest CT should be performed; CT findings frequently are suggestive of PCP. Gallium scanning of the chest also offers improved sensitivity over that of chest radiography, but this modality generally is not available in the ED and is associated with a high false-positive rate. Serum LDH often is elevated in patients with PCP (sensitivity of approximately 90%) but, again, has poor specificity so it cannot be used for definitive diagnosis. In the ED, a presumptive diagnosis of PCP can be made in a patient with later-stage HIV (CD4+ count less than or in the range of 200 cells/µL) with unexplained hypoxia when other causes (e.g., pulmonary embolism) have been eliminated. The organism cannot be grown in the laboratory, so diagnosis relies on indirect IFA staining using monoclonal antibodies. Studies using induced sputum (often not practical to obtain in the ED) have a relatively low sensitivity. Accordingly, bronchoscopy (bronchoalveolar lavage, brush biopsy, or transbronchial biopsy) often is required for establishing the diagnosis. Establishment of a definitive diagnosis is not necessary before the initiation of treatment. Treatment should begin as early as possible with 15 to 20 mg/kg per day of trimethoprim and 75 mg/kg per day of sulfamethoxazole (TMP-SMX), given either orally or intravenously in two or three daily divided doses for a total of 21 days (e.g., two Bactrim DS tablets every 8 hours). Indications for intravenous therapy include tenuous respiratory status, an alveolar-arterial gradient above 45 mm Hg, and Pao2 below 60 mm Hg. Other therapeutic agents that can be used with PCP include pentamidine isethionate, dapsone, clindamycin plus primaquine, atovaquone, and trimetrexate. Steroid treatment (prednisone 40 mg PO, twice daily for five days with a 3-week taper) is recommended for patients with a Pao2 less than 70 mm Hg, or an alveolar-arterial gradient
greater than 35 mm Hg.29 Most patients (60 to 80%) respond to therapy, although Pneumocystis persists in the lungs of two thirds of patients. All patients requiring steroids should be hospitalized because clinical status in those with PCP typically will worsen a few days after the initiation of therapy. Adverse effects of TMP-SMX occur in up to 65% of patients with AIDS and are 20 times more common than in the general population (Table 130-2); such effects generally become apparent after 7 to 14 days of therapy. The most common adverse effects are nausea, vomiting, rash, fever, neutropenia, thrombocytopenia, hyponatremia, and hepatitis. Pentamidine can cause nausea, vomiting, diarrhea, neutropenia, hypoglycemia, hyperglycemia, renal impairment, hepatic toxicity, and orthostatic hypotension.30 Because sterile abscesses may develop at the injection site, intravenous infusion is preferred. Prophylaxis (with Bactrim DS, one tablet by mouth once daily) against PCP may be an important step in preventing reinfection and is recommended for patients with CD4+ cell counts below 200 cells/µL.31 The mortality rate for PCP-associated respiratory failure is close to 60%. Patients requiring ventilatory support should be maintained on low tidal volumes and plateau pressures, because PCP is associated with an increased risk of pneumothorax. The presence of a pneumothorax in a patient with a low CD4+ count should be presumed to be caused by PCP, although KS, intravenous drug use, toxoplasmosis, and viral, fungal, and mycobacterial infections also can cause pneumothoraces. Asymptomatic patients with a small pneumothorax (involving less than 20% of lung volume) may be treated with observation or insertion of a Heimlich valve.
Mycobacterium Tuberculosis Infection The incidence of MTB infection in HIV-infected patients has increased dramatically, and it is estimated that over 10 million patients worldwide are co-infected with HIV and tuberculosis.32 HIV-infected patients have an estimated 50- to 200-fold increased risk of acquiring tuberculosis over the general population.33 The increase in tuberculosis among the HIV-infected population is thought to be due to a number of factors, including increased risk of reactivation of latent infection, high rates of infection after exposure, overlap in at-risk groups, and rapid progression to clinically significant disease. Tuberculosis may be a very early manifestation of AIDS. Common presenting signs and symptoms include fever, cough, and hemoptysis, but in patients with immunosuppression, clinical manifestations are more atypical and extrapulmonary findings more common. Classic radiographic abnormalities are upper lobe alveolar lesions with cavitation accompanied by pleural effusions and mediastinal adenopathy.34 Findings may vary considerably, however, and atypical features and absence of radiographic abnormalities are more common among patients with lower CD4+ cell counts.35 Central nervous system (CNS), bone, visceral, skin, pericardial, eye, pharynx, and lymph node involvement also may occur. The diagnosis of tuberculosis is based on a number of factors, including risk of infection, clinical presentation, direct examination of patient specimens, and identification of mycobacteria from cultures.36 Because a definitive diagnosis cannot be made in the ED, and the disease is transmitted by the aerosol route, isolation and hospital admission are indicated for any patient with suggestive clinical factors. Definitive laboratory diagnosis can be made by a nucleic acid amplification test (NAAT), an AFB smear, or culture evaluation of induced sputum specimens or samples obtained on bronchoscopy. NAAT has a higher sensitivity than AFB smear, and bronchoscopy or tissue biopsy a higher yield than induced sputum.
1739
Table 130-2 Common Drug Reactions in HIV-Infected Persons*
Acyclovir Amphotericin Atovaquone Azithromycin Clarithromycin Clindamycin Clotrimazole Dapsone Didanosine Fluconazole Foscarnet Ganciclovir Ibuprofen Indinavir Isoniazid Itraconazole Ketoconazole Lamivudine Narcotics Pentamidine Pyrimethamine Rifabutin
X X
RASH
N/V
DIARRHEA
X
X X X X X
X
X X
X X X
H/A
DMS
NEUROPATHY
≠ LFT
X X X X
Ø WBC
Ø HCT
Ø PLT
X X
X X
X
X
X
Hepatitis Pancreatitis
X
Nephrotoxicity, seizures
OTHER
Vertigo Nephrotoxicity
X X
X X X
X X
X
X X X X
X
X X X X X X X X X X X X X
X X X X
X X X
X
X X
X X
X X
X X X
X X
X X X
X X X X
X X
X X
X X
X
Zalcitabine Zidovudine
X X
X X
X
X
X
Nephrolithiasis Hepatitis
Cough
X X
X
X X
X X
X
X
Ritonavir Saquinivir TMP-SMX
X X
X
X X X
X
X
Metallic taste
X X X
Skin discoloration Paresthesias
X
Hepatotoxicity, ↓K
X
X X
X X
X
X
*This table represents only a partial list of adverse drug reactions. An authoritative source should be consulted whenever adverse drug reactions are suspected.
Purified protein derivative (PPD) skin testing generally is not helpful, particularly in patients with more advanced immunosuppression, because negative PPD test results are common among those infected. Dissemination of pulmonary infection results in miliary tuberculosis, which can affect nearly every organ system. Treatment of patients with suspected tuberculosis should be determined in conjunction with an infectious disease specialist, taking into consideration local resistance as well as individual susceptibility tests. Patients with AIDS found to have tuberculosis should receive a four-drug regimen consisting of isoniazid, rifampin, pyrazinamide, and ethambutol, for 6 months.37,38 Second-line agents include ciprofloxacin, ofloxacin, kanamycin, amikacin, capreomycin, ethionamide, cycloserine, and para-aminosalicylic acid (PAS). Multidrug-resistant tuberculosis strains, resistant to multiple pharmacologic agents including isoniazid and rifampin, remain a concern.39 Empirical treatment for tuberculosis should be provided for HIVinfected persons with close contact with a patient with active tuberculosis.40 All HIV-infected patients with a positive result on PPD testing should receive tuberculosis prophylaxis with a regimen of isoniazid plus pyridoxine or rifampin plus pyrazinamide. Steps toward prevention of tuberculosis and its spread include the use of HAART, early identification of tuberculosis, early initiation of multidrug therapy, the use of respiratory isolation, and the use of personal respiratory protection devices.
Other Pulmonary Complications Fungal pulmonary infections other than PCP may be seen in patients with AIDS. Such infections may include cryptococcosis, histoplasmosis, coccidioidomycosis, aspergillosis, nocardiosis, and blastomycosis.41 Cryptococcus neoformans is the most common fungal pathogen in patients with AIDS after Pneumocystis jiroveci, typically causing infection in patients with CD4+ counts less than 100 cells/µL. Radiographic findings often are nonspecific and may include consolidation, reticulonodular infiltrates, and nodules. Diagnosis is with serum cryptococcal antigen assay. Geographic regions with identified predilections for infection with specific pathogens include the eastern and central United States for Histoplasma, the south central and central United States for Blastomyces, and the southwestern United States for Coccidioides. Each of these pathogens is seen more commonly in late-stage HIV disease. In patients with cavitary lesions on chest radiographs, aspergillosis as well as tuberculosis and methicillin-resistant Staphylococcus aureus infection should be suspected. Viral respiratory infections are common among HIV-infected patients. CMV infection is the most frequent, typically occurring with advanced immunosuppression. Radiographic findings may include alveolar consolidation or ground-glass opacities. Pulmonary malignancies include Kaposi’s sarcoma, nonHodgkin’s lymphoma, Hodgkin’s disease, and bronchogenic carcinoma. Kaposi’s sarcoma typically is associated with hilar
Chapter 130 / AIDS and HIV Infection
FEVER
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1740
peribronchovascular thickening, lower lobe reticulonodular opacities, adenopathy, pleural effusion, or focal consolidation.42 Pulmonary Kaposi’s sarcoma is treated with cytotoxic agents and HAART. Lymphoproliferative disorders also may have a pulmonary presentation among HIV-infected patients. Such disorders include lymphocytic interstitial pneumonia, nonspecific interstitial pneumonia, and bronchiolitis obliterans. Patients hospitalized with pulmonary involvement in whom the diagnosis cannot be determined may require bronchoscopy, bronchial lavage, and possibly biopsy. If the clinical probability of PCP is high, treatment should begin before diagnostic bronchoscopy.
Neurologic Involvement Neurologic diseases are the initial manifestation of AIDS in 10 to 20% of patients. The frequency of neurologic complications increases over the course of HIV infection, with crosssectional studies showing a 75 to 90% prevalence of neurologic disorders in patients with AIDS.43 The overall incidence rates of HIV-associated neurologic diseases and CNS opportunistic infections have been decreasing since the introduction of HAART, although this trend is expected to change as resistance to antiretroviral drugs emerges.44 Neurologic complications in the HIV-infected patient may be caused by both direct effects of HIV infection on the CNS and opportunistic infections and neoplasms occurring as a result of immunosuppression. In the early stages of HIV infection, aseptic meningitis, herpes zoster radiculitis, and inflammatory demyelinating polyneuropathy are common. Later stages of HIV infection are associated with cognitive dysfunction, dementia, opportunistic infections, cancers, and sensory neuropathies. The most common AIDS-defining neurologic complications are HIV encephalopathy (dementia), C. neoformans infection, toxoplasmosis, and primary CNS lymphoma. Less common CNS complications include bacterial meningitis, histoplasmosis (usually disseminated), CMV infection, progressive multifocal leukoencephalopathy, herpes simplex virus (HSV) infection, neurosyphilis, and tuberculosis. Noninfectious CNS processes include CNS lymphoma, cerebrovascular accidents, and metabolic encephalopathies. Clinical presentations in patients with serious neurologic complications can be nonspecific, making the diagnosis and disposition challenging. The most common clinical manifestations of CNS pathology are seizures, meningismus, focal neurologic deficits, altered mental status, and headache (new or persistent). Infection accounts for a majority of neurologic disorders and most often is accompanied by fever. Patients with CD4+ cell counts greater than 200 cells/µL who present with fever and meningismus in the absence of focal neurologic deficits should have an immediate lumbar puncture performed. For those with focal deficits or new seizures, neuroimaging is recommended first, followed by lumbar puncture if neuroimaging is unrevealing. For patients with altered mental status or headache, diagnostic evaluation should proceed as in the non-HIV-infected population, with neuroimaging and lumbar puncture reserved for those cases in which another cause for the symptoms is not identified or with a clear indication for workup (e.g., patient complaint of “the worst headache of my life”).44 For patients with CD4+ cell counts less than 200 cells/µL, a more aggressive approach is advocated, with any of the aforementioned findings demanding emergent imaging, usually followed by lumbar puncture45,46 (Fig. 130-2). Generally speaking, for those CNS processes that require immediate identification, CT without contrast is considered
adequate.47,48 If the entire ED evaluation is unrevealing, more advanced diagnostic imaging should be pursued immediately, usually in an inpatient setting, if the patient’s symptoms are severe or if new neurologic findings are present. For all other cases, close follow-up is indicated with the patient’s primary provider, because it has been demonstrated that more subtle lesions may be identified by contrast CT scan or magnetic resonance imaging (MRI). Cerebrospinal fluid (CSF) analysis should include determination of opening and closing pressures, cell count, measurement of glucose and protein, Gram’s stain, and bacterial, viral, and fungal cultures. Testing for toxoplasmosis and cryptococcal antigens and coccidioidomycosis titer also are appropriate, particularly in patients with laterstage disease. A prudent measure is to direct the laboratory to hold excess CSF for further testing if the preliminary workup is unrevealing.
HIV Encephalopathy HIV encephalopathy, or AIDS dementia complex, occurs in up to one third of patients with HIV, and is the initial manifestation of AIDS in 3% of affected adults.43 It is a progressive process caused by direct HIV infection and commonly is heralded by impairment of recent memory or subtle cognitive deficits, such as difficulty concentrating. Traditionally, symptoms are expected to occur in patients with CD4+ counts less than 200 cells/µL, although since 1996, increasing numbers of cases are being seen in patients with CD4+ counts greater than 200 cells/µL.3,49 Early stages of dementia may be easily confused with depression, the effects of psychoactive substances, or anxiety disorders. Deficits become more debilitating in later stages of disease and can include more obvious changes in mental status, seizures, frontal release signs, and hyperactive deep tendon reflexes; in such cases, physical examination usually reveals the hallmarks of advanced AIDS, including wasting, alopecia, generalized dermatitis, and lymphadenop athy. AIDS dementia is a diagnosis of exclusion: Even among patients with AIDS presenting to the ED with an established diagnosis of AIDS dementia, the appearance of progressive signs or symptoms requires immediate further evaluation to rule out other CNS processes. Neuroimaging findings in patients with HIV encephalopathy typically show atrophy and diffuse deep matter hyperintensities; MRI may reveal patchy punctate lesions in the white matter. Lumbar puncture findings typically are normal. Controlled trials in adults and children with HIV dementia have demonstrated benefit of high-dose zidovudine.49
Cryptococcus Neoformans Infection C. neoformans is the agent of a fungal CNS infection that causes either focal cerebral lesions or diffuse meningoencephalitis. It occurs in up to 10% of patients with HIV infection but most commonly in those with CD4+ counts less than 100 cells/µL. The most frequent initial symptoms are fever and headache, often accompanied by nausea and vomiting. Less frequent manifestations are visual changes, dizziness, seizures, and cranial nerve deficits.50 The brainstem and basal ganglia are typical locations; high intracranial pressure and sudden clinical deterioration from herniation are relatively common. The mortality rate approaches 30%. Patients with C. neoformans infection usually have no significant changes on CT. Definitive diagnosis relies on finding cryptococcal antigen in the CSF, which is nearly 100% sensitive and specific; other diagnostic tests include India ink staining (60 to 80% sensitive), fungal culture (95% sensitive), and serum cryptococcal antigen (95% sensitive). All patients with a positive result on serum cryptococcal antigen assay should undergo lumbar puncture to rule out neurologic involvement.
1741
Results of imaging
Figure 130-2. Approach to the evaluation of patients presenting to the emergency department with advanced HIV disease plus altered mental status, new-onset seizures, headache (severe or persistent), or focal neurologic deficits. CMV, cytomegalovirus; CSF, cerebrospinal fluid; CT, computed tomography; EBV, Epstein-Barr virus; FA, fluorescent antibody; HAART, highly active antiretroviral therapy; HSV, herpes simplex virus; IFN, interferon; MRI, magnetic resonance imaging; PCR, polymerase chain reaction; PML, progressive multifocal leukoencephalopathy; SV40, simian virus 40; VDRL, Venereal Disease Research Laboratory. (Modified from McArthur J, Bartlett JG: Headache in patients with AIDS. In Bartlett JG [ed]: 1999 Medical Management of HIV Infection. Baltimore, Port City Press, 1999, p 333.)
Multiple enhancing lesions
Atypical lesions for toxoplasmosis
Empirical treatment for toxoplasmosis*
Periventricular lesions, rapid-onset encephalopathy
CMV treatment if CSF shows CMV by PCR assay or culture
No response clinically and/or by MRI at 2 weeks
Brain biopsy (stereotactic)
Viral culture (HSV and CMV) FA stain for HSV Immune peroxidase stain for SV40 (cause of PML)
No lesions
Other lesions Lumbar puncture
Cell count, protein, glucose, VDRL cryptococcal antigen, cytology (lymphoma) rarely positive Experimental: PCR assay for toxoplasma, CMV, JC virus (PML), EBV (lymphoma) PML: HAART and consider alpha IFN
*Serology for Toxoplasma gondii is positive 85–90% with toxoplasmic encephalitis. Clinical response to empirical treatment is anticipated within 1 week. Negative serology, atypical presentation, and/or delayed clinical response should prompt early biopsy.
Additional findings associated with cryptococcal infection include elevated CSF opening pressure and a mononuclear pleocytosis. Treatment requires hospital admission for administration of intravenous amphotericin B (0.7 mg/kg per day) plus 5-flucytosine (100 mg/kg per day) for 2 weeks, followed by oral fluconazole (400 mg per day) for 8 weeks or until the CSF is sterile. The most clinically significant adverse effect of treatment for cryptococcal meningitis is bone marrow suppression due to flucytosine; amphotericin B also may cause fever and renal dysfunction. After successful treatment, chronic suppressive therapy with lower doses of oral fluconazole is indicated because of the high relapse rate (approximately 50%). This therapy can be discontinued in patients with immune reconstitution.
Toxoplasma Gondii Infection T. gondii is the most common cause of focal intracranial mass lesions in patients with HIV infection, with an incidence of 3 to 4%.43,49 In most cases, symptomatic disease is a result of reactivation of latent infection. Common signs and symptoms include headache, fever, altered mental status, and seizures. Focal neurologic deficits are found in up to 80% of cases. Serologic testing is not useful, because up to 30% of the U.S. population has antibodies to T. gondii. Diagnosis most often is made by the presence of multiple subcortical lesions on CT. Noncontrast CT often is used as the initial study in the ED, because addition of contrast has been shown to be of marginal
value in patients with completely normal findings on noncontrast CT scans.48 In those patients with suspicious lesions, or those with clinical findings strongly suggestive of a pathologic process but equivocal or negative findings on noncontrast scans, a contrast CT or MRI study may be helpful. In the presence of contrast, toxoplasmosis lesions are ring-enhancing with surrounding edema. MRI is considered even more sensitive than contrast CT in delineating the extent of lesions but usually is not indicated in the ED setting.51 Clinical and radiologic features often cannot reliably distinguish CNS toxoplasmosis from a wide variety of other potential causative disorders (e.g., lymphoma, cerebral tuberculosis, fungal infections, progressive multifocal leukoencephalopathy, CMV infection, KS, hemorrhage). Toxoplasmosis more typically is characterized by a greater number of lesions with a predilection for the basal ganglia and corticomed ullary area, whereas lymphomas more often are singular lesions located in the periventricular matter or corpus callosum. Tuberculosis is characterized by an inflammatory appearance on the CT scan, with a thick isodense exudate filling the basal cisterns. Patients with suspected toxoplasmosis should be hospitalized and treated with pyrimethamine (200 mg loading dose, then 50 to 75 mg/day) plus sulfadiazine (4 to 6 g/day). Folinic acid (leucovorin, 10 mg/day) should be added to reduce the incidence of pancytopenia. Alternative agents to sufadiazine include sulfisoxazole, clindamycin, azithromycin, atovaquone,
Chapter 130 / AIDS and HIV Infection
CT scan ± contrast or MRI (MRI with gadolinium preferred) Toxoplasmosis serology, serum cryptococcal antigen, serum VDRL assay
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1742
or doxycycline and often are required due to the relatively high frequency of side effects associated with sulfadiazine. Dexamethasone (4 mg IV) may be used in cases with the radiographic finding of midline shift, critically elevated intracranial pressure (ICP), or clinical deterioration. Seizure prophylaxis is not recommended. Bactrim is indicated for chronic suppressive therapy after initial treatment as well as for prophylaxis in patients with a positive result on serologic testing and a CD4+ cell count less than 100 cells/µL. Failure to respond to treatment suggests an alternate diagnosis, which may necessitate biopsy.
Primary Central Nervous System Lymphoma A previously rare disorder, primary CNS lymphoma occurs in up to 3% of patients with HIV infection, typically in those with CD4+ cell counts less than 50 cells/µL. Incidence has decreased slightly since 1996 with the introduction of HAART.49 Primary CNS lymphomas originate from B cells that express EpsteinBarr virus (EBV). Patients present with headache, aphasia, memory loss, hemiparesis, or seizure. Diagnosis usually is based on CT findings, which show hyperdense or isodense round or multiple lesions that enhance with contrast, and have a predilection for the periventricular region. Differentiation from toxoplasmosis can be challenging and often is made after failure to respond to therapy for that infection. PCR assay for EBV is a helpful diagnostic adjunct, but definitive diagnosis often necessitates biopsy. Prognosis for lymphoma is poor, with median survival time of less than 1 month. Life expectancy may be extended to several months with whole-brain irradiation along with corticosteroids and chemotherapy.49
Progressive Multifocal Leukoencephalopathy Progressive multifocal leukoencephalopathy occurs in approximately 1 to 3% of patients with AIDS and is caused by reactivation of the polyomavirus (JC virus). The most common presenting features are weakness, speech disturbances, cognitive dysfunction, and headaches. CT or MRI scans show hypodense white matter disease. JC virus PCR assay is approximately 80% sensitive. Progressive multifocal leukoencephalopathy carries a poor prognosis unless immune reconstitution can occur.
Tuberculosis Meningitis Mycobacterium avium-intracellulare (MAI) infection is the most common cause of tuberculosis meningitis; it occurs in less than 1% of patients with AIDS and may be associated with intracranial abscesses or spinal cord absesses. CT findings in MAI infection may suggest toxoplasmosis. Because findings on CSF analysis may be negative, definitive diagnosis often requires brain biopsy. Four-drug therapy for at least 9 months is required for cure.49
Human Immunodeficiency Virus Neuropathy HIV infection also is associated with a variety of disorders of the peripheral nervous system. These disorders rarely are emergent but necessitate appropriate referral. The most common peripheral nervous system disorder is HIV neuropathy, which occurs in up to 50% of HIV-infected patients and is characterized by painful sensory symptoms in the feet. Treatment in the ED should be directed toward analgesia. Ibuprofen may be used for first-line therapy, although narcotics may be required in more severe cases. Amitriptyline and phenytoin have been shown to be helpful but should be used judiciously because of their potential for causing delirium in patients with concurrent HIV dementia.
Gastrointestinal Involvement Most patients with AIDS have gastrointestinal signs or symptoms at some time during the course of their illness. The most common clinical manifestations are diarrhea, weight loss, malabsorption, abdominal pain, bleeding, esophageal symptoms, and hepatobiliary symptoms.52 Nonspecific findings may include nausea, vomiting, and abdominal pain as common adverse effects of antiretroviral therapy.53 Evaluation for a specific causative disorder often is difficult until objective studies are performed. More than one source of infection often is present, which may further complicate the diagnosis. Treatment in the ED focuses on supportive care, fluid and electrolyte repletion, and obtaining appropriate studies for further investigation.
Oropharynx Oral involvement is common in AIDS and may manifest as a variety of problems, including fungal infections (oral candidiasis, histoplasmosis, cryptococcosis, penicillinosis), viral lesions (herpes simplex, herpes zoster, cytomegalovirus, hairy leukoplakia, papillomavirus infection), bacterial lesions (periodontal disease, necrotizing stomatitis, tuberculosis, Mycobacterium avium complex [MAC], bacillary angiomatosis), neoplasms (Kaposi’s sarcoma, lymphoma, Hodgkin’s lymphoma), and autoimmune or idiopathic lesions (salivary gland disease, aphthous ulcers). Presence of oral lesions may be an indicator of disease progression.54 Oral candidiasis affects more than 80% of patients with AIDS. Candida albicans infection, the most common fungal infection in HIV-infected patients, typically involves the tongue and buccal mucosa and may be asymptomatic. Symptoms may include soreness, burning, and dysphagia. Candidiasis can be distinguished from hairy leukoplakia by its characteristic whitish, lacy plaques, which are easily scraped away from an erythematous base. Any of three forms of candidiasis may be seen: pseudomembranous candidiasis (thrush), erythematous candidiasis, and angular cheilitis. Microscopic examination with a potassium hydroxide smear can confirm the diagnosis in the ED. Most oral lesions can be managed symptomatically on an outpatient basis. Preferred treatment is with clotrimazole troches, 10 mg, PO, five times daily for 14 days. Other treatment options include nystatin vaginal tablets, one tablet dissolved slowly in the mouth four times daily, and nystatin pastilles, two dissolved in the mouth five times daily. Systemic therapy with fluconazole, ketoconazole, or itraconazole may be used for resistant lesions. Hairy leukoplakia also is commonly seen, typically manifesting as white, corrugated or filiform, thickened lesions on the lateral aspects of the tongue. Because it often is asymptomatic, therapy is not necessary, but when indicated, treatment is with acyclovir, 800 mg PO five times a day for 2 to 3 weeks. Painful oral and perioral ulcerations may be caused by HSV. HSV infection can be diagnosed in the ED by the identification of multinucleated giant cells in scrapings of the lesions. Definitive diagnosis is by culture. Therapy is with acyclovir, 400 mg PO three times daily for 7 to 10 days. MAC also may cause painful oral ulcerative lesions. Diagnosis is by acid-fast stain. Oral Kaposi’s sarcoma may appear as nontender, wellcircumscribed, slightly raised, violaceous or erythematous lesions anywhere in the oropharynx. Definitive diagnosis requires biopsy. Treatment may include surgical excision, localized chemotherapy, sclerosing agents, or radiation therapy. Periodontal disease, including gingival erythema and necrotizing periodontal disease, may be seen in up to 10% of
Esophagus In HIV infected patients with dysphagia or odynophagia and a CD4+ count greater than 200 cells/µL, non-HIV-related causes of esophagitis, such as gastroesophageal reflux disease or medications, must be considered. In patients with a CD4+ count below 200 cells/µL, Candida is responsible for 50 to 70% of esophagitis cases. Other etiologic disorders include HSV and CMV infection, Kaposi’s sarcoma, Mycobacterium avium complex disease, and reflux esophagitis, as well as idiopathic esophagitis. The most cost-effective approach to the evaluation of patients with esophageal complaints is to initiate empirical therapy with fluconazole (100 to 200 mg PO daily for 2 to 3 weeks). Alternative agents include clotrimazole, ketoconazole, and itraconazole. Endoscopy, fungal stains, viral cultures, and occasionally biopsy may be required to definitively establish diagnosis in nonresponders. On endoscopy, Candida infection is associated with an ulcerative pattern with plaques separated by normal mucosa compared with herpes, which typically produces “punched-out” ulcerations without plaques. Relapses are common after cessation of treatment, and intravenous amphotericin B is recommended in these cases. Disseminated candidiasis is managed with intravenous amphotericin B and flucytosine. Fluconazole has been shown to be effective for prophylaxis against fungal infections in patients with a CD4+ count less than 100/µL, although survival is unaffected by prophylactic therapy.55
Diarrhea Diarrhea is the most common gastrointestinal complaint in AIDS, occurring in 50 to 90% of patients. Diarrhea can vary in severity, ranging from a few loose stools per day to massive fluid loss with prostration, fever, chills, and weight loss. Medication side effects should be considered, because use of antiretroviral agents is associated with a high incidence of gastrointestinal adverse effects. Potential pathogens include parasites (Cryptosporidium parvum, Enterocytozoon bieneusi, Isospora belli, Giardia lamblia, Entamoeba histolytica, Microsporidia, Cyclospora, and others), bacteria (Salmonella, Shigella, Campylobacter, Helicobacter pylori, MTB, Mycobacterium avium complex, Clostridium difficile, and others), viruses (CMV, herpes simplex virus, HIV, and others), and fungi (Histoplasma capsulatum, Cryptococcus neoformans, Coccidioides immitis, and others). Opportunistic infections are more commonly seen among patients with CD4+ cell counts less than 100 cells/µL. Significant gastrointestinal bleeding and dehydration have been associated with many pathogens, particularly CMV. Salmonella infection can be of particular concern in HIV-infected patients because it often produces recurrent bacteremia. Neoplastic gastrointestinal tract involvement with Kaposi’s sarcoma or lymphoma may produce dysphagia, obstruction, intussusception, or diarrhea.
1743
ED management of diarrhea should be directed toward stabilization, hydration, and obtaining appropriate diagnostic studies. Initial studies should include blood cultures (for MAC and salmonellae), stool cultures, microscopic examination of stool for ova and parasites, trichrome stain for microsporidia, Giardia EIA, and C. difficile assay for toxins A and B. If indicated, colonoscopy or sigmoidoscopy (with or without biopsy) may be arranged for patients who require further evaluation.56 Often, no definitive diagnosis of the cause of diarrhea can be made in patients with AIDS. Management of severe diarrhea not necessitating specific therapy may include attapulgite (Kaopectate), psyllium (Metamucil), diet modification, or diphenoxylate hydrochloride with atropine (Lomotil). Cryptosporidium and Isospora infections commonly are associated with HIV infection, and both organisms may produce prolonged watery diarrhea.57 Diagnosis may be sought using acid-fast staining of stool samples, monoclonal antibody assay, or ELISA. Treatment of these disorders is variably successful. Symptoms may be managed with diet modification or loperamide. Cryptosporidium infections may be treated with some success using paromomycin or azithromycin. Isospora infections often are successfully treated with TMP-SMX. Pyrimethamine or metronidazole may be used as alternative agents. HAART also may reduce duration and severity of symptomatology. Viruses causing diarrhea include CMV, adenovirus, astrovirus, rotavirus, and others.58 CMV colitis most commonly is seen in patients with CD4+ cell counts below 50 cells/µL. Diarrhea is associated with weight loss and abdominal pain; CT shows colonic thickening. Complications include hemorrhage and perforation. Treatment is with intravenous ganciclovir, valganciclovir, or foscarnet. C. difficile is responsible for approximately 50% of cases of diarrhea in HIV-infected patients. The typical history is one of watery diarrhea and recent antibiotic use. C. difficile testing should be performed. Treatment is with oral metronidazole (Flagyl) or vancomycin. MAC consists of M. avium and M. intracellulare, acquired by ingestion or inhalation. The typical presentation is that of fever, night sweats, and diarrhea in patients with CD4+ counts less than 100 cells/µL. Diagnosis can be made by stool or blood cultures, results of which may take up to weeks to turn positive. Treatment is with oral clarithromycin, 500 mg twice a day, plus ethambutol, 15 mg/kg per day. Prophylaxis is indicated in patients with CD4+ cell counts less than 50 cells/µL. Malabsorption syndromes are relatively common with HIV infection. Delayed gastric emptying and intestinal infections may contribute and lead to significant weight loss. Treatment includes nutritional counseling, parenteral nutrition, and adjunctive agents such as dronabinol, megestrol acetate, and human growth hormone.59
Liver Involvement Hepatomegaly is seen in up to 50% of patients with AIDS. Jaundice is less common. Hepatitis B and hepatitis C are common among these patients, especially among intravenous drug users. Previous hepatitis B virus infection may become reactivated after HIV infection or may be acquired with increased frequency after HIV infection. Several opportunistic organisms, including CMV, MAI, M. tuberculosis, Histoplasma capsulatum, and Cryptosporidium, also can produce hepatitislike disease in patients with HIV infection. Typically, an elevation in the alkaline phosphatase level occurs that is disproportionate to levels of other liver enzymes. Hepatotoxi city also may result from a variety of medications, including indinavir.
Chapter 130 / AIDS and HIV Infection
patients. Outpatient treatment, including local irrigation and mouth rinses and oral antibiotics such as amoxicillin-clavulanate or clindamycin, may be instituted. Dental follow-up care is essential. Aphthous ulcerations, often painful and recurrent, are small crateriform ulcers with white to yellow membranes surrounded by an erythematous ulcer. The etiology is unknown but is thought to involve immune deficiency. Other potential causes of ulcerations, such as fungal or mycobacterial infection, HSV or CMV infection, and lymphoma, should be excluded. Aphthous ulcers usually respond to topical steroids, such as 0.05% flucinomide ointment mixed 50-50 with an oral topical anesthetic such as benzocaine preparations (e.g., Orabase).
1744
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
Anorectal Disease Complete examination of the anus and rectum is important in diagnosing such disorders. Fissures, masses, infection, and inflammation will be detected by inspection, palpation, digital examination, anoscopy, and, when indicated, sigmoidoscopy. Proctocolitis is common in patients with AIDS and may be caused by any of several organisms, including Campylobacter jejuni, Shigella species, Salmonella species, Giardia, HSV, Entamoeba histolytica, Chlamydia, and Neisseria gonorrhoeae. Diagnostic tests include anoscopy with evaluation of stool for blood, leukocytes, ova, and parasites. Additionally, bacterial cultures, an HSV culture or Tzanck preparation, a rapid plasma reagin (RPR) test, and appropriate assays for detection of N. gonorrhoeae and Chlamydia may be useful. The diagnosis of anal gonorrhea can be confirmed on a Gram’s stain of stool showing leukocytes and intracellular organisms. HSV infection can be diagnosed by viral cultures or by identification of multinucleated giant cells on scrapings of anal lesions.
Cutaneous Involvement Several common cutaneous manifestations of AIDS are likely to be seen in the ED. Preexisting dermatologic conditions may be exacerbated by HIV infection. Common infections and conditions may manifest in an atypical fashion. Generalized cutaneous complaints such as xerosis (dry skin) and pruritus are common and may be manifested before an AIDS-defining illness. Treatment for these conditions is identical to that in patients who do not have AIDS. Xerosis may be treated with emollients. Pruritus may be treated with oatmeal baths and, if necessary, antihistamines. Kaposi’s sarcoma is the second most common manifestation of AIDS. It is found commonly among homosexual or bisexual men and is caused by human herpesvirus-8 (HHV-8). The disease usually is widely disseminated with mucous membrane involvement. Kaposi’s sarcoma typically manifests in HIV-infected patients with any variation of mucocutaneous involvement, lymph node involvement, or involvement of the gastrointestinal tract or other organs. The typical appearance includes pink, red, or purple papules, plaques, nodules, and tumors. Treatment is based on site and extent of involvement. Kaposi’s sarcoma is incurable but rarely is fatal. Palliative therapies include cryotherapy, radiotherapy, infrared coagulation, sclerosing agents, intralesional vinblastine, and systemic chemotherapy with doxorubicin (Adriamycin), bleomycin, and vincristine.60 Varicella-zoster (VZ) eruptions are nearly 27 times more likely in HIV-infected patients than the general population, and multidermatomal involvement is more frequent in those with AIDS.61 In the HIV-infected patient with simple dermatomal zoster infection, outpatient management options include oral famciclovir (500 mg three times a day), acyclovir (800 mg fives times daily), and valacyclovir (1000 mg three times daily).62 Hospital admission to an isolation bed for administration of intravenous acyclovir (10 mg/kg every 8 hours) is warranted for any patient with systemic involvement, ophthalmic zoster, or severe dermatomal zoster.63 Varicella immune globulin may be useful in patients with primary infection and visceral involvement. HSV infections are highly prevalent among HIV-infected patients. Both HSV-1 and HSV-2 infections may occur as localized infection or systemic disease. HSV infections commonly manifest with fever, adenopathy, malaise, and ulcerative lesions of mucosal and cutaneous sites. Common sites of involvement include oral mucosa, genital areas, and rectum. HSV and HZV infections may be difficult to distinguish clini-
cally, and cultures may be required for differentiation of the two conditions. Reactivation is common. Mucocutaneous HSV infection responds well to oral famciclovir (750 mg three times daily) or acyclovir (200 mg five times daily for 10 days). For disseminated infection or neurologic involvement, intravenous acyclovir is recommended (5 to 10 mg/kg every 8 hours for 7 to 21 days). Famciclovir, penciclovir, foscarnet, or valacyclovir also may be used. Suppressive therapy is effective in decreasing rates of recurrence. Patients with these viral infections should be assigned to isolation beds in the hospital. Molluscum contagiosum manifests with small flesh-colored papules with a whitish core (see Fig. 151-4). The condition is difficult to cure; cryotherapy or curettage is reserved for symptomatic lesions. Intertriginous infections with either Candida or Trichophyton are common and may be diagnosed by microscopic examination of scrapings in potassium hydroxide. Treatment may include topical imidazole creams (e.g., clotrimazole, miconazole, ketoconazole). Scabies should be considered in all HIV-infected patients, particularly those with dermatitis complicated by excoriations or pruritus. Microscopic identification of mites is diagnostic. Preferred treatment is with single application 5% permethrin. Sexual and household contacts also should be treated. Norwegian scabies is particularly resistant to treatment and should be considered if lesions consistent with scabies fail to respond to traditional therapy. Treatment should be undertaken in consultation with an infectious disease specialist. Seborrheic dermatitis is common, particularly among patients with AIDS-associated dementia. Lesions are erythematous, hyperkeratotic scaling plaques involving the scalp, face (especially the nasolabial folds), ears, chest, and genitalia. Treatment with topical steroids often is effective, although less so than in the general population. Alternative therapy includes topical or oral ketoconazole. Human papillomavirus infections occur with increased frequency in immunocompromised patients. Treatment is cosmetic or symptomatic and may include cryotherapy, topical agents, or, in extreme cases, laser therapy. Other dermatologic disorders, including psoriasis, atopic dermatitis, and alopecia, occur with increased frequency in patients with AIDS. Any preexisting dermatologic disorder may be exacerbated by HIV infection.
Ophthalmologic Manifestations Ocular findings are common in the HIV-infected patient. Cotton-wool spots in the retina are the most common eye finding and do not require intervention. Other common ophthalmologic manifestations of HIV include CMV retinitis, herpes zoster ophthalmicus, and Kaposi’s sarcoma of eyelids or conjunctiva. CMV retinitis occurs in 10 to 30% of HIV-infected patients and is the most common cause of blindness in patients with AIDS. With advances in HAART, reduced incidences of CMV retinitis have been observed, but discontinuation of HAART may result in intraocular inflammation.64 CMV retinitis typically produces severe necrotic vasculitis and retinitis. This may be asymptomatic or may manifest as blurred vision, a change in visual acuity, “floaters,” flashes of light, photophobia, scotoma, redness, or pain.65 Funduscopic examination typically shows fluffy white perivascular lesions with areas of hemorrhage that may be confused with retinal cotton-wool spots, a benign condition with no prognostic implications for those with AIDS. Considerations in the differential diagnosis also include toxoplasmosis, syphilis, HSV infection, VZV infection, and tuberculosis. Because of the risk of rapid progression and blindness, any patient in whom ophthalmic CMV
Cardiovascular Manifestations Cardiac manifestations of AIDS may include pericardial effusion, cardiomyopathy, increased left ventricular mass, myocarditis, endocarditis, malignancy, and cardiotoxicity of medications.69 The pericardium is the most common site of cardiac involvement, although many patients have clinically insignificant effusions. Pericardial effusions may be secondary to malignancies, uremia, lymphatic obstruction, or infections such as with Mycobacterium tuberculosis, Streptococcus pneumoniae, Staphylococcus aureus, or a host of other bacterial, viral, fungal, or protozoal pathogens. Infective endocarditis occurs commonly in HIV-infected patients with a history of intravenous drug use and should be considered in all injection drug users presenting with febrile illnesses. Cardiac neoplasms also may occur, typically as either Kaposi’s sarcoma or lymphoma. Such neoplasms may be clinically silent or may manifest congestive heart failure, tamponade, or arrhythmias or other clinical syndromes. Some antiretroviral agents are associated with a fat redistribution syndrome and diabetes, which may increase the risk of coronary artery disease. Patients with HIV infection also have higher rates of dilated cardiomyop athy. Etiologic categories include primary HIV infection; viral, mycobacterial, fungal, or protozoal infection; drug-induced; immunologic; and ischemic. Patients present with typical signs and symptoms of congestive heart failure; echocardiography shows left ventricular diastolic dysfunction with decreased ejection fraction. These patients are at an increased risk for arrhythmias.
Renal Manifestations Renal insufficiency in the patient with AIDS may be associated with a variety of underlying disorders, but initial ED presentation may include general malaise, edema, or oliguria. Prerenal azotemia is the most common renal abnormality, especially in conjunction with volume loss related to systemic or gastrointestinal infection. It is diagnosed and treated by evaluation and therapy of fluid status. Acute renal failure also
1745
may occur and often is secondary to drug nephrotoxicity (e.g., from pentamidine, aminoglycosides, sulfa drugs, foscarnet, rifampin, dapsone, or amphotericin B). HIV-associated nephropathy (HIVAN) typically is a cause of chronic renal insufficiency in the late stages of immunosuppression, but may occur earlier in disease progression.70 Vasculitis, tuberculosis, or other systemic infections also may contribute to renal insufficiency. Postrenal azotemia may result from tubular, ureteral, or pelvis obstruction or from lymphoma, stones, fungus ball, blood clot, or sloughed papillae. ED evaluation should include urinalysis, assessment of fluid status, and determination of blood urea nitrogen and serum creatinine. If indicated, ultrasonography or intravenous pyelography may demonstrate the site and degree of obstruction. Renal biopsy may be indicated for patients with proteinuria and undiagnosed renal disease. Treatment depends on the causative agent. Therapies that have demonstrated limited benefit in HIVAN include corticosteroids, angiotensin-converting enzyme inhibitors, and dialysis and should be initiated in consultation with a nephrologist.
Psychiatric Considerations HIV-infected patients may present with a variety of social and emotional issues complicated by neuropsychiatric and cognitive impairments. The diagnosis of AIDS may dramatically alter interactions with family and friends, and patients may be devastated by the prospect of confronting chronic illness and death. Although psychiatric issues are common among HIVinfected patients, many do not receive optimal care.71 Depression is common among AIDS patients and often is responsive to hospitalization and psychosocial intervention. It has been estimated that 60% of HIV-infected patients experience depression during their illness.71 Patients with depression generally have lower CD4+ counts and may report more AIDSrelated symptoms.72 Referral for antidepressant therapy should be considered if symptoms have lasted for longer than 2 weeks. Depression may result in suicidal ideation and may bring the patient to the attention of the ED for medical treatment after a suicide attempt. Other psychiatric disorders may be seen, including personality disorders, addiction disorders, and adjustment disorders. Delirium suggests the presence of a primary physiologic disease state. Considerations in the differential diagnosis include CNS, toxic, and metabolic derangements. AIDS psychosis commonly manifests with psychiatric symptoms such as hallucinations, delusions, or other abnormal behavioral changes. Treatment should be undertaken with traditional antipsychotic agents.
Sexually Transmitted Diseases Sexually transmitted diseases (STDs) are common among HIV-infected patients, and the prevalence of STDs, including syphilis, is increasing.73 Syphilis is associated with increased susceptibility to HIV seroconversion by an unknown mechanism.74 For EDs that are able to institute screening for HIV, combined testing for both HIV and other STDs may be most cost-effective.75 Common STDs include gonorrhea, chlamydial infection, herpes, and syphilis. Serologic testing for syphilis should be performed for all HIV-infected patients with a suspected STD. Empirical therapy for syphilis may be instituted even without laboratory proof of infection. Recommended treatment for primary or secondary syphilis of less than 12 months’ duration is with a single intramuscular dose of benzathine penicillin (2.4 million units). For latent syphilis or unknown
Chapter 130 / AIDS and HIV Infection
infection is a possibility requires immediate evaluation by an ophthalmologist. Treatment is with intravenous ganciclovir (5 mg/kg every 12 hours for 2 weeks, followed by 6 mg/kg/day maintenance therapy) or foscarnet (90 mg/kg every 12 hours). Intravitreal injections of fomivirsen also may be used for patients unresponsive to traditional therapy.66 Similar rates of efficacy are achieved with ganciclovir and foscarnet. Ganciclovir-containing intravitreal implants constitute another therapeutic option that provides higher intravitreal concentrations and is associated with reduced risk for CMV-related retinal detachment. Immune recovery uveitis may occur as a complication of treatment during the recovery phase.67 Chronic suppressive therapy with ganciclovir or foscarnet may be indicated. Patients with serum anti-Toxoplasma antibodies and CD4+ cell counts below 100 cells/µL should receive prophylaxis with TMP-SMX.68 Herpes zoster ophthalmicus is another common cause of ocular damage in patients with HIV infection. The typical presentation is pain or paresthesia in the distribution of cranial nerve V1, followed by the emergence of the vesicular zoster skin rash. Complications include conjunctivitis, episcleritis, iritis, keratitis, secondary glaucoma, and, rarely, retinitis. As with CMV infection, early recognition and treatment can prevent morbidity. All patients with suspected zoster ophthalmicus require immediate ophthalmologic consultation and may need hospital admission. Treatment should be initiated with oral or intravenous acyclovir, famciclovir, or valcyclovir.
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1746
duration of secondary syphilis, three weekly injections are recommended. Patients with known or suspected syphilis should be evaluated for the presence of neurosyphilis, which has an increasing incidence among HIV-infected persons. Patients with neurosyphilis should be treated with penicillin G, 12 to 24 million units IV daily for 10 to 14 days.
Drugs Approved by FDA for Table 130-3 Antiretroviral Treatment of HIV Disease DRUG CLASS
GENERIC NAME
TRADE NAME
NRTIsa
Zidovudine (AZT, ZDV)
Retrovir Combivir (AZT + 3TC) Trizivir (AZT+3TC + ABC) Videx and Videx EC Hivid Zerit Epivir Epzicom (3TC + ABC) Ziagen Emtriva Atripla (FTC + EFV + TDF) Truvada (FTC + TDF) Viread Viramune Rescriptor Susitiva Crixivan Aptivus Prezista Lexiva Norvir Invirase Viracept Reyataz Kaletra Fuzeon Selzentry Isentress
Hematologic Complications Hematopoiesis may be adversely affected by HIV infection, tumor, infection, or HIV medications. Anemia in AIDS is independently associated with an increased risk of death. Chronic anemia in AIDS characteristically is of the normocytic, normochromic type, with a low reticulocyte count and low erythropoietin level. HAART usually results in improvement.
Didanosine (ddI) Zalcitabine (ddC) Stavudine (d4T) Lamivudine (3TC) Abacavir (ABC) Emtricitabine (FTC)
Pediatric Considerations HIV/AIDS in pediatric patients may have a variety of ED presentations, including recurrent or severe bacterial infections, chronic diarrhea, candidiasis, opportunistic infections, and numerous other clinical syndromes. In addition to stabilization, diagnostic tests, and definitive management, close follow-up and communication with family members and primary physicians are imperative in this population.
NNRTIsb PIsc
Drug Reactions Drug reactions are extremely common among HIV-infected patients; these patients commonly are treated with a variety of drugs known to produce adverse effects in some people. In addition, for unclear reasons, HIV-infected persons often experience more frequent or more severe reactions to commonly used medications than noninfected patients. Dermatologic reactions are particularly common. Antimicrobial drugs frequently are implicated. Drug reactions must always be considered as a possible cause of new symptoms in HIV-infected patients. Table 130-2 presents a brief summary of common drug reactions in the HIV-infected patient.
■ MANAGEMENT Antiretroviral Therapy and Chemoprophylaxis The introduction of HAART in 1996 has had dramatic effects on the clinical consequences of HIV infection in the developed world. The incidence of AIDS-defining illnesses and death rate declined rapidly through 1998.3 Reports of high levels of treatment failure due to serious adverse effects, emergence of drug resistance, and difficulties in maintaining long-term adherence have raised concerns regarding the continued success of HAART; however, recent studies suggest that the reduction in morbidity and mortality associated with HAART has been sustained.76 The U.S. Department of Health and Human Services (DHHS) has published guidelines for use of antiretroviral agents in HIV-infected adults and adolescents.77 Antiretroviral therapy for HIV infection is constantly evolving, and optimal decisions regarding such therapy will require a basic understanding of the classes of drugs available, the rationale for initiating treatment, and the common adverse drug reactions. The five classes of antiretroviral drugs currently in use are listed in Table 130-3. Each class of drugs independently interrupts the normal life cycle of the HIV. When used with appropriate timing and in combination, these agents have been shown to significantly delay progression of disease and prolong life.
EIsd IIe
Tenofovir (TDF) Nevirapine (NVP) Delavirdine (DLV) Efavirenz (EFV) Indinavir Tipranavir (TPV) Darunavir (DRV) Fosamprenavir (FPV) Ritonavir (RTV) Saquinavir (SQV) Nelfinavir (NFV) Atazanavir (ATV) Lopinavir/ritonavir Enfuvirtide (T20) Maraviroc (MVC) Raltegravir (RAL)
a
Nucleoside analog reverse transcriptase inhibitors Nonnucleoside reverse transcriptase inhibitors c Protease inhibitors d Entry inhibitors e Integrase inhibitor b
The first drug demonstrated to have antiretroviral activity was an nucleoside analogue reverse transcriptase inhibitor (NRTI), a competitive inhibitor of the viral enzyme reverse transcriptase. Several controlled trials showed that zidovudine (azidothymidine [AZT], Retrovir) decreases the number and severity of opportunistic infections.78,79 Although zidovudine also was found to decrease the rate of AIDS progression in patients with early symptomatic HIV infection, no significant change in survival was found.80 This finding, coupled with the recognition of the emergence of drug resistance and the appearance of significant side effects, led to the development of other NRTIs. Combination therapy, with zidovudine and another NRTI, resulted in not only prevention of disease progression but also decreased mortality.81 The FDA has since approved multiple agents in this class, each with its own unique adverse effect profile. The most common side effects are bone marrow suppression with zidovudine; distal sensory peripheral neuropathy with didanosine (Videx), stavudine (Zerit), and zalcitabine (Hivid); and pancreatitis with didanosine.81 The non-nucleoside reverse transcriptase inhibitors (NNRTIs) are noncompetitive inhibitors of reverse transcriptase and block RNA-dependent and deoxyribonucleic acid (DNA)–dependent DNA polymerase activity. Three NNRTIs are currently available; the most commonly used agents are nevirapine (Viramune) and efavirenz (Sustiva). Target organ-
1747
some evidence suggesting that higher CD4+ counts can be achieved with early therapy; however, initiation of treatment in these patients should be individualized in accordance with the willingness and readiness of the person to begin therapy, the potential benefits and risks of initiating therapy in an asymptomatic person, and the likelihood of compliance with the prescribed treatment regimen. Selection of an appropriate combination of drugs also is a complex issue for which no definitive recommendations exist. Twenty-eight antiretroviral drugs are currently approved by the FDA. A complete list and up-to-date guide for their use can be found on the NIH website (AIDSinfo.nih.gov). The current DHHS recommended first-line HAART regimens include NNRTI-based regimens (one NNRTI and two NRTIs), or protease inhibitor–based regimens (one or two protease inhibitors and two NRTIs).77 Triple-NRTI regimens are recommended as second-line regimens. Therapy should be individualized with consideration of tolerability, comorbid conditions, adverse effect profile, likely drug-drug interactions, convenience, and likelihood of adherence. Pregnancy should not preclude women from receiving optimal treatment regimens; however, issues relating to prevention of mother-to-child transmission as well as maternal and fetal safety deserve special considerations. Previous studies have shown that HAART reduces perinatal transmission to 1 to 2%, and the rate is strongly correlated with viral load at the time of delivery.86,87 On the basis of these observations, HAART should be recommended for any pregnant woman. Selection of antiretroviral combinations should take into account known safety efficacy and pharmacokinetic data for each agent during pregnancy. Efavirenz-containing regimens should be avoided in pregnancy or in women of reproductive age owing to potential teratogenic effects. Elective cesarean section has established merit in reducing prenatal transmission if done at 38 weeks of gestation with a maternal viral load greater than 1000 copies/mL.88 The goal of the antiretroviral therapy is to produce longterm viral suppression. Clinical situations that should prompt consideration for changing therapy include drug toxicity or intolerance, difficulty with adherence, and failure to suppress viral infection. Decisions regarding alternative treatment regimens should be made in consultation with infectious disease experts to assess potential cross resistance from previously used drugs. With advances in genotypic and phenotypic analysis of HIV strains, selection of a drug regimen based on drug resistance patterns will soon become an essential part of therapeutic decision-making. Chemoprophylaxis is directed toward preventing initial and subsequent episodes of certain opportunistic infections (i.e., primary and secondary prophylaxis). Emphasis on measures to prevent opportunistic infections is critical because of the inherent limitations of HAART and the recognition that these infections constitute a cause of significant morbidity and mortality in the HIV-positive population. The CD4+ cell count is the best predictor of the risk for opportunistic infections and is used most often in making decisions about initiating or maintaining antimicrobial prophylaxis. The most serious and common infections for which antimicrobial prophylaxis has been shown to be effective include PCP, toxoplasmosis, tuberculosis, and MAC infection. Specific timing and choice of agents are described in earlier clinical sections of this chapter; a more comprehensive review can be found in the Public Health Service and Infectious Disease Society Revised Guidelines for the Prevention of Opportunistic Infections.89 The emergency physician can play a critical role in recognizing those patients requiring initiation of chemoprophylaxis and then should work closely with the patient’s
Chapter 130 / AIDS and HIV Infection
isms have a high propensity for developing resistance to these agents, which are recommended for use only as part of a threedrug (or more) regimen. Rash is the most common side effect associated with use of the NNRTIs, with the development of Stevens-Johnson syndrome seen in a small minority of patients (less than 5%).82 Symptomatic hepatitis, including fatal hepatic necrosis, has been reported with use of nevirapine.83 The enzyme HIV protease activates the HIV proteins that are required for infectivity by cleaving the inactive viral polypeptide precursors. Protease-inhibiting agents block this step, thereby preventing HIV particles from becoming infectious. Ten protease inhibitors currently are approved for clinical use in the United States. Introduction of this class of drugs is believed to be responsible in large part for the marked decline in mortality rates for HIV infection, which was first realized in 1996. Protease inhibitors are expensive, however, and they also have been associated with a high frequency of side effects. Short-term effects are principally gastrointestinal (including nausea, diarrhea, and bloating); long-term effects are metabolic, the most common of which are hyperglycemia, hyperlipidemia, and fat redistribution. Additional newer classes of drugs include entry inhibitors and integrase inhibitors. Entry inhibitors prevent HIV entry into cells by targeting specific viral surface proteins or their corresponding receptors.84 Enfuvirtide and maraviroc are two entry inhibitors currently approved for use in combination with other antiviral agents only in treatment-experienced patients. Major side effects include local injection site reactions and increased rate of bacterial pneumonia (with enfuvirtide). Integrase inhibitors work by blocking integrase, a protein required by HIV to allow it to insert its viral genetic material into the genetic material of an infected cell.85 Raltegravir is the only integrase inhibitor currently approved for use, restricted to patients who have limited or no treatment options. Common side effects include diarrhea, nausea, headache, and fever. The DHHS recently published updated guidelines for use of antiretroviral agents in HIV-infected adults and adolescents.77 In general, the goals of antiretroviral therapy are virologic, immunologic, clinical, and therapeutic. Because virologic (HIV RNA levels) and immunologic (CD4+ cell count) parameters are independent predictors of clinical outcomes, therapeutic recommendations are based on both of these factors. The virologic goal is to reduce viral load as much as possible, halt disease progression, and prevent development of resistant HIV variants. Immunologic goals are to achieve both quantitative (CD4+ cell count) and qualitative (pathogen-specific immune response) immune reconstitution. The principal clinical goals are to prolong and improve the quality of life. The therapeutic goal is to achieve the other three goals by choosing a sequence of drugs that maintains therapeutic options, minimizes side effects, and optimizes the likelihood of patient compliance with the chosen regimen. Expert consensus on the timing of initiation of HAART continues to evolve. The current consensus recommends mandatory treatment for HIV-infected patients with CD4+ counts below 350 cells/µL or with history of AIDS-defining illness. Other patient populations in whom initiation of antiretroviral therapy is indicated regardless of CD4+ count include pregnant women, patients with HIVAN, and patients with HBV co-infection requiring treatment. Similarly, in patients with recognized primary HIV infection, antiretroviral therapy is recommended, because early treatment is believed to decrease the number of infected cells, maintain or restore immune response, and perhaps lower the viral “set point,” resulting in improved course of the disease.81 Therapy may be considered in some patients with CD4+ greater than 350 cells/µL, given
1748
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
primary care doctors or an infectious disease consultant to begin therapy.
Immunizations for Human Immunodeficiency Virus–Infected Patients Response to immunizations may be variable among HIVinfected patients. Many such patients mount an adequate antibody response to immunizations, but the immune response is not predictable.90 Most routine immunizations recommendations are the same as for the non-HIV-infected patients.91 However, HIV-infected patients should not receive live virus or live bacterial vaccines. Pneumococcal vaccine is recommended for all patients older than 2 years of age92; however, immunization is recommended early in the disease course to optimize antibody development.93 Hepatitis B vaccine is indicated for patients at risk of exposure, although owing to variable immune response, follow-up serologic testing is indicated. Hepatitis A vaccination also should be considered because of increased risk of severe liver damage among patients previously infected with hepatitis B or C.94 Influenza vaccination is considered safe and is routinely recommended.95 Measlesmumps-rubella (MMR) vaccine may be considered because studies have not documented an increased incidence of adverse effects. If polio vaccine is indicated, enhanced inactivated polio vaccine may be administered. Although evidence suggests that the expression of HIV may be transiently increased by administration of tetanus toxoid,96 the clinical significance of this observation is unknown; current recommendations include providing a booster every 10 years for patients who have completed their primary series. Because the smallpox vaccine has not been rigorously studied in the HIV-infected population, the adverse effects and immune response are unknown, and some experts currently advise against its use.97 The potential risks and benefits of immunization in the HIVinfected patient should be considered in decisions regarding immunization.
■ DISPOSITION When questions remain about specific diagnostic or management options, consultation with specialists is appropriate. Consultations with an infectious disease specialist, neurologist, psychiatrist, AIDS specialist, and others may be indicated. Although symptomatic patients are cared for predominantly by AIDS specialists, the increasing numbers of symptomatic patients are shifting the focus of primary care to nonspecialists. Disposition decisions for HIV-infected patients are based, as for any patient, on clinical condition, availability of outpatient resources, and ability to arrange adequate follow-up observation. Any patient to be discharged must demonstrate capability for self-care or have sufficient in-home assistance available. In the AIDS population, particular attention should be given to ability to ambulate and to tolerate oral intake, as well as availability of timely and appropriate medical followup care. Although the AIDS epidemic has raised concerns regarding the economic impact of the disease, financial considerations should not be a factor in determining management or disposition. Guidelines for hospital admission and discharge are presented in Box 130-3.
■ ETHICAL CONSIDERATIONS Numerous ethical issues arise in the management of HIVinfected patients. General issues relevant to many patients
General Emergency Department (ED)
BOX 130-3 Discharge Requirements
Stable medical condition Normal or baseline vital signs Appropriate follow-up arrangements have been made Appropriate consultations and referrals have been made Patient understands discharge instructions Patient or caregiver is able to comply with discharge instructions Patient or caregiver understands warning signs indicating the need for repeat ED evaluation
may include issues of confidentiality, discrimination, access to health care, justice, informed consent, respect for autonomy, and advance directives. Additionally, concerns specific to HIV infection may arise, such as questions related to prenatal testing, abortion, euthanasia, suicide, access to experimental therapies, and role in clinical trials. In general, commonly accepted principles of medical ethics may be applied, which include principles of beneficence, nonmaleficence, respect for autonomy, and justice. Additionally, codes of ethical conduct developed by the define and the Society for Academic Emergency Medicine (SAEM) may provide general guidance.98,99 Testing of patients to detect HIV infection has some controversial aspects. Routine HIV testing initiated in the ED often is not appropriate because of difficulties in ensuring appropriate pre- and post-test counseling and confidentiality issues. However, recommendations and referral for testing often are indicated for patients with risk factors or with clinical evidence of HIV infection. Each institution should have appropriate mechanisms arranged for these referrals. Occupational exposures to blood and body fluids may necessitate testing of patients and health care workers in the ED to expedite initiation of antiretroviral therapy. In such cases, institutions not only must comply with state guidelines but should implement uniform policies and procedures for testing that ensure pretest and post-test counseling and confidentiality of results. Confidentiality of the patient’s identity and medical data are of paramount importance in the ED, particularly for HIVinfected patients, for whom breached confidentiality may have numerous clinical, social, psychological, career, and insurability effects. Public health responsibilities may at times override the duty of the physician to maintain strict confidentiality. AIDS is a reportable disease in most states, and state guidelines for reporting should be followed as a public health measure, even if this breaches confidentiality, as in cases of child abuse, gunshot wounds, or other infectious diseases. Moreover, the physician who is aware of potentially contagious practices of an infected patient has an obligation to provide appropriate counseling for that person. Additionally, infected patients should be encouraged to divulge their disease state with sexual or needle-sharing partners. In many states, the physician has the discretion to inform public health officials about the culpable practices, to allow partners potentially at risk to be informed.100 The potential value of aggressive interventions in critical care settings must be determined on an individual case basis. Some clinicians believe that in the advanced stages of AIDS, resuscitative measures are not appropriate because of the uniformly poor prognosis. Many patients may agree as they approach the terminal stages of their disease. Appropriate advance directives should be completed before the patient
1749
at a Baltimore inner city hospital found that up to 11% of patients were infected with HIV and nearly 24% were infected with HIV or hepatitis B or C.105 Numerous studies have demonstrated that a substantial number of patients in the ED have previously undiagnosed HIV infection, and HIV seroreactivity cannot be accurately predicted even with the aid of risk factor assessment. Because asymptomatic persons who are HIV antibody–positive can transmit the disease, all contacts with patients’ blood or body secretions must be considered to be potentially infectious by ED personnel. HIV transmission by health care workers to patients appears to be extremely rare. Only seven cases have been reported to date, six of which occurred from a single dentist’s practice, and one from a patient who apparently acquired HIV during orthopedic surgery. At present, routine screening of health care personnel is not indicated. Numerous studies have demonstrated that health care workers can significantly reduce their risk of exposure to blood-borne pathogens by following universal precautions. CDC guidelines for universal precautions include the use of protective equipment (including gloves, gown, mask, and eye protection) for any situation in which the potential for exposure exists. Protective equipment is indicated for most ED procedures, including examination of the bleeding patient, chest tube placement, lumbar puncture, and other commonly performed procedures in which contact with blood or body fluids is likely. Although significant improvement has been made in observance of universal precautions in the ED setting, studies have indicated that continued education and improvements in work environments are required to ensure consistent compliance.106,107
Postexposure Prophylaxis Occupational Exposures
■ PRECAUTIONS AND POSTEXPOSURE PROPHYLAXIS FOR HEALTH CARE WORKERS Precautions and Exposures Health care workers often are exposed to the blood and body secretions of HIV-infected patients or of other persons who are at high risk of harboring HIV and other infectious pathogens. The overall risk of having any occupational blood exposure is not insignificant, with more than one half of emergency physicians reporting at least one occupational exposure during a 2-year period.102 The overall risk of contracting HIV infection remains small. As of December 2006, the CDC had received reports of 57 documented cases of HIV seroconversion that were temporally associated with occupational exposure to HIV among U.S. health care workers.103 An additional 140 infections among health care workers were considered to represent possible cases of occupational transmission. No new documented cases of occupationally acquired HIV/AIDS have been reported since December 2001. Global surveillance data is less reliable, so the overall rates of occupational transmission are not known. A majority of cases occurred in nurses; less frequently affected were laboratory technicians and physicians.102 Of all transmissions, a majority were percutaneous, followed by mucocutaneous or both. There have been no confirmed seroconversions to date with exposures to a suture needle. Efficacy of transmission is estimated at 0.3% for percutaneous exposure and 0.09% for mucocutaneous exposure.104 The proportion of patients infected with a pathogen varies by geographic setting and practice locale. A survey conducted
Postexposure prophylaxis (PEP) reduces the risk of HIV transmission and seroconversion.108 The CDC provides explicit guidelines for institution of PEP for occupational exposure to HIV.104 Current guidelines advise case-by-case determination of exposure risk to resolve whether PEP should be recommended. Recommendations are based on two primary factors: (1) type of exposure and (2) HIV status of the source (or, if the source status is unknown, evaluation of risk status of the source). Separate recommendations are provided by the CDC for percutaneous and mucus membrane or nonintact skin exposures. Exposures involving contact between intact skin and blood or other body fluids contaminated by HIV are not indications for therapy. Higher-risk percutaneous exposures associated with an increased likelihood of transmission include those involving deep injuries, visible blood on a device, and injuries sustained during placement of a catheter in a vein or artery; lower-risk percutaneous exposures are superficial or involve solid needles. High-risk sources are patients with symptomatic HIV infection, AIDS, acute seroconversion, or high viral load; low-risk sources are patients with asymptomatic HIV infection or viral load of less than 1500 copies/ mL.109 When the serostatus of the source is not known (i.e., no recent positive or negative results on serologic tests), rapid testing should be performed. A negative result on EIA (using either SUDS or OraQuick) is adequate for a decision to withhold or discontinue therapy if initiated. Some states allow testing the source patient without informed consent. Confidentiality should be rigorously protected while still ensuring that the appropriate information is provided to all exposed persons. In unusual circumstances in which the source patient has an illness consistent with acute
Chapter 130 / AIDS and HIV Infection
enters the resuscitation setting. However, many patients fail to provide such documentation. Decisions regarding the withholding of extraordinary resuscitation efforts may be difficult to make in the ED because of insufficient information about an individual patient, his or her wishes, the specific disease state, prognosis, and the judgment and intentions of the primary care and consultant physicians. Although some ethicists argue against the excessive use of extensive resources for this class of patients, decisions in the ED should be largely unbiased and based on the appropriate factors relevant to the individual case. As with all patients with clinical indications for invasive monitoring or interventions, decisions should be based on factors including the patient’s wishes (if known) or a surrogate’s assessment of the patient’s wishes, expected outcome of the intervention, and potential risks of the intervention. Interventions should not be withheld or discontinued merely because of the presence of AIDS. If certain diagnostic and therapeutic interventions are withheld, particular attention should be made to ensure adequate control of pain and other symptoms. Psychosocial, religious, and cultural needs also should be addressed. The courts have addressed increasing numbers and varieties of cases regarding the treatment of AIDS and HIV-related illness. The AIDS Litigation Project (a review of cases) has shown increasing cases of litigation involving areas of AIDS education, blood supply, epidemiologic surveillance, criminal law, public places, products and fraud, torts, court system, family law, confidentiality, prisons, military, fear of exposure, homelessness, and discrimination.101 In general, the same ethical principles of respect for autonomy, beneficence, nonmaleficence, justice, confidentiality, communication, informed consent, and research ethics should be honored in the treatment of HIV-infected patients as for all ED patients.
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1750
HIV infection, testing should include assay of HIV RNA levels. Current public health guidelines recommend a 4-week regimen of two drugs for most HIV exposures given by percutaneous or mucous membrane routes.104 Two-drug therapy options include zidovudine plus lamivudine (available as Combivir), lamivudine plus stavudine, and didanosine plus stavudine. For highest-risk exposures, an expanded PEP regimen with the addition of a protease inhibitor (preferably lopinavir plus ritonavir) is advised. When the source is known to be infected with a resistant HIV strain, the selection of PEP drugs to which the source’s virus is unlikely to be resistant is recommended. PEP should be initiated as soon as possible, preferably within hours rather than days of exposure; in general, antiretroviral therapy is not indicated in patients presenting more than 36 hours after exposure. The optimal duration of PEP is 4 weeks, if tolerated. Constitutional and gastrointestinal side effects may be significant and often lead to early termination of treatment. Initial treatment should never be delayed during the wait for information regarding final determination of overall risk of exposure, because therapy subsequently can be altered or stopped after the first dose. If the source person’s HIV infection status is unknown at the time of exposure, use of PEP should be decided on a case-by-case basis, with type of exposure and likelihood of HIV infection in the source taken into consideration. If the findings suggest a possibility for HIV transmission and the result of HIV testing of the source person is pending, a two-drug PEP regimen should be initiated until laboratory results become available. PEP should be discontinued if the source patient is determined to be HIV-seronegative. In addition to the evaluation and management of HIV exposure risk, all patients should be tested and treated for other highly infectious diseases, such as hepatitis. Patients often present to the ED seeking PEP because services are available at any time and early initiation of PEP is critical for efficacy. Many EDs are developing protocols and starter treatment packets for PEP. If possible, however, the choice of intervention and regimen usually is best accomplished in consultation with an infectious disease specialist and the patient’s primary physician, to allow arrangement for appropriate medical follow-up and counseling.
Nonoccupational Exposure Recent interest in the use of PEP for nonoccupational exposure has emerged, because the probability of HIV transmission by certain sexual or injection drug exposures is of the same order of magnitude as for percutaneous exposures, for which the CDC recommends PEP. The current DHHS recommendations regarding nonoccupational postexposure prophylaxis (nPEP) are as follows110: (1) For persons seeking care at 72 hours or earlier after nonoccupational exposure to blood, genital secretions, or other potentially infectious body fluids of a person known to be HIV-infected, when that exposure represents a substantial risk for transmission, a 28-day course of a HAART regimen is recommended, and antiretroviral medications should be initiated as soon as possible after exposure. (2) For persons seeking care at 72 hours or earlier after nonoccupational exposure to blood, genital secretions, or other potentially infectious body fluids of a person of unknown HIV status, when such exposure would represent a substantial risk for transmission if the source were HIV-infected, no recommendations are made for the use of nPEP. (3) For persons with an exposure history that represents no substantial risk for HIV
transmission or who seek care later than 72 hours after exposure, DHHS does not recommend the use of nPEP. (4) Clinicians may consider prescribing nPEP for exposures conferring a serious risk for transmission, even if the person seeks care later than 72 hours after exposure if, in their judgment, the diminished potential benefit of nPEP outweighs the risks for transmission and adverse events. All patients seeking care after HIV exposure should be tested for the presence of HIV antibodies at baseline and at 4 to 6 weeks, 3 months, and 6 months after exposure to determine whether HIV infection has occurred. In addition, testing for STDs, hepatitis B and C, and pregnancy should be offered. Early experience with implementation of these guidelines indicates that challenges exist in ensuring routine implementation.111 For most cases in which a patient with recent exposure is likely to have continuing risk for exposure, the CDC recommends providing basic risk reduction counseling and referral to risk reduction programs rather than offering PEP. Additional resources should be used whenever possible to assist with decision making and follow-up services; in-house infectious disease consultation should be sought. Other useful resources for information on both occupational and nonoccupational exposure include the CDC/University of California– San Francisco (UCSF) National Clinicians PEP Hotline (1-888-448-4911), providing 24-hour assistance, and the University of California at Los Angeles (UCLA)’s online decisionmaking support (http://www.needlestick.mednet.ucla.edu).
KEY CONCEPTS ■
The seroprevalence of HIV infection and AIDS among patients presenting to EDs serving large metropolitan areas is 2 to 15%. Many of these are undiagnosed cases, so compliance of ED personnel with universal precautions is extremely important. ■ Acute HIV seroconversion syndrome commonly follows exposure by 2 to 6 weeks and manifests with common nonspecific signs and symptoms such as fever, fatigue, diarrhea, weight loss, adenopathy, and rash. Patients fitting this profile should be screened for HIV risk factors and appropriately referred for HIV testing. ■ PCP is the most common opportunistic infection in patients with AIDS. It often manifests as progressive dyspnea on exertion associated with a nonproductive cough. The chest radiograph commonly shows a diffuse interstitial infiltrate but may be normal in appearance. Blood gas analysis usually reveals hypoxemia that often is more pronounced after exercise. ■ CNS disease is common in HIV-infected patients and may be caused by the disease itself, opportunistic infections, or malignancy. An algorithm for the evaluation of HIV-infected patients with severe or prolonged headache, altered mentation, new-onset seizures, or focal neurologic deficits is presented in Figure 130-2. ■ The evaluation and management of HIV-infected patients with acute symptoms often are complex and best accomplished either in the hospital or in the outpatient setting with close follow-up. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 131
Parasitic Infections
Bruce M. Becker and John D. Cahill
■ PERSPECTIVE Challenge Parasitology has become increasingly important in the practice of emergency medicine. Recent years have seen a dramatic increase in immigration from South East Asia, Central and South America, and Africa into the United States. Many of the people in transit have left their countries of origin under dire circumstances, fleeing civil unrest, war, famine, economic hardship, political persecution, and environmental devastation; they often lived in regions where parasitic infections were endemic. Business and adventure travel, including ecotourism, frequently transports immunologically naive and vulnerable hosts to sites rich in parasitic disease (Fig. 131-1). Patients with human immunodeficiency virus (HIV) infection or acquired immunodeficiency syndrome (AIDS) who travel to countries where parasitic illnesses are endemic are at higher risk of contracting these illnesses. Patients with AIDS who emigrate or travel to the United States or Europe may harbor a number of devastating parasitic illnesses. There is a significant prevalence of endemic parasitic disease in many rural areas of the southeastern and southwestern United States and in some parts of Europe. Often, patients with parasitic illness initially seek treatment in the emergency department (ED). Correct diagnosis and chemotherapy given early in the course of parasitic illness often result in rapid recovery (Table 131-1); mismanagement can be disastrous. Osler wrote, “Early in the course of disease, diagnosis is difficult and treatment easy; late in the course, diagnosis is easy and treatment difficult.” Parasitic illness often begins insidiously and, without appropriate treatment, commonly pursues a chronic course, eventuating in end-organ damage and severe morbidity and even death. To diagnose parasitic infection, the emergency physician will need to play detective, obtaining a thorough travel history, performing a detailed physical examination, ordering appropriate laboratory studies, and integrating these findings with a strong understanding of the basic life cycles of parasites, usual and unusual presentations of infection, and the intersecting geography of the organism and the host. An important point is that the incubation period for the development of symptoms for parasitic diseases ranges from days (falciparum malaria) to months (vivax malaria) to years (filariasis). Uncovering parasitic illness depends heavily on Osler’s principle—to make the diagnosis, one must first think of the diagnosis.
A detailed account of individual parasites can be found in Bell’s Tropical Medicine1 and Guerrant and colleagues’ Tropical Infectious Diseases.2
Travel History Parasitic illness should be considered in the differential diagnosis for almost every sign or symptom imaginable, particularly in patients who recently have spent time in areas of the world with endemic parasitic illnesses (Table 131-2). Accordingly, a travel history should be included in the evaluation of most if not all patients presenting to the ED. Important questions are summarized in Box 131-1. For patients who recently emigrated to the United States, the history should elicit additional information specific to the country of origin, also summarized in Box 131-1.
■ PRINCIPLES OF THERAPY New and more effective antiparasitic agents are continually being developed. The list of drugs used to treat parasitic infestations is large and varied (Table 131-3; see also Table 131-1). Table 131-3 includes some of the newest pharmaceutical agents, in addition to many medications that, although still recommended, have become almost obsolete because of toxi city or mediocre efficacy. The newer antiparasitic drugs are less toxic to the patient and more effective. Parasite biochemical pathways are sufficiently different from those in the human host to permit selective interference by relatively small doses of chemotherapeutic agents. In many instances, single-dose treatment can eradicate an entire parasite burden, and this approach has led to implementation of mass treatment programs in infected populations in endemic areas. Treatment and disposition in the ED focus on the individual patient and particular disease entity. The evolutionary goal of the successful parasite is to live with and at the expense of the living host; a parasite that kills its host has no survival advantage. Most parasitic infections (with certain important exceptions, such as falciparum malaria) pursue a chronic course and are not acutely life-threatening. Alterations in host immune function can change the virulence and morbid course of more benign infections (e.g., strongyloidiasis can become fulminantly disseminated in patients receiving immunosuppressive medication after organ transplantation or after the initiation of long-term steroid therapy). Despite the subacute or chronic nature of most parasitic infections, 1751
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1752
when a diagnosis (or a diagnostic plan) has been made and chemotherapy instituted, arrangements for careful follow-up and repeat laboratory examinations will be needed to ensure a cure. When the parasites are not eliminated promptly, repeat doses or alternative drugs should be considered, because drug resistance is becoming increasingly common. In such cases, referral to a geographic medicine clinic or an infectious disease clinic is indicated. Any patient who appears clinically ill or has presumptive falciparum malaria (by symptoms or travel history) should be admitted to the hospital for initial diagnosis, treatment, and observation.
■ FEVER Malaria Principles of Disease. The febrile patient with shaking chills and a time-appropriate history of travel to an endemic region should be evaluated for malaria. Plasmodium falciparum, Plasmodium ovale, Plasmodium vivax, and Plasmodium malariae are the species responsible for human malaria. More than 41% of the
world’s population lives in malaria-endemic areas (e.g., parts of Africa, Asia, Oceania, Central America, and South America). Approximately 300 million to 500 million clinical infections occur annually, resulting in 1.5 million to 2.7 million deaths.3 Approximately 1500 cases of malaria are diagnosed yearly in the United States. The female Anopheles mosquito is the arthropod vector that can transmit malaria after ingesting gametocytes from infected persons. After sexual reproduction in the gut of the mosquito, sporozoites are released from the salivary glands of the arthropod into the human host during a blood meal. Sporozoites rapidly penetrate the liver parenchymal cells of their host. The protozoans, now termed cryptozoites or exoerythrocytic schizonts, rapidly multiply. Eventual lysis of the hepatic cells results in the release of merozoites into the bloodstream, where they invade erythrocytes. In P. vivax and P. ovale infection, dormant hypnozoites can reside in hepatocytes; recrudescence of infections can occur many months to years later. After invading red blood cells (RBCs), the merozoites transform into trophozoites, which feed on the cells’ hemoglobin. Text continued on p. 1758
Table 131-1 Drug Classes and Modes of Action for Agents Used for Treatment of Parasitic Disease USEFUL IN THE TREATMENT OF
LIKELY TARGET(S) IN THE PARASITE
Thiabendazole Mebendazole Albendazole
Ascaris, Enterobius, hookworm, Strongyloides, Trichuris, hydatid disease (longterm therapy)
Tubulin polymerization
Ivermectin* (Stromectol)
Many nematodes of humans (except hookworms) Filariasis Onchocerciasis Schistosomes Most other flukes, such as Clonorchis, Paragonimus, Fasciolopsis (many tapeworms of humans)
GABA-sensitive neuromuscular interface
TYPE OF DRUG
EXAMPLE(S)
Anthelmintics
Trematodicides
Praziquantel (Biltricide)
Antiprotozoals
Metronidazole (Flagyl) Tinidazole Niridazole
Amebiasis Balantidiasis Giardiasis S. haematobium
Molecular electron transport systems Acetylcholine recycling systems
Antimalarials
Chloroquine phosphate (Aralen)
Many species of susceptible malaria
Chloroguanide Pyrimethamine Trimethoprim and combinations of antifolate and sulfa drug (e.g., sulfadoxine/ pyrimethamine [Fansidar])
Many species of susceptible malaria Various malaria species partially or totally refractory to chloroquine
Parasite digestive vacuole hemoglobinase Dihydrofolate reductase step in folate synthesis or incorporation of PABA in folic acid
Surface structure Carbohydrate metabolism
PROPOSED EFFECTS ON TARGETS
Blocks cellular structural integrity and egg production; secondary effects on mitochondrial fumarate reductase and on glucose uptake Flaccidity or contraction (a tightbinding drug effective at low dose)
Vacuolization and surface disruption followed by immune attacks by the host; contraction of the muscles due to flooding of calcium through a permeable tegument; initial increase of glucose metabolism followed by shutdown Failure to sustain energyproducing systems Binds to acetylcholinesterase, inactivating normal neuromuscular function Local pH is changed so that enzyme becomes inoperative Blocks normal folate synthesis and eventually one-carbon metabolism
*Available at present from CDC Drug Service, Centers for Disease Control and Prevention, Atlanta, Georgia 30333, telephone: 404-639-3670 (evenings, weekends, and holidays: 404-639-2888). GABA, γ-aminobutyric acid; PABA, para-aminobenzoic acid.
1753
Table 131-2 Parasites Causing Human Disease: Geographic Location and Portal of Entry
Protozoa Entamoeba histolytica Balantidium coli Giardia lamblia Trichomonas vaginalis Leishmania tropica Leishmania braziliensis Leishmania donovani Trypanosoma gambiense Trypanosoma rhodesiense Trypanosoma cruzi Plasmodium vivax Plasmodium malariae Plasmodium falciparum Nematodes Trichinella spiralis Trichuris trichiura Strongyloides stercoralis Necator americanus Ancylostoma duodenale Enterobius vermicularis Ascaris lumbricoides Wuchereria bancrofti Brugia malayi Onchocerca volvulus Loa loa Dracunculus medinensis Cestodes Taenia saginata Taenia solium 1. Adult worm 2. Cysticercus stage Echinococcus granulosus Echinococcus multilocularis Hymenolepis nana Hymenolepis diminuta Diphyllobothrium latum Trematodes Fasciola hepatica Fasciolopsis buski Clonorchis sinensis Opisthorchis felineus Opisthorchis viverrini Paragonimus westermani Schistosoma japonicum Schistosoma mansoni Schistosoma haematobium
GEOGRAPHIC DISTRIBUTION
COMMON INFECTIVE STAGE AND PORTAL OF ENTRY
Cosmopolitan, especially prevalent in warm climates Warm climates Cosmopolitan, especially prevalent in warm climates Cosmopolitan, United States Mediterranean area to western India Mexico to northern Argentina China, India, Africa, Mediterranean area, continental Latin America West and Central Africa Central and East Africa Continental Latin America Warm and cooler climates Warm climates Warm climates
Cyst via mouth Cyst via mouth Cyst via mouth Trophozoite via vulva or urethra Leptomonad via skin Leptomonad via skin Leptomonad via skin Trypanosome via skin Trypanosome via skin Trypanosome via skin Sporozoite via skin Sporozoite via skin Sporozoite via skin
Cosmopolitan, common in the United States Warm, moist climates Warm, moist climates Common in warm climates Western South America Cosmopolitan, common in the United States Cosmopolitan, common in the United States Prevalent in warm climates Asia Tropical Africa, Mexico, Central America, and northern South America
Encysted larva in pork via mouth Embryonated egg via mouth Filariform larva via skin Filariform larva via skin Filariform larva via skin Embryonated egg via mouth Embryonated egg via mouth Filariform larva via skin Filariform larva via skin Filariform larva via skin
Tropical West Africa Tropical Eastern Hemisphere
Filariform larva via skin Ingestion of larva via copepod via mouth
Cosmopolitan, United States
Cysticercus in beef via mouth
Cosmopolitan, United States Cosmopolitan, United States Cosmopolitan, United States Central Europe, Asia, Alaska
1. Cysticercus in pork via mouth 2. Eggs in human infections via mouth Eggs from canines via mouth Eggs from foxes via mouth
Warm climates Warm climates North Temperate Zone, Argentina, Chile, Australia
Eggs in human infections via mouth Larva in arthropod host via mouth Sparganum larva in fish flesh via mouth
Sheep-raising countries Asia Asia Europe, Asia Thailand Primarily Asia, also South America and Africa Asia Africa, Latin America Africa to India, southern Portugal
Larva on vegetation via mouth Larva on water nuts Larva encysted in freshwater fish Larva encysted in freshwater fish Larva encysted in freshwater fish Larva encysted in crabs or crayfish via mouth Cercarial larva in water via skin Cercarial larva in water via skin Cercarial larva in water via skin
Modified from Beaver PC, et al: Clinical Parasitology, 9th ed. Philadelphia, Lea & Febiger, 1984.
Chapter 131 / Parasitic Infections
PARASITE
1754
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
BOX 131-1 Comprehensive Travel History for Evaluation for Parasitic Disease in the Emergency Department Questions for All Patients ■ What were the exact dates of travel? ■ What countries did the patient visit? ■ How much time was spent in each country? ■ What was the patient doing in the country, and where was he or she living? ■ Was the patient a tourist, an adventure traveler, or a worker? ■ Did the patient stay in cities or rural villages? ■ Was the patient sleeping in hotels or tents? ■ Did the patient engage in protected or unprotected sexual intercourse? ■ What did the patient eat and drink? ■ What were the patient’s activities (e.g., swimming in freshwater leads to schistosomiasis)? ■ Did the patient receive prophylactic immunizations before travel? ■ Did the patient take malaria chemoprophylaxis and comply with the regimen? ■ Did the patient use mosquito repellent and netting? ■ Does the patient have underlying chronic medical problems? ■ What medications does the patient take? ■ When did symptoms start, and what has been the chronology of symptoms, particularly fever and diarrhea? Questions for Patients Who Are Recent Immigrants to the United States ■ When did the patient arrive and from where? ■ What acute and chronic illnesses did the patient have previously while living in the country of origin? ■ What treatment did the patient receive there? ■ If a refugee, what countries did the patient pass through, and what were the living conditions (especially relevant for persons who have lived in numerous refugee camps)? ■ What was the season during the patient’s stay or travel in the countries (e.g., monsoon versus dry)? ■ What animal exposures and bites has the patient experienced? ■ Has the patient had exposure to fresh water, in either work or recreational activities?
Table 131-3 Drug Regimens for Treatment of Parasitic Infections INFECTION
Amebiasis (Entamoeba histolytica) Asymptomatic DRUG OF CHOICE: ALTERNATIVES: Mild to moderate intestinal disease DRUG OF CHOICE: ALTERNATIVE: Severe intestinal disease, hepatic abscess DRUG OF CHOICE: ALTERNATIVE: Amebic meningoencephalitis, primary (Naegleria spp.) DRUG OF CHOICE: Anisakiasis (Anisakis) TREATMENT OF CHOICE: Ascariasis (Ascaris lumbricoides): roundworm DRUGS OF CHOICE: Balantidiasis (Balantidium coli) DRUG OF CHOICE:
ALTERNATIVES:
DRUG
ADULT DOSAGE
PEDIATRIC DOSAGE
Iodoquinol Diloxanide furoate or Paromomycin
650 mg tid × 20 days 500 mg tid × 10 days 25–30 mg/kg/day in 3 doses × 7 days
30 mg/kg/day in 3 doses × 20 days 20 mg/kg/day in 3 doses × 7 days 25–30 mg/kg/day in 3 doses × 7 days
Metronidazole Tinidazole
750 mg tid × 10 days 2 g/day × 3 days
35–50 mg/kg/day in 3 doses × 10 days 50 mg/kg (maximum 2 g) qd × 3 days
Metronidazole Tinidazole
750 mg tid × 10 days 600 mg bid or 800 mg tid × 5 days
35–50 mg/kg/day in 3 doses × 10 days 50 mg/kg or 60 mg/kg (maximum 2 g) qd × 3 days
Amphotericin B
1 mg/kg/day IV, uncertain duration
1 mg/kg/day IV, uncertain duration
Mebendazole or Pyrantel pamoate
100 mg bid × 3 days 11 mg/kg once (maximum 1 g)
100 mg bid × 3 days 11 mg/kg once (maximum 1 g)
Tetracycline
500 mg qid × 10 days
Iodoquinol Metronidazole
650 mg tid × 20 days 750 mg tid × 5 days
40 mg/kg/day in 4 doses × 10 days (maximum 2 g/day) 40 mg/kg/day in 3 doses × 20 days 35–50 mg/kg/day in 3 doses × 5 days
Surgical or endoscopic removal
1755
Table 131-3 Drug Regimens for Treatment of Parasitic Infections—cont’d DRUG
ADULT DOSAGE
Cutaneous larva migrans (creeping eruption) DRUG OF CHOICE:
Ivermectin
200 µg/per kg once daily × 1 or 2 days
Dracunculus medinensis (Guinea worm) DRUG OF CHOICE:
Metronidazole
750 mg tid × 5–10 days
Thiabendazole
50–75 mg/day in 2 doses × 3 days
Albendazole
A single dose of 400 mg; repeat after 2 wk A single dose of 100 mg; repeat after 2 wk
11 mg/kg once (maximum 1 g); repeat after 2 wk A single dose of 100 mg; repeat after 2 wk
ALTERNATIVE:
Enterobius vermicularis (pinworm) DRUGS OF CHOICE:
Mebendazole
PEDIATRIC DOSAGE
25 mg/kg/day (maximum 750 mg/day) in 2 doses × 10 days 50–75 mg/kg/day in 2 doses × 3 days
Filariasis Wuchereria bancrofti, Brugia malayi DRUG OF CHOICE:
Diethylcarbamazine
Day 1: 1 mg/kg PO Day 2: 1 mg/kg tid Day 3: 1–2 mg/kg tid Days 4–21: 6 mg/kg/day in 3 doses
Loa loa DRUG OF CHOICE:
Day 1: 50 mg PO Day 2: 50 mg tid Day 3: 100 mg tid Days 4–21: 6 mg/kg/day in 3 doses
Diethylcarbamazine
Day 1: 1 mg/kg PO Day 2: 1 mg/kg tid Day 3: 1–2 mg/kg tid Days 4–21: 6 mg/kg/day in 3 doses
Onchocerca volvulus DRUG OF CHOICE:
Day 1: 50 mg PO Day 2: 50 mg tid Day 3: 100 mg tid Days 4–21: 9 mg/kg/day in 3 doses
Ivermectin*
150 µm/kg PO once, repeated every 3–12 mo
150 µm/kg PO once, repeated every 3–12 mo
Fluke, hermaphroditic Clonorchis sinensis (Chinese liver fluke) DRUG OF CHOICE:
Praziquantel
75 mg/kg/day in 3 doses × 1 day
Fasciola hepatica (sheep liver fluke) DRUG OF CHOICE:
75 mg/kg/day in 3 doses × 1 day
Bithionol
Fasciolopsis buski (intestinal fluke) DRUG OF CHOICE:
30–50 mg/kg on alternate days × 10–15 doses
30–50 mg/kg on alternate days × 10–15 doses
Praziquantel
75 mg/kg/day in 3 doses × 1 day
Opisthorchis felineus DRUG OF CHOICE:
75 mg/kg/day in 3 doses × 1 day
Praziquantel
75 mg/kg/day in 3 doses × 1 day
Paragonimus westermani (lung fluke) DRUG OF CHOICE:
75 mg/kg/day in 3 doses × 1 day
Praziquantel
75 mg/kg/day in 3 doses × 2 days
Bithionol
75 mg/kg/day in 3 doses × 2 days 30–50 mg/kg on alternate days × 10–15 doses
Metronidazole Furazolidone Tinidazole
250 mg tid × 5 days 100 mg qid × 7–10 days 2 g as a single daily dose for 1–3 days
15 mg/kg/day in 3 doses × 5 days 6 mg/kg/day in 4 doses × 7–10 days 50 mg/kg as a single daily dose for 1–3 days
Albendazole Mebendazole or Pyrantel pamoate
400 mg × one dose 500 mg × one dose 11 mg/kg (maximum 1 g) × 3 days
500 mg × one dose 11 mg/kg (maximum 1 g) × 3 days
ALTERNATIVE:
Giardiasis (Giardia lamblia) DRUG OF CHOICE: ALTERNATIVES:
Hookworm infection (Ancylostoma duodenale, Necator americanus) DRUGS OF CHOICE:
30–50 mg/kg on alternate days × 10–15 doses
Chapter 131 / Parasitic Infections
INFECTION
1756
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
Table 131-3 Drug Regimens for Treatment of Parasitic Infections—cont’d INFECTION
DRUG
ADULT DOSAGE
PEDIATRIC DOSAGE
Leishmaniasis (Leishmania braziliensis, Leishmania mexicana, Leishmania tropica, Leishmania donovani [kala-azar]) DRUG OF CHOICE:
Miltefosine
2.5 mg/kg/day PO × 28 days
or Stibogluconate sodium Amphotericin B
Not indicated in those 12 years or younger 20 mg/kg/day IV or IM × 20–28 days 0.25–1 mg/kg by slow infusion daily or every 2 days for 8 wk
Chloroquine phosphate
600 mg base (1 g), then 300 mg base (500 mg) 6 hr later, then 300 mg base (500 mg) at 24 and 48 hr
10 mg base/kg (maximum 600 mg base), then 5 mg base/kg 6 hr later, then 5 mg base/kg at 24 and 48 hr
Quinine dihydrochloride
20 mg/kg loading dose in 10 mg/kg 5% dextrose over 4 hr followed by 10 mg/kg over 2–4 hr q8h (maximum 1800 mg/day) until oral therapy can be started 10 mg/kg loading dose (maximum 600 mg) in normal saline slowly over 1–2 hr, followed by continuous infusion of 0.02 mg/kg/min for 3 days maximum
Same as adult dose
200 mg base (250 mg) IM q6h if oral therapy cannot be started
0.83 mg base/kg/hr × 30 hr continuous infusion or 3.5 mg base/kg q6h IM or SC
Quinine sulfate plus Doxycycline or Clindamycin Mefloquine Atovaquone/proguanil
650 mg tid × 3 days 100 mg bid × 7 days 900 mg tid × 3–5 days 1250 mg once 1000/400 mg qd × 3 days
25 mg/kg/day in 3 doses × 3–7 days
Same as above
Same as above
Prevention of relapses: P. vivax and P. ovale only DRUG OF CHOICE:
Quinine dihydrochloride or Quinidine gluconate or Artesunate
Same as above Same as above
Same as above Same as above
Primaquine phosphate
15 mg base (26.3 mg)/day × 14 days or 45 mg base (79 mg)/ wk × 8 wk
0.3 mg base/kg/day × 14 days
Malaria, prevention of DRUG OF CHOICE:
Chloroquine phosphate
300 mg base (500 mg salt) PO, 5 mg/kg base (8.3 mg/kg salt) once a once a week beginning 1 wk week, up to adult dose of 300 mg before and continuing for base, same schedule as for adult 4 wk after last exposure
ALTERNATIVE:
Malaria, treatment of (Plasmodium falciparum, P. ovale, P. vivax, and P. malariae) All Plasmodium species except chloroquine-resistant P. falciparum Oral DRUG OF CHOICE:
Parenteral DRUGS OF CHOICE:
or Quinidine gluconate
ALTERNATIVE:
Chloroquine-resistant P. falciparum Oral DRUGS OF CHOICE:
ALTERNATIVE:
Parenteral
DRUGS OF CHOICE:
or Artesunate for treatment failure or adverse reactions from quinidine/ quinine, available from the CDC Chloroquine hydrochloride
20 mg/kg/day IV or IM × 20–28 days 0.25–1 mg/kg by slow infusion daily or every 2 days for 8 wk
Same as adult dose
20–40 mg/kg/day in 3 doses × 3–5 days 25 mg/kg once (45 kg: 1 tablet
or Doxycycline
>8 yr: 2 mg/kg/day PO, up to 100 mg/day
Schistosomiasis Schistosoma haematobium DRUG OF CHOICE:
Praziquantel
40 mg/kg/day in 2 doses × 1 day
Schistosoma japonicum DRUG OF CHOICE:
40 mg/kg/day in 2 doses × 1 day
Praziquantel
60 mg/kg/day in 3 doses × 1 day
Schistosoma mansoni DRUG OF CHOICE:
60 mg/kg/day in 3 doses × 1 day
Praziquantel
40 mg/kg/day in 2 doses × 1 day 15 mg/kg once
40 mg/kg/day in 2 doses × 1 day
Praziquantel
60 mg/kg/day in 3 doses × 1 day
60 mg/kg/day in 3 doses × 1 day
Thiabendazole or Ivermectin
50 mg/kg/day in 2 doses (maximum 3 g/day) × 2 days 200 µg/kg/day × 1–2 days
50 mg/kg/day in 2 doses (maximum 3 g/day) × 2 days 200 µg/kg/day × 1–2 days
Praziquantel
5–10 mg/kg once
5–10 mg/kg once
Praziquantel
25 mg/kg once
25 mg/kg once
Albendazole
400 mg bid × 28 days, repeated as necessary
15 mg/kg/day × 28 days, repeated as necessary
50 mg/kg/day in 3 doses × 15 days
50 mg/kg/day in 3 doses × 15 days
200–400 mg tid × 3 days, then 400–500 mg tid × 10 days
Same as adult
Metronidazole
2 g once or 250 mg tid or 375 mg bid PO × 7 days
15 mg/kg/day PO in 3 doses × 7 days
Mebendazole or Albendazole
100 mg bid × 3 days 400 mg once
100 mg bid × 3 days 400 mg once
ALTERNATIVE: Schistosoma mekongi DRUG OF CHOICE: Strongyloidiasis (Strongyloides stercoralis) DRUGS OF CHOICE:
Tapeworm infection—adult (intestinal stage) Diphyllobothrium latum (fish), Taenia saginata (beef), Taenia solium (pork), Dipylidium caninum (dog) DRUG OF CHOICE: Hymenolepis nana (dwarf tapeworm) DRUG OF CHOICE: Tapeworm infection—larval (tissue) stage Echinococcus granulosus (hydatid cysts) DRUG OF CHOICE:
Oxamniquine
Echinococcus multilocularis TREATMENT OF CHOICE: Cysticercus cellulosae (cysticercosis) DRUG OF CHOICE:
Praziquantel
Surgery
ALTERNATIVE:
Trichinosis (Trichinella spiralis) DRUGS OF CHOICE:
Trichomoniasis (Trichomonas vaginalis) DRUG OF CHOICE: Trichuriasis (Trichuris trichiura, whipworm) DRUGS OF CHOICE:
20 mg/kg/day in 2 doses × 1 day
Surgical excision
Steroids for severe symptoms plus Mebendazole
Chapter 131 / Parasitic Infections
INFECTION
1758
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
Table 131-3 Drug Regimens for Treatment of Parasitic Infections—cont’d INFECTION
DRUG
ADULT DOSAGE
PEDIATRIC DOSAGE
Trypanosomiasis Trypanosoma cruzi (South American trypanosomiasis, Chagas’ disease) DRUG OF CHOICE:
Nifurtimox
8–10 mg/kg/day PO in 4 doses × 120 days
ALTERNATIVE: Trypanosoma brucei gambiense, Trypanosoma brucei rhodesiense (African trypanosomiasis, sleeping sickness) hemolymphatic stage DRUG OF CHOICE:
Benznidazole
5–7 mg/kg/day × 30–120 days
1–10 yr: 15–20 mg/kg/day in 4 doses × 90 days 11–16 yr: 12.5–15 mg/kg/day in 4 doses × 90 days Same as adult
Suramin
Pentamidine isethionate
100–200 mg (test done) IV, then 1 g IV on days 1, 3, 7, 14, and 21 4 mg/kg/day IM × 10 days
ALTERNATIVE:
Late disease with central nervous system involvement DRUG OF CHOICE:
ALTERNATIVES:
only)
(T. b. gambiense
Melarsoprol
Tryparsamide
plus Suramin
Visceral larva migrans (toxocariasis) DRUG OF CHOICE:
ALTERNATIVES:
Diethylcarbamazine Mebendazole or Albendazole
20 mg/kg on days 1, 3, 7, 14, and 21 4 mg/kg/day IM × 10 days
2–3.6 mg/kg/day IV × 3 days; 18–25 mg/kg total over 1 mo; initial dose after 1 wk 3.6 mg/kg/day IV of 0.36 mg/kg IV, increasing gradually × 3 days; repeat again after to maximum 3.6 mg/kg at intervals of 1–5 days for total of 9–10 doses 10–21 days One injection of 30 mg/kg Unknown (maximum 2 g) IV every 5 days to total of 12 injections; course may be repeated after 1 mo One injection of 10 mg/kg IV Unknown every 5 days to total of 12 injections; course may be repeated after 1 mo
6 mg/kg/day in 3 doses × 7–10 days 100–200 mg bid × 5 days 400 mg bid × 3–5 days
6 mg/kg/day in 3 doses × 7–10 days Same as adult 400 mg bid × 3–5 days
*Some drugs available from CDC Drug Service, Centers for Disease Control and Prevention, Atlanta, 30333, telephone: 404-639-3670 (nights, weekends, and holidays: 404-639-2888). Modified from Drugs for parasite infections. Med Lett Drug Ther 37:99, 1995.
Trophozoites mature into schizonts, which may divide asexually into additional merozoites. The RBCs undergo lysis, releasing merozoites into the blood. Although some merozoites are destroyed by the host’s immune apparatus, many enter new erythrocytes. After several repetitions of this erythrocytic cycle, the cyclic process changes, and male microgametocytes or female macrogametocytes may develop instead of merozoites. These gametes subsequently complete the reproductive cycle by fusion, accomplished sexually, within the gut of a new female Anopheles mosquito after she has taken a blood meal from an infected host. Most people contract malaria after being bitten by an infected vector mosquito in an endemic region. Other means of transmission have been reported, including blood transfusions, injection drug use with contaminated syringes, perinatal transmission, organ transplantation, and socalled airport malaria. Airport malaria has been reported in people who have never been in an endemic area but live near or work in an international airport. The infected mosquito is transported from the endemic region and released when the plane arrives at its destination.4,5 Clinical Features. Most patients with malaria present with irregular fevers. Other signs and symptoms include anemia, headache, nausea, chills, lethargy, abdominal pain, and upper
respiratory complaints.6 The important difference between P. falciparum and the other malaria species is the capacity of P. falciparum to cause severe organ system damage and death. Acute falciparum infection may have any of the following manifestations: cerebral malaria with cerebral edema and encephalopathy (Fig. 131-2), hypoglycemia (especially in children), metabolic acidosis, severe anemia, renal failure, pulmonary edema, disseminated intravascular coagulation, and death. In chronic malaria, hepatosplenomegaly may develop because of increased cellularity from the host’s immune response. Within the liver, parasites and malarial pigment distend the Kupffer cells. Parasitized RBCs also adhere to the sinusoidal system of the spleen, reducing its immunologic effectiveness. Anemia results from acute and chronic hemolysis. So-called blackwater fever—hemoglobinuria caused by severe hemolysis—may occur in patients with either chronic or acute falciparum malaria. Diagnostic Strategies. Light microscopic examination of thick and thin blood films is the “gold standard” modality for the diagnosis of malaria. The clinician may have to view several slides to make the diagnosis if the parasite burden is small. Peripheral blood smears are stained with Giemsa or Wright stains and examined with ordinary light microscopy. The diag-
1759 Schistosoma Schistosoma Schistosoma Clonorchis Paragonimus haematobium mansoni japonicum sinensis westermani
0
50
100
Chapter 131 / Parasitic Infections
Red blood cell
Fasciola and Fasciolopsis Taenia spp. hepatica buski
150
200 µm
Ascaris lumbricoides
Normal
Decorticated Embryonated
Unfertilized
Ancylostoma duodenale
Trichuris Enterobius Fresh Developed trichiura vermicularis Figure 131-1. Eggs of major parasitic worms causing disease in humans.
nosis can be made in a simply equipped laboratory. Even if the parasite is not visualized in the smear, treatment for malaria is nevertheless indicated if the disease is suspected on clinical grounds. The U.S. Food and Drug Administration (FDA) has approved the use of an antigen-based rapid diagnostic test for screening patients. Microscopy must still be performed in all patients with positive results on such tests, to determine species and the severity of parasitemia. Management. In the past, chloroquine phosphate was the treatment of choice for acute, uncomplicated attacks of malaria. Resistance to chloroquine has been steadily increasing, and now the drug should be used only in regions of known chloroquine sensitivity: Haiti, Dominican Republic, Central America, and limited regions of the Middle East. For uncomplicated malarial infections in patients from chloroquineresistant regions, oral quinine and doxycycline given together may be used. Another suitable alternative combination is proguanil-atovaquone. For complicated P. falciparum infection (e.g., cerebral malaria, involvement of multiple organ systems, unability to tolerate oral medication), intravenous quinine (not available in intravenous form in the United States) or quinidine is used. Rapid infusion of intravenous quinine can cause profound hypoglycemia. Patients should not receive intravenous quinine without cardiac monitoring. The artemisinin agents are excellent antimalarials and are available in enteral and parenteral preparations. They have a rapid onset of action and are well tolerated. Most are not available in the United States; however, artesunate (which is an artemisinin agent) is now available as an investigational new drug for patients who have complicated malaria not responding to quinidine. To obtain this drug, physicians can contact the Centers for Disease Control and Prevention (CDC) Malaria Hotline at 770-488-7788 or, during off hours, at 770-448-7100.7
Figure 131-2. Patient in coma from falciparum cerebral malaria. Primaquine is used to expunge the hepatic phases of P. ovale and P. vivax, to prevent recrudescent disease. Before starting primaquine therapy, the patient must be tested for glucose-6phosphate dehydrogenase enzyme deficiency, to avoid precipitating severe hemolysis. Prompt diagnosis and appropriate therapeutic intervention are essential for patients infected with falciparum malaria, to prevent coma and death.8,9
Babesiosis Babesiosis is a malaria-like illness that is becoming increasingly prevalent in the northeastern United States (Babesia microti), the northwestern United States (Babesia gibsoni), and Europe (Babesia divergens). Babesiosis is endemic on Martha’s Vineyard and Nantucket. The organism is a protozoan similar in structure and life cycle to the plasmodia. It is transmitted by the deer tick, Ixodes dammini, the vector of Lyme disease. Several cases have been attributed to transfusions with infected blood.10 Patients with babesiosis experience fatigue, anorexia, malaise, and emotional lability, with myalgia, chills, high spiking fevers, sweats, headache, and dark urine. Other manifestations include hepatosplenomegaly, anemia, thrombocytopenia, leukopenia, elevated liver enzyme levels, and signs of hemolysis with hyperbilirubinemia and decreased haptoglobin. In an otherwise healthy person, the disease may remit spontaneously. In asplenic, elderly, and immunocompromised patients (especially patients with AIDS and patients taking corticosteroids), 85% of RBCs may contain organisms. Clinical syndromes include massive hemolysis, jaundice, renal failure, disseminated intravascular coagulation, hypotension, and adult respiratory distress syndrome.11 Diagnosis is based on clinical suspicion, multiple thin and thick blood smears, and serologic testing (results of which may not turn positive for several weeks after the infection). The treatment of choice consists of quinine plus clindamycin. Patients infected with B. divergens tend to be sicker and require more supportive care. Co-infection with Borrelia burgdorferi, the agent of Lyme disease, results in more severe and prolonged illness.12 Babesia organisms resemble plasmodia in blood smears.
1760
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
Other Parasites Causing Fever Other parasitic illnesses that commonly cause significant fever include schistosomiasis, fascioliasis, African and American trypanosomiasis, leishmaniasis, toxoplasmosis, and amebic liver abscess. Katayama fever may be the initial phase of schistosomiasis. Infected patients report brief exposures to fresh water in endemic areas. Clinical manifestations include spiking fevers, diaphoresis, and cough. Eosinophilia is common.13 Fascioliasis, caused by the liver fluke, Fasciola hepatica, is endemic throughout Asia, the former Soviet Union, southern Europe, and South America. Infection begins with ingestion of the metacercariae often found in watercress. Within 6 weeks, patients exhibit right upper quadrant abdominal pain, fever, and eosinophilia.14 American trypanosomiasis (Chagas’ disease) is endemic to Central and South America. The vector, the reduviid bug, sheds trypomastigotes in its feces proximal to the bite site, leading to local infection and subsequent systemic spread in the host. Acute Chagas’ disease begins with the chagoma, the infected and swollen bite site often periorbital in location, and quickly progresses to fever, malaise, facial swelling, and pedal edema. Parasitization of cardiac muscle leads to the dysrhythmias and ventricular dysfunction that are classically found in late disease (chronic Chagas’ cardiopathy).15 Leishmaniasis is spread to humans by the sandfly and is found in the Middle East, India, and East Africa; along the Mediterranean coast; and in Brazil. Although leishmaniasis can involve the skin (cutaneous) and the mucosa (mucosal), fever is seen only in visceral leishmaniasis in immunocompetent persons. Signs and symptoms also include massive hepatosplenomegaly, neutropenia, and weight loss.16 Amebic liver abscesses frequently manifest with high fevers, right upper quadrant pain, and an elevated white blood cell count.17
■ NEUROLOGIC SYMPTOMS Cerebral Malaria Principles of Disease and Clinical Features. Cerebral malaria is a common, life-threatening complication of P. falciparum infection. Parasitized RBCs express malarial cell surface glycoproteins called knobs that are sticky. They adhere to capillary walls, causing sludging in the cerebral microvasculature, localized ischemia, capillary leak, and petechial hemorrhages. Clinical manifestations include fever, altered mentation including obtundation and coma (see Fig. 131-2), and occasionally seizures. A careful history and early diagnosis and therapy are essential to prevent severe morbidity and death. Management. Treatment of cerebral malaria consists of intravenous quinine, quinidine, or artemisinin (if available); supportive care, including mechanical ventilation for comatose patients and patients with noncardiogenic pulmonary edema; antiepileptics; and correction of acidosis and hypoglycemia (associated with quinine use and cerebral malaria). The mortality rate is high, especially in children (30%), but if the patient recovers, neurologic sequelae are rare (seen in less than 10%).18,19 Corticosteroids, including dexamethasone, are not beneficial and are potentially harmful in cerebral malaria.
gut wall and end up anywhere in the body. The most common sites include the CNS, muscle, and soft tissue.20,21 Clinical Features. In the brain, the cluster of larvae of T. solium form an expanding cyst that induces an intense immunologic reaction from the host, including inflammation, fibrosis, and ultimately calcification. Neurologic abnormalities develop when the involved neural tissue cannot accommodate the enlarging cyst. Seizure activity often is the first indication of cysticercosis, which should be considered in any adult patient with undiagnosed seizures. The diagnosis of T. solium infection is established by finding characteristic proglottids (gravid segments) or scolices (worm heads) in stool preparations. Diagnostic Strategies and Management. Cranial computed tomography (CT) scan with contrast or magnetic resonance imaging (MRI) may reveal an enhancing ring lesion. These lesions can mimic a CNS abscess, metastasis, or a primary tumor such as glioblastoma multiforme. Albendazole is the therapeutic agent of choice, and corticosteroids may be necessary during therapy, particularly if CNS cysts are present.22 Neurosurgical consultation should be sought in treating neurocysticercosis because acute obstructive hydrocephalus can occur.
Echinococcosis Principles of Disease and Clinical Features. Echinococcus granulosus is another tapeworm capable of causing CNS disease. Cerebral hydatid cysts are loculated structures containing E. granulosus scolices (heads) and remains of germinal epithelium, termed hydatid sand. Common types of exposure include ingestion of food or water contaminated by the ova from feces of sheep or cattle infected by the adult worm or close contact with a sheepherding dog. Infection results in the liberation of the embryo oncosphere into the small intestine. After penetrating the intestinal wall, the larvae travel through the bloodstream to multiple sites for encystment. The liver is the target organ in nearly two thirds of cases, but 7% of patients have brain involvement; infected patients may present with seizures or focal neurologic signs. Diagnostic Strategies and Management. The diagnosis of hydatid cyst disease is suggested by the appearance and localization of the cyst on ultrasound or CT scan. Serologic evaluation of serum or cerebrospinal fluid (CSF) may help confirm the diagnosis. Aspiration of the cyst should not be attempted because of the risk of seeding the host’s body with metastatic cysts. Treatment options include albendazole and surgical resection. Resection of the cyst may cause an anaphylactoid reaction if there is spillage of hydatid sand23 (Figs. 131-3 and 131-4).
Cysticercosis Principles of Disease. Cysticercosis is caused by the larval form of Taenia solium, a common central nervous system (CNS) pathogen in many tropical areas. It is acquired by humans who eat pork containing the larval cysts. The adult worm matures in the small intestine; the larval forms may penetrate through the
Figure 131-3. Hydatid cysts removed surgically.
1761
Chapter 131 / Parasitic Infections
Figure 131-4. Additional hydatid cysts removed surgically.
African Trypanosomiasis Principles of Disease. African sleeping sickness is caused by Trypanosoma brucei gambiense and Trypanosoma brucei rhodesiense. This infection is endemic in limited areas of West and East Africa.24 Several recent cases have been reported in returned travelers who were on safari in East Africa. The motile organisms are transmitted by the bite of the Glossina (tsetse) fly, which introduces the infective form of the trypanosome into the host’s blood. A small lesion or boil may develop and persist for several days. The flagellated organism travels throughout the bloodstream, invading the lymph nodes and spleen. Clinical Features. Winterbottom’s sign, which is posterior cervical lymphadenopathy, usually is apparent at the time treatment is sought. The patient often is febrile, and trypanosomes are visible in a thick peripheral blood smear. A maculopapular rash may be noted in fair-skinned persons. After invasion of the CNS, severe headache may result from cerebral inflammation. Patients may display psychiatric symptoms, progressing eventually to extreme sleepiness and lethargy. Coma and death from starvation and trypanotoxins are inevitable in untreated patients.25,26 Diagnostic Strategies and Management. Diagnostic suspicion is aroused by an appropriate exposure history and characteristic symptoms. Trypanosomes in peripheral blood, CSF, or lymph node and bone marrow aspirates establish the diagnosis. The presence of parasites in the CSF indicates advanced progression of the disease. Suramin sodium is the treatment of choice for early infection with T. b. rhodesiense. Pentamidine isethio nate is the preferred treatment for early T. b. gambiense. Trivalent arsenicals, such as melarsoprol, which can penetrate the blood-brain barrier, are used in advanced disease with neurologic sequelae.2
Other Parasites Causing Neurologic Symptoms CNS involvement with Trichinella spiralis has been reported in severe cases, after larval migration of this parasite into the brain and meninges. Serious consequences are meningitis, encephalitis, seizures, paresis, coma, and death. The pathophysiology may reflect obstruction of small arterioles by migrating larvae, with subsequent vasculitis or cerebral edema from immunologic reaction to the larvae or larval fragments. Therapy for trichinosis with severe muscle or CNS involvement includes mebendazole or thiabendazole coupled with steroids, which depress the host immune response to infection.27,28 Amebic abscess of the brain or meningoencephalitis caused by Entamoeba histolytica is a rare complication of infection with
Figure 131-5. Five-year-old child with severe life-threatening anemia (hematocrit of 9) in association with chronic malaria. this intestinal parasite. Infestation occurs with ingestion of food or drink contaminated with cysts of this protozoan. Spread of amebae to the brain or meninges from the colonized large bowel wall is rare but should be considered in any patient with amebiasis and subsequent neurologic impairment. The diagnosis may be made by microscopic identification of trophozoites (motile amebae) in CSF; however, biopsy of affected tissue is more specific. CNS amebiasis is treated with in travenous metronidazole but may require neurosurgical intervention. Naegleria and Acanthamoeba are free-living amebae in fresh water that can be acquired while swimming and diving in ponds and lakes. The amebae invade the CNS through the olfactory neuroepithelium or cornea (violated by abrasion or associated with contact lens wear), causing an amebic meningoencephalitis. The combination of amphotericin B and miconazole is the pharmacologic regimen of choice when these motile amebae are identified in CSF.29 Strongyloides stercoralis infection is a common disease in the tropics. The worm enters through the skin and migrates to the small bowel. Infection with Strongyloides is more clinically significant in an immunosuppressed patient, who may suffer larval dissemination throughout the body with subsequent encephalitis and pyogenic meningitis in the CNS. Strongyloides infection is treated with thiabendazole or albendazole.30 Granulomas may occur in the brain from egg deposition by Schistosoma. Generally, they do not cause major symptoms; however, several cases of transverse myelitis with paraplegia have been reported.
■ ANEMIA Malaria Malaria infection often is associated with anemia, especially in children younger than 5 years of age (Fig. 131-5). Anemia may develop quickly, from massive hemolysis in acute infection, or it may have a more insidious onset, developing over months. Mature merozoites lyse parasitized RBCs. Uninfected RBCs
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1762
undergo immune destruction from cell surface antibodies produced in response to parasite-associated changes in RBC surface proteins. This process of destruction is abetted by increased reticuloendothelial activity. The reticulocyte response in infected persons is blunted by inhibition of erythropoietin secretion.31,32 The antimalarial drug primaquine can precipitate hemolysis in patients who have glucose-6phosphate dehydrogenase deficiency, which is common in black Africans and some Asians.
Whipworm and Hookworm Infestation by the whipworm Trichuris trichiura and, especially, the two human hookworms Necator americanus and Ancylostoma duodenale is a major cause of iron deficiency anemia worldwide. Adult worms penetrate into intestinal mucosa and feed, causing significant ongoing luminal blood loss. The host defecates eggs that mature in the soil through a rhabditiform larval form to the infective filariform larva. These larvae penetrate the human skin, usually through the feet. In trichuriasis, anemia is seen only with massive parasite infestation. Ova from the whipworm are ingested through stool-contaminated food and water. Diagnosis of these infections requires identification of characteristic ova in the stool. As with most helminthic infections, peripheral eosinophilia is common. Mebendazole or albendazole effectively controls trichuriasis and hookworm infections in adults and children. Anemic patients should receive iron supplementation.
Tapeworm Infection with the fish tapeworm Diphyllobothrium latum is associated with pernicious anemia. This tapeworm competes with the human host for absorption of vitamin B12. When the host ingests raw freshwater fish that contains the embryo plerocercoid larvae in its muscle fibers, the large adult tapeworm develops within the human small intestine. The diagnosis is made by identifying the ova in the feces. Praziquantel is the drug of choice in adults and children.
■ PERIPHERAL EDEMA Elephantiasis Principles of Disease. Elephantiasis, or filariasis, is manifested in the host by the development of massive peripheral edema with distention and thickening of the overlying skin, which acquires the appearance and texture of elephant skin. It is caused by infection with the filarial worm Wuchereria bancrofti or Brugia malayi. The infection is confined to humans and is widely distributed in the equatorial regions of the world, including Africa, Asia, South America, and Oceania. More than 90% of all infections are found in Asia, where the disease has reached epidemic proportions. Even in endemic regions in which most residents are infected, the disease is rare among travelers. Infected mosquitoes introduce microfilariae into the bloodstream of the human host during a blood meal. After infecting the host, the worms migrate into the lymphatic system and mature into coiled, gravid adults. The adult worm triggers a robust inflammatory reaction in the lymphatic vessels, particularly in the lower extremities and genitalia. The macrophages, lymphocytes, plasma cells, giant cells, and eosinophils migrate to the inflamed and fibrotic lymphatic vessel, which becomes erythematous, edematous, and tender, suggesting the diagnosis of filariasis. Clinical Features. Chronic manifestations of filariasis include fibrosis of a lymphatic vessel containing a dead or calcified
worm. Subsequent mechanical blockage of the lymphatic system leads inevitably to severe lower extremity and genital edema accompanied by thickening of the skin. Recurrent cellulitis is common in these patients; prevention requires meticulous skin care.33 Diagnostic Strategies and Management. The adult female worm produces microfilariae, which reach the peripheral blood through the lymphatics, whereupon the patient experiences shaking chills and fever. Thick peripheral blood smears may show infection, particularly at night, when the release of microfilariae is most likely. Diethylcarbamazine rapidly clears the microfilariae from the peripheral blood and slowly sterilizes the gravid female nematode. Established elephantiasis of the scrotum can be successfully treated surgically. Chronic lymphatic obstruction of the limbs rarely responds to operative intervention.34
■ DERMATOLOGIC SYMPTOMS Cutaneous Leishmaniasis Principles of Disease. Cutaneous leishmaniasis is one of the most important causes of painless chronic ulcerating skin lesions in the world. Leishmania braziliensis and Leishmania mexicana responsible for New World leishmaniasis, whereas Leishmania tropica and Leishmania major commonly cause Old World leishmaniasis. The female Phlebotomus sandfly transmits the promastigotes during a blood meal, which are ingested by host macrophages and survive in their leishmanial form in the skin. Clinical Features. Skin papules and nodules are seen early in the course of infection at the site of the insect bite. A raised macule also can appear, which subsequently develops painless central ulceration and a raised border. Lymphocyte and macrophage invasions of the epidermis and dermis cause the induration that occurs at the ulcer border. Secondary bacterial infections of these ulcers increase the associated scarring. L. braziliensis braziliensis (subspecies of L. braziliensis) attacks the mucocutaneous skin borders (i.e., in tissues of the nose and mouth). Mutilation of the face occurs after massive tissue and nasal cartilage destruction. The larynx and trachea also can be involved, compromising the airway. Disseminated cutaneous leishmaniasis (L. mexicana amazonensis in South America and L. tropica aethiopica in Ethiopia) is characterized by diffuse nodules and papules resembling those of lepromatous leprosy (Fig. 131-6). Persons with this manifestation of leishmaniasis are thought to have a defect in their cell-mediated immunity response.35,36
Figure 131-6. Cutaneous leishmaniasis.
Dracunculiasis Principles of Disease and Clinical Features. Dracunculus medinensis, the “fiery serpent,” appears in the host as the adult worm migrates through the subcutaneous tissues of the leg. The head of the gravid adult female erodes through the skin of the leg and releases larvae into the water when the host wades in a pond or open well. The larvae promptly infect the Cyclops water flea. Humans who drink water containing the infected crustacean complete the cycle of infection. The patient may complain of rash, intense pruritus, nausea, vomiting, dyspnea, and diarrhea before the female worm erupts through the skin. Management. The “classic” treatment in developing countries has been to wind the worm around a stick and slowly extract the parasite from the skin over the course of 1 or 2 days. If the worm breaks while being extracted, the patient experiences an intense inflammatory reaction with cellulitis along the worm tract. The diagnosis is confirmed when microscopic larvae are found in the fluid of the cutaneous ulcer or when the adult female worm is identified extruding from the skin. The use of metronidazole to shorten the time of extraction is controversial. The World Health Organization set a goal to eradicate this disease through public health awareness: encouraging covering wells, filtering well water to remove the fleas, and keeping infected persons with active skin lesions out of potable water. These efforts have had a tremendous impact on the eradication of dracunculiasis from Africa.
Other Parasites Causing Dermatologic Symptoms Cutaneous larva migrans, the “creeping eruption,” occurs in the host’s epidermis when the skin is penetrated by Ancylostoma braziliense (dog or cat hookworm) larvae. Exposure usually occurs after walking barefoot or lying on beaches or other warm soil contaminated by animal feces. The diagnosis is suggested by the presence of a characteristic meandering erythematous track on the skin surface caused by larval migration. Visceral larva migrans occurs in young children after ingestion of soil containing ova from the dog ascarid Toxocara canis. Thiabendazole, ivermectin, or albendazole may be used for treatment of cutaneous larva migrans, and antipruritics give symptomatic relief. Diethylcarbamazine treats visceral larva migrans. An alternative is thiabendazole.38 “Swimmer’s itch” is a dermatitis that occurs when skin is penetrated by the nonhuman schistosome of avians and mammals, usually from swimming in northern U.S. freshwater lakes. The infection spontaneously resolves when the nonhuman schistosome is destroyed by the human host’s immune system. A similar dermatitis also can occur after infection with schistosome species that are trophic for humans. Treatment is symptomatic.
1763
Strongyloides can cause a transient, pruritic rash that may appear and then disappear within hours. Taenia solium can cause cysts in the soft tissues and muscles. These cysts often are an incidental finding. Onchocerciasis (from Onchocerca volvulus), which is commonplace in West Africa and parts of South America, can cause severe pruritus and the development of nodules on bony protuberances.
■ VISUAL SYMPTOMS Onchocerciasis Principles of Disease. Onchocerciasis is a major cause of blindness in the world. Ninety-five percent of all cases occur in Africa.39 The parasite is found only in humans and is transmitted by the bite of the Simulium fly. These flies live near rivers—hence the common name of the disease, “river blindness.” Micro filariae of O. volvulus are released by adult nematodes, which coil in subcutaneous nodules in the infected host; the microfilariae then migrate through the dermis and epidermis. The presence of adult worms stimulates a brisk immune response, including the infiltration of lymphocytes, macrophages, plasma cells, and eosinophils. Clinical Features. The skin becomes chronically edematous and pruritic; it then atrophies, resulting in loose, thin folds of skin. River blindness is more likely to develop in patients with nodules in proximity to the eyes. When the microfilaria dies during its migration in the eye, the foreign tissue that is deposited in the iris musculature incites an immune sclerosing keratitis, which is a major cause of the ocular destruction with subsequent blindness (Fig. 131-7). Diagnostic Strategies and Management. The diagnosis of onchocerciasis requires identification of characteristic microfilariae in skin snipped from the patient. Ivermectin is the therapeutic drug of choice. In many countries in which the disease is endemic, the manufacturers of ivermectin have donated the drug in an attempt to eradicate the disease. Surgical excision of the subcutaneous nodules is recommended when they are located on the head.
Loiasis Principles of Disease and Clinical Features. Another filarial infection that causes ocular problems is loiasis. Loiasis is confined to forest areas in West and Central Africa. Transmission of Loa loa occurs through the bite of flies of the genus Chrysops. The edema initially associated with migration of the worm is called a Calabar swelling. The disease is caused by a migrating adult worm in the subcutaneous tissue. The adult worm occasionally
Figure 131-7. Patient with onchocerciasis, or “river blindness.”
Chapter 131 / Parasitic Infections
Diagnostic Strategies and Management. Definitive diagnosis of leishmaniasis is made by direct visualization of the parasite with light microscopy. Diagnosis also can be made by an indirect fluorescent antibody test. Results of intradermal skin testing often are negative during the acute stages of the disease. Many forms of cutaneous leishmaniasis, especially L. tropica and L. mexicana infection, are self-limited and require no treatment, unless the wounds become secondarily infected. Treatment options for advanced disease include sodium stibogluconate, meglumine antimonate, and amphotericin B. An oral drug, miltefosine, has been used with success for treating both the visceral and the cutaneous forms of leishmaniasis.37 These treatments rarely are initiated in the ED setting.
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1764
migrates through the subconjunctival tissues of the eye and can be surgically excised from the conjunctiva. Although upsetting to the patient, the disease often is fairly benign. The adult worm releases sheathed microfilariae into the peripheral bloodstream during the daytime. Diagnostic Strategies and Management. Microfilariae can be detected within a thick blood smear, securing the diagnosis of loiasis. The treatment of choice for L. loa infection is diethylcarbamazine. Corticosteroids or antihistamines often must supplement specific chemotherapy because of the intense allergic reaction that occurs when the killed adult worms and microfilariae disintegrate.40,41
Other Parasites Causing Ocular Symptoms T. canis has a trophism for the host’s eyes. Toxocariasis is a roundworm infection found in urban dogs. Humans ingest eggs by the fecal-oral route. The larvae migrate and often enter the retina, where they become trapped. They stimulate an immune response that culminates in granuloma formation. These granulomas can impair vision and sometimes are mistaken for retinal tumors. There is no means of direct diagnosis except tissue biopsy. Although serologic tests are available, results need to be interpreted with caution. Infection is treated with albendazole and steroids; larvae visible in the retina can be destroyed with a laser. Toxoplasma gondii infection can precipitate a vitreal inflammation with retinal hemorrhages. Immunocompromised patients may develop chorioretinitis and optic neuritis with visual field defects and ocular palsies. Erythrocytes with sticky “knobs” from P. falciparum infection can cause retinal vascular congestion and ischemia with hemorrhage, exudate, infarction, and macular destruction. Cerebral malaria can produce cortical blindness. Mucocutaneous leishmaniasis can involve the eyelids, tear glands, retina, or iris and may result in total ocular destruction. Acanthamoeba can cause a dangerous keratitis in contact lens wearers. The patient complains of severe pain, tearing, and photophobia. Early infection may be misdiagnosed as herpetic keratitis. The infection may become chronic and necessitate keratoplasty for preservation of vision. Numerous worms migrate to or through the eye, causing inflammation, tissue destruction, and blindness. Echinococcus and Cysticercus can initiate destructive cystic lesions in the eye.
■ PULMONARY SYMPTOMS Patients with P. falciparum malaria initially may seek treatment for fever and cough. Early in the course of treatment for severe malaria, noncardiogenic pulmonary edema or acute respiratory distress syndrome may develop that necessitates mechanical ventilation with positive end-expiratory pressure.42,43 E. histolytica can cause sympathetic pleural effusions, pulmonary or pleural involvement by direct extension or rupture of an amebic liver abscess, or direct hematogenous seeding of the lungs, leading to considerable additional morbidity and mortality among patients with underlying amebic infection.44 Pneumocystis pneumonia, caused by Pneumocystis jiroveci (formerly Pneumocystis carinii), is one of the most common respiratory opportunistic infections in patients with human immunodeficiency virus (HIV) infection in the United States and Europe; however, it is responsible for less than 10% of pulmonary opportunistic infections in Africa and the developing world. The reason for this discrepancy is unclear. Many patients with AIDS in these countries die with CD4+ cell counts that are higher than those associated with P. jiroveci pneumonia in the United States.45
Löffler’s syndrome, characterized by persistent and nonproductive cough, substernal chest pain, wheezing, rales, pulmonary infiltrates on the chest radiograph, and marked eosinophilia,46 often is seen when larvae from the roundworm Ascaris lumbricoides, the hookworms N. americanus and A. duodenale, and the threadworm S. stercoralis transit the lungs as part of their developmental cycles. Ascaris larvae penetrate the small intestinal wall to gain entry into the small venules of the gastrointestinal tract and then migrate to the lungs. Strongyloides and the hookworm filariform larvae penetrate through the skin of the feet, entering small cutaneous venules before migration to the lungs. The pulmonary infiltrates and symptoms are transient, resolving within 2 weeks. Diagnosis depends on discovering larvae in sputum or gastric aspirates. Negative stool examinations initially are nondiagnostic, because eggs do not appear in the stool for at least 1 month after initial infection. The patient’s immune response to the microfilariae of W. bancrofti and B. malayi is the cause of tropical eosinophilic pneumonia.47 Affected persons present with malaise, weight loss, new-onset nocturnal wheezing and asthma, shortness of breath, and chest discomfort. Chest radiographs may show nodular or interstitial infiltrates, consolidations, or cavitation. Microfilariae can be seen in lung biopsy material. Untreated infection may result in obstructive or restrictive lung disease. Patients have marked eosinophilia and elevations of serum IgE.48 Paragonimus westermani and echinococcal species are trophic for the lungs in their human hosts. P. westermani eggs are shed in stool, hatch in fresh water, and, as miracidia, infect a snail intermediary. After further development, cercariae are released from the snail, penetrating and encysting in freshwater crab or crayfish. After consumption by the human host, metacercariae from the crab or crayfish excyst within the duodenum, penetrating the duodenal wall into the abdominal cavity. The larvae migrate from the peritoneal cavity through the diaphragm into the pleural cavity, finally migrating to the lungs, where they cause hemorrhage, necrosis, and a granulomatous response. Early in the process, patients may have infiltrates and eosinophilia; later disease is marked by bronchiectasis, chronic bronchitis, fever, hemoptysis, and cachexia. Pulmonary nodules and cysts may cavitate.49 Many of these patients may have a positive result on purified protein derivative (PPD) testing, and their symptoms and chest radiographic findings may mimic those in tuberculosis. Sputum often is blood-streaked and flecked with dark brown particles containing diagnostic ova. Radiography, stool examination, and immune testing of sputum and blood can help make the diagnosis.50 Praziquantel is the therapeutic agent of choice. E. granulosus causes pulmonary hydatid cyst disease that remains asymptomatic until a cyst grows large enough to cause a mass effect, becomes superinfected, or leaks cyst material, which is highly immunogenic, causing a severe anaphylactoid reaction. Pulmonary hydatid cysts also can be associated with cough, expectoration of sandlike material, chest pain, and hemoptysis.51 Primary hydatid disease in the liver can metastasize to the lungs or brain. A thoracic CT scan shows a unilocular lung cyst; on a plain radiograph, a ruptured cyst is said to resemble a water lily—a pathognomonic finding. Cysts can be treated with careful surgical excision and pharmacotherapy. Early schistosomal disease, or Katayama fever, can manifest with fever, cough, eosinophilia, and diffuse pulmonary nodules as the schistosomula pass through the lungs. In long-standing disease, ova shed from worm pairs can lodge in the vasculature of the lungs, causing pseudotubercles, granulomatous lung disease, pulmonary hypertension, and cor pulmonale. In
■ CARDIOVASCULAR SYMPTOMS Chagas’ Disease Principles of Disease. T. cruzi infection causes acute and chronic myocarditis. T. cruzi is endemic in South and Central America and causes Chagas’ disease. The vector is the reduviid or “kissing bug” that inhabits the walls and roofs of thatched dwellings built adjacent to forest. Previously a disease of rural populations, urban transmigration has expanded the epidemiologic scope of Chagas’ disease. The disease is not seen commonly in travelers. The reduviid’s bite is no longer the only source of T. cruzi infection; transfusion with blood containing live trypanosomes from infected hosts is a growing source of infection.55 Oral transmission also has recently been noted as well.56 The reduviid bites the patient, often around the eye, and excretes feces containing the trypomastigote of T. cruzi. The trypanosome enters the inflamed bite wound or other mucosal or conjunctival surfaces, causing a local swelling, called a chagoma. Romaña’s sign (painless unilateral periorbital edema) is pathognomonic but rarely seen. The trypomastigote migrates to trophic tissues, including smooth muscle, cardiac muscle, and autonomic ganglia in the heart, esophagus, and colon, causing local inflammation and tissue destruction. Clinical Features. Acute infection is heralded by fever, facial and dependent extremity edema, hepatosplenomegaly, lymphadenopathy, malaise, lymphocytosis on peripheral blood smear, and elevated liver transaminases. At this stage, fatal left ventricular dysfunction and dysrhythmias are uncommon. Early illness lasts 1 to 2 months and resolves spontaneously, resulting in a latency known as the indeterminate phase, which can persist throughout the patient’s lifetime. In approximately 25% of the cases, the infection progresses to chronic Chagas’ disease, principally with cardiopathy and gastrointestinal disease. Amastigotes invade cardiac muscle and the cardiac conduction system. Chronic inflammation, mononuclear cell infiltration, and fibrosis are other findings. Additional features may include atrial bradydysrhythmias, right and left bundle branch blocks, complete heart block, and ventricular dysrhythmias including ventricular fibrillation. With development of right and left ventricular dysfunction with dilated cardiomyopathy, cardiac muscle is replaced by fibrosis and scarring. Mural thrombi are common; thromboembolic disease manifesting as pulmonary embolism, stroke, or peripheral arterial embolism can be the first indication of long-standing asymptomatic infection. Congestive heart failure is rapidly progressive and fatal within months unless aggressively treated with pharmacologic intervention and transplantation.57 Diagnostic Strategies. Acute Chagas’ disease can be diagnosed by the presence of motile trypomastigotes in anticoagulated blood specimens. The organism also can be cultured in special liquid media. Chronic Chagas’ disease can be diagnosed using several serologic tests, including complement fixation, enzyme-
1765
linked immunosorbent assay (ELISA), and indirect immunofluorescence testing. The assays are nonspecific, cross-reacting with malaria, syphilis, leishmaniasis, and some collagen vascular diseases. Polymerase chain reaction technology is improving and soon will provide the “gold standard” modality for diagnosis.58 Management. Nifurtimox and benznidazole are used for treating T. cruzi infection. Cure rates rarely exceed 50%. The duration of treatment with nifurtimox is prolonged, and the drug has many serious side effects. Its production has been discontinued; however, it is the only antitrypanosomal medication available in the United States today (it can be obtained from the CDC by calling 404-639-2888). Benznidazole has fewer side effects. It is now recommended for indeterminate-phase treatment. Late complications of chronic diseases are modulated by autoimmune activity and do not respond to antiparasitic pharmacotherapy. Use of nifurtimox and benznidazole has been associated with lymphoma in an animal model.59,60 Chronic Chagas’ disease of the heart, esophagus, or colon is treated symptomatically. Automated implantable cardioverterdefibrillators have been demonstrated to decrease the incidence of sudden death in this patient population.61 Patients receiving immunosuppressive therapy to prevent rejection after cardiac transplantation have shown recurrent disease in the transplanted myocardium.
Other Causes Aberrant migration of Ascaris to the myocardium is well described, causing myocarditis and pericardial effusions. E. histolytica abscesses of the liver also may cause pericardial effusions if they erode through the diaphragm.
■ GASTROINTESTINAL SYMPTOMS Diarrhea Diarrhea is one of the most common symptoms for which travelers seek medical attention. Gorbach62 wrote, “Travel expands the mind and loosens the bowels.” Diarrhea also is the leading cause of death in children younger than 5 years of age in developing countries and a major source of morbidity for older children and adults (Fig. 131-8). Most diarrheal disease is viral or bacterial; however, some clinically significant diarrheal disease is caused by parasites. Cryptosporidium parvum and Cyclospora cayetanensis are foodborne and water-borne coccidians that cause watery diarrhea.
Figure 131-8. Fecal-oral transmission of diarrheal agents occurring in a developing country.
Chapter 131 / Parasitic Infections
patients with long-standing, latent, and asymptomatic S. stercoralis infections who are started on corticosteroids or immunosuppressive therapy, the helminth disseminates widely. Fatal, massive pulmonary infections with radiographic whiteouts and unsupportable respiratory failure have been reported in patients who have received organ transplants; this clinical disaster occurs more commonly in patients who originally came from developing countries and were never evaluated for Strongyloides infection before transplantation.52,53 Strongyloides infection can be misdiagnosed as bronchospasm and asthma, prompting the clinician to prescribe steroids, which precipitate dissemination.54
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1766
Both are particularly significant causes of morbidity in malnourished children and patients with AIDS. In these populations in the developing world, the prevalence may approach 50%.63,64 Cryptosporidial oocysts can be seen in stool when an acid fast–based stain is used. ELISA and immunofluorescent assays of stool also are available. Paromomycin decreases diarrheal frequency in patients with AIDS, who often have prolonged illness. Treatment is symptomatic for immunocompetent hosts.65 Cyclospora oocysts can be detected in stool samples using a Ziehl-Neelsen stain. Trimethoprimsulfamethoxazole treats the infection.66 E. histolytica causes an invasive or inflammatory diarrhea. Patients complain of fever, tenesmus, abdominal pain, and watery stool containing blood and mucus. Untreated disease can progress to widespread colitis and perforation of the bowel wall with peritonitis and death.67 Stool examination reveals mobile trophozoites containing ingested RBCs. Cysts noted on stool studies do not necessarily reflect active infection. Immune assays of stool can now differentiate between E. histolytica and nonpathogenic ameba species. Serologic tests may be useful in an infected patient from a nonendemic region but take a month for results to turn positive. Metronidazole is the drug of choice for treatment of amebiasis. Balantidium coli is the other protozoan that can cause invasive diarrhea. It has trophism for the terminal ileum, sometimes resulting in a clinical picture suggestive of appendicitis. Tetracycline and metronidazole are active against B. coli. Giardia lamblia can cause persistent diarrhea, abdominal bloating, cramps, flatulence, and weight loss. The organism is ingested and reproduces exponentially in the small bowel. In severe infection, the entire jejunum becomes covered with organisms, and the patient has malabsorption with steatorrhea. The organisms are rarely seen in fresh stool preparations because they quickly break down and become indiscernible. Accordingly, an antigen test often is used to confirm the diagnosis. Giardia has many animal reservoirs, including the beaver. Campers who drink unfiltered, pure mountain spring water in the United States commonly contract Giardia infection. Metronidazole and tinidazole treat the disease. S. stercoralis, Capillaria philippinensis, T. trichiura, and Schistosoma all have been associated with diarrhea. Hyperinfection or dissemination of Strongyloides can cause persistent diarrhea, weight loss, and abdominal pain. Trichuris causes diarrhea when the parasite load in the intestine is high. Schistosomiasis can cause a chronic granulomatous colitis, which may resemble inflammatory bowel disease, or an acute, bloody, febrile colitis associated with Katayama fever in the immunologically naive patient.
In chronic schistosomiasis, worm pairs in patients’ mesenteric and portal venous systems lay eggs that become ensnared in the liver, causing intense local inflammation and scarring and the classic “pipestem” cirrhosis with periportal fibrosis. Clinical manifestations in these patients include portal hypertension, ascites, and esophageal varices (Figs. 131-9 and 13110). Upper gastrointestinal bleeding is not as common as in patients with alcoholic cirrhosis; however, a high number of patients are infected with schistosomiasis in endemic regions, so variceal bleeding is an important cause of gastrointestinal hemorrhage in these populations.68,69
Figure 131-9. Pipestem cirrhosis with extensive ascites in a patient with
Figure 131-10. Extensive ascites in a child, which may be from
chronic schistosomiasis.
Abdominal Pain In an extensive review of a number of cases of appendicitis, parasitic infection was found to be the cause in 3%. Pathologic examination revealed enterobiasis, amebiasis, ascariasis, trichuriasis, and taeniasis.70 A. lumbricoides can cause significant persistent or recurrent abdominal pain in adults and partial intestinal obstruction in children with significant worm loads. Antihelminthics and conservative, supportive therapy usually eliminate the problem, thereby avoiding surgical intervention.71,72 The diagnosis of ascariasis is made by identifying eggs in the stool. Patients with large worm loads may excrete adult worms, especially after therapy is started. Severe intestinal amebiasis can be complicated by colonic perforation and peritonitis. Angiostrongylus costaricensis, a nematode known as the rat lung worm, is common in Central America. Infected children may appear clinically to have Meckel’s diverticulum or acute appendicitis. Manifestations of the infection include nausea, vomiting, fever, abdominal pain localized to the right lower quadrant, and a tender mass. Surgical exploration may uncover abscesses, obstruction, or intestinal infarction.73 Anisakiasis is characterized by severe abdominal pain after ingestion of raw fish (sushi and sashimi primarily). Anisakis
schistosomiasis or kala-azar (leishmaniasis).
Pruritus Ani Enterobius vermicularis, or pinworm, causes pruritus ani, a syndrome of intense perianal itch occurring primarily in children. Autoinfection is common because children (and adults) scratch the pruritic anal area and then bite their nails or put their fingers in their mouth. The worm has a worldwide distribution. Diagnosis is clinical and is confirmed by finding the small adult worms wiggling about on the anal verge. Eggs are rarely seen in the stool but can be visualized using the tape test: Transparent tape touched to the perianal region collects eggs, which can be seen with light microscopy. Albendazole or mebendazole is the drug of choice.
■ PARASITIC CO-INFECTIONS IN PATIENTS WITH HUMAN IMMUNODEFICIENCY VIRUS INFECTION AND AIDS Perspective HIV infection and AIDS are prevalent in developing countries. Heterosexual and perinatal transmission are common; young children and young adults of both sexes are primarily
1767
infected. Patients presenting to the ED may be co-infected with HIV or any other infectious agent, including all of the parasites discussed in this chapter. HIV co-infection may worsen the symptoms and outcome, alter the presentation, increase the virulence, or assist the infective process. AIDS causes abnormalities in almost every aspect of a host’s immune response to infection; cell-mediated immunity (which is important in combating parasitic infection) is most affected.79 The diagnosis and response to therapy of many parasitic infections are monitored serologically. HIV infection interferes with this response, rendering many of these tests unreliable. Therapies that are extremely effective in the normal host may be ineffective in a patient with HIV infection. Pharmacologic agents may have to be given for long periods or for the patient’s entire life.
Specific Parasitic Infections Malaria is not an opportunistic infection in patients with AIDS; however, many patients, especially children, with recurrent malaria and anemia from hemolysis have required transfusions from blood supplies not screened for HIV and have become infected.80 Treating febrile patients for malaria in areas in which it is endemic is a common practice. Patients with AIDS have more severe allergic reactions to drugs, especially sulfonamides, that are antimalarials. In patients with AIDS, fever alone is not predictive of malaria; diagnosis should precede therapy. Patients with HIV infection are at greater risk for severe clinical manifestations of babesiosis.81 Visceral leishmaniasis is more commonly disseminated and fatal in patients with AIDS. Latent leishmanial infections may reactivate, and a prolonged febrile illness in an HIV-positive patient with a lifetime history of travel in leishmaniasisendemic areas of the world should prompt consideration of this co-infection.82 Cutaneous infection also may become diseminated in these patients. Several clinical trials are currently examining the role of chemoprophylaxis for leishmaniasis in HIV-positive persons who live in endemic regions of the world. Chagas’ disease in the indeterminate phase can be reactivated in patients infected with HIV. These patients frequently have CNS involvement with meningoencephalitis and severe myocarditis.83 Single-drug therapy may be insufficient, because benznidazole penetration into the CSF is minimal. T. gondii infection is well recognized throughout the world as a common opportunistic infection of patients with AIDS, with a particular trophism for the CNS. The coccidial organisms Isospora belli, C. parvum, and C. cayetanensis all have been associated with prolonged diarrhea in patients with AIDS. These organisms cause infections that are difficult to treat and are almost impossible to eradicate in these patients. The diarrhea is extremely debilitating and can be as profuse as that seen in cholera. E. histolytica has a high prevalence among homosexual men who practice unprotected anal intercourse; however, invasive amebiasis is not an opportunistic infection associated with HIV infection. Schistosomiasis enhances the pathogenesis of HIV infection and is more difficult to treat and eradicate in patients who are HIV-positive.84 S. stercoralis infection is more likely to manifest as hyperinfection and disseminated disease in patients who are HIV-positive.85 In patients who are at risk for HIV infection and parasitic illness, it is essential to consider co-infection in the differential diagnosis.
Chapter 131 / Parasitic Infections
marina, a nematode that burrows into the intestine, is the pathogen.74 The liver fluke, F. hepatica, causes a syndrome that mimics viral hepatitis: right upper quadrant pain, fever, nausea and vomiting, jaundice, a tender enlarged liver, and elevated transaminases. Patients also have eosinophilia and urticaria. Imaging studies, including CT, show the tracts of burrowing flukes. Serologic testing establishes the diagnosis; the patient’s stool may not contain eggs for several months after ingestion.75 The eggs of schistosomes become trapped in the portal venules, where they trigger an inflammatory response, leading to granulomatous liver disease, fibrosis, and cirrhosis. Hepatic granulomas also are seen in disseminated strongyloidiasis and aberrant biliary ascariasis. E. histolytica can cause hepatic abscesses. Affected patients typically do not have amebic dysentery and do not shed Entamoeba in their stool, but results of serologic studies almost always are positive. Patients have fever, weight loss, anorexia, and right-sided abdominal pain, but no jaundice. Treatment is with metronidazole or tinidazole and a luminal amebicide, such as iodoquinol.76 E. granulosus produces hydatid cysts of the liver that, on CT, contain septations and so-called daughter cysts. Pharmacotherapy with albendazole and careful excision remain the treatments of choice. Leaking cyst material can initiate a severe anaphylactoid reaction in the host. Jaundice may result from hemolysis secondary to direct infection of RBCs with Plasmodium or Babesia or from biliary obstruction with pigmented stones. Ascaris can cause biliary colic, pyogenic cholangitis, pancreatitis, or liver abscess. Dead worms can be the nidus for gallstone formation. Biliary imaging and endoscopic retrograde cholangiopancreatography will show worms in the biliary tree. Mechanical removal by endoscopy combined with anthelminthic therapy is curative.77 Clonorchis sinensis and F. hepatica are trophic for the biliary tree. These worms can be present without producing symptoms for years before eventually precipitating cholecystitis, cholangitis, or cholangiocarcinoma.78
1768
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
KEY CONCEPTS ■
Parasitic diseases may manifest with almost any symptom or constellation of symptoms. Accordingly, a travel history should be obtained in all patients with clinically significant signs and symptoms of unclear etiology. The combination of presenting symptoms and signs and a history of recent travel to specific geographic regions can lead to early diagnosis of most parasitic infections. ■ Parasitic co-infections are particularly common in patients with HIV infection and AIDS. ■ Acute malaria should be suspected in patients with irregular high fevers associated with headache,
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
abdominal pain, or respiratory symptoms. Patients who are clinically ill or who are suspected of having falciparum malaria should be hospitalized for evaluation and treatment. ■ Cysticercosis should be considered in the differential diagnosis for new-onset seizures. ■ Giardiasis should be suspected in patients with diarrhea who recently have been camping or drinking unfiltered mountain spring water.
Chapter 132
Tick-Borne Illnesses
Edward B. Bolgiano and Joseph Sexton
■ PERSPECTIVE
Physiology of Tick Feeding
Ticks are hematophagous parasites of humans and animals, distributed worldwide. They transmit rickettsial, bacterial, spirochetal, viral, and protozoal diseases and cause disease by means of their own toxins (Table 132-1). As vectors of human disease, ticks rank second in importance only to mosquitoes. People who travel during the summer months may return from endemic areas with tick-borne disease. In addition, reports of infection acquired within urban areas emphasize the need to consider tick-borne illness even in the absence of a history of travel to high-risk areas.1 Tularemia (category A) and Q fever (category B) are now considered by the Centers for Disease Control and Prevention (CDC) to be significant threats during biologic warfare. For this reason, research involving ticks and their diseases has become increasingly important. Reports on ticks, their feeding habits, and their possible relation to disease can be found from early history.2 Pliny (ce 77), in Historia naturalis, referred to “an animal living on blood with its head always fixed and swelling, being one of the animals which has no exit [anus] for its food, it bursts with over-repletion and dies from actual nourishment.” Tick-borne illness was first recognized on the North American continent by Native Americans. According to legend, Shoshone men avoided the “evil spirits” that caused illness by sending only women into certain areas of the Rocky Mountain region known to be especially hazardous. The etiologic association of the tick vector with Rocky Mountain spotted fever (RMSF) was noted by missionaries and by early settlers, who named the affliction “tick fever.” Physicians in Idaho and Montana recorded the classic clinical descriptions of the disease in 1899.
An understanding of the physiology of feeding in arthropods is more essential than species identification in assessing the risks of transmission of diseases. Blood-sucking arthropods are divided into two groups according to their method of acquiring blood. The solenophagic feeders insert their mouthparts directly into capillaries to obtain blood. Telmophagic feeders insert their mouthparts indiscriminately, lyse tissue along with capillaries, and feed on the resultant pool of blood, extracellular fluid, and tissue. Ticks and deer flies, for example, are telmophagic feeders, whereas mosquitoes are mostly solenophagic. Argasid ticks (soft-bodied), are short, rapid feeders with preformed distensible endocuticles. They therefore need to feed for only minutes to hours to acquire a full meal. As a result, they tend to be found in nests and burrows where their hosts visit frequently. The genus Ornithodoros is the vector for relapsing fever. Ixodid ticks (hard-bodied) include the genera Ixodes, Dermacentor, Amblyomma, and Rhipicephalus, which are those responsible for the remainder of human tick-borne diseases discussed in this chapter. These ticks need to form a new exocuticle (phase I of feeding) and thus feed slowly during the first 12 to 24 hours. Once fully formed, the new endocuticle allows rapid feeding (phase II) and significant engorgement. In the capitulum of ticks, the sucking structure, consisting of the chelicerae, is surrounded by a sheath from which it protrudes during feeding. Sense organs on the capitulum, or podomeres, help locate a host by means of chemoreceptors. Hair, or setae, on the legs act as tactile and temperature receptors. A special sensory structure, Haller’s organ, is located on the first set of legs and is a humidity and olfactory receptor.3 When a suitable location is found, adjacent cheliceral digits incise the skin, and the chelicerae and barbed hypostome are inserted. Two mechanisms prevent the tick from being removed from the skin: the barbed hypostome and a cementlike salivary secretion from the base of the hypostome, composed of lipoproteins and glycoproteins. This allows ixodid ticks to attach for as long as 2 weeks. Because argasids are much faster feeders, they secrete no cement substance. During a bite, trauma and salivary gland products can cause local inflammation, hyperemia, edema, hemorrhage, and skin thickening. The saliva injected during feeding contains many different substances. Both hard and soft ticks produce a histolytic secretion that liquifies tissue, which is then sucked into the gut. Eventually, the secretion breaks down the walls of the
■ PRINCIPLES OF DISEASE Identification of Ticks Ticks are arthropods but not insects. They have eight legs instead of six and generally two fusing body parts—a capitulum (head) and an opisthosoma (abdomen)—instead of three. Identification of an arthropod as a tick and subsequent categorization into family and some genera are not difficult (Figs. 132-1 and 132-2). Speciation requires a trained acarologist. However, tick identification has limited importance in clinical decision-making. Color, which varies seasonally, and size, which varies by amount of blood ingested at the time of presentation, are unreliable criteria for identification purposes.
1769
1770
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
Table 132-1 Tick-Borne Illnesses TYPE
DISEASE
PATHOGEN
ARTHROPOD VECTOR
GEOGRAPHIC DISTRIBUTION
Bacterial (including spirochetal)
Lyme disease
Borrelia burgdorferi
Ixodes scapularis I. pacificus I. ricinus
Tularemia
Francisella tularensis
Rocky Mountain spotted fever
Rickettsia rickettsii
Q fever Human monocytic ehrlichiosis Human granulocytic anaplasmosis
Coxiella burnetii Ehrlichia chaffeensis
I. scapularis Amblyomma americanum Dermacentor variabilis D. andersoni D. variabilis Rhipicephalus sanguineus D. andersoni A. americanum
Northeastern U.S. Upper midwestern U.S. Pacific coast Europe Southwest central U.S.
Parasitic (protozoal) Viral*
Babesiosis Colorado tick fever
Babesia microti Orbivirus
Miscellaneous
Tick paralysis
Ixobotoxin
Rickettsial
Anaplasma phagocytophilum
I. scapularis I. pacificus I. scapularis D. andersoni D. andersoni D. variabilis A. americanum I. scapularis I. pacificus I. holocyclus
Predominantly southeastern U.S. Arizona Worldwide South central and southeastern U.S. New England and north central U.S. Northern California Coastal New England Mountain areas of western U.S. and Canada Worldwide
*Many other viruses are transmitted to humans by ticks. In the United States, only Colorado tick fever occurs with any significant frequency.
dermal blood vessels and the released blood is ingested. To prevent hemostasis, the saliva contains a thrombokinase inhibitor, apyrase, which prevents platelet aggregation by depleting adenosine diphosphate, prostaglandin E2, and prostacyclin (prostaglandin I2) to prevent vasoconstriction, and cytolysins. Ixodes scapularis also secretes a carboxypeptidase that destroys other inflammatory mediators such as anaphylatoxins and bradykinin, as well as anti–complement C3 factor. These other mediators normally would cause further inflammation, which would enhance hemostasis. All infectious agents, as well as excretory liquids from some argasids, are transmitted through this saliva. Transmission of a disease from Ixodes ticks is unlikely if the tick is not yet engorged with blood at the time of removal. Likewise, a tick removed within a few hours after attachment is unlikely to transmit disease. The neurotoxins responsible for tick paralysis also are found in tick saliva. The local physiologic changes associated with tick feeding produce the characteristic 1- to 4-mm erythematous mark typically seen on the skin after a tick bite. This is a common finding from most blood-sucking arthropods. The mark should not be confused with certain rashes associated with disease progression—for example, erythema migrans. Informing patients of this difference may be reassuring.
■ LYME DISEASE Perspective Lyme disease, the most common vector-borne disease in the United States, is a tick-borne illness caused by the spirochete Borrelia burgdorferi. The story of Lyme disease begins in 1975, when health officials at Connecticut’s State Department of Health and physicians at Yale University were alerted by two skeptical mothers to an unusually large number of cases of apparent juvenile rheumatoid arthritis (JRA) occurring in their
small coastal community of Old Lyme, Connecticut. Investigation led to the description of a “new” entity called Lyme arthritis.4 Lyme disease occurs worldwide and has been reported on every continent except Antarctica.5 It now accounts for more than 95% of all reported cases of vector-borne illness in the United States.6 The actual overall incidence of Lyme disease is unknown, because many cases go unreported. Lyme disease occurs in people of all ages but is more common in children younger than 15 years and in adults 30 to 60 years of age.7 Persons at greatest risk live or vacation in endemic areas. In the United States, three distinct endemic foci are recognized: the northeastern coastal, mid-Atlantic, and north central states. During 2000, a total of 17,730 cases of Lyme disease were reported from 44 states and the District of Columbia. Twelve states—Connecticut, Rhode Island, New Jersey, New York, Delaware, Pennsylvania, Massachusetts, Maryland, Wisconsin, Minnesota, New Hampshire, and Vermont—accounted for 95% of cases reported in the nation (Fig. 132-3).7 The principal tick vectors are I. scapularis in the Northeast and Midwest and Ixodes pacificus in the West. Nymphal Ixodes ticks satisfy all known epidemiologic requirements for the zoonosis as it exists in nature. There is no compelling evidence for alternate arthropod vectors of infection. The I. scapularis population density depends on that of its preferred hosts: the white-footed field mouse, Peromyscus leucopus, for the larval and nymphal forms and the white-tailed deer, Odocoileus virginianus, for the adult form. The whitefooted mouse readily becomes infected after being bitten by infected ticks and remains highly infectious for periods of time that approach its life span in nature, thereby providing an important reservoir for B. burgdorferi.8 Adult I. scapularis ticks feed primarily on deer, which are key hosts in the tick life cycle and in whose fur the adult tick may survive the winter. The relatively recent repopulation of several areas in the
1771
Chapter 132 / Tick-Borne Illnesses
Figure 132-1. Scanning electron micrographs of two tick
species. A, Dorsal view of adult female Dermacentor variabilis. B, Dorsal view of adult female Ixodes scapularis. C, Dorsal close-up view of D. variabilis head. D, Dorsal close-up view of I. scapularis head. (Courtesy of Dr. J.E. Keirans, Georgia Southern University, Statesboro, Georgia.)
A
B
C
D
KEY TO IDENTIFICATION OF IXODIDAE AND ARGASIDAE TICKS No scutum, head ventral
Scutum present, full head visible from above
Argasidae (soft ticks) (Ornithodoros, etc.)
Ixodidae (hard ticks)
Long, oval bodies
Long mouth parts
Short mouth parts
Ixodes
Dermacentor Rhipicephalus
Short, stout body
Amblyomma
Figure 132-2. Identification scheme for genera within Ixodidae and Argasidae, the two primary disease-transmitting families of ticks.
United States by white-tailed deer preceded the recent emergence of Lyme disease in those regions.9 Although all stages of the tick may feed on humans, the nymph is primarily responsible for the transmission of Lyme disease. It is not surprising that more than two thirds of patients with Lyme disease do not recall a tick bite, in view of the small size (1 to 2 mm) of nymphs (Fig. 132-4). The nymph feeds in the spring and summer, which correlates with a peak incidence of early Lyme disease occurring between May and August. In addition, recreational and occupational exposure is greatest during this time. Later manifestations of Lyme disease may appear throughout the year.
Principles of Disease The spirochete B. burgdorferi persists and multiplies in the midgut of its tick vector, I. scapularis. Transmission of the spirochete to humans occurs during feeding, generally 2 days after attachment.10 The mechanism of transmission probably is inoculation with infectious saliva or, alternatively, with tick gut fluids periodically regurgitated during the feeding process.11 After an incubation period that lasts several days to several weeks, spirochetemia develops, and Borrelia organisms may
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1772
Figure 132-3. Number of newly
reported Lyme disease cases, by county in the United States in 2005. N = 23,174; the county was unknown for 131 other cases. One dot was placed randomly within the county of patient residence for each reported case. (From Lyme disease—United States, 2003–2005. MMWR Morb Mortal Wky Rep 56:573, 2007.)
disease. The variable severity of Lyme disease may in part result from genetic variations in the human immune system. Patients with chronic Lyme arthritis have an increased frequency of human leukocyte antigen (HLA) specificity—in particular, for HLA-DR4 and, less often, HLA-DR2. = 1 mm
A
Clinical Features
A. americanum
I. scapularis
1 mm
Unengorged Engorged
B
D. variabilis
I. scapularis nymph
Figure 132-4. A (Left to right), Larva, nymph, adult male, adult female, and
engorged adult female Ixodes ticks and adult male and female Dermacentor ticks. Actual size. B, Adult female Amblyomma americanum (Lone Star tick), adult female and nymphal Ixodes scapularis (deer tick), and adult female Dermacentor variabilis (dog tick). (From Hayes EB, Piesman J: How can we prevent Lyme disease? N Engl J Med 348:2424, 2003.)
migrate outward in blood or lymph to virtually any site in the body. The spirochete appears to be tropic for synovial tissue, skin, and cells of the nervous system, but the mechanism of this tropism is not yet understood. Infection by the spirochete itself accounts for early clinical manifestations. It remains unclear whether late disease manifestations require the continued presence of viable spirochetes or whether an ongoing host immune response to initial infection is sufficient to cause some late disease manifestations. Although the exact roles of infecting spirochetes, spirochetal antigens, and host immune responses are unknown, it is likely that persistent live spirochetes are responsible for most later manifestations of the
Lyme disease, a multisystem disorder, can be classified into three stages: early localized, early disseminated, and late disease. Virtually any clinical feature may occur alone or recur at intervals, and some patients who had no early symptoms may have late symptoms. The disorder usually begins with a rash and associated constitutional signs and symptoms, suggesting a “viral syndrome” (early Lyme disease). Neurologic, joint, or cardiac manifestations may emerge weeks to months later (early disseminated Lyme disease), and chronic arthritic and neurologic abnormalities may appear weeks to years later (late Lyme disease). The time course for the clinical features of untreated Lyme disease is illustrated in Figure 132-5.
Early Lyme Disease Ticks may attach to human hosts at the initial point of contact (generally around ankle level) or may move about until they encounter an obstruction. The groin, popliteal fossae, gluteal folds, axillary folds, and ear lobes are common sites of attachment. After transmission of B. burgdorferi through a tick bite, the initial site of infection is the skin at the site of the bite. After an incubation period of approximately 1 week (range, 1 to 36 days), the spirochetes cause a gradually spreading localized infection in skin and a resultant skin lesion, erythema migrans. Erythema migrans is the most characteristic clinical manifestation of Lyme disease and is recognized in 90% or more of patients. Erythema migrans may go unnoticed if the entire skin surface is not examined.12 The characteristic rash begins at the site of the tick bite with an erythematous papule or macule. The lesion expands gradually (1 to 2 cm/day, a rate of expansion slower than that of cellulitis). The patch of erythema may be confluent or may have bands of normal-appearing skin. Central clearing may occur but is not an invariable feature. The lesion borders usually are flat but may be raised. The lesions generally are sharply demarcated and blanch with
1773 Clinical features
Figure 132-5. Natural history of serologic response,
with clinical features, in untreated Lyme disease. IgG, immunoglobulin G; IgM, immunoglobulin M. (From Rahn DW: Natural history of Lyme disease. In Rahn DW, Evans J [eds]: Lyme Disease. Philadelphia, American College of Physicians, 1998, pp 35–48.)
• Neurologic Cranial neuropathy Meningitis Radiculoneuropathy • Joint Acute inflammatory large joint arthritis • Carditis
Late Lyme disease • Neurologic Peripheral neuropathy Encephalopathy • Chronic arthritis
Relative frequency (%)
• Erythema migrans Localized erythema migrans Flulike illness Multiple erythema migrans
Early disseminated Lyme disease
Years
Serology
Months
lgG
lgM
Months
pressure. Most lesions are oval or round, but triangular and elongated patches may occur. In patients presenting 1 to 7 days after the appearance of lesions, the average lesion size is approximately 8 by 10 cm (range, 2 by 3 cm to 25 by 25 cm). In some cases, the center of some early lesions becomes red and indurated or vesicular and necrotic. The lesion is warm to the touch and may be described by the patient as nontender to minimally tender (Fig. 132-6). Hematogenous spread of viable spirochetes (not additional tick bites) may result in one or more secondary lesions. These secondary lesions are smaller, migrate less, and typically spare the palms and soles. In all, 10 to 15% of patients have more than 20 such lesions; on rare occasions they may number more than 100. Blistering and mucosal involvement do not occur. The primary and secondary skin lesions generally fade after approximately 28 days (range, 1 week to 14 months) without treatment and within several days of antibiotic therapy. Recurrent lesions may develop in patients who do not receive antibiotic therapy, but apparently not in those who receive appropriate antibiotics. Constitutional signs and symptoms commonly appear in early Lyme disease. Malaise, fatigue, and lethargy are most common (seen in approximately 80% of patients) (Table 132-2) and may be severe. Fever typically is low grade and intermittent. Lymphadenopathy usually is regional in the distribution of erythema migrans or may be generalized; splenomegaly may occur. Musculoskeletal complaints, such as arthralgias and myalgias, are common, and the discomfort typically is short-lived and migratory, sometimes lasting only
Years
hours in one location. Frank arthritis may occur at this stage but is rare. Clinical manifestations of meningeal irritation are frequently seen. Headache, the most common symptom, usually is intermittent and localized. Nausea, vomiting, and photophobia occasionally accompany the headache. Kernig’s and Brudzinski’s signs typically are absent, and neck stiffness usually is noted only on extreme forward flexion. At this stage, the neurologic examination and cerebrospinal fluid (CSF) assessment (usually) both yield normal findings. Signs and symptoms of hepatitis, including anorexia, abdominal pain, right upper quadrant tenderness, nausea, and vomiting, may be present. Mild pharyngitis also may be present, but other upper respiratory symptoms such as rhinorrhea do not occur. Although the systemic symptoms of early Lyme disease often are described as “flulike,” this term can be misleading because clinically significant cough usually does not occur. Conjunctivitis develops in approximately 10% of patients. The incidence of Lyme disease without erythema migrans appears to be approximately 10%.12 Because of the variety of nonspecific signs and symptoms at this stage, in the absence of the characteristic rash or history of tick bite, early Lyme disease may be easily confused with a viral or collagen-vascular disease. The intermittent and rapidly changing nature of the early signs and symptoms of Lyme disease may be a helpful distinguishing feature, especially in a patient from an endemic area. In untreated disease, early symptoms usually last for several weeks but may persist for months.
Chapter 132 / Tick-Borne Illnesses
Early Lyme disease
1774
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
Table 132-2 Early Clinical Manifestations of Lyme Disease MANIFESTATION
A
NO. OF PATIENTS (%)
Signs Erythema chronicum migrans* Multiple annular lesions Lymphadenopathy Regional Generalized Pain on neck flexion Malar rash Erythematous throat Conjunctivitis
128 63 52 41 38 35
(41) (20) (17) (13) (12) (11)
Symptoms Malaise, fatigue, lethargy Headache Fever and chills Stiff neck Arthralgias Myalgias Backache Anorexia Sore throat Nausea Dysesthesia Vomiting
251 200 185 151 150 135 81 73 53 53 35 32
(80) (64) (59) (48) (48) (43) (26) (23) (17) (17) (11) (10)
314 (100) 150 (48)
*Required for inclusion in this study. From Steere AC, et al: The early clinical manifestations of Lyme disease. Ann Intern Med 99:76, 1983.
B Figure 132-6. Lyme disease usually begins with a slowly expanding skin
lesion, erythema migrans, which occurs at the site of the tick bite. A, A classic “bull’s-eye” or target lesion with partial central clearing, a bright red outer border, and a “target” center. B, Expanding erythematous lesion, with central clearing, on trunk. (A, Courtesy of Michael O. Murphy, MD. From Malawista SE, Bockenstedt LK: Lyme disease. In Goldman L, et al [eds]: Cecil Medicine: Expert Consult, 23rd ed. Philadelphia, Saunders, 2008, pp 2289–2294; B, courtesy of John Cook, MD. From Goldstein BG, Goldstein AE [eds]: Practical Dermatology, 2nd ed. St. Louis, Mosby, 1997, p 63.)
Acute Disseminated Infection Shortly after disease onset, hematogenous spread can cause a variety of systemic signs and symptoms and result in secondary sites of infection. Organ systems commonly affected are the nervous system, heart, and joints. Less commonly, the eyes, liver, skeletal muscle, subcutaneous tissue, and spleen are infected.
Neurologic Manifestations A relatively symptom-free interval usually occurs between early and disseminated infection; however, neurologic signs and symptoms may be the presenting manifestations of Lyme disease or may overlap with early or late manifestations. Beginning at an average of 4 weeks (range, 0 to 10 weeks) after the onset of erythema migrans, neurologic involvement occurs in approximately 15% of untreated patients. The most common neurologic manifestation of Lyme disease is a fluctuating meningoencephalitis with superimposed symptoms of cranial neuropathy, peripheral neuropathy, or radiculopathy. A triad of meningitis, cranial neuropathies
(usually Bell’s palsy), and radiculopathy has been described, but each entity may occur alone. Headache of variable intensity usually is present; other signs and symptoms of a mild meningoencephalitis may be noted, including lethargy or irritability, sleep disturbances, poor concentration, and memory loss. At this point, the disease often is misdiagnosed as viral meningitis. As in early disease, Kernig’s and Brudzinski’s signs are absent and computed tomography (CT) findings are normal. Unlike in early disease, however, findings on CSF examination often are abnormal, with a lymphocytic pleocytosis and elevated protein level. CSF glucose concentration usually is normal. Intrathecal B. burgdorferi antibody (usually immunoglobulin G [IgG] or IgA) is present in 80 to 90% of patients. CSF polymerase chain reaction (PCR) assay results are positive in less than one half of patients.13 Cranial neuropathies are common, occurring in approximately 50% of patients with Lyme meningitis; usually the seventh nerve is involved. Other cranial nerves are affected less often. Bell’s palsy is bilateral in approximately one third of patients. Its duration usually is from weeks to months, and the condition generally resolves spontaneously without treatment. Peripheral nervous system manifestations also may occur in early disseminated Lyme disease. The spinal root and plexus and the peripheral nerves may be involved in the form of thoracic sensory radiculitis, brachial plexitis, mono neuritis, and motor radiculoneuritis in the extremities. Patients may complain of weakness, pain, or dysesthesia. Examination may reveal loss of reflexes. Involvement of the extremities usually is asymmetrical, but cervical and thoracic dermatomes may be affected. Other rare neurologic abnormalities described in association with Lyme disease include chorea, transverse myelitis, ataxia, and pseudotumor cerebri.
Cerebral vasculitis associated with Lyme disease also has been reported. Cardiac involvement in Lyme disease is uncommon. Estimates of the incidence of carditis in untreated patients who have Lyme disease range from 4 to 10%.14 The average time from initial illness to the development of carditis typically is 3 to 5 weeks (range, 4 days to 7 months). Direct myocardial invasion has been demonstrated with endomyocardial biopsy.15 Electrophysiologic testing has demonstrated widespread involvement of the conduction system.16 The most common cardiac manifestation of Lyme disease is atrioventricular (AV) block, although conduction defects may involve any level of the conducting system. Myopericarditis, tachydysrhythmias, and ventricular impairment occur less often. In a review of 105 reported cases of Lyme carditis, 49% of cases were third-degree AV block, 16% were seconddegree, and 12% were first-degree.16 The degree of AV block seen in a specific patient may fluctuate rapidly.14 A commonly observed feature of AV block in patients with Lyme carditis is its gradual resolution, resembling that occurring after an acute inferior wall myocardial infarction and presumably related to the resolution of inflammation. Assessment of the level of the AV block is important to determine the prognosis of a patient with Lyme carditis. In most cases, block appears to be at or above the level of the AV node; therefore, the prognosis is favorable.17 However, infranodal AV block does occur and may be characterized by slow escape rhythms of wide QRS pattern, asystole, or fluctuating left and right bundle branch block. Other electrocardiographic findings include nonspecific ST- and T-wave abnormalities and intraventricular conduction delay.17 Patients with high-degree AV block usually are sympto matic. Symptoms include light-headedness, palpitations, syncope, chest pain, and dyspnea on exertion. Physical examination may reveal flow murmurs and murmurs of mild mitral regurgitation, pericardial friction rub, or evidence of congestive heart failure. Associated left ventricular dysfunction may be present and has been documented by two-dimensional echocardiography and radionuclide studies; in most reported cases, it has been mild and transient.
Arthritis Although classically considered a sign of late Lyme disease, acute arthritis may begin during the acute disseminated stage. Monarticular or oligoarticular arthritis, primarily affecting large joints, especially the knee, may develop weeks to months after the onset of initial illness. In an early study of the natural history of Lyme arthritis, approximately 50% of untreated patients experienced one episode or multiple intermittent attacks of arthritis. Acute arthritis typically is monoarticular, with involvement of only one knee. The shoulder, elbow, temporomandibular joint, ankle, wrist, hip, and small joints of the hands and feet are involved less commonly. Episodes of arthritis typically are brief (lasting weeks to months) and separated by variable periods of remission. Arthrocentesis generally is nondiagnostic, yielding an inflammatory synovial fluid with a mean white blood cell (WBC) count of approximately 25,000 cells/µL (75% polymorphonuclear leukocytes). Higher WBC counts have been reported, simulating septic arthritis.18 The synovial glucose concentration usually is normal, and protein levels are variable, ranging from 3 to 8 g/dL. Cultures of the fluid rarely identify the causative spirochete.19 Complement level generally is greater than one third that of serum. Synovial biopsy reveals hypertrophy, vascular proliferation, and a mononuclear cell infiltrate. Find-
Ophthalmic Manifestations Ocular involvement also may be seen in early disseminated disease, with manifestations including conjunctivitis, keratitis, choroiditis, retinal detachment, optic neuritis, and blindness.20 These findings also may be seen in late disease.
Late Lyme Disease The chronic phase of Lyme disease is characterized by arthritic and, less commonly, neurologic symptoms. Transition from a pattern of episodic inflammation in early disease to a more indolent persistent inflammation is observed over time. The term chronic (or late) Lyme disease is used to describe continuous inflammation in an organ system for more than 1 year.21 A pattern of exacerbation and remission of arthritis may extend over several years, with a gradual tendency toward less frequent and less severe occurrences. The spontaneous longterm remission rate approximates 10 to 20% annually in untreated patients. However, patients commonly have episodes of periarticular involvement, arthralgias, or fatigue interspersed between attacks of frank arthritis. During the second or third year of illness, attacks of joint swelling sometimes become longer in duration, lasting months rather than weeks. Chronic arthritis eventually develops in approximately 10% of patients.22 Late neurologic complications include a wide variety of abnormalities of the central and peripheral nervous systems, as well as fatigue syndromes. Diagnosis may be difficult because of the large number of other neurologic conditions that Lyme disease may imitate and because late neurologic symptoms may be the first symptoms of the disease. The manifestations of chronic neuroborreliosis usually appear months to years after the onset of infection. The most common late neurologic manifestation of Lyme disease is a chronic encephalopathy that manifests as a mild to moderately severe impairment of memory and learning. Hypersomnolence and mild psychiatric disturbances (depression, irritability, or paranoia) also may develop.13 Peripheral nervous system manifestations often are seen in late disease, with involvement of cranial nerves, spinal roots, spinal plexuses, and peripheral nerves. A predominantly sensory polyradiculoneuropathy that manifests as either radicular pain or distal paresthesia is common. Significant overlap occurs with early symptoms. Less commonly, a demyelinating condition resembling multiple sclerosis may appear in late disease. Symptoms are variable and, as in multiple sclerosis, may undergo exacerbations and remissions. CT and magnetic resonance imaging (MRI) may reveal multiple white matter lesions.23 Chronic inflammation also may occur in the skin, causing a seldom-recognized late cutaneous manifestation of Lyme disease, acrodermatitis chronica atrophicans.24 This condition usually involves the skin of distal extremities at the site of a tick bite. It is characterized in its initial stages by an edematous infiltration, which progresses to an atrophic lesion resembling localized scleroderma in its more established form. B. burgdorferi has been demonstrated in the skin of patients with acrodermatitis chronica atrophicans as well as positive findings on serologic studies.
Chapter 132 / Tick-Borne Illnesses
Cardiac Manifestations
1775
ings therefore are similar to those in rheumatoid arthritis, except that rheumatoid factor and antinuclear antibody assays yield a negative result in Lyme arthritis. Radiography may reveal nonspecific abnormalities such as juxta-articular osteoporosis, cartilage loss, cortical or marginal bone erosions, and joint effusions.
1776
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
Diagnostic Strategies The diagnosis of Lyme disease is based primarily on clinical and epidemiologic features, and identification of the disorder often is difficult, especially in the early stage. A history of tick bite is elicited in only approximately one third of cases. Erythema migrans is present in most patients and, in endemic areas, is considered diagnostic. Isolated late symptoms may emerge months after the initial infection, however, and the patient may not recall the rash. The disease should be considered in patients who live in or have visited an endemic area and who present during the summer months with nonspecific symptoms suggesting a viral illness or meningitis. In addition, the development of monarticular arthritis, multiple neurologic abnormalities, or heart block in previously healthy patients should raise the suspicion of Lyme disease. Results of routine laboratory studies are nonspecific and such studies generally are not helpful in diagnosing Lyme disease. Abnormalities may include an elevated erythrocyte sedimentation rate, mild anemia, total WBC count in the normal range with a decreased absolute lymphocyte count, microhematuria, proteinuria, and increased alanine transferase level.25 Cultures of blood, tissue, and body fluids (including CSF and synovial fluid) for B. burgdorferi and direct visualization techniques are difficult to perform properly and have such a low yield that they are not clinically useful.26,27 Serologic testing is the most practical and useful means of confirming a clinical diagnosis of Lyme disease, but it is not without limitations. Results of serologic tests must be interpreted cautiously within the clinical context, and such tests should be regarded as only adjuncts in the diagnostic process. Current serologic tests measure host antibody response (for both IgG and IgM) to B. burgdorferi. Problems with the performance of these tests and interpretation of findings often result in diagnostic confusion. False-negative and especially false-positive results are common. The antibody response to B. burgdorferi develops slowly. The peak of IgM titers appears between 3 and 6 weeks after the onset of illness. Earlier in the course of the illness, IgM titers may be negative. IgM usually returns to nondiagnostic levels 4 to 6 weeks after their peak, but elevations may persist. IgG antibody may be detectable 2 months after exposure and peaks at approximately 12 months. Early antibiotic therapy may blunt or even abolish the antibody response. During the first month of illness, both IgM and IgG titers should be determined, preferably in acute and convalescent serum samples. In approximately 20 to 30% of patients, a positive response occurs with acute samples, whereas even after antibiotic treatment, a positive response occurs with convalescent samples obtained 2 to 4 weeks later in approximately 70 to 80% of patients. After that time, most patients demonstrate a positive IgG antibody response, and a single test usually is sufficient. In patients with illness lasting longer than 1 month, a positive IgM test result alone is likely to be false positive. Therefore, a positive IgM response should not be used to support the diagnosis after the first month of infection. Testing with the enzyme-linked immunosorbent assay (ELISA) is the cornerstone of laboratory diagnosis of Lyme disease. Although ELISA alone has a sensitivity of 89% and a specificity of 72%, a positive test result in patients with a low pretest probability (i.e., less than 0.20) of Lyme disease is more likely to be a false positive than a true positive.28 In patients with a positive or equivocal ELISA result, a confirmatory Western blot assay should be ordered.29 Specimens that yield a negative result on ELISA are not tested further. Criteria for positive Western immunoblotting (requiring the presence of bands at particular locations) have been adopted by the CDC.29
IgG (and occasionally IgM) antibody may persist for several years after adequate treatment and symptom resolution. Persistent seropositivity is not diagnostic of ongoing infection. Even an IgM response cannot be interpreted as a demonstration of recent infection or reinfection unless the appropriate clinical characteristics are present. IgG antibody developed after natural infection does not always confer immunity against future infection by B. burgdorferi. Patients who are treated for erythema migrans may become reinfected; patients with Lyme arthritis, however, usually have high antibody titers to many spirochetal proteins and seem not to become reinfected.30 False-positive ELISA results are common. Serologic crossreactivity can occur between B. burgdorferi and other spirochetes, most notably Treponema pallidum. False-positive results for Lyme disease also can occur with relapsing fever, gingivitis, leptospirosis, enteroviral and other viral illnesses, rickettsial diseases, autoimmune diseases, malaria, and subacute bacterial endocarditis.31 In addition, it is estimated that up to 5% of the normal population will “test positive” for Lyme disease by ELISA. Bayes’ theorem states that if the pretest likelihood of the disease is low, then the positive predictive value is low: A positive test result is more likely to be a false-positive result. For this reason, screening serologic tests are not indicated in the absence of objective clinical evidence of Lyme disease.31 Patients suspected of having acute Lyme neuroborreliosis should be evaluated with serologic tests and routine CSF examination. Paired serum and CSF samples should be obtained to evaluate for intrathecal production of antibody, although most patients with neuroborreliosis have positive results on serum serologic testing, thereby making additional laboratory confirmation with CSF serology unnecessary.32 Polymerase chain reaction (PCR) assay is superior to culture for the detection of B. burgdorferi in synovial fluid33 and has a sensitivity of 73% and specificity of 99% in untreated Lyme arthritis.34
Differential Considerations Although Lyme disease manifests in many ways, each stage has characteristic clinical findings that are helpful in narrowing the scope of a differential diagnosis that at first may seem overwhelmingly broad. Early Lyme disease (erythema migrans and associated constitutional symptoms) may be easily confused with a variety of other diseases, especially if the characteristic rash of erythema migrans is absent. A common clinical presentation is a flulike illness with headache, nausea, fever, chills, myalgias, arthralgias, stiff neck, and anorexia, occurring during the summer months. Even in endemic areas during the summer months, most patients with such symptoms do not have Lyme disease. When headache and stiff neck are the predominant symptoms, the principal diagnostic distinction to be made is between Lyme disease and the enteroviral diseases (and other causes of aseptic meningitis). The enteroviral diseases also have their peak incidence during the summer months; however, diarrhea, commonly associated with enteroviral infection, is not a feature of Lyme disease. Abdominal pain, anorexia, and nausea suggest hepatitis; sore throat, adenopathy, and fatigue suggest mononucleosis; and myalgias and arthralgias suggest connective tissue diseases. The rash of erythema migrans is characteristic of but not pathognomonic for Lyme disease. Some patients are not aware of having had such a rash, and in others, its appearance is atypical. Secondary lesions may be confused with the target lesions of erythema multiforme, which generally are smaller and nonexpanding. Erythema multiforme also may involve the mucous membranes, palms, and soles; erythema migrans does not. The presence of a malar rash in association with Lyme
1777
Although previous expert consensus has recommended that persons bitten by deer ticks (I. scapularis) should not routinely receive antimicrobial chemoprophylaxis,5 this recommendation should be modified in accordance with the findings of a well-designed trial in which a single 200-mg dose of doxycycline effectively prevented Lyme disease when given within 72 hours after tick bite.37 A single 200-mg dose of doxycycline should be considered for adult patients and children 8 years of age and older (4 mg/kg, up to a maximum dose of 200 mg) when all of the following criteria are met: (1) the tick is an adult or nymphal I. scapularis, (2) the tick has been attached for 36 hours or more as indicated by certainty of the time of exposure or the degree of engorgement, (3) proplylaxis can be started within 72 hours after tick removal, (4) the local rate of infection of these ticks with B. burgdorferi is 20% or greater, and (5) doxycycline is not contraindicated. Infection rates of 20% or greater of ticks with B. burgdorferi generally are reported from highly endemic areas such as New England, parts of the mid-Atlantic region, and parts of Minnesota and Wisconsin. Most other areas of the United States do not have infection rates high enough to warrant prophylaxis.5 The efficacy of single-dose doxycycline in patients who present more than 72 hours after removing a tick is unknown. In children, dosing and efficacy of prophylactic treatment have not been evaluated. The effectiveness of doxycycline for the prevention of other infections transmitted by I. scapularis ticks (e.g., babesiosis, human granulocytic ehrlichiosis) is unknown and should not be assumed.37 Other antimicrobial agents that are effective for the treatment of Lyme disease (e.g., amoxicillin) and even other regimens of doxycycline (e.g., 100 mg twice daily) have unknown efficacy for Lyme disease prophylaxis. Bites from Dermacentor variabilis and Amblyomma americanum do not require prophylactic treatment. Any patient who has been bitten by a tick should be instructed to seek medical evaluation if symptoms of tick-borne illness develop.
Management
Early Disease
Prompt treatment of early disease can shorten the duration of symptoms and prevent progression to later stages of disease. Most of the various manifestations of Lyme disease can be treated successfully with oral antibiotic therapy, with the
Prompt antibiotic therapy is essential in early Lyme disease because it generally shortens the duration of the rash and associated symptoms and, more important, prevents later illness in most patients. Some patients with severe early
Vaccination No vaccine against Lyme disease is currently available in the United States. The LYMErix vaccine (SmithKline Pharmaceuticals, Philadelphia), initially licensed in 1999, was withdrawn from the market in 2002. The vaccine, directed against the outer surface protein A of B. burgdorferi (OspA), was apparently safe and efficacious but required multiple and repeated doses for optimal protection. Ongoing questions about its safety and cost-effectiveness dampened demand for the vaccine.37 A history of vaccination with the previously licensed vaccine should not change the approach to ED management. Because protective immunity produced by the vaccine is short-lived, it is unlikely that previous vaccination will provide any residual protective effect. Vaccination may cause a persistently positive ELISA result but a negative Western blot result.
Prophylaxis and Asymptomatic Tick Bites
Chapter 132 / Tick-Borne Illnesses
exception of neurologic abnormalities, which usually require intravenous therapy. Treatment of Lyme disease is summarized in Table 132-3.
disease suggests systemic lupus erythematosus. Erythema nodosum generally causes more painful induration than erythema migrans and has a predilection for the extensor surfaces of the legs. Erythema marginatum of acute rheumatic fever also is in the differential diagnosis for erythema migrans; the Lyme disease rash differs in comprising generally fewer, larger, less evanescent lesions that migrate more slowly.35 Atypical erythema migrans manifesting as an urticarial rash may suggest hepatitis B infection or serum sickness. Other cutaneous entities in the differential diagnosis for erythema migrans include cellulitis, fungal infection, fixed drug-related eruptions, plant dermatitis, and insect or spider bites. Lyme disease must be considered in a patient with any atypical rash accompanied by a “viral syndrome” or meningitis-like illness, especially during the months of peak incidence. Acute rheumatic fever, coronary artery disease, or viral myocarditis may be suggested by the cardiac manifestations of Lyme disease. The carditis of Lyme disease, like the carditis of rheumatic fever, may follow pharyngitis and migratory polyarthritis. Erythema marginatum usually occurs with the onset of arthritis, in contrast with erythema migrans, which usually precedes the carditis. Although some patients with Lyme disease may satisfy the clinical aspects of the Jones criteria for acute rheumatic fever, they lack evidence of a preceding streptococcal infection; in addition, valvular involvement is not a prominent feature of Lyme carditis. The differential diagnosis of the neurologic manifestations caused by Lyme disease is extensive. Considerations include aseptic meningitis, herpes simplex encephalitis, Bell’s palsy of other causes, multiple sclerosis, Guillain-Barré syndrome, dementia, primary psychosis, cerebral vasculitis, and brain tumor. Neurologic symptoms often occur in the absence of any epidemiologic clues or preceding clinical symptoms suggestive of Lyme disease, making the diagnosis particularly challenging. Lyme arthritis may mimic other immune-mediated disorders. The arthritis of Lyme disease generally is asymmetric, oligoarticular, and episodic. In contrast to patients with rheumatoid arthritis, those with Lyme arthritis rarely have symmetric polyarthritis, morning stiffness, a positive result on rheumatoid factor assay, or subcutaneous nodules. Lyme arthritis commonly is mistaken for seronegative rheumatoid arthritis; however, Lyme arthritis is most similar to the spondyloarthropathies, particularly reactive arthritis.36 Lyme disease and Reiter’s syndrome both commonly cause huge knee effusions, but in Lyme disease, absence of the extraarticular features of Reiter’s syndrome (conjunctivitis, urethritis or cervicitis, balanitis, keratosis blennorrhagica) at the time of the arthritis helps distinguish it from Reiter’s syndrome. In children, Lyme arthritis may mimic juvenile rheumatoid arthritis, but joint involvement in Lyme disease usually occurs in short, intermittent attacks, and iridocyclitis typically is absent. Rheumatoid factor titers will be negative in both juvenile rheumatoid arthritis and Lyme disease. The diseases resemble one another closely enough to have been confused at the time of the initial description of Lyme disease. Other diseases in the differential diagnosis for Lyme arthritis include acute gouty arthritis, septic arthritis, gonococcal arthritis, rheumatic fever, polymyalgia rheumatica, and the temporomandibular joint syndrome.
1778
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
Table 132-3 Treatment of Lyme Disease SYNDROME/MANIFESTATION
DRUG
ADULT DOSAGE
Early Lyme disease
Doxycycline† or Amoxicillin Alternative: Cefuroxime axetil or Erythromycin (less effective than doxycycline or amoxicillin)
100 mg PO bid for 21 days
Neurologic disease Facial nerve paralysis Lyme meningitis‡
Cardiac disease Mild§ More severe
250–500 mg PO tid for 21 days
25–40 mg/kg/day divided tid
500 mg PO bid for 21 days
250 mg bid
500 mg PO qid for 14–21 days
With an isolated deficit, oral regimens for early disease, used for at least 30 days, may suffice. For a deficit associated with other neurologic manifestations, intravenous therapy is warranted (see below). Ceftriaxone 2 g IV by single dose for 14–28 days 75–100 mg/kg/day IV Penicillin G 20 million units daily in divided 300,000 U/kg/day IV doses for 10–14 days Alternative: 1 g IV every 6 hr for 10–21 days Chloramphenicol Doxycycline† or Amoxicillin Ceftriaxone or Penicillin G
Arthritis
PEDIATRIC DOSAGE*
Oral: Doxycycline† or Amoxicillin Parenteral: Ceftriaxone or Penicillin G
100 mg PO bid 250–500 mg PO tid 2 g IV daily by single dose for 14–21 days
25–50 mg/kg/day divided tid 75–100 mg/kg/day IV
20 million units daily in divided doses for 14–21 days
300,000 U/kg/day IV
100 mg PO bid for 30 days 500 mg PO tid for 30 days
50 mg/kg/day divided tid
2 g IV by single dose for 14–21 days
75–100 mg/kg/day IV
20 million units daily in divided doses for 14–21 days
300,000 U/kg/day IV
*Pediatric dosage should not exceed adult dosage. † Tetracycline, 250 to 500 mg PO qid, may be substituted for doxycycline. Neither doxycycline nor any other tetracycline should be used for children younger than 8 years of age or for pregnant or lactating women. ‡ Regimens for radiculoneuropathy, peripheral neuropathy, and encephalitis are the same as those for meningitis. § Oral regimens are reserved for mild cardiac involvement (see text). Modified from Abramowitz M (ed): Med Lett 42:37, 2000; and Wormser GP, et al: The clinical assessment, treatment, and prevention of Lyme disease, human granulocytic anaplasmosis, and babesiosis: Clinical practice guidelines by the Infectious Diseases Society of America. Clin Infect Dis 43:1089, 2006.
disease, however, progress to later stages despite appropriate antibiotic regimens. The drug of choice for men, nonpregnant and nonlactating women, and children older than 8 years of age is doxycycline, 100 mg twice daily for 3 weeks.5 An advantage of doxycycline is that it also is effective for treatment of human granulocytic ehrlichiosis, which is transmitted by the same tick that transmits Lyme disease. Pregnant or lactating women and children younger than 8 years of age should receive amoxicillin, 500 mg PO (20 to 40 mg/kg per day in three doses for children). Cefuroxime axetil has been shown to be as effective as doxycycline38 and may be used in children of any age, but cephalexin is ineffective in Lyme disease. Macrolide antibiotics are not recommended as first-line agents for therapy for early Lyme disease.38 They should be reserved for patients who cannot tolerate doxycycline, amoxicillin, and cefuroxime axetil. Macrolide regimens5 for adults include azithromycin, 500 mg PO daily for 7 to 10 days; erythromycin, 500 mg PO four times daily for 14 to 21 days; and clarithromycin, 500 mg PO twice daily for 14 to 21 days. A Jarisch-Herxheimer–type reaction may occur in the first 24 hours of antibiotic treatment, consisting of fever, chills, myalgias, headache, tachycardia, increased respiratory rate, and mild leukocytosis.39 Defervescence usually takes place
within 12 to 24 hours, and the patient’s symptoms can be managed with bedrest and aspirin. The pathogenesis of this reaction is controversial, but it probably is caused by the killing of spirochetes with release of pyrogens. The Jarisch-Herxheimer reaction occurs more commonly with penicillin and doxycycline than with erythromycin, probably because of their superior spirocheticidal activity.
Early Disseminated Infection Neurologic Disease For patients with relatively mild symptoms (e.g., solitary facial nerve palsy with normal findings on CSF examination), doxycycline or amoxicillin can be used in the same dosage as for early disease, but the duration of therapy should be extended to 30 days. The use of prednisone for facial nerve palsy from Lyme disease has been suggested but is not currently recommended. For patients with other objective neurologic abnormalities (e.g., meningitis or encephalitis, peripheral neuropathies, cranial neuritis other than facial nerve palsy) or evidence of the spirochete in the CSF, parenteral antibiotic therapy is required. Ceftriaxone, 2 g/day IV for 14 days (75 to 100 mg/kg/ day for pediatric patients), or penicillin G, 18 to 24 million
Cardiac Disease Patients with mild cardiac conduction system involvement (first-degree AV block with a PR interval less than 0.30 second) and no other significant symptoms usually can be treated safely on an outpatient basis with oral doxycycline or amoxicillin for 21 to 30 days.38 Patients with higher degrees of AV block, including first-degree block with a PR interval greater than 0.30 second or evidence of global ventricular impairment, should be hospitalized for cardiac monitoring and treatment with parenteral antibiotics. Either penicillin G, 18 to 24 million units IV, or ceftriaxone, 2 g daily for 21 days (50 to 80 mg/kg/ day for children), may be used. The benefit of adjuvant use of aspirin or prednisone in treating Lyme carditis is uncertain. Temporary cardiac pacing may be necessary in patients who have severe heart block with hemodynamic instability. The block generally resolves completely with antibiotic treatment, so the recognition of Lyme carditis in young patients with unexplained heart block is critical to avoid unnecessary permanent pacemaker implantation.
Late Infection Arthritis In established Lyme arthritis, the response to antibiotic therapy may be delayed for several weeks or months.5 Thirtyday oral regimens such as doxycycline, 100 mg PO twice daily, or amoxicillin, 500 mg three times daily, usually are effective and, for reasons of cost and convenience, may be selected as first-line therapy given on an outpatient basis before use of parenteral antibiotic therapy is considered.38 Persistent or recurrent joint swelling after recommended courses of antibiotic therapy can be treated with another 4-week course of oral antibiotics or with a 2- to 4-week course of intravenous ceftriaxone.5 A small percentage of patients with Lyme arthritis, particularly those with HLA-DR4 specificity or antibody reactivity with OspA, may have persistent joint inflammation despite treatment with either oral or intravenous antibiotics.22 Such patients often do not respond to any antibiotic therapy and may require arthroscopic synovectomy. Neurologic Disease Patients with late neurologic disease affecting the central or peripheral nervous system should be treated with ceftriaxone (2 g once a day intravenously for 2 to 4 weeks). Alternative parenteral therapy may include cefotaxime (2 g IV every 8 hours) or penicillin G (18 million to 24 million units daily, given in divided doses every 4 hours). Response to treatment is usually slow and may be incomplete.
Lyme Disease and Pregnancy Similar to the spirochetal agents of syphilis and relapsing fever, B. burgdorferi can be passed transplacentally. In rare cases, Lyme disease acquired during pregnancy may lead to infection of the fetus and possibly to stillbirth, but adverse effects on the fetus have not been documented conclusively. Counseling the termination of a pregnancy because of maternal Lyme disease is unwarranted. Lyme disease contracted during pregnancy can be treated and cured. Treatment for pregnant patients can be identical to that for nonpregnant patients with the same disease mani-
1779
festations, except that doxycycline should be avoided.5 Most women give birth to normal infants despite documented Lyme borreliosis during their pregnancies.41
■ RELAPSING FEVER Perspective Relapsing fever is caused by bacteria of the Borrelia species, order Spirochaetales. Human Borrelia infections occur worldwide and all are associated with arthropod vectors. The epidemic (louse-borne) form of relapsing fever is caused solely by Borrelia recurrentis and is found mostly in Africa, where mortality rates can reach 70% with outbreaks. The endemic form, tick-borne relapsing fever, is caused by a group of closely related Borrelia species, their names derived from the species names of Ornithodoros tick vectors that carry them. The more common ones in North America are Borrelia hermsii, Borrelia turicatae, and Borrelia parkeri.42 B. burgdorferi has been recognized as the etiologic agent of the third and most recently described borrelial disease, Lyme disease.
Principles of Disease Tick-borne relapsing fever is maintained in an animal reservoir consisting primarily of wild rodents, including squirrels, mice, rats, chipmunks, and rabbits. It is found predominantly at altitudes of 2000 to 7000 feet in coniferous forest habitats.43 The tick vectors are argasids belonging to several species of the genus Ornithodoros, which routinely reside in the nests and burrows of their mammalian hosts. Ticks acquire the infection by feeding on a spirochetemic rodent. The borreliae remain viable in the ticks for several years and can be passed transovarially to the next generation; thus, the tick is a major reservoir and vector. These soft ticks feed for brief periods (15 to 30 minutes), usually at night, and their painless bite generally is unnoticed by the sleeping victim. Transmission occurs by injection of infected saliva through the bite site or intact skin. Less common modes of transmission (e.g., by way of venipuncture equipment in intravenous drug users) have been reported. In the United States, relapsing fever occurs primarily in the western Mountain and Pacific states, including Montana, Wyoming, Nevada, Colorado, California, and Washington. Persons who come in contact with infected ticks from wild rodents are at greatest risk. Outbreaks have been reported among groups of persons sleeping overnight in hunting cabins inhabited by wild rodents.42-44 In tick-borne relapsing fever, the initial febrile episode lasts 3 days. This is followed by an asymptomatic period of variable duration but usually approximately 7 days, during which patients generally feel better and may return to their usual daily activity levels under the assumption that they have recovered from another viral illness. Relapse then occurs, with symptoms that mimic those of the original illness. With tickborne relapsing fever, this cycle repeats itself three to five times. Each successive relapse usually is less severe. Relapse is caused by the spirochete’s unique ability to undergo antigenic variation within the body of the infected host. Each successive antigenic variation is cleared from the bloodstream by specific host antibodies, and a characteristic relapsing febrile course results.42 Clinical illness manifests in two classic stages as each fever episode resolves. The first stage is called the “chill” phase (high fevers with reported temperatures of to 106.7° F, mental status changes, tachycardia, and tachypnea) lasting approximately 30 minutes, followed by a “flush” phase (rapid tem-
Chapter 132 / Tick-Borne Illnesses
units daily IV for 10 to 14 days, may be used.38 Ceftriaxone may be more effective than penicillin, and many experts recommend longer courses (e.g., up to 4 weeks).40 In cases of penicillin or cephalosporin allergy, oral doxycycline for 30 days may be used.
1780
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
perature decrease, sweats, and hypotension), which can be confused with a Jarisch-Herxheimer reaction.45
Clinical Features After a postbite incubation period of 4 to 18 days, during which time the host concentration of spirochetes increases, fever of abrupt onset occurs, often accompanied by shaking chills, headache, arthralgias, myalgias, nausea, and vomiting. Occasionally a pruritic eschar may be noted at the site of the tick bite, but this usually is absent by the onset of clinical symptoms. Consequently, the nonspecific nature of the clinical presentation often leads to misdiagnosis of the disease as a viral illness. The patient’s temperature is high, and generalized muscle weakness and lethargy are common. Hepatomegaly, splenomegaly, and jaundice are sometimes seen. Neurologic involvement is less common but can manifest as delirium, nuchal rigidity, peripheral neuropathy, or pupillary abnormalities. Uveitis has been described.46 A macular or petechial skin rash, more apparent on the trunk than on the extremities, may be present. Recent severe cases of tick-borne relapsing fever resulting in ARDS in California and Nevada near the Lake Tahoe area and in the state of Washington prompted a comprehensive epidemiologic investigation of cases in those areas over a 10year period. This study showed that ARDS may be more common than was previously suspected. Reported occurrence rates for Jarisch-Herxheimer reaction were between 6% and 21%; for hypoxia, 16%; for elevated liver function test values, 8%; and for ARDS, 6%. Forty-six percent of the patients required hospitalization.43
Diagnostic Strategies The definitive diagnosis of relapsing fever depends on the demonstration of spirochetes in the peripheral blood during a febrile episode. This is not a typical finding with other spirochetal diseases.47 In most cases, spirochetes are readily visible on a routine blood smear prepared with Wright or Giemsa stain. Thick or thin blood smears, such as those prepared for malaria evaluation, also are satisfactory. The organisms are seen within the plasma spaces between blood cells or may overlie the blood cells. Several organisms per high-power field typically are visible in smears from febrile patients with relapsing fever.42 Blood specimens for the smears should be obtained as the temperature curve swings up, and repeated samples may be required before a positive result is observed, because sensitivity approaches only 70%. Spirochetes also may be visible in wet mounts with the use of phase contrast microscopy. Cultures, although the most sensitive diagnostic method available, require a special medium and do not yield rapid results and so are not commonly performed. Genus-specific PCR testing has now been successfully used and may be higher in sensitivity than either serology or blood smear, especially in the acute phase of disease.48 Serologic testing offered by the CDC can be accessed through local and state health departments. Nonspecific laboratory findings may include mildly increased bilirubin and liver function levels, thrombocytopenia, and elevated erythrocyte sedimentation rate.45
Differential Considerations On initial presentation, the differential diagnosis is extensive; however, it narrows with the occurrence of relapse. A history of possible soft tick exposure together with recurrent fever should suggest the diagnosis. Other conditions that initially may be considered include malaria, typhus, dengue, yellow
fever, Colorado tick fever, and tularemia. Careful examination of blood smears, together with clinical data and other laboratory tests, will aid in making the correct diagnosis.
Management Relapsing fever is effectively treated with tetracycline or erythromycin. Tetracycline should be avoided in children younger than 8 years and in pregnant women. Tetracycline or erythromycin should be given in an oral dose of 500 mg for 7 days; single-dose therapy is also effective.42 Other treatment regimens, including doxycycline and chloramphenicol, have been recommended. Treatment with penicillin G has been associated with an increased rate of relapse. Success with ceftriaxone has been reported in a patient with relapsing fever that did not respond to penicillin. Prophylaxis with doxycycline for tick-borne relapsing fever in exposed subjects in high-risk infested areas has been shown to be effective.49 As many as one third of patients will experience a JarischHerxheimer–type reaction during treatment with antibiotics. The reaction can be severe, especially with louse-borne relapsing fever. This phenomenon may be related to release of high levels of cytokine intermediaries or endogenous opioids. Approximately 4 hours after antibiotic treatment and coinciding with the clearance of spirochetes from the blood, the patient usually experiences an increase in temperature and severe rigors, accompanied by a drop in leukocyte and platelet counts and onset of hypotension. Anticipation of this reaction is crucial because intravenous volume expansion with saline solution may be required to maintain the blood pressure; the reaction can be more threatening than the disease itself. Meptazinol, an opioid antagonist with agonist properties, has been proposed for use in treatment of this reaction. Prognosis is good in treated patients with relapsing fever, with approximately 95% achieving complete recovery. Bad prognostic signs include the presence of jaundice, high spirochete counts in the blood, and hypotension.42 Transplacental transmission can occur in infected pregnant women. Perinatal death of the fetus or infant and spontaneous abortions occur in nearly 50% of the cases in pregnant women.50 Death is rare in tick-borne relapsing fever and is limited to infants and the elderly.
■ TULAREMIA Perspective Tularemia was first characterized in 1837 by Soken, who described a febrile illness with generalized lymphadenopathy in people who had eaten infected rabbit meat.51 In 1912, McCoy first isolated Bacterium tularense, now known as Francisella tularensis, from rodents in Tulare County, California, giving rise to the name of the disease. Edward Francis, for whom the genus Francisella was later named, contributed much to the understanding of the bacteriology and epidemiology.52 Tularemia occurs worldwide and is endemic between 30 and 71 degrees north latitude. The incidence of tularemia is low. There were 247 reported cases of tularemia in the United States between 2004 and 2005, although it is not a notifiable disease in all states.53 Tularemia has been seen in every state but is most common in the southwest central region (Arkansas, Louisiana, Oklahoma, Texas, and Mississippi). Fifty-six percent of reported cases have come collectively from Missouri, Oklahoma, South Dakota, and Arkansas. It is more common in men than in women. Persons at increased risk for infection include hunters, trappers, butchers, agricultural
Principles of Disease Francisella tularensis is a small pleiomorphic gram-negative coccobacillus and is a facultative pathogen of macrophages, neutrophils, and nonphagocytic cells such as hepatocytes and alveolar epithelial cells.64 Two serologically identical types of F. tularensis organisms are responsible for human disease and can be distinguished from each other on the basis of geographic distribution, fermentation reactions, and virulence. Jellison type A (F. tularensis biovar tularensis), the predominant biovar in North America, is associated with ticks and rabbits and causes severe disease in humans. Strain B (F. tularensis biovar holarctica) occurs in Asia, Europe, and, to a minor extent, North America; it is associated with rodents and causes milder disease in humans.64 Tularemia manifests in different ways, depending on the portal of entry of the organism. The primary route of infection by F. tularensis is through the skin. Entry can occur through hair follicles or through small cuts and abrasions that may be contaminated by exposure to an infected animal; tick exposure can also introduce the bacteria.54 Because the bacterium has not been isolated from the salivary glands of ticks, it is thought that they transmit the organism through their feces.51 Scratching after a tick bite introduces the infected feces into the skin. Inhalation or ingestion of the organism or transmission through the conjunctivae also can cause infection. The incubation period is approximately 2 to 6 days, depending on the size of the inoculum. After penetration of the skin or epithelial membrane, the organism usually spreads to the regional lymph nodes. An erythematous tender papule develops at the primary infection
1781
site, followed by inflammation and skin ulceration. The regional nodes enlarge, necrose, and may rupture. The necrotic, purulent, painful lymph node is termed a bubo. In the ulceroglandular form of the infection, the organism may not spread further than the regional lymph nodes. If the inoculum is sufficiently large or the host defenses are inadequate, bacteremia ensues, with dissemination to phagocytic cells of the reticuloendothelial system. Pulmonary tularemia may result from inhalation of smallparticle aerosols containing F. tularensis or may be secondary to hematogenous dissemination. Small areas of localized pneumonitis most commonly are seen, although chest radiographic findings are nonspecific; lobar consolidation or abscess formation is rare. Oculoglandular tularemia occurs when the conjunctiva becomes infected from contact with material from an ulcer or a contaminated finger. Typhoidal tularemia follows systemic spread of F. tularensis from the oropharynx and probably the gastrointestinal tract when a large inoculum is swallowed.
Clinical Features Tularemia has six clinical presentations, depending on whether disease is localized to an entry site and its regional lymph nodes—ulceroglandular, glandular, oculoglandular, and oropharyngeal forms—or is more invasive and generalized— typhoidal and pulmonary forms. Ulceroglandular tularemia is the most common form of the disease (accounting for approximately 80% of the cases). Typically, a skin lesion on an extremity at the site of primary inoculation begins as an erythematous papule, which then ulcerates 2 to 3 days later.54 The ulcer is slow to heal and often is still present when the subsequent regional lymphadenopathy and fever develop. The distribution of the regional adenopathy reflects the primary entry site; patients with tick-borne tularemia usually have inguinal or femoral adenopathy, whereas those who acquire rabbit-associated tularemia have axillary or epitrochlear nodal involvement. Generalized lymphadenopathy also may be seen. Occasionally, nodes suppurate and drain.54 Glandular tularemia, the second most common form, is characterized by the development of lymphadenopathy (usually cervical) without an associated skin ulcer. Oculoglandular tularemia is seen in 1 to 2% of cases and is characterized by unilateral conjunctivitis with regional adenopathy involving preauricular lymph nodes. Oropharyngeal tularemia manifests as severe exudative pharyngitis with associated cervical lymphadenitis. It has been known to cause acute glaucoma.52,65 Typhoidal tularemia is a systemic form of the disease in which no obvious entry site can be found; it occurs in approximately 10% of cases. Only 10 to 50 organisms are required to induce disease; incubation time is 2 to 10 days.58 Symptoms and signs may include fever, chills, constipation or diarrhea, abdominal pain, and weight loss. A 30 to 60% case-fatality rate is associated with untreated typhoidal tularemia.60 Pulmonary tularemia is common and has symptoms similar to those of other bacterial pneumonias: fever and chills, cough (usually nonproductive), substernal burning, dyspnea, malaise, and prostration. It may result from either direct inhalation of aerosolized organisms or bacteremic spread from another site. Uncommon complications of tularemia include pericarditis, meningitis, endocarditis, peritonitis, appendicitis, perisplenitis, and osteomyelitis.54 Guillain-Barré syndrome associated with tularemia also has been reported.66 Tularemia is one of the most widely studied diseases with respect to potential biologic warfare. The United States devel-
Chapter 132 / Tick-Borne Illnesses
workers, campers, sheepherders, mink farmers, and laboratory workers.54,55 Ticks, lagomorphs (hares, rabbits), and rodents (mice, rats) are believed to be the most important sources of transmission to humans; however, the organism has been recovered from animals of more than 100 different species with significant epidemics linked to contact with a variety of them, including domestic cats.52,56 A large number of commercially distributed prairie dogs from Texas died from tularemia in 2002.57 The ticks most commonly involved in transmission in the United States are the deer tick (I. scapularis), the Lone Star tick (A. americanum), and the dog tick (D. variabilis), all of which have been associated with other tick-borne illnesses. Whereas mosquitoes are major vectors in many European countries,58,59 horse fly and deer fly bites have been implicated in endemic situations in the United States.60 Transmission to humans most commonly occurs through tick bites or handling infected animals. It also can occur with ingestion of infected food or water, with inhalation of dust or water aerosol, and through insect bites.58,60 Nonimmune laboratory workers who work with F. tularensis can acquire the disease. Person-to-person transmission is rare. Tularemia has a bimodal prevalence in the United States, with an increased incidence in May to August associated with tick-borne transmission and a December to January peak associated with hunting and skinning of infected mammals (primarily rabbits). Francisella tularensis has been found to coexist in reservoir populations harboring the agent responsible for Lyme disease.61 Eleven cases of pneumonic tularemia, found to be from aerosolization of contaminated vegetation clippings, were discovered in Martha’s Vineyard.62 Outside of the United States, tularemia has been confirmed in hundreds of cases in Kosovo, through rodent contamination of food. In addition, Sweden has reported a high number of cases, usually associated with aquatic environments and mosquitoes.52,59,63
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1782
oped an aerosolized form in the 1950s, and the Japanese allegedly contaminated prisoners with the disease in the 1930s.67 It was removed from the national list of notifiable diseases in 1995 but then was reinstated in view of the heightened biologic weapons threat.55 It is now classified by the CDC as one of the six category A critical biologic diseases.68 An aerosolized form of the bacterium would be the most likely delivery mechanism used in biologic warfare.69 With release of aerosolized particles, disease would manifest clinically as acute fever, progressive pneumonia, pleuritis, and hilar lymphadenopathy, beginning as early as 3 to 5 days after delivery.70 Only approximately 55% of emergency departments have been adequately educated on the recognition of and preparedness for tularemia.71
Diagnostic Strategies Diagnosis of tularemia is based on clinical findings and serologic testing. Antibody titers begin to rise approximately 7 to 10 days after exposure and peak in 3 to 4 weeks. In a patient with a clinical presentation suggesting tularemia, an antibody titer of 1 : 160 or greater in a single specimen is diagnostic. Confirmatory evidence is provided by a fourfold or greater rise in titer in a second sample obtained 2 weeks later. Unfortunately, titers of IgG and IgM can continue to be high for up to 10 years, and cell-mediated immunity can be maintained for up to 25 years.72,73 Rapid testing with PCR assay is available.74,75 A rapid point-of-care test using an immunochromatographic approach is currently being evaluated.76 Aspiration of affected lymph nodes for culture is not routinely recommended because of the associated risk to laboratory personnel. If tularemia is suspected, the laboratory should be alerted so that appropriate precautions can be taken in specimen handling and so that enriched culture medium can be used.
Management Isolation of patients with tularemia is not required. Streptomycin is the drug of choice for treatment of all forms of tularemia. When given intramuscularly in a dose of 30 to 40 mg/kg per day in two divided doses every 12 hours, streptomycin usually produces symptomatic improvement and resolution of fever in 1 to 2 days.54 After the third treatment day, one half of the dose is given for a total course of 7 to 14 days.54,58 With this regimen, relapses are unusual. Ulcers and tender lymph nodes usually heal within 7 to 10 days; however, enlarged nodes occasionally develop into fluctuant sterile buboes, requiring incision and drainage after completion of the course of antibiotics. Gentamicin also is effective for treatment (3 to 5 mg/kg/day for 10 to 14 days).58 Tetracycline and chloramphenicol are effective; however, the risk of relapse is greater than that associated with the aminoglycosides. Imipenem-cilastatin, an antibiotic without nephrotoxicity, has been used successfully to treat pulmonary tularemia in a patient with acute renal failure. Ceftriaxone is not effective against F. tularensis infections.77 Prophylaxis for possible exposure requires doxycycline, 100 mg twice a day for 14 days. Doxycycline or ciprofloxacin prophylaxis is recommended for a large biologic attack.70 An effective live vaccine strain has been available for vaccination against tularemia for nearly 50 years, but licensure approval has been delayed owing to a lack of understanding of the vaccine’s mechanism of action and attenuation.64 Because of recent interest in biologic warfare, research on tularemia vaccines has resurged. The overall mortality rate in untreated tularemia ranges from approximately 5 to 30%; the higher figure is associated
with severe disease or significant pulmonic involvement. With appropriate antibiotic treatment, death is rare (mortality rate of less than 1%).
■ ROCKY MOUNTAIN SPOTTED FEVER Perspective RMSF is an acute, febrile, systemic tick-borne illness caused by Rickettsia rickettsii. There are 11 other rickettsial species that cause human disease in other parts of the world. RMSF is found in North, South, and Central America. The number of reported cases in the United States more than tripled between 2001 and 2005, especially in suburban areas.78,79 RMSF ranges in clinical severity from mild or even subclinical illness to a fulminant disease with vascular collapse and death within several days of onset. It is the only rickettsiosis still associated with significant mortality, causing approximately 40 deaths in the United States each year,80 with a mortality rate ranging from 3 to 5% despite appropriate treatment. Before tetracycline and chloramphenicol were available, death occurred in as many as 30% of cases in the 1930s.79 The age-specific incidence is highest in children younger than 10 years, and casefatality rates are highest in persons older than 60 years.81 RMSF is a nationally reportable disease, with all cases to be registered with the respective state department. The recorded history of RMSF suggests that the disease was present at least before the white settlement of western North America, in afflicted Native American inhabitants of wooded Rocky Mountain regions. Early terms used to name the disease included “tick fever” and “black measles.” In 1899, RMSF was described as “an acute, endemic, noncontagious but probably infectious, febrile disease, characterized by a continuous moderately high fever, severe arthritic and muscle pains, and a profuse petechial or purpuric eruption in the skin, appearing first on the ankles, wrists, and forehead, but rapidly spreading to all parts of the body.” In 1906, the causative organism, Rickettsia rickettsii, was identified by Howard T. Ricketts, who also described the importance of the tick vector in transmission to humans. Although RMSF was first described in Montana and Idaho, it is now relatively rare in the Rocky Mountain states. Endemic in all 48 contiguous states except Maine, the disease continues to be most prevalent in the southeastern United States. RMSF has been reported in Canada, Central America, Mexico, and South America but never outside the Western Hemisphere. In 1987, four cases of RMSF were reported among residents of the Bronx in New York City; none of the affected persons had recently traveled to an area known for endemic disease—raising the possibility that other urban foci of RMSF may exist. RMSF also tends to be focally endemic, with clustering of cases within a larger endemic area that may correspond to “islands” of infected ticks. These areas, ecologically distinct from surrounding areas, may be ideally suited to ticks; they usually consist of wild open fields, deciduous forests with thick ground cover and poor water drainage, or uncultivated areas. In areas with frequent occurrence of RMSF (Oklahoma, North and South Carolina, Tennessee, and Pennsylvania), an infectivity rate of 2 to 15% of the tick population has been reported. North Carolina and Oklahoma carry the highest incidence rates (35% of all cases) for RMSF.79 Rickettsia rickettsii organisms are obligate intracellular bacteria that often occur in pairs and possess a cell wall similar in structure and chemical composition to that of gram-negative bacteria.82 R. rickettsii contain both RNA and DNA and, in contrast with other rickettsial organisms, can invade the nucleus as well as the cytoplasm.
Principles of Disease After introduction of R. rickettsii into the host by the tick vector, the organisms invade and multiply in the vascular endothelial cells. They then enter deeper areas of the vessel walls and infect vascular smooth muscle. Rickettsial organisms move from cell to cell by actin-based motility.86 Damage to endothelial cells not only exposes subendothelium but also releases tissue plasminogen activator and von Willebrand factor, thereby causing microhemorrhage, microthrombus formation, and increased vascular permeability. In addition, antibody forms, with antigen activating the complement system (type III immune response), and a cellular response is recruited. These widespread vascular lesions form the basis for most of the clinical features associated with RMSF. Hypotension, edema, and increased extravascular fluid space result from the increased small-vessel permeability. The early rash results from the vasculitis and the associated changes in permeability; later petechial and hemorrhagic lesions are secondary to the vasculitis and thrombocytopenia. Microinfarcts and focal lesions develop in various organs, including the brain, heart, lungs, kidneys, adrenal glands, liver, and spleen. Rickettsial encephalitis and diffuse microinfarcts are usual features of central nervous system involvement. An interstitial pneumonitis caused by direct lung invasion by the organism may occur, and acute respiratory distress syndrome (ARDS) can ensue. Acute renal failure and hypovolemic shock, the primary causes of death, can occur as early as the second week of illness.
Clinical Features Children from 5 to 9 years of age are the most common victims of RMSF. Two thirds of all cases are in children younger than 15 years. More than 90% present with a fever and rash. A history of tick bite or presence in possible tick-infested areas is elicited in 60 to 70% of all patients with RMSF, although only 49% of the pediatric population reports a tick bite.87 The incubation period ranges from 2 to 14 days, with a mean of 7 days.79 A short incubation period may indicate a more serious infection. Onset of symptoms usually is abrupt but is gradual in approximately one third of patients. Early symptoms are nonspecific and similar to those of many acute infectious diseases, making early diagnosis very difficult. “Typical” patients experience sudden onset of fever, severe headache, myalgias, prostration,
1783
and Signs in 262 Persons with Table 132-4 Symptoms Rocky Mountain Spotted Fever FREQUENCY DURING ILLNESS (%) SYMPTOM OR SIGN
Fever (temperature of 37.8° C– 38.9° C [100° F–102° F]) Headache, mild to moderate Fever (≥102° F) Any rash Myalgia, mild to moderate Rash, maculopapular Rash, palms/soles Triad of fever, rash, history of tick exposure Nausea/vomiting Headache, severe Abdominal pain Rash, petechial and hemorrhagic Myalgia, severe Conjunctivitis Lymphadenopathy Stupor Diarrhea Edema Ataxia Meningismus
ANY TIME
FIRST 3 DAYS
99
73
91 90 88 83 82 74 67
71 63 49 57 46 28 3
60 57 52 49 47 30 27 26 19 18 18 18
38 40 30 13 25 13 13 6 9 3 7 5
From Helmick CG, Bernard KW, D’Angelo LJ: Rocky Mountain spotted fever: Clinical, laboratory, and epidemiological features of 262 cases. J Infect Dis 150:480, 1984.
nausea, and vomiting. Tenderness may be noted in large muscle groups (Table 132-4). As many as 80% of patients may have gastrointestinal symptoms, secondary to myositis of the abdominal wall. Fever (temperature usually higher than 102° F) is nearly always present during the first 2 to 3 days of illness and may precede other signs by 1 week or more.79 Occasionally, the onset of illness is mild, with lethargy, headache, anorexia, and low-grade fever; these patients may remain ambulatory. Although the triad of fever, rash, and tick exposure traditionally has been seen in only approximately 3 to 18% of cases, newer data show that it is found in up to 45% of children with the disease.87 An extreme complication of RMSF is gangrene, which probably is induced by small-vessel occlusion.88
Cutaneous Manifestations Vasculitis secondary to rickettsial invasion of vascular endothelial cells causes the rash commonly associated with RMSF; however, the rash reportedly is absent in 4 to 16% of laboratory-confirmed cases, referred to as “Rocky Mountain spotless fever.” In addition, the rash may go unnoticed in dark-skinned patients. It usually appears on the third to fifth febrile day but can emerge as early as the second and as late as the sixth day. The initial lesions generally are restricted to the ankles and wrists, spreading to the palms and soles. The rash then spreads centripetally to the forearms, arms, legs, thighs, and trunk. Although usually spared, the face can be involved. Despite the common belief that the palms and soles are critical for diagnosis, they are not consistently involved (rash on the palms and soles is reported in approximately 50% of cases). Involvement of the scrotum or vulva can be an evasive clue for RMSF.89
Chapter 132 / Tick-Borne Illnesses
The American dog tick, Dermacentor variabilis, and the Rocky Mountain wood tick, Dermacentor andersoni, have been the responsible vectors of human RMSF cases in the United States to date79; however, the common brown dog tick, Rhipicephalus sanguineus, has recently emerged as a third vector.83 R. sanguineus has been the main RMSF vector in Mexico and Central America. The ticks feed on virtually any available warm-blooded animals and humans; the occurrence of R. rickettsii in the United States does not depend on the presence of any given order of mammal. Domestic dogs infected with R. rickettsii can demonstrate clinical illness similar to that seen in humans. Although dogs do not play an important role in the amplification cycle of RMSF, they can serve as a conduit for infected ticks, carrying them into close contact with pet owners. Dogs may serve as sentinels for RMSF in humans. Communication between physicians and veterinarians is important when cases of zoonotic diseases are detected.84 Humans serve only as accidental participants in the cycle of infection. A retrospective study revealed that none of 10 recipients of blood products found to be from donors with either confirmed or probable RMSF contracted the disease.85
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1784
A
B
Figure 132-7. A, Exanthem of Rocky Mountain spotted fever. B, Exanthem of Rocky Mountain spotted fever, close-up view. (From McGinley-Smith DE, Tsao SS: Dermatoses from ticks. J Am Acad Dermatol 49:363, 2003.) The rash of RMSF typically begins as 1- to 5-mm blanchable pink to bright red discrete macules that may be pruritic (Fig. 132-7). At this initial stage, the lesions fade when pressure is applied and are not palpable. A warm compress applied to the area enhances the rash. After 6 to 12 hours, the rash spreads centripetally. After 2 to 3 days, the rash becomes maculopapular and changes to a deeper red; at this stage, skin changes can be appreciated on light palpation. By approximately the fourth day, the rash becomes petechial and no longer fades with applied pressure. Applying tourniquets for several minutes or taking the blood pressure may cause additional petechiae to form distal to the site of occlusion (Rumpel-Leede phenomenon). Occasionally, the lesions coalesce to form large ecchymotic areas that may slough and form indolent ulcers (Fig. 132-8). Prompt institution of specific therapy can cause the initial nonfixed lesions to disappear rapidly, unlike the later fixed lesions. Patients who have had the typical rash may exhibit brownish discolorations at the site during the convalescent period.
Cardiopulmonary Manifestations Echocardiographic evidence of decreased left ventricular contractility secondary to myocarditis commonly is seen and often is detectable even before clinical signs of RMSF appear. Clinical manifestations of left ventricular dysfunction are uncommon, however, and hypotension and pulmonary edema, when present, usually have noncardiogenic causes. Chest radiographs may demonstrate cardiac enlargement. Electrocardiographic changes include low-voltage, nonspecific ST-T changes, firstdegree AV block, dysrhythmias (sinus and nodal tachycardia, paroxysmal atrial tachycardia, atrial fibrillation), and left ventricular hypertrophy. Most cardiac abnormalities are transient, but persistent echocardiographic changes have been described. Decreased systolic function, elevated serum cardiac markers, no finding of vascular lesions, and a fourfold rise in antibody titers are consistent with myocarditis from RMSF.90
Figure 132-8. Late appearance of rash: Rocky Mountain spotted fever
manifesting on lower extremity. (Courtesy of Theodore Woodward, MD.)
Interstitial pneumonitis and increased pulmonary capillary permeability may result from infection of the pulmonary capillaries with rickettsiae. Nonproductive cough and dyspnea secondary to pneumonitis are sometimes seen on presentation.91 Chest radiographic abnormalities are identified in approximately 25% of patients. These abnormalities include interstitial infiltrates, patchy alveolar infiltrates, pleural effusions, and
Neurologic Manifestations Neurologic manifestations of RMSF range from mild headache and lethargy to seizures and coma. Acute disseminated encephalomyelitis has been described.92 Headache, generally severe, is common, occurring in 50 to 90% of cases. Meningismus is present in 16 to 29% of patients. The CSF may be normal or show slight protein elevation and pleocytosis of both lymphocytes and polymorphonuclear cells (usually 8 to 35 cells/mL). CSF glucose level and opening pressure usually are normal. Resolution of eosinophilic meningitis during RMSF after appropriate antibiotic treatment has been reported. Fewer than 40% of patients have a positive CSF finding. Cerebral thrombovasculitis may cause focal neurologic deficits, which usually are transient. Seizures can occur, especially during the acute phase of the illness. Generalized cerebral dysfunction ranging from lethargy to coma can occur secondary to systemic toxicity (fever, hypotension, hyponatremia) or to vasculitic lesions involving the central nervous system. Coma is a late finding in patients with severe disease and is seen in less than 10% of cases. Some reports have described patients who remain alert but are amnesic for their illness after recovery. Other reported neurologic manifestations include transient deafness, tremor, rigidity, athetoid movements, paralysis, ataxia, opisthotonos, aphasia, and blindness.79,82 Generally, neurologic signs abate without residua; permanent neurologic deficits are rare. Behavioral disturbances and learning disabilities have been reported after recovery from RMSF-associated coma in children.
Diagnostic Strategies Most immediately available laboratory tests provide little help in diagnosing RMSF. Early in the course of the illness, the diagnosis is based primarily on clinical evidence, so epidemiologic features must be correlated with clinical signs and symptoms. The initial presentation of RMSF is similar to that of many acute febrile infectious diseases, and almost invariably a therapeutic decision must be made on clinical grounds alone, without the luxury of confirmatory laboratory evidence.93 Abnormalities such as thrombocytopenia and hyponatremia may be detected by routine laboratory tests, but they are nonspecific and unhelpful diagnostically. Up to 30% of patients present with anemia.94 A definitive diagnosis of RMSF requires positive results on one or more of several tests: serologic study, skin biopsy, or direct isolation and identification of the organism (Box 132-1).
Serology Rickettsial infection can be confirmed by demonstrating antibody rise in paired sera. Even with the most sensitive serologic tests, however, elevations in antibody titers do not occur until approximately 5 to 7 days after the onset of initial symptoms. Accordingly, serodiagnosis is retrospective. It is achieved by comparing acute serum, which typically yields negative findings, with convalescent serum, which yields positive results for antibodies. The indirect immunofluorescence assay (IFA) generally is considered the reference standard for diagnosis of RMSF and is the test currently used by the CDC and most state public health laboratories. It has a high specificity and sensitivity (94%).89 IFA can be used to detect either IgG or IgM antibodies. A RMSF latex agglutination test that report-
1785
Diagnostic Criteria for Rocky Mountain
BOX 132-1 Spotted Fever
Laboratory Criteria ■ Serologic evidence of a significant change in serum antibody titer reactive with Rickettsia rickettsii antigens between paired serum specimens, as measured by a standardized assay conducted in a commercial, state, or reference laboratory ■ Demonstration of R. rickettsii antigen in a clinical specimen by immunohistochemical methods ■ Detection of R. rickettsii DNA in a clinical specimen by the polymerase chain reaction (PCR) assay ■ Isolation of R. rickettsii from a clinical specimen in cell culture For confirmed cases, a significant change in titer must be determined by the testing laboratory; examples of commonly used measures of significant change include, but are not limited to, a fourfold or greater change in antibody titer as determined by indirect immunofluorescent antibody (IFA) assay or an equivalent change in optical density measured by enzyme-linked immunosorbent assay (EIA or ELISA). Case Classification (CDC Case Definition, 2004) Confirmed: The patient has a clinically compatible illness that is laboratory confirmed. Probable: The patient has a clinically compatible illness and serologic evidence of antibody reactive with R. rickettsii in a single serum sample at a titer considered indicative of current or past infection (cutoff titers are determined by individual laboratories). CDC, Centers for Disease Control and Prevention. From Diagnosis and management of tickborne rickettsial diseases: Rocky Mountain spotted fever, ehrlichioses, and anaplasmosis—United States. MMWR Recomm Rep 55(RR-4):18, 2006.
edly gives a turnaround result in less than 24 hours is available in selected laboratories.95 A recent study showed a 12% seroprevalence, with antibody titers of 1 : 64 or greater, in the pediatric population in the southeastern and south central regions of the United States.87 Accordingly, clinical correlation with titers in these regions is critical. Convalescent-stage blood samples are best obtained 2 to 3 weeks after the onset of clinical illness. Antibiotic therapy does not affect the time of appearance of antibodies or their ultimate titer if such treatment is begun several days after the onset of illness. However, if antibiotic therapy is initiated earlier in the course of the illness, the rise in titers can be delayed for 4 weeks or more. Under these circumstances, antibody titers should be tested again at 4 to 6 weeks after the onset of illness.
Skin Biopsy Identification by immunofluorescent assay and immunoperoxidase staining of R. rickettsii in biopsy specimens of the skin rash from patients with suspected RMSF are the best rapid diagnostic tests currently available.79 In experienced laboratories, the diagnosis of RMSF can be confirmed as soon as 4 hours after the specimen is obtained. The organisms can be detected as early as day 3 of clinical illness and as late as day 10. Unfortunately, this technique can be used only when a rash is visible for accurate localization of the biopsy site. Biopsy specimens generally are obtained with a 3-mm punch in the
Chapter 132 / Tick-Borne Illnesses
cardiomegaly with pulmonary edema. Pulmonary consolidation is rare. In severe cases, progression to noncardiogenic pulmonary edema and ARDS may occur.
PART III ■ Medicine and Surgery / Section Twelve • Infectious Diseases
1786
center of the skin lesion. Immunofluorescent demonstration of rickettsiae in frozen sections of skin biopsies has a sensitivity of 70%. Results of immunohistochemical staining of tissues at autopsy were positive in all fatal cases in one study, whereas IFA results were negative in the majority of cases.96 Failure to obtain a biopsy specimen of a rickettsial cutaneous lesion or failure to obtain sections through its center is associated with false-negative results. Treatment with antirickettsial drugs for 24 hours does not appreciably alter the sensitivity of the test; however, after 48 hours, rickettsiae are substantially reduced in numbers.
Isolation of Organism For most pathogenic infections, the standard diagnostic criterion is isolation and identification of the etiologic organism from the patient’s blood or tissues. This is seldom attempted in rickettsioses, however, because the isolation procedures are time-consuming, expensive, and hazardous to laboratory personnel. In addition, primary isolation of rickettsiae by inoculation in the yolk sac of a chick embryo usually fails because of the small number of organisms in the patient’s blood.
Differential Considerations Delayed diagnosis or misdiagnosis is the principal reason for the significant mortality associated with RMSF. Clinical diagnosis is difficult, especially early in the course of the illness, because of the nonspecific presentation. To prevent avoidable deaths, a diagnosis of RMSF must be considered in any patient with an unexplained febrile illness (with or without a rash and headache), even in the absence of a history of tick bite or travel to an area known to be endemic for the disease.97 An atypical presentation or manifestation of RMSF also must be considered during the differential diagnosis, including (1) absence of a rash (“Rocky Mountain spotless fever”) or late appearance of a rash, (2) predominant gastrointestinal features or abdominal pain suggestive of an acute condition in the abdomen, (3) cough and pulmonary congestion suggestive of pneumonitis, and (4) meningismus suggestive of viral meningitis.81 A presumptive diagnosis must be made and specific therapy initiated well before specific confirmatory laboratory values are available. A wide variety of other infections with similar exanthems can be confused with RMSF.89 The most common include meningococcal infection, measles (rubeola) and atypical measles, gonococcemia, infectious mononucleosis, toxic shock syndrome, and enteroviral infections. Less common diseases include dengue, leptospirosis, murine typhus, and epidemic typhus.
rate rises to 25%.89 For a select group of early-stage, mildly ill patients, outpatient therapy with oral antibiotics can be successful if the patient is reliable and close follow-up observation is arranged. More severely ill patients in whom the diagnosis is uncertain should be hospitalized for administration of intravenous antibiotics.
Antibiotics Antibiotic therapy is most effective when initiated during the early stages of disease, coincident with the initial appearance of the rash. Although data from randomized clinical trials regarding antibiotic selection for RMSF are lacking, doxycycline is still widely regarded as the therapeutic agent of choice for most patients.98 Chloramphenicol should be considered only for patients in whom the tetracyclines have caused significant adverse events and for pregnant women (except those who are near term). The recommended doses of doxycycline and chloramphenicol are summarized in Table 132-5. Although previous treatment guidelines recommended avoiding doxycycline in children younger than 8 years of age, the American Academy of Pediatrics and the CDC currently recommend doxycycline therapy as the agent of choice for treatment of RMSF in children of all ages.79,97,99 The risk of cosmetically perceptible tooth staining appears to be small for a single course of treatment and is subordinate to the potential lethality of this illness.93 The effectiveness of therapy depends on both the duration of therapy and the interval between the onset of illness and the initiation of therapy. Treatment should begin as early as possible and continue for 7 to 10 days or until the patient is afebrile for 2 to 5 days. Patients who are clinically ill should be hospitalized for parenteral antibiotic treatment. Response to treatment, as manifested by decreasing fever and subsiding rash, generally occurs 36 to 48 hours after beginning antibiotic therapy. Resistance to chloramphenicol or the tetracyclines has not been reported.81 Penicillin, erythromycin, cephalosporins, aminoglycosides, clindamycin, and sulfonamides are ineffective against RMSF. In fact, empirical use of these agents for presumed bacterial infections may potentially permit progression of the illness. Occasionally, secondary bacterial infection from the RMSF rash may occur. Although sulfonamides have recently become a mainstay for empirical treatment of MRSA for skin infections, the use of these agents should be avoided in RMSF patients, since their mechanism of inhibiting para-aminobenzoic acid may worsen the primary RMSF infection. The role
Table 132-5 Antibiotic Therapy for Rocky Mountain Spotted Fever*
Management Treatment of RMSF consists of antibiotic therapy, supportive care, and possibly administration of steroids. An understanding of the underlying pathophysiologic changes and an appreciation of the systemic complications that can occur in the patient afflicted with RMSF are necessary for the formulation of a balanced therapeutic regimen. The course of the disease can be complicated by circulatory collapse, coma, renal failure, and electrolyte imbalances. Although often absent in the mildly ill patient, in whom antibiotic therapy alone usually suffices, these complications should be anticipated in the seriously ill patient, especially if first seen late in the disease course.93 The most important factor contributing to the persistent case-fatality rate of 5% is delayed administration of specific antibiotic therapy. Without appropriate treatment, the fatality
PATIENT
DOXYCYCLINE† (ORAL/INTRAVENOUS)
CHLORAMPHENICOL‡ (ORAL/INTRAVENOUS)
Adult
100 mg bid
Child (10% bands on a peripheral smear) represents the release of immature cells from the bone marrow and is considered to be a sign of infection and inflammation. Like the white blood cell count, it is an imperfect indicator of infection. The absence of leukocytosis or bandemia does not preclude the possibility of severe sepsis, nor does their absence have any proven predictive value in determining mortality risk. The hemoglobin and hematocrit should be obtained to ensure adequate oxygen delivery in shock. Patients should be maintained with a hematocrit greater than 30% and hemoglobin greater than 10 g/dL. Platelets are an acute-phase reactant and may be elevated in the presence of infections. Conversely, a low platelet count has been found to be a significant predictor of bacteremia in patients with shock.81,83,84 Thrombocytopenia, elevated pro-
Chemistry Electrolyte abnormalities should be identified and corrected. Low bicarbonate level suggests acidosis and inadequate perfusion. An elevated anion gap acidosis in the setting of sepsis syndrome commonly represents lactic acidosis or diabetic ketoacidosis, but other causes need to be ruled out. Elevated serum creatinine or decreased glomerular filtration rate is indicative of renal dysfunction or failure, which, if due primarily to sepsis, indicates organ failure and a worse prognosis. Calcium, magnesium, and phosphorus levels should be checked. The presence of an elevated lactate level is associated with inadequate perfusion, shock, and a poorer prognosis.85 One ED-based study showed a progression in mortality rate with increasing venous lactate level: a lactate level of 0 to 2.5 mg/dL was associated with a 5% mortality rate; a lactate level of 2.5 to 4.0 mg/dL, 9% mortality; and greater than 4 mg/dL, 28% mortality.85 An arterial blood gas assessment may be helpful in identifying and classifying acid-base disturbances. Metabolic acidosis suggests inadequate tissue perfusion. Liver function tests can be used to identify liver failure or dysfunction. An elevated bilirubin level may suggest the gallbladder as a cause of sepsis. An elevated amylase and lipase level may represent pancreatitis as the cause of noninfectious SIRS.
Microbiology Obtaining proper blood, sputum, urine, cerebrospinal fluid, and other tissue culture samples is important in guiding therapy. Although not helpful in the initial management, culture samples should be obtained before or soon after the administration of antibiotics in the patient with sepsis syndrome. The initiation of antibiotic therapy should not be delayed while waiting for culture samples to be obtained. One well-designed prospective study suggests the following factors as predictive of a positive blood culture: fever greater than 38.3 °C, the presence of a rapidly (100 msec), or rightward deviation of the terminal 40-msec QRS axis (R wave in aVR > 3 mm or R/SaVR ratio of >0.7) should be considered to have CA poisoning until
1966
PART IV ■ Environment and Toxicology / Section Two • Toxicology
Peripheral Nervous System Effects of
BOX 149-2 Cyclic Antidepressants Anticholinergic Tachycardia Hyperthermia Mydriasis Anhydrosis Red skin Decreased bowel sounds Ileus Urinary retention Distended bladder Alpha1-Blockade Reflex tachycardia Miosis or midrange pupils
Central Nervous System Effects of
BOX 149-3 Cyclic Antidepressants Excitation Agitation Delirium Myoclonic jerks Hyper-reflexia Clonus Seizures Hyperthermia Inhibition Sedation Coma
proven otherwise (Fig. 149-2).18 With QRS prolongation greater than 100 msec in the limb leads, 30% of patients experience seizures; for prolongation greater than 160 msec, 50% develop dysrhythmias. CAs can induce an electrocardiogram (ECG) pattern consistent with the Brugada’s syndrome, a genetic disorder resulting in sodium channel dysfunction. A right bundle branch block and ST segment elevation in the right precordial leads (V1 through V3) are seen.19,20 CA toxicity is not fully excluded with normal ECG findings. Quantitative serum toxicology tests for CAs are not usually available in a timely fashion and do not predict adequately the degree of toxicity. Qualitative laboratory tests may document only exposure. Neither serum toxicology tests nor laboratory tests are useful for clinical decision-making. Diagnosis, treatment, and disposition should be based on a clinical basis and with ECG and cardiac monitoring.9,10
Differential Considerations The differential diagnosis of CA poisoning is extensive and includes intoxications by anticholinergic, psychiatric, and cardiac medications, especially type I antidysrhythmics. Many conditions cause sinus tachycardia and hypotension with a wide pulse pressure; however, QRS widening, seizures, or coma suggests CA poisoning.
Management Treatment begins with assessment of airway and breathing. Endotracheal intubation should be performed if the patient is
exhibiting a markedly decreased level of consciousness or if the level of consciousness is rapidly deteriorating. Respiratory depression, with attendant hypoxia and hypercarbia, significantly increases morbidity and mortality from CA poisoning. Patients with overdoses severe enough to require admission to the intensive care unit (ICU) have a high incidence of airway and breathing complications, with aspiration pneumonitis reported in 13 to 18% of patients.21 The initial benign appearance of the patient’s presentation may be deceptive; rapid deterioration with cardiac dysrhythmias, generalized seizures, and death can occur despite appropriate management. Ventilatory support to avoid respiratory acidosis is crucial because acidosis inhibits conductance through fast sodium channels. Any patient with a significant CA overdose should have continuous pulse oximetry. Arterial blood gas determination or capnography may be helpful when ventilatory function is questionable despite normal oxygen saturations on pulse oximetry. Patients with low minute volumes, hypoxia, and acidosis can appear clinically to be ventilating adequately. Gastric lavage and administration of activated charcoal within 60 minutes from time of significant ingestion is theoretically appealing, and despite the general lack of evidence for efficacy, the toxicity of these drugs is high and any decrease in the absorption may be helpful. If the patient is thought to have ingested sufficient CA to prompt gastric lavage, this should be preceded by intubation for airway protection, since the risk of deterioration is high. The use of physostigmine is not recommended. Seizures, cardiac arrest, and death have occurred when physostigmine has been used in CA overdose.22 Hypertension is usually mild and transient and requires no treatment (Table 149-2). Treatment of hypotension begins with isotonic crystalloids. If the QRS is greater than 100 msec, and the patient is symptomatic, with hypotension or a dysrhythmia, or if the patient is acidemic, intravenous (IV) sodium bicarbonate (NaHCO3) also should be administered.23 NaHCO3 produces an alkaline pH and provides a sodium load and hypertonicity that increase sodium conductance through myocardial fast sodium channels.11 Hyperventilation and hypertonic sodium chloride also enhance sodium conduction.24 Hypertonic sodium chloride has been used for treating hypotension with QRS widening caused by CA-induced cardiotoxicity unresponsive to standard therapies.25 Serum alkalinization is clinically effective in decreasing CA-induced intraventricular conduction delays. The major effect of increasing pH seems to be increased sodium conductance through myocardial sodium channels rather than the increase in plasma protein binding.11 NaHCO3 is administered by IV boluses of 1 to 2 mEq/kg until the QRS narrows or until serum pH increases to 7.50 to 7.55. After obtaining the desired endpoint with IV NaHCO3 boluses, a continuous infusion can be maintained by adding three ampules of 8.4% NaHCO3 (50 mEq/ampule, 100 mOsm/ampule) to 1 L of 5% dextrose in water. The initial IV infusion is started at a usual maintenance rate for IV fluids, based on the patient’s weight. This initial NaHCO3 infusion rate often must be maintained for at least 4 to 6 hours before the rate can be decreased. For a hypertonic continuous IV solution, four ampules of 8.4% NaHCO3 can be added to 1 L of 5% dextrose in water. This infusion is occasionally necessary for refractory hypotension with a prolonged QRS or refractory ventricular dysrhythmias. Excess NaHCO3 and saline can worsen heart failure and should be avoided. Excessive alkalemia from combined use of hyperventilation and NaHCO3 can result in complications, including death.26 Repeat boluses and continuous IV infusion should be guided by serial measurements of arterial pH and QRS duration. When symptoms are refractory, hypertonic
1967 aVR
V1
V4
II
aVL
V2
V5
III
aVF
V3
V6
I
aVR
V1
V4
II
aVL
V2
V5
III
aVF
V3
V6
I
aVR
V1
V4
II
aVL
V2
V5
III
aVF
V3
V6
A
B
C Figure 149-2. Electrocardiographic (ECG) changes in cyclic antidepressant poisoning. A, ECG of a 24-year-old woman who ingested 2.5 g of desipramine
shows ventricular bigeminy, right axis deviation, wide QRS complex, long QTc interval, and right deviation of terminal 40-msec QRS vector in limb leads, with prominent R wave in aVR. B, ECG of same patient after receiving bolus and continuous intravenous infusion of sodium bicarbonate shows sinus tachycardia, 90-degree axis, slight prolongation of QRS and QTc intervals, and persistent right axis deviation of terminal 40 msec in limb leads. C, ECG of same patient 25 hours after ingestion and 4 hours after termination of sodium bicarbonate infusion shows persistent but improved sinus tachycardia, normalization of QRS axis and width, normalization of QTc interval, and resolution of prominent R wave in aVR.
sodium chloride boluses can be used to treat hypotension and a wide QRS interval with ventricular ectopy.25 If hypotension does not resolve, norepinephrine or dopamine is recommended.8 High-dose dopamine (20–30 µg/kg/min) and norepinephrine (0.1–1 µg/kg/min) may be necessary for the direct alpha1-agonist effect.27 For inotropic support alone, dobutamine is controversial.11 Vasopressin has been used to treat
refractory hypotension following amitriptyline poisoning; and intralipid infusion has been used in an animal model to successfully treat hypotension due the fat-soluble drug clomipramine.28,29 Sinus tachycardia is usually well tolerated and does not require specific therapy. Beta-receptor antagonists and physostigmine should not be used.23 Wide-complex tachycardia is
Chapter 149 / Antidepressants
I
1968
PART IV ■ Environment and Toxicology / Section Two • Toxicology
Table 149-2 Cardiovascular Complications of Cyclic Antidepressant Toxicity and Treatments MECHANISM: CAUSE COMPLICATION
CARDIAC
PERIPHERAL VASCULAR
TREATMENT
Hypertension (early and transient)
Positive chronotropism: Anticholinergic vagolytic effects
Initial vasoconstriction: Increased circulating catecholamines caused by reuptake inhibition
Not indicated
Vasodilation: Alpha1adrenoreceptor blockade
IV isotonic crystalloid IV NaHCO3 if QRS >100 msec Norepinephrine or dopamine Not indicated
Hypotension Sinus tachycardia
Ventricular tachycardia (monomorphic) Ventricular tachycardia (polymorphic) (torsades de pointes) Bradydysrhythmias (late and uncommon)
QRS prolongation, QT prolongation
Positive inotropism: Increased circulating catecholamines caused by reuptake inhibition Negative inotropism: Fast sodium channel inhibition with impairment of excitation-contraction coupling Positive chronotropism: Anticholinergic vagolytic effects Positive chronotropism: Increased circulating catecholamines caused by reuptake inhibition Negative dromotropism: Fast sodium channel inhibition with QRS prolongation Negative dromotropism: Fast sodium channel inhibition with QRS prolongation and resultant QT prolongation, and potassium efflux inhibition Negative dromotropism: Fast sodium channel inhibition Negative chronotropism: Impaired automaticity from threshold voltage elevation and slowing of phase 4 depolarization Negative dromotropism: Fast sodium channel inhibition (potassium efflux inhibition), PR prolongation, rightward terminal 40-msec QRS axis deviation
Reflex tachycardia: Alpha1adrenoreceptor blockade
IV NaHCO3 Synchronized cardioversion Overdrive pacing Magnesium sulfate for torsades de pointes
ACLS algorithm for bradycardia
IV NaHCO3 for symptomatic QRS prolongation
ACLS, advanced cardiac life support; IV NaHCO3, intravenous sodium bicarbonate.
not always ventricular tachycardia and can represent sinus tachycardia with aberrant conduction; however, treatment in either case is IV NaHCO3.11 Lidocaine is not consistently effective.30 Phenytoin has been shown to increase the frequency and duration of episodes of ventricular tachycardia and is not recommended as an antidysrhythmic agent.31 Type IA antidysrhythmics (quinidine, disopyramide, procainamide) and type IC antidysrhythmics (flecainide, moricizine, propafenone) are contraindicated because they also inhibit fast sodium channels. A transvenous pacemaker and overdrive pacing can be used for CA-associated polymorphic ventricular tachycardia (torsades de pointes) that is not responsive to magnesium. CA poisoning can result in increased threshold requirements for ventricular pacing and decreased electric cardioversion and defibrillation effectiveness. Bradydysrhythmias are rare and late in CA overdose. QT prolongation, PR prolongation, and rightward terminal 40-msec QRS axis deviation do not mandate specific therapy.11,23 Treatment with NaHCO3, hypertonic sodium chloride, and hyperventilation does not completely correct QT prolongation, which involves not only sodium channel blockade, but also protracted repolarization from potassium efflux blockade.
Treatment of neurologic complications of CA poisoning includes early intubation with mechanical ventilation for coma (Table 149-3). Benzodiazepines should be used for agitation because they do not have the anticholinergic or cardiac conduction effects of some antipsychotic medications. Status epilepticus or prolonged seizures account for 20 to 30% of the seizures caused by CAs.14,32 These seizures usually respond to IV lorazepam or diazepam.14 Seizures refractory to other benzodiazepines have terminated with IV midazolam boluses of 2.5 to 10 mg and continuous IV infusions.33 If benzodiazepines fail to terminate prolonged or repetitive seizures, phenobarbital may be administered in a loading dose of 20 mg/kg, given at a rate of up to 50 mg/min in adults or up to 1 mg/kg/min in children. Propofol is also used to treat refractory seizures successfully.34 A loading dose of 2.5 mg/kg is followed by continuous infusion of 25 to 200 µg/kg/min.34 Phenytoin may cause more and longer episodes of ventricular tachycardia and is unlikely to be effective for seizures.31 If maximal doses of benzodiazepines, phenobarbital, or propofol are ineffective, neuromuscular blockade and general anesthesia with continuous electroencephalogram monitoring are recommended to prevent rhabdomyolysis and hyperthermia caused by excessive muscle activity.
1969
COMPLICATION
TREATMENT
Coma
Endotracheal intubation Mechanical ventilation Supportive care Lorazepam or diazepam Phenobarbital, continuous intravenous midazolam infusion, or propofol Cessation of seizures with benzodiazepines and phenobarbital Neuromuscular blockade with vecuronium during phenobarbital loading General anesthesia with continuous EEG monitoring Evaporative cooling Ice water bath
Seizures Hyperthermia
EEG, electroencephalogram.
Flumazenil is contraindicated, even if benzodiazepines are known coingestants.35,36 Flumazenil counteracts the anticonvulsant activity of the co-ingested benzodiazepines. Seizures, ventricular tachycardia, and ventricular fibrillation can occur when flumazenil is used in mixed benzodiazepine-CA overdoses.36 Life-threatening hyperthermia (rectal temperature > 40° C) is best treated with control of seizures and neuromuscular blockade. A nondepolarizing neuromuscular blocker (e.g., rocuronium) is recommended if rhabdomyolysis and hyperkalemia with ECG changes are present. Evaporative cooling should be used until core temperature reaches 38.5° C. Forced diuresis, hemodialysis, and hemoperfusion are not helpful because CAs have large volumes of distribution, are highly bound to plasma proteins, and are minimally eliminated by the kidneys.
Disposition Patients with known or possible CA overdoses require observation with continuous cardiac monitoring and pulse oximetry. After 6 hours of observation, patients may be discharged for psychiatric evaluation if they do not develop (1) ventilatory insufficiency, (2) desaturation on pulse oximetry, (3) QRS greater than 100 msec, (4) sinus tachycardia greater than 120 beats/min, (5) dysrhythmias, (6) hypotension, (7) decreased level of consciousness, (8) seizures, or (9) abnormal or inactive bowel sounds. Patients who exhibit any of these findings should be admitted to an ICU (see Fig. 149-1).9,10
■ SELECTIVE SEROTONIN REUPTAKE INHIBITORS SSRIs approved in the United States for treatment of depression are fluoxetine (Prozac), paroxetine (Paxil), sertraline (Zoloft), citalopram (Celexa), and escitalopram (Lexapro). Fluvoxamine (Luvox) is approved in the United States for treating obsessive-compulsive disorder. Combination medications also exist such as olanzapine with fluoxetine (Symyax).
Principles of Disease SSRIs are well absorbed from the gastrointestinal tract and reach peak plasma concentrations 3 to 8 hours after achieving
Clinical Features Overdose The signs, symptoms, and morbidity and mortality rates for SSRI poisoning are much less than for CA poisoning. The organ systems most affected by excessive serotonin are the gastrointestinal tract, cardiovascular system, and CNS (Table 149-4). Overdose can cause sedation, agitation, tremor, hyper-reflexia, tachycardia, bradycardia, nausea, vomiting, abdominal pain, facial flushing, and dizziness. Syndrome of inappropriate secretion of antidiuretic hormone and priapism can occur rarely but are unique to specific substances. Severe overdose can cause seizures and cardiac toxicity, and the frequency of severe symptoms increases with coingestants.37 Fluoxetine. About 45% of adults and 90% of children who overdose on fluoxetine alone develop no symptoms.3,38 Common symptoms of fluoxetine overdose include agitation, anxiety, restlessness, confusion, and hypomania.39 Other milder symptoms are tachycardia, drowsiness, tremor, nausea, and vomiting. In children, 5% develop diarrhea, and 5% develop sleepiness.3 QTc prolongation, torsades de pointes, QRS widening, ventricular bigeminy, ventricular tachycardia, and seizures, with delayed onset 10 hours after ingestion, are rare with fluoxetine overdose.39-41 Fluvoxamine. Tachycardia, bradycardia, hypotension, ECG abnormalities, seizures, liver function abnormalities, coma, and death have been reported with fluvoxamine.39 Status epilepticus and refractory hypotension also have been described with fluvoxamine toxicity.42,43
Chapter 149 / Antidepressants
of Neurologic Complications of Table 149-3 Treatment Antidepressant Poisoning
therapeutic doses. Controlled-release paroxetine reaches peak serum concentration 6 to 10 hours after ingestion. Because SSRIs have no significant anticholinergic effects, they are less likely than CAs to have delayed absorption. SSRIs are highly lipophilic, are extensively bound to plasma proteins, and have large volumes of distribution (12–97 L/kg).1 SSRIs are metabolized predominantly by the liver. The elimination half-life is about 15 hours for fluvoxamine, about 1 day for paroxetine and sertraline, 27 hours for escitalopram, 35 hours for citalopram, and 1 to 4 days for fluoxetine. Fluoxetine, sertraline, citalopram, and escitalopram have active metabolites that increase the duration of their effect. Fluoxetine is demethylated to its clinically active metabolite, norfluoxetine, with a half-life of 4 to 9 days.1 The long half-lives of the SSRIs and their active metabolites explain the long interval necessary to avoid the life-threatening serotonin syndrome when stopping SSRI treatment and beginning MAOI treatment. MAOI therapy should not be initiated until at least 5 weeks after stopping fluoxetine and for at least 2 weeks after stopping the other SSRIs. Likewise, none of the SSRIs should be used for at least 2 weeks after stopping MAOI treatment. Minimal amounts of SSRIs are eliminated in the urine as unchanged drugs. Enterochromaffin cells of the gastrointestinal tract contain 90% of the total serotonin in adults. Most remaining serotonin is in platelets and the CNS. In the CNS, serotonin is involved with mood, depression, anxiety, obsession, compulsion, perception of pain, migraine headaches, sleep, circadian rhythms, temperature regulation, and regulation of blood pressure. Serotonin’s roles in the periphery involve regulating gastrointestinal motility and facilitating hemostasis by promoting vasoconstriction and platelet aggregation. The antidepressant effects of SSRIs are thought to be due to the blockade of presynaptic reuptake of serotonin at the 5-hydroxytryptamine type 1 (5-HT1) autoreceptors, resulting in an increased concentration of intrasynaptic serotonin.1
1970
PART IV ■ Environment and Toxicology / Section Two • Toxicology
Table 149-4 Clinical Manifestations of Acute Fluoxetine Overdoses in Adults NEUROMUSCULAR SYSTEM
CARDIOVASCULAR SYSTEM
GASTROINTESTINAL SYSTEM
MANIFESTATIONS
%
MANIFESTATIONS
%
MANIFESTATIONS
%
Drowsiness Tremor Euphoria Headache
21 8 2 2
Sinus tachycardia Hypertension Trigeminy Junctional rhythm
22 6 2 2
Nausea Emesis Abdominal pain
6 6 2
From Borys DJ, et al: Acute fluoxetine overdose: A report of 264 cases. Am J Emerg Med 10:115, 1992.
Citalopram. Gastrointestinal upset; mild CNS changes, such as dizziness and somnolence; and mild autonomic symptoms, such as tachycardia, are observed with ingestions of citalopram less than 600 mg.2,8,44 QTc prolongation has been observed after an ingestion of 400 mg.44 At higher doses, QTc prolongation, QRS widening, ventricular fibrillation, seizures, and death have been reported.2,3,44 Seizures have been reported to start 14 hours after a citalopram and fluoxetine overdose.45 Seizures and QTc prolongation followed by torsades de pointes with cardiac arrest have been observed 13 hours following overdose.46,47 A prolonged observation period beyond the routine 6-hour observation period is suggested following citalopram overdose. Escitalopram. Limited information is available about this enantiomer of citalopram. Ingestion of 600 mg is associated with only minor effects, including drowsiness, agitation, and tachycardia. No seizures were reported.48 Clinical toxicity is similar for overdose of citalopram when compared with escitalopram, with the most frequent symptoms being drowsiness, lethargy, tachycardia, and hypertension.49 Sertraline and Paroxetine. Mild symptoms after sertraline or paroxetine overdose are similar to symptoms with other SSRIs. Seizures and death have been reported with large doses of sertraline.50 In general, deaths attributed to SSRI overdose usually are associated with polydrug poisonings involving significant co-ingestants.8,50
Serotonin Syndrome Serotonin syndrome is a constellation of signs and symptoms manifesting as autonomic, neuromuscular, and mental status changes (Box 149-4).1,51,52 Various drugs increase serotonin concentrations and serotoninergic neurotransmission (Table 149-5). Serotonin syndrome can occur when (1) a serotoninergic agent is added to an established medication regimen, (2) the dose of a serotoninergic agent is increased, or (3) high but usually therapeutic doses of a serotoninergic agent are taken.51 Sternbach suggested diagnostic criteria for the serotonin syndrome (Box 149-5); these also could include signs and symptoms of severe disease, such as muscle rigidity, clonus, hypertension, and tachycardia.51,53 The Hunter criteria are diagnostic decision rules for diagnosing serotonin toxicity (Box 149-6).54
Diagnostic Strategies A history of SSRI dosage increase, SSRI overdose, or SSRI use with an incompatible drug is the most helpful diagnostic indicator. Urine and blood toxicology tests for SSRIs are not readily available nor clinically useful.
BOX 149-4 Clinical Manifestations of Serotonin Toxicity Neuromuscular Agitation Akathisia Anxiety Ataxia Bilateral Babinski’s signs Clonus Coma Confusion Delirium Diaphoresis Dysarthria Euphoria Headache Hyper-reflexia Hyperthermia Hypomania Insomnia Mania Mydriasis Myoclonus Nystagmus Piloerection Rhabdomyolysis Rigidity Seizures Shivering Tremor Cardiovascular Cutaneous flushing Hypertension Hypotension Sinus tachycardia Ventricular tachycardia (rare) Gastrointestinal Abdominal cramps Diarrhea Salivation
Differential Considerations The differential diagnosis of SSRI poisoning includes intoxications and pathologic conditions that cause sinus tachycardia, hypertension or hypotension, gastrointestinal upset, and seizures. In addition to these signs and symptoms secondary to SSRI overdose, SSRIs can produce serotonin syndrome (see
1971
Table 149-5 Drugs Associated with Serotonin Toxicity INCREASED SEROTONIN RELEASE
L-Tryptophan*
Amphetamines Cocaine Codeine Dextromethorphan Fenfluramine Levodopa Pentazocine Reserpine
Decreased Serotonin Degradation (Monoamine Oxidase Inhibitors) Amphetamine metabolites* Clorgyline* Iproniazid* Isocarboxazid Moclobemide* Pargyline* Phenelzine Selegiline Tranylcypromine
Decreased Serotonin Reuptake Amphetamines Carbamazepine Citalopram Cocaine Cyclic antidepressants Dextromethorphan Fluoxetine Fluvoxamine Meperidine Methadone Paroxetine Sertraline Trazodone Venlafaxine
Direct or Indirect Serotonin Receptor Agonists Buspirone Electroconvulsive therapy Lithium LSD and other indoles* Mescaline and other phenylalkylamines* Sumatriptan *Not marketed in the United States. LSD, d-lysergic acid diethylamide.
Boxes 149-5 and 149-6). CNS and other infections, intoxications (e.g., methamphetamine, cocaine, other sympathomimetics), metabolic derangements (e.g., thyroid storm), sedative-hypnotic withdrawal, and strychnine poisoning should be considered in the differential diagnosis of serotonin syndrome. Clinically, serotonin syndrome can be difficult to distinguish from neuroleptic malignant syndrome. A history of precipitating medication use (SSRI vs. neuroleptic), more rapid onset of symptoms, and presence of clonus helps differentiate serotonin syndrome from neuroleptic malignant syndrome (Table 149-6).
Management Activated charcoal may be considered, but SSRI toxicity is mild and rare so the overall potential risk-benefit consideration would argue against this. Hemodialysis and hemoperfusion are not indicated because SSRIs have large volumes of distribution and are highly bound to plasma proteins.1 Forced diuresis is not helpful because minimal amounts of SSRIs and their active metabolites are eliminated in the urine. Cardiovascular complications of serotonin toxicity include hypertension, sinus tachycardia, hypotension, and, rarely, ventricular dysrhythmias (Table 149-7).51 Hypertension and
Adding a serotoninergic agent to a patient’s established medication regimen or increasing the dose of a patient’s serotoninergic agent. At least three of the following signs and symptoms: Agitation Ataxia Diaphoresis Diarrhea Hyper-reflexia Hyperthermia Mental status changes (e.g., confusion, hypomania) Myoclonus Shivering Tremor A neuroleptic has not been started or increased in dosage before the onset of the above signs and symptoms. Other etiologies, such as infections, intoxications, metabolic derangements, and withdrawal, have been ruled out. From Sternbach H: The serotonin syndrome. Am J Psychiatry 148:705, 1991.
Hunter Serotonin Toxicity Criteria:
BOX 149-6 Decision Rules
In the presence of a serotonergic agent: 1. If (spontaneous clonus = yes), then serotonin toxicity = yes 2. Else If (inducible clonus = yes) and [(agitation = yes) or (diaphoresis = yes)], then serotonin toxicity = yes 3. Else If (ocular clonus = yes) and [(agitation = yes) or (diaphoresis = yes)], then serotonin toxicity = yes 4. Else If (tremor = yes) and (hyper-reflexia = yes), then serotonin toxicity = yes 5. Else If (hypertonic = yes) and (temperature > 38° C) and [(ocular clonus = yes) or (inducible clonus = yes)], then serotonin toxicity = yes 6. Else serotonin toxicity = no From Dunkley EJC, et al: The Hunter serotonin toxicity criteria: Simple and accurate diagnostic decision rules serotonin toxicity. QJM 96:635, 2003.
tachycardia are usually mild and transient and require no treatment.3 Hypotension is treated with isotonic crystalloids, and vasopressors are rarely necessary. Ventricular dysrhythmias should be treated with standard antidysrhythmic agents (e.g., lidocaine) (see Table 149-7).3 QRS prolongation with fluoxetine toxicity has responded to IV NaHCO3 treatment.40 Neurologic complications of SSRI overdose and serotonin syndrome are treated with benzodiazepines. Cyproheptadine, methysergide, chlorpromazine, and propranolol have been proposed as therapies, but have inconsistent effects. They have been used in isolated case reports of serotonin syndrome; however, none of these should be considered a proven therapy or diagnostic modality.51 Cyproheptadine is available as a liquid that can be given through a nasogastric tube, beginning with an adult dose of 4 to 8 mg followed by 4-mg doses every 1 to 4 hours as needed to a maximum of 32 mg/day. For children, 0.25 mg/kg/day is given in divided
Chapter 149 / Antidepressants
INCREASED SEROTONIN SYNTHESIS
Sternbach’s Diagnostic Criteria for
BOX 149-5 Serotonin Syndrome
1972
PART IV ■ Environment and Toxicology / Section Two • Toxicology
Table 149-6 Comparison of Serotonin Toxicity and Neuroleptic Malignant Syndrome FACTOR MALIGNANT SYNDROME
SEROTONIN TOXICITY
NEUROLEPTIC
Dopamine antagonists Serotonin agonists Onset of symptoms Hyperthermia, altered level of consciousness, autonomic dysfunction, muscle rigidity Leukocytosis, metabolic acidosis Elevated creatine kinase Hyper-reflexia, myoclonus Treatment Resolution of symptoms
No Yes Within minutes to hours Present in varying degrees
Yes No Usually over days to weeks, may occur immediately Almost universal for each sign
Unusual Present in varying degrees Present in varying degrees Benzodiazepines, cyproheptadine Resolution of symptoms begins but not complete in 460 msec) and all patients with significant ingestions of thioridazine or mesoridazine should have at least 12 hours of cardiac monitoring.76 Patients with less severe signs of toxicity should be observed in the emergency department for a minimum of 4 hours from the time of ingestion, with hospitalization for persistent or worsening signs and symptoms. Criteria for hospital discharge include return to normal mental status and resolution of any vital sign, metabolic, and ECG abnormalities. Psychiatric consultation may be necessary to assess the risk of suicide.
PART IV ■ Environment and Toxicology / Section Two • Toxicology
■ MANAGEMENT Acute Overdose
KEY CONCEPTS
QT Prolongation and Torsades de Pointes Correction of hypokalemia, hypomagnesemia, and hypocalcemia shortens the QT interval. Treatment of TDP includes IV magnesium sulfate, overdrive pacing, and possibly isoproterenol. Administration of antiarrhythmic drugs that prolong the QT interval should be avoided.
Neuroleptic Malignant Syndrome Treatment of NMS consists of supportive care and discontinuation of the offending medication. Agitation, psychomotor hyperactivity, and muscular rigidity should be treated with liberal doses of IV benzodiazepines. Lorazepam can be administered IV 1 or 2 mg every 3 minutes until muscle rigidity improves, to a maximum dose of 10 mg. In refractory cases or cases at risk of aspiration, rapid sequence induction, endotracheal intubation, and neuromuscular blockade with a non depolarizing agent (e.g., rocuronium and vecuronium) are required. Hyperthermia should be managed with IV fluids and active external cooling with mist and fans. If rhabdomyolysis is present, IV hydration and urinary alkalization are used to prevent renal damage.
■
Extrapyramidal movement disorders are a common complication of antipsychotic medications and are treated with benztropine, diphenhydramine, and benzodiazepines. ■ The most common finding in antipsychotic overdose is central nervous system depression. Treatment centers on supportive care, airway control, and cardiac monitoring. ■ QT prolongation and torsades de pointes are potential complications from overdose of many antipsychotic medications, and they can occur with therapeutic doses of some agents. ■ The neuroleptic malignant syndrome is characterized by altered mental status, hyperthermia, muscular rigidity, and autonomic instability. Supportive care includes airway management, benzodiazepines, neuromuscular blockade, and active cooling. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 160
Opioids
Christina Hantsch Bardsley
■ PERSPECTIVE Opioid is a term that applies to all natural, synthetic, and semisynthetic agents with morphine-like actions. It is more inclusive than the term opiate, which refers only to natural agents. Both terms are derived from opium, the Greek word for juice in reference to poppy juice. Poppy juice contains more than 20 distinct natural alkaloids with morphine-like activity. The term narcotic refers to any agent that induces sleep and is nonspecific. Although the term narcotic persists, primarily in legal contexts, opioid is more precise and is the correct medical term for the family of agents that act on opiate receptors in the body. Finally, the term endorphin applies to any of the three endogenous opioid families: enkephalins, beta-endorphins, and dynorphins.1 Pharmacologic actions of opioids involve the gastrointestinal system, genitourinary system, cardiovascular system, pulmonary system, and central nervous system (CNS) and cause characteristic clinical effects. Sedation and analgesia are the most common therapeutic goals of opioid medications, which are available alone or in combination with other agents (e.g., acetaminophen and salicylates) for these purposes. Additional therapeutic goals of opioids and combination preparations include antitussive and antidiarrheal effects. Misuse of pharmaceutical opioid preparations and use of illicit opioids are significant problems in the United States. According to the 2006 National Survey on Drug Use and Health, 506,000 Americans age 12 years or older have used heroin, including 338,000 current heroin users.2 Injection is the most common route, followed by inhalation, smoking, and ingestion. Since the early 1990s, there has been a trend toward inhalation rather than injection. Prescription pain medications are used for nonmedical means by 5.2 million users in comparison to only 2.4 million cocaine users.2 Opioids are second to sedatives/hypnotics/antipsychotics for substances involved in reported fatal exposures.3 A multistate epidemic of nonpharmaceutical fentanyl-related overdoses resulted in more than 1000 deaths from April 2006 to March 2007.4
■ PRINCIPLES OF DISEASE Anatomy and Physiology Although opioids have been used for more than 5000 years, receptors and endogenous opioids have been recognized only
since the 1970s.1,5 Opioids have an incompletely understood physiologic role. There are three established opioid receptors, known as mu or OP3, kappa or OP2, and delta or OP1.5 Opioid receptors are distributed throughout the CNS, concentrated in pain pathways and in areas associated with the perception of pain (e.g., periaqueductal gray matter, locus ceruleus, limbic system, and nucleus raphe magnus). Systemically, opioid receptors are localized in sensory nerve endings, on mast cells, and in the gastrointestinal tract.5 Genetic variations in opioid receptors account for some of the interindividual differences in response to both endogenous and exogenous opioids. Several polymorphisms of the mu receptor have been identified, including A118G polymorphism, which is present in more than 34% of individuals.6 The role of the opioid receptors in pain perception and the analgesic effects of exogenous opioids are discussed in detail in Chapters 186 and 187.
Pathophysiology and Pharmacology Toxicity Opioids include therapeutic agents and illicit substances. Toxicity occurs as a result of intentional overdose, intentional abuse, or adverse effect of therapeutic use. Although different opioids have receptor preferences in therapeutic or low doses, this specificity is lost at higher doses. Opioids are well absorbed after gastrointestinal (oral and rectal) or parenteral administration but also through nasal, buccal, pulmonary, and transdermal routes, depending on the lipid solubility of the specific opioid.1 Heroin is usually abused through intravenous and subcutaneous routes, but it is also absorbed after nasal administration because it is lipid soluble.7 In general, opioid toxicity is less pronounced but more prolonged with oral ingestion than with parenteral administration.1 Absorption of opioids after oral administration occurs in the small intestine. With therapeutic doses, absorption is complete within 1 or 2 hours. Absorption and clinical effects of toxicity may be prolonged after overdose, however, because gastric emptying is delayed. Most opioids have a large volume of distribution. Clinical effects depend on lipid solubility, which affects the ease with which opioids and their metabolites cross the blood-brain barrier. All opioids undergo hepatic metabolism and renal elimination, and variations in hepatic and renal function are 2047
PART IV ■ Environment and Toxicology / Section Two • Toxicology
2048
important because metabolite activity may contribute to clinical effects and toxicity.1 The half-lives of meperidine and normeperidine are prolonged with cirrhosis, and the elimination of normeperidine is decreased with renal insufficiency. These conditions increase the likelihood of normeperidinerelated seizures, as do multiple or large doses.8 The pharmacokinetics of the specific agent determine the clinical course of opioid toxicity. Heroin peaks in the serum within 1 minute of intravenous injection, 3 to 5 minutes of intranasal administration, and 10 minutes of subcutaneous injection. Heroin’s lipophilic nature allows for rapid transport across the blood-brain barrier into the CNS. In the CNS and blood, heroin is rapidly hydrolyzed to 6-monoacetylmorphine and then morphine (less lipid soluble). In the liver, morphine undergoes conjugation with glucuronic acid to form more water-soluble compounds that are excreted by the kidneys.1
arthria, ataxia, tremor, and other neurologic abnormalities.12 This syndrome is incompletely understood but is thought to be related to a combination of mitochondrial injury and hypoxia.13 One report also associates heroin-induced movement disturbance with basal ganglia lesions.14 Serotonin syndrome is a clinical triad of mental status changes, autonomic instability, and neuromuscular changes (see Chapter 149) and may be fatal. Most cases involve an interaction between a serotoninergic agent and a second agent, usually a selective serotonin reuptake inhibitor or a monoamine oxidase inhibitor. Meperidine and dextromethorphan have serotoninergic properties and have been associated with serotonin syndrome.8
Respiratory
The toxidrome of opioid toxicity is CNS depression, respiratory depression, and miosis. Other potential findings in opioid toxicity are associated with toxicity from any opioid, but some features are unique to a specific agent or route of exposure.
Opioids decrease respiratory rate and tidal volume in a dosedependent manner by suppressing the sensitivity of the medullary respiratory center to hypercapnia.1 Although it initially remains intact, the hypoxic drive is overridden in severe poisoning or when antagonistic stimuli (e.g., pain) are blocked. Overdose of an agonist-antagonist agent produces less significant respiratory depression, presumably because of mu receptor antagonism (Table 160-1). Central sleep apnea is associated with long-term opioid use and also occurs in those with acute increased opioid use from baseline. Continuous positive airway pressure is usually ineffective for treatment of sleep apnea in these patients.15 Bronchospasm is rare with heroin use in asthmatic and nonasthmatic patients and occurs mostly after inhalational exposure, but other routes are also implicated. The bronchospasm is often severe, prolonged, and refractory to beta-agonist therapy. Patients may require mechanical ventilation for several days. It is unclear whether the heroin, an adulterant, or a combination triggers the bronchospasm and whether the response is histamine mediated or the result of direct irritation.7 Acute lung injury occurs with therapeutic opioid use but is much more common after overdose.16 The capillary leak is likely from hypoxia rather than a direct drug effect.
Neurologic
Ophthalmologic
CNS depression is a well-recognized manifestation of opioid toxicity. Hypoxia from CNS depression and respiratory depression also causes many neurologic complications. Dysphoria and acute psychosis may occur with an agonist-antagonist opioid. Excitatory effects may occur with opioid toxicity. Hypertonicity, myoclonus, and seizures have been reported with overdose of the synthetic opioids meperidine and propoxyphene. Meperidine-related seizures are probably caused by accumulation of normeperidine, especially after multiple or large doses or in patients with hepatic or renal insufficiency. Seizures may also result from hypoxia with overdose of any opioid. Parkinsonian symptoms in intravenous drug abusers have been attributed to 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), an unintended side product produced during synthesis of a meperidine analogue in street laboratories. MPTP injection is associated with accumulation of an MPTP metabolite in CNS cell mitochondria, focal lesions in the substantia nigra, and a syndrome clinically indistinguishable from idiopathic parkinsonism. The syndrome is irreversible in some patients.11 Spongiform leukoencephalopathy has been associated with inhalation of heated heroin, a practice known as “chasing the dragon.” Patients present with psychomotor retardation, dys-
In more than 90% of heroin overdoses, stimulation of mu receptors in the Edinger-Westphal nuclei of the third nerve results in miosis.17 Miosis is not typically seen with meperidine, propoxyphene, or diphenoxylate-atropine (Lomotil) overdose. Toxicity from agonist-antagonists (e.g., pentazocine) or multiple agents may not produce miosis.
Withdrawal Because the half-life of heroin is 30 minutes and the half-life of methadone is 15 to 40 hours, withdrawal symptoms occur 4 to 6 hours after discontinuation of heroin compared with 24 to 48 hours after discontinuation of methadone.9,10 Duration of symptoms is 7 to 10 days and 2 weeks, respectively. The degree of physical dependence that has developed is also important. With chronic opioid exposure, cellular adaptation results in upregulation of cyclic adenosine monophosphate (cAMP). When either the exposure is discontinued or an antagonist is administered, the result is a temporary elevation of cAMP levels and increased sympathetic activity above a normal baseline.
■ CLINICAL FEATURES Toxicity
Otolaryngolic Rapidly progressive sensorineural hearing loss has been reported with the use of hydrocodone. Genetic polymorphism producing altered metabolism and/or comorbidities have been suggested as causative factors.18
Cardiovascular Opioids cause mild hypotension and relative bradycardia. Hypotension seems to be from histamine release and can be blocked by antihistamines (H1 antagonists).19 The hypotension is typically orthostatic and resolves with supine positioning. Propoxyphene and its metabolite, norpropoxyphene, may cause sodium channel blockade similar to that of type IA antidysrhythmic agents to produce widening of the QRS complex.20
2049
Table 160-1 Select Opioid Doses and Associated Respiratory Depression
Codeine Dextromethorphan Dihydrocodeine Diphenoxylate Fentanyl Heroin Hydrocodone Hydromorphone Levorphanol Meperidine Methadone Morphine Oxycodone Oxymorphone Paregoric Propoxyphene
ORAL DOSE (MG)*
200 700 150 300 — 15 100 6 1 250 20 70 30 6 175 600
INTRAMUSCULAR DOSE (MG)*
120 — 60 — 0.125 3 — 1.5 2 100 10 10 10 1 — —
TIME TO ONSET OF RESPIRATORY DEPRESSION†
Fast Fast Fast Slow (or even more delayed) Very fast Fast Fast Fast Fast Fast Slow (or even more delayed) Fast Fast Fast Fast Fast
*Equivalent to 10 mg of intramuscular morphine. † Varies with the drug and route of administration. In addition, the effects of a dose in any particular patient depend on multiple factors, including age, weight, and comorbid conditions. After intramuscular administration, very fast means 5 to 30 minutes, fast means 15 to 60 minutes, and slow means 1 to 4 hours. After oral administration, these time definitions are approximately doubled.
Gastrointestinal
Metabolic
Nausea and vomiting are common with therapeutic opioid use and also with overdose. Mechanisms include opioid-induced delayed gastric emptying, direct stimulation of the chemoreceptor trigger zone, and vestibular stimulation. Antihistamines and dopamine antagonists (e.g., chlorpromazine) may be effective in treatment. Decreased gastrointestinal motility is a common finding with therapeutic use and overdose of opioids. Severe cases may develop ileus.1 Increased biliary tract pressures and choledochoduodenal sphincter spasm occur with therapeutic dosing of many opioids, including morphine, meperidine, and codeine. Spasm is not always reproducible within the same patient but seems related more to individual susceptibility than to a specific agent. Presenting clinical symptoms mimic biliary colic and may respond to naloxone or glucagon.21
Hypoglycemia occurs after opioid overdose, but the mechanism is unclear. Co-ingestants, especially ethanol, may contribute to this finding. Hypothermia has been reported, but the mechanism is unclear. Hyperthermia should prompt a search for infectious complications, particularly in intravenous drug users, and for co-ingestants (e.g., cocaine) or adulterants (e.g., tripelennamine and scopolamine).23 “Cotton fever” is reported in intravenous drug users who strain suspended drug through cotton balls or cigarette filters to remove particulates. Filters are boiled to extract residual drug when supply is low. Cotton is a known pyrogen and can cause a benign fever in patients who subsequently “shoot the cottons” or inject the extracted residue from the filters.24
Genitourinary Opioids can cause urinary retention from urethral sphincter spasm and decreased detrusor tone. Alpha-adrenergic antagonists may reverse this effect. Glomerulosclerosis and renal amyloidosis are seen in end-stage “heroin nephropathy” of chronic opioid addicts.22
Dermatologic Pruritus, flushing, and urticaria occur after administration of certain opioids that release histamine (e.g., morphine) and do not represent a true allergy. Pruritus and erythema are often localized to the area of injection (e.g., along the vein through which the morphine was administered). Symptoms typically are controlled easily with antihistamines. Although all opioids have the potential to stimulate mast cell degranulation and histamine release, some (e.g., fentanyl) release only clinically negligible amounts of histamine and thus have good hemodynamic stability profiles.
Withdrawal Opioid withdrawal occurs in tolerant individuals when opioid use is discontinued or an antagonist is administered. Increased sympathetic discharge and adrenergic hyperactivity are responsible for the clinical symptoms and signs. In contrast to the typical toxidrome of opioid toxicity (CNS and respiratory depression and miosis), withdrawal is associated with CNS excitation, tachypnea, and mydriasis. Pulse and blood pressure are also increased. Although these can be uncomfortable, they are typically not life-threatening.9,10 Neurologic manifestations are prominent in opioid withdrawal. Restlessness, agitation, and anxiety are virtually universal, and seizures may rarely occur. Cognition and mental status are unaffected. Dysphoria and drug craving may be severe and prolonged. Nausea, vomiting, diarrhea, and abdominal cramps are common in withdrawal. They can be significant and lead to dehydration and electrolyte abnormalities. Other symptoms may also include diffuse myalgias and insomnia with piloerection, yawning, lacrimation, rhinorrhea, and diaphoresis.
Chapter 160 / Opioids
DRUG
2050
PART IV ■ Environment and Toxicology / Section Two • Toxicology
■ DIAGNOSTIC STRATEGIES The diagnosis of opioid intoxication is usually based on history and physical examination. Diagnostic studies rarely assist in evaluation of patients suspected of opioid overdose. Other than hypoglycemia, specific laboratory abnormalities are not seen. When the patient has hypoxemia and pulmonary rales, a chest radiograph should be obtained to evaluate for acute lung injury. In the appropriate circumstances, an abdominal radiograph may identify packets of opioids or other illicit substances in a body packer. A case of QRS widening is thought to be associated with propoxyphene overdose but has not been validated.20 If cardiac monitoring shows a prolonged QRS, a 12-lead electrocardiogram is advisable. With ingestion of an unknown opioid preparation, acetaminophen and salicyl ate levels should be checked because many prescription opioids are available in combination products. Likewise, many illicit opioid users are exposed to additional drugs and contaminants.23,24 Although opiates are detected on most qualitative urine toxicology screens, these are rarely helpful in acute situations. On some assays, several synthetic opioids are also detected because they cross-react or because they are metabolized to opiates, which are then excreted. Other agents, such as fentanyl and its derivatives, are missed on urine screens. Poppy seed ingestion can lead to a positive opiate screen for morphine and codeine; however, detection of 6-monoacetylmorphine, a specific met abolite of heroin, can confirm heroin use.25 As with opioid toxicity, no diagnostic test exists for opioid withdrawal.
■ DIFFERENTIAL CONSIDERATIONS The diagnosis of opioid intoxication is usually obvious, based on history and physical examination, although patients with other intoxications or nontoxicologic conditions may have a similar physical examination. Other drugs that should be considered are clonidine (or a related drug), tramadol, valproic acid, γ-hydroxybutyrate, and sedative-hypnotic agents. The differential diagnosis encompasses all causes of depressed mental status, but the coexistence of miosis and respiratory depression greatly narrows the possibilities. Opioid withdrawal is usually a straightforward diagnosis, and the patient often reveals it as the chief complaint. Simultaneous intoxication with, or withdrawal from, other classes of agents, especially CNS depressants and stimulants, may be seen.
■ MANAGEMENT Attention to the airway, oxygenation, and ventilation is of vital importance in patients with opioid toxicity. If reversal is not achieved with antidote therapy, appropriate interventions include airway protection and ventilatory support. Patients with acute lung injury may require oxygen and positivepressure modalities, such as bilevel positive airway pressure, continuous positive airway pressure, or mechanical ventilation with positive end-expiratory pressure. Circulatory support usually does not require more than a crystalloid infusion. Most opioids have a large volume of distribution and cannot be cleared by dialysis. There are no clinically effective techniques for enhanced elimination of opioids.
Gastrointestinal Decontamination Gastrointestinal decontamination is often unnecessary because the antidote can reverse the effects. Whole-bowel irrigation
can hasten passage of drug packets from body packers or patients with ingestions involving opioid combination products or multidrug ingestions. A single early dose of activated charcoal (1 g/kg in children and 50–100 g in adults) may be beneficial in some patients because gastrointestinal motility may be reduced.26
Antidote Naloxone, a pure opioid antagonist, is the antidote most frequently used to reverse opioid toxicity. Naloxone has a rapid onset of action. For reversal of systemic opioid toxicity it is ineffective after oral administration secondary to the first-pass effect, but intravenous, subcutaneous, intramuscular, inhalational, and endotracheal routes can be used. Naloxone competitively binds opioid receptors and can reverse all the receptor-mediated actions of opioids. Naloxone is indicated for patients with opioid intoxication who have significant CNS or respiratory depression. The initial intravenous dose is 0.4 to 2 mg for adults and children, but 10 mg may be required to obtain a clinical response for synthetic opioids. Naloxone can precipitate acute withdrawal in chronic opioid users. In this population, the dose should be started at 0.2 mg and slowly titrated. The duration of action of naloxone is 1 or 2 hours. Consequently, either repeat doses or a continuous infusion of two thirds of the effective initial dose per hour may be required.1 Naloxone has an excellent safety profile. Acute lung injury, hypertension, and dysrhythmia have been associated with use of naloxone after general anesthesia and in patients with underlying cardiac or pulmonary disease.27 Whether naloxone is the cause of these complications is unproven. Idiosyncratic reactions and sympathetic discharge with precipitation of acute withdrawal have been proposed as explanations. The risk may be greater in those with ongoing hypoventilation prior to naloxone administration. Complete clinical recovery in response to naloxone is strongly suggestive of opioid overdose. Other intoxications, including valproic acid, clonidine, tramadol, captopril, and ethanol intoxication, may improve to lesser degrees with naloxone. Naloxone has been given to patients who have ingested these agents because of a presentation similar to opioid intoxication or suspicion of a mixed exposure that included opioids. The mechanism of these responses to naloxone is not established. Some of these drugs may have activity at opioid receptors.27 Nalmefene, another opioid antagonist, has a long half-life (8– 11 hours) and duration of clinical effect. Intravenous nalmefene has a rapid onset of action in reversing opioid-induced CNS and respiratory depression. Alternate administration routes are oral, subcutaneous, and intramuscular. The initial intravenous dose is 0.5 to 1.5 mg (pediatric dose not established). Higher doses have been used but are associated with increased risk of adverse effects.28,29 When a clinical response has been achieved with nalmefene, repeat doses or continuous infusions are generally not required; however, the duration of withdrawal symptoms may be longer with nalmefene. Naloxone remains the preferred antidote in patients at risk for withdrawal or other adverse effects and in patients with anticipated short duration of opioid toxicity. Seizures associated with opioid toxicity resolve with correction of hypoxia or administration of benzodiazepines.
Withdrawal Opioid withdrawal is not life-threatening, but the potentially serious manifestations mandate attention to supportive and symptomatic care. When withdrawal is produced by adminis-
■ DISPOSITION Toxicity Patients with opioid toxicity are usually treated successfully in the emergency department, sometimes in conjunction with the emergency department observation unit. Patients who receive naloxone should be observed for 2 hours to assess the extent of re-sedation. Asymptomatic patients can be observed in the emergency department until 4 hours after an ingestion, at which time they can be discharged with appropriate psychiatric evaluation or drug abuse counseling. Patients who have ingested an opioid with a longer half-life, active metabolite, or
2051
modified release preparation as well as patients with multiple drug ingestions involving an opioid may require longer observation periods. Patients who have a known or potential overdose of diphenoxylate-atropine (Lomotil) should be observed in a monitored unit even if asymptomatic; this includes small children who may have ingested only a single tablet (2.5 mg of diphenoxylate or 0.025 mg of atropine). The metabolite of diphenoxylate, difenoxin, has a longer half-life and is five times more active than diphenoxylate, and it may cause delayed and prolonged toxicity. Delayed absorption, caused by each component of Lomotil and enterohepatic recirculation, may contribute as well.32 Patients with a known or potential ingestion of packets of illicit opioid drugs should be admitted until the packets are passed. Body packers and body stuffers remain asymptomatic until one or more of the packets leak.
Withdrawal Patients with opioid withdrawal may be managed as outpatients. Clonidine may alleviate some of the symptoms of withdrawal but has a high incidence of postural hypotension. Patients with refractory complications (e.g., vomiting, dehydration, and electrolyte abnormalities) and patients with an uncertain diagnosis may require hospitalization. Some patients may need to be detoxified before entering chemical treatment programs. Substance abuse counseling and establishment of outpatient program referral should be completed before discharge.
KEY CONCEPTS ■
Diagnosis of opioid intoxication is based on history, physical examination, and response to naloxone. ■ The opioid toxidrome includes three prominent findings—CNS depression, respiratory depression, and miosis—but these may not be present in every patient. ■ Early administration of reversal agents and airway management and attention to oxygenation and ventilation are crucial to management of patients with opioid toxicity. ■ The duration of action of many opioids, especially after overdose, is significantly longer than that of naloxone. Patients responsive to naloxone should be observed for recurrence of opioid toxicity after the effect of naloxone has resolved. ■ Opioid withdrawal syndrome does not include altered cognition. Patients with known or suspected opioid withdrawal who also have altered cognition should be evaluated for another etiology of the altered cognition. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 160 / Opioids
tration of naloxone, the symptoms are of short duration, and replacement opioids should be avoided.9,10 Other patients with withdrawal can receive opioid replacement or other medications to alleviate symptoms. In addition, fluid and electrolyte replacement is important for patients with dehydration from gastrointestinal symptoms. Methadone, a long-acting opioid, and l-α-acetyl-methadol (LAAM), an even longer acting opioid, provide opioid replacement to treat or prevent withdrawal in chronic heroin users admitted for other medical conditions. The onset of action of an initial dose of methadone of either 20 mg orally or 10 mg intramuscularly is 30 to 60 minutes. This initial dose typically controls significant symptoms.9 Some symptoms, particularly drug craving, may be ongoing and require a subsequent dose after several hours. Maintenance methadone therapy requires doses every 24 hours or may be tapered daily. LAAM, a three dose per week opioid replacement, was discontinued by the manufacturer due to cardiac adverse events, including QTc prolongation.9 In October 2002, the Food and Drug Administration approved a buprenorphine monotherapy product (Subutex) as well as a buprenorphine/naloxone combination product (Suboxone) for treatment of opioid addiction. As opposed to methadone, these treatments are not limited to use in traditional opioid treatment programs.30 However, neither treatment is indicated in routine emergency department care. Clonidine, a central alpha2-agonist, is used for treatment of opioid withdrawal without opioid replacement.9 Clonidine controls symptoms by suppressing sympathetic hyperactivity. It may also shorten the duration of withdrawal. The initial dose is 0.1 mg orally. Repeat doses may be given every 30 to 60 minutes; relatively large total doses may be required. As with opioid replacement, clonidine therapy is titrated to individual clinical response. Hypotension may limit the treatment but is not common in the setting of opioid withdrawal treatment. Patches for transdermal administration of clonidine are available, but onset of action is delayed 24 hours. Oral doses still must be given initially. Buspirone is an azaspirone compound used to treat ethanol and nicotine addiction. Studies suggest that it may also be an option to treat withdrawal in opioid addicts, but further investigation is needed.31
Chapter 161
Pesticides
Cynthia K. Aaron, James W. Rhee, and Bram A. Dolcourt
■ PERSPECTIVE Pesticides, a generic term used to refer to all pest-killing agents, include numerous chemicals intended for use as insecticides, herbicides, rodenticides, fungicides, and fumigants. Many of these chemicals are general protoplasmic poisons affecting a wide range of organisms, including humans. Although space does not allow a comprehensive discussion of each individual chemical that may produce human toxicity, numerous chemical classes are commonly used as pesticides. These classes have associated characteristic clinical pictures that are important to recognize because patients with acute (and occasionally chronic) exposures to these agents come to the emergency department. In addition, other pesticides with particularly unique mechanisms of toxic effects are described.
■ ORGANOPHOSPHATE AND CARBAMATE INSECTICIDES The organophosphate insecticide triethyl pyrophosphate was first synthesized in 1859 but was not used to replace nicotine as a pesticide until World War II. After World War II, these compounds were used as chemical warfare agents, as organophosphorus and carbamate insecticides, and as medicinal agents. After the negative publicity associated with the organochlorine dichlorodiphenyltrichloroethane (DDT), organophosphate insecticides soon became some of the most common pesticides for home and industrial use. Since the late 1990s, with the increased awareness of terrorism, nerve agents have gained prominence as weapons of mass destruction.1
Principles of Disease Organophosphorus insecticides are highly lipid soluble and are readily absorbed via dermal, gastrointestinal (GI), and respiratory routes.2 This lipid solubility results in the storage of organophosphorus compounds in body fat, making toxic systemic levels possible from gradual or rapid accumulation from repeated low-level exposures. The parent compound and its metabolites are acetylcholinesterase inhibitors, and many parent organophosphorus compounds are less potent than their metabolites (e.g., parathion to paraoxon), which may result in delayed onset of clinical toxicity. Organophosphorus pesticides work by persistently inhibiting the enzyme acetylcholinesterase, the enzymatic deactivator of the ubiquitous neurotransmitter acetylcholine. Because 2052
of the global penetration of organophosphorus compounds, inhibition occurs at tissue sites (true acetylcholinesterase and represented by erythrocyte or red blood cell [RBC] cholinesterase) and in plasma (circulating pseudocholinesterase).3,4 Inhibition of cholinesterase results in the accumulation and subsequent prolonged effect of acetylcholine at a variety of neurotransmitter receptors, including the sympathetic and parasympathetic ganglionic nicotinic sites, postganglionic cholinergic sympathetic and parasympathetic muscarinic sites, skeletal muscle nicotinic sites, and central nervous system sites (Fig. 161-1).2
Clinical Features Signs and Symptoms The accumulation of acetylcholine results in a classic cholinergic syndrome, manifested by hyperactivity of cholinergic responses at the receptor sites indicated previously. The clinical syndrome of muscarinic acetylcholinesterase inhibition is commonly called the SLUDGE syndrome (Table 161-1). This syndrome represents postganglionic acetylcholine-induced hollow end-organ general hypersecretion,2 resulting in clinical findings that include miotic pupils, lacrimation, rhinorrhea, sialorrhea, bronchorrhea, vomiting, diarrhea, and urinary incontinence. Bradycardia is a classic sign of the cholinergic syndrome, but the increased release of norepinephrine from postganglionic sympathetic neurons precipitated by excess cholinergic activity at sympathetic ganglia may result in normal or even tachycardic heart rates (nicotinic effect). Sympathetic hyperactivity can cause diffuse diaphoresis, although this response is mediated by cholinergic receptors at preganglionic (nicotinic) and postganglionic (muscarinic) sites. The most lethal components of acetylcholinesterase inhibition occur in the brain and neuromuscular junction. A combination of sympathetic stimulation, involvement of the N-methyl-d-aspartate receptor, and enhanced acetylcholine concentrations can lead to seizures.5 At the neuromuscular junction, excess acetylcholine causes hyperstimulation of the muscles with secondary paralysis. Because the diaphragm is affected, cholinesterase poisoning leads to respiratory arrest.6 Although the usual clinical picture of acute organophosphorus poisoning is impressive, toxicity from gradual, cumulative exposure may be much more subtle. These patients commonly exhibit vague confusion or other central nervous system complaints; mild visual disturbances; or chronic abdominal cramping, nausea, and diarrhea.7
2053
Autonomic nervous system
ACh
ACh
ACh
NE
ACh
ACh
Hollow end organs
Sympathetic response
Neuromuscular junction
Central nervous system
SLUDGE Hypertension Fasciculations Salivation Tachycardia Respiratory paralysis Seizures Lacrimation Mydriasis Coma Urinary incontinence Diarrhea Figure 161-1. The autonomic nervous system (ANS) comprises the sympathetic and parasympathetic nervous systems. The sympathetic nervous system is also known as the thoracolumbar outflow, where the cell body lies in the spinal cord and the first synapse occurs in the sympathetic ganglia. The neurotransmitter in this first synapse is acetylcholine (ACh) (preganglionic), and the neurotransmitter in the postganglionic neuron with the target organ is norepinephrine (NE). In the parasympathetic nervous system (craniosacral outflow), nerves from the medulla and sacrum use ACh as the neurotransmitter in preganglionic and postganglionic target organs. The ANS is divided further into the muscarinic and nicotinic receptors, where atropine can block the muscarinic receptors but not the nicotinic receptors. The neuromuscular junction uses ACh as the effector neurotransmitter. In the brain, ACh is just one of several active neurotransmitters.
Table 161-1 SLUDGE Symptoms or DUMBELS Salivation Lacrimation Urinary incontinence Defecation Gastrointestinal cramps Emesis
Diarrhea/Diaphoresis Urination Miosis Bradycardia/Bronchorrhea/Bronchospasm Emesis Lacrimation Salivation
Complications Seizures and pulmonary hypersecretion, or bronchorrhea and bronchoconstriction, are prominent mechanisms of early morbidity and mortality in cases of poisoning from acetylcholinesterase inhibitors. Bronchorrhea is often incorrectly called noncardiogenic pulmonary edema because the origin of the excessive pulmonary fluids is from airway secretions and not transudation of fluid across the alveolar-capillary membrane. The obstruction of upper and lower airways, the potential intrusion of these bronchial secretions into alveolar sacs, and bronchoconstriction produce hypoxia, which is the primary concern in the initial stages of poisoning.2 Nicotinic hyperstimulation of skeletal muscle determines the ultimate morbidity and mortality of acetylcholinesterase inhibitors. Signs of skeletal muscle hyperactivity include involuntary twitches, fasciculations, and hyperactive reflexes. Muscle hyperactivity eventually gives way to muscle fatigue and paralysis, including the respiratory musculature and particularly the diaphragm.6,8 Respiratory insufficiency may be delayed and result in death if not anticipated and corrected by mechanical or pharmacologic means.
Diagnostic Strategies Any patient with a full-blown cholinergic syndrome should be treated empirically without waiting for laboratory confirmation of decreased cholinesterase activity. Known or suspected exposure to cholinesterase inhibitors should be confirmed by ordering plasma and erythrocyte (RBC) cholinesterase levels. In acute exposures, the plasma cholinesterase levels decrease first, followed by decreases in RBC cholinesterase levels. The RBC cholinesterase level is more indicative of what is occurring at the nerve terminal.3 Patients with chronic exposures may show only reduced RBC cholinesterase activity, with a normal plasma cholinesterase level. The true reflection of depressed cholinesterase activity is found in the RBC activity, and even a mild acute exposure may result in severe clinical poisoning. RBC cholinesterase levels recover at a rate of 1% per day in untreated patients and take approximately 6 to 12 weeks to normalize, whereas plasma cholinesterase levels may recover in 4 to 6 weeks. Other studies should focus on the evaluation of pulmonary, cardiovascular, and renal function and fluid and electrolyte balance. Patients presenting with no acidosis, or only a metabolic acidosis on the arterial blood gas, have lower mortality than those presenting with a respiratory or mixed acidosis.10
Differential Diagnosis Few toxins or other clinical conditions produce the same symptoms as acetylcholinesterase inhibitors. A species of mushroom, Amanita muscaria, historically has been mentioned in the differential diagnosis, but it actually contains alkaloids that usually produce an anticholinergic (antimuscarinic) syndrome. A variety of conditions that induce excessive vagal responses (e.g., inferior wall myocardial infarction) may also produce some signs suggesting acetylcholinesterase inhibition, but other symptoms should make the primary cause apparent.
Management Treatment is directed toward four goals: (1) decontamination, (2) supportive care, (3) reversal of acetylcholine excess at muscarinic sites, and (4) reversal of toxin binding at active sites on the cholinesterase molecule. Decontamination should start in the out-of-hospital phase of care to prevent greater absorption and subsequent toxicity and to protect care providers. Decontamination is particularly important in cases of dermal exposure; removal and destruction of clothing and thorough flushing of exposed skin may limit absorption and subsequent toxicity. Alternatively, dermal decontamination can be done with dry agents, such as military resins, flour, sand, or bentonite. Caregivers are at risk for contamination from splashes or handling of contaminated clothing. Treating personnel may be rotated
Chapter 161 / Pesticides
Sympathetic ganglia
Acetylcholinesterase inhibitors produce direct toxic effects on the central nervous system leading to neurologic signs of confusion, combativeness, seizures, and coma. Status epilepticus may occur in severely poisoned patients. Structural central nervous system damage may occur if seizures are not terminated rapidly.9 A unique effect of organophosphorus insecticides results from “aging,” the irreversible conformational change that occurs when the organophosphorus agent is bound to the cholinesterase enzyme for a prolonged time. On average, for commercial organophosphorus agents some aging will occur by 48 hours, but aging may take longer. Once the enzyme has aged, an oxime antidote cannot regenerate the cholinesterase.
PART IV ■ Environment and Toxicology / Section Two • Toxicology
2054
to limit their exposure to the organophosphates.6 Caregivers should use universal precautions, including eye shields, protective clothing, and nitrile or butyl rubber gloves. In the case of ingestion, GI decontamination procedures are of questionable benefit because of the rapid absorption of these compounds. Profuse vomiting and diarrhea are seen early in ingestion and may limit11 or negate any beneficial effect of additional GI decontamination.12,13 Equipment, but not tissues, may be washed with a 5% hypochlorite solution to inactivate the cholinesterase inhibitor. Because death is from airway and respiratory failure, supportive care should be directed primarily toward airway management, including suctioning of secretions and vomitus, oxygenation, and, when necessary, ventilatory support. Succinylcholine can be used for intubation but may have an extremely prolonged duration.3 It is preferable to use a competitive neuromuscular blocking agent, such as rocuronium, for rapid sequence intubation in these patients, but increased dosing may be necessary. Although some authors have advocated the use of beta-blockers to control tachycardia, this may increase cardiovascular instability and worsen bronchospasm.12,14 Most cardiovascular complications that occur in this setting rarely require specific therapy. The definitive treatment of acetylcholinesterase inhibition starts with atropine.2 A competitive inhibitor of acetylcholine at muscarinic receptor sites, atropine reverses the clinical effects of cholinergic excess at parasympathetic end organs and sweat glands. Large doses of atropine may be required.15 Data suggest that the more rapid the atropinization, the faster control is obtained.6,12,13 Suggested dosing is 1 or 2 mg of atropine (0.02–0.05 mg/kg) intravenously, with doubling of each subsequent dose every 5 minutes until there is control of mucous membrane hypersecretion and the airway clears.6,12,13,16 If intravenous access is not immediately available, atropine may be administered intramuscularly. Patients may require 200 to 500 mg of atropine intravenously during the first hour, followed by prolonged continuous infusions of 5 to 100 mg/hr to maintain adequate secretion control.16 Tachycardia and mydriasis may occur at these doses, but they are not indications to stop atropine administration. The endpoint of atropinization is drying of respiratory secretions, easing of respiration, and a mean arterial pressure greater than 60 mm Hg.16 Animal evidence suggests that early rapid atropinization may limit seizure propagation and, in conjunction with diazepam, prevent status epilepticus.9,17 Atropine is not active at nicotinic sites and does not reverse the skeletal muscle effects (e.g., muscle fatigue and respiratory failure).3,13 Other anticholinergic medications such as diphenhydramine or ophthalmic agents may have benefit if atropine is scarce or unavailable; however, optimal intravenous dosing is not known.18-20 The second part of acetylcholinesterase inhibition treatment is the use of an oxime, such as pralidoxime (2-PAM, Protopam) or obidoxime (Toxigonin), to regenerate the organophosphate-acetylcholinesterase complex and restore cholinesterase activity at muscarinic and nicotinic sites.3,6,13,21 There are various dosing regimens; the most common dose of pralidoxime is 1 or 2 g intravenously (pediatric dose, 25–50 mg/kg); additional doses may be given based on clinical response and serial cholinesterase levels. The medication may be given in a bolus of 1 or 2 g intravenously over 30 to 60 minutes every 4 to 8 hours or 500 mg/hr (pediatric dose, 10–25 mg/kg/hr).21,22 The World Health Organization recommends an initial dose of 30 mg/kg followed by 8 mg/kg/hr continued for at least 24 hours or, if an infusion cannot be used, 30 mg/kg every 4 hours.23 The infusion may be continued for several days with no adverse effects attributable to the pralidoxime; however, rapid administration can lead to hypertension, vomiting, and a
transient reversible neuromuscular blockade.24 The ideal dose of pralidoxime should be determined by monitoring the clinical condition of the patient and serial cholinesterase levels; the patient may require higher doses of oxime than recommended here. The World Health Organization recommended infusion dose of obidoxime is 4 mg/kg followed by 0.5 mg/kg/hr; alternatively, intermittent intravenous doses of 4 mg/kg, then 2 mg/kg, every 4 hours are given.23 Pralidoxime and obidoxime can be administered by intramuscular injection. Indications for oxime therapy include respiratory depression/apnea, fasciculations, seizures, arrhythmias, cardiovascular instability, and use of large amounts of atropine. Oxime therapy can be used whenever the patient requires more than a limited amount of atropine (2–4 mg) to completely reverse the signs and symptoms of intoxication or in any patient who requires repeated doses of atropine. Oxime therapy and atropine are synergistic. In the past, pralidoxime was only used within the first 24 hours because of aging of the organophosphate-acetylcholinesterase complex,2 but not all organophosphates behave in a similar manner. Dimethyl and diethyl phosphoryl insecticides react differently at variable rates with acetylcholinesterase and oxime therapy. Many organophosphates are highly lipid soluble and slowly leach out of fat stores for up to 6 weeks, resulting in newly formed complexes with excellent reversal of the cholinesterase inhibition by pralidoxime clinically and by measurements of cholinesterase activity. Pralidoxime can also combine with unbound organophosphates and prevent their subsequent binding to nerve terminals. Even with optimal treatment, seriously intoxicated patients may require longterm supportive care, including ventilator support.6,22 In conjunction with atropine and oxime pralidoxime, patients with agitation, seizures, and coma should be treated with adequate doses of a benzodiazepine after the airway has been secured.6,7,9,25 Although diazepam is most studied, any parenteral benzodiazepine may be used. The military has classically used diazepam autoinjectors for intramuscular use, but midazolam is the best intramuscular agent, with lorazepam as an alternate. Sarin, soman, tabun, and VX are nerve agents that might be used in a terrorist attack. These agents have important differences from the common household or commercial organophosphorous insecticides. These agents tend to age very quickly, with tabun (GA) aging in 14 hours, sarin (GB) in 5 hours, soman (GD) in 5 or 6 minutes, and VX in 48 hours. Due to this rapid aging, reversal of nerve agent poisoning is very time sensitive. VX is an oily but highly toxic agent with low volatility. It does not readily vaporize, and because it has a low risk of inhalation, exposure is predominately transcutaneous. The other agents can be mostly dispersed into the air by explosion or vaporization, resulting in inhalation exposure. These agents do not require the extremely large doses of atropine but do require pralidoxime.26-28 See Chapter 194 for further information on treatment. New therapies for treatment of organophosphorus poisoning, including the use of N-acetylcysteine and exogenous acetylcholinesterase, show promise in research studies.29,30 When added to anticholinergics, NMDA receptor antagonists may decrease organophosphorus compound–induced seizures.31
Disposition Because of the prolonged effects of acetylcholinesterase inhibition, most patients with significant exposures require hospital admission. Occasionally, a person with chronic exposure, depressed cholinesterase levels, and mild visual or GI symptoms may be followed on an outpatient basis; however, some
■ CARBAMATE INSECTICIDES Carbamate insecticides are another class of acetylcholinesterase inhibitors and are differentiated from the organophosphorus compounds by their relatively short duration of toxic effects. Carbamates inhibit acetylcholinesterase for minutes to 48 hours, and the carbamate-cholinesterase binding is reversible.2 Although the clinical picture of acute carbamate poisoning may be identical to that of organophosphate poisoning, the toxic effects are limited in duration and patients may require only decontamination, supportive care, and treatment with adequate doses of atropine. Although the duration is limited in scope, patients may become just as sick and require assisted ventilation and seizure therapy. The use of pralidoxime is controversial in carbamate poisoning; an animal study suggests that pralidoxime administration may produce greater toxicity in cases of carbaryl (Sevin) poisoning, although the author has used pralidoxime in carbaryl-poisoned humans without adverse events.35 Nevertheless, if doubt exists as to whether a severe poisoning is due to a carbamate or organophosphate, pralidoxime should be administered. It is the author’s practice to use oximes when patients present with a cholinergic toxindrome and a history of exposure of organophosphorus or carbamate insecticides.
■ CHLORINATED HYDROCARBON INSECTICIDES DDT, the prototype of chlorinated hydrocarbon insecticides (sometimes referred to as organochlorine insecticides), was first used extensively during World War II for the successful control of typhus and malaria and was used widely in the United States as a general insecticide after the war. Because of the effectiveness of DDT, many other chlorinated hydrocarbon insecticides were developed. These insecticides were used extensively in agricultural, commercial, and residential pest control. However, although these insecticides were very
2055
effective, their widespread use, long half-life, and persistence had negative ecologic repercussions. Many of these insecticides have been targeted as persistent organic pollutants by international agencies, leading to their restricted use.36 Although chlorinated hydrocarbon insecticides are no longer used in the United States for agricultural use, γ-hexachlorobenzene, better known as lindane (Kwell), is still used as a topical medicinal agent for the treatment of head lice and scabies. As a result, lindane is probably the most common cause of toxicity from an organochlorine compound in the United States. Given its toxicity, lindane is no longer a firstline agent for the treatment of scabies.37 In 2001, California issued a ban on the use and sale of lindane, and other states are considering a ban on lindane.36
Principles of Disease Chlorinated hydrocarbon pesticides are highly lipid soluble. They are readily absorbed through dermal, respiratory, and GI routes.38 Dermal and GI exposures account for most clinical poisonings, including inappropriate external use of lindane or other compounds and the occasional accidental oral administration of lindane. Because they are so lipid soluble, these compounds are stored in fatty tissues, and repeated small exposures result in accumulation and eventual clinical toxicity.39 Chlorinated hydrocarbon insecticides primarily affect axonal membranes, resulting in neuronal irritability and excitation. Toxicity occurs in central and peripheral neurons.40 Some of the organochlorines can inhibit the chloride channel of γ-aminobutyric acid (GABA) receptors, leading to decreased inhibition of the central nervous system.41 Chlorinated hydrocarbons induce hepatic microsomal enzymes and produce hepatic tumors in some animals. This potential carcinogenicity is the basis for current human health concerns, but it is only theoretical. Chlorinated hydrocarbon insecticides, including chlorinated hydrocarbon solvents, may sensitize the myocardium to circulating catecholamines and increase susceptibility to ventricular dysrhythmias, such as tachycardia and fibrillation.40
Clinical Features Signs and Symptoms The primary clinical picture of acute or cumulative toxicity from chlorinated hydrocarbon pesticides is related to their neurotoxicity. Premonitory peripheral signs and symptoms, such as tremor or paresthesias, may be absent, and the first sign of toxicity may be the acute onset of seizure activity.42 Additional signs include confusion, combativeness, and muscle twitching. Untreated, continued muscle activity can lead to hyperthermia, metabolic acidosis, and rhabdomyolysis with secondary acute tubular necrosis.43 Because many of these agents are halogenated, ventricular dysrhythmias may occur from catecholamine sensitization and direct myocardial toxic effects. Immediate hepatotoxicity is unlikely without secondary hyperthermia or other metabolic complications.40,44 Longterm exposure may result in neuropsychiatric symptoms.45 Diagnosis may be difficult in chlorinated hydrocarbon pesticide exposure because the patient may be unable to provide a history. Nonhospital personnel are often in the best position to obtain information concerning pesticide availability and use and the situation surrounding the exposure. Another clue is the solvent odor and oily feel of the hydrocarbon solvent containing the highly lipid-soluble chlorinated hydrocarbon pesticides.
Chapter 161 / Pesticides
patients, particularly those exposed to fenthion, initially present with signs and symptoms of mild exposure and progress to severe, life-threatening toxicity over time.32 If plasma cholinesterase levels are available, they may be useful for treatment and disposition decisions. Asymptomatic or minimally symptomatic patients with normal or minimally depressed levels may be discharged after 4 to 6 hours with close outpatient follow-up to ensure that progressive toxicity does not occur. Patients with severely depressed levels (usually associated with significant symptoms) require admission and close monitoring, usually in a high-intensity care unit. Patients may develop rebound toxicity several days after apparently satisfactory response to initial treatment. Rebound toxicity may occur for many reasons, including persistent release of organophosphates from lipid stores. A secondary syndrome, the intermediate syndrome (IMS), occurs 24 to 96 hours after exposure and consists of proximal muscular weakness specifically of the respiratory muscles. It is believed to be an abnormality at the neuromuscular junction. Patients with IMS present with respiratory failure several days after the acute cholinergic symptoms have resolved and may require several weeks of ventilatory support. It is theorized that this may occur as a result of inadequate initial oxime treatment or premature discontinuation of oxime therapy.6,33 Oximes may be beneficial for IMS; however, this is controversial.34 Finally, organophosphorus-delayed neuropathy has been reported as a different entity and affects an axonal enzyme, neurotoxic esterase, and leads to a peripheral sensorimotor neuropathy 7 to 21 days after exposure.6
2056
PART IV ■ Environment and Toxicology / Section Two • Toxicology
Diagnostic Strategies Diagnosis must be confirmed by history or by investigation at the site of the exposure to establish the offending agent with certainty. No specific tests are readily available to confirm the diagnosis of chlorinated hydrocarbon pesticide poisoning. Some reference laboratories can measure fat and plasma levels, but results are difficult to interpret and seldom available during the acute phase of toxicity.39,44 Ancillary laboratory and other studies should be based on the clinical condition, complications, and consideration of alternative diagnoses on an individual basis.
Differential Considerations The differential diagnosis includes virtually every condition that produces seizures. The specific diagnosis depends on obtaining the history of significant acute or chronic chlorinated hydrocarbon pesticide exposure.
Management and Disposition Skin decontamination with soap and water may reduce toxicity in acute dermal exposure. High lipid solubility results in rapid absorption, and delayed GI decontamination is not of benefit. Elimination of some chlorinated hydrocarbon insecticides can be increased, and repeat doses of cholestyramine (4 g orally every 8 hours) given during a mass exposure of chlordecone (a chlorinated hydrocarbon insecticide) enhanced the fecal elimination of this compound.45-47 The primary therapeutic objective is seizure control, which is best accomplished with short-acting benzodiazepines or barbiturates. Recurrent seizures or status epilepticus may require high-dose barbiturates and paralyzing agents (e.g., pancuronium or vecuronium) to prevent secondary morbidity from continuous motor activity in prolonged seizures. The seizure activity is usually self-limiting, lasting only 1 or 2 days even in severe cases.42,46,48 Continuous cardiac monitoring during the acute phase is indicated because of the potential for myocardial sensitization. Ventricular dysrhythmias are most likely to occur during seizure activity because of the high circulating catecholamine levels and other metabolic abnormalities present during seizures. Dysrhythmias should be treated with beta-adrenergic antagonists, such as propranolol, metoprolol, or esmolol, to reduce the effect of catecholamines on the myocardium. Additional treatment should focus on the complications of prolonged seizure activity, such as rapid external cooling measures for hyperthermia. Metabolic acidosis is almost always transient and resolves spontaneously without treatment. Rhabdomyolysis and myoglobinuria should be anticipated. Other complications of seizures should be treated as indicated. Because of their high lipid solubility, chlorinated hydrocarbon pesticides are distributed largely in tissues and are not amenable to hemoperfusion, dialysis, or other attempts to enhance elimination. Patients who have acute or cumulative chlorinated hydrocarbon pesticide toxicity require hospitalization until their seizures are controlled, complications have resolved, and they have returned to their neurologic baselines; this usually occurs within 1 or 2 days. Severe complications, such as renal failure from rhabdomyolysis, may prolong the clinical course.
■ SUBSTITUTED PHENOLS The substituted phenols include dinitrophenol (DNP), pentachlorophenol, and dinitrocresol. These compounds have
been used since the 1930s as insecticides, termiticides, herbicides, and wood preservatives. They are currently used in agricultural, commercial, and residential applications, including over-the-counter preparations for home gardeners. Substituted phenols such as DNP are abused as weight-reduction agents and occasionally are used in illegitimate weightreduction operations.49
Principles of Disease Substituted phenols are readily absorbed through the skin and GI tract, and aerosols may be absorbed through the respiratory tract. There is some potential for cumulative toxicity with repeated exposures, but much less so than with the organophosphorus and chlorinated hydrocarbon pesticides previously discussed. Substituted phenols produce their toxicity by uncoupling cellular oxidative phosphorylation; this leads to inefficient production of high-energy phosphate substrates and increased cellular use of oxygen, glucose, and water, with subsequent excessive heat production.50 These compounds are commonly used during the summer when the external heat predisposes users to increased toxicity.51,52 In addition, nitro-substituted phenols may produce methemoglobinemia.
Clinical Features Patients with substituted phenol toxicity present hypermetabolic and hyperthermic, tachycardic, tachypneic, and profusely diaphoretic. They may also have a relative hypovolemia from excessive insensible fluid losses through sweating and metabolic consumption. Loss of energy production in the brain results in neurologic changes ranging from confusion to seizures and coma. Renal and hepatic injury, and rhabdomyolysis with myoglobinuria, is common.43 Because phenols are generally corrosive, patients with dermal exposures often have irritation or chemical burn, and some substituted phenols, such as dinitrophenol, produce a characteristic yellow staining of the skin or mucous membranes at the site of absorption. This same staining can be found throughout the internal organs at autopsy.52 Cataracts are a complication of long-term exposure. This condition was common in patients who used substituted phenols as part of a weight-reduction regimen and was partially responsible for the banning of this substance. The cataracts regress spontaneously after exposure is discontinued.53
Diagnostic Strategies Laboratory evaluation of patients with substituted phenol toxicity is aimed at identifying deficiency of aerobic metabolic substrates, including oxygen, glucose, and water. A complete blood count may reveal hemoconcentration and a nonspecific leukocytosis. Electrolyte abnormalities depend on the duration and severity of symptoms, environmental factors, and complications or underlying disease states. Arterial blood gas measurements show varying degrees of acidosis, depending on the extent of anaerobic metabolic activity due to oxidative phosphorylation uncoupling and associated tissue hypoperfusion from dehydration. Serum enzyme determinations document the extent of hepatic, renal, and skeletal muscle injury. The presence of phenolic compounds in the urine of a patient with this clinical picture strongly suggests substituted phenol pesticides as the causative agent.
2057
Differential Considerations
Management and Disposition Initial treatment is directed toward control of body temperature; treatment of acidosis; protecting the kidneys, brain, and liver from hyperthermic damage; and providing the basic substrates for excessive metabolic activity—oxygen, glucose, and water.52 If the chemical exposure is known or recognized, early decontamination of affected sites is important. Therapy should be directed toward prevention or minimization of the associated complications discussed previously. Patients with mild toxicity can usually be stabilized after a few hours and discharged from the emergency department. Patients with significant organ system injury or a high likelihood of complications, such as prolonged or recurrent seizures, significant alteration of consciousness, and rhabdomyolysis, require admission, usually to the intensive care unit.
■ CHLOROPHENOXY COMPOUNDS The chlorophenoxy pesticides were developed in the early 1940s and hailed as a selective herbicide particularly effective against broadleaf weeds. This class of herbicide developed a special notoriety during the Vietnam War as Agent Orange, a defoliant used in aerial spraying. Agent Orange consisted of a mixture of 2,4-dichlorophenoxyacetic acid (2,4-D) and 2,4,5trichlorophenoxyacetic acid (2,4,5-T). 2,4,5-T is almost always contaminated with isomers of tetrachlorodibenzo dioxin. This concern regarding dioxin exposure has led to the extensive medical investigations of Vietnam veterans and severe restrictions on the production and use of 2,4,5-T.54 Because of the relative safety and broadleaf selectivity of 2,4-D, however, most home gardeners have at least one chlorophenoxy compound on a shelf in their garages, and some old cans may contain 2,4,5-T or a mixture of both compounds.
Principles of Disease Chlorophenoxy compounds may be absorbed through the skin, GI tract, and respiratory tract, but almost all significant poisonings occur as a result of accidental or intentional ingestion. The lipid solubility of these compounds is low, and excretion is fairly rapid, so cumulative toxicity from repeated exposures does not occur.55 Although skeletal muscle is the target organ for chloro phenoxy herbicides, the exact mechanism is obscure.40,56 Depending on severity, muscular abnormalities may range from generalized muscle weakness to acute rhabdomyolysis. Higher doses may also uncouple oxidative phosphorylation and cause a hypermetabolic state similar to that seen with the substituted phenols.40
Clinical Features Similar to most organic pesticides in an organic solvent, the chlorophenoxy herbicides may produce mild, nonspecific
Diagnostic Strategies There are no specific tests for the detection of the chlorophenoxy compounds. Laboratory evaluation should be aimed at evaluating skeletal muscle injury and its complications. Severely poisoned patients require generalized organ system evaluation, including hepatic and renal function, because of the effects of rhabdomyolysis and hyperthermia.
Differential Considerations Differential diagnostic possibilities include other causes of acute myopathy. The manifestation of chlorophenoxy compound toxicity is extremely rare, however, and without a definite history or strong suspicion of exposure, other explanations for acute myopathy should be pursued.
Management and Disposition Treatment consists of initial skin decontamination, activated charcoal or gastric lavage with early presentation, and basic supportive care. Serious toxic effects develop within 4 to 6 hours after ingestion, and treatment can be directed toward the specific problems of muscle weakness, airway and ventilatory support, and rhabdomyolysis. Treatment of hyperthermia and acidosis has been discussed previously. Asymptomatic or minimally symptomatic patients may be discharged with reassurance after 4 to 6 hours of observation. Patients with significant toxicity should be admitted for close observation and monitoring.
■ BIPYRIDYL COMPOUNDS The bipyridyl (also called dipyridyl) compounds, paraquat and diquat, were first investigated in the late 1950s and early 1960s. They are extremely effective contact herbicides that are rapidly inactivated by the surrounding soil in the event of overspray. Paraquat is activated when exposed to sunlight, which led to its use as the herbicide of choice during aerial spraying of marijuana by the U.S. and Mexican governments. After spraying, however, growers simply would harvest the crops before the plants were exposed to enough sunlight to damage the plants, resulting in an apparently healthy harvest but one contaminated with paraquat. The burning of marijuana pyrolyzes paraquat into a nontoxic form, a fact that was lost in the warning messages dispensed by the government at that time.58
Principles of Disease Of the two bipyridyl compounds in use, paraquat is the most clinically significant in terms of number of cases and toxic effects. Paraquat use is tightly regulated in the United States but is widespread throughout the world. Diquat is less regulated in the United States and is included in some formulations of Roundup. Paraquat is absorbed through the skin, GI tract, and respiratory tract. Almost all fatal exposures have resulted from the ingestion of paraquat, although a few case reports have involved extensive skin contamination.59 Toxicity has occurred, but no fatal cases have been reported from inhalation of paraquat vapor or aerosols. Diquat is poorly
Chapter 161 / Pesticides
Acute toxicity from substituted phenol poisoning is difficult to distinguish from typical environmental heat-related emergencies or toxicity from sympathomimetics or salicylates. Continued evidence of hypermetabolic activity and metabolic acidosis after routine cooling measures, rehydration, and other supportive care should lead to a consideration of toxin-induced states. Persistent hyperthermia and acidosis in a weight lifter should trigger concern for DNP abuse. The presence of yellow staining virtually clinches the diagnosis.52
dermal and GI irritation with nausea, vomiting, and GI distress. Large exposures are likely to cause systemic symptoms ranging from diffuse myotonia and muscle fasciculations progressing to rhabdomyolysis, hyperthermia, and a hypermetabolic state with metabolic acidosis.57
PART IV ■ Environment and Toxicology / Section Two • Toxicology
2058
absorbed through intact skin, and most cases of toxicity result from ingestion.60 A formulation change in Sri Lanka, adding an emetic—magnesium sulfate—and an alginate binder, decreased mortality by 9.5% in cases of intentional ingestion of paraquat.61 Paraquat’s toxic effect results from the production of superoxides created during cyclic oxidation-reduction reactions of the compound in tissues. Lipid peroxidation of cellular membranes seems to be one significant pathway of cellular injury.62,63 Paraquat selectively concentrates in the lungs because of an amine uptake mechanism in alveolar cells. In addition, high concentrations of oxygen significantly increase the extent of paraquat-induced injury so that the lungs are the major target organ. The pathophysiologic lesions include direct injury to the alveolar-capillary membrane followed by surfactant loss, adult respiratory distress syndrome, progressive pulmonary fibrosis, and respiratory failure.64 Paraquat damages other major organ systems by the same cellular membrane effects, including the liver, kidneys, heart, and central nervous system. Diquat has similar effects, with most of its toxicity concentrated in the kidneys and not preferentially in lung tissue.60
Clinical Features Signs and Symptoms Both agents are extremely corrosive and cause nausea, vomiting, and severe chemical burns of the oropharynx soon after ingestion. Patients who ingest concentrated paraquat frequently die as a result of esophageal perforation and mediastinitis before development of the characteristic progressive pulmonary injury. Patients with dermal paraquat exposures show significant skin irritation, and ocular exposures may produce severe corneal injury.65 The classic finding of paraquat-induced progressive pulmonary injury usually occurs over 1 to 3 weeks, although the clinical course varies considerably with severity of poisoning, involvement of other organ systems, and underlying medical problems.64 This is not a factor in the emergency department, and the delayed pulmonary injury is not discussed here. In contrast to paraquat, diquat usually spares the lungs but produces similar toxicity in all other organ systems.60
Diagnostic Strategies Paraquat is measurable in the blood, and the nomogram provides a fairly accurate prognosis. The assay is not readily available in the United States, and in most cases, by the time the results are obtained, nothing more can be done to change the eventual outcome. There is a qualitative bedside test that uses the reduction of paraquat or diquat in alkalinized urine by sodium dithionite, but the reagent frequently is not available.57 Studies other than evaluation of caustic GI injury and pulmonary and renal damage should be directed toward secondary effects of the poisoning.
Differential Considerations A person with acute paraquat or diquat ingestion is likely to present with the initial complaint of an acute corrosive injury; the differential diagnosis should encompass all corrosive agents. Successful therapeutic intervention for paraquat toxicity is extremely time dependent, and patient outcome depends on the history. Any patient who has evidence of pulmonary or other organ injury caused by paraquat exposure is probably already beyond recovery.
Management and Disposition There are no studies comparing various treatment strategies, but the key to successful treatment of an acute paraquat exposure likely depends on early decontamination measures to limit absorption. Thorough skin cleansing is obvious and straightforward in dermal exposures. Careful gastric lavage and administration of activated charcoal may be lifesaving, but these measures should be undertaken in consultation with a poison center and may even be hazardous in the context of a corrosive ingestion. Early endoscopy and surgical intervention may be necessary if there is evidence of esophageal perforation and mediastinitis. Although Fuller’s earth and bentonite are recommended as adsorbents in paraquat ingestions, activated charcoal is much more readily available in the United States and has equal, if not greater, efficacy.66 Although controversial, many toxicologists recommend rapid initiation of charcoal hemoperfusion to rapidly lower plasma paraquat levels and to limit pulmonary and other organ system uptake of paraquat. Many also recommend serial and combined hemoperfusion and hemodialysis, particularly during the first 24 hours after exposure.67,68 There are multiple suggested treatment protocols for paraquat, such as N-acetylcysteine, low fraction of inspired oxygen, and cytoprotective agents such as amifostine, but no single therapy has proven consistently successful.69,70 Patients with any significant dermal paraquat exposure and all patients with ingested paraquat require hospitalization and consideration of enhanced elimination therapy. These patients should be observed and treated expectantly until paraquat levels are reported to be nonexistent or nontoxic.
■ PYRETHRINS AND PYRETHROIDS Pyrethrins are naturally occurring insecticides of the yellow Chrysanthemum cinerariifolium and Tanacetum cinerariifolium and are among the oldest known insecticides, first used in the 1800s. Extracts of the dried flowers contain the active compound pyrethrum, which contains six naturally occurring pyrethrins. In addition, numerous synthetic derivatives, pyrethroids, have been produced and have greater chemical stability than the natural pyrethrins. Type II pyrethroids contain a cyano substituent and are among the more toxic formulations of this class. These present a potential danger to humans, but type II pyrethroids are generally less toxic than many of the other classes already discussed and are being used more commonly.
Principles of Disease Because pyrethrins and pyrethroids are most commonly aerosolized, inhalation is the most likely route of exposure. The patient may not be aware of an exposure because pyrethrin and pyrethroid aerosols are used frequently as automated insect sprays in public areas, such as in airplanes. In these situations, concentrations rarely reach levels likely to produce symptoms in any but the most sensitized patient. Occasional ingestions have been reported, and significant toxicity is possible via this route. Systemic absorption via the dermal route is unlikely, but topical effects are possible. Most pyrethrins and pyrethroids are rapidly metabolized and deactivated in human exposure, so cumulative toxicity is not a problem. Piperonyl butoxide, which is added as an insect “knockdown” agent, may increase the toxicity of the pyrethrum derivatives. Pyrethrins and pyrethroids have a variety of effects in humans and other mammals.71,72 Clinically, the naturally
Clinical Features Allergic manifestations, including potentially life-threatening events, may occur after acute inhalation or dermal exposure. Inhalation exposure often occurs with the use of a pyrethrinbased aerosol in an enclosed, poorly ventilated space. Local effects include lacrimation, rhinitis, rhinorrhea, sneezing, throat irritation, and pharyngeal and laryngeal edema. Lower respiratory effects include cough, shortness of breath, chest pain, and wheezing. Skin rashes, consistent with a contact or allergic dermatitis, and photosensitivity may contribute to the dermatologic picture. There is potential for allergic cross-reactivity in patients who are allergic to ragweed. Sodium channel-mediated and GABA-mediated chloride channel effects mediate neurologic signs and symptoms. Facial paresthesias have been reported, and seizures occur with massive ingestions.71,72 Nonspecific symptoms, such as headache, fatigue, dizziness, and weakness, have been reported.
Diagnostic Strategies No laboratory tests are available to measure pyrethrins or pyrethroids in a clinical setting.
Differential Considerations The differential diagnosis of the signs and symptoms of pyrethrin or pyrethroid toxicity includes the usual causes of bronchospasm and seizures and other acute neurologic complications.
Management and Disposition Decontamination, including removal from a contaminated environment or washing, should be the first step. Definitive treatment is supportive and directed at the respiratory and neurologic complications. Disposition of a patient with exposure to pyrethrins depends on the severity of the underlying complications. If discharge from the emergency department is anticipated, the patient should be counseled with regard to the possibility of recurrent allergic phenomena on reexposure.
■ GLYPHOSATE Glyphosate (Roundup) was introduced as a broad-spectrum nonselective herbicide in 1971 by the Monsanto Agricultural Company. It is the isopropyl ammonium salt of a noncholinesterase-inhibiting organophosphate herbicide. It is sold mixed with the surfactant polyoxyethylene amine (POEA). Because it is effective on broadleaf weeds and does not undergo photodecomposition, it is popular in the home market. Newer formulations of Roundup may contain diquat.
2059
Principles of Disease Glyphosate is poorly absorbed through the skin so that most exposures result from ingestion. The concentrated solution is extremely irritating, and patients may vomit with subsequent aspiration. The concentrated solution is provided as 41% glyphosate in 15% POEA. The directions state that it should be diluted to a 1% glyphosate solution. Glyphosate is toxic to plants by inhibition of the enzyme 5enolpyruvylskikimate-3 phosphatase-synthetase in the shikimic acid metabolic pathway. After application of glyphosate on the leaves, it is transported to the roots, where the enzyme is active. Humans lack this enzyme and are unlikely to develop toxicity. Reported toxicity is believed to result largely from the surfactant POEA and may reflect the direct corrosive effect from the amine salt, or it may uncouple oxidative phosphorylation.73
Clinical Features Most ingestions of the dilute solution cause only minimal symptoms, including GI distress. Patients ingesting large volumes of dilute solutions or moderate volumes of concentrated solutions complain of sore throat, nausea, abdominal pain, and fever. They may develop vomiting, diarrhea, respiratory distress, noncardiogenic pulmonary edema, dysrhythmias, shock, coma, and renal failure. Acidosis reflects poor tissue perfusion and cardiovascular compromise.73 Negative prognostic indicators include shock, acidosis, and persistent hyperkalemia.73
Diagnostic Strategies The critical element in diagnosis is history of ingestion. Laboratory analysis may demonstrate an anion gap metabolic acidosis, hypoxemia, and hyperkalemia. Elevated transaminases may occur in 30% of ill patients, and signs of renal failure may develop in persistent shock states. The electrocardiogram may show ventricular dysrhythmias and secondary signs of hypoxemia.73
Differential Diagnosis The differential diagnosis includes most corrosive ingestions and causes of shock. The findings of hyperkalemia and metabolic acidosis may suggest hydrofluoric acid ingestions. A normal ionized calcium level may help rule out hydrofluoric acid exposure. Any cause of aspiration should also be considered. The history is the most useful factor in the differential diagnosis.
Management and Disposition Treatment is supportive. The patient may require positivepressure ventilation to overcome the noncardiogenic pulmonary edema. POEA may also be a direct cardiac depressant; inotropic agents can be useful. Hyperkalemia should be treated in the usual manner with fluids, medications to shift potassium into the cell (e.g., bicarbonate, calcium, and beta-adrenergic agonists), and kayexalate. If there is an indication of significant corrosive ingestion, early endoscopy with placement of stent, high-dose steroids, and laparotomy may be considered. Asymptomatic patients with small ingestions of dilute substances may be observed for 6 hours and discharged. Patients with complaints consistent with corrosive ingestions require admission and GI evaluation. Any patient with pulmonary complaints requires admission and intensive supportive care.
Chapter 161 / Pesticides
occurring pyrethrins can cause sensitization and allergic phenomena. This property does not occur with the synthetic pyrethroids. Both classes are associated with sodium channel blockade, slowing the rate of activation of the sodium channel and extending the time during which the channel is open. In addition, both classes affect GABA receptors, inhibiting chloride channel function. Less significant effects include potentiation of nicotinic cholinergic neurotransmission, enhancement of norepinephrine release, and inhibition of calcium adenosine triphosphatase interference with sodium-calcium exchange across membranes.71,72
2060
PART IV ■ Environment and Toxicology / Section Two • Toxicology
■ DEET N,N-diethyl-m-toluamide or N,N-diethyl-3-methylbenzamide (DEET) is not a pesticide but an insect repellent. It is the most widely used chemical insect repellent in the United States. DEET was developed by scientists at the U.S. Department of Agriculture in 1946, patented by the U.S. Army soon thereafter, and released to the general public in 1957.74 With the prevalence of Lyme disease and other concerning arthropod-borne diseases, the use of DEET has increased greatly. Formulations containing DEET range from 5 to 100%. The U.S. Army routinely used 75% solutions until 1987 but now uses a 35% time-release, polymer-based formulation. The American Academy of Pediatrics (AAP) recommends 30% as the maximum concentration that should be used in infants and children. The AAP does not recommend use of DEET in infants younger than 2 months.75
Principles of Disease DEET is lipophilic and can be absorbed through the skin. Skin absorption and toxicity increase with repeated applications, increased ambient temperatures, sweating, and abraded, thin skin. Ingestion may lead to toxicity.76 DEET primarily affects the central nervous system. Its mechanism of action is unknown. It may sensitize the skin and cause allergic reactions.
Clinical Features Prolonged skin contact may lead to contact dermatitis, and prolonged contact with high concentrations has led to skin blisters. Patients who have ingested DEET or have repeated skin applications in a hot enclosed environment that enhances absorption have developed liver function test abnormalities and neurologic findings, including encephalopathy, seizures, movement disorders, and coma.76 Most exposures to DEET result in no, or minimal, toxicity and should not preclude its use in susceptible populations in which significant arthropodborne diseases are prevalent.77
Diagnostic Strategies Exposure history is central to the diagnosis. Although DEET can be detected in urine, most laboratories are not able to do this testing during the acute toxicity phase. An electroencephalogram may be useful in a patient with coma or encephalopathy and seizures.
Differential Diagnosis The differential diagnosis includes conditions that may cause encephalopathy, seizures, and movement disorders. Such conditions include drug intoxication, infectious causes, drug interactions, and structural defects.
tion. After DEET ingestion, milk products and oil-containing foods should be avoided until the GI tract has eliminated the offending agent. Seizures should be treated with benzodiazepines. Asymptomatic patients who have ingested DEET-containing repellents should be observed for 4 to 6 hours. Patients who develop neurologic symptoms should be admitted and observed.
KEY CONCEPTS ■
All patients exposed to cholinesterase inhibitors should have skin decontamination. Emergency department personnel need to be protected during this process. ■ Morbidity in cholinesterase inhibitor exposure results from early airway compromise secondary to copious secretions, status epilepticus, and late respiratory failure. ■ Cholinesterase inhibitor exposure may include bradycardia or tachycardia, hypertension or hypotension, and miosis or mydriasis. ■ The clinical endpoint for atropine administration is drying of airway secretions. ■ Pralidoxime should be given to all organophosphoruspoisoned patients who require atropine regardless of time since exposure. ■ With chlorinated hydrocarbon exposures, skin decontamination with protection of personnel is indicated. ■ In chlorinated hydrocarbon exposures, catecholamine administration is avoided. ■ Supportive care, temperature control, and seizure control are important. ■ Rapid cooling and substrate provision (glucose) are the two most important therapies in substituted phenol toxicity. ■ Diagnosis of chlorophenoxy compound toxicity depends on a history of accidental or deliberate ingestion. ■ Rapid GI decontamination may be indicated in paraquat and diquat ingestions despite the corrosive injury. ■ The predominant form of pyrethrin and pyrethroid toxicity is allergic. ■ Small ingestions of dilute glyphosate solutions are GI irritants. Large or concentrated ingestions may cause acidosis, hyperkalemia, and noncardiogenic pulmonary edema. ■ DEET should not be applied over abraded or raw skin. ■ DEET applications to children should be restricted to 30% solutions, should not be used under occlusive clothing, and should be washed off completely between applications.
Management and Disposition Treatment is supportive. If DEET exposure is suspected, the skin should be thoroughly decontaminated. Oils or lipophilic agents should be avoided because they enhance skin absorp-
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 162
Plants, Mushrooms, and Herbal Medications
Richard D. Shih
■ PERSPECTIVE
■ BOTANICAL IDENTIFICATION
Botanicals such as plants, herbal products, and mushrooms have a long-standing and important place in medical history. Their use as therapeutic agents has been documented in the earliest of medical writings. The extraction of alkaloids from the opium poppy in the 1800s was a forerunner of modern pharmacology. The general public increasingly is using herbal products for medicinal purposes. Despite the tremendous growth in popularity of herbal products, data on herbal efficacy and toxicity are limited. This chapter does not focus on the medicinal use of natural products, but rather the toxicology related to their exposure or utility.
Although plants, mushrooms, and herbal products all are derived naturally, they have different exposure patterns and epidemiology.
Identification of the botanical in question is most useful, but most patients and medical professionals do not have any knowledge of botany or plant identification. The name of the plant or mushroom is often confusing because the scientific name is not typically known, and common names often overlap. Most emergency department personnel cannot identify common plants, such as mistletoe, holly berries, philodendron, and others. Several resources may be helpful, including plant atlases, CD-ROM plant databases, a local botanical expert, botanical garden personnel, or a poison center. Alternatively, digital photography of the plant or mushroom in question can be quickly e-mailed to a local expert to aid in identification.4 With herbal products, the product name or herbal plant may be known. Because of the limited Food and Drug Administration (FDA) regulation, however, the purported herb might have been harvested from the wrong plant or contaminated with other toxic material.
Unintentional Childhood Exposure
Plants
Unintentional childhood exposure occurs most commonly with plants. Approximately 5% of all poison center calls involve plants. Of these, about 75% involve children younger than 6 years. Most of these cases involve household plants with a limited amount of plant or toxin ingested, resulting in little or no toxicity.1
Among more than 100,000 plant exposures reported to U.S. poison centers annually, most plant exposures occur in children (6 hours postingestion), or no symptoms (edible). Mushrooms with serious toxicity and potential for death are in the group that have late onset of symptoms. Particular attention should be paid to the timing of initial symptoms.96–98 Mushroom species often grow together, and foragers frequently pick and eat more than one species of mushroom. The onset of early symptoms does not preclude the diagnosis of a more serious poisoning.
Early Onset of Symptoms Gastrointestinal symptoms of nausea, vomiting, diarrhea, and abdominal cramps are common among many of the mushroom groupings (see Table 162-3). These symptoms are most predominant, however, in the gastrointestinal toxin group. This group contains numerous diverse mushrooms, many with unknown toxins. They cause symptoms rapidly (0.5–3 hours) after ingestion, and symptoms typically last 24 hours. Treatment is supportive, with good outcome expected. CNS effects are associated with two groups of mushrooms, ibotenic acid/muscimol and psilocybin.97–99 Psilocybin is structurally related to serotonin and lysergic acid diethylamide (LSD). Similar to LSD, hallucinations and CNS effects are prominent. This mushroom is most often used as a drug of
Late Onset of Symptoms Three groups of mushrooms cause late onset of symptoms (>6 hours postingestion): cyclopeptide, gyromitrin, and orelline/orellanine. The orelline/orellanine-containing mushrooms are infrequently found in the United States and there have not been any reported cases of toxicity in North America. The cyclopeptide group is responsible for most mushroomrelated deaths in the United States. The cyclopeptide mushrooms contain many species, of which Amanita phalloides is the most well known. Several cyclopeptide toxins have been identified (e.g., amatoxins, virotoxins, phallotoxins) that are thought to be responsible for toxicity.100,101 Initial manifestations, such as severe nausea, vomiting, diarrhea, and abdominal cramping, begin 6 to 24 hours postingestion. Hydration and supportive care often lead to initial relief of symptoms and a relatively quiescent period. Hepatic toxicity followed by other end-organ involvement may ensue over the next several days to weeks. Progressive elevation of hepatic transaminases, jaundice, and hepatic encephalopathy can lead to death. Many cases are misdiagnosed as gastroenteritis. Numerous noninvasive therapies have been suggested, including silibinin, thioctic acid, activated charcoal, high-dose penicillin, dexamethasone, vitamin C, cytochrome c, cimetidine, Nacetylcysteine, kutkin, and aucubin.102,103 None of these therapies has been rigorously tested in human-controlled studies. Multidose administration of activated charcoal seems to be reasonable because of its ability to bind the toxins, availability, and relative safety; however, its effectiveness is unclear. Numerous invasive therapies also have been proposed for use in cyclopeptide poisoning, including forced diuresis, hemodialysis, hemoperfusion, hemofiltration, plasmapheresis, and hepatic transplantation.104–109 Similar to the noninvasive modalities, it is not clear if any of these therapies are effective. There have been several reports of successful transplantation in severe cases of poisoning107–109; however, it is uncertain what criteria should be used for selecting candidates.110,111 Patients developing severe hepatic signs and symptoms should be considered for transfer to a transplant center. Gyromitrin-containing mushrooms commonly are mistaken for edible mushrooms as they look similar to Morchella species (morel) mushrooms. The metabolites of this toxin cause GABA
Chapter 162 / Plants, Mushrooms, and Herbal Medications
GROUP/TOXIN
abuse.2,3 Ibotenic acid and muscimol are toxins that are structurally related to glutamic acid and GABA. Glutamic acid is an excitatory neurotransmitter, whereas GABA is an inhibitory one. Lethargy, hallucinations, seizures, or severe agitation begins within 1 to 2 hours after ingestion. Treatment for seizures and supportive care produce a good outcome. The cholinergic toxidrome is associated with muscarinecontaining mushrooms. Muscarine is structurally related to acetylcholine. Because of muscarine’s quaternary structure, it does not cross the blood-brain barrier. Symptoms include salivation, lacrimation, urination, defecation, gastroenteritis, and emesis (SLUDGE). Atropine can be used for severe symptoms. Pralidoxime is not indicated because acetylcholinesterase inhibition is not involved. The final mushroom group that causes early onset of symptoms occurs only with simultaneous ingestion of ethanol. These coprine-containing mushrooms cause a disulfiram-like reaction by blocking acetaldehyde dehydrogenase. The symptoms include flushing, nausea, vomiting, and headache. The onset of this reaction associated with ethanol can occur 30 minutes to several days after mushroom ingestion. Treatment is supportive and similar to disulfiram reactions from other causes.
PART IV ■ Environment and Toxicology / Section Two • Toxicology
2070
neurotransmitter depletion similar to isoniazid toxicity, leading to excitatory CNS effects, such as headaches, agitation, and seizures. Effects also include nausea, vomiting, and possible hepatotoxicity. The onset of symptoms is at least 6 hours after ingestion. Because of its similarity to isoniazid toxicity, pyridoxine has been proposed as an antidote. It is unclear how effective this antidote is, but it is useful for gyromitrin-induced CNS effects because of its availability and safety profile. Orelline/orellanine-containing mushrooms have been found in North America.112 There have been no reports of toxicity associated with cases in the United States, however, with most
reported cases in Europe. Symptoms begin 1 to 2 days after ingestion, with nausea, vomiting, abdominal pain, and headache. Renal toxicity manifests days to weeks after these initial symptoms and can progress to chronic renal failure.
Disposition Initial management is aimed at ruling out mushroom groups associated with early onset of symptoms. If the patient remains asymptomatic after a 3-hour period of observation, the patient should be discharged with instructions to return if any symptoms manifest over the next 72 hours.
KEY CONCEPTS ■
Plant exposures occur commonly in children and most commonly involve household plants. Most exposures cause little or no toxicity. ■ Plants and mushrooms are often ingested for their mindaltering properties. ■ Misidentification of plants and herbal products is a common cause for plant-induced and herbal product– induced toxicity. ■ Natural plant and mushroom gathering for personal ingestion is a popular activity. Mistakes while foraging occur commonly, with the potential for serious toxicity and numerous fatalities.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
■
Herbal medicines increasingly are being used by the general public. Limited information is available regarding the efficacy and toxicity of these products. ■ In the assessment of a patient exposed to a mushroom, the timing of initial symptoms and the assessment for associated symptoms constitute the most important data needed to make a differential diagnosis. ■ A patient who has eaten or been exposed to a wild mushroom may have another medical condition actually responsible for the symptoms.
Chapter 163
Sedative Hypnotics
Leon Gussow and Andrea Carlson
■ BARBITURATES Perspective Barbiturates are discussed in do-it-yourself suicide manuals and have been implicated in the high-profile deaths of Marilyn Monroe, Jimi Hendrix, Abbie Hoffman, Margaux Hemingway, and the mass suicide of 39 members of the Heaven’s Gate cult in 1997. Although barbiturates are still used for seizure disorders, their use as sedatives has declined significantly with the availability of safer alternatives, such as benzodiazepines, resulting in a decline from approximately 1500 barbiturate deaths per year in the 1950s to only six fatalities in 2006.1 Barbiturates are addictive, producing physical dependence and a withdrawal syndrome that can be life-threatening. While tolerance to the mood-altering effects of barbiturates develops rapidly with repeated use, tolerance to the lethal effects develops more slowly, and the risk of severe toxicity increases with continued use.
Principles of Disease Barbiturates depress the activity of all excitable cells, especially those in the central nervous system (CNS) by enhancing the activity of γ-aminobutyric acid (GABA), the major central inhibitor. In acute overdose, barbiturates decrease neural transmission in autonomic ganglia, the myocardium, and the gastrointestinal tract and also inhibit the response to acetylcholine at the neuromuscular junctions. The GABAA receptor is a protein complex found on postsynaptic membranes in the CNS. Structurally, it consists of several distinct receptor sites surrounding a chloride ion (Cl−) channel (Fig. 163-1). GABA opens the chloride channel. The resulting flow of Cl− into the cell increases the negative resting potential, hyperpolarizing and stabilizing the membrane. There are separate receptor sites for barbiturates and for benzodiazepines and a third site that binds GABA, ethanol, and meprobamate. Although barbiturates and ethanol can directly increase Cl− conductance, benzodiazepines require the presence of GABA to affect Cl− flow, which may account for the relative safety of benzodiazepines in comparison with barbiturates. Barbiturates produce dose-related depressive effects, from mild sedation to coma and fatal respiratory arrest. In the early stages of intoxication, some patients experience euphoria. Barbiturates have no analgesic effect and can paradoxically increase the reaction to pain at low doses.
Barbiturates act directly on the medulla to produce respiratory depression. In therapeutic doses, this respiratory depression mimics that of normal sleep. Starting with doses approximately three times therapeutic, the neurogenic, chemical, and hypoxic respiratory drives are progressively suppressed. Since airway reflexes are not inhibited until general anesthesia is achieved, laryngospasm can occur at low doses. Therapeutic oral doses of barbiturates produce only mild decreases in pulse and blood pressure, similar to sleep. With toxic doses, more significant hypotension occurs from direct depression of the myocardium along with pooling of blood in a dilated venous system. Peripheral vascular resistance is usually normal or increased, but barbiturates interfere with autonomic reflexes, which then do not adequately compensate for the myocardial depression and decreased venous return. Barbiturates can precipitate severe hypotension in patients whose compensatory reflexes are already maximally stimulated, such as those with heart failure or hypovolemic shock. Barbiturates also decrease cerebral blood flow and intracerebral pressure. Although hypnotic doses of barbiturates do not affect gastric emptying, higher doses can decrease gastrointestinal smooth muscle tone and peristaltic contractions and delay gastric emptying. Barbiturates are classified according to their onset and duration of action: (1) ultrashort-acting (onset immediate after intravenous dose, duration minutes), (2) short-acting (onset 10–15 minutes after oral dose, duration 6–8 hours), (3) intermediate-acting (onset 45–60 minutes, duration 10–12 hours), and (4) long-acting (onset 1 hour, duration 10–12 hours) (Box 163-1). Only long-acting preparations have anticonvulsant effects in doses that do not cause sedation. Short- and intermediate-acting preparations are almost completely metabolized to inactive metabolites in the liver, while 25% of a phenobarbital (long-acting) dose is excreted unchanged through the kidney. Because phenobarbital is a weak acid (pKa 7.2), alkalinizing the urine will increase the amount of drug present in ionized form, minimizing tubular reabsorption and increasing drug clearance. Short- and intermediate-acting barbiturates are not significantly affected by pH changes in this range. Barbiturates cross the placenta, with fetal levels approaching those of the mother. They are also excreted in low concentration in breast milk. Use during pregnancy is associated with birth defects (category D). 2071
rb ba ro, OH , ET
BA
me p
GA
PART IV ■ Environment and Toxicology / Section Two • Toxicology
2072
Chloride channel
Hypotension is common in patients with severe intoxication, along with a normal or increased heart rate. Barbiturate overdose has been associated with noncardiogenic pulmonary edema. Altered pulmonary capillary permeability can be caused by hypoperfusion, hypoxia, or a direct effect of the drug. Pneumonia may be delayed. A barbiturate withdrawal syndrome includes tremors, hallucinations, seizures, and delirium (similar to the delirium tremens of ethanol withdrawal). However, severe withdrawal occurs only following dependence on short- or intermediateacting barbiturates (e.g., pentobarbital, secobarbital, amobarbital, or butalbital). Because these drugs are not commonly used, this syndrome is now rare.
Diagnostic Strategies BZ
Figure 163-1. The γ-aminobutyric acid (GABA) receptor complex. BZ, benzodiazepine binding site; GABA, GABA binding site; ETOH, mepro, barb, binding sites for ethanol, meprobamate, and barbiturates, respectively.
BOX 163-1 Barbiturates Ultrashort-acting Methohexital (Brevital) Thiopental (Pentothal) Short- and intermediate-acting Pentobarbital (Nembutal) Secobarbital (Seconal) Amobarbital (Amytal) Aprobarbital (Alurate) Butabarbital (Butisol) Butalbital (Fiorinal) Long-acting Phenobarbital (Solfoton, Luminal) Mephobarbital (Mebaral)
Clinical Features Mild barbiturate toxicity mimics ethanol intoxication, presenting with drowsiness, slurred speech, ataxia, unsteady gait, nystagmus, emotional lability, and impaired cognition. In severe acute intoxication, CNS depression progresses from stupor to deep coma and respiratory arrest. Although pupils are usually normal or small and reactive, concomitant hypoxia can cause pupils to be fixed and dilated. Corneal and gag reflexes may be diminished or absent, muscle tone flaccid, and deep tendon reflexes diminished or absent. Flexor (decorticate) and extensor (decerebrate) posturing can occur in patients comatose from barbiturate intoxication. These neurologic signs are variable and do not always correlate with severity of intoxication or depth of coma. A fluctuating level of consciousness is commonly seen. High barbiturate levels depress gastrointestinal motility, delaying drug absorption. As the drug is metabolized and blood levels drop, peristalsis and drug absorption may increase, causing drug levels to rise. The life threat of severe barbiturate toxicity is respiratory depression. Because respirations can be rapid but shallow, the degree of hypoventilation may not be apparent on clinical examination, but pulse oximetry or capnography will detect the ventilation compromise.
The therapeutic level of phenobarbital is 15 to 40 µg/mL (65–172 µmol/L). A serum level greater than 50 µg/mL can be associated with coma, especially in a patient who is not a chronic user. Levels greater than 80 µg/mL are potentially fatal. Serial phenobarbital levels may be helpful in monitoring effectiveness of treatment. Because barbiturates other than phenobarbital have high volumes of distribution, serum levels do not accurately reflect CNS concentrations or correlate with clinical severity. A positive urine screen establishes exposure to a barbiturate but does not prove that the drug is present in toxic amounts and should not be relied upon to explain decreased mental status. Chest radiographs can detect noncardiogenic pulmonary edema or pneumonia. Computed tomography of the head should be obtained in comatose patients with evidence of trauma, focal neurologic signs, papilledema, or uncertain diagnosis. Other causes of stupor and coma must be considered and ruled out. Since the electroencephalogram may be silent as a result of barbiturate overdose, no patient should be declared “brain dead” if barbiturates are present at therapeutic levels or greater.
Management Since barbiturates have no specific antidote, management is based on supportive care, particularly with respect to the cardiovascular and respiratory systems. Severely intoxicated patients are unable to protect their airway adequately and have decreased ventilatory drive. Supplemental oxygen may suffice for patients with mild to moderate overdose, but intubation is often required. Long-term induced paralysis is rarely necessary, and additional sedation usually is unnecessary for mechanical ventilation. Careful fluid replacement should maintain a systolic blood pressure above 90 mm Hg and adequate urine output. Patients must be monitored for fluid overload and pulmonary edema. If vasopressors are necessary, dopamine is preferable to norepinephrine because of its renal vasodilating effects. Active warming should be initiated if the rectal temperature is less than 30° C.
Gastrointestinal Decontamination Gastric emptying by lavage is not indicated. For large overdoses, there is evidence that clearance of phenobarbital is markedly increased with multidose activated charcoal (MDAC).2–4 The dose of activated charcoal is 25 g every 2 hours in an adult; the pediatric dose is 0.5 g/kg every 2 hours. If vomiting occurs, a smaller dose or antiemetics should be used. MDAC can also be administered slowly through a nasogastric tube. Contraindications to MDAC include an unpro-
Disposition An asymptomatic patient who arrives in the emergency department (ED) after ingesting barbiturates should be observed until 6 hours postingestion and monitored for mental status changes, slurred speech, ataxia, hypotension, and respiratory depression. Onset of symptoms generally occurs within 1 hour of ingestion. Patients who remain asymptomatic and have no significant complicating co-ingestants or medical problems can be discharged or referred for psychiatric care. Patients who are still symptomatic 6 hours after arrival should be admitted for observation.
2073
success inspired the development of other nonbarbiturate anxiolytics. With chlordiazepoxide in 1960 and diazepam in 1963, benzodiazepines emerged as the principal agents for the treatment of anxiety. Cardiac effects and fatalities from pure benzodiazepine overdose are rare, and respiratory depression is less pronounced than with barbiturates. Additionally, drug-drug interactions involving benzodiazepines are uncommon. Benzodiazepines remain among the most widely prescribed class of drugs (Table 163-1). With nearly 50 individual agents available worldwide, they account for two thirds of all psychotropic drug prescriptions.7 Benzodiazepines are the most common prescription drugs for attempting drug-assisted suicide. Despite such frequent misuse, the vast majority of benzodiazepine overdoses follow a relatively benign clinical course. Children make up 10% of benzodiazepine overdose cases.
Principles of Disease Benzodiazepines produce sedative, hypnotic, anxiolytic, and anticonvulsant effects by enhancing the inhibitory actions of GABA. Binding of a benzodiazepine to a specific benzodiazepine receptor potentiates GABA effects on the chloride channel at the GABAA receptor, increasing intracellular flux of chloride ions and hyperpolarizing the cell. The net effect is a diminished ability of the nerve cell to initiate an action potential, resulting in inhibition of neural transmission. Three unique benzodiazepine receptors have been identified. The distribution of these receptors varies throughout the central and peripheral nervous systems. Classic benzodiazepines are nonselective, producing a broad range of clinical effects. Newer benzodiazepines interact selectively with a single receptor subtype to achieve a desired result, such as sedation, while minimizing unnecessary effects.
Pharmacokinetics ■ BENZODIAZEPINES Perspective Prior to 1950, drug options for treating anxiety were limited. While meprobamate, first synthesized in 1950, ultimately proved no safer than the barbiturates, its commercial
Benzodiazepines are rapidly absorbed orally. Intramuscular use of chlordiazepoxide and diazepam is limited by erratic absorption, but both lorazepam and midazolam are predictably absorbed after intramuscular injection. Following absorption, benzodiazepines distribute readily and penetration of the blood-brain barrier is facilitated by their highly
Table 163-1 Benzodiazepines GENERIC NAME
BRAND NAME
USUAL DOSE
ORAL PEAK (HR)
HALF-LIFE (HR)
PARENT METABOLITE ACTIVITY
Alprazolam Chlordiazepoxide Clonazepam Clorazepate Diazepam Estazolam Flurazepam Halazepam Lorazepam Midazolam Oxazepam Quazepam Temazepam Triazolam
Xanax Librium Klonopin Tranxene Valium ProSom Dalmane Paxipam Ativan Versed Serax Doral Restoril Halcion
0.25–0.5 mg 5–25 mg 0.25–0.5 mg 7.5–15 mg 2–10 mg 1–2 mg 15–30 mg 20–40 mg 0.5–2 mg 0.025–0.1 mg/kg 10–30 mg 7.5–15 mg 7.5–30 mg 0.125–0.25 mg
1–2 0.5–4 1–2 1–2 0.5–1 2 0.5–1 1–3 2–4 1–2 2–4 2 1–2 1–2
6–27 5–30 18–50 1–3 20–50 8–28 2–3 14 10–20 1.5–3 5–20 39–41 3–19 1.5–5.5
Inactive Active Inactive Active Active Inactive Active Active Inactive Active Inactive Active Inactive Inactive
Chapter 163 / Sedative Hypnotics
tected airway, so the patient is intubated before MDAC is initiated. Gastrointestinal obstruction or perforation is also a contraindication. Decreased peristalsis, such as that caused by opiate or anticholinergic overdose, is a relative contraindication to MDAC.4 Although MDAC may shorten the duration of the intoxication, there is no evidence for improved outcome over that provided by supportive management of the airway and ventilation, and hemodynamic support. Although alkalinizing the urine with NaHCO3 has been recommended in the past, a recent nonrandomized study suggested that MDAC alone is most effective at increasing the drug’s clearance.5 The authors of that study hypothesize that alkalinization may interfere with the ability of the drug to diffuse across intestinal mucosa. Hemodialysis or charcoal hemoperfusion is rarely needed but may increase clearance of phenobarbital in the presence of renal or cardiac failure, acid-base or electrolyte abnormalities, unstable cardiorespiratory status, or inadequate response to less invasive measures. Because phenobarbital is 40 to 60% protein-bound, hemoperfusion is sometimes advocated over hemodialysis; however, newer high-efficiency dialyzers using high blood flow rates may provide drug clearance greater than those achieved by hemoperfusion.6
PART IV ■ Environment and Toxicology / Section Two • Toxicology
2074
lipophilic structure. In plasma, benzodiazepines are highly protein-bound.7 Metabolism of all benzodiazepines occurs in the liver. Oxazepam, temazepam, and lorazepam are directly conjugated to an inactive, water-soluble glucuronide metabolite that is excreted by the kidney. Other benzodiazepines must first be converted by the hepatic cytochrome P-450 system. Chlordiazepoxide, diazepam, flurazepam, and clorazepate are metabolized to active derivatives that are then slowly conjugated and excreted. The long elimination half-lives of these intermediates can cause accumulation in the body with repeated dosing. Triazolam, alprazolam, and midazolam are converted to hydroxylated intermediates that, although active, are very rapidly conjugated and excreted and do not contribute significantly to the drug’s overall pharmacologic effect.7 Cytochrome P-450 processes may be significantly impaired in elderly patients or those with liver disease, leading to prolonged elimination of some benzodiazepines. Co-ingestion of drugs that also undergo cytochrome P-450 metabolism (e.g., cimetidine, ethanol) prolongs the half-lives of these benzodiazepines, but the clinical significance of these interactions is unclear.8
Clinical Features Central nervous system depression is common in patients with benzodiazepine poisoning and ranges from mild drowsiness to coma. Significant respiratory depression is rare, but can be seen with large oral overdoses or during intravenous conscious sedation, particularly when the benzodiazepine is combined with an opioid such as fentanyl.9 Hypotension is uncommon. Other complications include aspiration pneumonia and pressure necrosis of skin and muscles. The vast majority of children develop symptoms within 4 hours of benzodiazepine ingestion. Ataxia is the most common sign of toxicity, occurring in 90% of patients. In children, respiratory depression occurs in less than 10% of cases and hypotension has not been reported.
Diagnostic Strategies Any patient with altered mental status should have a blood glucose level rapidly determined. Qualitative immunoassays for benzodiazepines in urine are available but do not aid management decisions. Many of these tests detect only benzodiazepines that are metabolized to oxazepam glucuronide; therefore, clonazepam, lorazepam, midazolam, and alprazolam are not detected on a urine drug screen.10 Serum drug concentrations are not routinely available and do not correlate with clinical severity. The benzodiazepine antagonist flumazenil should not be routinely administered to patients with coma of unknown origin or suspected benzodiazepine overdose.11 Any possibility of concomitant tricyclic overdose contraindicates flumazenil use.9
Differential Considerations Benzodiazepine overdose is usually suspected or diagnosed because of the clinical presentation. Many patients are arousable and can provide supporting information. Atypical or focal findings can be clues to the presence of other conditions. Profound coma or cardiopulmonary instability with pure benzodiazepine overdose is rare, and the presence of either should prompt the search for a co-ingestant. Nontoxicologic causes of CNS depression should also be considered.
Management General Initial stabilization, including endotracheal intubation, must not be delayed by administering antidote. The vast majority of benzodiazepine overdoses can be managed expectantly. Activated charcoal is generally not beneficial in overdose.12 MDAC, hemodialysis, and whole bowel irrigation are not indicated or effective in benzodiazepine overdose.
Antidote Flumazenil, a nonspecific competitive antagonist of the benzodiazepine receptor, can reverse benzodiazepine-induced sedation after general anesthesia, procedural sedation, and overdose, but is not recommended for the reversal of benzodiazepine overdose in the ED. Although theoretical benefits of flumazenil use include cost savings and avoidance of procedures and tests such as endotracheal intubation and lumbar puncture, several studies have not been able to demonstrate an actual benefit.13 Seizures and cardiac dysrhythmias can occur with flumazenil administration, and fatalities have been reported.14–16 Flumazenil is especially hazardous when given to patients who are habituated to benzodiazepines, in whom acute benzodiazepine withdrawal, including refractory seizures, can be induced, and also when seizure-causing drugs (such as cocaine or a tricyclic antidepressant) have also been ingested, due to loss of the benzodiazepine’s protective anticonvulsant properties. Co-ingestants that cause dysrhythmias, such as carbamazepine or chloral hydrate, may increase the likelihood of cardiac effects.9 Other risk factors are summarized in Box 163-2. One study found that 12% of patients receiving flumazenil after known pure or mixed benzodiazepine overdose actually had a contraindication to its use.17
BOX 163-2 Use of Flumazenil Indications Isolated benzodiazepine overdose in nonhabituated user (e.g., accidental pediatric exposure) Reversal of conscious sedation Absolute contraindications Known or suspected co-ingestant that lowers seizure threshold Tricyclic antidepressants, cocaine, lithium, methylxanthines, isoniazid, propoxyphene, monoamine oxidase inhibitors, bupropion, diphenhydramine, carbamazepine, cyclosporine, chloral hydrate Patient taking benzodiazepine for control of a potentially life-threatening condition (e.g., seizures) Concurrent sedative-hypnotic withdrawal Seizure activity or myoclonus Hypersensitivity to flumazenil or benzodiazepines Patient with neuromuscular blockade Relative contraindications Chronic benzodiazepine user, not taking for control of life-threatening condition Known seizure disorder not treated with benzodiazepines Head injury Panic attacks Chronic alcoholism
Disposition Patients remaining asymptomatic after 4 to 6 hours of ED observation may be medically cleared. For cases of deliberate overdose, appropriate psychiatric consultation should be obtained.
Benzodiazepine Withdrawal Syndrome Abrupt discontinuation of a benzodiazepine in a chronic user results in a characteristic constellation of symptoms (Box 163-3). Risk for withdrawal is a function of both the dose of benzodiazepine and the duration of its use. Continuous treatment for more than 4 months is generally required before a patient is at risk for withdrawal. With abrupt discontinuation of a benzodiazepine, the most severe withdrawal symptoms are expected within several days to a week.18 Use of flumazenil can precipitate immediate withdrawal symptoms. Treatment of withdrawal consists of restarting benzodiazepines.
■ FLUNITRAZEPAM Flunitrazepam (Rohypnol) has been used in Europe, Asia, and Latin America for insomnia and preoperative sedation since 1975. Although never manufactured or sold in the United States, flunitrazepam has been implicated in many reports of “date rape” incidents. Flunitrazepam has been an active agent in the illicit drug market, where it is used to alter the effects of other drugs, including heroin and cocaine.19 Flunitrazepam has 10 times more affinity than diazepam for certain benzodiazepine receptors. Onset of CNS depression occurs within 30 minutes. The drug is most frequently ingested with alcohol, producing disinhibition and amnesia. Despite marked CNS depression, patients can usually be aroused with noxious stimuli. The half-life of the drug is 16 to 35 hours, but coma can be prolonged for up to 48 hours.19 Flunitrazepam is easily obtainable outside the United States. The drug is not
BOX 163-3 Benzodiazepine Withdrawal Symptoms Nonspecific Anxiety, depression, insomnia, tremor, tachycardia, sweating Severe (rare) Visual hallucinations, delirium, seizures
2075
detected on routine urine drug screens but, if needed as evidence, urine should be collected and refrigerated or frozen and the local or state police crime laboratory contacted to arrange specific testing. Metabolites of flunitrazepam are detectable in the urine up to 72 hours after exposure.20
■ BUSPIRONE Buspirone (BuSpar) has been used for generalized anxiety since 1986. Unlike benzodiazepines, buspirone does not have any effect on GABA. Rather, it acts as a partial serotonin (5HT1A) agonist. To a lesser extent, it also antagonizes dopamine (D2) receptors. Buspirone has no hypnotic, anticonvulsant, or muscle relaxant effects. Buspirone has several advantages over benzodiazepines. The drug causes minimal CNS depression, even in combination with ethanol. Dosage adjustment is not needed for elderly patients. Chronic administration does not cause tolerance, and dependence does not occur. A withdrawal state after discontinuation has not been reported. Only one case of isolated buspirone overdose has been published. That patient was lethargic and had a tonic-clonic seizure but recovered fully.21
■ ZOLPIDEM AND ZALEPLON Zolpidem (Ambien) and zaleplon (Sonata) differ in structure from both the benzodiazepines and buspirone, and neither is detected on a benzodiazepine toxicology screen. They act selectively at a specific benzodiazepine receptor, producing sedation without many of the side effects seen with benzodiazepines. They have modest anxiolytic, muscle relaxant, and anticonvulsant properties. Significant drug interactions are rare. Compared with zolpidem, zaleplon causes less memory loss and sedation at therapeutic doses and is more rapidly eliminated.22 Transient visual disturbances and hallucinations can occur in patients with normal levels of consciousness with both zolpidem and zaleplon.22,23 Abuse of zolpidem is limited by vomiting, which may occur after a supratherapeutic dose. Both zolpidem and zaleplon are rapidly eliminated and lack active metabolites.24 Patients with zolpidem overdose do well with supportive care alone. Fatalities from isolated zolpidem overdose are rare. All published cases involve individuals found dead at home and are often associated with co-ingestants, particularly other sedative hypnotics or antipsychotics.25 Drowsiness is by far the most common symptom. Coma and respiratory failure are rare, despite overdoses of up to 40 times the normal dose, although intubation may be required, particularly if there are co-ingestants.26 Zolpidem overdose in children follows a similarly benign course. Drowsiness, ataxia, and hallucinations resolve within 10 hours.27 Overdose information for zaleplon is limited. In one case series, patients had CNS depression and mild hypotension. Arousal was temporally associated with flumazenil use in one patient.28 The only published fatality involves a mixed drug overdose with unknown quantities of zaleplon and butalbital, with postmortem serum zaleplon concentration 40 times greater than therapeutic.29 Adverse effects with therapeutic use include headache, anterograde amnesia, and transient visual hallucinations.24
■ ESZOPICLONE Eszopiclone (Lunesta) has been marketed in the United States since 2005 for treating insomnia. It is the S-isomer of racemic zopiclone, which has been used for decades outside the United
Chapter 163 / Sedative Hypnotics
The initial adult dose of flumazenil is 0.2 mg given intravenously over 30 seconds. A second dose of 0.3 mg may be given, followed by 0.5-mg doses at 1-minute intervals, to a total of 3 mg. Most patients respond within 3 mg. In children, the initial dose is 0.01 mg/kg (up to 0.2 mg). Because the duration of action of flumazenil is short (0.7–1.3 hours), resedation occurs in up to 65% of patients and requires either redosing or continuous infusion (0.25–1.0 mg/hr). In summary, benzodiazepine overdose requires only supportive care (including, in some cases, intubation). Flumazenil may precipitate seizures or acute withdrawal. It should be used only in highly selected cases, such as small children with accidental poisoning or for reversal of accidental overdose of benzodiazepines during procedural sedation. When flumazenil is used, careful monitoring is necessary because of the risk for persistent respiratory depression or resedation. Use of flumazenil has not consistently altered outcome, complication rate, number of costly procedures performed, or duration of hospital stay in ED patients.17
PART IV ■ Environment and Toxicology / Section Two • Toxicology
2076
States. Eszopiclone has a structure unrelated to those of benzodiazepines, barbiturates, zolpidem, and zaleplon.30–34 The mechanism of eszopiclone’s action is not completely described but may involve a specific GABAA receptor close to or coupled with the benzodiazepine receptor.30–33 Eszopiclone is rapidly absorbed, with a peak serum level at 1 hour and a half-life of 6 hours. It is metabolized in the liver to minimally active metabolites. The usual bedtime dose is 3 mg. It is recommended that elderly patients and those with hepatic insufficiency be treated with a lower (1 mg) dose. Adverse effects seen with therapeutic use of eszopiclone include drowsiness, dizziness, dry mouth, unpleasant taste, nausea, and vomiting. Auditory and visual hallucinations have been reported.35 Experience with eszopiclone overdose is limited, but treatment is supportive. A retrospective case review described 525 eszopiclone ingestions, but 259 of these patients had also ingested other drugs or chemicals.32 The ingestions involved eszopiclone doses up to 210 mg and presented with mild to moderate symptoms at most. Two deaths occurred, both involving significant co-ingestants. A single case report described a 52-year-old man who developed coronary vasospasm and a ventricular fibrillation arrest after ingesting 45 to 60 mg of eszopiclone.34 However, the arrest occurred approximately 20 hours after ingestion, and it is unclear what role, if any, eszopiclone played in causing the arrest.
■ CHLORAL HYDRATE Perspective Deaths related to chloral hydrate overdose were first reported in the medical literature in 1890. Chloral hydrate has a low therapeutic ratio and can produce significant, potentially fatal, toxicity. While chloral hydrate use is rare today, it is still occasionally prescribed as a sedative in the elderly and for sedation in children undergoing medical procedures. The hypnotic oral adult dose is 0.5 to 1.0 g. The toxic oral dose in adults is approximately 10 g and may be as little as 1.5 g in a child.36 The toxic effects of chloral hydrate include CNS depression, gastrointestinal irritation, cardiovascular instability, hepatitis, and proteinuria. The primary active metabolite of chloral hydrate, trichloroethanol, has a barbiturate-like effect on GABAA receptors and is responsible for most of the CNS depression seen with significant overdose. Chloral hydrate is rapidly absorbed from the gastrointestinal tract and almost immediately metabolized to trichloroethanol by the enzyme alcohol dehydrogenase. Onset of action is 20 to 30 minutes.37 Trichloroethanol is long-acting, with a half-life that can be significantly prolonged after overdose as metabolic pathways become saturated. The combination of chloral hydrate and ethanol (the “Mickey Finn”) potentiate each other’s action to produce rapid loss of consciousness. Chloral hydrate increases the halflife of ethanol by competitively inhibiting the enzyme alcohol dehydrogenase, and the metabolism of ethanol generates NADH, a cofactor for the conversion of chloral hydrate to trichloroethanol.
Clinical Features Chloral hydrate toxicity causes CNS and respiratory depression, gastrointestinal irritation, cardiovascular instability, and dysrhythmias. The combination of deep coma and cardiac dysrhythmia without hypoxia is characteristic of severe cases. Mild chloral hydrate toxicity can mimic ethanol or barbitur ates, with drowsiness, ataxia, and lethargy. A pear-like odor to the patient’s breath or gastric contents may suggest the
diagnosis. More severe toxicity includes miosis, muscle flaccidity, diminished deep tendon reflexes, hypoventilation, hypotension, and hypothermia.36 Chloral hydrate is corrosive and causes nausea, vomiting, esophagitis, hemorrhagic gastritis and, more rarely, gastrointestinal perforation or necrosis.36 Transient hepatic or renal dysfunction can also occur. Dysrhythmias from chloral hydrate toxicity can be fatal. Chloral hydrate decreases myocardial contractility, shortens the cardiac refractory period, and increases the sensitivity of myocardium to catecholamines. Dysrhythmias include atrial fibrillation, supraventricular tachycardia, ventricular tachycardia, multifocal premature ventricular contractions, torsades de pointes, ventricular fibrillation, and asystole.37 Hypotension results from inhibition of central neurovascular regulatory centers as well as impaired myocardial contractility.
■ MANAGEMENT FOR SEDATIVES The key to management for all of these agents is support of cardiorespiratory function. Intubation may be required for airway protection or to support ventilation and oxygenation. Avoid naloxone or flumazenil, which may precipitate ventricular dysrhythmias.36 Because chloral hydrate sensitizes myocardium to catecholamines, epinephrine and norepinephrine should also be avoided. Standard antidysrhythmic agents such as lidocaine do not appear effective against chloral hydrate– induced cardiac ectopy. The treatment of choice is a betablocker.38 Intravenous propranolol can be given in adult doses of 0.5 mg until ectopy is suppressed, followed by an infusion of 1 to 2 mg/hr, titrated to a heart rate of 80 to 100 beats per minute. A short-acting agent such as esmolol can also be used. Torsades de pointes should be treated with intravenous magnesium or overdrive pacing. Type I antidysrhythmic agents such as quinidine should be avoided. Unstable patients not responding to conservative therapy can be treated with hemoperfusion or hemodialysis.36
■ OVER-THE-COUNTER SLEEP AIDS Perspective In the past, most over-the-counter (OTC) sleep aids contained a combination of an antihistamine (either methapyrilene or pyrilamine) and scopolamine. Some preparations also contained a bromide. For safety reasons, these products were reformulated in the late 1980s to contain diphenhydramine or doxylamine, now the only two drugs found in nonprescription hypnotics. Many preparations also contain acetaminophen or aspirin, added to create a nighttime pain reliever (Table 163-2). The availability and frequent use of these agents may explain why overdose is so common.
Principles of Disease Diphenhydramine and doxylamine are antihistamines that also have hypnotic, anticholinergic, and weak local anesthetic properties. They act as competitive antagonists of H1 histamine receptors and cause sedation by inhibiting the actions of acetylcholine on muscarinic receptors in the CNS. The pharmacokinetic profiles of diphenhydramine and doxylamine are similar. Both are rapidly absorbed, with peak plasma levels occurring at 1 to 2 hours after administration. In the systemic circulation, they are highly protein-bound, with large volumes of distribution. Extensive metabolism occurs in the liver by the cytochrome P-450 system. The elimination half-life is 4 hours for diphenhydramine and 9 hours for doxylamine.
2077
BRAND NAME
ACTIVE INGREDIENT(S)
Alka-Seltzer PM effervescent tablets Bayer PM Aspirin Plus Sleep Aid caplets
38 mg diphenhydramine citrate 325 mg aspirin 25 mg diphenhydramine hydrochloride 500 mg aspirin 38 mg diphenhydramine citrate 500 mg acetaminophen 38 mg diphenhydramine citrate 500 mg acetaminophen 50 mg diphenhydramine hydrochloride 25 mg diphenhydramine hydrochloride 25 mg diphenhydramine hydrochloride 50 mg diphenhydramine hydrochloride 25 mg diphenhydramine hydrochloride 50 mg diphenhydramine hydrochloride 25 mg doxylamine succinate
Excedrin PM tablets, caplets or geltabs Goody’s PM powder Maximum Strength Nytol QuickGels softgels Nytol QuickCaps caplets Simply Sleep caplets Sleepinal Night-Time Sleep Aid capsules and softgels Sominex Night-Time Sleep Aid tablets Unisom Maximum Strength SleepGels Unisom SleepTabs tablets
Clinical Features Impaired consciousness is the most frequent finding with diphenhydramine overdose. Somnolence, psychotic behavior, and agitation are common.39 Anticholinergic effects may be apparent, as noted in Chapter 148. Apart from a lower incidence of psychosis, doxylamine has toxicity similar to that of diphenhydramine.40 Seizures and rhabdomyolysis may occur with severe toxicity. Serious cardiotoxicity is rare.41
Diagnostic Strategies Some comprehensive urine drug immunoassays will detect diphenhydramine. Quantitative serum levels of diphenhydra mine or doxylamine are neither routinely available nor clinically useful. Serum acetaminophen and salicylate concentrations should be measured in patients with OTC sleep aid overdoses, because many preparations contain both a hypnotic and an analgesic. Measuring serum creatine phosphokinase and urinary myoglobin may help detect myoglobinuria.
Management Management of mild to moderate toxicity from OTC sleep aid overdose is generally supportive. Specific details regarding anticholinergic toxicity are discussed in Chapter 148.
Disposition Patients with minor sedation or anticholinergic effects that are resolving or who remain asymptomatic or are minimally symptomatic after a 4-hour observation period can be medically cleared. If the ingestion was in the context of self-harm, psychiatric evaluation is indicated. Other patients require inpatient observation in a monitored setting.
BOX 163-4 g -Hydroxybutyrate Street Names GHB Grievous bodily harm GBH Georgia home boy Gib Natural sleep-500 Gamma-OH Gamma hydrate Liquid X Organic quaalude Liquid E
Liquid ecstasy Liquid G Somatomax Soap Salty water Scoop Sodium oxybate Easy lay Cherry menth Fantasy G-Riffick
■ g-HYDROXYBUTYRATE Perspective Originally synthesized in the 1960s as an anesthetic in Europe and Japan, researchers later discovered that γ-hydroxybutyrate (GHB) was a naturally occurring metabolite of GABA. Since 1970, GHB has been used to treat narcolepsy and alcohol addiction, as well as alcohol and opiate withdrawal.42 A 1977 report that GHB may enhance the effects of steroids and the release of growth hormone resulted in marketing of the agent as a natural aid for increasing muscle mass. Numerous reports of adverse effects followed. In 1989, the U.S. Food and Drug Administration called for a voluntary withdrawal of the drug from store shelves. Its sale and manufacture were banned in 1990; however, illicit use of GHB increased along with the emergence of GHB precursors, γ-butyrolactone (GBL) and 1,4-butanediol (1,4-BD) (Box 163-4).43 The Hillory J. Farias and Samantha Reid Date-Rape Drug Prohibition Act of 2000 made GHB a schedule I controlled substance. GHB has since been approved for the treatment of narcolepsy, under the trade name Xyrem (sodium oxybate, 0.5 mg/mL)44 as a schedule III drug. GHB remains a popular drug of abuse.45 Recipes for home synthesis are widely available. Some individuals take GHB for its purported muscle-building and fat-burning actions, others for its psychoactive effects. The drug’s euphoria-producing properties make it popular at “raves” (large, crowded youth parties with energetic dancing to rhythmic music for many hours).46–48 Self-treatment of insomnia with GHB has been reported and can cause dependence.49 CNS depression, amnesia, and disinhibition with GHB, especially mixed with ethanol, make this combination a potential agent in “date rape” situations.50,51 Chemical precursors to GHB are also commonly abused. GBL is rapidly converted to GHB by plasma lactonases. 1,4BD is metabolized to γ-hydroxybutyraldehyde by the enzyme alcohol dehydrogenase, and then to GHB by aldehyde dehydrogenase.52,53
Principles of Disease GHB binds to specific GHB receptors and at high concentrations to GABAB receptors.54 The complex interaction between these two receptors may explain the sometimes paradoxical manifestations of GHB toxicity of somnolence and agitation. Through its action on the GABAB receptor, GHB decreases55,56 release of dopamine.42 As underground laboratories often synthesize liquid GHB by mixing and heating butyrolactone and sodium hydroxide, careless preparation can result in residual unreacted base, causing significant caustic injury when the liquid is ingested.43
Chapter 163 / Sedative Hypnotics
Sedative Hypnotics, Table 163-2 Nonprescription United States
PART IV ■ Environment and Toxicology / Section Two • Toxicology
2078
GHB is liphophilic and rapidly absorbed. Onset of symptoms occurs within 15 to 30 minutes and peak plasma levels within 20 to 60 minutes.49 Unlike GABA, it readily crosses the blood-brain barrier. The half-life of GHB is 27 minutes but may increase at high doses. GBL is an industrial solvent that is rapidly absorbed after ingestion and metabolized within minutes to GHB by peripheral and hepatic lactonases.43 Before conversion to GHB, GBL itself is inactive and has no sedating effects.42 It produces a clinical syndrome similar to that of GHB ingestion, but its effects are greater and more prolonged. In fact, GBL is more efficient at delivering GHB to the CNS than GHB itself.57 GBL is available under a number of street names (Box 163-5). 1,4-Butanediol (1,4-BD) is converted after ingestion to GHB by the enzyme alcohol dehydrogenase.53 Like GBL, it is used as an industrial solvent. Unlike GBL, 1,4-BD itself has sedative-hypnotic effects. Clinical findings are similar to GHB. When 1,4-BD and ethanol are ingested together, ethanol acts as a competitive inhibitor of alcohol dehydrogenase, so the toxic effects of 1,4-BD are delayed and prolonged, and the risk of death is increased.53 1,4-BD is available under a number of street names (Box 163-6). In 2007, a children’s toy marketed under the names Aqua Dots and Bindeez Beads was contaminated when 1,4-BD was substituted for a more expensive industrial solvent during the manufacturing process. The toy consisted of tiny brightly colored spheres that were readily ingested by toddlers,
BOX 163-5 g-Butyrolactone Street Names Blue Nitro Blue Nitro Vitality Enliven Fire Water Gamma G GH Revitalizer GHRE (growth hormone release extract) Nitro NRG3 Remforce RenewTrient Revitalize Plus Revivarant SomatoPro Verve 5.0
BOX 163-6 1,4-Butanediol Street Names Blue Raine Dream On Fubar Inner G Pine needle extract Pine needle oil ReJuv@Nite Revitalize Plus Serenity SomatoPro Thunder nectar Weight belt cleaner Zen
causing decreased levels of consciousness, coma, or apparent seizures.58
Clinical Features Diagnosis of GHB intoxication is based on the history and clinical course. Rapid recovery from coma, or periods of agitation alternating with periods of decreased level of consciousness, is characteristic. Signs and symptoms are generally consistent with poisoning by other sedative-hypnotic agents. Hypothermia may occur.47 In the presence of coma, bradycardia with or without hypotension may be seen and occasionally responds to stimulation alone.47 Eye examination may reveal miosis with or without nystagmus. Behavioral changes are most common and range from aggression and delirium to coma.47 A distinctive feature of GHB intoxication is respiratory depression with apnea, interrupted by periods of agitation and combativeness, especially stimulated by attempts at intubation that do not use RSI drugs. Emesis occurs in 50% of cases.47 Generalized seizures may actually represent random myoclonic movements of the face and extremities. The dose-response curve of GHB is steep. An oral dose of 10 mg/kg results in hypotonia and amnesia, whereas 25 mg/kg induces sleep. A dose of 50 to 60 mg/kg produces anesthesia, and higher doses may cause coma associated with bradycardia, respiratory depression, vomiting, and myoclonic activity. The severity is also dependent on the dose and the concurrent use of alcohol or other psychoactive drugs.49
Diagnostic Strategies GHB is not detected on most urine toxicology screens. If laboratory confirmation is required, specimens must be collected early to capture the parent compound, and gas chromatography-mass spectroscopy must be performed. The drug may be detected in urine up to 12 hours after ingestion.59 Poisoning with another sedative hypnotic can produce a similar clinical picture to that seen with GHB. Unique to GHB, however, is the relatively rapid resolution of symptoms. In the absence of a co-ingestant such as ethanol, most patients will awaken within 3 to 4 hours. Nearly all patients recover fully within 8 hours. Prolonged coma should prompt a search for another cause. Cardiac effects and refractory seizures are rare and suggest the presence of other agents.
Management Because of the high incidence of emesis with GHB overdose, intubation for airway protection should be seriously considered in patients with significant CNS depression. In the absence of an identified difficult airway, rapid sequence intubation is the method of choice. Bradycardia unresponsive to stimulation can be treated with atropine. Treatment of isolated GHB ingestion is supportive. Patients should be protected from self-injury until resolution of symptoms. Physostigmine had been used as an antidote for GHB when used as an anesthetic agent, but the use of physostigmine is not generally recommended.52,60
Withdrawal Similar to other sedatives and hypnotics, patients who suddenly stop GHB or its precursors after chronic, frequent use can experience a severe and potentially life-threatening withdrawal syndrome.42,61,62 Because of the short half-life of GHB, symptoms of withdrawal usually begin within several hours of
Disposition Because of GHB’s short half-life, symptoms often resolve while the patient is still in the ED. The patient generally regains consciousness spontaneously. No delayed toxicity is expected. Patients should be counseled about the seriousness of GHB intoxication.
2079
KEY CONCEPTS ■
Most patients with barbiturate intoxication will recover with observation and meticulous supportive care. The vast majority will not require gastric lavage, hemodialysis, or hemoperfusion. ■ A urine toxicology screen positive for barbiturates does not prove that the patient’s clinical condition is caused by the drug. Quantitative levels confirm the diagnosis. Serum barbiturate levels do not necessarily correlate with depth of coma or clinical outcome. ■ Flumazenil should not be used in most benzodiazepine overdoses, because it can precipitate seizures. If it is used, patients must be monitored closely for persistent respiratory depression and for resedation after its use. ■ Chloral hydrate–induced dysrhythmia should be treated with beta-blockade. ■ Endotracheal intubation to protect against emesis and respiratory depression should be strongly considered for patients with GHB overdose and significant CNS depression. ■ Withdrawal from GHB or its precursors can manifest with anxiety, tremor, and insomnia but can progress to a severe syndrome characterized by delirium and autonomic instability. Management of this syndrome often requires high-dose benzodiazepines or barbiturates.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 163 / Sedative Hypnotics
the last dose. The typical patient will have been using these products for weeks or years, every 1 to 3 hours around the clock to avoid withdrawal symptoms. Mild withdrawal presents with anxiety, tremor, and insomnia. This can progress to confusion, delirium, overt psychosis, paranoid ideation, hallucinations (visual, aural, and/or tactile), and autonomic instability. Diagnosis relies on a history of symptoms beginning after abruptly ceasing use of these products. The differential diagnosis includes withdrawal from other sedatives or hypnotics, delirium tremens, sympathomimetic toxicity, serotonin syndrome, neuroleptic malignant syndrome, CNS infection, and thyroid storm. Initial treatment usually begins with high-dose benzodiazepines. However, GHB withdrawal may involve depleted levels of GABA.61 Since the effect of benzodiazepines requires the presence of GABA, they may not be effective in controlling GHB withdrawal. Barbiturates, such as pentobarbital, which do not need GABA to be effective, are often required in cases of severe intoxication.63 These patients often require intensive care admission for high-dose sedatives to manage agitation and to monitor fluctuating vital signs. Rhabdomyolysis and severe hyperthermia should be ruled out. Deaths have been reported, sometimes many days after presentation and after apparent improvement.61
PART V
Special Populations
•
Section One The Pediatric Patient Chapter 164
General Approach to the Pediatric Patient
Robert A. Wiebe and Susan M. Scott
■ PERSPECTIVE Background Assessment of pediatric patients from the newly born through adolescence offers unique and varied challenges to the emergency care provider. Approximately 30% of all visits to a general emergency department (ED) are for issues related to pediatric patients. The vast majority of children in crisis are not seen in pediatric specialty hospitals, but in community hospital EDs. Although most pediatric visits to the ED are not serious, it is not uncommon to see true emergencies in infants and young children. Serious and life-threatening pediatric emergencies result from a wide variety of causes and require the health care provider to understand the unique anatomic, physiologic, immunologic, and developmental differences that make serious problems often difficult to recognize and the differential diagnosis dependent on the age of the patient. Emergency care does not stop or start in the ED. An integrated emergency medical system that is capable of responding to the needs of children is a critical part of pediatric emergency care. Likewise, many critically ill or injured children cannot receive definitive care in small community hospitals. Therefore, a support network that includes interfacility transport resources and definitive care for pediatric patients must be part of an integrated system of emergency care for children. This chapter focuses on the role of the emergency physician in recognizing and assessing children needing emergency care.
Epidemiology According to the National Center for Health Statistics, there were approximately 115,300,000 ED visits in the year 2003. In 2005, 20.5% of ED visits were for patients less than 18 years of age, and 28.2% of all children under 6 years of age in the United States had at least one visit to an ED in the past 12 months.1 The National Hospital Ambulatory Medical Care Survey in 2005 demonstrated that the age group with the highest annual per capita ED visit rate was infants under 12 months of age, with 91.3 visits per 100 infants.2 A recent survey revealed that pediatric patients were seen in an adult facility 89% of the time, with only 4% of departments having a separate area designated for the care of children.3 The Emergency Pediatric Services and Equipment Supplement to the National Hospital Ambulatory Medical Care Survey revealed that 23% of EDs have a Pediatric Emergency Medicine specialist available, 62% have pediatric attending physicians available, and 71% have
emergency medicine physicians on staff.4 Only 6% of EDs have all the supplies as recommended by the joint American Academy of Pediatrics/American College of Emergency Physicians policy, “Care of Children in the Emergency Department: Guidelines for Preparedness.”5,6 Fifty percent of departments have at least 85% of the recommended supplies. Data continue to suggest that health insurance status is not a significant cause of ED overcrowding.7 Access to pediatric care is associated with a marked decrease in ED utilization regardless of insurance status, especially among the uninsured. The convenience of the ED, access without appointment, and lack of understanding by parents of the meaning of an emergency are the major factors related to pediatric nonurgent visits.8 Respiratory emergencies and trauma are the most common reasons for visits to an ED. Table 164-1 provides a list of the most common problems presenting to a pediatric ED. The most common diagnoses include acute respiratory infection, fever, otitis media and other head and neck infections, enteritis, and minor cuts and contusions. Injury is the most common cause of serious morbidity and mortality in children younger than 15 years and is responsible for 14.1 ED visits per 100 persons per year. In the pediatric age group, approximately 95% of these visits are for unintentional injuries that are largely predictable and preventable. Although the reasons for pediatric ED visits are many and varied, the care of critically ill or injured children should always focus on two physiologic events: shock and respiratory failure.
■ PRINCIPLES OF DISEASE Pathophysiology Anatomic and Physiologic Differences Physical assessment of a pediatric patient requires attention to a variety of anatomic, physiologic, and developmental differences that vary with the age of the patient. It is sometimes difficult to separate anatomic and physiologic issues; for example, the large surface area-to-weight ratio in young infants can result in heat loss and temperature instability. Thus, it is important to maintain a neutral thermal environment during the physical assessment and stabilization process. The relatively large head-to-body ratio and the small, weak neck of infants and young children make them particularly prone to head injuries. Blunt trauma to the chest and abdomen often results in injury to internal organs with minimal or no 2083
PART V ■ Special Populations / Section One • The Pediatric Patient
2084
Pediatric Complaints from Emergency Table 164-1 Chief Department Triage COMPLAINT
Fever Trauma/injury Breathing difficulty Upper respiratory symptom Vomiting/diarrhea Abdominal pain Rash Ear complaint Genitourinary problem Soft tissue infection Seizure Eye complaint Sore throat Oral/dental complaint Headache Irritability Chest pain Foreign body Behavior issue Sexual abuse suspected Total visits to triage Top 20 visits
NO. OF PATIENTS
19,754 11,650 9979 9004 7657 4460 4287 3428 3235 1754 1745 1712 1562 1236 1229 780 707 547 392 354 102,453
PERCENTAGE OF TOTAL VISITS
19.3 11.4 9.7 8.8 7.5 4.4 3.3 3.3 3.2 1.7 1.7 1.7 1.5 1.2 1.2 0.8 0.7 0.5 0.4 0.3 82.6%
From Children’s Medical Center of Dallas: 2003 ED Triage Statistics.
external signs of trauma. The elasticity of growing bones creates unique problems in pediatrics. Soft and pliable growing ribs will bend rather than break and transmit the forces of blunt trauma to the thoracic and upper abdominal organs. The weakest part of growing bones is the physeal plate or growth plate, and this area is injured more frequently than the surrounding ligaments. In a growing child, sprains are uncommon and physeal fractures result in nearly 20% of pediatric fractures. Recognition of growth plate injuries in children is critical to avoid imbalance in bone growth. Anatomic differences between the pediatric and adult airway are important to understand for appropriate assessment and emergency support of the airway when required. The small airways of infants and young children are more prone to obstruction from secretions, which can result in relatively rapid deterioration ranging from respiratory distress to failure. Simple maneuvers such as deep suction of the upper airway can frequently have dramatic results in improving air movement in small infants. Because infants are often preferential nose breathers, nasal obstruction from secretions can result in significant airway compromise. An irritable and crying infant may just be learning to mouth breath when the nose is obstructed. Youth also has its distinct advantages. Compensatory mechanisms to protect the infant or young child from decompensated shock, including increasing heart rate and shunting blood from the peripheral to more needed central circulation, are more effective in protecting the pediatric patient from serious consequences from shock.
Vital Signs Table 164-2 provides normal ranges of vital signs. It is imperative for the provider to be familiar with these different values
Table 164-2 Normal Pediatric Vital Signs for Age* AGE (YR)
RESPIRATORY RATE (BREATHS/MIN)
12 12–16 Lower limits of systolic blood pressure†: 0–28 days: 60 mm Hg 1–12 months: 70 mm Hg 1–10 years: 70 mm Hg + (2 × age in years)
HEART RATE (BEATS/MIN)
100–160 90–150 80–140 70–120 60–100
*From Dieckmann R, Brownstein D, Gausche-Hill M (eds): Pediatric Education for Prehospital Professionals. Sudbury, Mass, Jones & Bartlett, American Academy of Pediatrics, 2000, pp 43–45. † From American Heart Association ECC Guidelines, 2000.
as they vary with age. Recognition and explanation for abnormal vital signs is one of the keys to success in the treatment of the ill patient. Normal heart rate varies with age. Tachycardia can be a product of fever, anxiety, pain, or fear but is also the first and most sensitive sign of cardiovascular compromise in the pediatric patient. When measuring the heart rate, the quality of the pulse can be extremely helpful, as can a comparison of the strength of the central and peripheral pulses in the same extremity. The quality of the brachial and radial pulses or the femoral and dorsalis pedis pulses palpated concurrently provides important information to differentiate cardiovascular compromise from benign causes of tachycardia. Bradycardia can be an ominous sign in the ill patient heralding cardiopulmonary failure and impending cardiac arrest. Blood pressure measurement is a key component of the assessment of cardiovascular function and should be obtained in ill patients of all ages. Infants and young children have excellent compensatory measures for maintaining blood pressure in the presence of significant loss of circulatory volume. Compensatory mechanisms include an increase in heart rate and systemic vascular resistance. When these compensatory mechanisms fail and blood pressure drops below normal, the patient moves from a state of compensated to decompensated shock. Obtaining an accurate blood pressure in infants and small children can be challenging. The most common hurdle is the lack of cooperation by infants and small children. Another hurdle is the appropriate selection of the blood pressure cuff. The properly sized cuff should cover approximately two thirds of the circumference of the upper arm and extend at least 50% of the length of the upper arm. The lower limit for acceptable blood pressure in children older than 1 year can be quickly estimated by using the following formula: systolic blood pressure (mm Hg) = 70 + (2 × age [in years]). The pulse oximetry waveform can also be used to determine systolic blood pressure. Observing for the return of a plethysmographic waveform of the pulse oximeter as the blood pressure cuff is deflated has been shown to correlate closely with conventional methods of blood pressure measurement.9 Whenever possible, blood pressure should be obtained in all patients. An alert and crying infant with good peripheral pulses and normal mental status can be assumed to have adequate blood pressure, but this assumption can be misleading, and obtaining a blood pressure measurement in patients of all ages should be attempted. As with the heart rate, respiratory rate varies with the age of the patient. Respiratory rate alone cannot be used to deter-
Developmental Issues Knowledge of basic behavioral and developmental differences by age is important when assessing a pediatric patient. Table 164-3 summarizes age-related pediatric differences in motor function, problem-solving, language, and social/adaptive milestones during the first 2 years.
Neonates During the neonatal period and early infancy, normal behavior consists of sleeping, feeding, and crying when hungry or experiencing discomfort. There is little or no eye contact and no social smile. Discomfort is usually nonspecific, and the cause of the irritability or crying may be difficult to interpret. It is essential in the assessment of the infant for the provider to listen to the parents and their concerns. A mother’s “sense” of her child is often accurate and should be considered seriously.
Infants (12 Months or Younger) By 2 to 3 months an infant has a social smile and responds to a friendly voice. Lack of appropriate social interaction can be worrisome. An infant with a glassy-eyed, “nobody home” stare can be easily distinguished from a normal infant who tracks lights or has a smiling face. Infants at this age have little or no understanding of language, but they will certainly respond to a calm and soothing voice. Of concern is a 6-month-old who does not acknowledge your presence. Normal behavior for this age includes any expression of curiosity or anxiety, such as crying. Children older than 6 months should be approached with caution. Infants at this age will cry when taken from the safety of their caregiver’s arms. The emergency physician should anticipate stranger anxiety and examine the patient, whenever possible, in the lap of the caregiver. Distractions such as toys and penlights are useful to provide emotional control of the infant during the examination.
Toddlers (13 Months to 3 Years of Age) During the toddler age, as language develops, it is important to talk directly to the child. It is important to realize that toddlers and preschoolers have more extensive receptive language than expressive language and can pick up fears and concerns
2085
from their parents and caregivers, making them apprehensive and fearful. If the patient appears stable, sitting and talking or playing with the patient first rather than performing the examination immediately upon entering the examination room can help establish a rapport. Praise and reassurance during the examination can go a long way in maintaining this rapport with the child.
Preschool Children (4 to 5 Years of Age) Preschool children often fantasize, and such fantasies may result in irrational assumptions and even nightmares. The child may sense the fears and concerns of their caregivers and this may contribute to these assumptions. The cause of injuries or illnesses in preschoolers or loved ones may be misinterpreted as being a result of their own misbehavior. The provider can expedite the examination in a toddler by allowing the child to control some part of his or her experience, such as picking which ear the provider examines first.
School-Age Children (6 to 12 Years of Age) As children reach school age, it becomes particularly important to explain procedures, answer questions, and address fears and concerns honestly. Privacy and modesty should be respected. Whenever possible, the child should be included in conversations, and historical information should be taken from both the child and parent. The powers of reasoning begin to mature, and a school-age child will often attempt to negotiate control over having painful or distasteful procedures performed. Choices and behavior limits should be given only when they do not compromise care. A child may be given some autonomy by allowing them to make choices in their care such as the choice to have blood drawn from the left arm or the right arm or be told that it is okay to cry but that it is important to keep the arm still.
Adolescents (13 to 19 Years of Age) With adolescence comes independence and autonomy. Peer pressure becomes far more important than the behavior boundaries provided by the caregiver. Adolescents are risk takers and often have no fear of danger or injury. They rarely anticipate consequences and may lack common sense. Privacy and confidentiality should always be respected, and it is wise to separate adolescents from the caregiver when obtaining information and performing the physical examination. Adolescents should be given the opportunity to express their opinions and concerns about their own medical care.
■ CLINICAL FEATURES Recognition of the pediatric patient at risk for decompensation and deterioration is a challenge to the health care provider forced to multitask in a busy ED. Despite the advanced technology available to the provider, the bedside evaluation of the patient is the key component to accurate evaluation and management. The initial assessment of the patient begins with the providers’ examination of respiratory effort and circulation of the skin concurrently with their sense of the child’s acuity, that is, “sick, not sick.”
Pediatric Assessment Triangle The pediatric assessment triangle (PAT) offers a sensible, orderly approach that can be used to assess children of all ages, identify abnormal cardiopulmonary physiology, and define the
Chapter 164 / General Approach to the Pediatric Patient
mine adequacy of ventilation. The respiratory rate must be compared with the adequacy of air exchange and work of breathing when assessing ventilation. Tachypnea can be the first and most sensitive sign of respiratory compromise in the pediatric patient. Tachypnea, however, is nonspecific and can also be a product of fever, fear, anxiety, or pain. In a febrile infant, the respiratory rate will increase by up to five respirations per minute for every degree centigrade in temperature elevation. Either a slow or rapid respiratory rate can be a sign of impending respiratory failure. Periodic breathing is a common reason for an ED visit by concerned, often first-time parents. It is not at all unusual for infants to have periodic breathing with episodes of apnea lasting up to 20 seconds. To be considered abnormal, periodic breathing must be associated with a drop in heart rate or oxygen saturation. Vital signs at one given point in time may be quite difficult to interpret. Repeated measurement of the respiratory rate, heart rate, and blood pressure over time will provide a more accurate assessment of a patient’s physiologic condition. Again, it is imperative that the provider find an explanation for any abnormal vital sign before disposition of the patient.
2086
PART V ■ Special Populations / Section One • The Pediatric Patient
Table 164-3 Developmental Milestones AGE
GROSS MOTOR
VISUAL-MOTOR AND PROBLEM SOLVING
1 mo
Raises head slightly from prone position, makes crawling movements
2 mo
Holds head in midline, lifts chest off table
3 mo
Supports on forearms in prone position, holds head up steadily
Holds hands open at rest, follows in circular fashion, responds to visual threat
4 mo
Rolls front to back, supports on wrists and shifts weight Rolls back to front, sits supported
Reaches with arms in unison, brings hands to environment midline Transfers objects Unilateral reach, uses raking grasp
7 mo
Sits unsupported, puts feet in mouth in supine position Creeps
8 mo 9 mo
Comes to sit, crawls Pivots when sitting, pulls to stand, cruises
5 mo 6 mo
Birth: visually fixes 1 mo: has tight grasp, follows to midline No longer clenches fist tightly, follows object past midline
7–8 mo: inspects objects 7–9 mo: finger-feeds Uses pincer grasp, probes with forefinger, gestures, waves byebye, holds bottle, throws objects
12 mo
Walks alone
Uses mature pincer grasp, releases voluntarily, marks paper with pencil
15 mo
Creeps up stairs, walks backward
Scribbles in imitation, builds tower of 2 blocks in imitation
18 mo
Runs, throws objects from standing without falling
Scribbles spontaneously, builds tower of 3 blocks, plays in company of other children
21 mo
Squats in play, goes up stairs
Builds tower of 5 blocks
24 mo
Walks up and down steps without help
Imitates stroke with pencil, builds tower of 7 blocks, turns pages 1 at a time, removes shoes, pants, etc.
Modified from Gunn KL, Nechyba C (eds): The Harriet Lane Handbook, 16th ed. St. Louis, CV Mosby, 2003.
LANGUAGE, SOCIAL, AND ADAPTIVE
Alerts to sound Regards face Smiles socially (after being stroked or talked to) Recognizes parent Coos (produces long vowel sounds in musical fashion) Reaches for familiar people or objects, anticipates feeding Laughs, orients to voice Enjoys looking around Says “ah-goo,” orients to bell (localizes laterally) Babbles Recognizes strangers Orients to bell (localizes indirectly) “Dada” indiscriminately “Mama” indiscriminately, understands “no” Starts to explore environment, plays gesture games (e.g., patty cake) 10 mo: “Dada” and “Mama” discriminately, orients to bell (directly) 11 mo: 1 word other than “Dada” and “Mama,” follows 1-step command with gesture Uses 2 words other than “Dada” and “Mama,” immature jargoning (runs several unintelligible words together) Imitates actions, comes when called, cooperates with dressing 13 mo: uses 3 words 14 mo: follows 1-step command without gesture Uses 4–6 words 15–18 mo: uses spoon, uses cup independently 17 mo: uses 7–20 words, points to 5 body parts, uses mature jargoning (includes intelligible words in jargoning) Uses 2-word combinations Copies parent in tasks (sweeping, dusting), turns 2–3 pages at a time 19 mo: knows 8 body parts Uses 50 words, 2-word sentences Asks to have food and to go to toilet Uses pronouns (I, you, me appropriately), follows 2-step commands Parallel play
2087
Work of breathing
Work of Breathing Circulation to skin
Figure 164-1. Pediatric assessment triangle. urgency and need for lifesaving interventions. Before touching the patient, one should observe the child from a distance for visual and auditory clues. Figure 164-1 defines the three arms of the PAT: appearance, work of breathing, and circulation to the skin. This brief assessment rarely takes more than 30 seconds and adds to the initial discernment of “sick from well” (Table 164-4).10,11
Appearance From a distance, one should quickly determine the general appearance of the child. A “hands-off ” assessment of infants and young children can allow the examiner to gather critical information from a distance before upsetting the child with an invasive physical examination. The evaluation of appearance can be remembered by the mnemonic TICLS: Tone, Interactiveness, Consolability, Look/Gaze, and Speech/Cry. Is the child interacting normally with the environment? Is the level of consciousness appropriate? Is the patient irritable, somnolent, lethargic, or appropriately responsive? When the brain is not adequately perfused, irritability is usually the first sign. This is followed by alternating irritability and lethargy, which, if left untreated, can progress to coma. Infants can be the most difficult to assess since their ability to interact with their surroundings is limited. The provider may need to rely on parents or caregivers, who can often recognize normal or abnormal behavior for their infant. The glassy-eyed, “nobody home” stare of a septic or braininjured infant is not difficult to recognize. A high-pitched or cephalic cry is characteristic of any insult to the central nervous
Assessment of the work of breathing in infants and young children is best done from a distance. Once an infant begins to cry, it is difficult to make any reasonable interpretation of oxygenation and ventilation or to interpret breath sounds. One must listen carefully for audible abnormal airway sounds such as grunting, wheezing, stridor, and snoring. Grunting is an infant’s or child’s way of providing self-administered positive end-expiratory pressure to recruit collapsed or fluid-filled alveoli. The presence of inspiratory stridor alerts the examiner to upper airway obstruction. Muffled, hoarse, or abnormal speech can occur with trauma to the larynx or from a peritonsillar or peripharyngeal abscess. Assessment of the severity of wheezing should be determined by the presence of sounds that occur during both inhalation and exhalation, or just as expiratory sounds, and whether there is a prolonged expiratory phase. Observing for abnormal positioning can help determine the cause and severity of airway obstruction. A child assuming the “sniffing position” is attempting to best position the airway to overcome obstruction. Tripoding is often seen with severe respiratory distress in an attempt to maximize use of the accessory muscles of breathing (Fig. 164-2). The presence of intercostal, supraclavicular, and substernal retractions indicates an increased work of breathing (Fig. 164-3). Infants for the first several months of life may normally have abdominal breathing. Seesaw movements of the chest and abdomen are always abnormal. Nasal flaring and head bobbing can be seen in infants and young children with significantly increased work of breathing (Fig. 164-4).
Assessment Table 164-4 Pediatric Triangle—Initial Assessment APPEARANCE
WORK OF BREATHING
Tone
Abnormal sounds: stridor, grunting, snoring, wheezing Abnormal positioning: sniffing, tripoding, refusal to lie down Retractions Head bobbing Nasal flaring
Irritable, interactive Consolable Look/gaze Speech/cry
CIRCULATION TO THE SKIN
Pallor Mottling Cyanosis Petechiae
Modified from Dieckmann R, Brownstein D, Gausche-Hill M (eds): Pediatric Education for Prehospital Professionals. Sudbury, Mass, Jones & Bartlett, American Academy of Pediatrics, 2000, pp 36–40.
Figure 164-2. Tripod position in a child with airway obstruction.
Chapter 164 / General Approach to the Pediatric Patient
Appearance
system. Simple observation of older children for proper tone, motor movements, and reaction to environmental stimulation is a good assessment of the appearance arm of the PAT. If the child is unusually irritable, allowing the patient to remain with the parent can help differentiate between behavior and pathology. Again, when in doubt a parent or caregiver can confirm whether the behavior is abnormal for the patient. Appearance, when normal, can insinuate that ventilation, oxygenation, and brain perfusion are at least adequate.
PART V ■ Special Populations / Section One • The Pediatric Patient
2088
A
Figure 164-3. Intercostal retractions in a child with respiratory distress.
B Figure 164-5. Cutis marmorata (A) and mottling of the skin (B).
Effortless tachypnea, or rapid respirations with no increased work of breathing, is characteristic of a child compensating for metabolic acidosis by increasing the respiratory rate and driving the pH toward normal. It is important to remember that as a patient slips from respiratory distress to respiratory failure, work of breathing and the respiratory rate may both decrease. When this occurs, the appearance will change and a decrease in the level of consciousness will be evident.
become mottled. Mottling is manifested by areas of vasoconstriction and vasodilation in a random pattern on the skin. It reflects loss of small vessel integrity and may be similar to what is seen in vital organs during multiple organ system failure. Mottling is usually an ominous sign. It is important to not confuse cutis marmorata with mottling in young infants (Fig. 164-5). Cutis marmorata is a lacy marbling of the skin caused by vascular instability. It is a normal finding and is commonly seen in infants in a cool ambient environment. Cyanosis may occur in the late stages of shock or with respiratory failure. Unless the child is cyanotic as a result of chronic primary cardiopulmonary problems or congenital heart disease, the development of cyanosis is an indication of respiratory failure or decompensated shock. Table 164-5 summarizes how the PAT can be used to interpret specific physiologic abnormalities and the clinical condition of the patient.
Circulation to the Skin
Initial Hands-on Assessment
Visual inspection of the skin can be used to quickly assess peripheral perfusion. Young children have excellent compensatory reserves, and with early and compensated shock, blood is shunted from the skin to vital organs. Compensated shock can be recognized by the presence of pallor. A pale child with a rapid heart rate should always be considered to be in shock until proved otherwise. As cardiac output is further compromised and perfusion to vital organs is decreased, the skin may
Two critical issues should be remembered when interpreting vital signs in children: first, one must remember to use ageappropriate standards (see Table 164-2); second, changes in vital signs over time are far more important than any single recording. A monitored and sleeping infant with an increasing heart rate cannot be ignored. The initial hands-on assessment should be done in an orderly fashion by performing a stepwise assessment of airway,
Figure 164-4. Nasal flaring in a child with respiratory distress from lower airway obstruction.
2089
Table 164-5 Interpretation of the Pediatric Assessment Triangle APPEARANCE
WORK OF BREATHING
CIRCULATION TO THE SKIN
Respiratory distress Respiratory failure Compensated shock Decompensated shock Brain injury/dysfunction Cardiopulmonary failure
Normal Abnormal Normal Abnormal Abnormal Abnormal
Abnormal Abnormal Normal Normal/abnormal Normal Abnormal
Normal Normal/abnormal Abnormal Abnormal Normal Abnormal
Modified from Dieckmann R, Brownstein D, Gausche-Hill M (eds): Pediatric Education for Prehospital Professionals. Sudbury, Mass, Jones & Bartlett, American Academy of Pediatrics, 2000, pp 30–57.
breathing, and circulation and resolving issues related to each before progressing to the next step. Assessment of neurologic status or disability provides an opportunity for a more detailed objective measure of the child’s appearance from the PAT. The Glasgow Coma Scale or its pediatric modification provides methods for assessing disability that can be used to monitor changes in mental status over time. The AVPU (alert, verbal, painful, unresponsive) scale is a simple alternative to assess whether the child is alert, responsive to verbal commands, responsive only to painful stimuli, or unresponsive. Exposure of infants and young children can usually be performed more effectively with the parent’s assistance. One should make every effort to maintain a neutral thermal environment to avoid unnecessary heat loss during the examination. Modesty can be preserved and cooperation can be improved by exposing body parts one area at a time.
Triage The purpose of triage is to rapidly assess a patient and determine the urgency of evaluation and management. A number of triage scores have been devised, including the recently developed Canadian Pediatric Triage and Acuity Scale. This triage tool uses a five-level system. Table 164-6 uses a similar triage tool to list the 10 most common signs, symptoms, or problems categorized as triage level I in a busy inner-city ED in the United States. A triage level I patient is defined as one in shock or respiratory failure, unresponsive, or with absent or unstable vital signs. Traumatic injuries rank first for triage level I visits and rank third for triage level II visits. The second most common cause for a triage level I ED visit is seizures, which represent 16% of all triage level I visits. Respiratory failure is responsible for nearly 10% of all triage level I ED visits, and respiratory distress is the most common cause of triage level II visits. Table 164-7 lists the 10 most common causes for triage level II visits. A triage level II patient is one who is considered to be an emergency with a potential threat to life, limb, or function; one who is lethargic, with significant respiratory distress; or one with severe pain. A level II patient requires physician assessment within 15 minutes of arrival to triage.
Clinical Interview The initial contact with the child and parent will often determine the ultimate level of cooperation received and parent satisfaction with the visit. Parents who bring children to an ED for care usually perceive that their child has an emergency. Treating a family with respect, gentleness, and kindness goes a long way in ensuring a trusting patient-doctor relationship. If the family has been waiting a long time to see the doctor, start
Table 164-6 Triage Level I Emergency Department* Visits CATEGORY
NO. OF PATIENTS
Trauma Seizures Respiratory failure Altered consciousness Sepsis Cardiac problem Diabetic ketoacidosis Toxic ingestion SIDS/CPR Hypovolemic shock Percentage of total Total level I visits
PERCENTAGE OF LEVEL I VISITS
369 107 64 19 15 15 11 9 8 4
56.3 16.3 9.8 2.9 2.3 2.3 1.7 1.4 1.2 0.6 94.8 64.0
655
*Children’s Medical Center of Dallas, 2003; N = 102,453 visits. CPR, cardiopulmonary resuscitation; SIDS, sudden infant death syndrome.
Table 164-7 Triage Level II Emergency Department* Visits CATEGORY
Respiratory distress Suspected sepsis/meningitis Trauma/injury Sickle cell disease complications Diabetes complications Genitourinary complaints Ventriculoperitoneal shunt complications Seizures Hypovolemic dehydration Oncology patient with fever Percentage of total Total level II visits
NO. OF PATIENTS
PERCENTAGE OF LEVEL II VISITS
4132 881 786 727 447 413 388
33.0 7.0 6.3 5.8 3.6 3.3 3.1
332 309 255
2.6 2.5 2.0 69.2 (12.2)
12,540
*Children’s Medical Center of Dallas, 2003; N = 102,453 visits.
the conversation with a simple apology that expresses regret for the long wait. Introducing oneself to both the parents and child will facilitate a relaxed atmosphere for the interview. Toddlers and early school-age children may be expected to say little during the interview process, but they should be allowed to provide answers to questions when appropriate. In an emergency setting, the chief complaint and present illness are the main focus for information gathering, but the past,
Chapter 164 / General Approach to the Pediatric Patient
PHYSIOLOGIC STATE
2090
PART V ■ Special Populations / Section One • The Pediatric Patient
BOX 164-1 Focused “SAMPLE” History Signs/symptoms Allergies Medications Past medical problems Last food or liquid Events leading to injury/illness Modified from Dieckmann R, Brownstein D, Gausche-Hill M (eds): Pediatric Education for Prehospital Professionals. Sudbury, Mass, Jones & Bartlett, American Academy of Pediatrics, 2000, p 51.
family, and social history pertinent to the child’s condition must also be explored. It is wise to not begin the interview with “What’s the problem?” Parents do not like to consider their children or their conditions as “problems.” A better approach would be “What brings you and your child to the emergency department today?” The “SAMPLE” mnemonic may be used to systematically obtain a focused history (Box 164-1). Signs and symptoms that describe the onset and nature of the chief complaint should be detailed. Allergies or drug reactions are recorded with information describing the characteristics of the reaction. Medications that the patient is currently taking are recorded, including the time and amount of the last dose. Past medical problems and special health care needs are detailed, including information about the pregnancy, labor and delivery, and current immunization status. Knowledge of the last food and liquid given is important if sedation, analgesia, airway management, or surgery is necessary. Events leading up to the injury or illness should be recorded. Gentle and soothing conversation throughout the physical examination often facilitates information gathering and reduces anxiety. The history is generally obtained while performing the secondary assessment.
result in ligamentous instability, which if ignored may result in significant morbidity or mortality. A history of neck pain, paresthesias, numbness, tingling, or focal neurologic findings must not be ignored even if cervical spine films are normal.13 Because most out-of-hospital emergency medical systems require out-of-hospital providers to transport patients in complete spinal immobilization, it is important to obtain information from the transport professionals. Was the immobilization simply a precautionary measure by protocol? Was the child up, walking, and moving the neck before immobilization? Was there any history of pain, paresthesias, or evidence of neurologic injury? Such information can be helpful in determining whether the child needs radiographic imaging or can be clinically cleared from cervical spine immobilization. The decision to clinically clear a patient from immobilization can be considered if the patient is awake, alert, and cooperative with no distracting injuries and denies cervical spine tenderness. If at any time during the examination the patient complains of pain or has midline cervical tenderness, clinical clearance cannot be completed. As the secondary trauma survey is systematically performed, continued attention to the PAT and primary survey is critical to identify ongoing bleeding or progressing respiratory problems. A pale trauma victim should have vascular access and blood available before or during the secondary survey. Intentional injuries to children still result in more than 1200 deaths per year. Because these injuries are most commonly associated with blunt trauma, it is possible that no external evidence of injury will be apparent.14 When historical indicators suggest possible child abuse, consultation with child protective services is mandatory. Box 164-2 lists historical indicators that should alert the heath care provider to the possibility of child abuse.15 Examination of the skin for burns or bruises consistent with child abuse should be a part of any trauma survey. Box 164-3 summarizes bruises typically found in abused infants and children. See Chapter 64 for a more extensive discussion of these issues.
Physical Examination In infants and young children, the physical examination is not a stepwise process. It is neither a head-to-toe nor a toe-to-head evaluation. The examination should be performed in the least traumatic fashion by leaving painful or frightening components until the end and concentrating on high-value components initially. Children in late infancy through the toddler age should be left in the caregiver’s lap during the majority of the examination.
BOX 164-2 Historical Indicators of Child Abuse Unexplained delay in seeking medical care. History does not explain the injury. History changes with time. History is not consistent with the child’s developmental abilities. Child has “magical” injuries.
Specific Disorders Trauma
BOX 164-3 Bruises Suggestive of Child Abuse
The initial assessment of the injured pediatric patient starts with a primary survey that includes careful assessment of the airway, adequacy of oxygenation and ventilation, as well as a search for signs of circulatory compromise. Once life-threatening abnormalities are identified and appropriate resuscitation efforts are begun, a careful and detailed secondary trauma survey is necessary to identify subtle but potentially lifethreatening injuries. A systematic approach to pediatric trauma that includes a continuum of care from first responders through ED stabilization, interfacility transport, and definitive care will save lives.12 Attention to assessment of the cervical spine is a critical part of pediatric assessment. Because of the elasticity of the cervical spine in young children, spinal cord injuries without radiographic abnormalities (SCIWORA) can occur. These injuries
Any unexplained bruising in an infant under 9 months of age Multiple bruises of different ages Pattern injuries Handprints Belt marks Cord loop marks Linear marks from rigid objects Bite marks Unusual distribution of bruises Neck Groin Inner aspect of thigh Restraint marks on wrists or ankles
2091
BOX 164-4 Clinical Croup Score 0 1 2 0 1 2 3 0 1 2 0 4 5 0 5
From Westley CR, Cotton EK, Brooks JG: Nebulized racemic epinephrine by IPPB for the treatment of croup. Am J Dis Child 132:484, © 1978. Reprinted by permission of Wiley-Liss, a division of John Wiley & Sons, Inc.
Medical Disorders Seizures. Seizures are a common presenting complaint in the ED. Although most seizures in children are benign and selflimited, assessment must include attention to the airway and ventilation. An actively convulsing child must be quickly assessed for adequacy of oxygenation and ventilation. A nasal airway properly inserted can dramatically relieve airway obstruction and allow for assisted bag/mask ventilation when appropriate. As the child arouses from the postictal phase, the nasal airway should be removed to avoid vomiting. As a part of the secondary survey, the emergency physician should observe carefully for the character of the seizure and the presence or absence of any focal findings. Fever, central nervous system infections, and brain injury must always be considered in children with seizures and no history of childhood epilepsy.
Table 164-8 Pediatric Asthma Severity Score CLINICAL FINDING
DEFINITION
0
1
2
Wheezing
High-pitched expiratory sound absent wheezing heard by auscultation because of poor air exchange Intensity of inspiratory sounds
None or mild
Moderate
Severe
Normal or mildly diminished None or mild
Moderately diminished
Severely diminished
Moderate
Severe
Moderately prolonged
Severely prolonged
Air entry Work of breathing Prolongation of expiration Tachypnea Mental status
Observed use of accessory muscles, retractions, or in-breathing Ratio of duration of expiration to inspiration Respiratory rate above normal for age Observation of the child’s state of alertness
Normal or mildly prolonged Absent
Present
Normal
Depressed
From Gorelick MH, Stevens MW, Schultz TR, Scribano PV: Performance of a novel clinical score, the Pediatric Asthma Severity Score (PASS), in the evaluation of acute asthma. Acad Emerg Med 11:10, 2004.
Chapter 164 / General Approach to the Pediatric Patient
Stridor None Audible with stethoscope (at rest) Audible without stethoscope (at rest) Retractions None Mild Moderate Severe Air entry Normal Decreased Severely decreased Cyanosis None With agitation At rest Level of consciousness Normal Altered
Other less common problems to rule out include intoxication, metabolic derangements, intracranial vascular anomalies, and tumor. Neonatal seizures can be very subtle and difficult to recognize. A seizure in this age group may be manifested as lack of an appropriate response to environmental stimuli, nystagmus, blinking eyes, or any fine repetitive movements.16 The Difficult Airway and Airway Obstruction. Recognition of a “difficult airway” can be a critically important issue in the assessment of seriously ill and injured children. Although scientific evidence for defining a difficult airway in children is limited, certain critical observations may be helpful. Dysmorphic features or physical conditions that limit the ability to completely open the mouth or that reduce neck mobility should suggest the possibility of a difficult airway. Although Mallampati grades have not been well studied in children, if the uvula cannot be visualized when opening the mouth, the patient should be considered high risk. Any evidence of upper airway obstruction in a child who is in impending respiratory failure should alert the emergency physician to the need of alternate airway devices or the possibility of surgical airway support.17 Scores that quantify the severity of upper and lower airway obstruction and disease are helpful tools to monitor response to therapeutic interventions. An example of a clinical croup score is detailed in Box 164-4.18 The pediatric asthma severity score seen in Table 164-8 has been shown to be reliable, reproducible, and valid for assessing the severity of acute asthma in children 1 to 18 years of age.19 Altered Level of Consciousness. Altered level of consciousness accounts for approximately 3% of all critical ED visits. An orderly approach to the assessment of a child with altered level of consciousness is important to minimize organ damage and provide timely initial management. Quick assessment of the respiratory pattern can identify the irregular appearance of a Cheyne-Stokes breathing pattern, which consists of alternating periods of hyperpnea, followed by slowing ventilatory effort and periods of apnea. This respiratory pattern is typically found in children with increased intracranial pressure. Midbrain dysfunction will generally cause hyperventilation that is regular and rapid with normal oxygen saturation and low arterial partial pressure of carbon dioxide (Paco2). Examination of the eyes is helpful in sorting out causes of an altered level of consciousness. Pupils that are fixed and dilated in an unresponsive patient are always suggestive of
PART V ■ Special Populations / Section One • The Pediatric Patient
2092
serious intracranial pathology and are never seen in patients unresponsive because of metabolic derangements. A unilaterally dilated pupil is secondary to increased intracranial pressure from mass lesions causing uncal compression on the third cranial nerve or due to direct ocular trauma. With significantly increased intracranial pressure, transtentorial herniation will increase pressure on the brainstem and result in asymmetrical pupils that become fixed and dilated. One should observe closely for equality of extraocular movements, nystagmus, and deviation of the eyes at rest. Funduscopic examination is helpful in identifying papilledema and retinal hemorrhages. Assessment of motor movements including strength and tone is an important part of the secondary assessment for altered consciousness. One should observe closely for symmetry of movements as the absence of symmetry should signal the possibility of stroke and the presence or absence of abnormal posturing should alert the provider to the possibility of intracranial pathology and/or seizure. Shock. Patients with trauma, sepsis, cardiac problems, diabetic ketoacidosis, diarrhea, vomiting, and ingestion of toxic substances may arrive at the ED in shock. The secondary survey will help differentiate causes of the shock state. Assessment of four organ systems in systematic fashion will help determine the etiology and severity of the shock and monitor response to treatment. The heart is the first organ to respond to a shock state and does so with an increased rate. Cardiac output is a function of stroke volume and heart rate, and when stroke volume decreases for any reason, the heart rate will increase. Careful examination of the heart by listening closely for a gallop rhythm, murmurs, indistinct or muffled heart sounds, or any dysrhythmia will help identify cardiogenic shock as the cause. The presence of these signs with development of hepatomegaly, especially after the administration of intravenous fluids, can also indicate the presence of cardiogenic shock. Tachycardia may be absent in patients with distributive shock. In the late stages of decompensated shock, bradycardia will occur as the shock progresses and can herald cardiopulmonary failure. The second organ to respond to shock is the skin. In addition to the manifestations covered in the PAT, one should observe for skin temperature and signs of dehydration. The reverse thermometer sign can be of assistance in roughly judging the degree of hypovolemic shock. The examiner assesses skin temperature by running the fingers up the extremity to determine the point of cool/warm demarcation. During resuscitation, as reperfusion of the skin occurs, this point of warm/cold demarcation progresses peripherally. With early septic shock, the skin may appear warm and flushed. One must look closely for signs of dehydration, including tenting of the skin, a hollow-eyed appearance, or dry mucous membranes, which may suggest dehydration and hypovolemic shock. Capillary refill must be measured at or above the level of the heart to avoid erroneous results from venous flushing. In a neutral thermal environment, capillary refill longer than 2 seconds will be present in all forms of shock. The third organ that objectively responds to the shock state is the brain. Irritability or a decrease in mental status is consistent with progressing shock. The lungs respond to shock with tachypnea and hyperpnea. Hypovolemic shock will often be manifested as effortless tachypnea when the patient has become acidotic. Cardiogenic shock, distributive shock, and septic shock may also give rise to increased work of breathing, crackles, or wheezing (or any combination of the three). Close and repeated monitoring of these four organ systems must be continued throughout the secondary survey and stabilization processes.
Children with Special Health Care Needs Children with special health care needs account for a large number of emergency and urgent ED visits. In a child with developmental delay or significant neurologic problems, clinical assessment can be difficult and routine assessment tools may be limited in their usefulness. The caretaker is usually quite experienced and can often provide valuable information to assist in assessment. A behavior change recognized by the caretaker may be the only clue to a potentially serious illness or complication. For example, children with life-sustaining hardware such as ventriculoperitoneal shunts commonly come to the ED with nonspecific findings such as vomiting and headache, and change in behavior may be the only clue leading to an accurate diagnosis. It is important to engage the parent to assist with baseline information and to seek specialty consultation from providers who are familiar with children who have special health care needs.
■ DIAGNOSTIC STRATEGIES Noninvasive Monitoring Noninvasive monitoring techniques such as oxygen saturation and end-tidal carbon dioxide measurements have become a routine part of pediatric assessment for a variety of illnesses and injuries. Pulse oximetry provides a valuable noninvasive and continuous measurement of arterial hemoglobin oxygen saturation, which has rapidly become a new “vital sign” for many illnesses. The role of pulse oximetry in monitoring children with respiratory distress and failure has been well defined. The use of pulse oximetry in children with cardiorespiratory problems is a valuable assessment tool and helps the clinician titrate the need for oxygen supplementation and respiratory support. End-tidal CO2 measurement is a noninvasive method for monitoring a variety of critically ill children that helps avoid repeated blood gas analysis. Studies have demonstrated that analysis of exhaled end-tidal CO2 can be used to monitor respiratory failure, accurate endotracheal tube placement, and perfusion to peripheral tissues during the resuscitation of patients in shock. Measurement of end-tidal CO2 and pulse oximetry can provide valuable assistance in monitoring children in both shock and respiratory failure. It is also a critical component of assessing correct placement of a tracheal tube after intubation has been performed.
The Pediatric-Ready Emergency Department ED preparedness for the care of children is a special need that requires limited additional resources, but significant professional support and advocacy from both physicians and nursing staff. It is quite clear that the majority of hospitals with EDs will serve as pediatric receiving facilities for emergency care independent of size, volume, or availability of comprehensive resources. Hospitals without all the necessary resources must clearly be prepared to stabilize critical pediatric illnesses and injuries within their limitations and ensure timely transfer to facilities with available necessary resources. Guidelines for preparedness for the care of children in the ED are available.5,6 The availability of protocols and policies that ensure timely transfer of critically ill and injured patients after stabilization is a part of the pediatric assessment process.
2093
KEY CONCEPTS The physical and developmental stages of growth in pediatric patients provide challenges to the assessment of children in crisis. ■ An understanding of age-related developmental issues and differences is important for performing an adequate pediatric assessment. ■ One must establish rapport with pediatric patients to get cooperation in performing the examination. ■ Abnormal vital signs always require explanation and resolution. ■ The PAT provides a rapid assessment of severity and physiologic status and will provide a road map for the initial assessment of all pediatric patients. ■ Continuing close monitoring and reassessment, especially following any intervention, is a key component to the evaluation and management of the ill or injured child. ■ A neutral thermal environment should be maintained during the physical examination to avoid heat loss in infants who are critically ill. ■ The presence of a difficult airway can be assessed in seriously ill or injured children by determining whether
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
they are able to open their mouth wide enough for visualization of the uvula and normally extend and flex their neck. ■ With special needs children, the caregiver can assist in determining baseline status. ■ Beware the pale child: Pallor is a consistent and objective early sign of shock that cannot be ignored, especially in those with blunt trauma. ■ Cervical spine: A large head with an elastic cervical spine may result in SCIWORA. Patients with this condition may not have any neurologic deficits, and cervical spine images may be normal. A history of transient paresthesia, numbness, tingling, or any focal neurologic findings, even if radiographs are normal, should be interpreted in this light and clinical clearance from immobilization should be delayed until further evaluation is performed. ■ ED preparedness for pediatric emergencies: Guidelines for having a pediatric-ready ED are available through the American College of Emergency Physicians and the American Academy of Pediatrics.
Chapter 164 / General Approach to the Pediatric Patient
■
Chapter 165
Pediatric Fever
Nathan W. Mick
Fever is the most common presenting chief complaint in pediatric patients presenting to the emergency department (ED), accounting for up to 20% of ED visits. Most cases of fever are viral in origin, benign in course, and resolve spontaneously. Fever tends to be of a higher clinical importance in younger children as they are immunologically immature and incompletely vaccinated. Management of children presenting to the ED varies dramatically depending on the age of the child with the following common, albeit arbitrary, divisions (0–28 days, 1–2 months, 2–3 months, 3–6 months, 6–36 months, 3 years to adulthood).1 These divisions reflect differing immunologic and vaccination milestones as well as the spectrum of agespecific pathogens.
■ PERSPECTIVE Definitions and Epidemiology Fever is defined as any elevation in body temperature equal to or above 38.0° C and typically is the result of infection. The most reliable method to measure temperature is via a rectal thermometer, particularly in high-risk groups such as infants 0 to 3 months of age as axillary, oral, or tympanic thermometers are unreliable in this age group. The rectal route should not be used in patients who are potentially immunocompromised (i.e., children with fever who are receiving cytotoxic chemotherapy) due to the risk of mucosal damage, bacteremia, or transmission of infection. The cutoff for a clinically significant fever (i.e., one that triggers a laboratory evaluation) varies with the age and immunologic status of the child. A rectal temperature of 38.0° C is generally considered to be a clinically significant fever in an infant less than 3 months old, necessitating a thorough laboratory workup, whereas a toddler with a temperature of 39.5° C and an upper respiratory infection may not need any workup beyond a thorough history and physical. Fever is to be distinguished from hyperthermia, which is an elevation in the body’s “set point.” Causes of hyperthermia include bundling, heatstroke, salicylate ingestion, malignant hyperthermia as a complication of inhalational anesthetics, and elevation in temperature secondary to hypothalamic central nervous system (CNS) damage. The cause of fever varies depending on the age of the child (Table 165-1). The vast majority of pediatric fever is due to infections, and the vast majority of infections are attributable to a viral source. Upper respiratory infections, viral gastroenteritis, croup, bronchiolitis, stomatitis, roseola, infectious mononucleosis, and varicella are all known causes of 2094
fever. Most viral illnesses are benign and self-limited, but infection with herpes simplex virus (HSV) at any age, or respiratory syncytial virus (RSV), particularly in the first month of life, can lead to significant morbidity and mortality. Bacterial disease is also an important cause of fever in children. Serious bacterial illness (SBI) is typically defined as the presence of pathogenic bacteria in a previously sterile site and includes urinary tract infection (UTI), bacteremia, meningitis, osteomyelitis, bacterial gastroenteritis, bacterial pneumonia, cellulitis, or septic arthritis. Studies have found the risk of SBI in febrile infants less than 3 months of age to be between 6 and 10%, with children less than 28 days of age having the highest incidence.2 Pathogens change during early infancy, with vertical transmission of organisms such as group B Streptococcus, Listeria monocytogenes, and HSV being more common in neonates. By 1 to 2 months of age, organisms such as Streptococcus pneumoniae, Neisseria meningitidis, and urinary pathogens (Escherichia coli or Enterococcus) become more common. In all children less than 3 months of age, the urinary tract is the most common site of infection, followed by bacteremia and meningitis. UTIs are more common in white children, particularly females, when compared with other races and is of higher prevalence in patients in whom no source for infection is found and who have higher temperatures (i.e., >39.0° C).3 Children under the age of 3 months may present with an apparent viral syndrome and still harbor serious bacterial illness. Levine and colleagues studied 1248 infants less than 60 days of age who had temperatures above 38.0° C. Of these children, 22% were positive for RSV. Although, overall, children with documented RSV had a lower incidence of concomitant SBI than those without RSV (12.5 vs. 7%), there was no significant difference in rates of SBI in children less than 28 days of age (14.2% in RSV-negative neonates vs. 10.1% in RSV-positive infants). Most of the bacterial infections were UTIs.4 Older children 3 to 36 months of age with recognizable viral syndromes (e.g., croup, bronchiolitis, varicella, stomatitis) generally have a very low incidence of bacteremia. Greenes and coworkers found that among 1347 patients with fever above 39.0° C who had a recognizable viral syndrome, the risk of bacteremia was 0.2%.5 Occult bacteremia describes the presence of pathogenic bacteria in the bloodstream of a well-appearing febrile child in the absence of a focus of infection and was first described as a clinical entity in the 1970s.6 The term typically refers to children 3 to 36 months of age who are highly febrile (>39.0° C) but appear well. Prior to the adoption of the conjugate vaccines
2095
Table 165-1 Etiology of Fever in Children BACTERIAL CAUSES
VIRAL CAUSES
OTHER
0–28 days of age
Group B Streptococcus Listeria monocytogenes Escherichia coli Chlamydia trachomatis Neisseria gonorrhoeae Haemophilus influenzae Streptococcus pneumoniae Neisseria meningitidis E. coli S. pneumoniae N. meningitidis E. coli
Herpes simplex virus Varicella Enteroviruses Respiratory syncytial virus Influenza Varicella Enteroviruses Respiratory syncytial virus Influenza Varicella Enteroviruses Respiratory syncytial virus Influenza Mononucleosis Roseola Adenovirus Norwalk virus Coxsackievirus Varicella Enteroviruses Respiratory syncytial virus Influenza Mononucleosis Roseola Adenovirus Norwalk virus
Bundling Environmental
3 years to adulthood
S. pneumoniae N. meningitidis E. coli Group A Streptococcus
against Haemophilus influenzae type b (HIB) and S. pneumoniae, the incidence of bacteremia in this population was approximately 5%.7,8 Vaccination has proven remarkably effective, nearly eradicating HIB as a significant pathogen and greatly reducing the burden of pneumococcal disease (Fig. 165-1).9–11 Currently, the rate of occult bacteremia is below 1%, with pathogens such as N. meningitidis becoming proportionally more prevalent. Continued surveillance is ongoing to ensure that there is not a rise in invasive disease caused by nonvaccine serotypes. Urinary pathogens continue to be an important source of bacterial illness in infants and children, occurring in 2% of febrile children less than 5 years of age. Rates are highest in males less than 6 months of age (2.7%) and females less than 12 months of age (6–8%).12,13 Bacterial illness in school-aged children and adolescents includes focal infections such as streptococcal pharyngitis, cellulitis, and pneumonia, as well as bacteremia and meningitis. N. meningitidis has a bimodal distribution, with the highest incidence in children under 12 months of age (9.2/100,000 population). A second peak occurs during adolescence, when the rate of illness is 1.2/100,000 population, with a significant proportion of cases occurring in college students who reside in a dormitory setting (3.2/100,000 population).14 Although much less common than viral or bacterial infection, fever can also be a presenting sign of autoimmune diseases such as juvenile rheumatoid arthritis or Kawasaki disease (KD). CNS lesions such as brain tumors also can infrequently present with fever.
■ DISTINGUISHING PRINCIPLES OF DISEASE Pathophysiology The body’s ability to fight infection varies with age. Maternal antibodies confer some protection after birth, but the infant’s
Bundling Environmental Leukemia Lymphoma Neuroblastoma Wilms’ tumor
Leukemia Lymphoma Neuroblastoma Wilms’ tumor Juvenile rheumatoid arthritis
120 1989 1990 1991
100 Cases/100,000
3–36 months of age
80 60 40 20 0 0–14
15–23 Age (months)
A
200
Cases/100,000
1–3 months of age
Chapter 165 / Pediatric Fever
AGE
100 50 0
B
1998–1999 2001
150
12 wk >20 wk
Blood pressure > 140/90 mm Hg; proteinuria, edema, HELLP syndrome Respiratory distress, cardiovascular collapse Uterine tenderness; vaginal bleeding; fetal distress
DIC, disseminated intravascular coagulation; hCG, human chorionic gonadotropin; HELLP, hemolysis, elevated liver enzymes, low platelets; PIH, pregnancyinduced hypertension.
the differential diagnosis and guiding the selection of ancillary studies.
Appendicitis Perspective. Appendicitis is the most common surgical emergency in pregnant patients. The incidence of appendicitis in
pregnant patients is the same as that in nonpregnant patients, but delays in diagnosis contribute to an increased rate of perforation, which results in significant fetal mortality and maternal morbidity.61,62 During the first half of pregnancy, diagnostic findings are usually similar to those in the nonpregnant woman, but the clinical picture becomes less classic during the second half of pregnancy.
Chapter 176 / Acute Complications of Pregnancy
DIAGNOSIS
PART V ■ Special Populations / Section Two • The Pregnant Patient
2292
8 mo. 7 mo. 6 mo. 5 mo. 4 mo. 3 mo. P.P. McBurney’s point
Figure 176-8. Locations of the appendix during succeeding months of
pregnancy. In planning an operation, it is better to make the abdominal incision over the point of maximal tenderness unless there is great disparity between that point and the theoretical location of the appendix. P.P., postpartum. (From Gabbe SG, et al [eds]: Obstetrics: Normal and Problem Pregnancies. New York, Churchill Livingstone, 2007; as modified from Baer JL, Reis RA, Arens RA: Appendicitis in pregnancy. JAMA 98:1359, 1932.)
Principles of Disease. Traditionally, the appendix was thought to displace counterclockwise out of the right lower quadrant after the third month of gestation, with its ultimate location deep in the right upper quadrant, superior to the iliac crest (Fig. 176-8). However, a study has questioned this finding; it found that in only 23% of pregnant patients did the location change from the right lower quadrant, even in the third trimester.62 Displacement of the abdominal wall away from the abdominal viscera can result in difficulty palpating organs and loss of signs of parietal peritoneal irritation. The physiologic increase in white blood cell count and erythrocyte sedimentation rate in pregnancy must also be considered when evaluating the patient with possible appendicitis because these may confuse the overall clinical picture. Clinical Features. The gastrointestinal symptoms of appendicitis, such as anorexia, nausea, and vomiting, mimic those of pregnancy, particularly during the first trimester, making such symptoms relatively nonspecific. Right-sided abdominal pain is the most constant finding, although this is less reliable later in pregnancy. Peritoneal signs are likewise most common during the first trimester. Lack of fever, leukocytosis, or tachycardia has been reported.63,64 The cause of these differences in clinical findings in pregnant patients with appendicitis may be a blunted inflammatory response from elevated maternal levels of pregnancy-related steroids. Pyuria without bacteruria is seen in up to 58% of patients.65 Because of confounding factors, the misdiagnosis rate for appendicitis is 30 to 35% overall in pregnancy, with a 40 to 50% rate of removal of normal appendix during the third trimester.63,65,66 In contrast to the relative safety of performing an exploratory laparotomy or laparoscopy during pregnancy, the risk of fetal loss and maternal morbidity from failure to diagnose appendicitis and perforation is considerable, so clinical vigilance is required even in the absence of classic signs. In
later pregnancy, when peritoneal signs are often absent and the uterus obscures normal physical findings, diagnosis is frequently delayed and the perforation rate may approach 25%. Differential Considerations. Pyelonephritis, cholecystitis, nephrolithiasis, and pregnancy-related diseases such as ectopic pregnancy, broad ligament pain, corpus luteum cyst leakage, and ovarian torsion must be considered in the patient who has right-sided abdominal pain. Pyelonephritis is the most common condition that is confused with appendicitis. During its migration, the appendix takes up a position very near the kidney, resulting in a high incidence of pyuria and flank pain (see Fig. 176-8). In cases of appendicitis, unless there is coincident urinary tract infection, the urine is free of bacteria, a feature distinguishing it from pyelonephritis. Salpingitis, another common misdiagnosis, is very rare in pregnancy, although it can occur before 12 weeks of gestation. Diagnostic Strategies. Leukocytosis is common in pregnant patients with appendicitis, although it is rarely high enough to distinguish it from the physiologic leukocytosis of pregnancy. Pyuria in a catheterized urine specimen suggests pyelonephritis, but it is also seen in 20% of patients with appendicitis.63 Bacteruria is uncommon. Ultrasonography, using graded-compression techniques, may reveal a noncompressible tubular structure in the right lower quadrant consistent with appendicitis. Studies regarding the diagnostic utility of ultrasonography in the diagnosis of appendicitis are limited but suggest that it has a high positive predictive value but a low negative predictive value.65,67 Given the low radiation risk to the fetus, the noninvasive nature of the test, and its utility in evaluating other complications, some authors have recommended that ultrasonography should be the initial test of choice when appendicitis is suspected, especially in the first and second trimesters.64,65 Castro and colleagues68 reported the utility of helical CT in diagnosis, with fetal radiation exposures of only 300 mrad, using rectal contrast and a limited study. Otherwise, laparoscopy or laparotomy is the diagnostic procedure of choice in the pregnant patient suspected of having appendicitis. Early exploration is to be encouraged even more in pregnant than in nonpregnant patients because of the variability of clinical signs and the increased fetal risk if diagnosis is delayed. Management. The pregnant patient with suspected appendicitis should be admitted to the hospital after appropriate consultation with surgeons and obstetricians. Ultrasonography or CT scan should be considered as diagnostic options. The patient should be kept on nothing by mouth (NPO) status, with intravenous fluid hydration to maintain intravascular volume. Although prompt surgery is required if the diagnosis is clear, in unclear cases a period of inpatient observation may allow clarification of signs and symptoms.
Gallbladder Disease Perspective. Cholelithiasis is present in approximately 5% of pregnant women and is the second most common nonobstetric surgical condition in pregnant patients. The natural history of asymptomatic cholelithiasis is believed to be similar to that in nonpregnant women, with less than half of patients with gallstones developing symptoms.69,70 Principles of Disease. Changes in gallbladder kinetics are believed to be due to high pregnancy-related steroid levels. Progesterone decreases smooth muscle tone and induces gallbladder hypomotility and cholestasis, causing an increased risk of stone formation. In addition, pregnancy induces changes in bile composition and increased cholesterol secretion, thus increasing the incidence of cholesterol stone formation.71 Clinical Features. The signs and symptoms of acute cholecystitis during pregnancy are the same as those in nonpregnant
Liver Disorders Perspective. Pregnancy is associated with several unique liver abnormalities in addition to more usual hepatic diseases. Clinicians should recognize the various symptoms of liver disease during pregnancy as well as the hepatic diseases unique to pregnant women. Liver metabolism increases during pregnancy, but hepatic blood flow is unchanged and little change occurs in liver function. Bilirubin, transaminases, lactate dehydrogenase, and prothrombin times are unchanged from the nonpregnant state. Albumin levels decrease secondary to an increase in maternal circulating plasma volume. Alkaline phosphate levels may be up to double the nonpregnant values, and amylase levels may also be slightly elevated.74-76 Hepatitis. Hepatitis is the most common cause of liver disease in pregnancy, accounting for 40% of cases of jaundice in pregnancy. Management and treatment are supportive and unchanged from those for nonpregnant patients. Hepatitis E,
2293
however, has been reported to have a more aggressive course in pregnancy.76 Maintaining adequate nutrition is a priority. Vertical transmission of hepatitis B can occur if the disease is not recognized. Prophylaxis should be administered to the newborn. Acute Fatty Liver of Pregnancy. Acute fatty liver of pregnancy is a disorder of the third trimester that can result in hepatic failure, complicated labor, and fetal mortality. The disease is rare, occurring most often in primiparous patients and patients with twin gestations. Principles of Disease. The cause of acute fatty liver of pregnancy is unknown, although studies suggest that a deficiency in the fetus’s fatty acid metabolism leads to an accumulation of hepatotoxic metabolites in the maternal circulation.76-79 Microscopically, fatty infiltration of the hepatocytes with edema and vacuolization can be seen, but there is no necrosis or inflammation. Liver function returns to normal after delivery if the patient can be supported through the acute phase. Although up to 50% of patients have signs of preeclampsia, the two are not clearly related.76 The diagnosis must be differentiated from viral hepatitis and HELLP syndrome, which have similar disease presentations and laboratory findings but, again, are not clearly related.76,77 Clinical Features. Nausea and vomiting or liver dysfunction during the third trimester should trigger consideration of a diagnosis of acute fatty liver. In addition, nonspecific flulike symptoms, such as anorexia, fatigue, and headache, occur initially.80 The right upper quadrant and/or epigastrium is usually tender. The disease may progress to coagulopathy, jaundice, seizures, disseminated intravascular coagulation, and hepatic encephalopathy. The diagnosis is often delayed secondary to the multiple differential considerations.81 Diagnostic Strategies. Typically, leukocytosis is present, the platelet count and fibrinogen level are low, prothrombin time and partial thromboplastin time are elevated, and fibrin split products are present. Hypoglycemia may occur. Serum transaminase levels are elevated, although rarely above 1000 U/L (a distinguishing feature from hepatitis), and should be measured in all patients who have systemic gastrointestinal symptoms during the third trimester. In contrast to Reye’s syndrome, the serum ammonia level is only mildly elevated. Hyperuricemia is usually present. Bilirubin is elevated late in the course of the disease. The CT scan is usually normal, as is the sonogram. Liver biopsy is used to make the definitive diagnosis.82 Differential Considerations. Liver tenderness and coagulopathy most often suggest preeclampsia during the third trimester.82 Jaundice and increases in aminotransferase level are distinguishing features because they are unusual in cases of liver disease associated with pregnancy-induced hypertension. Similarly, rapid progression of hepatic failure, hypoglycemia, and coagulopathy is unlikely in cases of preeclampsia. The patient with viral hepatitis is likely to have more marked elevations in transaminase levels. Drug-induced hepatic failure should be excluded by history and toxicologic screen for acetaminophen or other toxins if appropriate. Cholecystitis may be distinguished by ultrasound, but it may also be characterized by right upper quadrant pain; it is not associated with coagulopathy or progressive liver failure. Management. The patient with acute fatty liver of pregnancy requires immediate stabilization if experiencing seizure or coma. Hypoglycemia may occur, which should be rapidly corrected with dextrose. Coagulation parameters should be assessed. Fluid resuscitation and replacement of clotting factors may be required, and the patient should be admitted to an obstetric service capable of managing this serious disease. The diagnosis is usually made with liver biopsy if the disease
Chapter 176 / Acute Complications of Pregnancy
women. Epigastric or right upper quadrant pain and tenderness and nausea predominate. Leukocytosis must be interpreted carefully because of the increased white blood cell count seen normally in pregnancy. Likewise, a slightly elevated amylase level can be normal during pregnancy, and alkaline phosphatase, which is produced by the placenta, may be twice the nonpregnant level. A history of self-limited pain episodes associated with food intake is useful in suggesting the diagnosis. Diagnostic Strategies. Ultrasonography is a reliable means of recognizing stones within the gallbladder, although it may not differentiate symptomatic from asymptomatic stones. In the patient with right upper quadrant pain, simultaneous sonographic evaluation of the liver is useful but technically difficult, particularly during the third trimester, when subcapsular liver hematomas and other intrinsic hepatocellular disease can occur but the liver may be obscured under the ribs. Differential Considerations. Pyelonephritis should always be considered in the patient with right upper quadrant pain with or without fever. During the third trimester, appendicitis can also be associated with right upper quadrant pain. Hepatitis and fatty liver infiltration occur in pregnancy; liver distention and inflammation associated with pregnancy-induced hypertension can also cause right upper quadrant pain. In addition, spontaneous intrahepatic bleeding can occur in late pregnancy, mimicking cholecystitis. Because of the potential for other serious diseases, diagnostic studies should always be performed to verify a clinical diagnosis of symptomatic cholelithiasis and cholecystitis in pregnancy. Management. The patient who has fever, leukocytosis, prolonged pain, or evidence of cholecystitis should be made NPO and given IV fluid hydration, adequate pain control, and broadspectrum antibiotics. These patients must be admitted for inpatient management. Some patients can be managed medically for prolonged or complicated cholecystitis. Patients with obstructive jaundice, gallstone pancreatitis, or sepsis or patients who fail conservative management should have surgery. Discharge should be considered only in patients with uncomplicated and sonographically proven cholelithiasis who do not meet admission criteria after consultation with an obstetrician. Pregnant patients with symptomatic cholelithiasis have a high rate of symptomatic relapse and increased severity of disease with each relapse.72 Early follow-up should be arranged and the patient should be given careful instructions to return if she experiences fever, vomiting, or persistent pain. In one study, one third of pregnant women with biliary colic failed conservative treatment but were treated safely, often using newer laparoscopic techniques.73
PART V ■ Special Populations / Section Two • The Pregnant Patient
2294
has not progressed to severe coagulopathy. Rapid delivery is usually advisable when the diagnosis has been established. The route of delivery is dictated by the patient’s clinical course and hemodynamic status. Fresh frozen plasma, platelet transfusions, and glucose may be needed to sustain the patient until delivery can be accomplished. After delivery, infants of mothers with acute fatty liver of pregnancy are at risk for postpartum hypoglycemia and liver dysfunction and therefore should be monitored closely.82 Intrahepatic Cholestasis of Pregnancy. Intrahepatic cholestasis of pregnancy, also termed idiopathic jaundice of pregnancy, icterus gravidarum, or pruritus gravidarum, is a rare syndrome that occurs during the third trimester of pregnancy. It is the second most common cause of jaundice in pregnancy, after hepatitis. Histologically, the disease is characterized by cholestasis and dilated canaliculi in the biliary tree. The liver is normal. It is more common with increasing maternal age, with multiple gestations, and in the winter months.76,80 Clinical Features. Generalized pruritus and mild jaundice are the hallmarks of intrahepatic cholestasis of pregnancy. Only 20% of patients present with this combination, however, and 80% present with pruritis alone. The pruritis usually begins in the palms and soles and ascends to the trunk. Although insomnia and fatigue occasionally accompany the pruritus, the patient appears nontoxic, without fever, vomiting, diarrhea, or significant malaise. The bilirubin level is rarely above 5 mg/dL, the alkaline phosphatase level can be elevated 7- to 10-fold, and transaminase levels are in the normal range. Resolution occurs when the woman delivers. Although maternal outcome is favorable, women with intrahepatic cholestasis of pregnancy are at increased risk for preterm delivery, meconium passage, and intrauterine fetal demise.76,80,83 Differential Considerations and Management. Exclusion of more serious entities, such as viral hepatitis, acute fatty liver, drug-induced cholestasis, or complicated cholecystitis, is required. Outpatient management is appropriate, provided the diagnosis is clear and the patient has close obstetric follow-up. Some authors advocate aggressive fetal surveillance and delivery after fetal lung maturity to improve fetal outcome.76 Sympto matic treatment with antihistamines, ursodeoxycholic acid, bile salts, guar gum, benzodiazepines, and other medications has been tried with variable success.80,84
Hyperemesis Gravidarum Nausea and vomiting are common in pregnancy, particularly from 6 to 20 weeks of gestation. Hyperemesis gravidarum is defined as nausea and vomiting that causes starvation metabolism, weight loss, dehydration, and prolonged ketonemia and ketonuria. It occurs in a small minority of pregnant patients. The cause of hyperemesis gravidarum is not clear; associations have been made with increasing estradiol and hCG levels, as well as with maternal cytokines.85,86 Several studies have suggested an increased infection rate with Helicobacter pylori in patients with hyperemesis gravidarum.87-89 Initial management involves rehydration with intravenous fluids, antiemetics, and demonstration of ability to take oral hydration. Patients may require enteral nutrition. Most standard antiemetics are in Food and Drug Administration category C and are used successfully to treat hyperemesis gravidarum. A short course of oral prednisolone has been reported to be therapeutic for intractable hyperemesis.90,91 Oral vitamin B6 has also been reported to be helpful.89 Bilirubin and alkaline phosphatase levels can be mildly elevated but should return to normal levels after delivery. Hyperemesis may be complicated by liver disease and abnormal liver function tests, which are expected to resolve with supportive treatment.92
Thromboembolic Disease in Pregnancy Principles of Disease Thromboembolic disease accounts for almost 20% of obstetric mortality, making it the leading cause of death in pregnancy.16 Pregnancy is a hypercoagulable state, with increased coagulation factors and stasis as pregnancy progresses and significant vascular trauma at the time of delivery. The risk of venous thrombosis increases during pregnancy to five or six times that of nonpregnant women. Although the risk is increased throughout pregnancy, it is highest during the puerperium. Women who smoke, are overweight, are older than 35 years, have varicose veins, or have a prior superficial venous thrombosis or history of a hypercoagulable state, as well as women who deliver prematurely or have postpartum hemorrhage, are at higher risk.93-96
Clinical Features As in nonpregnant patients, clinical signs of pain, tenderness, and swelling are poor predictors of deep vein thrombosis in pregnancy. The clinical diagnosis of pulmonary embolus is likewise difficult. Although tachypnea, tachycardia, dyspnea, and pleuritic pain are commonly associated with pulmonary embolus, the symptoms are nonspecific and may be associated with such diverse diseases as hepatic inflammation, pyelonephritis, and diaphragmatic impingement from a normal gravid uterus.
Diagnostic Strategies An arterial blood gas analysis should be undertaken, although it may be difficult to interpret. Arterial blood gases in pregnancy normally show a respiratory alkalosis from progesterone-induced respiratory stimulation, and the alveolararterial (A-a) oxygen difference is normal (the pregnant normal A-a gradient can be as high as 20 mm Hg) in the majority of patients with pulmonary embolus.97 A chest radiograph (shielding the pelvis and uterus) should be obtained to exclude other disease processes that may mimic a pulmonary embolus. The diaphragm is normally symmetrically elevated during late pregnancy. Noninvasive impedance plethysmography is highly accurate for the exclusion of deep vein thrombosis but lacks sensitivity for the exclusion of nonobstructive thrombi. Due to its widespread availability, Doppler ultrasonography has superseded noninvasive impedance plethysmography as the first-line test for the diagnosis of deep venous thrombosis. Both tests, however, are useful in the diagnosis of deep vein thrombosis of the femoral or popliteal veins in pregnancy, and these studies provide the least risk to the patient. Abnormal flow study results can be found in the normal patient studied in the supine position during late pregnancy, so positive results should be confirmed with the patient positioned on her left side. An unequivocally abnormal flow study finding is sufficient reason to treat the pregnant patient in most cases. However, normal leg study results can be seen with isolated iliac vein disease, which is common in pregnancy and requires imaging with magnetic resonance imaging or CT for diagnosis. If thromboembolic disease is suspected, serial indirect Doppler testing or CT may be required.98 The risk of anticoagulation usually outweighs the risk of definitive studies when the diagnosis is equivocal. Technetium-labeled ventilation-perfusion scans expose the fetus to less than 50 mrad of radiation, making them safe in all trimesters. One study found that only 2% of pregnant women
Management Warfarin (Coumadin) is contraindicated during pregnancy because of its teratogenic effects and high risk of abortions and fetal hemorrhage. Heparinoids are used to treat thromboembolic disease during pregnancy. Unfractionated heparin carries a poorly understood risk of fetal osteoporosis, thrombocytopenia, prematurity, or miscarriage. In general, acute anticoagulation with intravenous heparin is followed by subcutaneous heparin twice daily, usually continued for 3 to 6 months postpartum in patients who have deep vein thrombosis or pulmonary embolus during pregnancy. Patients receiving this treatment require laboratory testing every 1 or 2 weeks, and the efficacy of anticoagulation may be variable in pregnancy. Low-molecular-weight heparin is considered safe in pregnancy and offers several advantages over unfractionated heparin: decreased bleeding risk, reliable pharmacokinetics, decreased risk of heparin-induced thrombocytopenia, fixed dosages, less frequent dosing, and decreased risk of osteoporosis and thrombocytopenia.101-105 In patients with a history of deep vein thrombosis or pulmonary embolus, prophylaxis for subsequent gestations is usually recommended.106-108
Genitourinary Infections Urinary Tract Infection Principles of Disease. Although the risk of asymptomatic bacteruria (9%) does not increase in pregnancy, it appears that pregnancy predisposes the patient to develop symptomatic lower and upper tract genitourinary infections.109,110 Uterine pressure exerted on the bladder and ureters, poor emptying of the bladder with voiding, and progesterone-induced smooth muscle relaxation that inhibits ureteral peristalsis all appear to contribute to increased risk of infection during pregnancy. Identification of patients with asymptomatic bacteruria by prenatal screening in early pregnancy identifies approximately 95% of individuals at risk for subsequent bacteruria during the pregnancy. Because up to 30% of women who have asymptomatic bacteruria will develop pyelonephritis if untreated, treatment of bacteruria is cost-effective and important.111,112 Clinical Features and Diagnostic Strategies. The pregnant patient who comes to the emergency department with lower urinary tract symptoms (e.g., dysuria, frequency, and urgency) or upper tract symptoms (e.g., fever, malaise, or back pain) should have a pelvic examination and evaluation of an uncontaminated urine specimen (preferably catheterized). There is a predominance of right-sided symptoms during pregnancy, probably the result of increased mechanical forces on the right ureter, but left-sided flank pain or bilateral symptoms may be caused by pyelonephritis. Rarely, urinalysis may yield normal results or cultures may produce negative findings, either because of failure to report lower colony counts or because of complete obstruction of the involved ureter.
2295
The major risk of asymptomatic and lower urinary tract infection is spread to the renal parenchyma. Acute pyelonephritis carries considerable morbidity in pregnancy, including maternal sepsis, permanent renal injury, and premature labor.112 The risk of prematurity can be minimized by effective treatment and continued monitoring for recurrence. The development of premature labor in the woman who has pyelonephritis is ominous and can be prevented only by aggressive recognition and treatment earlier in pregnancy. Differential Considerations. Vaginitis, herpes genitalis, chlamydial infection of the urethra, or ovarian torsion can masquerade as urinary tract symptoms. A history of external dysuria (burning at the perineum with urination) suggests herpes or vaginitis. A pelvic examination should be performed to obtain cervical cultures and to identify perineal or vaginal causes of dysuria. Appendicitis, cholecystitis, pancreatitis, and liver diseases in pregnancy must be considered in the differential diagnosis of upper urinary tract infection. Back pain may also be a sign of premature labor. Careful evaluation of an uncontaminated catheterized urine specimen is essential to making the correct diagnosis. Management. Patients with asymptomatic bacteruria or lower urinary tract signs and symptoms should be treated with 7 to 10 days of an antibiotic that is active against usual urinary pathogens and is safe in pregnancy.112,113 The most common choices are a cephalosporin, a nitrofurantoin, or a sulfonamide (except during the third trimester). Single-dose therapy for these infections during pregnancy has been proposed but may not be appropriate in an emergency department population with questionable follow-up and a relatively high incidence of occult upper urinary tract infection. Patients with fever, back pain, and evidence of acute pyelonephritis in pregnancy are usually admitted for intravenous antibiotic administration, although outpatient parenteral therapy can be effective and safe in selected patients.112,114 In such cases, aggressive intravenous hydration, obstetric consultation, and urine cultures should be initiated. At least one parenteral dose of antibiotics should be given, with antibiotic coverage guided by known organism susceptibilities in a given hospital. Because the resistance of Escherichia coli to ampicillin is considerable in most regions, cephalosporin or a combination of a penicillin and an aminoglycoside (which must be carefully monitored because of variable clearance by infected kidneys) is the usual intravenous medication administered. Cultures must be performed to ensure that the original choice of antibiotic was correct, and the patient must have a repeat culture and be followed closely after treatment.
Vaginitis Bacterial Vaginosis. Bacterial vaginosis (formally known as Gardnerella vaginitis or Haemophilus vaginalis vaginitis) is an overgrowth of multiple endogenous vaginal bacteria, in some cases producing excessive discharge and vaginal malodor. Prevalence rates for bacterial vaginosis in pregnancy are estimated at 15 to 20%. Bacterial vaginosis is associated with an increased risk of chorioamnionitis, subclinical PID, premature rupture of membranes, fetal prematurity, and postpartum endometritis after vaginal delivery. Therapy during the second trimester is recommended even when the patient is asymptomatic to prevent the sequelae of premature rupture of membranes. Management includes a 7-day course of metronidazole or a 7day course of clindamycin. Intravaginal treatment is not recommended in pregnant patients.113,115 Candida albicans Vaginitis. The incidence of vulvovaginal candidiasis is increased during pregnancy by high levels of estrogen and other steroids. It is not increased in the pregnant patient
Chapter 176 / Acute Complications of Pregnancy
had high-probability scans; no adverse events occurred during follow-up in 104 women who were not heparinized and had normal or nondiagnostic scans.99 Helical CT scanning provides an alternative for the diagnosis of pulmonary embolus in pregnancy. The average fetal radiation dose in helical CT scans is less than that from ventilation-perfusion scans, making it a potentially attractive alternative, although studies of its accuracy in pregnancy have not been done.100 A pulmonary angiogram may be required if the diagnosis of pulmonary embolus is unclear after less invasive studies have been performed.
PART V ■ Special Populations / Section Two • The Pregnant Patient
2296
who has recently been on antibiotics,116 and there is no association of Candida colonization with adverse pregnancy outcomes.117 Oral azoles are contraindicated in pregnancy because of an association with adverse fetal outcomes.118 Treatment with vaginal azoles for 7 days during pregnancy is considered safe, with an estimated 85 to 100% cure rate.113 Recurrent disease may require a vaginal culture to confirm diagnosis and to identify unusual candidal species (e.g., Candida glabrata) that may be resistant to conventional treatment. Longer treatment or treatment of a potential Candida reservoir in the patient’s partner may also be required. Trichomoniasis. Trichomoniasis is a sexually transmitted vaginitis caused by a protozoan parasite, Trichomonas vaginalis. Of patients who have trichomoniasis, 50% are asymptomatic. Symptoms include vaginal itching, malodorous discharge, or vaginal irritation. Diagnosis is made by direct visualization or protozoans on wet mount. The recommended treatment is metronidazole, a one-time dose of 2 g, for symptomatic patients only.113 The organism is rarely aggressive during pregnancy but is associated with adverse pregnancy outcomes if untreated.119
Sexually Transmitted Diseases in Pregnancy Sexually transmitted diseases are treated in pregnant patients according to the latest Centers for Disease Control and Prevention guidelines. In general, the tetracyclines and quinolones are contraindicated in pregnant patients. Treatment of genital tract infections may be important in preventing preterm labor and the morbidity of prematurity. Chlamydia trachomatis Infection. Chlamydia trachomatis infection is the most common sexually transmitted disease, both in the United States and worldwide. Its prevalence is currently three to five times that of Neisseria gonorrhoeae infection.113 Clinical diagnosis is difficult during pregnancy because cervical mucus is usually cloudy and contains white blood cells. Routine chlamydia screening during pregnancy is important to prevent complications of preterm labor and postpartum endometritis, both of which are more common in patients who have chlamydial cervical infections.120 Chlamydial infections of infants born to infected mothers include conjunctivitis and pneumonitis. Treatment during pregnancy or breast-feeding is azithromycin (single 1-g dose), which improves compliance and decreases gastrointestinal side effects.121 Treatment with a 7-day course of erythromycin base or amoxicillin is an acceptable alternative. Tetracyclines are contraindicated in pregnancy.113 Herpes Simplex Infection. Herpes simplex virus infections pose a significant risk in pregnancy to both the mother and the newborn. Women who have genital herpes during the third trimester have a 30 to 50% increased risk of transmission compared with women with herpes simplex virus infection in the first trimester (1%). The virus can be transmitted prenatally via transplacental infection or ascending vaginal infection and via vaginal delivery, particularly when herpetic lesions are present. Infections in the neonate often are disseminated or involve the central nervous system, causing significant morbidity and mortality. In the emergency department, culturing of new suspected herpetic lesions of the cervix, vagina, or perineum
identifies patients at risk for perinatal complications. Although the risk of oral acyclovir and valacyclovir use in pregnancy is not well-known, it is recommended for first-episode genital herpes. Suppressive therapy can reduce the need for cesarean section in women whose first clinical episode of genital herpes simplex virus occurred during pregnancy but may not eliminate the need for cesarean section in women with recurrent herpes simplex virus.113,122-124 Treatment should be undertaken with obstetric consultation and careful patient monitoring. Neisseria gonorrhoeae Infection. Gonococcal infection of the cervix occurs during pregnancy in 1% of women. Symptoms are similar to those in nonpregnant women. Salpingitis is rare but may develop during the first trimester from upper genital extension of cervical infection. Some practitioners believe that the incidence of the disseminated infection is increased in pregnant patients because of elevated progesterone levels and increased vascularity in the area of the cervix. Gonococcal arthritis is the most common manifestation of gonococcal dissemination. Diagnosis and treatment of gonococcal infections are unchanged by pregnancy, and treatment includes cephalosporins or spectinomycin.113 Treatment for possible coexistent chlamydial infection is recommended for pregnant and nonpregnant women. The major complications of thirdtrimester gonococcal infection are neonatal gonococcal ophthalmia and sepsis.113,119
Upper Genital Tract Infection Pelvic Inflammatory Disease. Pelvic inflammatory disease is very rare in pregnancy and does not occur after the first trimester. Differential diagnosis includes ectopic pregnancy, septic abortion, and appendicitis, all of which are more common. In the patient with suspected infection, smears or cultures for chlamydia and gonorrhea should be obtained. Given the risk of endometrial infection in pregnancy and the need to consider other diagnoses, pregnant patients who have suspected PID require hospital admission and intravenous antibiotics.113 Chorioamnionitis. Chorioamnionitis is the infection or inflammation of the placenta and fetal membranes. After 16 weeks of pregnancy, the chorioamniotic membranes adhere to the cervical os and may become infected. The risk is increased in women with preterm labor. Chorioamnionitis is diagnosed by the findings of fever, maternal and fetal tachycardia, and uterine tenderness in a patient past 16 weeks of pregnancy. Leukocytosis can be suggestive of chorioamnionitis but is not diagnostic. The diagnosis is confirmed by amniocentesis. Patients should have blood cultures drawn. Vaginal and cervical cultures for group B strep, E. coli, chlamydia, and gonorrhea should also be obtained. Urgent obstetric consultation should be obtained, and hospital admission for intravenous antibiotics is required. Patients are usually treated with intravenous ampicillin and gentamicin. Vancomycin, clindamycin, or erythromycin may be substituted in the penicillin allergic patient.125,126
Acknowledgment We thank Dr. Jean Abbott for her work on prior editions of this chapter.
2297
KEY CONCEPTS The emergency physician must determine the type of miscarriage (incomplete, complete, inevitable, and threatened) because management priorities for these patients are different. ■ The history and the physical examination of the patient with ectopic pregnancy are insensitive and nonspecific; therefore, ancillary studies are required to locate the pregnancy in any emergency department patient who has abdominal pain or vaginal bleeding and a positive pregnancy test result. ■ The major conditions associated with vaginal bleeding in the second half of pregnancy include abruptio placentae
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
and placenta previa. Patient history, physical examination, and results of ultrasonography must be used together by the emergency physician to distinguish these conditions. ■ Appendicitis is the most common surgical emergency in pregnancy, and the presentation of these patients may not be classic, leading to a misdiagnosis rate of 30 to 35% overall in pregnant patients. ■ The treatment of eclampsia includes magnesium sulfate for seizures, reduction of blood pressure, and delivery of the fetus.
Chapter 176 / Acute Complications of Pregnancy
■
Chapter 177
Chronic Medical Illness during Pregnancy
Kirsten K. Calder and Edward J. Newton
This chapter focuses on the care of the pregnant patient with selected chronic medical illnesses. Recognition of chronic disease is an important aspect of screening in these patients and can prevent adverse outcomes for both the mother and the fetus.
■ PRINCIPLES OF DISEASE The physiologic changes that occur in pregnancy may exceed the patient’s underlying compensatory mechanisms, resulting in initial symptom onset or rapid decompensation of medical illness during pregnancy. Certain chronic medical conditions also pose a serious threat to the mother’s health or result in a poor fetal outcome. Finally, some medical illnesses result in a difficult delivery or the need for special resuscitation measures in the neonate. The incidence of pregnancy in chronically ill patients is increasing because of improved survival of patients with diseases such as diabetes, epilepsy, renal failure, and various cancers. Also, the demographics of pregnancy are changing in that maternal age at the time of first pregnancy is increasing. Advances in assisted reproduction, including in vitro fertilization and oocyte donation, have made it possible for older women—including those who are postmenopausal—to become pregnant. Older pregnant women experience an increased rate of antepartum and intrapartum complications and are more likely to have comorbid conditions such as cardiovascular disease.1
■ SPECIFIC DISORDERS Asthma Asthma is the most common pulmonary problem in pregnancy, affecting up to 8.4% of gestations, and the prevalence appears to be increasing.2 The overall effect of pregnancy on asthma seems to be that one third of patients will experience a worsening of the disease, one third will experience improvement, and the other third remain unchanged. The exact effect of asthma on the fetus remains controversial. Some studies show an increased risk of complications including preeclampsia, gestational diabetes mellitus, prematurity, and intrauterine growth retardation.3,4 Several changes in respiratory physiology during normal pregnancy affect the management of asthma in this context; both the tidal volume and the minute ventilation rise, increasing up to 50% by term. This increase in minute ventilation, 2298
or “hyperventilation of pregnancy,” lowers the resting Pco2 by an average of 8 to 10 mm Hg to a Pco2 of 32 mm Hg. In response to chronic hyperventilation, there is a compensatory increase in the renal excretion of bicarbonate, with serum levels averaging 19 mEq/mL. The net result is a slight respiratory alkalosis, with a serum pH ranging from 7.40 to 7.45. It is important to keep these “normal” values in mind when gauging the severity of an asthma exacerbation in the pregnant patient. The treatment strategy in the care of the pregnant asthmatic patient is to achieve normal pulmonary function, prevent acute exacerbations, and minimize use of short-acting rescue betaagonists. Acute asthma exacerbations pose a significant maternal and fetal risk, and the pregnant patient should be managed as aggressively as the nonpregnant patient.5 In fact, patients whose asthma is well controlled are generally expected to have normal pregnancies, and the risks from inadequately treated disease outweigh the risks of therapy.5,6 The inhaled beta-agonist agents are generally considered as safe in pregnancy (Tables 177-1 and 177-2).5,7 As in the nonpregnant patient, inhaled short-acting beta-agonists are firstline agents for acute asthma exacerbations. Long-acting agonists are also recommended for maintenance use in patients with moderate or severe disease.5 Although safe from a teratogenic perspective, these medications are powerful tocolytics and will usually halt active labor. Selective agonists are preferred because nonselective agents such as epinephrine theoretically decrease uteroplacental blood flow via beta-mediated effects. Corticosteroids remain a core component of maintenance treatment for asthma in the pregnant patient. In fact, current data support the safety of inhaled corticosteroids,7 and their chronic use is recommended in all asthmatic patients except those with mild intermittent disease.5 Use of systemic corticosteroids is sometimes necessary on a chronic basis in patients with severe persistent asthma and as a short burst in patients with exacerbations that exhibit inadequate or poor response to initial beta-agonist therapy. However, corticosteroids have a possible association with cleft deformities, preterm delivery, low birth weight, and preeclampsia.3,5,7-10 These complications may also be related to uncontrolled asthma, but inhaled delivery is preferred over systemic dosing. The methylxanthine theophylline has been shown to be safe in pregnancy at therapeutic doses and is an acceptable alternative medication for maintenance therapy.5,7,11 There are limited data for chromolyn sodium and the newer leukotriene receptor antagonists (montelukast and zafirlukast),12 but all are
2299
Table 177-1 Gestational Effects and Treatment of Medical Illnesses during Pregnancy GESTATIONAL CONCERNS
TREATMENT
Asthma
Fetal—IUGR, PTD, hypoxia, meconium-stained amniotic fluid, fetal loss Maternal—preeclampsia, gestational hypertension, gestational diabetes, hyperemesis gravidarum, need for labor induction
Acute coronary syndrome
Fetal—perinatal death, PTD Maternal—uterine hemorrhage and abruption (related to use of antiplatelet, antithrombotic, and fibrinolytic agents)
Valvular heart disease
Fetal—perinatal death, PTD Maternal—decompensated heart failure, thromboembolism, death
Hypertension
Fetal—perinatal death, IUGR, PTD Maternal—progression of target end-organ damage, superimposed preeclampsia, abruption Fetal—low birth weight, PTD, low fetal iron stores, fetal loss (with severe anemia) Maternal—preeclampsia, highoutput heart failure (rare); effects on maternal mortality unclear Fetal—fetal loss, IUGR, PTD, and premature rupture of membranes Maternal—↑ need for cesarean section, preeclampsia, infection, heart failure, pulmonary infarction, ↑ incidence of painful crises; maternal mortality low with treatment; false-positive Apt and Kleihauer-Betke tests 2° to persistent hemoglobin F Fetal—various congenital malformations associated with AEMs, fetal hypoxia and bradycardia, fetal loss Maternal—variable changes in seizure frequency; alterations in AEM levels; increased seizure frequency secondary to voluntary medication noncompliance; abruption, anemia, hyperemesis gravidarum, preeclampsia, need for labor induction and cesarean section, premature rupture of membranes
Treatment of acute exacerbations is the same as for the nonpregnant patient. Fetal monitoring for exacerbations during the third trimester is recommended even in the absence of maternal hypoxia. Maintenance therapy also unchanged with the following precautions: Corticosteroids: inhaled agents preferred to limit side effects, but oral route may be necessary with severe, persistent disease; patients on long-term steroids require “stress dose” hydrocortisone during labor and delivery Methylxanthines: safe but debatable benefit; use only in refractory disease; reduced clearance during pregnancy may result in maternal toxicity and fetal tachycardia Leukotriene receptor antagonists: avoid zileuton Standard medical therapy is the same as for the nonpregnant patient, although antiplatelet, antithrombotic, and fibrinolytic agents should be avoided when delivery is imminent. Avoid maternal hypotension when using nitrates—may result in fetal distress. Avoid beta-blockers in the first trimester (teratogenic). Coordinate definitive care (fibrinolytics vs. PCI) with cardiology. Subcutaneous heparin therapy indicated for patients with prosthetic valves and atrial fibrillation. Avoid warfarin. Mitral stenosis: diuresis and beta-blockade; valvuloplasty or open cardiac surgery for severe symptomatic disease; consider pregnancy termination in patients with atrial fibrillation or severe stenosis Mitral and aortic regurgitation: diuresis in patients with pulmonary congestion; surgical therapy for acute regurgitant lesions; consider pregnancy termination in patients with symptomatic disease Aortic stenosis: avoid hypotension and the supine hypotensive syndrome— position left lateral decubitus; vigorous fluid replacement during delivery; valvuloplasty or open cardiac surgery for severe symptomatic disease Fetal cyanide poisoning might develop after several hours of sodium nitroprusside—avoid prolonged infusions. The most commonly used agents include methyldopa, labetalol, and hydralazine. Avoid beta-blockers in the first trimester (teratogenic). Avoid ACEIs and ARBs. Oral iron supplementation is indicated to improve maternal iron stores. Several over-the-counter preparations are available. There is a delay from onset of therapy to increase in serum hemoglobin. Parenteral iron replacement is safe and effective, although rarely required. Transfusion is rarely required but may be necessary with severe symptomatic anemia.
Iron-deficiency anemia
Sickle cell anemia
Epilepsy
Management of pain crises and infections is the same as for the nonpregnant patient, with rest, hydration, narcotic analgesia, supplemental oxygen, and antibiotics as indicated. Narcotic analgesics should not be withheld, but the need for neonatal respiratory support should be anticipated if delivery is imminent. Prophylactic transfusion is not indicated. Fetal monitoring and assessment of fetal well-being are indicated for viable pregnancies. Chronic maintenance care includes pneumococcal vaccination, supplemental folate, and usually iron. Hydroxyurea has been discouraged during pregnancy, but few adverse fetal effects have been reported in humans. Management of status epilepticus is the same as for the nonpregnant patient. Maintenance therapy should be coordinated by the patient’s neurologist (or primary care practitioner) and obstetrician—in general, a single AEM given at the lowest effective dose is recommended. Patients with pregravid seizure control with AEM monotherapy should continue their current regimen. There is limited experience with newer AEMs, but these are likely to have a more favorable side effect profile. Folate supplementation is mandatory for patients taking older AEMs. Administer oral vitamin K to the mother during the last month of pregnancy and parenteral vitamin K to the newborn.
Chapter 177 / Chronic Medical Illness during Pregnancy
MEDICAL ILLNESS
Table 177-1 Gestational Effects and Treatment of Medical Illnesses During Pregnancy—cont’d MEDICAL ILLNESS
GESTATIONAL CONCERNS
TREATMENT
Myasthenia gravis
Fetal—transient neonatal myasthenia syndrome Maternal—variable changes in disease severity; arrest of labor; disease exacerbation in the postpartum period
Diabetes mellitus— insulindependent, non-insulindependent, and gestational
Fetal—congenital malformations; macrosomia; IUGR; fetal loss; neonatal hypoglycemia, jaundice, hypomagnesemia, and hypocalcemia Maternal—preeclampsia, “brittle” diabetes, ↑ need for cesarean section; mothers with GDM have ↑ risk for development of postpartum diabetes
Hyperthyroidism
Fetal—PTD, low birth weight, fetal thyroid dysfunction, fetal loss Maternal—preeclampsia, heart failure
Hypothyroidism
Fetal—congenital malformations, low birth weight, fetal loss, fetal thyroid dysfunction, and goiter Maternal—preeclampsia, abruption, postpartum hemorrhage, ↑ need for cesarean section Fetal—fetal loss, low birth weight, PTD; fetal and neonatal tuberculosis Maternal—preeclampsia; potential for delayed diagnosis and treatment out of concern for the fetus
Management is the same as for the nonpregnant patient. Ventilatory support is the most important aspect of therapy; note that patients with myasthenia gravis are relatively resistant to depolarizing paralytic agents—higher doses may be required. Pyridostigmine therapy can be continued. Patients on maintenance corticosteroids require “stress dose” hydrocortisone during labor and delivery. Plasmapheresis is safe during pregnancy. Every attempt should be made to maintain maternal serum glucose of 100 mg/dL. Note that insulin requirements decrease during the immediate postpartum period and the mother may not need insulin for 24–48 hours after delivery. Insulin therapy—using either intermittent dosing or continuous subcutaneous infusion—is standard care for patients with both IDDM and NIDDM. Management of DKA is the same as for the nonpregnant patient with the addition of an assessment of fetal well-being and continuous fetal heart monitoring. GDM is often manageable with diet alone but may require insulin therapy when fasting plasma glucose remains >105 mg/dL. Management of thyroid storm is the same as for the nonpregnant patient and includes a search for the underlying precipitant. Therapy for hyperthyroidism in the absence of thyroid storm: Reversal of sympathetic effects: propanolol in standard doses is useful until thyroid hormone synthesis has been blocked by thioamides Thioamides: both propylthiouracil and methimazole at the lowest effective dose are acceptable Surgical therapy: thyroidectomy is useful in refractory cases Other: avoid iodide if possible; hydrocortisone decreases peripheral conversion of T4 to the more active T3 and can be used during pregnancy Radioactive iodine is absolutely contraindicated. Maintenance therapy includes levothyroxine 0.15 mg/day. Appropriate treatment prevents adverse obstetric and fetal outcomes. Myxedema coma is rare, but when present, treatment is the same as for the nonpregnant patient.
Tuberculosis
HIV/AIDS
Fetal—HIV infection, PTD, fetal loss; neonatal drug withdrawal if the mother uses intravenous drugs Maternal—postpartum endometritis, uterine bleeding (in the setting of thrombocytopenia)
Syphilis
Fetal—congenital syphilis, fetal loss, PTD, IUGR, nonimmune hydrops
PPD positive/chest radiograph negative: 6- to 9-month course of isoniazid (starting after the first trimester) for patients with recent conversion (2 years may defer treatment until after delivery but should still be offered a course of isoniazid since it is safe in pregnancy Active tuberculosis: 9-month course (starting immediately) of isoniazid plus rifampin or ethambutol Multidrug-resistant tuberculosis: warrants aggressive therapy without regard to potential teratogenicity Pyridoxine is mandatory for all patients receiving isoniazid. Antiretroviral therapy: 1. Highly active antiretroviral therapy (HAART) should be offered to all pregnant patients with HIV and viral load >1000. The HAART regimen should include zidovudine (AZT) to prevent vertical transmission of the virus. 2. There are specific HAART drug-related concerns during pregnancy— decisions regarding therapy are best made by appropriate specialists. 3. AZT monotherapy is not recommended except in those patients with a low viral load who do not wish to take HAART. In these cases, AZT is appropriate to reduce disease transmission. Cesarean section is recommended with viral load >1000. Opportunistic infections require standard therapies despite the potential fetal effects. Primary, secondary, early latent (1 year or unknown duration): BPG, 2.4 million units IM weekly × three doses Neurosyphilis: aqueous penicillin G, 2–4 million units IV q4h × 10–14 days or procaine penicillin G, 2.4 million units IM, and probenecid, 500 mg PO q6h × 10–14 days
ACEI, angiotensin-converting enzyme inhibitor; AEM, antiepileptic medicine; AIDS, acquired immunodeficiency syndrome; ARB, angiotensin II receptor blocker; DKA, diabetic ketoacidosis; GDM, gestational diabetes mellitus; HIV, human immunodeficiency virus; IDDM, insulin-dependent diabetes mellitus; IUGR, intrauterine growth retardation; NIDDM, non-insulin-dependent diabetes mellitus; PCI, percutaneous coronary intervention; PTD, preterm delivery.
2301
Table 177-2 Drugs Used in the Treatment of Acute Asthma Exacerbations during Pregnancy EXAMPLES
DOSAGE
COMMENTS
Inhaled beta-agonists
Albuterol Levalbuterol
2.5–5 mg every 20 min 1.25–2.5 mg every 20 min
Injectable beta-agonists
Epinephrine
Systemic corticosteroids
Terbutaline Prednisone Prednisolone
0.3–0.5 mg SC (1 : 1000 or 1 mg/mL) every 20 min 0.25 mg SC (1 mg/mL) every 20 min Dosage applies to all preparations Initial inpatient therapy: variable dosing; need at least 120–180 mg/day Outpatient burst therapy: 40–60 mg/day for 3–10 days 0.5 mg every 20 min
Inhaled beta-agonists first-line therapy May also be administered via MDI Up to three doses in first hour; continuous use in severe exacerbations No proven benefit over inhaled dosing Up to three doses in first hour
Methylprednisolone Inhaled anticholinergics
Ipratropium bromide
Smooth muscle relaxants
Magnesium sulfate
No benefit of intravenous dosing over oral except in patients with impending respiratory failure who can’t take oral medications
Not first-line therapy; should be used with beta-agonists Consider for use in patients with severe exacerbations Limited data on its use for asthma in pregnancy
From NAEPP Expert Panel Report: Managing asthma during pregnancy: Recommendations for pharmacologic treatment—2004 update. J Allergy Clin Immunol 115:34, 2005.
Table 177-3 Hypertensive Disorders of Pregnancy
Definition
CHRONIC HYPERTENSION
GESTATIONAL HYPERTENSION
1. Hypertension* that antedates pregnancy
Hypertension diagnosed after 20 weeks of gestation in the absence of proteinuria or other evidence of preeclampsia
PREECLAMPSIA
Hypertension that begins after 20 weeks of gestation occurring in association with new-onset proteinuria (≥300 mg/24 hr)
2. Hypertension diagnosed prior to 20 weeks of gestation
Comment: rarely, preeclampsia presents prior to 20 weeks of gestation
CHRONIC HYPERTENSION WITH SUPERIMPOSED PREECLAMPSIA
1. Hypertension that antedates pregnancy in association with new-onset proteinuria
2. Sudden increase in proteinuria in a woman with chronic hypertension* and proteinuria prior to 20 weeks of gestation 3. Hypertension that antedates pregnancy in association with sudden increase in blood pressure 4. Hypertension that antedates pregnancy in association with decreased platelets or elevated liver transaminases
Comment: may progress to preeclampsia; may also represent previously undiagnosed hypertension
*Defined as blood pressure ≥140 mm Hg systolic or ≥90 mm Hg diastolic. From Roberts JM, et al: Summary of the NHLBI working group on research on hypertension during pregnancy. Hypertension 41:437, 2003; and the National Institutes of Health: The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. NIH Publication No. 04-5230, August 2004.
considered class B (presumed safety) during pregnancy. Magnesium sulfate has received a good deal of attention in the acute treatment of refractory asthma exacerbations because it has a wide safety profile and potent bronchodilator properties. Although large doses of magnesium have been used for the treatment of eclampsia, at this time there are insufficient data to support routine use of magnesium in this patient population. Perhaps the most important tenet of treatment to recognize is that if the mother is hypoxic, the fetus is hypoxic as well. In addition, the fetus is more sensitive to hypoxia; thus, a normal maternal oxygen saturation does not preclude fetal distress. Supplemental oxygen should be administered to all pregnant patients with an acute asthma exacerbation.
Hypertension Chronic Hypertension and Hypertensive Emergencies Chronic hypertension is defined as elevated blood pressure that is present before the onset of pregnancy or that begins prior to the 20th week of gestation. It affects up to 5% of pregnancies and represents a significant source of maternal and fetal mortality and morbidity (see Table 177-1).13,14 The emergency physician may be required to provide therapy for hypertensive emergencies in patients with chronic hypertension and to differentiate between the various hypertensive disorders of pregnancy (Table 177-3). In the treatment of chronic hypertension in pregnancy, the physician must balance the goal of reducing maternal blood
Chapter 177 / Chronic Medical Illness during Pregnancy
PHARMACOLOGIC CLASS
PART V ■ Special Populations / Section Two • The Pregnant Patient
2302
pressure with the requirements to maintain cardiac output and minimize adverse effects for both mother and fetus. Precipitous and marked decreases in blood pressure may significantly diminish uteroplacental blood flow. Treatment of patients with mild disease does not have an appreciable effect on perinatal outcomes, nor does it reduce the incidence of superimposed preeclampsia. In addition, these patients are likely to have a decrease in blood pressure without pharmacologic agents due to normal physiologic changes that occur in pregnancy, and they may be successfully managed without medications.14-16 On the contrary, patients with more severe hypertension, patients with evidence of end-organ damage, and patients on more than one pregestational antihypertensive medication probably need maintenance medical therapy (see Table 177-1).15,16 Lifestyle modification is a potential mode of therapy, but it is unclear whether dietary and weight restrictions are safe in the gravid patient.14 Decisions regarding maintenance therapy are most appropriately made by the patient’s primary care practitioner in conjunction with her obstetrician. However, the emergency physician may be called upon to make treatment decisions for patients with severely elevated blood pressure and for patients with superimposed eclampsia. Nearly all of the major classes of antihypertensive agents are acceptable in the pregnant patient, with the exception of angiotensin enzyme inhibitors and angiotensin II receptor blockers. Diuretics are also considered second-line agents because of concerns regarding plasma volume constriction (see Table 177-1).15,16 Hydralazine and labetalol are the agents most commonly used for hypertensive emergencies associated with eclampsia and are also appropriate for such emergencies in the patient with chronic hypertension. Sodium nitroprusside can cause fetal cyanide toxicity after several hours of infusion and is considered a second-line agent (Table 177-4).15 Diagnosis of preeclampsia in the pregnant patient with chronic hypertension is challenging but necessary (see Chapter 176). Patients with chronic hypertension are more likely to develop preeclampsia, a situation that results in increased morbidity and mortality compared with either process in isolation. Coexistence of the two disorders should be suspected in the following situations: (1) new onset of proteinuria after 20 weeks of gestation in a patient with known hypertension and (2) pregnant patients with hypertension and proteinuria prior to 20 weeks of gestation who develop thrombocytopenia, increased transaminase levels, or an acute increase in proteinuria or blood pressure (see Table 177-3).14
Pulmonary Hypertension Primary (idiopathic) pulmonary hypertension, Eisenmenger’s syndrome (pulmonary hypertension associated with intracardiac left-to-right shunts), and secondary vascular pulmonary
hypertension have extremely high mortality rates in association with pregnancy, ranging from 30 to 52%. The majority of deaths are a result of heart failure.17 Unfortunately, mortality is unaffected by peripartum management in most cases, although some patients benefit from selective pulmonary artery vasodilators such as prostacyclin and inhaled nitric oxide.17 The primary goals of the emergency physician are to ensure high right ventricular filling pressures by maintaining adequate volume status and to obtain early consultation with obstetrics and cardiology. Patients with early gestations should be referred for elective pregnancy termination.
Cardiac Disorders Acute Coronary Syndromes Coronary artery disease is rare in pregnant women, with a population-based study noting a diagnosis of acute myocardial infarction (AMI) in up to 6.2 per 100,000 deliveries.18 Prior mortality rates were reported to be approximately 20%, but more recent studies reveal rates ranging from 5.1 to 7.3%.18-20 Normal physiologic changes of pregnancy such as increased cardiac output and reduced oxygen-carrying capacity secondary to physiologic anemia have the potential to exceed the threshold for angina if a fixed coronary artery stenosis is present. Certain conditions are also associated with an increased risk of pregnancy-related AMI, including advanced maternal age, thrombophilia, hypertensive disorders, anemia, diabetes mellitus, and tobacco use.18,19 AMI can occur anytime during the gestational period but peaks during the last trimester and peripartum period. Twenty-seven to 41% of events occur in the 6 weeks after delivery.18,19 Because increases in cardiac output peak during labor and delivery, maternal mortality from AMI is higher intrapartum than during the antepartum or postpartum periods.19 A significant percentage of pregnant women with AMI have normal coronary arteries—reported to be 29% of those patients who underwent angiography.20 Other pathophysiologic mechanisms besides atherosclerosis that cause AMI in the pregnant patient include coronary artery dissection, coronary artery aneurysm, and vasospasm. The relative significance of these lesions varies according to the gestational period. Atherosclerotic disease causes the majority of AMIs in the antepartum period, whereas the incidence of coronary artery dissection is increased postpartum.20 Pulmonary embolus, reflux esophagitis, biliary colic, and aortic dissection are all more common than myocardial ischemia during pregnancy and need to be considered in the differential diagnosis. Unfortunately, some women continue to use illicit drugs during their pregnancy, and it is important to consider cocaine use in the pregnant patient presenting with chest pain.
Table 177-4 Antihypertensive Agents for Hypertensive Emergencies ANTIHYPERTENSIVE AGENT
DOSE
COMMENT
Hydralazine
5 mg IV, then 10 mg IV every 20–30 min until target blood pressure is achieved 20 mg IV; subsequent dosing with 40 mg, then two doses of 80 mg every 10 min until target blood pressure is achieved Infusion rates vary—start at 0.25 µg/kg/min
Consider another agent if inadequate response despite administration of 25 mg Consider another agent if inadequate response despite administration of 220 mg Standard contraindications apply Use only if other agents not effective—fetal cyanide toxicity is possible after several hours
Labetalol Sodium nitroprusside
National Institutes of Health: The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. NIH Publication No. 04-5230, August 2004.
2303
as physiologic demands lessen, but its occurrence during pregnancy merits a complete cardiac evaluation for acute coronary syndrome.
Valvular Heart Disease Maternal valvular heart disease can be congenital or acquired and is one of the leading causes of nonobstetric death. Acquired valvular disease is mainly the result of rheumatic fever and endocarditis. In the United States, most cases of significant congenital heart disease are identified and corrected surgically before puberty. The ability of patients to tolerate pregnancy without significant adverse effects depends on the type and severity of the lesion. Mild to moderate lesions (New York Heart Association classes I and II) are often associated with good outcomes. On the other hand, mitral stenosis (beyond class I), advanced aortic stenosis, aortic and mitral lesions associated with moderate to severe ventricular dysfunction or pulmonary hypertension, and mechanical prosthetic valves requiring anticoagulation can result in significant maternal mortality and require directed therapy (see Table 177-1).16,23 Mitral Stenosis. Mitral stenosis is the most commonly encountered valvular lesion in pregnancy and is also the most important lesion to detect in early pregnancy because maternal mortality is appreciable.16 The increased resting heart rate seen in normal pregnancy shortens left ventricular diastolic filling time and consequently results in decreased stroke volume. The demand for increased cardiac output during pregnancy creates a vicious cycle in which further acceleration of heart rate occurs. This tachycardia, combined with the expanded plasma volume of pregnancy, ultimately produces high left atrial pressures, pulmonary vascular congestion, and the symptoms and physical findings typical of left ventricular failure. Even patients with asymptomatic pregestational mitral stenosis warrant close cardiology follow-up with echocardiography at frequent intervals during gestation.16 Treatment is aimed at reducing plasma volume and slowing the heart rate (see Table 177-1). Surgical intervention is indicated in patients with refractory symptoms despite optimal medical management and in patients with pulmonary hypertension. The preferred procedure is percutaneous balloon valvotomy, which is associated with good maternal and fetal outcomes when performed in experienced centers.16,23 Aortic and Mitral Regurgitation. In most cases, chronic regurgitant lesions are well tolerated during pregnancy and may even improve as the reduced systemic vascular resistance of pregnancy allows more forward and less regurgitant flow. In addition, this effect is aided by the increase in heart rate and shortened diastole that occur in pregnancy.16,23 When necessary, medical therapy consists of diuresis and, possibly, vasodilators. Patients with acute mitral regurgitation due to ruptured chordae do not fare as well and may require surgical therapy.16,23 Aortic Stenosis. Symptomatic aortic stenosis during pregnancy usually occurs in the setting of a congenital bicuspid valve and is uncommon.16,23 Patients with aortic stenosis tend to have good pregnancy outcomes, and conservative management is often possible. Severely symptomatic patients may need percutaneous valvotomy in an experienced institution, although this procedure is risky for both the mother and the fetus.16,23 Prosthetic Heart Valves. Anticoagulation for the pregnant patient with a prosthetic heart valve is complicated. Warfarin is considered to be contraindicated during weeks 6 through 12 of gestation, although the risk of embryopathy is likely low when the dose is ≤5 mg/day.16,23 In addition, because warfarin crosses the placenta, there is a risk of fetal bleeding. Neither unfractionated heparin (UFH) nor low-molecular-weight heparin
Chapter 177 / Chronic Medical Illness during Pregnancy
The diagnosis of angina is often clinical. Because normal pregnancy is often associated with electrocardiographic changes such as ST wave depression and T wave inversion, additional evaluation may be necessary. Echocardiography is useful in correlating suspicious electrocardiographic findings with wall motion abnormalities. The enzymatic diagnosis of myocardial infarction is unchanged except during and immediately after delivery, when troponin is preferred over creatinine kinase and myoglobin, which are both elevated above baseline during this time period. Angiography for the patient with suspected angina is generally avoided because the large dose of radiation poses a risk to the fetus, but it can be used in certain clinical situations with the preferred time of intervention being the second trimester.16 Radionuclide studies carry a lower radiation risk than does angiography but are also avoided unless absolutely necessary. Treatment of AMI during pregnancy is similar in most respects to treatment of the nonpregnant patient. Survival of the mother is the primary concern, and therapy to improve maternal outcome should not be withheld. Standard treatment includes antiplatelet agents, nitroglycerin, beta-blockers, and antithrombotic agents. However, there are potential adverse effects of such therapy on the mother and fetus (see Table 177-1), and emergent consultation with cardiology is recommended. Regarding antiplatelet agents, aspirin remains the first-line agent. There is limited experience with clopidogrel and eptifibatide, but both have been used successfully and are considered safe from a teratogenic standpoint. Heparin has long been considered the antithrombotic agent of choice in pregnant patients, but newer low-molecular-weight agents, specifically dalteparin and enoxaparin, appear to be efficacious and safe and are deemed appropriate for use.21 Heparin is preferable for patients in the late third trimester because there is a more predictable response to protamine sulfate should labor begin. Experience with thrombolytic therapy in pregnancy is extremely limited and is much more extensive in the setting of stroke and pulmonary embolism.22 Although such therapy may reduce maternal and fetal mortality, pregnancy is considered a relative contraindication to its use. Reported adverse effects have included maternal hemorrhage, maternal death, placental abruption, preterm delivery, fetal death, and fetal intracranial hemorrhage, although the causal relationship in many of these cases is unclear because neither tissue plasminogen activator nor streptokinase cross the placenta. The majority of patients had favorable maternal and fetal outcomes, but most were being treated for indications other than AMI.20,22 Because thrombolytic therapy precludes major surgery and epidural anesthesia in the hours to days immediately after administration, the emergency physician must carefully consider whether to use these agents in pregnant women who are close to term, especially if the need for cesarean delivery is anticipated. Because of the lack of data and potential risks of fibrinolytics in gravidas, emergent percutaneous intervention is the treatment of choice for definitive management of AMI. In the setting of peripartum AMI, labor should be conducted with continuous monitoring of both the mother’s hemodynamic status and fetal well-being. There are benefits and risks of both vaginal and operative delivery. Cesarean section avoids prolonged exertion by the mother but subjects the patient to general anesthesia if use of antithrombotic agents precludes epidural catheter placement. In addition, surgical delivery places the patient at risk for typical postoperative complications, such as infection, hemorrhage, and thromboembolism. Therefore, assisted vaginal delivery is preferred unless there is an obstetric reason for cesarean section.20 Resolution of angina may occur during the postpartum period
PART V ■ Special Populations / Section Two • The Pregnant Patient
2304
(LMWH) cross the placenta or are teratogenic, but there are problems with their use in pregnancy. The response to UFH is more unpredictable in the pregnant patient, and frequent monitoring of the activated partial thromboplastin time is indicated. Even with frequent monitoring, the risk of thromboembolism is significantly increased. LMWH is an acceptable alternative to UFH in pregnant patients with thromboembolic disease, but there are limited data on its use in the gravida with a mechanical valve.16,23 As a result, use of LWMHs in patients with prosthetic heart valves is not recommended at this time.16,21,23
Hematologic Disorders Anemia By far the most common medical complication of pregnancy is anemia. The physiologic adaptations to pregnancy include expansion of plasma volume in excess of the increase in red blood cell (RBC) mass. This results in the dilutional, physiologic anemia of pregnancy with nadir hemoglobin occurring near the beginning of the third trimester. Any type of anemia can complicate pregnancy, but the three types most commonly involved are iron deficiency, folate deficiency, and sickle cell hemoglobinopathy. Iron-Deficiency Anemia. Iron-deficiency anemia is common, occurring in approximately 20 to 25% of pregnancies in industrialized countries. Apart from chronically low or marginal iron stores in many women, diversion of maternal iron to the fetus for development of its own RBCs and iron stores and increased maternal demand for iron exacerbate the deficiency during pregnancy. Pregnant patients are also subject to other causes of iron-deficiency anemia, such as malnutrition, chronic underlying disease, and blood loss from the gastrointestinal or genitourinary tracts. Anemia in pregnancy is defined by the World Health Organization as a hemoglobin (Hgb) concentration less than 11 g/ dL. Because of physiologic dilution, this level may be slightly lower after the 25th week of gestation. Mild, dilutional anemia (Hgb between 9 and 11 g/dL) confined to the later stages of pregnancy has little impact on obstetric outcomes.24 On the other hand, severe anemia (11 g/dL) and normal iron stores (ferritin >20 µg/ L) is controversial. A trial found that iron supplements for such women prior to 20 weeks of gestation did not reduce the prevalence of preterm delivery or development of anemia later in pregnancy but did result in an increase in birth weight.27 Another trial found that the incidence of both preterm delivery and low birth weight was reduced in women on prophylactic iron.28
Folate-Deficiency Anemia. Folate deficiency is one of a number of causes of megaloblastic anemia, a condition characterized by abnormal DNA synthesis and ineffective RBC production. The incidence of folate deficiency in pregnancy is low (4%) in developed countries but remains higher in other populations.29-31 The risk of developing folate deficiency is increased in patients with multiple gestations, preexisting malnutrition, hyperemesis gravidarum, malabsorption syndromes, and alcoholism. Use of certain antiepileptic medications also places women at increased risk for deficiency. Iron-deficiency and folate-deficiency anemias often coexist, making the peripheral blood smear difficult to interpret. Measurement of serum and RBC folate levels is indicated in cases of suspected folate deficiency, but both have limitations. Serum folate is noted to exhibit a rapid response to folate intake, and low levels may normalize within days after a folaterich meal. Serum folate is also affected by the hemodilution of pregnancy. RBC folate levels are more static and are indicative of folate status several months prior to analysis. RBC folate levels may be misleading in more acute cases and are affected as well by the increase in RBC production that occurs in pregnancy. As is the case for iron deficiency, effects on the fetus depend on the degree of anemia, with the most significant complications being neural tube defects (NTDs) and preterm delivery.32-34 Folate supplementation reduces the risk of NTDs,32,33 and oral folate supplementation with 0.4 to 1.0 mg daily is routinely recommended during pregnancy and prior to conception.35 Women at higher risk for NTDs (e.g., NTDs in prior pregnancy) are advised to take higher doses of folate at 4 mg daily.35 Sickle Cell Anemia. Sickle cell disease (SCD) is one of the major sources of maternal and fetal complications in the United States. The details of the pathophysiology and genetics of SCD are discussed in Chapter 119, but it is useful to review the most common phenotypes that affect pregnancy. The sickle gene can be homozygous (hemoglobin SS or SCD), and this form of the disease is responsible for most pregnancy complications. The sickle gene can also be heterozygous with normal hemoglobin A (sickle cell trait or hemoglobin SA), in which case symptoms are rare except under extreme environmental conditions. The hemoglobin S can also be heterozygous with a large number of abnormal hemoglobins such as hemoglobin C, several variants of thalassemia, and other rare hemoglobin variants, and each variant has its own complication profile. Of these, the most relevant in terms of pregnancy complications is hemoglobin SC. Patients with SCD are subject to many chronic medical problems secondary to a variety of pathophysiologic mechanisms, including sickling of RBCs, anemia, immunosuppression caused by autosplenectomy, and the need for repeated transfusion. Median life expectancy is in the fifth decade for both sexes affected by SCD, and female fertility is generally unaffected, so it is likely that the emergency physician will encounter pregnant patients with the disease. Reported maternal complications in patients with SCD include preterm labor, premature rupture of membranes, maternal infections, more frequent pain crises, and an increased need for cesarean delivery.36,37 Despite these complications, the maternal mortality rate is less than 1% with current treatment.37 SCD also results in adverse effects on the fetus (see Table 177-1). Placental infarction is common, with small-forgestational-age and low-birth-weight infants resulting from placental insufficiency.38 A high rate of fetal loss has been noted in the past, although a recent study found no increase in the rate of perinatal death.37 An incidental complication is that the Kleihauer-Betke test to distinguish fetal from mater-
Neurologic Disorders Epilepsy Epilepsy is the most common neurologic complication of pregnancy but remains relatively rare, affecting less than 1% of all gestations.41 Epilepsy refers to a broad spectrum of seizure disorders that range from relatively benign and infrequent seizures to a disabling condition with daily, poorly controlled generalized convulsions; therefore, care must be individualized. Management of epilepsy during pregnancy must balance the risk of increased frequency and duration of seizures to both the mother and the fetus against the teratogenic risks of antiepileptic medications (AEMs). The effect of pregnancy on epilepsy is variable. Most (65%) epileptic patients experience no change in their seizure frequency, whereas 20% experience more frequent seizures and 15% experience less frequent events.42 A decrease in plasma drug concentrations is expected with the older AEMs due to alterations in the plasma volume, protein binding, and renal clearance that occur in pregnancy. In addition, some patients engage in voluntary noncompliance with medications to avoid teratogenic effects on the fetus. Data for the “newer” agents are less robust, but decreased drug concentrations also occur with lamotrigine, oxcarbazepine, and levetiracetam. The primary adverse fetal complication in these pregnancies is congenital malformations. Of primary concern is the risk of NTDs with valproic acid and carbamazepine and facial clefts/cardiac anomalies with phenytoin and the barbiturates. There is a two- or threefold increase in the incidence of serious congenital malformations in offspring of epileptic mothers taking these drugs, and the risk is even higher if the mother is taking more than one agent.41,43,44 Controversy exists regarding whether infants of epileptic patients not taking AEMs have an increased incidence of congenital malformations compared with the general population. Studies comparing these infants with infants born to mothers without epilepsy noted a similar incidence of malformations.43,44 Data are limited, but preliminary results suggest that the newer agents also have a rate of congenital malformations similar to the general population.45-47 Care should be provided by specialists in high-risk obstetrics and neurology disciplines; however, emergency physicians may be forced to confront this problem in several clinical sce-
2305
narios—the pregnant patient with a first-time seizure or status epilepticus and the patient with epilepsy who is found to be pregnant. New-Onset Seizure. Pregnant patients may seek treatment for idiopathic new-onset seizures; however, drug toxicity or withdrawal, head injury, meningitis, stroke, and eclampsia should be considered as possible causes. The most important of these is eclampsia. Patients in the immediate postictal phase often manifest hypertension resulting from massive sympathetic discharge, and even those with normal pregnancies may have mild edema of the lower extremities. Consequently, urinalysis should be performed to search for proteinuria, which may be the only differentiating factor in the initial assessment of these patients. After a period of observation, elevated blood pressure in the noneclamptic patient will likely revert to normal. If the patient remains hypertensive or manifests other signs of eclampsia, magnesium sulfate and other agents are indicated to prevent further seizures and to control blood pressure. In patients who do not manifest signs of eclampsia, investigation of the cause of the seizure should proceed as with the nonpregnant patient (see Chapter 100). Status Epilepticus. Any potential seizure etiology, including eclampsia, may result in status epilepticus. Despite this, status epilepticus in pregnancy is relatively rare, and limited data are available regarding its occurrence and therapy. Observations from a pregnancy registry note that status epilepticus may occur at any time during gestation and even at delivery. It may also occur in patients who have been seizure-free throughout their pregnancy.42 Older reports note a high fetal and maternal mortality, but recent data support a much lower complication rate.42 The risk of untreated status epilepticus to both the mother and the fetus clearly outweighs the potential for adverse teratogenic effects, and standard resuscitative measures as well as drug therapy are indicated. Continuous fetal monitoring should be instituted as soon as possible and the mother positioned in the left lateral decubitus position to avoid the supine hypotensive syndrome. Pregnant Epileptic Patient. Patients with epilepsy coming to the emergency department for unrelated reasons may be found to be pregnant. Although no immediate change in their therapeutic regimen needs to be made, these patients should be advised of the potential risk of AEMs in pregnancy and be referred to appropriate specialists. Unintentional pregnancy is seen even in patients taking oral contraceptives because AEMs can cause increased clearance of these medications, thereby reducing their efficacy. There are significant obstetric complications related to prolonged seizure activity, and long-term treatment with an AEM for most patients with seizures is warranted (see Table 177-1). Patients who have nonconvulsive seizure disorders or who are seizure-free for a sufficient period prior to conception are candidates for nonpharmacologic observation, but this decision should be deferred to the patient’s primary physician or neurologist. Because phenytoin, carbamazepine, valproate, and possibly other AEMs interfere with folate metabolism, oral supplementation with at least 0.4 mg/day is recommended for all women of childbearing age taking these drugs. Higher doses are recommended for women using valproic acid and carbamazepine, which are known to cause NTDs. Enzymeinducing AEMs such as carbamazepine, phenytoin, and phenobarbital have been reported to cause neonatal vitamin K deficiency and hemorrhagic complications. It has been recommended that mothers taking these certain AEMs be given oral vitamin K during the last month of gestation, but data indicate that the actual risk of hemorrhage is low and possibly related to prematurity.48
Chapter 177 / Chronic Medical Illness during Pregnancy
nal blood will yield false-positive results because of the persistence of hemoglobin F in the mother. Management of SCD during pregnancy is similar to that of the nonpregnant patient (see Table 177-1). Folate supplementation is standard even in the nonpregnant state because of the increased turnover of RBCs, although the recommended daily dose of folate increases to 4 mg during pregnancy.39 Supplemental iron is controversial because of the potential for iron overload. The need for therapy is best determined by the appropriate specialists after a complete analysis of relevant hematologic parameters. Prophylactic transfusion to achieve a predetermined hemoglobin level has not been found to improve pregnancy outcomes, although it may reduce the number of acute pain crises.40 Transfusion is reserved for patients with symptomatic anemia, cardiopulmonary instability, acute chest syndrome, and preeclampsia and possibly for patients with increasingly frequent pain crises. It is also used preoperatively for anticipated blood loss in patients undergoing cesarean section.39,40 Hydroxyurea is not recommended for use in pregnancy because of potential teratogenicity.39
PART V ■ Special Populations / Section Two • The Pregnant Patient
2306
Multiple Sclerosis
Myasthenia Gravis
Multiple sclerosis (MS) affects approximately 400,000 Americans and is twice as common in women as in men. The peak age at onset is 20 to 35 years of age, which overlaps peak childbearing years. The disease is characterized by intermittent episodes of central nervous system demyelination with consequent neurologic impairment that follows a relapsingremitting course. Progressive neurologic deficits and permanent disability develop in certain patients. The impact of pregnancy on the course of MS has been closely studied in various cohorts of women, and a pattern has emerged. Like other autoimmune diseases, the frequency and severity of exacerbations of MS improve because of the immunosuppressant effects of pregnancy. This effect is most pronounced in the third trimester. During the 3 months after delivery, the rate of relapse increases and then returns to the prepregnancy baseline.49,50 Pregnancy does not seem to have any significant long-term adverse effects on disease progression.49,50 On the other hand, children of parents with MS have an increased susceptibility to develop MS, reflecting at least a partial genetic component to the disease. Maternal relapse rate is unaffected by epidural anesthesia,49 and decisions regarding anesthesia should be based solely on obstetric considerations. Labor may be complicated by fatigue and uncoordinated voluntary motor activity in pushing, but generally pregnancies in these patients progress without undue complications.51
Myasthenia gravis is a rare disorder in which autoimmune destruction of the postsynaptic cholinergic receptor results in profound muscle fatigability. Treatment with cholinesterase inhibitors can supply additional acetylcholine to achieve virtually normal muscle contractility but must be titrated to avoid precipitating a cholinergic crisis. Adjustments in medication dosing are frequently required to preserve this balance. The effect of pregnancy on myasthenia gravis is unpredictable in the individual patient, but overall approximately one fifth to one third of patients experience exacerbation of disease, with the remainder having improvement or no change in disease severity.53,54 Several important pregnancy complications are associated with myasthenia gravis. Because of weight gain, anemia, and other physiologic adjustments of pregnancy that may result in fatigue, the distinction between normal pregnancy symptoms and myasthenia may be difficult. Most deliveries are accomplished vaginally without complication in treated patients, and studies have found that assisted and surgical delivery in myasthenia gravis patients is usually performed not for exacerbations but, rather, for obstetric reasons.54 Nonetheless, assisted delivery with vacuum extraction has been recommended to facilitate the second stage of labor.54 Up to 30% of neonates born to mothers with myasthenia gravis develop a transient neonatal myasthenic syndrome via placental transport of acetylcholine receptor antibodies.54 The onset of neonatal myasthenia is typically within the first hours of life but may be delayed by a period of days. Manifestations include poor feeding and suck, diminished reflexes, hypotonia, and respiratory failure. As with adults, the symptoms respond to cholinesterase inhibitors, but treatment should be done in an intensive care unit setting. The duration of the syndrome depends on the clearance of maternal antibodies, and these neonates should be observed for at least 1 week. Treatment in the emergency setting is no different than treatment for nonpregnant patients (see Table 177-1). Assessment of pulse oximetry, forced vital capacity, and arterial blood gas parameters will guide respiratory therapy. For patients presenting with weakness, a “Tensilon challenge test” to distinguish myasthenic from cholinergic crisis is appropriate after initiation of appropriate ventilatory support. Standard medical treatment should be continued during labor and delivery in order to maximize motor strength. Epidural anesthesia is also recommended to reduce pain and fatigue. Sedatives and other agents that increase fatigue should be avoided during this time. The physician should be aware that 30% of patients experience an exacerbation during the postpartum period as the protective immunosuppressant effect of pregnancy dissipates.53,54 In addition, puerperal infections place the patient at risk for disease exacerbation and warrant aggressive therapy.54
Spinal Cord Injury Because spinal cord injury (SCI) occurs mainly in young people and usually does not impair fertility, there is a relatively large population of paraplegic and quadriplegic patients who become pregnant. Although many of these pregnancies are uneventful, these patients are at risk for certain complications. The incidence of urinary tract infection is increased as a result of neurogenic changes and the need for catheterization. These infections may progress to pyelonephritis during pregnancy, with the subsequent increased risk of fetal loss, prematurity, and maternal sepsis. The increased coagulability of pregnancy combined with chronic immobilization results in an increased incidence of venous thrombosis and pulmonary embolus. A unique problem in the patient with SCI is the detection of the onset of labor, which may be painless and precipitous. Patients with spinal cord lesions below T10 to T12 have an intact uterine nerve supply and will experience labor pains; however, with lesions above T10, labor may be imperceptible or experienced as only mild abdominal discomfort. In addition, 85% of patients with high lesions (above T5 to T6) experience potentially life-threatening autonomic dysreflexia.52 This manifests as severe paroxysmal hypertension, headache, tachycardia, diaphoresis, piloerection, mydriasis, and nasal congestion. Because the response is not specific to labor and may be precipitated by distention of bowel or bladder, other causes must be pursued as well. Pregnant patients with SCI who have these symptoms should be assessed for cervical dilation and have uterine contractions monitored. Emergency department treatment is directed at restoring normal blood pressure with nitroprusside, nitroglycerin, or hydralazine. Definitive therapy is with regional anesthesia. Both spinal and epidural anesthesia obliterate and prevent this response and should be used as soon as possible during labor for all women with SCI.52 Because of the difficulty in detecting labor, pregnant patients with SCI are sometimes admitted for observation near term.
Renal Disorders Several alterations in renal hemodynamics occur during pregnancy. Both the glomerular filtration rate (GFR) and effective renal plasma flow increase by 30 to 50% compared with the nonpregnant state. Because no substantial alterations are present in the production of creatinine or urea nitrogen, levels of these solutes decrease from low normal nonpregnant values of 0.7 and 12.0 mg/dL to 0.5 and 9.0 mg/dL, respectively. Thus, blood urea nitrogen and creatinine levels considered normal in nonpregnant women (creatinine > 0.8 mg/dL and blood urea nitrogen > 12 mg/dL) indicate underlying renal impairment and warrant further investigation. The primary factors that determine how renal disease affects pregnancy outcome are the degree of underlying dysfunction
Metabolic and Endocrine Disorders Diabetes Three types of diabetes are involved in pregnancy: type I, or insulin-dependent diabetes mellitus (IDDM); type II, or noninsulin-dependent diabetes mellitus (NIDDM); and gestational diabetes mellitus (GDM). Although NIDDM is sometimes considered a more benign form of disease, the risk of malformations is the same for both NIDDM and IDDM.60,61 Maternal and fetal complications relate more to inadequate glycemic control as well as the presence of vascular complications or severe renal insufficiency rather than to the type of diabetes. All pregnant patients with diabetes are considered “brittle” and require regular follow-up by appropriate specialists. In addition, all patients, including those with GDM, should perform routine self-monitoring of glucose. Maternal Complications. The physiology of glucose regulation during pregnancy is complex. During the first half of gestation, the sensitivity to insulin increases as a result of increased circulating estrogen. When combined with emesis, increased use of glucose by the placenta and fetus, and a decrease in hepatic glucose production, hypoglycemia occurs more easily. Consequently, patients with IDDM are at risk for severe hypoglycemia during early pregnancy and insulin dosage should be decreased. During later gestation, however, there is progressive insulin resistance that peaks during the third trimester and then falls again during labor and the immediate postpartum period. Pregnancy also predisposes to ketosis, and this effect is exacerbated in the setting of emesis. Specific adverse pregnancy-related outcomes are more common and include preeclampsia, preterm delivery, and the requirement for cesarean delivery.62,63 The effects of pregnancy on underlying diabetes vary depending on the organ system. The data are limited, but pregnancy is not advised for diabetic patients with significant coronary artery disease because of the cardiovascular demands of pregnancy and the high mortality of AMI during pregnancy.62 Patients with diabetic nephropathy are at increased
2307
risk for preeclampsia and the subsequent requirement for preterm delivery.62 The renal effects of pregnancy depend on the severity of underlying disease. Patients with mild to moderate renal dysfunction do not seem to experience permanent disease progression. On the other hand, women with more severe renal dysfunction are at increased risk for progression to end-stage disease.60,62 Retinopathy may worsen acutely during pregnancy, especially in patients with high HbA1c and hypertension and in the setting of aggressive control of hyperglycemia.62,64 Laser therapy of preexisting retinopathy is recommended prior to conception.60,62 Autonomic neuropathy does not accelerate during pregnancy,64 but the combination of hyperemesis gravidarum and gastroparesis often causes problems. Frequent vomiting results in dehydration and inadequate intake of food that can result in hypoglycemic episodes if the insulin dosage is not adjusted accordingly.60 Diabetic ketoacidosis (DKA) occurs in up to 10% of diabetic patients during pregnancy and may represent the initial presentation of diabetes. DKA is most commonly seen in patients with IDDM but also complicates pregnancies in women with NIDDM and GDM.62 Common precipitating events include the typical factors seen in nonpregnant patients, such as insulin noncompliance and infection. Other pregnancy-specific factors are hyperemesis, use of beta-mimetic medications for tocolysis, and use of corticosteroids to hasten fetal lung maturity. The serum pH may be deceptively normal in a pregnant patient with DKA because the initial pH tends to be higher in pregnancy due to physiologic hyperventilation. Loss of gastric acid through vomiting will also counteract the metabolic acidosis of diabetic ketoacidosis. In addition, serum glucose may be only moderately elevated because the fetus continues to secrete insulin and use glucose. Maternal mortality is rare in appropriately treated DKA. Fetal mortality rates are relatively high, ranging from 10 to 35%.62 Fetal Complications. Diabetes has many deleterious effects on the fetus (see Table 177-1). The rate of congenital malformations in patients with prepregnancy diabetes is increased three- or fourfold compared to the nondiabetic population, with anomalies being more likely in pregnancies with poor glycemic control.62,63,65 Glucose crosses the placenta, and prolonged fetal exposure to maternal hyperglycemia induces fetal pancreatic hyperplasia and high insulin production. Elevated insulin levels in turn promote fetal growth, resulting in macrosomia. Conversely, preeclampsia and placental infarction secondary to vascular disease may result in impaired fetal development.62,63 After delivery, continued high insulin secretion in the absence of a maternal glucose supply results in a significant rate of neonatal hypoglycemia. Because these infants are more likely to be preterm, hyperbilirubinemia and respiratory problems are also more frequent.62,63 Long term, it is likely that poor metabolic control during pregnancy leads to an increased risk of impaired glucose tolerance, obesity, and NIDDM in affected offspring.60,62,66,67 Management. Treatment of NIDDM and IDDM requires individualized and carefully adjusted insulin administration with the goal to maintain strict glycemic control while avoiding hypoglycemia. Ideally, HbA1c values should not be greater than 6% (see Table 177-1).60,62 Frequent self-monitoring is recommended, and these patients should also have frequent office visits with their practitioner throughout the duration of pregnancy.62 It is preferable to achieve glycemic control prior to conception in order to minimize the risks of malformations and other complications.60 Treatment of DKA does not differ from treatment given in the nonpregnant state except that fetal viability and well-being should be assessed. The timing and mode of delivery depend on whether there exist obstetric or maternal complications. In the absence of
Chapter 177 / Chronic Medical Illness during Pregnancy
and the presence of hypertensive disorders. In general, patients with mild insufficiency (creatinine < 1.4 mg/dL) and no hypertension can expect good pregnancy outcomes and preserved renal function. On the other hand, patients with moderate to severe renal dysfunction have a much higher risk of further decline in renal function as well as adverse obstetric outcomes, including preeclampsia, placental abruption, fetal loss, preterm delivery, low birth weight, and an increased need for cesarean section.55-57 Worsening of underlying renal function is more likely in patients with decreased GFR who also have associated proteinuria and/or hypertension.56-58 Because worsening renal function is manifested by hypertension and proteinuria, differentiation from preeclampsia can be difficult. In this setting, it is best to treat the patient for presumed preeclampsia with the caveat that magnesium administration be performed judiciously based on serum magnesium levels. Pregnant women with chronic renal failure require aggressive and timely management to optimize their chances for a successful gestation without causing further deterioration in renal function. Baseline renal function studies are done early in pregnancy and then reassessed every 4 to 6 weeks. Evidence of renal function deterioration or the development or exacerbation of hypertension warrants admission for specialized inpatient care. Hemodialysis is indicated for creatinine levels greater than 3.5 to 5 mg/dL. In patients undergoing dialysis, sessions are more frequent and more prolonged to minimize azotemia.59
PART V ■ Special Populations / Section Two • The Pregnant Patient
2308
suspected problems, vaginal delivery at term is recommended.60 Elective delivery is indicated in the setting of poor metabolic control, significant diabetic complications, and fetal macrosomia with suspected birth weight greater than 4500 g.62 Delivery is timed to avoid fetal growth into the macrosomic range while allowing for sufficient fetal lung maturity. If corticosteroids are administered, very close glucose monitoring is required because insulin requirements are likely to increase. Progression beyond term is not recommended.
Gestational Diabetes GDM is defined as glucose intolerance that begins (or is first detected) during pregnancy. As the incidences of both NIDDM and IDDM have increased, so has that of gestational disease.68 GDM has little impact on perinatal mortality if detected and addressed; however, untreated disease increases the likelihood of fetal macrosomia and adverse perinatal outcomes.67,69,70 GDM is also a risk factor for the subsequent development of nongestational diabetes. Patients at increased risk for NIDDM include those with obesity, whereas patients with evidence of islet cell immunity are at risk for development of IDDM.67,71 The exact means of screening for GDM is controversial, but all pregnant patients should undergo some form of screening, even if it is accomplished simply by obtaining a clinical risk profile.72 The American Diabetes Association advocates selective laboratory assessment in patients at high risk for GDM. This group includes patients older than 25 years; certain highrisk ethnic groups; and patients with obesity, known firstdegree relatives with diabetes, a personal history of diabetes or glucose intolerance, and a personal history of poor pregnancy outcome.67 A positive screening test at 24 to 28 weeks of gestation via measurement of serum glucose 1 hour after a 50-g oral load is an indication for a formal 3-hour glucose tolerance test. The majority of patients with GDM achieve metabolic control with dietary therapy. Traditionally, insulin administration is indicated if dietary control is unsuccessful (see Table 177-1).67,72 Standard practice is to avoid oral medications because they cross the placenta and have the potential for teratogenicity. However, studies comparing insulin with glyburide in GDM found that the latter provided similar glycemic control without an increase in adverse pregnancy outcomes.73 Preliminary investigations of metformin have found this agent to be safe for use in pregnancy, but further study is recommended prior to its widespread use.74 There are very limited data for other oral medications.
Thyroid Disorders The peak incidence of thyroid disease is in women of childbearing age. Both hypoactivity and hyperactivity lead to obstetric complications and warrant specific therapy. As with other organ systems, there are physiologic changes in thyroid function that occur in normal pregnancies. Human chorionic gonadotropin (hCG) is structurally similar to thyroidstimulating hormone (TSH) and stimulates the thyroid gland. As a result, women with normal pregnancies may experience a temporary suppression in TSH production and slightly lower TSH levels during the first trimester, which mirrors the high circulating hCG levels seen at this gestational period.75,76 This hCG-mediated effect may lead to a transient hyperthyroidism if hormone levels are markedly increased. Estrogen increases the amount of thyroid hormone binding proteins, which ultimately leads to an increase in the amount of total T4 and T3 even though free hormone levels remain normal. After 10 to
12 weeks of gestation, the fetus makes its own thyroid hormone but still requires placental transport of iodine and adequate maternal intake of this element (200 µg/day).
Hyperthyroidism The most common cause of hyperthyroidism is Graves’ disease, in which autoimmune thyroid-stimulating immunoglobulin G results in increased production and release of thyroid hormone. Because the symptoms of hyperthyroidism resemble the physiologic changes expected during pregnancy in many respects, the diagnosis may not be immediately evident. Patients with Graves’ disease have disease-specific findings, including a diffusely enlarged, soft, mildly tender thyroid gland; exophthalmos; and dermopathy. Other symptoms, such as dyspnea, heat intolerance, hyperemesis, tachycardia, palpitations, systolic flow murmurs, increased appetite, and fatigue, are common to both conditions, making clinical diagnosis difficult. In cases of suspected hyperthyroidism, thyroid function studies are indicated and will confirm the presence of disease. There are several obstetric concerns for both mother and the fetus in the setting of untreated hyperthyroidism (see Table 177-1).76 Thyroid storm is the most serious manifestation of the disease. It may be precipitated by stressors such as infection and delivery, and it manifests with fever, dysrhythmias, myocardial dysfunction, mental status changes, and circulatory collapse. In addition to the more general complications stemming from thyroid hormone excess, Graves’ disease places the fetus at risk for autoimmune-mediated thyroid dysfunction via placental transfer of maternal thyroid-stimulating immunoglobulins (TSIs). Up to 17% of neonates of mothers with Graves’ disease and positive TSI values have transient hyperthyroidism lasting 3 to 12 weeks. The condition gradually clears as maternal antibodies are metabolized. Manifestations are potentially severe and include irritability, tachycardia, goiter, cardiomegaly, congestive heart failure, premature craniosynostosis, low birth weight, and failure to thrive.77 These infants also have an increased mortality rate, so the condition must be recognized promptly and treated aggressively with typical therapies. The mainstay of treatment of hyperthyroidism (see Table 177-1) consists of antithyroid drugs. Propylthiouracil has traditionally been preferred over methimazole because of an increased potential for adverse congenital drug effects from the latter. Both drugs cross the placenta and can cause fetal hypothyroidism. Consequently, maternal thyroid function should be assessed periodically during pregnancy with the goal of keeping free thyroxine in the high normal range. Most patients respond to pharmacologic manipulation, although thyroidectomy may be considered in severe cases. Use of iodine-131 radionuclide to ablate the maternal thyroid is contraindicated because it will also destroy the fetal thyroid gland. Additional therapy with beta-blockade to mitigate the hemodynamic effects of sympathetic stimulation may be required in certain cases pending adequate disease control with antithyroid medications. In the nonpregnant patient, iodides may be used to transiently block the release of stored T4 from the thyroid gland and inhibit organification of iodide. However, the fetal thyroid is extremely sensitive to iodide, which may result in neonatal goiter and hypothyroidism. Therefore, iodide is considered class D in pregnancy, and its use should be reserved for severe cases with duration of therapy limited to a period of days. As with other autoimmune conditions, transient improvement of Graves’ disease during pregnancy is common, with rebound and clinical deterioration occurring after delivery.
2309
Postpartum Thyroiditis
Hypothyroidism The most common cause of hypothyroidism is Hashimoto’s thyroiditis. Overt hypothyroidism is often associated with infertility, so most cases seen during pregnancy are less severe or subclinical forms or occur in patients already undergoing levothyroxine therapy for known disease. Undiagnosed subclinical hypothyroidism may manifest during pregnancy as physiologic adjustments result in a greater requirement for thyroid hormone production. In addition, iodine deficiency may be exacerbated by pregnancy because renal losses of iodine increase as the glomerular filtration rate increases and as the fetus diverts iodine for its own thyroid hormone synthesis. Approximately 3 to 5% of women of childbearing age have subclinical hypothyroidism81 and, unfortunately, these patients may remain asymptomatic. When signs and symptoms do occur, they are identical to those in the nonpregnant state. Myxedema coma is extremely rare but must be considered along with other causes of coma in a pregnant patient. As with hyperthyroidism, there is an increased incidence of adverse maternal and fetal effects (see Table 177-1).76,82,83 The majority of patients who are already undergoing treatment for hypothyroidism will require an increased dosage of levothyroxine during pregnancy, and close monitoring of thyroid function is recommended as soon as possible after conception.75 Treatment for newly diagnosed patients consists of the replacement of thyroid hormone with the goal being to achieve a normal TSH level.76 In women with preexisting hypothyroidism, levothyroxine dosage requirements may increase significantly during pregnancy, and it is suggested that doses be increased by 30% as soon as pregnancy is diagnosed.75 Undiagnosed and untreated subclinical hypothyroidism results in increased perinatal mortality and adverse neurologic and cognitive outcomes for affected infants.82-84 It is unclear whether this effect is directly related to abnormal thyroid function or whether it results from the associated increase in preterm delivery.85 The issue of whether to provide routine prenatal screening for hypothyroidism remains controversial. At the very least, screening should include anyone at risk for disease, and some sources recommend assessment of TSH levels in all women prior to conception or as soon as possible after becoming pregnant because targeted screening may miss up to one third of affected women.84,86-88 Congenital hypothyroidism, associated with cretinism and severe mental retardation, occurs in approximately 1 in 3000 births in the United States. It is most often the result of sporadic fetal thyroid agenesis or ectopy rather than maternal
Systemic Infections Tuberculosis Acquisition and presentation of tuberculosis is unchanged during pregnancy. However, the effect of tuberculosis on pregnancy is unclear. Some studies reveal a significant increase in gestational complications (see Table 177-1), but these outcomes are likely significantly influenced by the site of disease and specifics of treatment or lack thereof. One study found no significant increase in gestational complications in 111 pregnant patients with properly treated tuberculosis.90 Complications are more likely in patients with inadequate or delayed treatment, delayed diagnosis, and extrapulmonary (extranodal) tuberculosis.91,92 Neonatal tuberculosis acquired via exposure to undiagnosed and untreated active disease places infants at significant risk for acquiring tuberculosis during the first year of life, with significant mortality. In addition, congenital tuberculosis is possible after the fetus becomes infected via the placenta or aspiration of infected amniotic fluid. The latter is rare if the mother has received appropriate therapy. Current recommendations are to administer a tuberculin skin test early in pregnancy to all patients at high risk for disease and to obtain a chest radiograph if the purified protein derivative (PPD) skin test is positive or if the patient’s signs and symptoms suggest tuberculosis. Universal screening should be considered in urban areas with a high prevalence of disease. Definitive treatment varies depending on the duration of PPD positivity and whether the patient has active disease (see Table 177-1).93,94 Isoniazid, ethambutol, and rifampin in their usual doses have not been shown to be teratogenic to human fetuses and are acceptable during pregnancy. On the other hand, streptomycin causes fetal ototoxicity, and little is known about the safety of other second-line agents during pregnancy. These less commonly used agents should be avoided except in the case of multidrug-resistant disease.
Human Immunodeficiency Virus and Acquired Immunodeficiency Syndrome The human immunodeficiency virus (HIV) is one of the leading health problems in pregnancy. In 2005, 27% of reported cases of HIV/acquired immunodeficiency syndrome (AIDS) in the United States were in women, with the majority being in women of childbearing age.95 Estimates of the seroprevalence of HIV in pregnant women vary on a regional basis. In the United States, the overall prevalence is low but also varies depending on the population, with increased rates seen in inner-city and other high-risk populations. The mechanism of vertical transmission is multifactorial. The majority of cases are thought to occur during delivery through exposure to maternal blood and secretions; other infants are likely infected in utero or via breast-feeding. Various factors influence the rate of transmission. The most important is maternal viral load, although infection can occur even with maternal HIV RNA greater than 1000 copies/ mL.96-100 Other contributing factors for transmission include vaginal infections, intravenous drug use, low birth weight, and prolonged rupture of membranes.96,101,102 Vertical transmission
Chapter 177 / Chronic Medical Illness during Pregnancy
Postpartum thyroiditis (PPT) is a common but relatively benign condition that develops in approximately 7% of parturients within 9 months of delivery.78 Patients typically experience transient hyperthyroidism, transient hypothyroidism, or transient hyperactivity followed by transient hypoactivity.78 Symptoms are typical for type of impairment, although they are often mild. Pharmacologic treatment may be required; however, the need for medication seems to be confined to those patients with hypothyroidism.76 Occasionally, patients with transient hyperthyroidism require beta-blockers, but antithyroid drugs are not helpful because pathophysiology relates to release of preformed thyroid rather than excess thyroid hormone production. Approximately 30% of patients with PPT develop permanent thyroid failure, and this percentage is increased to more than 50% if one considers just the patients with postpartum hypothyroiditis.79,80
thyroid dysfunction.89 Diagnosis is often difficult because clinical signs and symptoms may be masked by maternal thyroid hormone. Congenital hypothyroidism is successfully treated with hormone replacement begun in the first days of life, and screening for congenital hypothyroidism is mandated in most developed countries.
PART V ■ Special Populations / Section Two • The Pregnant Patient
2310
of HIV in the United States has declined significantly since its peak in the early 1990s due to the implementation of a number of interventions, including routine voluntary testing, antiretroviral therapy (ART), use of elective cesarean section, and avoidance of breast-feeding.100,103 However, vertical transmission remains a concern in women without adequate prenatal care as well as a global concern. It is estimated that perinatal infection occurs in approximately 20% of deliveries if the mother is untreated, whereas the previously mentioned interventions reduce this rate to less than 1 or 2%.97,98,100,102,103 Because of this beneficial effect, it is recommended that all pregnant women undergo screening during their routine prenatal evaluation. Rapid HIV screening is recommended for women in labor whose HIV status is unknown.104 Treatment of the pregnant patient with HIV includes appropriate ART as well as standard therapy for opportunistic infections (see Table 177-1).105 Optimal therapy to prevent vertical transmission includes three stages: antepartum administration of highly active antiretroviral treatment (HAART), intrapartum ART dosing, and treatment of the infant with 6 weeks of zidovudine.96 In HIV-infected mothers without prenatal care, intrapartum ART followed by postexposure treatment of the infant reduces the likelihood of infection but is less effective than the recommended three-stage regimen.96 Elective cesarean section is recommended in mothers with a viral load greater than 1000 copies/mL. It is also reasonable to present cesarean delivery as an option for patients with lower viral loads because transmission does occur even at these low levels.100 Even in the setting of ART, commonly used serologic tests to detect antibodies for HIV yield positive results in virtually all neonates born to HIV-positive mothers because of placental transfer of maternal antibodies. These antibodies may be present for up to 18 months and are not necessarily indicative of infection. The preferred means of diagnosis of perinatal HIV is via use of assays to detect viral RNA or DNA. These infants should also be monitored clinically. Infants with AIDS typically present with lymphocytic interstitial pneumonitis, recurrent bacterial infections, Pneumocystis pneumonia, encephalopathy, thrush, and generalized wasting.106 Recent analysis reveals that HAART therapy in infants and children is extremely effective in reducing HIV-related morbidity and mortality.107 The effects of pregnancy in women with symptomatic HIV infection include an increased incidence of prematurity, stillbirth, and low birth weight and also gestational diabetes. Seropositive mothers who undergo cesarean delivery also have an increased risk of postpartum endometritis and other maternal infections, with the highest rate of infection seen in women who have an emergent procedure.108,109 It is also important to consider the effect of ART on pregnancy outcomes. A few agents, such as efavirenz, have teratogenic potential and are avoided in pregnancy. However, most antiretroviral medications are considered safe in this regard. Several studies have not found an increase in adverse outcomes with the possible exception of preterm delivery or low birth weight when protease inhibitors are used.110-114 The effect of pregnancy is also a matter of debate. An observational cohort study in which a majority of women were using HAART found that pregnancy had no impact on the rate of disease progression.115
Syphilis The incidence of primary and secondary syphilis among females in the United States steadily declined between 1990 and 2004 but increased during 2005 and 2006, with a current rate of 1 case per 100,000 women.116 Fortunately, the incidence
of congenital syphilis (CS) is declining, but the disease remains a concern among those patients without access to adequate prenatal care and prenatal syphilis screening.117 Syphilis causes numerous gestational complications (see Table 177-1), but its most significant sequela is CS. This syndrome is characterized by clinical abnormalities such as hepatosplenomegaly, osteochondritis, jaundice, rash, lymphadenopathy, rhinitis, Hutchinson’s teeth, and anemia. Perinatal mortality for cases occurring from 1992 to 1998 was 6.4%.118 If untreated, vertical transmission rates are high but can be reduced significantly with appropriate penicillin therapy.119 Screening for all pregnant patients is indicated at the first prenatal visit and again in the early third trimester and at delivery for high-risk patients.104 Either Venereal Disease Research Laboratory (VDRL) or rapid plasma reagin (RPR) testing can be used to detect nontreponemal antibody. Pregnancy may cause false-positive results for nontreponemal studies, so confirmation using specific treponemal tests is indicated for a positive VDRL or RPR study. Patients with latent syphilis and those whose titers fail to respond to therapy should undergo cerebrospinal fluid analysis to screen for tertiary syphilis. In addition, fetal ultrasonography prior to the 20th week of gestation is indicated to assess for abnormalities consistent with CS. Treatment is identical to that given to nonpregnant patients using benzathine penicillin G (see Table 177-1).104 Penicillinallergic patients should undergo skin testing and desensitization if the skin test result is positive because alternative therapy is not reliably effective in preventing congenital syphilis.104 Treatment failures are rare with appropriately administered penicillin but do occur. Failures leading to CS are more likely in mothers with secondary syphilis, high VDRL levels, and an interval from treatment to delivery of less than 30 days. They are also more likely in the setting of preterm delivery.119,120
Hepatitis Hepatitis B. The prevalence of hepatitis B virus (HBV) infection among pregnant women varies depending on the population studied. In U.S. urban areas, 0.14 to 5.79% of pregnant women are positive for HBsAg, with Asians having the highest sero prevalence.121 The rate of vertical transmission depends on the acuity of maternal infection and when during the gestation it occurs. Perinatal transmission approaches 90% in mothers who are seropositive for HBsAg and HBeAg and is also more likely if the mother has acute infection during the third trimester or first few months postpartum or if she is a chronic carrier.104,122 Of infants who develop HBV infection, up to 90% become chronic carriers as adults and are at risk for complications such as cirrhosis and hepatocellular carcinoma.104 Routine screening for HBV during early pregnancy is recommended because treatment with hepatitis B immunoglobulin and hepatitis B vaccine is very effective in reducing the rate of vertical transmission.122,123 Treatment failures do occur in the setting of maternal seropositivity for HBeAg and HBV DNA.124 Women at risk for sexually transmitted disease (STD) and women with hepatitis should be retested at delivery.104 Vaccination against HBV is currently recommended for all infants in the United States. The treatment schedule and the need for additional therapy with immunoglobulin vary depending on maternal seropositivity. Infants of HBsAg-positive mothers should receive hepatitis B immunoglobulin and the first dose of vaccine within 12 hours of birth. Two additional doses of vaccine are administered at a later date. Pregnancy is not a contraindication to either therapy. All HBsAg-negative gravidas at risk for STDs or who are seeking STD treatment should receive the vaccine.104,122 Pregnant patients who are
Inflammatory Disorders Rheumatic diseases or collagen vascular diseases are characterized by sterile inflammation in multiple anatomic sites. The most common rheumatic diseases encountered in pregnancy are systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA). Patients with collagen vascular disease may have preexisting cardiovascular or renal compromise and may not tolerate the increased intravascular volume and other physiologic changes that occur during pregnancy. The following discussion focuses on SLE, which is the rheumatic disease responsible for the majority of gestational complications. Most of the treatment guidelines for SLE are relevant to other rheumatologic disorders as well.
Systemic Lupus Erythematosus SLE primarily affects women of reproductive age, and fertility is usually unaffected. The disease course during pregnancy is a matter of controversy, but analyses indicate that acute flares occur in less than one third of patients in clinical remission at the time of conception. These flares tend to be mild and involve the skin and musculoskeletal system.129,130 The gestational effects of SLE also depend on the underlying severity,131 and it is best for women to achieve good disease control prior to pregnancy. Many patients have acceptable outcomes, but lupus pregnancies are associated with an increased rate of complications, including hypertensive disorders, preterm delivery, intrauterine growth retardation, fetal loss, and need for cesarean section.129,130,132,133 The risk of preeclampsia is markedly increased in patients with preexisting lupus nephritis.134,135 As with other renal disease, increasing proteinuria warrants a careful evaluation to distinguish between lupus glomerulonephritis and preeclampsia. The presence of abnormal urine sediment, increasing titers of anti-DNA antibody, and decreasing levels of C3 and C4 point to lupus nephritis.129 Numerous other organ systems in addition to the kidneys are involved in SLE, and differentiation from pregnancyrelated changes may be difficult. Mild thrombocytopenia occurs in normal pregnancies and is also common in patients with SLE, although the clinical significance varies from patient to patient. Anemia is a frequent complication of lupus and magnifies the normal dilutional anemia of pregnancy. Various musculoskeletal and cutaneous symptoms associated with pregnancy, such as arthralgias and facial and palmar erythema, can also resemble active SLE. In addition, preexisting lupus rashes become more erythematous because of the increased cutaneous blood flow during pregnancy. Neurologic disease in SLE may manifest as psychosis, seizures, chorea, or peripheral neuropathy. The incidence of these complications is low
2311
during pregnancy, although the occurrence of seizures in late pregnancy in patients with coexistent hypertension and renal insufficiency may pose a diagnostic dilemma between the neurologic effects of SLE and eclampsia.
Other Rheumatologic Diseases Rheumatoid arthritis is characterized by chronic, destructive, symmetrical joint inflammation. Less common manifestations include the development of subcutaneous nodules, neuropathy, pleuropericarditis, and vasculitis. Systemic symptoms, including weight loss, lymphadenopathy, and fatigue, are common. Because the median age at onset is later with RA, this disorder is seen less frequently than SLE in the pregnant population.132 Approximately two thirds of patients with RA experience an amelioration of symptoms during pregnancy, although often an exacerbation follows delivery. Patients with other rheumatologic diseases, including scleroderma, Raynaud’s phenomenon, and polymyositis, tend to have good pregnancy outcomes in the setting of good disease control. Birth weights are lower in the offspring of these patients likely as a result of the underlying vasculopathy.136,137
Treatment The rheumatologic diseases are very similar with respect to the therapeutic approach. Corticosteroids form the basis of treatment for most disease complications and exacerbations. These drugs may result in a minimally increased risk of cleft deformities (data are inconclusive) but have not been found to have any other significant teratogenic effects and are considered relatively safe to use for disease control in these patients.9,10 Steroid use during pregnancy does have a doserelated effect on intrauterine growth retardation and predisposes the patient to GDM and hypertension, so close monitoring is advised. Aspirin and nonsteroidal anti-inflammatory drugs (NSAIDs) are also a mainstay of therapy in many patients with rheumatic diseases and may be used in pregnancy. There are potential adverse effects of aspirin and NSAIDs on gestation, including miscarriage when taken early in pregnancy and premature closure of the fetal ductus arteriosus, increased maternal hemorrhage, and prolongation of gestation and labor when taken later in pregnancy. There are also conflicting data regarding an increased risk for gastroschisis, cleft defects, and cardiac defects with NSAID use.138-140 If a true association exists, the risk for these anomalies remains extremely small in absolute terms. Use of these agents after 32 weeks of gestation should be avoided to minimize the risks of maternal and fetal hemorrhage and premature ductus closure. Acute flares often require institution of cytotoxic drugs. Cyclophosphamide and methotrexate are contraindicated during the first trimester and should be used only in extreme circumstances because of their teratogenicity and abortifacient properties. Azathioprine has a much more favorable safety profile because of its use in renal transplant patients, and it is the cytotoxic agent of choice during pregnancy. Although the drug and its metabolites cross the placenta, it does not appear to have any major teratogenic effects.141 Cyclosporine also seems to lack teratogenicity and is an acceptable alternative to azathioprine.142 The antimalarial agent hydroxychloroquine is also used in the treatment of SLE and RA. This medication has been found to be safe during pregnancy, and its use may allow for a decrease in corticosteroid dosage.143,144 In addition, patients who have been maintained on hydroxychloroquine prior to conception experience an increased rate of disease flares when the agent is discontinued.143
Chapter 177 / Chronic Medical Illness during Pregnancy
exposed to HBV should receive both hepatitis B immunoglobulin and vaccine.122 Hepatitis C. As with HBV, the prevalence of hepatitis C infection among pregnant women varies depending on the population, ranging from less than 1% to approximately 5%. Vertical transmission is rare in mothers with anti-hepatitis C antibodies and no circulating hepatitis C virus (HCV) RNA.125,126 However, perinatal transmission is significantly increased by the presence of HCV viremia, occurring in approximately 4 to 6% of cases. The transmission rate is even higher in the setting of co-infection with HIV.125-127 Unfortunately, there is little definitive data regarding the use of cesarean delivery to prevent HCV transmission.128 No available vaccine exists to prevent hepatitis C, although testing of potentially infected neonates is advised to identify those at risk for chronic hepatitis.
2312
PART V ■ Special Populations / Section Two • The Pregnant Patient
KEY CONCEPTS ■
The physiologic demands of pregnancy may cause previously occult medical conditions to become apparent. ■ The physiologic adjustments of pregnancy alter the normal ranges for certain laboratory values. The adjusted values need to be considered in interpreting results. ■ The possibility of pregnancy must be considered in the differential diagnosis of certain conditions, including new-onset seizures or status epilepticus (eclampsia), glucose intolerance (GDM), persistent vomiting (hyperemesis gravidarum), and thyroid disorders.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
■
The immunosuppressive effects of pregnancy may cause temporary improvement in inflammatory and autoimmune conditions. This beneficial effect is lost in the postpartum period, resulting in exacerbations of asthma, thyroid disorders, and myasthenia gravis. Medication requirements can change drastically during pregnancy and the postpartum period. ■ Certain medical conditions in the mother result in neonatal complications requiring special resuscitative measures.
Chapter 178
Drug Therapy in Pregnancy
Rania Habal
■ PERSPECTIVE The placenta was previously believed to act as a barrier, excluding toxins from the fetal circulation and protecting the fetus from environmental and pharmacologic exposures. In 1961, when an epidemic of amelia, a rare malformation characterized by an absence of limbs, was linked to the use of thalidomide during pregnancy, the vulnerability of the fetus to medications came into focus. Thalidomide was a sedativehypnotic agent introduced in 1956. It immediately became popular in the treatment of nausea and vomiting during the first trimester of pregnancy, but in the years that followed, it was established that thalidomide was the agent responsible for the amelia/phocomelia epidemic. By the time thalidomide was withdrawn from the market, an estimated 5850 children were affected worldwide.1 Similarly, it took an epidemic of debilitating congenital anomalies and deaths in the children of fishermen in Minamata, Japan, to recognize the teratogenicity of environmental pollutants.2,3 Minamata disease was due to the ingestion of fish contaminated by methylmercury, an industrial by-product that was dumped into Minimata Harbor in the early 20th century. These two events sparked the development of numerous control agencies to oversee the safety of drugs in pregnancy and numerous environmental protection laws. Thalidomide’s legacy continues to haunt physicians worldwide. Many physicians are reluctant to prescribe medications to pregnant women or to nursing mothers. However, only a few medications have been identified as teratogens, and medication use during pregnancy is extremely common. In a worldwide survey of more than 14,000 patients, the World Health Organization reported that more than 86% of women used at least one prescription drug while pregnant. In a similar survey in the United States, more than 80% of pregnant women reported using medications during pregnancy, with 30% using more than four drugs.4 The contribution of these substances to the incidence of birth defects is thought to be low, accounting for 1 to 3% of all live birth defects.5,6 The emergency physician must understand when the maternal benefit of prescribing a particular agent will outweigh any potential risk of fetal harm, and he or she must be able to discuss those risks and benefits with the patient.
■ PRINCIPLES OF DISEASE Major birth defects affect 3 to 5% of all live births.6 Most are of unknown etiology, but 1 to 3% of these are thought to be
due to pharmaceutical agents.5,6 A teratogen is any chemical, pharmacologic, environmental, or mechanical agent that can cause deviant or disruptive development of the conceptus.5-7 Included in this definition are physical malformations, growth retardation, fetal demise, and functional impairment.6 Although serious effects on the mother are identified immediately, a drug’s teratogenic effect may not be apparent for years. Malformations may range from subtle neurobehavioral effects to devastating physical deformities and physiologic effects, including death.5-7 Why one pregnancy would be affected and not another remains to be elucidated. Highly teratogenic medications seem to be few in number, estimated at well below 50 agents (Box 178-1).7,8 When examining the effects of substances on the outcome of pregnancy, it is important to keep in mind that the process of establishing the risk and safety of drugs in pregnancy is tedious and often flawed. For ethical reasons, few controlled prospective human studies analyzing the risk-benefit relationship for any given exposure are available. As a result, much current knowledge has been derived from case reports, casecontrolled studies, or cohort studies, which are inherently weak in establishing a causal relationship,1,5,6,9 and from animal research. Knowledge extrapolated from animal models, although valuable in determining risk initially, is not always applicable to humans.1,5,6,9 In evaluating data on the relationship between an exposure during pregnancy and a particular outcome, a multitude of confounding factors make the determination of a causal link difficult. The genetic background of the fetus, the timing and duration of the exposure, environmental factors, the occurrence of multiple exposures and the presence of nutritional deficits, maternal illness, and illicit drug use all contribute to the outcome of pregnancy.1,5-9 In the presence of maternal illness, for example, the outcome of pregnancy may be related to the medical condition and not the medication, and separating the risks of an anomaly from the expected background risk may be difficult. The study of teratogenicity is also hindered by several additional factors. First, the history of drug or environmental exposure is often obtained in retrospect, after 9 months of pregnancy and the delivery of an abnormal infant. By that time, significant recall bias may have been introduced, which may depend on the outcome of the birth.9 Second, because many pregnancies are spontaneously aborted before maternal knowledge that conception has occurred, the cited prevalence of druginduced birth defects may not be accurate.5,6,8,9 Finally, as in the case of diethylstilbestrol, teratogenicity may not be appar2313
2314
PART V ■ Special Populations / Section Two • The Pregnant Patient
Drugs and Agents Considered Human
Food and Drug Administration Classification:
BOX 178-1 Teratogens and Developmental Toxins
BOX 178-2 Teratogenic Risk of Drugs
Alkylating agents (busulfan, chlorambucil, cyclophosphamide, etc.) Lead* Aminopterin and methotrexate* Lithium* Amnioglycosides (streptomycin and others) Methimazole Amiodrone Methylmercury* Androgens* Methylene blue Angiotensin enzyme inhibitors* Misoprostol Carbamazepine Nonsteroidal anti-inflammatory agents Carbon monoxide Paramethadione, trimethadione* Chlorobiphenyls* Phenytoin* Cocaine* Polychlorinated biphenyls* Corticosteroids Progestins Coumarin derivatives* Quinine Danazole Diethylstilbestrol* Tetracycline* Ergotamine Ethanol (in large doses)* Tobacco Fluconazol (in high doses) Iodine Thalidomide * Ionizing radiation* Trimethadione* Isotretinoin (systemic) * Valproic acid* Statins
Class A: Controlled studies have shown no risk. Adequate well-controlled studies in pregnant women have failed to show risk to fetus. Class B: No evidence exists of risk for humans. Animal studies show risk or are negative, but no human studies have been done. Class C: Use may engender risk for fetus. Human studies are lacking, and animal studies may be positive or lacking. Potential for benefit may outweigh potential for harm. Class D: Positive evidence of risk is based on studies or postmarketing data. Potential for benefit may outweigh potential for harm. Class X: Drugs are contraindicated in pregnancy based on human or animal studies or postmarketing reports that indicate benefit is clearly outweighed by risk.
*Teratogen. Modified from Shepard TH: Catalog of Teratogenic Agents, 12th ed. Baltimore, Johns Hopkins University Press, 2008; and Fine JS: Reproductive and perinatal principles. In Goldfrank LR, et al (eds): Goldfrank’s Toxicologic Emergencies, 8th ed. New York, McGraw-Hill, 2006, pp 465–485.
ent for years after birth. Large population studies are needed to understand the connection between the outcome of a pregnancy and an associated in utero exposure.10
Classification of Teratogenic Risk To aid physicians in determining the teratogenic potential of a particular medication, the U.S. Food and Drug Administration (FDA) has published a classification system that assigns risk based on currently available human and animal studies and case reports. Drugs are assigned one of five letters—A, B, C, D, and X—depending on the strength of evidence for their safety or teratogenicity (Box 178-2). The FDA classification system has been criticized as oversimplistic and perhaps inaccurate because it relies on data that are generally of poor quality. In addition, using this classification, more than 90% of
drugs approved in the United States between 1980 and 2000 were assigned an undetermined teratogenic risk.11 Furthermore, some clinicians believe that the classification system conveys the incorrect impression that there is a gradation of reproductive risk from exposure across categories (i.e., that risk increases from A to B to C to D to X) and that the drugs within a given category present similar reproductive risks.12 The FDA has acknowledged these problems, and in 2008 it proposed new rules regarding drug labeling during pregnancy and the elimination of the current ABCDX pregnancy categories. A number of clinical teratology resources, such as TERIS, REPROTOX, and REPRORISK (Shepard’s catalogue of teratogenic agents), are now available online. These databases assign teratogenic risk to drugs based on a consensus of opinion of an expert panel.
Drug Transfer across the Placenta The degree to which the fetus is affected by a given pharmaceutical agent and the nature of that effect depend on multiple factors. The transport of maternal substrates to the fetus and of waste products from the fetus to the mother is established during week 5 of gestation.1,5,8,13 Drug transfer across the placenta occurs most commonly by simple passive diffusion or by protein transport. A thin layer of trophoblastic cells is all that separates maternal from fetal circulation. The degree to which a drug gains access to fetal circulation depends on molecular size, ionic state, lipid solubility, and the extent of protein binding. Drugs with a molecular weight of less than 5 kDa readily diffuse. Anionic forms diffuse through the lipid layer more readily than ionized forms. Free drug diffuses more readily than a drug that is bound to plasma proteins. Because fetal pH is slightly more alkalotic than maternal pH, weak organic acids (e.g., salicylate) may become ion trapped in the fetal circulation, increasing fetal exposure.1,5,8,13 Drugs may affect the fetus through a variety of mechanisms. Some drugs may alter the availability of substrates, such as vitamins, glucose, oxygen, and amino acids, needed for normal nutrition and growth.1,5,8,13 Others may directly affect cellular growth and differentiation. The age of the fetus is crucial in determining the impact of any given exposure. During the time of organogenesis (days 21–56 of fetal life), the fetus is much more vulnerable to toxic insults.1,5,8 The major body organs are formed during this period, and exposure to a terato-
Drug Transfer during Lactation For the most part, drugs and substances that are ingested or injected by the mother diffuse passively into milk and then back into the maternal circulation for excretion.14 The amount of drug diffusing into milk depends on many factors. Lipidsoluble and nonionic substances diffuse more readily, and highly protein-bound substances diffuse less readily.14 Whether a substance is concentrated in maternal milk or not, the neonate generally is able to detoxify it with no adverse effects, and only a few drugs pose a serious danger to a breast-feeding infant.14 The interruption of breast-feeding should not be advocated except in rare situations of known drug toxicity to the infant.15 Table 178-1 summarizes the compatibility of medications and their effects in pregnancy and lactation.
■ DRUG THERAPY DURING PREGNANCY In general, the health of the fetus is directly related to the health of the mother. Physicians should never withhold lifesaving medications from pregnant patients because of a reported risk to the fetus and should resuscitate pregnant patients according to Advanced Life Support Guidelines. Physicians may also prescribe any agent that presents maternal benefits that outweigh the risks to the fetus. Included in this category are therapeutic medications for asthma, arrhythmias, status epilepticus, and HIV.
Analgesic Agents Acetaminophen (paracetamol) is safe throughout pregnancy. It is widely used during pregnancy and has not been associated with an increase in the incidence of congenital malformations when therapeutic doses are used.16,17 Statements about its safety also apply to acute and chronic overdose conditions.18,19 However, there is an increase in the incidence of spontaneous abortion and fetal demise, especially when antidote treatment with N-acetylcysteine is delayed.18-20 Acetaminophen is safe during lactation because only a small amount is excreted into breast milk, and the amount that does get through is tolerated by the neonate’s sulfhydration pathway.15,16 Aspirin appears to be safe throughout pregnancy when used in small doses. Early studies of aspirin use during pregnancy linked it to an increased risk of perinatal and neonatal bleeding, increased risk of postmaturity, significant prolongation of labor, low birth weight, neonatal hypoglycemia, metabolic acidosis in the newborn, and neonatal death.16,17 However, in the Perinatal Antiplatelet Review of International Studies Collaboration (PARIS) study, low doses of aspirin were actually found to be beneficial and to reduce the risk of preeclampsia, premature birth, and adverse perinatal outcomes.23 Furthermore, a number of recent meta-analyses in humans failed to demonstrate a teratogenic effect to aspirin, although there was a trend toward a slightly increased incidence of gastroschisis when used in the first trimester.21,22
2315
Non-aspirin NSAIDs should be avoided in the first and third trimesters but are considered safe in the second trimester. Use of NSAIDs in the first trimester has been associated with a small increase in cardiac defects, oral clefts, and gastroschisis.16,17,24 Use of NSAIDs at or near term has been associated with premature closure of the ductus arteriosus, periventricular hemorrhages in the offspring, oligohydramnios, and fetal nephrotoxicity.16,17 Additionally, a number of population-based cohort studies have found that NSAIDs are associated with an increased risk of spontaneous abortion, preterm birth, and low birth weight.25 NSAIDs in general appear to be safe during lactation when used for short durations.16,17 The short-term use of opiates appears to be safe in pregnancy. Because opiates’ sedative effects extend to the fetus, caution should be used when prescribed at term. Chronic use of opiates is discouraged in general because it may result in maternal as well as fetal addiction. Because opiates are poorly concentrated in milk, opiate analgesia may be used safely during breast-feeding.15
Antibiotics First- through fourth-generation penicillins and their derivatives (including procaine, benzathine, clavulanate, sulbactam, and tazobactam) are considered safe for use in pregnancy, as is oral probenecid.16,17,26,27 Penicillins are considered safe during breast-feeding, but their use may interfere with culture results if a neonatal fever workup is required.15-17 First- through fourth-generation cephalosporins appear to be safe for use during pregnancy, although there are no controlled studies examining their safety.16,17,26-28 Some cephalosporins are excreted into breast milk and may have the same implications on the workup of neonatal sepsis as described for penicillin.15-17 Chloramphenicol is safe during pregnancy except at term. No relationship has been found between the use of chloramphenicol and congenital anomalies.16,17,27,29 Although it is considered safe throughout most of pregnancy, chloramphenicol should be used with caution at term. It has been associated with the development of cardiovascular collapse (the “gray baby” syndrome) in a neonate.16,17,26,27 The safety of chloramphenicol during breast-feeding is unknown; however, due to its potential toxicity, it is not recommended for use during lactation.15–17 The macrolides erythromycin, azithromycin, and clarithromycin are considered to be safe for use in pregnancy and compatible with breast-feeding, although there are no wellcontrolled studies examining their effects on the fetus.15–17 Some reports have linked erythromycin to pyloric stenosis, but these studies were not controlled.16,17,26,27 The estolate salt of erythromycin has also been associated with the development of hepatotoxicity in pregnant women and should be avoided during pregnancy.16,17,26,27 Clarithromycin has been associated with an increased risk of fetal and embryonic death as well as congenital malformations in animal species. To date, however, this has not been shown in humans. In addition, a prospective controlled multicenter study comparing the outcomes of pregnancies exposed to clarithromycin to matched controls did not find any differences in the types or patterns of malformations between the two groups.29,30 However, there appeared to be an increased number of spontaneous abortions in exposed women, which may have been due to confounding factors, and further studies are warranted. Azithromycin is poorly Text continued on p. 2320
Chapter 178 / Drug Therapy in Pregnancy
gen at this time may result in major anatomic defects. The central nervous system (CNS) develops over a longer period (10–17 weeks) so that later exposures may affect neurologic development and subsequent function. Exposure after the period of organogenesis may affect the growth and development of the fetus but does not have an impact on organogenesis; however, it most likely affects fetal growth.1,5,8
2316
PART V ■ Special Populations / Section Two • The Pregnant Patient
Table 178-1 Summary of Medication Safety in Pregnancy and Lactation MEDICATION
PREGNANCY
LACTATION
Analgesic agents Acetaminophen or paracetamol Safe (B) Safe Nonsteroidal anti-inflammatory drugs (NSAIDs) Salicylate Not recommended (D) Safe for short-term use Other NSAIDs Not recommended (D) Safe Opiates: Most opiates are considered safe for short-term use but are all reclassified in category D if used for prolonged periods or if used in high doses at term (due to respiratory depression in newborn). Do not use opiates combined with aspirin or NSAIDS. Morphine—short-term use Safe (C) Safe Fentanyl—short-term use Safe (C) Safe Methadone Safe (B/C) Safe Meperidine—short-term use Safe (C) Safe Codeine Possible risk (C) Not advised if longer than 2 days Oxycodone—short-term use Safe (C) Probably safe Hydrocodone—short-term use Safe (C) Probably safe Hydromorphone—short-term use Safe (B) Probably safe Oxycodode—short-term use Safe (B) Probably safe Antibiotics: Use of antibiotics near term may interfere with culture results in neonates. Penicillins Safe (B) Safe First-generation penicillins: penicillin G, benzathine penicillin, bicillin, penicillin VK Safe (B) Safe Second-generation penicillins: oxacillin, dicloxacillin, nafcillin Safe (B) Safe Third-generation penicillins: ampicillin, ampicillinsulbactam, amoxicillin, amocillin-clavulanate Fourth-generation penicillins: ticarcillin, ticarcillinSafe (B) Probably safe clavulanate, piperacillin, piperacillin-tazobactam, carbenicillin Cephalosporins First generation: cephalexin, cefazolin, cefadroxil Safe (B) Safe Safe (B) Safe Second generation: cefuroxime, cefaclor, cefoxitin, cefprozil Safe (B) Probably safe Third generation: cefdinir, cefotaxime, ceftazidime, ceftriaxone, cefpodoxime, ceftizoxime Safe until term (C) Possible toxicity Chloramphenicol: Do not use at term because it can cause “gray baby syndrome.” Macrolides Safe (B) Safe Erythromycin: Do not use estolate salt. Azithromycin Safe (B) Safe Clarithromycin Probably safe (C) Probably safe Sulfonamides: May cause kernicterus in newborn if Not recommended near term (C) Safe, except in premature infants given in third trimester. or infants with G6PD deficiency or hyperbilirubinemia Quinolones First generation: nalidixic acid Moderate risk (C) Safe Second generation: ciprofloxacin, norfloxacin, ofloxacin Small risk (C) Probably safe Third generation: levaquin Small risk (C) Probably safe Fourth generation: gatifloxacin, moxifloxacin Small risk (C) Probably safe Aminoglycosides Not recommended (D) Safe Clindamycin Safe (B) Safe Vancomycin Safe (B) Safe Linezolid: Maternal benefit may outweigh risks to fetus Unknown (C) Safety unknown or embryo. Tetracyclines Not recommended (D) Not recommended Nitrofurantoin: May cause hemolytic anemia in newborn Safe (B), except in third trimester Probably safe if used in third trimester. Metronidazole Contraindicated in first trimester, safe Not recommended in second and third trimesters (B) Antifungals Nystatin Safe (B/C) Safety unknown Clotrimazole Safe (B/C) Safety unknown Ketoconazole Probably safe (C) Probably safe Fluconazole Not recommened in high doses (C) Safe Terbinafine Low risk (B) Not recommended
2317
Table 178-1 Summary of Medication Safety in Pregnancy and Lactation—cont’d PREGNANCY
LACTATION
Antituberculous medications in general present a maternal benefit that is much greater than the fetal or embryonic risk and may be prescribed during any stage of pregnancy when indicated. INH Safe (C) Probably safe Rifampin Probably safe (C) Probably safe Ethambutol Safe (C) Probably safe Antiviral agents Antiherpetic medications Acyclovir Safe (B) Safe Valacyclovir Safe (B) Safe Famciclovir Small risk (B) Potentially toxic Anti-influenza agents Amantadine Possible risk (C) Safety unknown Oseltamivir Safe in animals (C) Safety unknown Anti-HIV medications in general present a maternal benefit that is much greater than the fetal or embryonic risk and may be prescribed during any stage of pregnancy when indicated. Reverse transcriptase inhibitors Zidovudine Caution in first trimester (C) Not recommended Lamivudine Caution in first trimester (C) Not recommended Didanosine Caution in first trimester (B) Not recommended Tenovir Caution in first trimester (B) Not recommended Indinavir Caution in first trimester (C) Not recommended Protease inhibitors Ritonavir Caution in first trimester (B) Not recommended Nelfinavir Caution in first trimester (B) Not recommended Anticoagulants Warfarin Contraindicated Safe Heparin Safe (B) Safe Low-molecular-weight heparin Safe (B) Safe Thrombolytics: Benefits to the mother generally outweigh the risk to the fetus. Alteplase Safe (C) Probably safe Reteplase Safe (C) Probably safe Urokinase Safe (B) Probably safe Streptokinase—no human data Safe in animals (C) Hold breast-feeding Tenectaplase—no human data Safe in animals (C) Hold breast-feeding Anticonvulsants: Benefits to the mother outweigh the risks to the fetus and embryo. Monotherapy is recommended. Use of highly teratogenic anticonvulsants is recommended in refractory cases only. Carbamazepine Teratogen (D) Probably safe Valproic acid Teratogen (D) Probably safe Phenobarbital Not recommended (D) Not recommended Phenytoin Teratogen (D) Safe Lamotrigine Small risk in animals (C) Not recommended Levetiracetam Small risk in animals (C) Safety unknown Topiramate Small risk in animals (C) Safety unknown Gabapentin Small risk in animals (C) Probably safe Sedative-hypnotics Benzodiazepines: Results are inconsistent. There may be a small risk of abnormalities. In the acute short-term treatment of status epilepticus, agitated delirium, and alcohol or benzodiazepine withdrawal, benefits to the mother outweigh risks to the fetus or embryo. Not recommended for long-term use. Diazepam—low risk in first and third trimesters Safe acutely; unsafe for chronic use (D) Potential toxicity Lorazepam—low risk in first and third trimesters Safe acutely; unsafe for chronic use (D) Potential toxicity Chlordiazepoxide—low risk in first and third trimesters Safe acutely; unsafe for chronic use (D) Potential toxicity Oxazepam—low risk in first and third trimesters Safe acutely; unsafe for chronic use (D) Potential toxicity Midazolam—low risk in first and third trimesters Safe acutely; unsafe for chronic use (D) Potential toxicity Barbiturates Methohexital Safe (C) Safety unknown Thiopental Safe (C) Safety unknown Ketamine: Risk mainly with high doses close to delivery. Safe (B) Probably safe after 12 hr Propofol Safe (B/C) Not recommended Etomidate Safe (B/C) Probably safe Paralytic agents Depolarizing agents Succinylcholine Low risk, especially around delivery (C) Probably safe Nondepolarizing agents Rocuronium Limited data (C) Probably safe Vecuronium Limited data (C) Probably safe Continued
Chapter 178 / Drug Therapy in Pregnancy
MEDICATION
2318
PART V ■ Special Populations / Section Two • The Pregnant Patient
Table 178-1 Summary of Medication Safety in Pregnancy and Lactation—cont’d MEDICATION
PREGNANCY
LACTATION
Antiarrythmics: In cases of refractory arrhythmias, the benefits to the mother may outweigh risks to the fetus or embryo. Adenosine Safe (C) Safe Amiodarone Not recommended (D) Contraindicated Digoxin—caution in third trimester (oxytocic) Safe (C) Safe Disopyramide Caution in third trimester (B) Probably safe Encainide Limited data (C) Probably safe Flecainide Limited data (C) Not recommended Quinidine Safe (C) Not recommended Procainamide Safe (C) Not recommended Ibutilide Risk in animals Not recommended Lidocaine Safe (B) Probably safe Sotalol Probably safe (C) Not recommended Antihypertensives Angiotensin antagonists Angiotensin-converting enzyme inhibitors Not recommended (D) Some are safe Angiotensin II receptor antagonists Not recommended (D) Safety unknown Beta-blockers Labetalol Probably safe (C) Probably safe Atenolol Caution in second and third trimesters Potential toxicity (D) Esmolol Caution in third trimester (C) Safety unknown Metoprolol Caution in second and third trimesters Potential toxicity (D) Propranolol Caution in second and third trimesters Probably safe (D) Calcium channel blockers Diltiazem—tocolytic Probably safe (C) Probably safe Verapamil Probably safe (C) Probably safe Amlodipine No data (C) No data Nifedipine—not sublingual Probably safe (C) Probably safe Diuretics: Caution for dehydration and electrolyte abnormalities. Furosemide Low risk (C) Probably safe Bumetanide Low risk (C) Probably safe Ethacrynic acid Low risk (B) Probably safe Torsemide Low risk (B) Probably safe Safe (B) Safe Hydrochlorothiazide—contraindicated in gestational HTN Nitrates Nitroprusside Some risk (C) Potential toxicity Nitroglycerin Probably safe (C) Probably safe Vasodilators Hydralazine Safe (C) Safe Alpha effectors Clonidine Caution in third trimester (C) No data Other Methyldopa Safe (B) Safe Fenoldopam No data (B) No data Medications used in the treatment of asthma, allergies, and upper respiratory infection: For asthmatic patients, maternal benefits outweigh risks to fetus or embryo. Beta-adrenergics considered safe for short-term use Epinephrine Risk (C) Potential toxicity Metaproterenol Safe (C) Probably safe Salmeterol Limited data (C) Probably safe Albuterol Safe (C) Probably safe Terbutaline Low risk (C) Probably safe Anticholinergic agents Ipratropium Safe (B) Probably safe Mast cell stabilizers Cromolyn sodium Safe (B) Probably safe Leukotriene inhibitors Zafirlukast Risk (B) No data Zileuton Risk (C) Not recommended Montelukast Risk (B) No data; probably compatible
2319
Table 178-1 Summary of Medication Safety in Pregnancy and Lactation—cont’d PREGNANCY
Corticosteroids safe for short-term use: Human data suggest an increased risk of orofacial clefts. Prednisolone Risk (C) Prednisone Risk (C) Methylprednisolone Risk (C) Antihistamines Chlorpheniramine Safe (B) Diphenhydramine Safe (B) Dimenhydramine Safe (B) Doxylamine Safe (B) Hydroxyzine Risk (C) Meclizine Safe (B) Cetirizine Probably safe (B) Fexofenadine Class C Loratadine Probably safe (B) Decongestants Pseudoephedrine Risk (C) Antiemetics Dopamine antagonists Promethazine Safe (C) Prochlorperazine Safe (C) Metoclopramide Safe (B) 5-HT3 antagonists—generally safe Dolasetron Low risk (B) Granisetron Low risk (B) Ondansetron Safe (B) Medications used in the treatment of diabetes Insulin Safe (B) Sulfonylureas Glipizide Low risk (C) Glyburide
Low risk (C)
Chapter 178 / Drug Therapy in Pregnancy
MEDICATION
LACTATION
Safe Safe Safe Not recommended Not recommended Not recommended Not recommended Not recommended Not recommended Not recommended Safe Safe Probably safe Not recommended Potential toxicity Potential tocxicity Probably safe Probably safe Probably safe Safe Probably safe. Caution: Nursing infants should be monitored. Probably safe. Caution: Nursing infants should be monitored. Probably safe Probably safe Probably safe
Metformin Moderate risk (C) Pioglitazone Moderate risk (C) Rosiglitazone Moderate risk (C) Antacids H2 blockers Famotidine Low risk (B) Probably safe Ranitidine Safe (B) Probably safe Nizatidine Low risk (B) Probably safe Cimetidine Safe (B) Safe Proton pump inhibitors Omeprazole Low risk (C) Safety unknown Esomeprazole Low risk (B) Safety unknown Lansoprazole Low risk (B) Safety unknown Pantoprazole Low risk (B) Safety unknown Antidotes and toxicology: When indicated, the benefits to the mother will outweigh the possible risks to the fetus. Antidote: acetaminophen overdose Safe (B) Safe N-acetylcysteine Universal antidote Charcoal Safe (B) Safe Antidote: iron toxicity Deferoxamine Probably safe (C) No data Antidote: digitalis DIG Fab No data (C) No data Antidote: benzodiazepines Flumazenil No data (C) No data Antidote: toxic alcohols Fomepazole No data (C) No data Antidote: cyanide poisoning Hydroxycobalamine Safe (B) Safe Antidote: methemoglobinemia Methylene blue Risk noted (C) Not recommended
Continued
2320
PART V ■ Special Populations / Section Two • The Pregnant Patient
Table 178-1 Summary of Medication Safety in Pregnancy and Lactation—cont’d MEDICATION
Antidote: narcotics Naloxone Universal antidote PEG-ELS Antidote: anticholinergics Physostigmine Antidote: organophosphates Pralidoxime Antidote: isoniazid overdose Pyridoxine Antidote: lead poisoning Succimer
PREGNANCY
LACTATION
Safe (B)
Safe
Safe (B)
Safe
No data
No data
No data
No data
Safe (B)
Safe
No data
No data
G6PD, glucose-6-phosphate dehydrogenase; PEG-ELS, polyethylenegycol-electrolyte lavage solution.
concentrated in breast milk and may be the preferred agent in lactating mothers.15-17,27 Sulfonamides are safe for use in the second trimester and possibly safe in the first trimester, but they should be avoided at term. The primary use of sulfonamides in the emergency department is in the treatment of uncomplicated urinary tract infection, and in this circumstance, sulfamethoxazole is combined with trimethoprim. Trimethoprim is a folate antagonist and has traditionally been contraindicated in pregnancy because of an increased risk of neural tube defects. The sulfonamides readily cross the placenta to the fetus during all stages of gestation. Fetal levels may reach 90% of maternal plasma concentrations. Although sulfamethoxazole has been associated with an increase in congenital malformations in animals, most reports of sulfonamide exposure during pregnancy in humans have failed to demonstrate such an association.16,17 Sulfonamides are contraindicated in pregnancy near term because they theoretically compete with bilirubin for protein-binding sites, leaving large amounts of free bilirubin to diffuse, be deposited in the infant’s brain, and cause kernicterus.16,17,31,32 To date, however, this complication has not been reported in neonates, presumably because free bilirubin is effectively cleared by the placental circulation. In contradistinction, kernicterus has occurred in newborns exposed to sulfonamides after birth.16,17 Sulfonamides are excreted into breast milk in low concentrations and are generally tolerated by a healthy neonate. They should be avoided, however, in ill or premature infants and in infants with hyperbilirubinemia or glucose-6-phosphate dehydrogenase deficiency.15 Aminoglycosides should be avoided during pregnancy. These drugs readily cross the placenta and their use in pregnancy has been linked to fetal ototoxicity and nephrotoxicity, especially when high doses are used.16,17 Gentamicin is secreted in small amounts in breast milk and is poorly absorbed from the gastrointestinal tract. It is compatible with lactation.15-17 Tetracycline should be avoided during pregnancy because it has been associated with the development of fatal fatty liver in pregnant women.16,17,26,27 It readily crosses the placenta and reaches the fetus, where it chelates calcium, causing abnormalities in bone growth and staining of decidual teeth. Tetracycline has also been associated with fetal genitourinary anomalies, inguinal hernias, and limb abnormalities.16,17,26,27 Doxycycline does not bind to calcium and is associated less with stained teeth than is tetracycline. Also, it does not appear to cause an increase in any type of congenital malformation. Despite these findings, doxycycline is not advocated for longterm use in pregnancy.16,17,26,27
Because tetracycline binds to breast milk calcium, only a small amount reaches the nursing infant, and it may be used for short periods (≤10 days) during breast-feeding.15-17 Because it does not bind to breast milk calcium, doxycycline is present in greater quantities in breast milk and is not recommended for use for prolonged periods.15-17,26,27 Fluoroquinolones have been linked to numerous toxic effects on bone and cartilage growth in animal models and have been discouraged from use during pregnancy, particularly during the first trimester.15-17,26,27 A 1998 prospective multicenter study, however, found no increase in premature birth, fetal distress, low birth weight (400 mg/day) and has been associated with an increased incidence of craniofacial and cardiovascular defects in offspring and multiple abnormalities of the skeleton and cartilage.16,17,27,37 These anomalies were not noted when lower doses were used or with single-dose (150 mg) therapy for vaginal candidiasis.16,17 Ketocanazole, fluconazole, and itraconazole are excreted into breast milk. Based on the safe use of ketocanazole in neonates and the lack of negative reports, it is considered compatible with breast-feeding.15-17
Amantadine appears teratogenic in some animals but not others. Its use in pregnant women is very limited, and one cannot draw any conclusions.16,17 Oseltamivir has no effect on embryonic or fetal development in animal studies. There appear to be no reports of its use in pregnancy, but based on its safety in animals, it may be used in human pregnancy, albeit cautiously.16,17
Antituberculous Agents Untreated tuberculosis places the mother and fetus at greater risk than does the use of antituberculous medications. In addition, in a review of antituberculous treatment during pregnancy, no association was found between these medications and congenital malformations.16,17,27,38 Rifampin crosses the placenta and occasionally has been implicated in case reports of congenital anomalies and with hemorrhagic disease of the newborn.16,17 Because there are no controlled studies documenting these effects, rifampin continues to be recommended as first-line therapy along with isoniazid for treatment of pregnant women with tuberculosis. Ethambutol crosses the placenta but has not been associated with any congenital defects.16,17,27 All three antituberculous medications are considered compatible with breast-feeding.15-17
Antiviral Agents Antiherpetic Drugs Acyclovir is a purine analogue commonly used in the treatment of herpesvirus infections. During pregnancy, acyclovir is indicated for life-threatening maternal herpes simplex virus infections, such as disseminated disease, herpes encephalitis, and varicella pneumonia, which carries a maternal mortality of 44% if untreated.16,17,27,39 The Centers for Disease Control and Prevention also recommends treatment of the first episode of genital herpes during pregnancy with oral acyclovir.39,40 In humans, acyclovir readily crosses the placenta and reaches higher concentrations in fetal circulation than in maternal circulation. There are no reports of teratogenicity or adverse effects in the fetuses or newborns of mothers using acyclovir or valaciclovir.16,17,27,39-41 Famciclovir was associated with congenital cardiovascular anomalies, hepatotoxicity, and death.16,17 Acyclovir is concentrated in milk, in which levels may be higher than in plasma. Because there are no reported adverse outcomes in infants of mothers taking acyclovir or infants treated with acyclovir for disseminated herpes, it is considered safe during breast-feeding.15-17
Anti-HIV Drugs Anti-HIV drugs may be indicated immediately after a needlestick injury or sexual contact with an infected individual. No specific pattern of birth defects has been described with the use of these drugs, but there are a number of unanswered questions relating to the drugs’ mutagenesis and carcinogenesis and their long-term effects on the liver, heart, and reproductive system.15-17 Animal and human data suggest that didanosine, lamivudine, stavudine, zidovudine, and zalcitabine present some risk, albeit small, of structural malformations and mitochondrial dysfunction in the developing fetus. However, even if a negative association is proven, the risk of morbidity and mortality from HIV infection far outweighs the risk of toxicity of most of these substances.15-17,40,42 Similarly, no specific pattern of birth defects has been described with protease inhibitors such as ritonavir and nelfinavir. When indicated, the benefits of treatment outweigh the drugs’ toxicities.15-17,42
Anticoagulants Warfarin (Coumadin) is a known human teratogen and affects 4 or 5% of exposed fetuses. The risk from exposure is greatest during 6 to 9 weeks of gestation and seems to be dose dependent.16,17,27 The fetal warfarin syndrome is associated with multiple abnormalities, such as hypoplasia of the nasal bones, midline dysplasia including agenesis of the corpus callosum, optic atrophy and blindness, mental retardation, seizures, and stippling of the bones with scoliosis and shortening of limbs.16,17,27,43 Because warfarin is so highly protein bound, only a little is secreted into milk, and use by breast-feeding mothers is acceptable.16,17 Caution should be used in breast-feeding premature infants because they may be at increased risk for intraventricular hemorrhage.16,17 Unfractionated heparin is a highly charged heterogeneous molecule with a molecular weight between 5 and 35 kDa. It does not cross the placenta and does not present a direct risk to the fetus. Early reports on the use of heparin for the prevention or treatment of venous thromboembolism during pregnancy concluded that the risks to the fetus from prematurity, stillbirth, and hemorrhage might affect one third of infants. Recently, however, the increased risks previously associated with heparin were determined to be related to underlying maternal medical problems rather than heparin.16,17,26,27 When anticoagulation during pregnancy is required, heparin is considered the agent of choice.16,17,44 Its use is sometimes associated with maternal osteopenia and immune-mediated thrombocytopenia.16,17,27 Patients need careful monitoring for these adverse effects. The risk of maternal hemorrhage at delivery is significant. Because of its high molecular weight, heparin is not excreted in breast milk and is compatible with breast-feeding.15-17 Low-molecular-weight heparin may be used during pregnancy and in the postnatal period for therapeutic or prophylactic anticoagulation.16,17,27 All currently available lowmolecular-weight heparin products have been used safely
Chapter 178 / Drug Therapy in Pregnancy
Antifungal Medications
PART V ■ Special Populations / Section Two • The Pregnant Patient
2322
during pregnancy. Data are limited, however, because of their relatively recent introduction. Thrombolytic agents have been used successfully in pregnant women in cases of life-threatening pulmonary embolus or myocardial infarction. Experience with these agents during pregnancy, however, remains limited. To date, no teratogenic effects have been reported in humans, but maternal hemorrhage occurred when alteplase was used during the intrapartum period.16,17,45 Most thrombolytics are thought to be compatible with breast-feeding because of their short half-life.16,17
Anticonvulsants The occurrence of generalized seizures during pregnancy has been associated with an increased risk of spontaneous abortion, hypoxic injury to the fetus, and impaired neuropsychological functioning.16,17 Anticonvulsants are known teratogens, however, and 30% of neonates exposed to commonly used anticonvulsants exhibit congenital anomalies.16,17,27,46 The risks for birth defects increase with the duration of exposure and with the number of agents used.16,17,27,46 In an observational study of 25 epilepsy centers, it was noted that valproate use during pregnancy was associated with the most frequent serious adverse effects on the pregnancy and fetus (20.3% incidence of serious adverse outcomes) compared with phenytoin, carbamazepine, and lamotrigine (10.7, 8.2, and 1.0%, respectively).47 Despite the risks, most practitioners believe that it is important to control seizures during pregnancy. Monotherapy is the most appropriate option and is recommended at the lowest effective anticonvulsant dose. Dividing the daily dose to decrease peak plasma levels may be considered. Adjustment of the dosage upward is often required to maintain adequate seizure control.16,17 Phenytoin is a human teratogen that readily crosses the placenta. The parent compound and all metabolites have been identified in fetal tissues. Of chronically exposed fetuses, 5 to 10% develop the fetal hydantoin syndrome.16,17,27,48 This syndrome is characterized by varying degrees of ossification abnormalities of the extremities and digits; craniofacial abnormalities including cleft lip and palate; impaired growth; delayed neurologic development; and cardiovascular anomalies, including atrial septal defects, ventricular septal defects, coarctation of the aorta, and endocardial cushion defects. Phenytoin has also been associated with hemorrhagic disease of the newborn, presumably because it competitively inhibits placental transport of vitamin K.16,17,27,48 To avoid this rare complication, some clinicians have advocated the use of vitamin K during the last month of pregnancy, but evidence does not support its use.47 Phenytoin has been linked to a variety of tumors in infants. Phenytoin use is considered safe in breast-feeding.15-17 Carbamazepine use during pregnancy is associated with a syndrome similar to fetal hydantoin syndrome, which is thought to be secondary to a toxic, teratogenic metabolite and not the parent compound.16,17,27,49 In one study, carbamazepine was associated with a twofold increase in major congenital abnormalities in exposed fetuses compared with nonexposed fetuses.16,17,27,49 These abnormalities include craniofacial defects, fingernail hypoplasia, neural tube defects, and developmental delay. Carbamazepine has also been reported to induce hemorrhagic disease of the newborn.39 The use of carbamazepine is considered compatible with breast-feeding.15-17 Valproic acid, a class D medication that should not be used in pregnancy, is an eight-carbon, branched-chain carboxyl acid that has been approved for the treatment of absence seizures since 1978.16,17,27,50 It is teratogenic in laboratory animals and in
humans. It readily crosses the placenta and concentrates in the fetus. Many authors have described a syndrome of defects associated with the use of valproic acid. The characteristics of the syndrome include multiple minor facial anomalies, low birth weight, delayed neurologic development, congenital heart defects, neural tube defects, hypospadias, strabismus, nystagmus, tracheomalacia, afibrinogenemia, and hyperglycemia.16,17,27,50 Valproic acid is present in breast milk in low levels and is considered safe during breast-feeding.15-17 Phenobarbital is considered a class D medication in pregnancy. It is associated with a slightly increased risk of congenital abnormalities, including congenital heart disease and cleft lip or palate and some minor malformations associated with the fetal hydantoin syndrome.16,17,46 It is occasionally associated with hemorrhagic disease in the newborn and may result in neonatal withdrawal. Breast-fed infants of mothers taking phenobarbital have developed toxicity characterized primarily by sedation. These infants must be monitored closely for sedation while breast-fed and after breast-feeding for symptoms of withdrawal.15-17
Newer Anticonvulsant Medications No adequate studies of human teratogenicity have been published regarding felbamate, levetiracetam, gabapentin, and lamotrigine, but they appear to be safe during pregnancy. Lamotrigine has been associated with a slightly increased incidence of cardiovascular, craniofacial, gastrointestinal, and genitourinary birth defects.16,17,51 Safety in lactation is unknown.
Cardiovascular Medications Antidysrhythmics Adenosine is a naturally occurring compound that is metabolized quickly in the body. It has been used safely throughout pregnancy and is the drug of choice for terminating maternal supraventricular tachycardia,52 despite the absence of largescale studies.16,17,27,48 Adenosine has also been used safely in terminating incessant tachycardia in the fetus.53 Adenosine is likely safe in lactation.16,17 Amiodarone, a class D agent, is not recommended in pregnancy, except in refractory cases of supraventricular and ventricular tachycardias of the mother or the fetus.54,55 It contains large amounts of iodine and has been associated with congenital goiter and transient neonatal hyperthyroidism and hypothyroidism.16,17,56,57 In addition, amiodarone use during pregnancy has been linked to many congenital abnormalities, including growth retardation, structural cardiac abnormalities, corneal deposits, and developmental delay.16,17,54 Because of its high iodine content, its excretion into milk, and its long elimination half-life, amiodarone should not be used in nursing mothers.16,17 Digoxin, disopyramide, and quinidine are all considered safe for use during pregnancy and lactation.16,17 None have been linked to congenital defects in humans or animals. Of the three agents, digoxin and quinidine have the longest safety records in pregnancy and are first-line agents for the treatment of significant maternal dysrhythmias.16,17 They have also been successfully used in fetal tachycardia.58,59 However, maternal overdoses of digoxin resulting in fetal death have occurred.60 Although considered safe during pregnancy, disopyramide has been associated with premature uterine contractions and labor.61 Lidocaine is a weak base. It rapidly crosses the placenta and becomes ion trapped in the fetus. There is no evidence of a
Vasopressors Dobutamine is an inotrope used in the setting of cardiac dysfunction and sepsis. Data from animal studies have not revealed any untoward reproductive effects. Effects in humans are not known, but one case report did not reveal any effects on the fetus.16,17,65 Dopamine and other vasopressors offer a maternal benefit that far outweighs the possible deleterious effects on the fetus and should not be withheld if indicated. Dopamine has been associated with increased uterine vascular resistance in animal studies, but no significant fetal side effects directly related to the drug have been reported. In addition to its use in maternal shock, dopamine has been successfully used in low doses to improve cardiac and urine output in patients with preeclampsia and oliguria.66 Epinephrine has been used to treat shock from any cause during pregnancy. However, it has been associated with anoxic injury to the fetus, intracranial hemorrhage, and an increased incidence of inguinal hernias.16,17,67 Safety during breastfeeding has not been studied. Norepinephrine is associated with an increased incidence of cerebral hemorrhage, skeletal abnormalities, and a significant decrease in placental blood flow and fetal oxygenation in animal studies.16,17,68 Its safety during breast-feeding has not been studied.
Antihypertensives Hypertension complicates 12% of pregnancies and accounts for 18% of maternal deaths in the United States.69 Previously,
2323
the most commonly used drug in hypertensive emergencies was hydralazine, but other medications currently available, such as labetolol and nifedipine, appear to be as effective and possibly safer than hydralazine. Angiotensin-converting enzyme (ACE) inhibitors are classified as category D drugs and are contraindicated for use during pregnancy. ACE inhibitors are embryocidal in animals and increase the rate of stillbirths in some species. Although they seem to be safe in humans during the first trimester of pregnancy, many adverse fetal effects have been noted with their use during the second and third trimesters, precluding their use.16,17,37,70 Reported adverse neonatal effects include oligohydramnios, anuria, renal agenesis resulting in death, increased risk of stillbirth, intrauterine growth retardation (IUGR), fetal skull abnormalities, pulmonary hypoplasia, respiratory distress syndrome, and fetal and neonatal hypotension. Captopril and enalapril are considered compatible with breast-feeding.15-17 Angiotensin II receptor antagonists should be avoided during pregnancy because their use has been reported to result in fetal abnormalities similar to the abnormalities seen with ACE inhibitors, including renal agenesis, neonatal anuria, oligohydramnios, IUGR, persistent patent ductus arteriosus, abnormal ossification, and death.16,17 Their safety in lactation is unknown. Beta-blockers have become a first-line treatment for hypertension in pregnancy.71,72 All beta-adrenergic blocking agents cross the placenta. The most experience with beta-blockers has been with women requiring treatment during the last trimester of pregnancy, at which time they seem to be safe. Long-term in utero exposure and first-trimester exposure have not been studied.16,17,37 Labetalol is the antihypertensive of choice during pregnancy.71,72 It has not been associated with any teratogenic effects in animal studies. Human reports of its use in the treatment of hypertension during pregnancy have not revealed any significant effect on the fetal birth weight or fetal heart rate.16,17,37 Transient neonatal hypotension and bradycardia may be observed when used at term. However, compared with traditional therapies for pregnancy-induced hypertension, labetolol appeared to reduce the blood pressure more smoothly than either hydralazine or diazoxide.16,17,71,73 Furthermore, it was associated with fewer cesarean sections than either of the two drugs.73 Atenolol and metoprolol are considered safe in pregnancy but only when used for short periods of time.16,17,73 There are reports of fetal harm when atenolol is used in the first trimester. Atenolol has also been associated with IUGR when used for prolonged periods during pregnancy, and when given near term, it is associated with persistent beta blockade in the newborn.16,17,73 Similarly, propranolol is associated with fetal and neonatal adverse effects, especially when doses exceeding 160 mg/day are used. These adverse effects include IUGR, hypoglycemia, bradycardia, respiratory depression at birth, and hyperbilirubinemia. Esmolol has also been associated with fetal bradycardia, neonatal bradycardia, and hypotonia as well as fetal distress requiring emergent cesarean section.74 It should therefore be used only if the benefits to the mother outweigh the risks to the fetus and if other options have failed. Beta-blockers are reportedly safe in breast-feeding, but close monitoring of the infant for adverse effects is recommended.15-17 Calcium channel blockers are indicated in the treatment of hypertension and a number of supraventricular rhythm disturbances during pregnancy.16,17,71 Nifedipine and cardizem have also been used as tocolytic agents. In addition, verapamil has been used to terminate maternal as well as fetal tachycardia.75 Despite negative reproductive studies in animals, the calcium channel blockers are used extensively during the second and
Chapter 178 / Drug Therapy in Pregnancy
link between lidocaine and any fetal malformations.16,17 High doses used near term are associated with neonatal CNS depression, apnea, hypotonia, seizures, and bradycardia. Lidocaine is considered compatible with breast-feeding.15-17 Procainamide is also well tolerated and should be considered a first-line treatment of wide-complex tachydysrhythmias during pregnancy.55,58 The use of procainamide in nursing mothers is controversial because it and its metabolite, N-acetyl procainamide, have been found in breast milk.15-17 Encainide and flecainide are newer class IC antiarrhythmic agents that are structurally related to procainamide. Both have been used safely to terminate maternal and fetal tachycardia.16,17 A few negative fetal effects have been noted with the use of flecainaide, including hyperbilirubinemia, hepatotoxicity, and loss of fetal heart rate variability.62,63 Both encainide and flecainide are found in breast milk. Although experience with encainide is limited, the AAP considers flecainide compatible with breast-feeding.16,17 Ibutilide is a class III antiarrhythmic used to terminate atrial fibrillation and flutter. Although there are no reports of its use in pregnancy in humans, when used in high doses, ibutilide was found to be teratogenic in rats.16,17 Ibutilide may be used in refractory cases in which the benefits of therapy outweigh any fetal risk.16,17 Sotalol does not appear to have teratogenic effects in animals.16,17 It has been used in pregnant women to treat hypertension. In these cases, bradycardia in the newborn was noted, which persisted for 24 hours. Sotalol has also been used successfully to terminate in utero fetal supraventricular tachycardia.64 Isoproterenol is indicated for refractory high-grade atrioventricular block and for torsades de pointes associated with prolonged QT interval. Data from animal studies have not shown any association between isoproterenol and developmental toxicity. It is also considered compatible with breast-feeding.16,17
PART V ■ Special Populations / Section Two • The Pregnant Patient
2324
third trimesters in humans and are considered safe for use during these stages of pregnancy.16,17,37,76 Whereas cardizem appears to be safe at all stages of pregnancy, nifedipine was associated with fetal distress secondary to maternal hypotension when it was used sublingually.16,17 In addition, when used in conjunction with magnesium, nifedipine appeared to potentiate the neuromuscular blocking effects of magnesium, resulting in profound muscle weakness, difficulty in swallowing, and paradoxical respirations.77 Calcium channel blockers are considered safe for use during breast-feeding.15-17 Thiazide diuretics have been used successfully for the treatment of hypertension in pregnancy but may result in electrolyte abnormalities in neonates when given near term.16,17,71 An increase in perinatal mortality and congenital defects possibly caused by volume depletion has been reported.16,17 Firsttrimester use has been associated with an increase in congenital anomalies.16,17 Diuretics are not recommended for the treatment of pregnancy-induced hypertension because of the maternal hypovolemia characteristic of this disease. Other risks to the pregnancy include higher rates of uterine inertia and meconium staining.78 In the neonate, there is a higher incidence of hypoglycemia, thrombocytopenia, hyponatremia, hypokalemia, and death from maternal complications.16,17 Moreover, thiazide diuretics may have a direct effect on smooth muscle and inhibit labor. Bendroflumethiazide, chlorthalidone, and hydrochlorothiazide are considered safe during breast-feeding.16,17 Loop diuretics are generally not used in pregnancy unless indicated for congestive heart failure. They have not been found to cause major adverse outcomes in the fetus. Hydralazine is safe in pregnancy. It was previously considered the drug of choice for the parenteral treatment of acute severe hypertension during pregnancy.71 However, it has been associated with higher rates of maternal hypotension compared to labetolol, which may affect perinatal outcome.16,17,37,71 It is also associated with a lupus-like syndrome, which has been reported in both mother and neonate.16,17,37,79 Because other agents, particularly labetalol, are safer and just as effective, hydralazine is no longer recommended as a first-line agent in the treatment of hypertensive emergencies in pregnant women.71 Hydralazine is safe in lactation.16,17,79 Methyldopa is considered safe in pregnancy, and most reviews have not linked it to any adverse effects on the pregnancy.16,17 Many clinicians still use it as first-line therapy to treat hypertension during pregnancy. Methyldopa is compatible with breast-feeding.31 Clonidine has been safely used throughout pregnancy, but experience during the first trimester is very limited.80 A few insignificant adverse fetal effects attributable to clonidine have been reported. Transient neonatal hypertension has been reported in neonates with in utero exposure to clonidine.16,17,37,80 Its effects on breast-feeding neonates are unknown, but it is considered to be compatible with breast-feeding.15-17 Nitroglycerin has not been shown to cause fetal harm in animal studies. Limited reports in humans do not show any major effects on the fetus or neonate. Nitroglycerin is rarely used during pregnancy, but it appears to be a safe, effective, rapidly acting and short-acting agent.16,17,81 It appears to be effective in relieving intrapartum fetal distress related to uterine hyperactivity.81 Nitroprusside use for the treatment of hypertensive emergencies in pregnancy has the same advantages and disadvantages seen in nonpregnant patients.16,17,71 Advantages include its rapid onset, rapid metabolism, and rapid excretion. Disadvantages of nitroprusside include the need for constant monitoring and cumbersome administration. During prolonged
administration of high doses, nitroprusside may result in cyanide toxicity. It readily crosses the placenta, and fetal levels of cyanide can increase as high as twice maternal levels. Standard doses do not seem to subject the fetus to major risk of toxicity, but with the availability of safer alternatives, notably labetalol, nitroprusside is considered a second-line agent.16,17,37 When used, it is recommended to monitor plasma and red blood cell cyanide and maternal pH. Nitroprusside is considered a category C medication. No data are available on its use during breast-feeding.
Asthma, Allergy, and Upper Respiratory Infection Medications Pregnant women with asthma are at risk of neonatal death, preterm birth, low-birth-weight infants, preeclampsia, and small-for-gestational-age infants.71,82 Asthmatic mothers may also have a higher rate of chorioamnionitis, hypertensive disorders of pregnancy, cesarean section, and prolonged hospital stay compared to control mothers.82,83 Better asthma control has been associated with an improved outcome.71 The beta-adrenergic medications albuterol, metaproterenol, and terbutaline are safe for use in pregnancy. None have been linked to congenital anomalies.16,17 Beta-adrenergic agents have also been used during the last trimester to treat premature labor. Adverse reactions are related to the drugs’ cardiovascular and metabolic effects, which are transient and generally well tolerated by the fetus.16,17,83,84 Transient hyperglycemia followed by insulin secretion may also occur, resulting in neonatal hypoglycemia, especially in diabetic patients.16,17 Long-term use of albuterol has not been associated with adverse effects. Albuterol is compatible with breastfeeding.15-17 Long-acting beta-agonists also appear to be safe during pregnancy.16,17,83 Ipratropium has not been found to be teratogenic in numerous animal models. Although there are few human data, ipratropium seems to be safe for use during pregnancy and lactation.15–17 Cromolyn sodium is safe in pregnancy. Cromolyn has not been associated with any significant risk of birth defects or negative perinatal outcomes.16,17,82,83 Corticosteroids are commonly used during pregnancy for the treatment of various disorders, including autoimmune diseases, hyperemesis gravidarum, and asthma. Inhaled corticosteroids are the main therapy for the prevention of asthma exacerbations during pregnancy. Oral corticosteroids are the mainstay of therapy for acute exacerbations of asthma. Although they are not considered human teratogens, there may be a slightly increased incidence of orofacial clefts when oral steroids are used during the first trimester.16,17,37,85 Furthermore, their use in the third trimester has been linked to an increased incidence of preterm delivery, low birth weight, preeclampsia, and cataracts in the newborn.16,17,37,85 Other authors have also raised concerns about the development of congenital adrenal hyperplasia in newborns.16,17 Prednisone is considered safe during breast-feeding.15-17 Data on the use of leukotriene antagonists in pregnancy are limited. One study did not find an association with congenital abnormalities, but there was a slight increase in intrauterine growth restriction. However, these results should be interpreted with caution because of the small sample size of the study.86 Zileuton is mutagenic in animal studies and should be avoided during pregnancy and lactation.16,17 Antihistamines have been safely used in the treatment of allergic reactions during pregnancy and as antiemetics in the treatment of nausea and vomiting during pregnancy. Antihis-
Gastrointestinal Medications Phenothiazines, such as promethazine, chlorpromazine, perphenazine, and metoclopramide, are dopamine antagonists commonly used in the treatment of nausea and vomiting during pregnancy, and they have not been linked to congenital abnormalities. Caution should be used with chlorpromazine because it may cause hypotension and also with promethazine at term because it may cause respiratory depression.16,17,89 Ondansetron, a serotonin 5-HT3 receptor antagonist, has not been linked to any fetal malformations, but it may not offer any additional antiemetic activity compared to the phenothiazines.16,17,89 Newer 5-HT antagonists, such as dolasetron and granisetron, also appear to be safe during pregnancy, although experience is limited.16,17,89 These agents are most likely compatible with breast-feeding.15-17 The H2 receptor antagonists ranitidine, famotidine, and cimetidine have not been linked to any congenital malformations and appear to be safe for long-term use during pregnancy and lactation.15-17 However, one report has linked the use of antacids during pregnancy to an increased incidence of asthma during childhood.90
Diabetes Medications Insulin has been used safely during pregnancy and lactation for many years and is the drug of choice for glucose control in pregnancy. Sulfonylurea drugs traditionally have not been used during pregnancy. They are regarded as possibly teratogenic and less effective than insulin in the control of gestational diabetes.16,17 Sulfonylurea drugs have also been associated with neonatal hypoglycemia when used at term.16,17 In reality, there is little information about their use during pregnancy, and in a randomized study, glyburide proved to be as effective and safe during pregnancy as insulin.16,17,91 Glyburide and glipizide are highly protein bound and are not likely to pass into breast milk; nursing infants should be monitored.16,17 Metformin has not been associated with fetal malformations in animals, and there are no controlled studies analyzing its effect in humans.16,17 Metformin has been associated with serious adverse effects in adults, including severe lifethreatening metabolic acidosis and hepatotoxicity. Because of its potential for serious effects in adults, metformin is not recommended for use in lactating mothers.16,17,92
2325
Anesthetics and Sedatives The short-term use of benzodiazepines during pregnancy appears to be safe. However, the safety of benzodiazepines during pregnancy has been debated because data on the fetal effects of these drugs have been inconsistent. Some case reports have linked their use during the first trimester of pregnancy to increased risk of oral clefts, but in a meta-analysis of pooled data from the cohort studies, no association was found between fetal exposure to benzodiazepines and the risk of oral clefts.93 Different benzodiazepines also have been linked to different effects and risks. Lorazepam, for example, has been linked to anal atresia; clonazepam has been associated with congenital cardiac abnormalities; and oxazepam and diazepam have been linked to specific dysmorphic features, CNS abnormalities, and growth defects.16,17 Midazolam, on the other hand, has not been linked to any developmental abnormalities. Neonates exposed to benzodiazepines may exhibit signs of toxicity, including apnea, cyanosis, unresponsiveness, hypotonia, poor feeding, and withdrawal symptoms characterized by irritability and tremulousness.16,17 Because of the reported risk of apnea, it is recommended that neonates exposed to benzodiazepines through breast-feeding be monitored closely.15-17 Ketamine is a rapidly acting dissociative anesthetic that is commonly used in pediatric procedural sedation and may be used in rapid sequence intubation (RSI). It has not been associated with any developmental malformations.16,17,94 Ketamine has a dose-related oxytocic effect, and in high doses it has been associated with uterine tetany, increases in maternal blood pressure and heart rate, and increased neonatal muscle tone. Neonatal depression has also been reported.16,17,94 Ketamine may remain in breast milk for 12 hours.16,17 Propofol is a rapidly acting sedative anesthetic that rapidly crosses the placenta. It has not been linked to any congenital defects when used in pregnancy.16,17,94 When high doses are used at term, it can cause neonatal respiratory and CNS depression. Propofol is excreted in breast milk in negligible amounts.16,17 Thiopental is an ultra-short-acting barbiturate that may be used during RSI or for persistent status epilepticus. It has not been linked to any congenital defects when used during pregnancy, but a slight reduction in birth weight has been noted when high doses are used.16,17,94 Etomidate is an ultra-short-acting hypnotic agent that is commonly used for procedural sedation or RSI. No reports on developmental effects of etomidate have been published. However, newborns of mothers undergoing cesarean section with etomidate were found to have significant reductions in serum cortisol concentrations 1 hour after delivery.16,17,95 The significance of this effect remains to be elucidated. No data on breast-feeding were found.
Paralyzing Agents Succinylcholine is a depolarizing neuromuscular blocking agent used in RSI for its rapid onset of action and short duration of paralysis. It has not been associated with any congenital defects, although there is limited experience with its use in early pregnancy in humans.16,17,94 In addition, it does not appear to have any effects on the newborn, except in rare cases of neonates with pseudocholinesterase deficiency.96 As occurs in adults with the same condition, newborns with cholinesterase deficiency exhibit prolonged respiratory depression and paralysis.96 Succinylcholine in lactation has not been
Chapter 178 / Drug Therapy in Pregnancy
tamines have been linked to the development of retrolental fibroplasia (retinopathy of prematurity) in premature infants when given during the last 2 weeks of pregnancy.16,17 A metaanalysis that reviewed 24 studies involving more than 200,000 patients confirmed the safety of antihistamines, including chlorpheniramine, diphenhydramine, doxylamine, hydroxyzine, and meclizine, during pregnancy.87 The newer generation antihistamines, such as cetirizine and loratadine, also appear safe during pregnancy.88 They may be acceptable alternatives for severe allergies if the first-generation antihistamines are not tolerated.16,17 First-generation antihistamines are not recommended during breast-feeding because they may inhibit lactation. In addition, neonates receiving antihistamines appear to develop serious adverse CNS effects, including seizures, especially when premature.15-17 Decongestants are not recommended during pregnancy.16,17 Decongestants with strong vasoconstrictive properties, such as phenylpropanolamine and pseudoephedrine, cause placental vasoconstriction, resulting in an increased incidence of abnormalities typically associated with placental vascular disruption, such as gastroschisis and intestinal atresia.16,17,83
PART V ■ Special Populations / Section Two • The Pregnant Patient
2326
studied; however, it is probably safe because it is hydrolyzed quickly.16,17,96 Rocuronium and vecuronium are nondepolarizing neuromuscular blocking agents used in RSI. The effects of neuromuscular blocking agents on organogenesis are not known, but these agents are not thought to pose a significant teratogenic risk.16,17,94 Because of their chemical properties, very little of either drug crosses the placenta, and very little is excreted in milk.16,17,94 Their effects on lactation are unknown but probably would be minimal.16,17
Antidotes N-acetylcysteine has been used successfully and without untoward effects in pregnant women who have overdosed on acetaminophen.18,19 No teratogenic effects have been reported. Deferoxamine is indicated for iron toxicity occurring from iron overdose or from multiple transfusions in thalassemia patients. It has been associated with developmental effects on ossification in some animal species.16,17 Experience in humans is limited, but it does not appear to affect the fetus.99 The effects of deferoxamine on the nursing infant are not known. Dimercaprol or British antilewisite is a metal chelating agent that is used as an antidote for acute mercury, lead, arsenic, and gold poisoning. It has also been used in Wilson’s disease. It is teratogenic in mice and has been associated with increased mortality, growth retardation, cleft facial features, cerebral herniation, and abnormal digits, but experience in humans is limited.16,17,97 In certain cases of heavy metal poisoning, the maternal benefits of dimercaprol use outweigh its potential risks to the unborn fetus. Breast-feeding is not recommended for patients poisoned by heavy metals. Flumazenil is a benzodiazepine antagonist. No teratogenic effects have been reported in animals, and data on humans are very limited.16,17 Its use in pregnancy depends on the potential maternal benefit compared to possible risks to the fetus. Fomepazole is a competitive inhibitor of alcohol dehydrogenase indicated in cases of methanol and ethylene glycol poisoning. Its use during pregnancy has not been studied in animals or humans. Its safety during pregnancy is not known.16,17 In cases of toxic alcohol poisoning, the benefits of treatment of the mother outweigh the possible risks to the fetus. Use of ethyl alcohol in these situations may also be considered. Activated charcoal is not absorbed and is probably safe for use in pregnancy and lactation, although there are no published studies regarding the effects of charcoal use during pregnancy.16,17 Digoxin fragment (DIG Fab) therapy is indicated for lifethreatening digoxin overdose and is being studied for treatment of preeclampsia. There are very few case reports of the use of DIG Fab immune globulin during pregnancy. A conclusion on the effects of DIG Fab cannot be made based on these reports. However, in cases of life-threatening digitalis overdose with arrhythmias, the benefits of treatment of the mother may outweigh the risk to the fetus. DIG Fab is not likely to be excreted in large amounts in milk, and it is probably safe for use during lactation.16,17 Hydroxycobalamin is a vitamin that is indicated in the treatment of cyanide toxicity. Studies in animals do not reveal an association with any developmental abnormality.16,17 Methylene blue is used in the treatment of methemoglobinemia. In the past, it was injected into the amniotic sac to identify twins and to detect rupture of the membranes, but
these practices were associated with hemolytic disease in the newborn, hyperbilirubinemia, and deep blue staining of the newborn.16,17,98 Methylene blue in pregnancy has also been associated with an increased incidence of intestinal obstruction and atresia in the newborn.99 The effects on nursing infants are unknown but probably minimal.16,17 Naloxone, used to reverse the effects of opiates in an overdose, readily crosses the placenta. Naloxone has not been associated with reproductive abnormalities; however, its use in opiate-addicted mothers may precipitate withdrawal in both mother and term fetus.16,17 It is compatible with breast-feeding.15 Physostigmine is an anticholinesterase agent indicated in cases of severe anticholinergic poisoning associated with delirium. Experience with the medication during pregnancy is limited, and its effects on the developing fetus are unknown.16,17 Use of physostigmine at term was associated with only mild decreases of Apgar scores at 1 and 5 minutes.100 Polyethylene glycol (PEG) is not absorbed systemically. It is probably safe for use in pregnancy and lactation, although there are no reports on the effects of PEG in pregnancy.16,17 Pralidoxime is indicated for organophosphate/cholinergic poisoning because it is able to reactivate cholinesterase. Experience with pralidoxime in pregnancy is limited, and its effects on fetal development are not known.16,17 In cases of organophosphate poisoning, the benefits to the mother generally outweigh the possible risk to the fetus. Pyridoxine is a vitamin required for good maternal health and good fetal development. It is indicated in isonizid poisoning and in gyromytrin mushroom poisoning. Its use has been advocated in some for nausea and vomiting of pregnancy and gestational hypertension and diabetes. It has not been associated with any adverse developmental effects, and it is safe in lactation.16,17 Succimer is a lead chelator that is indicated in lead poisoning. It has been linked to congenital defects in animal models, possibly due to effects on zinc and copper metabolism.16,17 Experience with the use of succimer in pregnancy is limited to case reports of women poisoned with lead. No conclusions can be drawn on its teratogenic effects.101
KEY CONCEPTS ■
Chemically induced birth defects are believed to be responsible for approximately 1 to 3% of anomalous births. ■ The age of the fetus is crucial in determining the impact of any given exposure; during the time of organogenesis (days 21–56 of fetal life) when major body organs are formed, exposure to a teratogen may result in major anatomic defects. ■ Certain medications, such as anticonvulsants, warfarin derivatives, NSAIDs, sulfonamides, fluoroquinolones, ACE inhibitors, and oral hypoglycemic agents, are known teratogens or cause potential toxic effects in the newborn and should be avoided, if possible, during pregnancy.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
Chapter 179
Labor and Delivery and Their Complications
Shoma Desai, Sean O. Henderson, and William K. Mallon
■ PERSPECTIVE Emergency department (ED) births are rare. In most cases, the emergency physician will identify patients in labor and triage them to the obstetric service for urgent management, maintaining a continuum of care with their primary providers. Because some births are precipitous and obstetric resources may not be immediately available, the emergency physician must possess the basic skills for intrapartum management of both normal and abnormal deliveries. In addition, a general knowledge of postpartum care is required in case of the occasional out-of-hospital delivery.1-3
Limitations of the Emergency Department The ED is a suboptimal location for the management of a complicated delivery. The obstetric suite has experienced personnel and better resources, including tocodynamometry, fetal scalp electrodes, intrauterine pressure monitors, vacuum extractors, and forceps. Also, the obstetrician usually has prenatal care information for each patient that helps to optimize maternal and fetal outcomes. Such information includes accurate gestational dates, the presence of multiple gestations, estimated fetal weights, detailed maternal pelvimetry, placental anatomy, amniocentesis results, maternal blood type and Rh factor, and prior documented obstetric complications. This information allows the obstetrician to anticipate complications of labor and delivery. It is difficult, if not impossible, to obtain these data in the ED while preparing for imminent delivery.4 Finally, in certain situations, cesarean section is the best way to ensure successful delivery. This surgical option is not available in the ED except in dire perimortem circumstances.
Epidemiology of Emergency Delivery In 2004, the perinatal mortality rate in the United States was 6.2 per 1000 live births and fetal deaths (>20 weeks of gestation).5 Extensive epidemiologic data regarding the subgroup of ED deliveries are lacking. However, it has been established that delivery complications and mortalities do occur with greater frequency in the ED. In fact, the perinatal mortality rate for ED deliveries has been approximated at 8 to 10%.6 There are multiple causes of the “high-risk” ED delivery profile. The ED is often selected by an obstetric population with unexpected complications. Antepartum hemorrhage, premature rupture of membranes (PROM), eclampsia, premature
labor, abruptio placentae, precipitous delivery, malpresentation, and umbilical cord emergencies are overrepresented in the ED population.6 Psychosocial factors further skew the epidemiology of ED deliveries. Women who present with precipitous deliveries often have had very little or no prenatal care.7 Pregnant women who have drug or alcohol problems or are victims of domestic violence represent a disproportionate number of patients who deliver in the ED. Women who are unaware or in denial of their pregnancies, or immigrants without access to other medical care, also present to the ED when labor begins.8-10 For each of these groups, all of which are compromised by psychosocial factors, inadequate prenatal care makes the subsequent delivery high risk.
Patient Transfer Considerations Because of the high risk associated with ED delivery, patients should be transported to a facility that has obstetric and neonatal resources as appropriate. The management of a premature infant may require highly specialized intensive care that is unavailable at many community hospitals. The desire to transfer a woman with an impending high-risk delivery to such a facility must be tempered, however, by clinical and medicolegal judgment.11,12
Medicolegal Considerations Transfer, with resultant en route delivery, can be disastrous for the mother and fetus. Such a transfer also violates federal law. The Consolidated Omnibus Reconciliation Budget Act (COBRA) of 1989 was based on an “inappropriate” obstetric transfer.13 Federal law has identified labor as a condition unsuitable for transfer because of its unstable nature. Although the intent of this legislation is to protect women from medical and financial “dumping,” COBRA also may force emergency physicians to perform difficult high-risk deliveries that might have better outcomes if transferred.14
Nursery Requirements For many ED deliveries, labor will have progressed to a point where tocolysis is contraindicated and delivery is imminent. Generally, this is the point when the mother feels the urge to push or the head is crowning. Whenever possible, a neonatologist or pediatrician should attend high-risk preterm (30 weeks’ gestation and >1200 g who require special attention short of circulatory or ventilator support and major surgical procedures. A high percentage of the patients in such nurseries have had obstetric complications (e.g., birth trauma, fetal distress, and obstetric anesthesia). Level 3 Nurseries These nurseries are staffed and equipped to care for all newborn infants who are critically ill, regardless of the level of support required. They are regional institutions serving as referral centers for other nurseries and for this reason are often linked with transport services. Perinatal Centers A perinatal center provides services to high-risk mothers and to infants requiring level 3 nursery care. Ample data show a higher neonatal survival rate for high-risk pregnancies cared for in such centers.
resuscitation and high-level nursery care should be initiated (Box 179-1).15
■ NORMAL DELIVERY Although the epidemiology and high complication rate associated with ED births demand caution, most are normal deliveries. Knowledge of normal labor and delivery mechanics aids safe vaginal delivery and facilitates the identification of complications. Whenever a woman in the third trimester of pregnancy seeks treatment in the ED, the possibility that she is in labor must be considered. A wide array of nonspecific symptoms may herald the onset of labor. Abdominal pain, back pain, cramping, nausea, vomiting, urinary urgency, stress incontinence, and anxiety can all be symptoms of labor. After 24 weeks’ gestation, any medical assessment should include the mother and the fetus because fetal viability becomes established near that time. In addition, given the generally high-risk nature of this patient population and the abundance of bodily fluids that the health care provider and newborn are exposed to during delivery, serologic testing for infectious disease may be warranted. With the development of rapid bedside testing technology, human immunodeficiency virus (HIV) and hepatitis screening before delivery is indicated in a significant group of patients presenting with active labor.
Distinguishing False from True Labor Braxton Hicks contractions, or false labor, must be differentiated from true labor. During the third trimester, the uterus develops into a contractile organ. After 30 weeks of gestation, the previously small and uncoordinated contractions of the uterus become more synchronous and may be perceived by the mother. Braxton Hicks contractions do not escalate in fre-
quency or duration, in contrast to the contractions of true labor. By definition, this muscular activity is associated with minimal or no cervical dilation or effacement. Examination should also reveal intact membranes. Care not to rupture the membranes is important to avoid inducing labor prematurely. If the diagnosis remains in doubt, external electrical monitoring of uterine activity can be utilized to rule out true labor. Any discomfort associated with false labor is usually relieved with mild analgesia, ambulation, or change in activity. Unlike false labor, true labor is characterized by cyclic uterine contractions of increasing frequency, duration, and strength, culminating in delivery of the fetus and placenta. In contrast to Braxton Hicks contractions, true labor causes cervical dilation to begin, marking the first stage of labor.
Bloody Show Early in pregnancy, the cervix becomes increasingly vascular and develops edema, giving the cervix a boggy texture. The vascularity of the cervix also increases, giving rise to Chadwick’s sign (a blue-violet coloration). At the onset of labor, the cervical mucus plug is expelled, resulting in what is called a bloody show. The bleeding associated with the process is slight, and usually only a few dark red spots are noticed. The dark color is due to its venous origin, and it is admixed with the mucous components of the cervical plug. The significance of a bloody show is that it is a fairly reliable indicator of the onset of true labor, specifically stage 1. Bloody show is not a contraindication to vaginal examination for determination of cervical effacement and dilation. If bleeding continues or is of a larger volume, more serious causes should be suspected, such as placenta previa and placental abruption, which are contraindications for a vaginal examination.
Stages of Labor First Stage of Labor The first stage of labor is the cervical stage, ending with a completely dilated, fully effaced cervix. It is divided into a latent phase, with slow cervical dilation, and an active phase, with more rapid dilation. The active phase begins once the cervix is dilated 3 cm.16 In multiparous women, the active phase can progress rapidly into stage 2 of labor (delivery of the fetus). Most women who deliver in the ED arrive while in the active phase of stage 1 or in early stage 2 labor (Fig. 179-1).6 The duration of the first stage of labor averages 8 hours in nulliparous women and 5 hours in multiparous women. During this time, frequent assessment of fetal well-being is important. For low-risk pregnancies, fetal heart tones should be auscultated approximately every 15 minutes. For higher risk pregnancies, continuous external electrical monitoring may help identify fetal distress, allowing appropriate intervention.17 Abdominal examination using Leopold’s maneuvers may confirm the lie of the fetus (Fig. 179-2). After labor has begun, particularly during the active phase of stage 1, Leopold’s maneuvers are difficult to use. The firm contractions of the uterus prevent the identification of fetal “small parts.” Other modalities of assessing the lie, such as ultrasonography, may be necessary if presentation remains in question.18 Maternal examination also provides a rough guide to gestational age. At 20 weeks’ gestation, the uterine fundus reaches the umbilicus. Approximately 1 cm of fundal height is added per week of gestation until 36 weeks. At that time, the fundal height decreases as the fetus “drops” into the pelvis (Fig. 179-3). These estimates help to establish gestational age rapidly.
2329 (False labor)
stage; stage 2, fetal expulsion; stage 3, placental expulsion (20 minutes); stage 4, uterine contraction (1 hour postpartum).
Cervical dilation
Figure 179-1. Stages of labor and delivery. Stage 1, cervical
5 0 cm Begin pain
5–8 hours Stage 1 nt Late e phas
Begin dilation
B
C
D
20
1 hr
Stage 2
Stage 3
Stage 4
Active phase Complete Placenta dilation/effacement delivered
Time
A
20–50
Delivery of infant
Figure 179-2. Leopold’s maneuvers. A, The first Leopold maneuver reveals what fetal part occupies the fundus. B, The second Leopold maneuver reveals
the position of the fetal back. C, The third Leopold maneuver reveals what fetal part lies over the pelvic inlet. D, The fourth Leopold maneuver reveals the position of the cephalic prominence. (Modified from Willson JR, et al: Obstetrics and Gynecology, 9th ed. St. Louis, Mosby, 1991.)
Chapter 179 / Labor and Delivery and Their Complications
10 cm
Labor begins
PART V ■ Special Populations / Section Two • The Pregnant Patient
2330 Iliac crest
Iliac crest
36
–5 –4 –3 –2 –1 0 +1 +2 +3 +4 +5
Ischial spine
40 32 28 24
20
Ischial tuberosity
Ischial spine
Ischial cm tuberosity Perineum Figure 179-4. Fetal stations. The level of the ischial spines is considered “0” station. The silhouette of the infant’s head is shown approaching station +1. (Courtesy of Ross Laboratories, Columbus, Ohio.)
16 12
Occiput Lambdoid suture
Biparietal diameter
Figure 179-3. Height of fundus by weeks of normal gestation with a
single fetus. Dotted line indicates height after lightening. (Modified from Barkaukas V, et al: Health and Physical Assessment. St. Louis, Mosby, 1994.)
The accurate determination of the stage of labor depends on examination of the cervix. A sterile approach using sterile gloves, a sterile speculum, and povidone-iodine (Betadine) solution is indicated to prevent ascending infection, such as chorioamnionitis. On pelvic examination, the clinician should determine the following: 1. Effacement refers to the thickness of the cervix. A paper-thin cervix is 100% effaced. 2. Dilation indicates the diameter of the cervical opening in centimeters. Complete, or maximum, dilation is 10 cm. 3. Position describes the relationship of the fetal presenting part to the birth canal. The most common position of the head is occiput anterior. 4. Station indicates the relationship of the presenting fetal part to the maternal ischial spines (Fig. 179-4). 5. Presentation specifies the anatomic part of the fetus leading through the birth canal. In 95% of all labors, the presenting part is the occiput, or vertex. On digital exam, a smooth surface with 360 degrees of firm bony contours and palpable suture lines is noted. Palpation of the suture lines and the fontanels where they join allows the examiner to determine in which direction the fetus is facing. Three sutures radiate from the posterior fontanel, and four radiate from the anterior fontanel (Fig. 179-5). The lateral margins should be examined carefully for fingers or facial parts that indicate compound or brow presentations. When the clinician suspects rupture of membranes, a sterile speculum examination should be performed. This may reveal pooling of amniotic fluid. Two tests used to confirm the pres-
Posterior fontanel (triangle)
Vertex Sagittal suture Anterior fontanel (quadrangle) Coronal suture Bitemporal diameter Frontal suture Sinciput
Figure 179-5. Bony landmarks of the fetal skull. (Modified from Willson JR, et al: Obstetrics and Gynecology, 9th ed. St. Louis, Mosby, 1991.)
ence of amniotic fluid include a fernlike pattern when the fluid is allowed to dry on a microscope slide and the use of nitrazine paper, which should turn blue, indicating an alkaline amniotic fluid (pH > 6). Although vaginal blood, cervical mucus, semen, and infection can interfere with results, sensitivities of both nitrazine paper and ferning in detecting amniotic fluid are nearly 90%.19 Of note, if vaginal bleeding is evident, both digital and speculum examination of the pelvis should be deferred until ultrasound can be obtained to rule out placenta previa.
Second Stage of Labor The second stage of labor is characterized by a fully dilated cervix and accompanied by the urge to bear down and push with each uterine contraction. The fetal station is advanced to +3, with crowning of the presenting part as expulsion begins. Stage 2 uterine contractions may last 1 or 2 minutes and recur after a resting phase of less than 1 minute. The median duration of this stage is 50 minutes in nulliparous women and 20
2331
Variable and early decelerations are common. Present on more than 50% of all tracings, these heart rate changes can represent physiologic reflexes associated with head compression in the birth canal or intermittent cord compression. Variable decelerations that are persistent and repetitive usually indicate repeated episodes of umbilical cord compression. The resultant hypoxia and acidosis may cause fetal distress. Attempts to shift maternal and fetal weight off the umbilical cord by changing position are indicated. If these variable decelerations continue, the situation calls for efforts to hasten the delivery or, if obstetric backup becomes available, to perform an emergency cesarean section. Late decelerations are more serious and most often indicate uteroplacental insufficiency. The tracing contours are generally smooth, with the heart rate nadir occurring well after maximal uterine contraction. The lag, slope, and magnitude of late decelerations correlate with increasing fetal hypoxia. Late decelerations are particularly ominous in association with poor variability, nonreactivity, and baseline bradycardia. Immediate delivery to prevent further hypoxia is indicated when these findings are present. The need for newborn resuscitation should be anticipated and preparation for critical care established for these deliveries. Overall, 30% of infants with late decelerations have good outcomes. The remaining 70% have suboptimal outcomes related to either the underlying pathologic condition or hypoxia. Finally, the clinician should be aware of the significance of sinusoidal tracings. Tracings of this type have low baseline heart rates and little beat-to-beat variability. The sinusoidal tracing is an ominous finding that is often premorbid. The differential diagnosis includes erythroblastosis fetalis, placental abruption, fetal hemorrhage (trauma), and amnionitis. Ultrasonography. Ultrasonographic techniques have wide application in obstetric care. In the third trimester or during labor, ultrasonography can provide crucial information pertaining to impending delivery. When a technician and radiologist are available, the gestational age, biophysical profile, amniotic fluid index, as well as a survey of fetal and placental anatomy may be discerned (Table 179-1).25-27 The parameters of Profile: 30 Minutes of Table 179-1 Biophysical Ultrasonographic Observation ELEMENT ASSESSED
Fetal heart rate reactivity Amniotic fluid index Fetal muscle tone
Body movements Breathing motions
NORMAL SCORE = 2
2 accelerations >15 beats/min for >15 sec 1 pocket >1 cm in orthogonal planes 1 episode of active flexion-extension with full return to flexed posture 3 discrete moves 1 episode of fetal breathing of at least 60 sec duration during 30 min of observation
ABNORMAL SCORE =0
5 cm Placental abruption
IV, intravenously; SC, subcutaneously; PO, orally. *Ritodrine is currently discontinued in the United States.
Table 179-3 Drug and Labor Interactions DRUG
EFFECT ON LABOR
Barbiturates Alcohol
In anesthetic doses can stop labor Decreases oxytocin release, smooth muscle relaxant Increased prematurity, placental infarction Increased duration of labor
Cocaine Caffeine or aminophylline Narcotics Atropine, scopolamine Halothane IV nitroglycerin
Increased latent phase, slow dilation (minimal effect once in active labor) Lower uterine segment relaxation, decreased frequency of contractions Strong inhibition of labor Profound uterine relaxation
IV, intravenous.
Magnesium Sulfate. Magnesium sulfate competitively inhibits calcium uptake into smooth muscle and allows relaxation. Women treated with magnesium require monitoring. Magnesium produces respiratory and neurologic depression at elevated levels, exacerbated by renal insufficiency. Pulmonary edema and cardiac dysrhythmias have also been reported.37 These effects can be reversed rapidly by the administration of calcium-containing solutions (i.e., 1 g of 10% calcium gluconate solution). Because women with premature labor are at risk for ascending infections, early treatment with antibiotics is often indicated during magnesium therapy (Table 179-3).43 Beta-Mimetics. Beta-mimetics (ritodrine and terbutaline) cause smooth muscle relaxation by activating enzymes that bind calcium to the sarcoplasmic reticulum. This effect is mediated by beta2-receptors that increase cyclic adenosine monophosphate concentrations in the myometrium. The beta-mimetic is titrated to effect since the dosage needed to eliminate uterine activity is unpredictable and varies. Only side effects limit beta-mimetic administration. They freely cross the placenta and cause fetal tachycardia. In one meta-analysis, beta-mimetics and magnesium sulfate had similar efficacy in eliminating contractions.42 Pulmonary edema is the main adverse effect of high-dose beta-mimetics. This complication is more likely to occur in mothers with preexisting cardiac disease, multiple gestations, and maternal infection. This form of pulmonary edema is high-output failure and tends to occur when there is sus-
tained maternal tachycardia greater than 120 beats/min. Betamimetics should be gradually titrated according to uterine activity and maternal heart rate. Eventually, tachyphylaxis and receptor down-regulation decrease the effectiveness of these drugs over 24 to 48 hours.40 Beta1-related side effects can be problematic in diabetic mothers. Beta1 stimulation can lead to diabetic ketoacidosis and the usual cascade of metabolic and electrolyte abnormalities.44 Surveillance of the urine for glucose and ketones is recommended. Fetal heart stimulation can result in increased cerebral perfusion pressures. A premature infant’s central nervous system vasculature is delicate and may not tolerate these changes. Beta-mimetics are associated with an increased incidence of fetal intraventricular hemorrhage.45 NSAIDs. The prostaglandin synthetase inhibitors, specifically indomethacin and sulindac, have been shown to be as or more effective than magnesium and the beta-mimetics in multiple trials. However, in the fetus, pulmonary hypertension, patent ductus arteriosus constriction, renal insufficiency, necrotizing enterocolitis, and intraventricular hemorrhage have been reported with NSAID use. Potential maternal side effects include a prolonged bleeding time and renal insufficiency.42,46 Calcium Channel Blockers. Calcium channel blockers have also been used as tocolytics with success. Nifedipine or nicardipine may be given. Onset is more rapid than that of magnesium, and the maternal and fetal side effect profiles are good.41,42 Aggressive titratable tocolytics are best for the initial 24 to 48 hours of preterm labor. After uterine contractions have been stopped, the patient can usually be maintained on oral agents, although the benefits of maintenance tocolysis, studied to date primarily with beta-mimetics and magnesium, have yet to be shown.42 The contraindications to tocolytics are important to review before initiating these therapies (Box 179-5). Any patient receiving tocolytics should be externally monitored (electrically) for signs of fetal distress.
Premature Rupture of Membranes Clinical Features. PROM, also known as amniorrhexis, is defined as rupture of the amniotic and chorionic membranes before the onset of labor. PROM affects 3% of all gestations.47 During pregnancy, the chorionic and amniotic membranes protect the fetus from infection and provide an environment that allows fetal growth and movement. The amniotic fluid is constantly exchanged by fetal swallowing and urination and umbilical cord transfer. The fetal airway also contains a secreted fluid that allows for fetal breathing movements, promoting fetal
Testing for Premature Rupture Table 179-4 Bedside of Membranes METHOD
RESULT
Nitrazine
Amniotic fluid pH 7.1–7.3 turns nitrazine paper yellow; >7.3 is blue Amniotic fluid crystallizes Amniotic fluid, when flamed, turns white and crystallizes Vaginal secretions caramelize and turn brown
Ferning Smear combustion
2337
The benefit of this strategy has been shown with preterm labor; however, this therapy is less well documented for PROM. In PROM, treatment with steroids seems to decrease the incidence and severity of hyaline membrane disease, but it may increase the risk of maternal infectious complications. Rupture of the membranes also stimulates fetal lung maturation, making it more difficult to establish a treatment benefit in PROM compared with preterm labor. When gestational age is less than 26 weeks, the latent interval to delivery is often 1 week. Tocolytics are an obvious choice, but their use is controversial. When tocolytics are used, the goal is to temporize, allowing time for therapy to take effect. These treatment decisions should be coordinated with the receiving obstetrician. All patients with PROM should be assessed for intra-amniotic infection. Infectious complications should be diagnosed and treated before the mother develops overt clinical signs of infection. Preterm PROM is usually treated with intravenous penicillin and erythromycin. Treatment of term PROM is indicated when the patient is group B streptococcus positive or has not been tested. The signs and symptoms of chorioamnionitis are late manifestations of advanced infection and are discussed next.48
Chorioamnionitis Chorioamnionitis occurs when vaginal or cervical bacteria ascend into the uterus, instigating an inflammation of the chorion and amnion layers of the amniotic sac.49 It occurs in 1 to 10% of all pregnancies, and risk factors include prolonged labor, premature rupture of membranes, excessive vaginal examinations, and recent amniocentesis. Box 179-6 summa-
BOX 179-6 Chorioamnionitis Evaluation Fluid in Vaginal Vault Phosphatidylglycerol Cervical Cultures Escherichia coli and other gram-negative bacteria Neisseria gonorrhoeae Vaginal Cultures Chlamydia spp. Mycoplasma hominis Group B streptococci Ureaplasma urealyticum Amniocentesis Studies Gram’s stain (group B streptococci) Culture Glucose Lecithin-to-sphingomyelin ratio Maternal Signs and Symptoms Premature rupture of membranes Uterine tenderness Fever Tachycardia Malodorous vaginal discharge Leukocytosis Fetal Signs and Symptoms Decreased activity Abnormal biophysical profile (ultrasonographic examination) Fetal tachycardia Decreased variability of fetal heart rate
Chapter 179 / Labor and Delivery and Their Complications
respiratory development. This fluid is produced at 5 mL/kg/hr at term gestation and is resorbed rapidly by the pulmonary lymphatics, blood vessels, and upper airway at birth. The word premature in PROM refers to rupture before labor, not to fetal prematurity. In 10 to 15% of PROM cases, the fetus is at or near term, and PROM may result in normal labor. When PROM is associated with fetal prematurity, there is significant fetal morbidity and mortality. PROM is the inciting event in one third of all preterm deliveries. After the membranes rupture, the period from latency to the onset of labor varies. Longer latent periods are common earlier in pregnancy, and the latency shortens as gestational age increases. At term, labor is a desirable result of PROM, but with fetal immaturity labor is problematic because delivery would result in fetal complications, such as hyaline membrane disease. Diagnostic Strategies. The diagnosis of PROM usually can be established by history and physical examination. The patient usually describes a spontaneous gush of watery fluid followed by a mild persistent seepage. In most cases, the patient suggests the diagnosis and usually is correct. Urinary incontinence or excess vaginal or cervical secretions are occasionally confused with PROM. Examination of women with potential PROM should be performed under sterile conditions to prevent ascending infection. Direct digital examination of the cervix should be avoided. The incidence of infection has been shown to be proportional to the number of examinations. The identification of amniotic fluid was previously discussed. Table 179-4 summarizes the bedside testing modalities available to confirm the diagnosis of PROM. Visualization of the cervix for prolapsed cord or abnormal fetal presentation (prolapse of a small part) should be done during the uterine evaluation for effacement and dilation. Cultures for group B streptococci, Chlamydia, and gonorrhea should be obtained. Management. When the diagnosis of PROM is established, management depends on several factors: gestational age and maturity of fetus, the presence of active labor, the presence or absence of infection, the presence of placental abruption, and the degree of fetal well-being or distress.48 Obstetric consultation and admission are indicated. Gestational age may be well known by menstrual history and previous ultrasonographic scans. In the absence of such data, immediate ultrasonography provides an estimated gestational age quickly. Fetal maturity is a more complex determination. Beyond 36 weeks of gestation, fetal pulmonary maturity is likely. If the gestational age is less than 36 weeks, testing the amniotic fluid for the lecithin-to-sphingomyelin ratio or for phosphatidylglycerol can establish maturity. Fluid pooling in the posterior vaginal vault can be used for this purpose. In the immature fetus (24–31 weeks of gestation), administration of corticosteroids can accelerate pulmonary maturation.
PART V ■ Special Populations / Section Two • The Pregnant Patient
2338
rizes the findings and evaluation of chorioamnionitis. Chorioamnionitis may result in prolonged first- and second-stage labor and decreased responsiveness to oxytocin. Early, aggressive treatment, even before evidence of infection occurs, decreases neonatal morbidity and delays delivery, allowing fetal maturation.50,51
Vertical Transmission of Human Immunodeficiency Virus Emergency deliveries may involve women who are known to be HIV positive in addition to women who are infected but have never been tested. The latter group generally includes pregnant women with little or no prenatal care who are at risk for precipitous delivery. Transmission may occur in the antepartum, intrapartum, or postpartum period (breast-feeding). Because intrapartum transmission accounts for up to 75% of vertically transmitted HIV infections, antiretroviral therapy upon presentation, even while labor progresses, can decrease vertical HIV transmission.52 Potential mechanisms of transmission include microtransfusion during contractions, absorption of virus through the mucous membranes, and even invasion through the epithelium. Risk factors for transmission include high viral loads, prolonged rupture of membranes, maternal drug use, vaginal delivery, and breast-feeding (Table 179-4).53,54 In November 2002, the Food and Drug Administration approved the OraQuick Rapid HIV-1 Antibody Test (OraSure Technologies, Bethlehem, Pa).55 Point-of-care testing for HIV with a median turnaround time of 45 minutes realistically allows the clinician to initiate intrapartum and neonatal antiretroviral therapy when the test is positive. As usual, serologic confirmation is recommended, but emergent interventions can proceed based on the bedside result.56 Since 1994, it has been known that immediate treatment during labor can significantly decrease vertical transmission to the newborn.57,58 A positive HIV test in some cases may allow for a change in the method of delivery. Cesarean section decreases the rate of HIV transmission compared with vaginal delivery methods. In a 1999 meta-analysis from the United States and Europe, vertical HIV transmission was decreased by cesarean section with an odds ratio of 0.43 (95% confidence interval, 0.33–0.56). The protective effect of surgical delivery over other delivery methods persisted even when the data were stratified for receipt of antiretroviral therapy.59 This finding was tempered by a reported increase in maternal morbidity and mortality in HIV-positive women undergoing cesarean section, namely an increased incidence of endometritis, maternal sepsis, pneumonia, and transfusion. Therefore, cesarean section may be reserved for only high viral load patients.60 Ideally, decisions regarding the mode of delivery and the need for antiretroviral therapy can be made before the rupture of membranes. The risk of transmission increases after that time and continues to increase as the fetus traverses the birth canal.61 When antiretroviral therapy and elective cesarean section are employed, the likelihood of vertical transmission is reduced by 87% compared with that for women receiving neither intervention.59 This relatively new diagnostic and therapeutic burden on the emergency physician faced with an imminent delivery is challenging. Data indicate that antiretroviral treatment and the mode of delivery make labor a true emergency for an HIVpositive patient. In addition, pregnant women with advanced HIV disease have a higher incidence of premature births, postpartum endometriosis, and perinatal mortality.54 Currently, the main obstacle to this type of care is the availability of point-of-care HIV testing. Since the late 1990s, the results supporting this approach have become increasingly robust and seem to justify the efforts.62,63
Complicated Delivery Perspective Complicated deliveries, involving dystocia, malpresentation, and multiple gestations, are potentially life-threatening emergencies. The emergency physician cannot “solve” these obstetric problems with cesarean section and will therefore face the prospect of an extremely high-risk vaginal delivery. As expected, these abnormal deliveries increase the risk of fetal and maternal complications. Aggressive attempts to obtain obstetric, neonatal, and anesthesia support are warranted. If the delivery proceeds in the ED, preparations for maternal and neonatal resuscitation should be made rapidly.
Principles of Disease Knowledge of abnormal labor and its anatomy and physiology is important for the clinician facing a complicated delivery. Intrapartum management skills will enable the emergency physician to proceed with delivery in an efficient, capable manner. Dystocia and Malpresentation. Dystocia, or abnormal labor progression, accounts for one third of all cesarean sections and half of primary cesarean sections. Because rapid surgical resolution is unavailable to the emergency physician, intrapartum management skills are important. Dystocia can be divided into three etiologic categories. Labor fails to progress when there are problems related to the pelvic architecture (the passage), when there are fetal size or presentation problems (the passenger), and when uterine expulsive forces are inadequate. Although it is useful to consider these causes independently, dystocia is usually caused by a combination of factors. Presentation problems are particularly important because they become apparent during stage 2 of labor and require immediate action. In order of increasing incidence, brow, face, shoulder, and breech presentations are the most common malpresentations (Table 179-5). True fetopelvic disproportion is much less common. Cesarean section is indicated when labor arrest or cord prolapse coexists with these presentations.64 Breech Delivery Perspective. Breech is the most common malpresentation, occurring in approximately 4% of all deliveries. Three types of breech presentation exist: frank, incomplete, and complete (Fig. 179-8; Box 179-7). The main mechanical problem with breech presentations is that the buttocks and legs do not provide a sufficient wedge, hindering cervical accommodation of the relatively larger head. In addition, because the presenting part does not completely occlude the cervical opening, umbilical cord prolapse may occur. The terminology for breech deliveries is complicated. By convention, the presentation (frank, incomplete, and complete) is followed by the relationship of the fetus to the birth canal, using the fetal sacrum as a reference point. The mode of delivery used also has specific terminology. Spontaneous breech delivery refers to vaginal delivery without manual aid,
Table 179-5 Relative Incidence of Malpresentations MALPRESENTATION
INCIDENCE
Breech presentation Shoulder dystocia Face presentation Brow presentation
1/25 live births 1/300 live births 1/550 live births 1/1400 live births
2339
A, Frank breech presentation. B, Complete breech presentation. C, Incomplete breech presentation. (Modified from Cunningham FG, et al: Williams Obstetrics, 19th ed. Norwalk, Conn, Appleton & Lange, 1993.)
A
B
C
BOX 179-7 Breech Presentations Frank Breech 60–65% of all breech presentations Hips flexed, knees extended Buttocks act as good dilating wedge Incidence of cord prolapse approximately 0.5% Complete Breech Least common, occurs in approximately 5% of all breech presentations Hips and knees flexed Buttocks act as good dilating wedge Incidence of cord prolapse is 5–6% Incomplete Breech 25–35% of all breech presentations Incomplete hip flexion, single or double footling Poor wedge Increased incidence of prolapsed cord (15–18%) almost always of a small premature infant. Assisted breech delivery and partial breech extraction are terms that apply when delivery is manually assisted after the umbilicus clears the perineum. The head may be delivered by intrapartum maneuvers or forceps. Complete breech extraction consists of the application of traction to the lower extremities or groin before delivery of the buttocks. This approach is never indicated for a singleton
breech because it increases the chances of head entrapment. Confusion arises because the word complete, as used here, describes the mode of delivery, not the presentation. Slightly less than 4% of deliveries are breech. Correlated with this abnormal presentation are several factors: prematurity, multiparity, fetal abnormalities, prior breech presentation, polyhydramnios, and uterine abnormalities.65 Overall, one third of breech fetal deaths are believed to be preventable. Asphyxia is often due to umbilical cord prolapse or entrapment of the head. Fetal head and neck trauma can occur if inappropriate delivery techniques are used. Scheduled cesarean section for these patients reduces the potential for emergency room presentation. Since the 1990s, however, obstetricians have been attempting to decrease cesarean rates. As part of this process, some centers have proposed a trial of labor and vaginal delivery for selected full-term infants with frank breech presentations.66 Therefore, the potential for the emergency physician to encounter a breech presentation may be increasing. Diagnostic Strategies. Before labor, Leopold’s maneuvers facilitate the diagnosis of breech presentation. In the case of breech presentation, Leopold’s first maneuver identifies a firm, round mass (the head) in the fundus of the uterus. The third maneuver reveals the softer breech at the pelvic inlet. For the emergency physician, active labor restricts the use of Leopold’s maneuvers; thus, vaginal examination is required. The differentiation of a vertex presentation from a breech by tactile vaginal examination is not always simple. Any time
Chapter 179 / Labor and Delivery and Their Complications
Figure 179-8. Breech presentations.
PART V ■ Special Populations / Section Two • The Pregnant Patient
2340
a fontanelle is not identified on examination, a breech presentation should be suspected. During the vaginal examination, it is helpful to remember that the face and the skull have a complete circle of bone, whereas the anus is flanked by bone on only two sides. If time allows, immediate ultrasonographic studies or plain radiographs are indicated to obtain information on the position of the fetal arms and neck. If the fetus has a hyperextended neck, vaginal delivery is associated with a 70% incidence of spinal cord injuries. If possible, labor should be delayed to allow cesarean section.67 Likewise, if the arms are over the head, they increase the dystocia when the head enters the birth canal. If ultrasound reveals anencephaly or massive hydrocephaly, vaginal breech delivery should be allowed to continue because cesarean section is undesirable. Management. Premature infants in the breech position often deliver spontaneously without difficulty. As the infant comes to term, dystocia becomes increasingly common. When committed to a vaginal delivery, knowledge of breech dystocia mechanics may allow atraumatic delivery. The key goals are to maximize the size of the passage and to minimize the dystocia of the aftercoming head. Box 179-8 summarizes the actions associated with successful vaginal breech delivery. The Mauriceau maneuver is the use of the fetal oral aperture to flex the fetal neck and draw in the chin. Because fetal neck extension is associated with cord injuries and worsening dystocia, this maneuver is crucial to successful vaginal delivery. While the Mauriceau maneuver is used, the fetal pelvis should be supported to avoid abdominal injuries. Generous episiotomy may even be necessary to facilitate the Mauriceau maneuver in a full-term infant. If the aftercoming head cannot be delivered quickly, the chances of good fetal outcome are poor. For term infants, labor arrest, asphyxia, and/or brachial plexus injuries are potential complications of vaginal breech deliveries.
BOX 179-8 Vaginal Breech Delivery Actions to Do as Able Monitoring fetal heart rate Focused history Diagnosis of breech lie Cervical dilation and station determination Ultrasonography or plain radiography Evaluation for prolapsed cord if spontaneous rupture of membranes Episiotomy Knee flexion and sweep out legs Pulling out a 10- to 15-cm loop of cord (room to work) after umbilicus clears perineum Use of bony pelvis as means of holding infant Keeping face and abdomen away from symphysis, and using rotation to deliver the more accessible arm Mauriceau maneuver Actions to Avoid Inappropriate transfer with delivery en route Misdiagnosis of cervical dilation Iatrogenic rupture of membranes (cord prolapse) Moving patients or leaving unmonitored Traction on the fetus during delivery Grasping fetus by the waist, causing abdominal organ injury Arm entrapment over head Neck hyperextension
Shoulder Dystocia Perspective. Shoulder dystocia is the second most common malpresentation, occurring in 1 in 300 deliveries. In contrast to a breech presentation, which may be diagnosed antepartum, shoulder dystocia develops intrapartum. Maternal and fetal factors are associated with shoulder dystocia. Maternal factors include diabetes, obesity, and prolonged second stage of labor; fetal factors include macrosomia, postmaturity, and erythroblastosis fetalis. The combination of prenatal data, estimated fetal weight, and fetal biometry cannot reliably identify most deliveries complicated by shoulder dystocia. The fact that shoulder dystocia responds well to a variety of intrapartum maneuvers means that delivery skill is an important determinant of fetal outcome. The consequences of shoulder dystocia can be devastating. As with breech presentation, infant complications are more common and severe than maternal complications. Asphyxia, traumatic brachial plexus injuries, and humeral and clavicular fractures contribute to a complication rate of 20%.68 Maternal complications are related to traumatic delivery and include vaginal, perineal, and anal sphincter tears as well as urinary incontinence.69 Diagnostic Strategies. Shoulder dystocia is diagnosed clinically by the inability to deliver either shoulder. The fetal head may appear to retract toward the maternal perineum. This finding is known as the “turtle sign.” Traction on the head extends and abducts the shoulders, increasing the bisacromial diameter and worsening the dystocia. Figure 179-9 shows the normal and abnormal relationship of the shoulders to the birth canal and illustrates why the bisacromial diameter is an important element of fetal biometry. Normally, the shoulders negotiate the maternal pelvis in sequential fashion, anterior shoulder first. With shoulder dystocia, both shoulders attempt to clear the maternal pelvis simultaneously. In addition to the turtle sign, examination often reveals that the fetal shoulders are on a vertical axis (rather than oblique). These findings in combination with an arrested delivery confirm the diagnosis of shoulder dystocia.70 Management. When shoulder dystocia becomes evident, knowledge of intrapartum delivery maneuvers can be lifesaving. Successful vaginal delivery is most likely when a directed sequential approach to each maneuver is used. Rapid resolution of shoulder dystocia is important to avoid fetal asphyxia and resultant central nervous system injury. A head-to-body time interval over 6 to 8 minutes is considered critical in the development of these sequelae.68 Obstetric and neonatology assistance may improve the outcome, and aggressive attempts to obtain help in these areas are warranted. Initial attempts to resolve shoulder dystocia involve increasing the anteroposterior diameter of the passage. An episiotomy may be used for fetal maneuvering by allowing access to the posterior shoulder. Anteriorly, draining the bladder with a Foley catheter can generate room. The most important first step is to use the McRoberts’ maneuver (Fig. 179-10). Maternal leg flexion to a knee-chest position may disengage the anterior shoulder, allowing rapid vaginal delivery to follow. This maneuver “walks” the pubic symphysis over the anterior shoulder and flattens the sacrum, helping the fetus to pass through the birth canal one shoulder at a time. This method, although requiring very little effort, is successful in 40% of shoulder dystocia cases when used alone.71 If the McRoberts’ maneuver does not free the anterior shoulder, suprapubic pressure may accomplish the goal. Pressure may slip the anterior shoulder beneath the pubis or cause the posterior shoulder to retreat into the hollow of the sacrum. Digital pressure on the posterior shoulder (via the episiotomy) may help facilitate posterior shoulder retreat.
2341 Anterior
O
bl
iq
ue
ax is
Chapter 179 / Labor and Delivery and Their Complications
Vertical axis
A
Posterior
Figure 179-10. McRoberts’ maneuver. Top, Bisacromial diameter pinned Bisacromial diameter
Pubis
behind pubic symphysis. Bottom, Removing the maternal legs from the stirrups and putting the knees up to the chest fulcrum the pubic symphysis over the impacted anterior shoulder.
Goal position Sacrum
Initial position
B Figure 179-9. A, Normal delivery. As the fetal head rotates, the shoulders assume an oblique position and enter the pelvis one at a time. B, Shoulder dystocia. Both shoulders attempt to clear the pelvis simultaneously, forcing the bisacromial diameter into the opening.
Use of these intrapartum maneuvers resolves most cases of shoulder dystocia. However, if delivery is still impossible, more drastic interventions are warranted. Decreasing the bisacromial diameter may be possible by pushing the most accessible shoulder toward the fetal chest (Rubin’s maneuver; Fig. 179-11). Often, both shoulders assume the same attitude, decreasing the bisacromial diameter and allowing delivery. Attempts to manipulate the shoulders for Rubin’s maneuver may be transabdominal, via the introitus (anterior shoulder), or via the episiotomy (posterior shoulder). If the shoulders remain undeliverable, the next step is to use Wood’s corkscrew maneuver. In this process, the impacted shoulders are released through rotation of the fetus 180 degrees. Fetal rotation is achieved by pushing the most accessible shoulder in toward the chest.72 The fetal axilla can be snared with a digit, or a hand can be slid in along the fetal spine to sweep the hips and generate rotation. Wood’s corkscrew maneuver is difficult to perform but should be attempted before reaching for an arm. A slightly oblique anteroposterior position for the shoulders provides the largest passage through the pelvic outlet. At this juncture, if the fetus remains trapped and several attempts have failed to yield delivery, consideration of deliv-
AP axis
Figure 179-11. Rubin’s maneuver decreases the bisacromial diameter. AP, anteroposterior.
ery of an arm is appropriate. A hand is introduced along the posterior aspect of the posterior shoulder. This would be a tight fit, requiring tactile identification of fetal anatomy. The posterior arm is swept across the chest, bringing the fetal hand up to the chin. Attempts to splint the humerus may prevent fractures and brachial plexus injuries. The fetal hand is grasped and pulled out of the birth canal across the face, delivering the posterior shoulder. The mnemonic HELPER (Box 179-9) has been proposed to help keep these steps organized and to
2342
PART V ■ Special Populations / Section Two • The Pregnant Patient
BOX 179-9 HELPER Mnemonic for Shoulder Dystocia Help Obstetrics, neonatology, anesthesia Episiotomy Generous, possibly even episioproctotomy Legs flexed McRoberts’ maneuver Pressure Suprapubic pressure, shoulder pressure Enter vagina Rubin’s maneuver or Wood’s maneuver Remove posterior arm Splint, sweep, grasp, and pull to extension
facilitate a sequential approach. These steps successfully deliver almost all cases of shoulder dystocia. Face, Brow, and Compound Presentations. Face and brow presentations yield a larger engaging aspect of the fetal head and predispose to labor arrest. Although these abnormal presentations can be diagnosed with ultrasonography or Leopold’s maneuvers, most are discovered during labor via vaginal examination. Approximately half are discovered during the second stage of labor. The engaging diameter of the head in vertex position is approximately 0.8 cm less than a face presentation and 1.5 cm less than a brow presentation. Face presentations are described using the chin as a reference point (e.g., mentum anterior). Face presentation is managed expectantly. The obstetric adage “if a face presentation is progressing, leave it alone” arises from the fact that mentum anterior presentations usually deliver vaginally and mentum transverse presentations frequently rotate to become mentum anterior. Brow presentations, occurring when the fetal head is partially flexed, also spontaneously convert to vertex or face presentations in more than 50% of cases.65 A persistent mentum posterior face and brow presentation cannot be delivered vaginally if the fetus is full term. The resultant labor arrest requires symphysiotomy or cesarean section. Prolongation of the second stage is the most common outcome of both of these malpresentations at term. For the emergency physician, this prolonged second stage may provide a window during which obstetric help may arrive. Compound presentations are those in which an extremity enters the birth canal with the head or breech. Small and premature fetuses generally proceed to vaginal delivery without incident. Labor arrest and umbilical cord prolapse are accepted indications for cesarean section in the setting of face, brow, and compound presentations. Manipulation of a compound presentation, including attempts to reduce the hand or arm, increases the rate of cord prolapse. Therefore, manipulation attempts are contraindicated. Cord prolapse rates are 10 to 20% even without manipulation. Close monitoring and careful examination are indicated. Multiple Gestations Perspective. Although twin gestations accounted for less than 1% of all deliveries historically, there has been a recent increase in the frequency of twin and triplet or higher births (65 and 500%, respectively) due to the increasing use of infertility treatments.73 In 2003, twin deliveries accounted for 31.5 per 1000 live births. Because multiple gestation deliveries have a higher incidence of preterm labor and low birth weights, both maternal and fetal complication rates are correspondingly increased.74
Diagnostic Strategies. Most women with multiple gestations have the situation identified well before the third trimester. In patients who have had little or no prenatal care, bedside ultrasonography allows rapid diagnosis and early preparation. The stages of labor for twins and other multiple gestations are similar to the stages for a singleton. Nulliparous women experience longer labors than multiparous women, with an overall labor duration that is similar to that of singleton pregnancies. Of importance to the emergency physician is a relatively short latent phase of labor, with rapid progression to the active phase. The active phase is usually longer, however. The prolongation of the active phase is due to overdistention of the uterus plus malpresentation and may allow time for obstetric assistance to arrive. Vertex twin A and vertex twin B occur in approximately 40% of deliveries. In the remaining 60%, one or both of the twins has a malpresentation, usually twin B.75 Delivery problems with twin B cause most of the preventable perinatal mortality of twin gestation.76 Management. The presentation of the twins is an important determinant for the safety of vaginal delivery. Twins who are vertex-vertex, the most common presentation, can be delivered vaginally. If twin B is nonvertex, some obstetricians recommend cesarean section to prevent delivery-related complications for twin B. External cephalic version may also be attempted to convert twin B to a vertex lie and then proceed to vaginal delivery. Breech extraction is a third, more difficult, option. The general consensus is that if twin A is nonvertex, cesarean section is the preferred route. In such cases, efforts should be made to delay delivery until an operative approach can be used.77,78 Proceeding vaginally can result in the “stuck twin” syndrome and lead to poor outcome. The interval between the delivery of twin A and twin B is variable. Although in most cases twin B delivers in minutes, prolonged interdelivery times with good fetal outcomes have been reported. When twin B does not follow rapidly, in utero assessment is important to document fetal well-being. If fetal heart tracings are reassuring, the delivery of twin B (especially nonvertex) should not be hastened. Repeat ultrasonographic evaluation may be utilized to confirm twin B’s presentation and well-being during the interdelivery period. After every emergency department delivery, particularly deliveries that are precipitous or that occur in the out-of-hospital setting, the mother should be examined for the possibility of twins. Ongoing labor may be confused with postpartum cramping, only to have twin B and all of the potential complications surprise the clinician. This situation is particularly relevant for women with inadequate prenatal care and low-birth-weight infants.
Umbilical Cord–Related Emergencies Perspective Umbilical cord–related complications can occur in normal and abnormal deliveries. Immediate intervention is required to prevent fetal morbidity and mortality. The spectrum of cordrelated emergencies includes prolapsed cord, nuchal loops of the umbilical cord, body coils, cord knots, and entangled cords in monoamniotic twins. Cord length is believed to be proportional to fetal activity in utero during the first and second trimesters. Excess cord length increases the potential for umbilical cord complications of all types. Because the umbilical cord supplies the fetus with all of its oxygen, interruption of cord circulation before establishment of fetal respiration is a life-threatening emergency. Fetal asphyxia caused by cord
2343
PRESENTATION
INCIDENCE (%)
Vertex Breech Frank breech Complete breech Incomplete breech Shoulder Compound Face or brow
0.14 2.5–3.0 0.4 5 10 5–10 10–20 Rare
circulation compromise is potentially preventable with appropriate delivery interventions.
Umbilical Cord Prolapse Clinical Features. Umbilical cord prolapse occurs when (1) the umbilical cord precedes the fetal presenting part or (2) the presenting part does not fill the birth canal completely. Most instances of cord prolapse are unexpected and develop during the second stage of labor. Cord prolapse has a variable rate of association with different fetal presentations. Compound, shoulder, and breech presentations all yield gaps and a relatively poor dilating wedge. Table 179-6 summarizes the rates of umbilical cord prolapse with various fetal presentations. Overall, malpresentations account for 50% of cord prolapse cases; therefore, prolapse may be the first indication of a malpresentation.79 The incidence of cord prolapse is reported to range from 0.3 to 0.6% of all deliveries and associated perinatal mortality ranges from 8.6 to 49%.80,81 Diagnostic Strategies. Umbilical cord prolapse may be overt or occult, requiring a pelvic examination to reveal the umbilical cord lying beside the presenting part. The diagnosis also may be made with Doppler ultrasonography. In most cases, the diagnosis is obvious, and the cord is encountered at the perineum or introitus. Management. Whenever a prolapsed cord occurs with a viable infant, cesarean section is the delivery method of choice. If surgical delivery is available, maneuvers to preserve umbilical circulation should be instituted immediately. The mother should be placed in the knee-chest position with the bed in Trendelenburg as the presenting part is digitally elevated off the umbilical cord.82 It is crucial that the mother be instructed not to push, to avoid further compression of the cord. Placement of a Foley catheter and instillation of 500 to 750 mL of saline into the bladder may help lift the fetus off the cord, particularly in the first stage of labor. Preparation for an emergency (“crash”) cesarean section should be underway. The time from prolapse to surgical intervention is an important factor in fetal outcome. Perinatal mortality rates of approximately 15% are reported for prolapsed cord deliveries; however, if cesarean section can be done within 10 minutes, the mortality rate decreases to 5%.83 These numbers have implications for the availability of obstetric and surgical backup for the ED. If surgical delivery cannot be done in a timely fashion, funic reduction (manual replacement of the cord into the uterus) and rapid vaginal delivery may be the only options available. The same maneuvers to decrease cord compression are used, and the umbilical cord is pushed gently, in a retrograde fashion, above the presenting part. Manipulation and cord trauma
Cord Entanglement The umbilical cord can also become entangled with itself, spontaneously knotting. Umbilical cord knots are related to intrauterine movements early in pregnancy. Approximately 4 or 5% of stillbirths are found to have knots that are believed to have caused fetal demise. Despite this association, cord knots can persist without problems as long as perfusion is maintained. Loose umbilical cord knots pulled tight at delivery may cause fetal distress. As with cord prolapse, this situation must be resolved quickly to prevent fetal asphyxia. Rapid delivery with avoidance of further cord traction optimizes fetal outcome. No specific interventions exist to deal with this problem. Long umbilical cords are associated with true knots as well as entanglements and prolapse. Because the fetal limbs are short and flexed in most presentations, they are rarely involved. Loops around the neck and body do occur, however. Umbilical cord loops can be single or multiple, and there are reports of six nuchal loops. Risk factors for excessive cord length include increasing parity and fetal weight. Although generally benign, they may result in fetal complications, such as nonreassuring fetal status and respiratory distress.85 During delivery, loose nuchal cords should be reduced at the perineum. Loose body coils usually disentangle spontaneously. The reduction process may be aided by slipping them over the extremities or forward over the head. Occasionally, loops are tight enough to impede delivery and cannot be reduced. The solution is to cut the clamped cord and deliver the infant rapidly. The high frequency of nuchal loops (one in five births) means that the emergency physician should expect to encounter this problem.
Maternal Complications of Labor and Delivery Perspective Maternal complications of labor and delivery include postpartum hemorrhage, uterine inversion and rupture, amniotic fluid embolism, infections, and more. Many of these problems can be managed nonsurgically. When severe, however, these complications threaten the reproductive future and the life of the mother, and they may require emergent surgical intervention.
Postpartum Hemorrhage Clinical Features. Postpartum hemorrhage is the most common complication of labor and delivery. Defined as hemorrhage greater than 500 mL, it affects 5 to 10% of all deliveries and accounts for up to 25% of obstetric deaths.86,87 Postpartum hemorrhage is divided into primary and secondary categories. The primary category includes blood loss that occurs within the first 24 hours, and the secondary category is hemorrhage 24 hours to 6 weeks after delivery. The clinical picture is as expected with any type of hemorrhage, although due to maternal adaptations during pregnancy, the patient may not show signs of shock until more than 1500 mL of volume is lost.88
Chapter 179 / Labor and Delivery and Their Complications
Associated with Umbilical Table 179-6 Conditions Cord Prolapse
should be kept to a minimum because the resultant vasospasm can cause fetal hypoxia. After funic reduction, the development of umbilical cord body coils or nuchal loops is common and should be anticipated.84 Subsequent to delivery, the physician should be prepared for resuscitation of a distressed newborn. If one is available, a neonatologist should attend the delivery.
PART V ■ Special Populations / Section Two • The Pregnant Patient
2344
Differential Considerations. The differential diagnosis of primary postpartum hemorrhage includes uterine atony, genital tract trauma, retained placental tissue, and coagulopathies, or the “four Ts”: tone, trauma, tissue, and thrombin. Uterine Atony. The most common cause of serious immediate postpartum hemorrhage is laxity of the uterus after delivery. It accounts for 75 to 90% of postpartum hemorrhage cases.89 Postpartum bleeding from the placental implantation site normally is limited by contraction of the myometrium, constricting the spiral arteries. If the uterus cannot or does not contract, ongoing hemorrhage will occur. Predisposing factors include overdistention of the uterus (multiple gestations and polyhydramnios), prolonged labor, chorioamnionitis, use of tocolytics, and general anesthesia with halogenated compounds. Despite the myriad causes, uterine atony is a diagnosis of exclusion. Physical examination to rule out obstetric trauma and retained products of conception must be done before reaching the diagnosis. On examination, the uterus is palpable as a soft, boggy mass. After other causes have been excluded, therapy to augment myometrial contraction should be instituted to prevent further hemorrhage. Two-handed uterine massage may stimulate uterine contraction. One hand exerts pressure transabdominally while the other supports the uterus via the introitus. Uterotonics in conjunction with massage usually provide enough stimuli to control bleeding. Blood should be typed, crossmatched, and available for resuscitation should these measures fail. Maternal Birth Trauma. Maternal birth trauma is the second most common cause of postpartum hemorrhage, accounting for 20% of the cases. Uncontrolled delivery, macrosomia, and malpresentation all may result in maternal birth-related trauma. Although genitourinary structures are most commonly involved, any part of the birth canal–associated anatomy may be injured, with resultant postpartum hemorrhage. Tears and lacerations may involve the perineum, rectum, cervix, vagina, vulva, and urethra. Blood vessels beneath the vulvar or vaginal epithelium can also be injured without frank hemorrhage, resulting in the formation of large, contained hematomas. These hematomas may go unrecognized for hours, gradually enlarging and eventually resulting in hemorrhagic shock. This type of hemorrhage should be suspected whenever there is evidence of ongoing blood loss and no identifiable obstetric site of bleeding (a firm uterus and negative examination for lacerations). Delayed postpartum hemorrhage at these sites can also occur and is often a diagnostic challenge. Tears. Vaginal tears, especially posteriorly over the ischial spines, should be diagnosed early by postpartum exploration rather than hours later when the mother is hemorrhaging. The vulva, including periurethral structures, should be examined. Although minor tears may be left, any laceration involving subcutaneous tissue should be repaired with absorbable suture. The most common sites of tears are the vaginal, perineal, and rectal structures. Tears are classified as first, second, third, and fourth degree. First-degree tears involve the perineal skin and vaginal mucous membranes without involvement of the underlying fascia and muscle. Second-degree lacerations extend through the skin into the fascia and muscles of the perineal body but not into the rectal sphincter. Third-degree tears involve the skin, mucous membranes, perineal body, and anal sphincter. Fourth-degree tear involvement extends through all layers, including the rectal mucosa. These lacerations are also associated with tears in the region of the urethra. The repair of these tears and the repair of episiotomies are virtually identical. Commonly, repair of the episiotomy is
delayed until the placenta is delivered, which allows an uninterrupted approach to potentially complicated repairs. The goal of these repairs is to restore anatomy and provide hemostasis with a minimum of suturing. Third-degree and fourthdegree tears should be repaired by the obstetrician in an operating room. Retained Products of Conception. Approximately 10% of postpartum hemorrhage cases are due to retained placental tissue. Normally, the plane of cleavage between the zona basalis and the zona spongiosa results in clean separation of the placenta from the uterus. When this occurs, the placental tissue delivers as a single unit, without evidence of fragmentation. Occasionally, accessory placental tissue exists as succenturiate placenta, but this also should cleave normally and deliver spontaneously. Any placental defect or evidence of accessory placental tissue may signify a retained cotyledon. Retained fragments prevent myometrial constriction and result in hemorrhage. Inappropriate traction on the placenta during stage 3 of labor can result in tears with retained products of conception, which may cause immediate and delayed postpartum hemorrhage. Ultrasonography may be utilized in the diagnosis of retained placenta, with an empty or fluid-filled uterus providing a high negative predictive value and an expanded endometrium or solid echogenic masses within the uterus providing evidence of retention.28,90 Treatment requires removal of the remnant placental tissue. Digital uterine exploration with blunt dissection of the fragments from the myometrium allows myometrial contraction. Normal placental tissue cleaves away easily, allowing removal. Abnormally adherent tissue is not freed by this mechanism. Placenta Accreta, Increta, and Percreta. Placenta accreta, increta, and percreta describe various degrees of abnormal placental attachment to the uterus. Placental villi may invade the myometrium at the site of implantation, firmly rooting the placenta and obliterating the normal cleavage plane. Thus, abnormal attachment results in retained products of conception and postpartum hemorrhage. When the placenta adheres to the myometrium without the intervening decidua basalis, it is termed placenta accreta. In placenta increta, the villi extend into the myometrium. In placenta percreta, the placenta penetrates the full thickness of the myometrium.91 The incidence of these placental disorders is 1 in 2000 to 7000 deliveries. Placenta accreta occurs in 80% of these cases. Associated risk factors include multigravidity, prior cesarean sections, placenta previa, previous curettage, and uterine infections. Management Uterine Exploration and Placental Removal. In the face of ongoing hemorrhage and retained products of conception, attempts to remove the placenta manually are indicated. The procedure entails risk of infection, perforation, and increased hemorrhage but may be the most expeditious way to control bleeding.92 Before beginning, the patient should be on a monitor, good vascular access should be established, and blood products should be available. The umbilical cord is traced through the cervical os to the placenta, allowing the identification of a placental margin. The placental membranes are digitally perforated, and the placenta is gradually divided from the myometrium. The palmar surface of the hand should be directed toward the placenta, taking care to avoid uterine perforation. After placental removal, the uterus should be explored for retained cotyledons. Removal of any further fragments still present requires curettage of the uterine cavity by an obstetrician. Placenta accreta, percreta, and increta may be diagnosed in this way because these are not digitally dissectible.
2345
signs of bleeding are associated with hypofibrinogenemia, thrombocytopenia, and elevated levels of fibrin split products and D-dimer.97 Appropriate management entails hemodynamic support as well as correction of coagulopathies. In fact, recent investigations have reported the successful use of recombinant factor VIIA for severe cases of postpartum hemorrhage.98
Uterine Inversion Uterine inversion is an uncommon but serious complication of delivery that occurs during stage 4 of labor. The resultant postpartum hemorrhage can be severe and life-threatening, accounting for a maternal mortality rate of up to 15%. Uterine inversion is relatively rare, complicating 1 in 2000 deliveries.99 It is classified by duration as well as degree of inversion. Risk factors include forceful traction on the umbilical cord (especially in conjunction with a fundal placenta), placenta accreta, maternal congenital abnormalities of the uterus, fundal pressure during delivery, use of magnesium sulfate in the antepartum period, and primiparity.99,100 Clinical Features. Clinically, the patient notes the sudden onset of severe abdominal pain. Abdominal examination reveals tenderness and an absence of the uterine corpus, which is potentially visualized at the cervical os or bulging from the introitus. Profuse bleeding leading to hemodynamic instability can also occur. Ultrasound may assist in making the diagnosis. Once uterine inversion is identified, the appropriate mobilization of resources should begin simultaneously with efforts to reestablish the correct anatomic position of the uterus. Management. The highest likelihood for successful repositioning of the inverted uterus is immediately after inversion occurs. If the placenta is still adherent, it should not be removed until after repositioning. Removing the placenta while the uterus is inverted is associated with excessive blood loss. The initial attempt to reposition the uterus should be to push the fundus upward via the introitus. Digital pressure should be directed toward the mother’s umbilicus along the long axis of the uterus. Contraction of the cervical uterine segments can create a muscular ring, preventing repositioning. Therefore, all uterotonic agents should be withheld immediately upon diagnosis of uterine inversion. If initial attempts fail and a cervical ring develops, pharmacologic attempts to relax the uterus are indicated. Sedation and tocolytics can be used to facilitate uterine replacement. Terbutaline and magnesium sulfate have been used successfully to relax cervical rings. When the uterus has been repositioned, the muscle relaxants should be halted, and oxytocin and prostaglandin therapy should be initiated. Firm manual pressure via the introitus should be maintained until uterine contraction begins, the cervical ring contracts, and the uterus can no longer invert. If all these measures fail and obstetric/anesthesia backup becomes available, halogenated anesthetics may be used to induce relaxation of the cervical rings with or without an attempt at surgical repair.99
Uterine Rupture Criticism of the high rate of cesarean delivery in the United States has led to an advocacy of vaginal birth after cesarean (VBAC). Prior cesarean section is no longer an automatic indication for repeat cesarean delivery. The high success rate and relative safety of VBAC are countered partly by the risk of uterine rupture. Dehiscence of a surgical scar occurs in 0.6% of VBAC deliveries.101 As more women have VBACs, emergency physicians can expect to encounter uterine rupture.
Chapter 179 / Labor and Delivery and Their Complications
Once emptied, the uterus can be stimulated to contract. Uterine massage, oxytocin, and prostaglandins all may be used. Uterine Packing. Uterine packing to decrease postpartum hemorrhage was widely used previously but now is uncommon. For the emergency physician, this technique may be used to create tamponade, preventing further blood loss. The procedure has limited morbidity and is straightforward. The physician introduces 15 to 20 yards of 4-inch gauze with a ring forceps and packs it into the uterus using a layering technique. A special “uterine packer” is available to help direct gauze high into the uterus but is not necessary. Opponents of packing point out that an atonic uterus may accommodate a large volume of packing and blood without effective tamponade. Packing may also increase the risk of postpartum infection even when prophylactic antibiotics are given. As with all uterine manipulation and instrumentation, some risk of perforation also exists. Because dilation and curettage and hysterectomy sometimes are not available to the emergency physician, the importance of uterine packing as an option is increased. This approach is a temporizing measure.93 Pelvic Vessel Embolization. Pelvic bleeding postpartum can be difficult to control. Hysterectomy as a solution results in infertility and brings with it all the complications of general anesthesia and major surgery. Radiographic embolization of the bleeding vessels by an interventional radiologist is another option. The procedure does not require an anesthesiologist, operating room, or obstetrician and may, in fact, be more available on an emergent basis. The success rate of embolization is estimated to be 90%.94 A catheter is placed in the aorta and fluoroscopically guided to the bleeding sites that are imaged by radiopaque dye. The vessels are embolized with absorbable gelatin sponges placed via the catheter. Common sites of bleeding include the uterine artery, pudendal artery, and hypogastric artery. Because only the smallest involved branches are embolized and recanalization usually occurs, future reproductive capability is generally preserved.95 Uterotonic Agents. The use of uterotonic agents, although commonly applied upon delivery of the placenta, also has special application in the case of postpartum hemorrhage. Uterotonics, such as oxytocin, ergot alkaloids, and prostaglandins, control bleeding by inducing myometrial contraction. Oxytocin is considered to be first-line treatment, given either intramuscularly or intravenously. Ergot alkaloids, such as methergine and ergotamine, may induce hypertension and are therefore contraindicated in patients with preeclampsia or other comorbid conditions. Finally, prostaglandins may also be used, although the F class is contraindicated in asthma.84 Hysterectomy. Most postpartum hemorrhages are controllable with uterotonics and massage or uterine exploration for products of conception. Rarely, hemorrhage continues despite the interventions outlined. Life-threatening obstetric bleeding may require emergency hysterectomy. The desire to preserve the patient’s reproductive capabilities must not be given priority if her life is in jeopardy.96 A speedy search for the refractory cause of hemorrhage is warranted because coagulopathies may complicate obstetric hemorrhage. Disseminated intravascular coagulation (DIC) can occur as a consequence of placental abruption, eclampsia, amniotic fluid embolism, postpartum infections, and dilution of clotting factors caused by aggressive volume resuscitation. Also, retained products of conception and dead fetal tissue contain excess thromboplastin, which can initiate DIC. All women with severe postpartum hemorrhage should be evaluated for DIC. As with DIC from nonobstetric causes, clinical
PART V ■ Special Populations / Section Two • The Pregnant Patient
2346
Clinical Features. Uterine rupture is an unpredictable event occurring late in pregnancy or as stage 1 of labor transitions to the active phase. It is defined as a full-thickness uterine wall perforation. The severity of rupture ranges from simple scar dehiscence to complete fetal extrusion. It may be spontaneous, but it is most often liked with previous uterine surgery.102 This diagnosis should be entertained when appropriate because significant fetal mortality is associated with the event. As the degree of fetal expulsion through the rupture increases, the fetal mortality rate increases as well. Minimal fetal extrusion results in a perinatal mortality rate of less than 1%, whereas complete extrusion results in a 10 to 20% mortality rate. Maternal death is rare, but significant hemorrhage is common, complicating one third of cases. Maternal genitourinary injury may also occur in association with uterine rupture. Diagnostic Strategies and Difficulties. The diagnosis of uterine rupture is sometimes difficult because pain is not always present. Risk for uterine rupture with VBAC generally cannot be predicted based on maternal characteristics except that women with a prior classic, or T-shaped, incision and women who have had more than three cesarean sections are at increased risk. Intrapartum vaginal bleeding may signal the problem, but its absence by no means precludes rupture. Prolonged fetal heart rate deceleration, indicating fetal distress, is the most reliable sign of fetal extrusion.28 Emergency ultrasonography may reveal a protruding amniotic sac, hemoperitoneum, and/or the site of myometrial rupture; however, good sensitivity data are lacking.102 Management. If uterine rupture is suspected, delivery should be hastened to limit fetal hypoxia. Emergency cesarean section is the best method to speed delivery and repair the injury. The American College of Obstetricians and Gynecologists guidelines for uterine rupture identify a 30-minute window of opportunity that maximizes fetal outcome.103 At surgery, the maternal condition dictates whether uterine repair or hysterectomy is indicated. In the absence of opportunity for emergency laparotomy, appropriate interventions remain speculative. Uterotonic agents (especially ergonovine) may enlarge the rupture and are contraindicated.
Amniotic Fluid Embolism Amniotic fluid embolism is a rare and catastrophic complication of labor and delivery. The incidence rate is 6.0 and 14.8 per 100,000 in primigravid and multiparous deliveries, respectively.104 Although the mechanism is not well understood, it is thought to involve the spread of amniotic fluid through the maternal vasculature, activating either a procoagulant or anaphylactic cascade. Uterine trauma at or around the time of delivery, amniocentesis, and miscarriage also may result in amniotic fluid embolism. The diagnosis is usually clinically evident with the sudden onset of dyspnea, hypoxia, altered mental status, seizure, or hemodynamic collapse. DIC frequently follows and maternal mortality is high. In more than half of amniotic fluid embolism patients, postpartum bleeding due to coagulopathy occurs. Central hemodynamic monitoring, vasopressors/inotropes, and DIC management may be needed.105-107
Postpartum Endometritis Perspective. Puerperal infections affect 5% of all vaginal deliveries and 10% of all cesarean sections. Operative delivery, prolonged rupture of membranes, lack of prenatal care, prolonged stage 2 labor, use of intrauterine monitoring, and frequent vaginal examinations have all been linked to these ascending
gynecologic infections.108 Of all puerperal deaths, 8% have infection and sepsis as a direct or contributing factor.109 Causative organisms for these infections include gram-positive cocci and gram-negative coliforms. Less commonly, Chlamydia and Mycoplasma species have been implicated. Clinical Features. Endometritis is the most common puerperal infection, usually developing on the second or third day postpartum. Typically, the lochia has a foul odor and the white blood cell count is elevated. Fever and abdominal pain indicate greater severity of infection, often warranting inpatient care and intravenous antibiotics. Often, a coexistent surgical wound infection is present. A search for retained products of conception is indicated, particularly if bleeding is ongoing. Management. Treatment is empirical and directed at the most likely organisms. Clindamycin in conjunction with an aminoglycoside is usually used, although second- and third-generation cephalosporins are acceptable alternatives.110 Most patients with postpartum endometritis require admission.
Postpartum Cardiomyopathy Perspective. For unclear reasons, the postpartum period is associated with the relatively sudden onset of cardiomyopathy in healthy women without evidence of prior cardiac disease. Estimates indicate that postpartum cardiomyopathy (PPCM) occurs in 1 of 4000 pregnancies and is more common in African American and multiparous women. Proposed etiologies include viral, immunologic, toxic, and genetic factors, but in most cases no specific cause is found.111 Mortality rates for PPCM range from 18 to 56%. Clinical Features. Symptom onset varies, as does the severity of the cardiomyopathy. Onset is usually days to weeks after delivery, and symptoms range from mild fatigue to acute pulmonary edema. PPCM is often unrecognized in its milder form, leading to the consensus that the condition may be more prevalent than reported. Dyspnea on exertion, orthopnea, and fatigue may be easily misinterpreted as normal in a mildly anemic woman who is breast-feeding a new infant at home. The clinician should not dismiss these symptoms because congestive heart failure and dysrhythmias may ensue. Management. Treatment with diuretics, vasodilators, and oxygen relieves the symptoms in many cases. Angiotensinconverting enzyme inhibitors are contraindicated if PPCM occurs during the last month of pregnancy (owing to teratogenicity) but should be considered a mainstay of treatment postpartum. Amlodipine (a dihydropyridine calcium channel blocker) may also have a role in the treatment of PPCM.112,113 Cardiac function returns to normal in half of patients with PPCM during the following 6 months. Others have residual left ventricular dysfunction and a cardiac mortality of 85% during the next 5 years. The presence of cardiomyopathy after one pregnancy does not predict recurrence during subsequent pregnancies.114 Most obstetricians recommend against future pregnancies, however, believing that there is some residual cardiac function impairment. If such a pregnancy cannot be avoided, it should be considered high risk and followed closely.
Postpartum Depression Perspective. Although likely underdiagnosed, it is estimated that postpartum depression affects 10 to 15% of mothers. Although in many cases it is self-limited, the condition has been recognized as having important consequences for the mother, infant, and family. Risk factors for postpartum depression include previously diagnosed depression and neuroticism, inadequate
2347
Management. Early identification and referral are the key components of therapy. Dismissal of postpartum fatigue as normal, without considering the diagnosis of postpartum depression, can be disastrous. Not only does this condition contribute to marital discord, maternal risk for suicide, and even infanticide, but also studies have shown that children of depressed mothers have an increased incidence of delayed cognitive, psychological, neurologic, and motor development.119 Therefore, sensitivity to the possibility of postpartum depression is crucial to successful treatment. The disposition of inpatient psychiatric care with suicide precautions may be required, as deemed appropriate.
KEY CONCEPTS ■
ED deliveries should be considered high risk. Antepartum hemorrhage, PROM, eclampsia, premature labor, precipitous delivery, malpresentation, and umbilical cord emergencies are all overrepresented in emergency deliveries. ■ Women in labor who present to the ED are generally best cared for in the obstetric suite. Women with the urge to push or with the head of the infant crowning are at imminent risk of delivery, which should take place in the ED. The transfer of a woman with an impending high-risk delivery to a perinatal center must be tempered by clinical and medicolegal judgment. ■ Most ED deliveries require only basic equipment to cut and clamp the umbilical cord and to dry and suction the
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
infant. However, the ED should have the equipment and staff available to care for a newborn requiring further resuscitation. ■ Maternal complications of labor and delivery include obstetric trauma, postpartum hemorrhage, uterine inversion and rupture, amniotic fluid embolism, coagulation disorders, and infections. Many of these problems can initially be managed nonsurgically in the ED. ■ Deliveries complicated by dystocia, malpresentation, or multiple gestations are life-threatening emergencies. The clinician must develop strategies to treat each of these potential complications of delivery.
Chapter 179 / Labor and Delivery and Their Complications
spousal support, adverse socioeconomic factors, recent life stressors, and emergency delivery.115 Clinical Features. Postpartum depression patients present with symptoms not unlike those of other major depressive disorders. These symptoms include depressed mood, anhedonia, loss of appetite, insomnia, fatigue, decreased concentration, feelings of guilt and worthlessness, and suicidal ideation.116 Most women with postpartum depression do not have vegetative signs or symptoms. Symptoms peak at 10 to 12 weeks postpartum, although some cases are diagnosed up to 1 year postpartum.117 When unrecognized, these women are at high risk for suicide and may come to the ED with overdoses or other manifestations of a suicidal attempt.118
•
Section Three The Geriatric Patient Chapter 180
The Elder Patient
Diane M. Birnbaumer
There are no diseases of the aged, but simply diseases among the aged. Leonard Larson, 1960
■ PERSPECTIVE The population of the United States is becoming more elderly, with the proportion of people older than age 65 years increasing at twice the rate of younger people. In 2000, 12.4% of the population was older than age 65 years, and by 2045 one in five people may be older than that age.1 With the first baby boomer turning 62 years old in January 2008, estimates are that 365 people per hour will reach this milestone during the next decade, and the fastest growing subset of the U.S. population is people age 85 years or older. This group is also the fastest growing population seen in the emergency department (ED). Approximately 35% of total health care dollars are spent on patients older than age 65 years.2 These changing demographics affect the practice of emergency medicine. Elders account for 6.4 million of the 110 million ED visits; this number reflects a 34% increase in ED visits by this age group from 1993 to 2003.3 Nearly half of these patients are admitted. This group accounts for nearly 43% of all ED admissions and 47% of all critical care admissions.4 Because the number of ED visits for the elderly is expected to nearly double by 2013, the health care impact of this age group will be enormous. In general, elderly patients presenting to the ED are more likely to have an emergent condition than younger patients and spend a longer time being evaluated in the ED.5
■ PRINCIPLES OF DISEASE Physiologic Changes of Aging Physiologic changes of aging affect virtually every organ system and have many effects on the health and functional status of elders (Table 180-1). Heart disease is the leading cause of hospitalization and death.2 Increased peripheral vascular resistance with aging leads to an increased risk of hypertension. Decreased inotropic and chronotropic cardiac functioning compromises the patient’s ability to respond to physiologic stressors. Atherosclerosis is common and contributes not only to the rate of heart disease but also to the risk of vascular conditions (e.g., stroke, mesenteric ischemia, peripheral vascular disease, aortic dissection, and abdominal aortic aneurysm). 2348
Elders are at higher risk for infections due to decreases in antibody titers. Prostate disease and also incomplete bladder emptying in women with pelvic floor abnormalities predispose to urinary tract infections. Microaspiration increases the risk of pneumonia, and fragile, aging skin prone to injury and breakdown increases the risk of infections of the skin and soft tissues. Immunosenescence of cell-mediated immunity predisposes patients to reactivation of latent diseases (e.g., tuberculosis) and may be associated with increased susceptibility to neoplasms. Cancer is the second most common cause of hospitalization and death in older patients.2 Fractures are the fifth leading cause of hospitalization, reflecting the high rate of osteoporosis, particularly in women. Arthritis is the most prevalent outpatient disease in elders because of the wear on the cartilaginous joints, particularly of the knees, hips, and hands.2 Arthritis greatly affects quality of life, and these patients report fair or poor health approximately three times more often than patients without arthritis.6 Although the physiology of aging often affects a patient’s functional status, laboratory values are usually within the normal range. Abnormal laboratory values in elders should be evaluated as abnormal findings and should not be attributed to “normal effects of aging.”
Pharmacologic Considerations Polypharmacy, drug interactions, and misuse and abuse of medications in elders are crucial health care issues. Elders currently consume more than 30% of the prescription drugs in the United States, and this figure is projected to increase to 50% by 2020. More than 40% of elders use 5 or more drugs weekly, and more than 10% use 10 or more.7,8 Although multiple medications may be necessary to treat the medical problems that occur with aging, significant adverse health effects may result. Underlying medical problems, multiple physicians, changing pharmacokinetics of aging, and treating side effects of one medication with another drug are all contributory. Twelve to 30% of admitted elders have adverse drug reactions or interactions as a primary or major contributing factor to their admission, and 25% of these drug reactions or interactions are serious or life-threatening.9 Pharmacokinetics may change with age. Altered gastrointestinal motility and blood flow, decreased lean body mass, increased proportion of adipose tissue, decreased creatinine clearance, and decreased hepatic blood flow all may alter the absorption, distribution, and clearance of medications. Despite these changes, the bioavailability of most medications is not
2349
Table 180-1 Physiologic Changes of Aging and Potential Effects
Nervous System Decreased efficiency of blood-brain barrier Decreased response to changes in temperature Alteration of autonomic system function Alterations in neurotransmitters Skin/Mucosa Atrophy of all skin layers
Sweat glands decreased in number or activity Musculoskeletal System Progressive bone loss Atrophy of fibrocartilaginous and synovial tissues Decrease in lean body mass Increase in proportion of adipose tissue Immune System Decrease in cell-mediated immunity
POTENTIAL EFFECT
Increased risk of meningitis Potential for exaggerated medication responses Impaired thermoregulation Variations in blood pressure; risk of orthostatic hypotension Reduced erectile function Urinary incontinence Slowing of complex mental functioning Decreased insulation Increased risk of skin injury Increased risk of infection Potential for hyperthermia Increased risk of fractures Joint instability and pain Impaired balance and mobility Alteration in pharmacokinetics Alteration in pharmacokinetics
Decreased antibody titers
Increased susceptibility to neoplasms Tendency to reactivate latent diseases Increased risk of infection
Cardiovascular System Decreased inotropic response Decreased chronotropic response Increased peripheral vascular resistance Decreased ventricular filling
Less efficient response to myocardial wall stress Decreased maximal heart rate Increased blood pressure Changes in organ perfusion
Pulmonary System Decreased vital capacity Decreased lung/airway compliance Decreased chemoreceptor response to hypercapnia/hypoxemia Decreased ventilatory drive Decreased diffusion capacity
Increased airway resistance Potential for rapid decompensation Decreased Pao2 and increased Paco2 Decreased Pao2
Hepatic Function Decrease in hepatic cell mass Decrease in hepatic blood flow Alterations in microsomal enzyme activity
Reduced ability to regenerate Alteration in pharmacokinetics Alteration in pharmacokinetics
Renal System Decrease in renal cell mass Thickening of basement membrane Reduced hydroxylation of vitamin D Decrease in total body water Decreased thirst response Decreased renal vasopressin response
Decreased drug elimination Decreased drug elimination Risk of hypocalcemia, osteoporosis Alteration in pharmacokinetics Risk of dehydration/electrolyte abnormalities Risk of dehydration/electrolyte abnormalities
Gastrointestinal System Decrease in gastric mucosa Decrease in bicarbonate secretion Decrease in blood flow to gastrointestinal system Decreased epithelial cell regeneration
Increased risk of gastric ulcer Increased risk of gastric ulcer Increased risk of perforation Longer healing times
significantly altered in elders. Medication interactions and side effects pose significant problems, however, particularly because so many elders take multiple medications. Emergency physicians unwittingly may contribute to this problem by adding a new medication at discharge that may have an adverse drug interaction with a patient’s preexisting medications. In addi-
tion, the altered pharmacokinetics in elders necessitates caution when medications are administered in the ED, particularly sedative-hypnotics and narcotics. A rule of thumb is “start low and go slow.” Several medications frequently used in the outpatient elder population are “potentially inappropriate medications.” This
Chapter 180 / The Elder Patient
PHYSIOLOGIC CHANGE
PART V ■ Special Populations / Section Three • The Geriatric Patient
2350
list includes, but is not limited to, narcotic analgesics, nonsteroidal anti-inflammatory agents, sedative-hypnotics, muscle relaxants, and antihistamines.10-12 These agents are generally not recommended for use in this patient population and should be used sparingly. The medications most often implicated in adverse reactions in elders in the ambulatory setting are cardiovascular medications, followed by diuretics, nonopioid analgesics, hypoglycemics, and anticoagulants.13 Caution should be used when prescribing any of these classes that can contribute to morbidity and mortality in this group. Narcotics and sedative-hypnotic agents can decrease cognition and increase the risk of falls and accidents. Diuretics may cause dehydration or electrolyte imbalance. Nonsteroidal antiinflammatory drugs (NSAIDs) may have serious and potentially lethal side effects. The toxicity of NSAIDs includes azotemia, worsened hypertension, and congestive heart failure as a result of sodium retention. Gastrointestinal toxicity ranges from bleeding to perforation.14 The data support using extreme caution when prescribing NSAIDs. These complications can be seen in patients taking widely available over-the-counter formulations of these drugs. There is a significant increase in cardiovascular risk with the use of COX-2 inhibitors. Due to these concerns, pain management is difficult. Although a “first, do no harm” philosophy is laudable, treatment of pain is one of the most important remedies physicians have to offer. Three major considerations influence the choice of medication for elderly patients. First, chronic pain (e.g., arthritis pain) may require different treatment from an acute painful condition (e.g., wrist fracture). Second, the patient’s underlying medical problems, social situation, and baseline functional status need to be considered. Third, starting with low doses, then titrating upward to effect, is the safest strategy. Overall, acetaminophen has the highest safety profile in elders and often is the drug of choice for chronic painful conditions (e.g., degenerative joint disease). Although inadvertent excessive dosing can occur, acetaminophen is safe and should be considered first-line management for chronic pain and acute conditions causing mild to moderate pain. NSAIDs effectively ease pain, but these agents have a ceiling analgesic effect. Increasing the dose increases the risk of side effects without increasing the analgesic benefit. These agents in low doses should be used as second-line agents after acetaminophen in chronic or acute painful conditions. Narcotics may be necessary for acutely painful conditions and chronic conditions refractory to acetaminophen or NSAIDs. The common side effect of constipation can be problematic, and older patients need to be instructed to exercise regularly, eat high-fiber foods or supplements, and remain well hydrated. Because these agents may decrease cognition, start with low doses and titrate upward as needed.
Psychosocial Issues The possibility of drug dependence and ethanol abuse should be considered when treating elders. Ethanol dependence is a factor in 14% of elders who present in the ED and is more prevalent than suspected.15,16 These patients often present with gastrointestinal complaints or after falls or other trauma. Iatrogenic dependence on prescription drugs, particularly sedative-hypnotics, is also more prevalent than suspected. The routine use of sedative-hypnotics should be avoided and duration limited. Dependence on these drugs may cause decreased cognition, and withdrawal from these agents can be life-threatening. Psychiatric disease in elders often manifests in atypical fashion. Depression, a common problem, may manifest as
agitation, anxiety, and somatic complaints, in addition to the typical depressive symptoms.17,18 Depression often follows chronic illness, loss of physical mobility, diminished cognitive function, bereavement from the death of a spouse or long-time friend, and financial pressures, all of which are common features of old age. Social isolation and loss of independence produce a sense of helplessness and hopelessness that may result in suicidal ideation or action. Certain types of depression, such as “late-life delusional depression” and involutional depression, occur exclusively in old age. In addition, depression may be caused by medication side effects or reversible physiologic conditions (e.g., thyroid disease and malnutrition). Elders often respond to pharmacologic treatment of depression but are prone to develop adverse effects from antidepressants; selective serotonin reuptake inhibitors seem to be safer. A phenomenon known as sundown syndrome occurs in elderly demented patients, who become highly agitated and disoriented after dark when visual sensory input is diminished and the environment becomes unfamiliar.
■ EVALUATION AND CLINICAL FEATURES The evaluation of elders is often difficult. Elders living independently have an average of 3 medical problems, which increases up to 10 for those living in care facilities. Underlying illnesses complicate the evaluation. Many emergency physicians are uncomfortable evaluating elders because a specific complaint is significantly more difficult than managing the same problem in a younger patient. Use of ancillary services can be increased by 50% in elders, most likely because of vague or atypical presentations and complicated medical backgrounds.
History Obtaining a medical history from an elderly patient requires meticulous and painstaking work. Cognitive and physical deficits must be recognized, and physicians often need to be creative and thorough to obtain adequate information. Cognitive deficits may compromise an elder’s recall and result in an inaccurate medical history. Enlisting family members, consulting with the patient’s primary care physician, and reviewing past medical records may be necessary. Particular attention should be paid to past medical and surgical problems and to the patient’s current medications, including over-thecounter and herbal preparations. Physical deficits also may impede the history-taking process. Sequelae from previous strokes (e.g., aphasia) are usually obvious. Hearing impairments affect communication and can lead to dangerous misunderstandings. Elders tend to lose high-frequency hearing earlier than other ranges, so physicians should lower the pitch of their voice and speak loudly while ensuring the patient’s privacy. Hearing loss can be embarrassing for the patient; thus, the physician should address the issue with sensitivity to allow for adequate communication while maintaining the patient’s dignity.
Physical Examination Due to the physiologic changes that occur with aging, the physical examination may be misleadingly benign in an elder, despite the presence of a potentially lethal illness. In addition, medications may alter the response of elders to physiologic stressors. Antihypertensives, particularly beta-blockers, may alter the patient’s ability to mount tachycardia in response to hypovolemia or sepsis or may predispose the patient to hypotension.
■ DIAGNOSTIC STRATEGIES Because of the difficulty in obtaining accurate histories and the often poor sensitivity of physical findings in elders with significant underlying disease, more diagnostic tests are generally obtained for these patients than for their younger counterparts who present with similar complaints.5 This increased use of resources is appropriate because the accuracy and timeliness of diagnosis are critical in elders, who suffer greater morbidity and mortality with delays in definitive treatment.
■ SPECIFIC DISORDERS Myocardial Infarction The incidence of atypical presentation of acute myocardial infarction (AMI) increases with age.21,22 In patients older than 85 years, atypical presentation of myocardial infarction may be anticipated, and a lack of chest pain may be the rule. Only 2 to 6% of elders with AMI, however, have an asymptomatic presentation.21 A painless myocardial infarction is more common with increasing age and occurs more often in women than men.22-24 Sudden onset of dyspnea may be the only presenting complaint in elders with myocardial infarction. Other presenting complaints include syncope, flulike symptoms, nausea, vomiting, confusion, and weakness.22-24 The prognosis for AMI in elders who present atypically is the same as that for elders who present typically; atypical presentations are not more benign.
Infections Elders are more prone to infectious diseases than are younger patients, with greater morbidity and higher mortality from these diseases. Common infections occur regularly, but aging causes this population to be more susceptible to unusual organisms. Although immunosenescence is a contributory factor, predisposing illnesses and institutionalization are more significant causes of this increased risk. Hospitalization increases the risk of nosocomial infection. Instrumentation and catheterization are also significant risk factors for acquiring infectious diseases. Evaluating infections may be difficult because 48% of elders with proven bacterial infections do not have a fever at presentation.20 In addition, the sensitivity of an elevated white blood cell count and bandemia is poor—approximately 44 and 32%, respectively.19 Several infections, particularly pneumonia, urinary tract infections, and sepsis, occur more often in elders. Pneumonia is one of the 10 leading causes of hospitalization and death.2 The elderly are more prone due to decreased vital capacity, decreased lung and airway compliance, less ventilatory drive, and poorer ciliary function. Combined with decreased cough
2351
response and increased esophageal reflux with microaspiration, pneumonia is the leading cause of hospitalization due to infection in the elderly. Pneumococcus is still a common cause, but gram-negative organisms are also common, as are mixed infections. Reactivation of tuberculosis also must be considered in this age group. Urinary tract infections are common in elders. The incidence of bacteriuria is 20% in men older than age 70 and 20% in women ages 65 to 70, increasing to 23 to 50%, respectively, in men and women older than age 80.25 Laxity of the pelvic floor and urinary incontinence are significant risk factors in women, whereas prostate enlargement is a significant risk factor in men. The incidence of urinary tract infections increases dramatically in patients with chronic indwelling catheters.
Abdominal Pain Abdominal pain may be the most difficult presenting complaint to evaluate in elderly patients. Despite this difficulty, most patients have a specific diagnosis made in the ED. Two thirds of elders with abdominal pain are admitted, and nearly one fifth go directly to the operating room. The differential diagnosis (Table 180-2) of abdominal pain differs significantly from that of younger patients, particularly in regard to the number of serious and potentially lifethreatening causes. The pathology in more than 60% is surgical in nature, a rate nearly double that of younger patients. There is a 10-fold higher risk of mortality compared with younger patients.26 Due to the physiologic changes of aging, even life-threatening causes of abdominal pain may present with few or no alarming findings. Elders may complain of vague abdominal pain despite the occurrence of a catastrophic process. The complication rates of typically benign processes in younger patients are dramatically higher in older patients. With aging, the abdominal musculature decreases, and patients are less able to manifest guarding and rebound. In addition, the omentum shrinks and is less able to contain intra-abdominal processes. Atherosclerotic disease, with its resultant decrease in blood flow, causes increased perforation rates in diseases such as cholecystitis and appendicitis. This increased rate of vascular disease also contributes to higher rates of vascular causes of abdominal pain, such as mesenteric ischemia and leaking or ruptured abdominal aortic aneurysms.27 The high prevalence of gallstones in elders leads to an increased risk of cholecystitis. Due to the vague presentation and high rate of serious disease, evaluating an elderly patient with abdominal pain often requires an extensive battery of laboratory and radiographic tests.28 Elders with potentially catastrophic intra-
Diagnoses of Abdominal Pain in Table 180-2 Differential Elderly Patients DISORDER
Cholecystitis/biliary colic Nonspecific abdominal pain Appendicitis Obstruction Hernia Perforation Pancreatitis Diverticular disease
INCIDENCE (%)
12–41 9.6–23 2.5–15.2 7.3–14 4–9.6 2.3–7 2–7.3 3.4–7
Chapter 180 / The Elder Patient
Of elders with a serious infection, 30% present with a blunted or absent fever response.19 The temperature measurement must be accurate; an oral temperature may be spuriously low, so a rectal temperature should be measured when fever is uncertain in a patient with a possible infectious disease. When fever is present, elders are much more likely to have a serious (nonviral) infection than are younger patients.20 Of febrile elders presenting to an ED, 89% have an infectious disease. Approximately one third are respiratory tract infections, one fifth are urinary tract infections, and nearly 20% are bacteremia or sepsis.20 Fever in an elder must be taken seriously, with appropriate ancillary testing and a low threshold for admission. A lack of fever does not exclude infection.
PART V ■ Special Populations / Section Three • The Geriatric Patient
2352
abdominal processes may not present with a fever or an elevated white blood cell count.29,30 As a result, ancillary diagnostic ultrasonography, computed tomography, radionucleotide studies, and, occasionally, angiography may be important.27 Since many elders with abdominal pain have a serious disease, consider admission and close observation when symptoms persist and the diagnosis remains unclear. If the patient is not admitted, a prolonged period of ED observation or reevaluation within 12 hours is prudent.
Major Trauma Major trauma in elders (see Chapter 36) is relatively uncommon, constituting 8 to 15% of cases in major trauma databases. Elder trauma patients, however, experience higher mortality and poorer functional recovery for a given trauma score.31,32 In any elderly trauma patient, the circumstances leading to the injury need to be determined. For any elder patient who pre sents with a fall, the circumstances of the fall should be questioned. Although mechanical falls occur, falls may be due to potentially serious or life-threatening medical causes such as syncope, hypovolemia due to dehydration or bleeding, cardiac or cerebrovascular disease, or medications.33 Motor vehicle crashes, particularly involving a single vehicle, may result from transient loss of consciousness due to dysrhythmias, syncope, medication side effects, transient ischemic attacks, strokes, or myocardial infarctions.34 These serious medical problems require simultaneous diagnosis and treatment in the setting of trauma and may be as important as or more important than the traumatic injuries. The presence of comorbid illness and the limited physiologic reserve in the cardiopulmonary and renal systems often complicate the trauma resuscitation in this age group. More aggressive ventilatory support and early use of invasive hemodynamic monitoring are indicated to guide volume resuscitation. Shock is poorly tolerated in elders because hemodynamic compensation is limited, and end-organ failure occurs earlier. Certain injuries are more common and severe in elders. A subdural hematoma may occur after relatively trivial head trauma, and chronic subdural hematoma may present as progressive dementia with only subtle neurologic findings. Increasing ankylosis of the spine, osteoarthritis, and decreased bone density as a result of osteoporosis render the geriatric cervical spine more susceptible to fracture. Preexisting pulmonary pathology and a brittle thoracic wall account for the greater severity of pulmonary contusions and the higher incidence of rib fractures and resultant complications (e.g., atelectasis and pneumonia). Skeletal fractures are more common and lacerations of atrophic skin are more difficult to repair and more prone to infection. This thin skin is also prone to decubitus breakdown in patients who are immobile, even for short periods such as during spinal immobilization on a backboard.
■ PREVENTIVE CARE Immunizations Pneumonia, influenza, accidents, and adverse medical events are among the top seven causes of death in elders.2 These conditions account for 15% of admissions and are potentially preventable. Annually, 36,000 adults die of complications from influenza and pneumococcal infections, and most of these deaths occur in elders. Immunization would reduce the incidence of clinical and serologic influenza by half in this patient population. The Centers for Disease Control and Prevention (CDC) has a goal of an 80% immunization rate for patients in high-risk groups, including elders. The actual vaccination rate
falls significantly below this goal. In patients 65 years old, 66 and 62% were vaccinated for influenza and pneumococcus, respectively, with the lowest rates among multiracial ethnic groups and people of lower socioeconomic means.35 As a result, the CDC recommends that potential vaccination sites be extended to include walk-in clinics and EDs. Elders who have not been hospitalized or seen their primary care physician in the previous 3 years average at least one visit to an ED, and many are willing to be vaccinated there. Despite this information, controversy exists regarding vaccination of elders in the ED. Emergency physicians can be reluctant to give vaccinations, citing time constraints and their concern regarding delivering primary care. Considering the low threshold in the ED to administer the tetanus vaccine, it seems feasible that influenza and pneumococcal vaccination programs could be relatively easy and effective in the ED.
Falls In combination with vaccinations, education about accident prevention in the home could have a considerable impact on the overall morbidity and mortality in elders. Falls and adverse medical events are the seventh leading cause of death in elders.2 A primary care or social service provider more appropriately supplies this instruction. The ED, however, may be an additional resource for providing educational services to elders on a case-by-case basis, particularly in reference to any specific incident that led to the need for emergency care. The leading cause of falls in elders is related to the use of pharmacologic agents, often prescription drugs.36 Reviewing the patient’s medications, searching for agents that might cause decreased cognition or dehydration, and addressing the need for these agents with the patient and the primary care provider could decrease significantly the risk for falls. In addition, informing the patient’s primary care provider that the patient has fallen may facilitate education by the primary physician.
KEY CONCEPTS ■
Polypharmacy and side effects of prescription and nonprescription drugs often occur in elders and should be considered in their evaluation. ■ Physiologic changes that occur with aging make the clinical evaluation of elders difficult. Because of the difficulty in obtaining accurate histories and the often poor sensitivity of physical findings in elders with significant underlying disease, more diagnostic tests are generally obtained for these patients than for their younger counterparts who present with similar complaints. ■ The immunosenescence that accompanies aging may blunt the fever response to infection and cause a less elevated white blood cell count. ■ Myocardial infarction frequently presents atypically in elders as dyspnea, syncope, weakness, confusion, or abdominal complaints, often in the absence of accompanying chest pain. ■ Abdominal pain in elders is often caused by surgical conditions that may require extensive evaluation in the ED, including radiographic studies. ■ Elderly trauma patients have a higher morbidity and mortality than their younger counterparts because of exacerbation of underlying medical problems. The references for this chapter can be found online by accessing the accompanying Expert Consult website.
•
Section Four The Patient with Compromised Immune Function Chapter 181
The Immunocompromised Patient
Michael J. Burns
■ PERSPECTIVE Compared to individuals with an intact immune system, infections in immunocompromised patients are more common, are more severe, progress more rapidly, are more often fatal, and are caused by a wider variety of microorganisms.1,2 Many factors, often interrelated, cause patients to become immunocompromised and predispose them to develop infections with potentially pathogenic microorganisms. These include disruption of the body’s protective surfaces such as skin and mucosal barriers (oral and respiratory mucosa and intestinal and genitourinary surfaces); disorders (e.g., lymphoma, asplenism, and myeloma) that directly impair the function of the body’s immune system; drugs and irradiation that suppress or alter immune function; alterations in body substances (hyperglycemia) or solid organ function (kidney and liver failure); as well as malnutrition, aging, and exposure to antimicrobial agents that inhibit the normal protective resident bacterial flora.
■ PRINCIPLES OF DISEASE The body’s defense mechanisms consist of surface barriers, such as skin, enzymes, and mucus, as well as innate (natural) and acquired (adaptive) responses. Innate responses occur to the same extent regardless of how often the body encounters the infectious agent, whereas acquired responses improve on repeated exposure.3 Innate immunity is activated immediately upon exposure to an infecting agent, rapidly controlling replication and allowing the requisite 3 to 5 days for the adaptive component to clone sufficient T and B cells to respond more specifically.4
Non-Microbe-Specific Immunity Physical Barriers The first line of defense against microorganisms consists of physical barriers, including skin, gastrointestinal and respiratory mucosa, and gastric acid. Cutaneous acidity inhibits dermal bacterial growth. In addition, mucous membranes are continually bathed in secretions that contain antimicrobial enzymes, other proteins, and immunoglobulin G (IgG) and secretory IgA.5 In the respiratory tract, mucociliary transport and the cough reflex remove particulate matter and microbes, but this mecha-
nism is impaired with smoking and ineffective cough. Mechanical ventilation or tracheostomy introduces large numbers of microbes that often overwhelm natural clearance.6 Gastric acid and pancreatic enzymes have antibacterial properties that prevent overgrowth in the upper gastrointestinal tract. Normal peristalsis and mucosal shedding help maintain normal gut flora. Alterations in these factors increase susceptibility to infection. Broad-spectrum antibiotics alter normal flora, permitting overgrowth of pathogens such as Candida species, multiantibiotic-resistant bacteria, and Clostridium difficile.
Acute Phase Response The metabolic changes that occur during systemic infection and inflammation constitute the acute phase response. The liver manifests a stress response by decreasing albumin synthesis while increasing production of proteins that enhance phagocytosis and assist in complement activation. One of these proteins is C-reactive protein (CRP), sometimes measured clinically as a nonspecific indicator of infection or inflammation. Muscle proteolysis is also part of this response and may result in the weight loss characteristic of chronic inflammatory or infectious states. Invading microorganisms activate the acute phase response, which is not microbe specific. The response delivers humoral and cellular immune components to sites of inflammation and initiates antibody responses. Cytokines, platelet-activating factor, and hormone-like proteins, including interferons, are secreted from various immune cells and play important roles in mediating this response.7 Clinically, these cytokines result in migration and adhesion of polymorphonuclear leukocytes and monocytes to sites of bacterial invasion. These cells release granules of substances that vasodilate and increase vascular permeability, leading to edema, warmth, and redness, but also allow both phagocytic cells and humoral components to concentrate at the site of infection. A family of distinct transmembrane proteins, called toll-like receptors, which mediate recognition of extracellular microbial products such as endotoxin, are vitally important in the innate immune response.5 They are present on many cell types, including macrophages, neutrophils, dendritic cells, mucosal epithelial cells, and endothelial cells. They recognize pathogen-associated molecular patterns, generate signals that lead to activation of innate immune responses, and help bridge innate and adaptive immune responses. 2353
2354
PART V ■ Special Populations / Section Four • The Patient with Compromised Immune Function
Reticuloendothelial System The reticuloendothelial system, composed of tissue macrophages and their blood-borne counterparts, monocytes, removes particulate matter, including microbes, from the lymph and blood. The tissue component is concentrated in the lymph nodes, spleen, liver, marrow, and lung and has particular affinity for encapsulated bacteria such as pneumococci, meningococci, and Haemophilus influenzae. The vital importance of this non-microbe-specific system is demonstrated by the overwhelming sepsis from encapsulated organisms in patients with asplenia.
Adaptive (Microbe-Specific) Immunity Humoral Immunity Antibodies. Antibodies are produced by B lymphocytes, and each B cell produces a single microbe-specific antibody type. Stimulation by an antigen (or microbe) causes proliferation of this particular B cell so that large quantities of a specific circulating antibody can be produced. Furthermore, B cells are active in presenting antigens to T lymphocytes, which promotes cell-mediated immunity. Immunoglobulins. IgM is the first immunoglobulin to appear in response to a new antigen. Although it has less affinity at binding antigens than IgG, IgM provides some recognition of antigens and begins B cell proliferation before the subsequent development of IgG.8 Clinically, IgM is detectable earlier in serum than IgG and serves as a marker for a patient’s early response to acute infection. Secretory IgA is the predominant immunoglobulin present in gastrointestinal fluids, nasal and oral secretions, tears, and other mucous fluids. IgA inhibits cell adherence of viral, bacterial, and protozoan pathogens and therefore prevents invasion by organisms through the respiratory or gastrointestinal tract.9,10 IgE, which is expressed in high concentration on the surface of mast cells and basophils, is responsible for immediate-type hypersensitivity responses. Mast cells and IgE are important in defense against helminthic pathogens. IgG, which accounts for 75% of the total immunoglobulin mass, is widely distributed in tissues. It crosses the placenta and provides fetal immunity during the first 6 months of life. Congenital or acquired deficiencies of IgG lead to infection with encapsulated organisms because the predominant subtype (IgG2) has affinity for the dense polysaccharides of bacterial cell capsules, such as those of Streptococcus pneumoniae and H. influenzae. Complement. The complement cascade, consisting of a complex interaction of 30 proteins, is another crucial component of humoral response. Complement is important in producing inflammation and leukocytosis and in recruiting leukocytes to sites of infection by production of chemoattractants. Complement also neutralizes viruses, enhances opsonization of bacteria, and produces bacterial cell wall and membrane lysis. Both IgG and IgM, when in contact with an antigen, activate the classical pathway, whereas molecules with repeating chemical structures (e.g., bacterial cell walls and capsules) activate the cascade through the alternative pathway. Components of C3, the merging point of the classical and alternative paths, provide opsonization and modulate the response of lymphocytes (cell-mediated immunity). Opsonization is important in defense against infection with S. pneumoniae, Streptococcus pyogenes, H. influenzae, and Staphylococcus aureus. The terminal leg of the cascade, C5 through C9, forms the membrane attack
complex, which inserts into cell walls and membranes and leads to cell death. Individuals with inherited complement deficiencies are predisposed to frequent and recurrent infections with S. pneumoniae, H. influenzae, and especially Neisseria meningitidis and Neisseria gonorrhoeae.11 The risk of meningococcal infection is increased several thousandfold and most often develops in people deficient in C3 and in late complement components (C5–C8). Paradoxically, the disease is usually milder with complement deficiency, and mortality is likewise reduced 5- to 10-fold.12 This suggests that the host response may be, in part, responsible for the severity of disease in normal individuals and is attenuated in complement deficiency. People with meningococcemia should be tested for inherited complement deficiencies because they may benefit from immunization. Acquired deficiencies of complement function may develop in people with rheumatologic diseases, especially systemic lupus erythematosus (SLE). Approximately 40% of patients with SLE have an inhibitor of C5a-derived chemotaxis in their serum that results in enhanced susceptibility to infection.13
Cell-Mediated Immunity Cell-mediated immunity (CMI) generally includes immune responses that are mediated by T lymphocytes, natural killer (NK) cells, and mononuclear phagocytes. CMI is vitally important in the control of infections caused by microbes that survive and replicate intracellularly, including most viruses, and some bacterial (obligate and facultative intracellular types), fungal, and protozoan pathogens. Only 5% of lymphocytes are in circulating blood. Most mature and are active in the marrow, thymus, spleen, and lymph nodes. The latter two sites expose T cells to circulating antigen from invading microbes.5 Specialized antigenpresenting cells in the lymphoid system sequester antigen and antigen-antibody complexes and present them to T cells. This process involves internalization and processing of the antigen, followed by formation of peptides that bind to a cell surface molecule called the major histocompatibility complex (MHC). Only with this specific presentation can a T lymphocyte become activated against a particular antigen. Two major types of T lymphocytes are CD4 (helper cell) and CD8 (suppressor cell), corresponding to type II and type I of MHC, respectively. CD4 lymphocytes provide help for other cells in the immune system, including enhanced B cell antibody production and production of cytokines. CD8 lymphocytes are generally cytotoxic and mediate the eradication of virally infected target cells and certain tumors. A decline in the number of CD4 cells, with predominance of CD8 cells, is responsible for the increased susceptibility to infection in patients with acquired immunodeficiency syndrome (AIDS).5 Despite the cytotoxicity of CD8 cells, immunity is reduced without adequate numbers of CD4 cells. Patients with defects in CMI are at increased risk for disseminated infection with intracellular bacteria, such as Mycobacterium tuberculosis, Listeria monocytogenes, and Salmonella species. The deoxyribonucleic acid (DNA) viral infections, such as cytomegalovirus, herpes simplex, and varicella zoster, also affect these patients more severely, as do fungal infections with Candida, Cryptococcus, Mucor, Aspergillus, and Pneumocystis. Finally, some protozoa are pathogenic without intact CMI, as infections with Toxoplasma gondii demonstrate.14,15 Some infections are seen only below a certain CD4 cell count. Pneumocystis pneumonia is seen almost exclusively in patients with counts less than 200 cells/mL, whereas almost all patients with
2355
once clear demarcation between cellular and humoral immunity is breaking down as more is understood about the interdependent immune system.
■ SPECIFIC IMMUNOCOMPROMISED STATES Solid Organ Transplants See Chapter 182.
Cancer Patients with cancer frequently have multiple immune defects that predispose them to infection, such as neutropenia and impaired T cell and B cell function, which are induced by cancer chemotherapy or by the disease process. Other factors leading to infection are defects in physical barriers (skin and mucous membranes), including cytotoxic effects of chemotherapy on cells lining the gastrointestinal tract. In addition, splenic dysfunction or splenectomy, use of long-term intravascular catheters, frequent use of complex invasive diagnostic and therapeutic procedures, toxic effects of radiation therapy, and frequent colonization with antimicrobial-resistant pathogens are predisposing factors. Cancer treatments (e.g., allogeneic bone marrow and autologous stem cell transplantation, platelet transfusion, granulocyte colony-stimulating factor, and implanted central venous catheters) increase survival during episodes of profound immunosuppression, allowing patients to receive more intense cytotoxic cancer chemotherapy regimens. This results in long survival of patients with neoplastic diseases that were formerly rapidly fatal. Despite many advances in supportive care, infections continue to result in serious morbidity and mortality. Furthermore, increasing resistance to antimicrobials is occurring among common pathogens, along with the emergence of new opportunistic pathogens. Infection is much more common in patients with acute leukemia and lymphoma (75% of patients) and multiple myeloma (50% of patients) compared to those with solid tumors.17 Factors predisposing to infection in immunocompromised patients are listed in Box 181-1.
Neutropenia Principles of Disease. Neutropenia is defined as a neutrophil count of less than 500 cells/mL, including band forms, or less than 1000 cells/mL and expected to fall to less than 500.18,19 It usually results from cytotoxic chemotherapy or radiation therapy or the disease process, especially in hematologic malignancies. In addition, cancer chemotherapeutic agents and radiation therapy can cause functional defects in granulocytes. The risk of febrile neutropenia and mortality is higher in the first one or two cycles of multicycle cytotoxic chemotherapy regimens.20 The incidence and severity of infection in cancer patients with neutropenia are inversely proportional to the absolute neutrophil count and directly proportional to the duration of neutropenia. Although the incidence begins to rise as the neutrophil count falls below 500 cells/mL, most severe infections and almost all bacteremias occur when the neutrophil count is less than 100 cells/mL.21 Fever in the neutropenic patient is defined as a single temperature of 38.3° C (101° F) or higher or a temperature of 38.0° C (100.4° F) or higher over 1 or 2 hours.19 In neutropenic patients, the temperature should be measured orally or tympanically, not rectally. Although fever can be suppressed or lessened by immunosuppressive agents such as corticosteroids and nonsteroidal anti-inflammatory
Chapter 181 / The Immunocompromised Patient
toxoplasmosis or cryptococcal meningitis have counts less than 100 cells/mL. NK cells, closely related to lymphocytes but neither B nor T cells, are important in the innate immune response and are found in high concentrations in blood and spleen.5 NK cells recognize infected cells and respond by directly killing these cells, and they secrete cytokines that activate macrophages to destroy phagocytosed microbes. These cells are very important in defense against intracellular microbes, particularly viruses and intracellular bacteria such as L. monocytogenes. Granulocytic Phagocytes. Granulocytic phagocytes are the cellular effectors of microbe killing, engulfing them and enzymatically lysing their cell membranes or walls. Two major types are polymorphonuclear leukocytes (also called neutrophils) and macrophages (the tissue version of circulating monocytes). Macrophages have surface receptors that recognize nonvertebrate carbohydrates such as mannose, which form the cell wall of some microorganisms. Hence, they can identify and attack “invaders” rather than “self.” Two other types of granulocytes, eosinophils and basophils, are less involved in the ingestion of organisms.16 Eosinophils mediate the destruction of certain parasitic helminths through release of toxic proteins. Normally only 3% of total granulocytes, this cell type can reach 20% during times of high parasite load. Basophils (rare in circulation) and their tissue counterparts, mast cells, have high affinity for IgE. When exposed to antigens, they release granules with histamine, prostaglandins, and leukotrienes, which affect the allergicinflammatory response with increased vascular permeability, bronchospasm, and vasodilation.3 Neutrophils constitute 90% of circulating granulocytes and spend only 6 to 8 hours of their average 4-day life in circulation (the remainder in tissues). Effective antibacterial activity depends on the ability of neutrophils to travel to sites of infection—a process known as chemotaxis. The locomotion of neutrophils along vascular endothelium is facilitated by adherence to cell surface proteins whose production is enhanced in the acute phase response.16 One half of all neutrophils that leave the bone marrow circulate in the plasma. The other half become marginated, adhering to endothelium, primarily in the lungs, liver, and spleen. During periods of stress or with endogenous or exogenous catecholamines or corticosteroids, these neutrophils demarginate and enter the circulation. As long as the patient is not neutropenic, demargination causes an increased peripheral neutrophil count composed of mature cells, whereas with infection, an increased proportion of immature (band) forms is typically seen. Neutrophils (and macrophages in tissue) bind to and ingest bacteria—a process called phagocytosis. This process is enhanced by proteins called opsonins that bind to bacterial surfaces. CRP, one of the acute phase response proteins, fulfills this function for certain bacteria, including S. pneumoniae. IgG and complement protein C3b also opsonize bacteria, again illustrating the interdependence of the immune system. Actual killing takes place within granulocytes when cytoplasmic granules enzymatically produce potent oxidants. Granulocytes further control bacterial proliferation at the site of infection by elaborating lactoferrin, which locally binds free iron necessary for bacterial replication. In addition to phagocytosis, macrophages (located in the spleen, alveoli, liver, and lymph nodes) modulate the immune response by presenting antigens to lymphocytes and releasing cytokines and complement components. Activation of macrophages to ingest bacteria depends on interaction with interferon-γ, a cytokine manufactured by T cells.5 Thus, the
2356
PART V ■ Special Populations / Section Four • The Patient with Compromised Immune Function
BOX 181-1 The Immunocompromised Patient: Factors Predisposing to Infection and the Most Common Pathogens Neutropenia Bacteria Gram-negative bacilli Escherichia coli Klebsiella pneumoniae Pseudomonas aeruginosa Enterobacter sp. Serratia sp. Citrobacter sp. Proteus sp. Stenotrophomonas maltophilia Gram-positive cocci Staphylococcus epidermidis Staphylococcus aureus Viridans streptococci Less common: Enterococcus sp. Gram-positive rods Corynebacterium sp. Less common: Bacillus sp. Fungi Candida sp. Aspergillus sp. Less common: Mucor sp., Rhizopus sp., Trichosporon beigelii, Fusarium sp., Pseudallescheria boydii Cellular Immune Dysfunction Bacteria Listeria monocytogenes Salmonella sp. Mycobacterium tuberculosis Mycobacterium avium-intracellulare Legionella sp. Nocardia sp. Fungi Cryptococcus neoformans Histoplasma capsulatum Coccidioides immitis
drugs, most cancer patients with infection manifest fever despite the use of these agents.22 Also, although it is uncommon, immunocompromised patients can have serious local or systemic infections without fever, manifested by unexplained tachypnea or tachycardia, mental status changes, metabolic acidosis, increased volume requirements, rapid changes in serum glucose or sodium, or acute abdominal pain. Because cancer patients with severe neutropenia or a history of splenectomy can develop the rapid onset of life-threatening infections, urgent evaluation and initiation of antimicrobial therapy in the emergency department are essential. The most common sites of infection in neutropenic patients are the lung (25%); mouth and pharynx (25%); gastrointestinal tract (15%); skin, soft tissue, and intravascular catheters (15%); perineum and anorectal area (10%); urinary tract (5%); and nose and sinuses (5%).23 Pneumonia and anorectal infection are more likely to be associated with bacteremia. Bacteremia may occur without an obvious source despite intensive investigation. Historically, the most important bacteria are three gram-negative bacilli—Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa—and four gram-positive cocci— Staphylococcus epidermidis, viridans group streptococci, Enterococcus species, and S. aureus. Many centers that treat large numbers of cancer patients have noted a decrease in these
Candida sp. Aspergillus sp. Pneumocystis jirovecii (formerly carinii) Viruses Herpes simplex Varicella zoster Cytomegalovirus Epstein-Barr Less common: Measles, adenovirus Parasites Toxoplasma gondii Cryptosporidium sp. Stronglyoides stercoralis Humoral Immune Dysfunction (Antibody Deficiency) Bacteria Streptococcus pneumoniae Haemophilus influenzae Neisseria meningitidis S. aureus Splenectomy/Functional Asplenia Bacteria S. pneumoniae H. influenzae N. meningitidis Capnocytophaga canimorsus Bordetella holmesii Parasites Babesia sp. Complement Deficiency Bacteria N. meningitidis S. pneumoniae H. influenzae
gram-negative bacilli and an increase in infections caused by others, such as Enterobacter, Citrobacter, and Serratia species, which are capable of rapidly developing resistance to cephalosporins and extended-spectrum penicillins. Anaerobes are uncommon but may be important in certain mixed infections (e.g., mouth, abdominal, and perianal). During the past 25 years, infection with gram-positive organisms (e.g., coagulase-negative staphylococci, S. aureus, viridans streptococci, and Enterococcus species) has increased and this is now the leading cause of bacterial infection (50–70% at some centers) in febrile neutropenic cancer patients in the United States, Canada, and Western Europe. Gram-negative organisms still predominate in developing countries.24,25 With the exception of viridans streptococci, most of these gram-positive organisms do not produce immediately life-threatening infections compared with the rapid lethality of many gram-negative infections. Life-threatening bloodstream infections caused by viridans streptococci (especially Streptococcus mitis) are common in many cancer centers and often respond poorly to penicillins and cephalosporins. Risk factors for serious viridans streptococcal infections include aggressive cytoreduction therapy for acute leukemia or allogeneic bone marrow transplantation (especially after high-dose cytosine arabinoside treatment), profound neutropenia, and severe oral mucositis. Other factors
2357
induration and redness and no purulent drainage. Tenderness may be the only finding in perineal and anal infections. The neutropenic patient with a documented infectious cause of fever may be difficult to distinguish from the patient with fever not caused by infection. The performance of a procedure before the onset of fever, the presence of chills, “toxic appearance,” and lack of localized findings do not help determine whether the patient is bacteremic.30 Only 20% of febrile neutropenic patients have a clinical focus of infection identified at presentation, and only 30% of patients have positive blood cultures. Mucositis involving the mouth and other mucous membranes is a painful and debilitating condition that commonly occurs in cancer patients receiving intense chemotherapy. It is a frequent prelude to viridans streptococcal bacteremia, which can produce sudden onset of acute respiratory distress syndrome, a toxic shock–like syndrome, rash, and pneumonia. Diagnostic Strategies. The emergency evaluation of the cancer patient with fever and neutropenia should include a meticulous search for subtle symptoms and signs of inflammation at common sites: oral cavity and pharynx, lower esophagus, lung, skin, perineum including anus, bone marrow aspiration sites, vascular catheter sites, and tissue around the nails.21 In nearly two thirds of patients, the initial evaluation does not identify a focus of infection.22 Two sets of blood cultures should be obtained. If the patient has a central venous catheter, cultures of blood should be obtained from one lumen of a multilumen catheter and from at least one peripheral site. Specimens for culture should also be obtained from any site of inflammation, including inflamed or draining catheter exit sites. Patients with severe mucositis should have herpes simplex cultures performed if not on antiherpes prophylaxis, and they should have a smear for Candida pseudohyphae. Complete blood count, electrolytes, transaminases, blood urea nitrogen, and creatinine should be obtained to plan management and to monitor the occurrence of drug toxicity. Urine culture should be obtained if symptoms or signs of a urinary tract infection are present, if a urinary catheter is in place, or if the urinalysis is abnormal. Examination of the cerebrospinal fluid is not recommended as a routine procedure unless subtle symptoms or signs of meningitis are present. A chest radiograph should be obtained even when symptoms or signs of pneumonia are absent. If the chest radiograph is negative or inconclusive but there is still suspicion for pneumonia, high-resolution CT or thin-section multislice CT scanning of the chest without contrast should be obtained because
Clinical Findings in Neutropenia That May Be Associated with Infection with Table 181-1 Characteristic Specific Pathogens CHARACTERISTIC CLINICAL FINDINGS
SUSPECT PATHOGENS
Ulcerative lesions in the mouth Necrotizing skin lesions Nontender subcutaneous nodules Nontender pink skin papules Black eschar of nose or palate Generalized macular red rash Right lower quadrant abdominal pain, tenderness, distention, bloody diarrhea Perineal pain and tenderness without inflammation or abscess Redness or pain at vascular catheter sites
Viridans streptococci, herpes simplex, Candida, anaerobes Pseudomonas aeruginosa, Aeromonas hydrophila, Aspergillus, Mucor Nocardia, Cryptococcus Candida Aspergillus, Mucor Viridans streptococci Typhlitis (neutropenic enterocolitis) caused by P. aeruginosa, Escherichia coli, Clostridium septicum Gram-negative bacilli, anaerobes Coagulase-negative staphylococci, Corynebacterium, Bacillus species
Chapter 181 / The Immunocompromised Patient
include prophylactic use of trimethoprim-sulfamethoxazole (TMP-SMX) or fluoroquinolones, use of antacids or H2 receptor antagonists, and childhood.26,27 Aspergillus and Candida species are the most common fungi producing infection in cancer patients with fever and neutropenia.23,28,29 They are most likely to develop in neutropenic patients treated with broad-spectrum antimicrobials and whose fever has persisted for more than 7 days. Aspergillus species usually produce necrotizing infections in the lung or sinuses. Pulmonary aspergillosis often presents with pleuritic pain, hemoptysis, and localized wheezing. The chest radiograph demonstrates pleural effusion or focal infiltrates. Computed tomography (CT) is more sensitive in detecting pulmonary infiltrates compatible with aspergillosis, and it may demonstrate a distinct halo of low attenuation surrounding a pulmonary infiltrate. This pattern is highly suggestive of invasive aspergillosis, although mucormycosis and other disorders may mimic the halo. Invasive aspergillosis originating in the paranasal sinuses may extend to the surrounding bone and brain. Often, an initial red-purplish lesion on the nasal turbinate or palate turns pale and then black as vascular invasion produces infarction of the mucosa and bone. The black eschar on the nose or palate is easily misdiagnosed as dried blood. Patients presenting with head or facial pain or swelling, or proptosis, should be rapidly evaluated for invasive aspergillosis and mucormycosis. Candida species produce infections of the skin, oral cavity, and esophagus as well as fungemia. The sudden onset of generalized rash consisting of pinkish-purple, nontender subcutaneous nodules is characteristic of candidemia. Less common fungi producing infection in these patients include Mucor and Rhizopus species (necrotizing pneumonia and sinusitis), Trichosporon species (pneumonia and fungemia), and Pseudallescheria boydii (soft tissue infection). Other than Candida species, these fungi are rarely found in blood cultures. Specific diagnosis requires biopsy. Clinical Features. Certain clinical findings are characteristic for specific pathogens (Table 181-1). Noninfectious causes of fever also need to be considered, such as drug toxicity, drug allergy, transfusion reactions, and pulmonary emboli.17 Fever is frequently the only sign of infection because these patients are unable to mount a full inflammatory response at a site of infection.23 Usual symptoms and signs of infection may not be present, especially when the neutrophil count is less than 100 cells/mL. When pneumonia develops, purulent sputum may be absent, and the initial chest radiograph may not show an infiltrate. Pyuria may be absent in the presence of urinary tract infection. Areas of cellulitis may have diminished or absent
PART V ■ Special Populations / Section Four • The Patient with Compromised Immune Function
2358
pneumonia is often detected by chest CT in febrile neutropenic patients with normal findings on chest radiograph.31 CT evaluation of the sinuses should be performed if facial pain or swelling is present. In patients with abdominal pain and tenderness, CT scanning of the abdomen is useful for diagnosing neutropenic enterocolitis (“typhlitis”), a necrotizing infection of the bowel wall that usually affects the cecum. This is more commonly seen in acute leukemia and is not generally treated surgically. Ultrasonography over a subcutaneous tunneled catheter tract and its vein of insertion may reveal the presence of an abscess or infected thrombus.32 In patients with diarrhea, stool testing for C. difficile toxin as well as culture for routine bacterial pathogens should be obtained. C. difficile colitis occasionally produces abdominal pain in the absence of diarrhea. Cryptosporidium parvum,
a protozoan, may cause profuse, watery diarrhea and is detected by acid-fast staining or immunofluorescence testing of stool. Management of Febrile Neutropenia Antibiotic Therapy. Broad-spectrum antimicrobial therapy should be initiated promptly in the febrile neutropenic patient if the neutrophil count is less than 500 cells/mL or if the neutrophil count is 500 to 1000 cells/mL and is expected to drop.19 Moreover, even afebrile neutropenic patients who have symptoms and signs (e.g., abdominal pain and tenderness) compatible with an infection should be treated empirically (Table 181-2). Use of a single antimicrobial agent is preferred in most patients because there is no conclusive evidence of a benefit from multiple drugs.33,34 Factors that must be considered when
Table 181-2 Selected Antimicrobial Agents Useful in the Immunocompromised Patient DOSAGE DRUG
Aminoglycosides Gentamicin
Tobramycin Amikacin
ADULT
CHILD (AGE > 28 DAYS)
PRECAUTIONS/COMMENTS
2 mg/kg loading dose, then 5 mg/kg/day IV q8–12h, or 5–7 mg/kg IV once daily Same as gentamicin 10 mg/kg/day loading dose, then 15 mg/kg/day IV q12h or once daily
Same as adult
Decrease maintenance dose in elderly or with renal dysfunction
Same as adult Same as adult
Extended-Spectrum Penicillins/β-Lactamase Inhibitors 4.5 g IV q6h Piperacillin/tazobactam
240–400 mg/kg/day IV q6h
For Pseudomonas aeruginosa, administer with aminoglycoside
Cephalosporins Ceftazidime Cefepime
1–2 g IV q6–8h 1–2 g IV q8h
150 mg/kg/day IV q8h 150 mg/kg/day IV q8h
Both active against P. aeruginosa, but cefepime more active than ceftazidime against gram-positive organisms and some resistant gramnegative bacilli
Carbapenems Imipenem/cilastatin
0.5–1 g IV q6h
60–100 mg/kg/day IV q6h
Meropenem
0.5–1 g IV q8h
60–120 mg/kg/day IV qh8
Adjust dose in elderly or with renal dysfunction. Seizures associated with imipenem. May be cross-allergenic with penicillin. Broad-spectrum against gram-positive and gram-negative organisms, including P. aeroginosa and anaerobes.
Other Aztreonam
1–2 g IV q8h
120 mg/kg/day IV q6h
Vancomycin
15 mg/kg IV q12h
40 mg/kg/day IV q6–12h
Amphotericin B
0.5–1.5 mg/kg/day IV once daily
Same as adult
5 mg/kg IV q8h or 400 mg PO three times a day
250 mg/m2 IV q8h or 15 mg/kg/ day PO q4h
800 mg PO five times a day 10 mg/kg IV q8h 10 mg/kg IV q8h
20 mg/kg PO four times a day 500 mg/m2 IV q8h 500 mg/m2 IV q8h
Acyclovir Herpes simplex, mucocutaneous* Herpes zoster Not severe† Severe Primary varicella
Active against gram-negative bacilli including P. aeruginosa Not active against gram-positive organisms Safe in penicillin-allergic patient Infuse over 2 hr (flushing, hypotension with rapid infusion) Refer to infectious disease or pharmacology text Infuse over 1 hr
Elderly: 7.5 mg/kg IV q8h
*Alternative for herpes simplex: famciclovir 250 mg three times a day, or valacyclovir 1 g two times a day PO. † Alternative for herpes zoster: famciclovir 500 mg three times a day, or valacyclovir 1 g three times a day PO q8–12h.
2359
advantage when vancomycin is in the initial therapy for all neutropenic patients, even those with indwelling catheters. Because most infections with gram-positive bacteria are indolent, vancomycin therapy can be safely delayed for 24 to 48 hours in most patients until a vancomycin-requiring gram-positive infection is identified.41 Indications for initial empirical vancomycin therapy include serious catheter-related infections, known colonization with penicillin-resistant pneumococci or methicillin-resistant S. aureus, and positive blood culture for gram-positive organisms before final identification and susceptibility testing. Other indications include shock, severe mucositis, prior fluoroquinolone prophylaxis, and institutions in which methicillinresistant S. aureus, vancomycin-susceptible enterococci, and S. mitis are frequent pathogens. Amphotericin B (and its lipid formulations) is the drug of choice for treatment of invasive fungal infections in patients with neutropenia.19,39,42 Up to one third of febrile neutropenic patients not responding to 1 week of antibiotics have systemic fungal infections, usually Candida or Aspergillus. Antifungal agents, such as caspofungin, voriconazole, or posaconazole, may be indicated in selected cases. Empirical use of fluconazole is not recommended because of lack of activity against Aspergillus and some Candida species. Cell Stimulation Therapy. To prevent and treat neutropenia, some centers routinely use human recombinant hematopoietic or colony-stimulating growth factors (G-CSF-filgrastim, pegfilgrastim; GM-CSF-sargramostim) to stimulate the proliferation and maturation of bone marrow progenitor cells and increase the number and function of these committed cell populations. Although safe and well tolerated, they are very expensive. Treatment with these agents after chemotherapy may shorten the hospital stay and the duration of fever, but no good evidence exists that they prolong survival or affect the frequency of severe infections. Many authorities recommend that use of these agents should be limited to high-risk neutropenic patients, such as those with severe sepsis, multiorgan failure, or recurrent febrile neutropenia and elderly patients.19,43,44 Risk Assessment for the Patient with Febrile Neutropenia, Including the Concept of Brief Observation and Early Discharge of the Low-Risk Patient. Febrile neutropenic cancer patients can be classified into high-risk and low-risk groups.19,45-47 Factors associated with high-risk patients include (1) status as inpatient when fever and neutropenia develop; (2) the presence of comorbid medical conditions; (3) uncontrolled cancer; (4) acute leukemia; (5) hemodynamically unstable; (6) evidence of organ failure; (7) the presence of pneumonia, severe soft tissue infection, infection of a central line, abdominal pain, neurologic or mental status abnormalities; and (8) neutropenia expected to last more than 10 days. These patients should always be treated in the hospital with intravenous antibiotics. Low-risk patients generally have an excellent outcome with therapy and are clinically stable outpatients with solid tumors, lymphomas, or chronic leukemia who lack any of the high-risk factors noted previously and who are not on fluoroquinolone prophylaxis. Low-risk patients with fever and neutropenia may be treated in the hospital with oral antibiotics, such as ciprofloxacin plus amoxicillin-clavulanate (or ciprofloxacin plus clindamycin in penicillin-allergic patients).47-50 Furthermore, oral antibiotic therapy with early discharge is safe and effective in carefully selected low-risk patients.51-53 Hospitalization exposes low-risk patients to potential iatrogenic complications and antimicrobial-resistant nosocomial pathogens, and early discharge followed by outpatient treatment allows an improved quality of life. Low-risk patients may be hospitalized initially, stabilized over 12 to 48 hours, and then discharged to continue parenteral or oral antibiotics. Some
Chapter 181 / The Immunocompromised Patient
choosing an antimicrobial agent include renal and hepatic function, drug allergies, suspected site of infection or organism, and cost. Antimicrobial resistance varies widely, and in the absence of a written hospital-specific protocol, immediate consultation with an oncologist or an infectious diseases specialist may be of great assistance. Monotherapy with intravenous cefepime, ceftazidime, imipenem, meropenem, or piperacillin-tazobactam is preferred, with an aminoglycoside (gentamicin, tobramycin, or amikacin) added for the more seriously ill patient.19,23,35,36 Monotherapy, without an aminoglycoside, may be advantageous in the patient with mild to moderate renal dysfunction or for patients on nephrotoxic agents such as cisplatin, cyclosporine, or amphotericin B. Ceftazidime-resistant gram-negative bacilli are common pathogens at some centers. In addition, ceftazidime is the least active against gram-positive organisms, compared with cefepime, cefotaxime, or ceftriaxone. Cefepime, a broad-spectrum cephalosporin with excellent activity against both gram-positive and gram-negative organisms including P. aeruginosa, is now preferred at many centers. The carbapenems imipenem and meropenem provide excellent activity against gram-negative organisms (including P. aeruginosa), gram-positive organisms, and anaerobic bacteria. None of the antimicrobial agents previously listed is active against vancomycin-resistant Enterococcus species or methicillinresistant staphylococci. For patients who are allergic to β-lactam antibiotics (e.g., penicillins, cephalosporins, imipenem, and meropenem), coverage of gram-negative bacilli, including P. aeruginosa, can be provided by aztreonam. Because aztreonam is not active against gram-positive or anaerobic bacteria, it should be combined with an antimicrobial such as vancomycin. If anaerobes are suspected (i.e., oral, abdominal, or perianal infection) in the β-lactam allergic patient or in the patient receiving cephalosporin monotherapy, an antianaerobic drug such as clindamycin or metronidazole should be administered. Empirical treatment with intravenous fluoroquinolones is not recommended in the febrile neutropenic cancer patient because of frequent prophylactic use of these agents in the cancer patient, risk for rapid emergence of resistance in gram-negative bacilli, and predisposition to C. difficile infection.19,37,38 When a focus of infection is identified, empirical therapy should cover the most likely pathogens causing infections at the site. For example, patients with pneumonia may need coverage for Legionella (fluoroquinolone, azithromycin, doxycycline, or erythromycin), Pneumocystis (TMP-SMX), or fungi (amphotericin B) in addition to standard antibacterial coverage. Agents effective against anaerobes (clindamycin, metronidazole, imipenem, meropenem, and piperacillin-tazobactam) should be considered for patients with perianal or oral infection and those with abdominal pain, who may have appendicitis, diverticulitis, or typhlitis (neutropenic enterocolitis). Acyclovir should be considered for patients with ulcerative or vesicular lesions who may have herpes simplex or varicella zoster virus infections. Ganciclovir may be required to treat cytomegalovirus infection, which is rare in febrile neutropenic cancer patients. Foscarnet may be needed for acyclovirresistant herpes simplex or ganciclovir-resistant cytomegalovirus infections. In patients with severe mucositis and febrile neutropenia, a carbapenem or extended-spectrum penicillin is preferred for empirical treatment rather than a highergeneration cephalosporin because of superior efficacy against viridans streptococci. Routine empirical use of vancomycin for the febrile neutropenic cancer patient is not recommended because of concern about the development of vancomycin-resistant organisms.26,39,40 Randomized clinical trials show no survival
PART V ■ Special Populations / Section Four • The Patient with Compromised Immune Function
2360
authorities recommend a period of observation less than 12 hours, which may allow discharge from the emergency department, an observation unit, or physician’s office.47 The patient should be observed for at least 2 hours after administration of the first dose of antibiotics and only discharged after consultation with the patient’s oncologist. Patients sent home need adequate instructions, family support, and easy access to a hospital in case of emergency.47,54,55
Non-neutropenic Conditions The Solid Cancer Patient without Neutropenia. Most solid cancer patients who develop fever and infection are not neutropenic. Infections often occur after surgical procedures and may include wound infection, deep abscess, or perforated viscus. Infections may be associated with central venous catheters, urinary catheters, stents, and prosthetic devices. In addition, solid tumor patients with large tumor lesions may develop obstructive infections (of bronchus, bile duct, or ureter). The spectrum of microorganisms includes a wide variety of indigenous organisms (bacterial, fungal, and viral) as well as nosocomial multiantibiotic-resistant pathogens. Prompt initiation of antimicrobial therapy in the febrile nonneutropenic solid cancer patient is not always indicated. Rapid surgical intervention may be more important than the urgent initiation of empirical antibiotics. In febrile non-neutropenic cancer patients who are not ill-appearing and have no identified focus of infection, it may be appropriate to obtain cultures and observe the patients. After consultation with an oncologist, some of these patients can be discharged home from the emergency department with close follow-up. Indications for urgent antibiotics in these patients include signs of sepsis, mental status changes, lactic acidosis, shock, abdominal pain, history of splenectomy, or identification of a focal site of infection.56,57 Impaired Cell-Mediated Immunity. The T cell defects resulting from impaired CMI in cancer patients usually result from cancer chemotherapy or corticosteroid treatment. The cancer impairs CMI in patients with Hodgkin’s disease, nonHodgkin’s lymphoma, and hairy cell leukemia. Bacterial Infections. Listeria monocytogenes is one of the more common bacterial organisms infecting cancer patients with impaired CMI.58-61 Listeria infection is also seen in patients with organ transplants, diabetes, cirrhosis, and AIDS and in those receiving high-dose corticosteroids. No early characteristics distinguish Listeria infection from bacteremias caused by other organisms. Meningitis, which may be accompanied by cerebritis or brain abscess, is the most common focus of infection and may present with personality changes or focal neurologic signs.62 Cerebrospinal fluid examination frequently does not reveal the organism on Gram’s stain, but protein is elevated and pleocytosis is present. Treatment should be with ampicillin and gentamicin. TMP-SMX is the alternative drug for patients with penicillin allergy. Vancomycin is not effective in treating Listeria infections even when showing in vitro susceptibility. Cephalosporins, such as ceftriaxone and cefotaxime, are not active against Listeria. Infections caused by Salmonella species are common in patients with impaired CMI and usually present with fever with or without enteritis.63 Bacteremia can result in infection of bones, joints, central nervous system, and endovascular devices. Multidrug-resistant Salmonella species are increasing. Treatment usually includes a third-generation cephalosporin or a fluoroquinolone because many isolates have become resistant to ampicillin and TMP-SMX.64 Patients with solid tumors, lymphoma, and leukemia (especially hairy cell leukemia) are at increased risk for pneumonia
from Legionella species, with the highest risk in cancer patients receiving high-dose corticosteroids.65,66 Non-pneumophila species of Legionella (e.g., Legionella micdadei and Legionella bozemanii) are particularly common in these patients.67 Clinical and radiographic manifestations of Legionella infection in the immunocompromised patient often differ from those in the immunocompetent host. For example, pleuritic chest pain may be a prominent symptom in the former and may mimic pulmonary embolism. These patients can have fever without any other symptoms of pneumonia despite the presence of radiographic pulmonary infiltrates. In addition, the chest radiograph may reveal an expanding pulmonary nodule or cavitation of a nodule or infiltrate rather than the usual lower lobe alveolar filling defects. Hyponatremia (serum sodium 38° C), shortness of breath, hypoxia, hypotension, poorly controlled hypertension, or new dysrhythmia should be admitted. Chest pain is rarely related to cardiac ischemia because the denervated heart is incapable of producing angina.29 Accelerated atherosclerosis of the graft vessels is the hallmark of chronic rejection. Ischemia is manifested as CHF, ventricular dysrhythmias, hypotension, syncope, or sudden death.30,31 CMV infection also appears to be a risk factor for accelerated atherosclerosis.9,18,32
Drug Toxicity Lifelong immunosuppression is required in these patients. Most centers use a three-drug regimen of cyclosporine, prednisone, and azathioprine. Each drug has potential for toxicity, and the combination of cyclosporine and prednisone aggravates the hyperlipidemias common to many of these patients.16,33
Rarely, cyclosporine toxicity can result in a neurotoxic syndrome of seizures, confusion, cortical blindness, and quadriplegia, possibly progressing to coma. Seizure management is with standard anticonvulsants.34-36
Rejection Acute rejection occurs in 75 to 85% of patients within the first 3 months. Its manifestations may be subtle with cyclosporine included in the immunosuppressive regimen. Previously, rejection was obvious, with decreased QRS voltage, a new S3 heart sound, or new-onset CHF or atrial dysrhythmias. These features are now present only during episodes of severe rejection, and the diagnosis of rejection is made by endomyocardial biopsy showing lymphocyte infiltration or myocyte necrosis.1,16 Because most episodes of early or mild rejection are asymptomatic, frequent biopsies are performed to monitor the success of the immunosuppressive regimen.37,38 Acute rejection is treated with increased doses of corti costeroids (methylprednisolone, 500–1000 mg/day) and cyclosporine or with OKT3 or antithymocyte globulin. Immunosuppression is continued lifelong, with endocardial biopsies at least every 3 to 6 months.38,39 The transplanted heart maintains a rate of 100 to 110 beats/ minute without parasympathetic tone. The electrocardiogram
2371
aVR
VI
V4
II
aVL
V2
V5
III
aVF
V3
V6
VI
II
V5
Figure 182-5. A 68-year-old male status postheterotopic heart transplant; native heart rate of 55 and donor heart rate of 75 beats per minute.
(ECG) typically demonstrates two P waves. One wave is from the native sinus node in the posterior right atrium, which is often left in place. The second P wave is from the donor sinoatrial node, which should conduct to the ventricles as usual with a normal PR interval. In the rare heterotopic heart transplant, bizarre ECGs may be seen that do not conform to normal standards (Fig. 182-5). The transplanted heart rate can increase with exercise or stress through the effects of endogenous catecholamines, up to 70% maximum for age. Exogenous pressor drugs work well in the transplanted heart. Up-regulation of beta-adrenergic receptors appears to occur in the graft, with a slightly enhanced response to norepinephrine and isoproterenol.34,39-41 Typical antihypertensive agents can be used as in the nontransplant patient. Atropine is ineffective due to vagal denervation.
Infection Infection in heart transplant patients has many causes. One fourth of deaths after transplant result from infection. The most vulnerable period is the first 3 months, when immunosuppression is maximal. Although one third of patients will have a major infection develop in the first year, life-threatening infection after 1 year is rare.1,18 In the first month, nosocomial infection predominates, with the usual gram-positive and gram-negative bacteria. After the first 3 months, patients experience an overall 20% per year rate of infection. The most common skin infection is herpes zoster, which is treated with high-dose acyclovir. Nausea, vomiting, or diarrhea should prompt a search for CMV by culture and serologic testing.2,16,37 All heart transplant patients with fever should be aggressively evaluated unless a minor source of infection such as an upper respiratory infection is obvious. The diagnostic workup in the ED can include blood and urine cultures, computed tomography (CT) scan, lumbar puncture, and bronchoscopy. In addition, complete blood count, glucose, serum chemistries, blood urea nitrogen and creatinine, chest radiograph, and ECG are recommended. In one series of 131 ED visits by such
patients, 23 patients were admitted with a diagnosis of “rule out sepsis,” and in 12 (52%) an infecting organism was identified.16,34,38 Any new headache with or without visual changes may be the first symptom of meningitis or brain abscess, and CT scan of the head and a lumbar puncture should be obtained. Fever, lethargy, headache, altered mental status, or seizures are presenting signs of Listeria, cryptococcal meningitis, Toxoplasma gondii, or brain abscesses from Nocardia or Aspergillus. A more definitive diagnosis can often be made through biopsy of a mass lesion or drainage of an abscess. This often obviates the need for lumbar puncture. Aseptic meningitis occurs in 10 to 14% of patients treated with OKT3 approximately 6 to 10 days after therapy.1,18 Due to the attendant risk of endocarditis, antibiotic prophylaxis should be provided for invasive procedures likely to cause bacteremia, such as abscess drainage and urethral catheterization. Endotracheal intubation requires no such prophylaxis. VZV immune globulin is recommended as soon as possible when seronegative patients are exposed to chickenpox or herpes zoster.2,16
Liver Transplant Liver transplants have 1- and 3-year survival rates of 86 and 76%, respectively.28 Complications are common, and loss of function of the graft organ is rapidly life-threatening.
Anatomic Considerations The typical liver transplant is connected to its host via five anastomoses (Fig. 182-6). The vessels are connected first, allowing organ reperfusion. The biliary system is then reconstructed and often stented with a T tube to prevent stenosis. In the early post-transplant period, T tubes may be replaced if restenosis develops. The most common vascular complication is hepatic artery thrombosis, which most commonly occurs early in the post-transplant period. Severe graft dysfunction is invariable and mortality approaches 75%.42
Chapter 182 / The Solid Organ Transplant Patient
I
PART V ■ Special Populations / Section Five • The Patient with an Organ Transplant
2372 Suprahepatic vena cava anastomosis Portal vein anastomosis Hepatic artery anastomosis Gallbladder fossa
Figure 182-6. Typical anastomoses of an orthotopic liver transplant. (From Powelson JA, Cosimis AB: Liver transplantation. In Ginnes LG, Cosimi AB, Morris PJ [eds]: Transplantation. Malden, Mass, Blackwell Science, 1999, p 352.)
Common bile duct anastomosis Infrahepatic vena cava anastomosis
T-tube
Rejection
Infection
Liver transplant rejection is the norm despite immunosuppressive therapy. Rejection often begins 1 or 2 weeks after surgery, with fever, right upper quadrant pain, and elevated bilirubin and transaminases. Leukocytosis may occur but is nonspecific. Related conditions that simulate graft rejection are mechanical biliary obstruction, primary nonfunctioning graft, ischemia from thrombosis of vascular anastomoses, viral infections, drug toxicity, and recurrent primary disease.43-46 As soon as transplant rejection is suspected, treatment with high-dose methylprednisolone should begin. Hospitalization is routine. If this fails to diminish the rejection episode, OKT3 monoclonal antibodies are again used in addition to polyclonal antilymphocyte globulin.43
The classic example of bacterial infection in transplanted organs is pyelonephritis in renal allografts. It often occurs in the first month after surgery during high-dose immunosuppression. During this period, an increased risk of sepsis exists. Gram-positive organisms associated with wound infections and gram-negative organisms are the predominant causes. Pyelonephritis occurs in 35% of patients during the first 4 months but can be prevented with prophylactic antibiotics. All patients with transplant who have pyelonephritis are admitted for aggressive antibacterial therapy.47,48 Hepatitis C virus (HCV) infection is the most common cause of hepatitis in renal transplant patients. It may be transferred to the recipient by blood transfusion or through the donor organ because pretransplant antibody testing is a poor predictor of latent HCV infection. Liver disease develops in approximately 50% of seropositive patients, occurring approximately 4 months after transplantation. Most patients develop chronic hepatitis with accompanying immune defects and susceptibility to sepsis and spontaneous bacterial peritonitis. Pretransplant screening is the only avenue of prevention.1,48
Infection After the first postoperative month, opportunistic infections replace the common postsurgical complications. Viral (CMV and HSV), fungal (Aspergillus, Candida, and Cryptococcus), protozoan (Pneumocystis and Toxoplasma), and unusual bacterial (Nocardia, Legionella, and Listeria) infections occur. In addition, ascending cholangitis due to colonization of biliary stents with staphylococcal species, enterococci, and gram-negative organisms is common. Subsequent injury to the graft during biopsy or cholangiogram can cause cholangitis or liver abscess.
Kidney (Renal) Transplant With survival rates of 96% at 1 year and 91% at 3 years, renal transplants are highly successful.28 Injury of the transplanted kidney is rare, despite its location in the retroperitoneal area of the anterior pelvis, where it may be at risk from direct blows as well as seat belt injuries.27
Rejection Early kidney transplant rejection is mediated through T lymphocytes against antigen donor tissues, including cytotoxic CD8 and CD4 cells. In addition, B lymphocytes, natural killer cells, and macrophages infiltrate the foreign tissues. The B lymphocytes manufacture specific antibodies, which results in microvascular lesions impairing perfusion. Chronic transplant rejection occurs after several years of adequate function and is a result of nephrosclerosis. This process involves proliferation of the vascular intima of renal vessels with marked decrease in the lumen size. Systemic
Lung Transplant The lung may be transplanted alone or in combination with the heart, with 1-year survival rates of 76 and 56%, respectively.28 Lung transplants are usually done unilaterally, except for cystic fibrosis. Unequal lung sounds are to be expected. There is no published experience with lung injury following transplant. Comparison of chest radiography with preinjury studies is critical in the evaluation of trauma. Chest tube placement on the transplanted side may be difficult because of adhesions and loculations.27
Rejection Although still rare, lung transplantation is increasing in frequency; more than 800 are performed annually in the United States.28 Most patients develop early rejection, and 25 to 40% have chronic rejection. An episode of acute rejection can occur as early as a few days after transplantation or as late as several years. Clinically, the patient presents with cough, dyspnea, and fever. Rales and rhonchi are heard, with deterioration in oxygenation and pulmonary function. Early rejection is often accompanied by infiltrates on chest radiograph. When rejection occurs more than a month after transplantation, 75% of radiographs are normal or unchanged. The diagnosis of rejection is made by transbronchial biopsy showing lymphocytic infiltration.18,49-51 Suspected episodes of acute rejection are treated with highdose methylprednisolone (500–1000 mg/day). This is successful in reversing most episodes, but OKT3 can be used in refractory cases.50 Chronic lung transplant rejection is a leading cause of late morbidity and mortality. Antecedent acute rejection and CMV pneumonia are risk factors. Pathologically, vascular sclerosis and progressive limitation to airflow occur from obliterative bronchiolitis. Rejection can occur several years after transplantation, but the mean time to onset is 8 to 12 months. Clinically, this rejection mimics an upper respiratory infection or bronchitis. If dyspnea is a component of the presenting complaint, a search for transplant rejection should be initiated.50,52 Chronic rejection is treated with high-dose methylprednisolone (500–1000 mg/day), with the first dose begun in the ED. Antilymphocyte antibodies are used as well, but relapse of the rejection episode is common.
Infection Transplanted lungs are highly susceptible to pneumonia because they can be colonized by bacteria during the ventila-
2373
tor stage of the brain-dead donor. After transplant, diminished mucociliary clearance, decreased cough reflex due to denervation, and defective function of alveolar macrophages are present. The most common infections are caused by gramnegative bacteria such as Pseudomonas and by Staphylococcus aureus. Antibiotic therapy should be aggressively directed toward any pathogenic bacteria cultured from the tracheobronchial tree. Community-acquired pneumonia must be considered as well. Pneumocystis pneumonia is uncommon because of routine prophylaxis with TMP-SMX.1,50 CMV pneumonia is the most common opportunistic pulmonary infection after lung transplantation. Patients are at highest risk between 3 weeks and 4 months. Clinically, CMV infection closely resembles transplant rejection. Tissue biopsy and viral culture are required to differentiate the two entities. Treatment with ganciclovir is effective. Colonization by Candida is common, but not invasive disease. Aspergillus provides the most significant fungal threat to the transplanted lung. Reactivation of tuberculosis is rare.50,51
Pancreas Transplant The pancreas may be transplanted singly or in combination with a kidney. This is typically secondary to diabetes. Pancreatic transplants have a high complication rate, with 1-year graft survival rates as low as 72%. Because the exocrine functions of the allograft pancreas are usually drained into the bladder, genitourinary complaints are also common. Duodenocystostomy fistula may form in the early post-transplant period. The clinical findings are abdominal pain, tenderness, hyperamylasemia, leukocytosis, and elevated serum creatinine. Other types of pancreatic transplant complications include urinary tract infections, hematuria, reflux pancreatitis, rejection, and pancreatic graft thrombosis.53-55 Anatomic considerations are important when these patients suffer major trauma because pancreas transplants are placed in the pelvis overlying the iliac vessels.27 Exocrine secretions are drained into the bladder for excretion, and patients have a chronic non–anion gap acidosis through loss of bicarbonate into the bladder. This should not be confused with lactic acidosis. These trauma patients should not require exogenous insulin, unless the graft is injured. Positive amylase on peritoneal lavage (CT is recommended if the patient is stable) can result from either native organ or graft trauma or from a ruptured bladder. CT scanning should be done with rectal contrast, in addition to IV and oral, to better define the native and transplanted pelvic organs.27 Islet cell transplantation is under investigation as a treatment for diabetes mellitus. If successful, this procedure may negate the need for future solid organ pancreas transplants and the resultant complications.
■ PSYCHOLOGICAL ASPECTS OF ORGAN TRANSPLANTATION The psychological issues that surface while under consideration for or after reception of an organ transplant impact many patients. Organ transplantation is no longer considered experimental, and it is a common therapeutic option for end-stage organ failure. Strong emotions naturally occur and should be acknowledged in donors, recipients, and transplanters. Transplant programs widely employ psychosocial selection criteria. Generally, the cardiac programs have the most stringent criteria. The side effects of lifelong immune suppression or steroid withdrawal can include anxiety, depression, and insomnia. On the other hand, successful transplantation often improves psychological well-being. Transplant social workers
Chapter 182 / The Solid Organ Transplant Patient
hypertension ensues as the graft fails from ischemia, with resultant tubular and glomerular atrophy. With cadaveric kidney transplants, histocompatibility differences are almost universal and routinely require long-term immunosuppressive therapy. This is accomplished with a combination of azathioprine, prednisone, and cyclosporine. On diagnosis of acute rejection, high-dose methylprednisolone (500–1000 mg) is begun daily and continued for 3 days. After 6 to 12 months, many patients only need lower doses (10–20 mg/day).47,48 Clinically, renal graft rejection presents as fever, swelling and tenderness over the allograft, and decreased urine output. A subtle rise in serum creatinine should prompt great concern. Renal ultrasound should be performed to rule out obstruction, abscess, vascular thrombosis, and perirenal collections of blood, pus, or lymph.48 Early consultation with the nephrologist is prudent.
PART V ■ Special Populations / Section Five • The Patient with an Organ Transplant
2374
can provide awareness of and access to the social network that should surround each transplant recipient. Long-term compliance with all aspects of treatment will minimize graft rejection.56
■ DISPOSITION Patients with solid organ transplants presenting to the ED have a much higher than average rate of hospitalization.28,46
The insidious nature of the diseases affecting this immunosuppressed population mandates a thorough approach to evaluation. If organ rejection, infection, or drug toxicity is evident, local transplantation specialists should be consulted. Physicians without significant transplant experience should contact the patient’s transplant center to obtain consultation and coordinate follow-up care. Patients who are discharged require careful instructions and close follow-up.
KEY CONCEPTS ■
The possibility of organ rejection, infection, or drug toxicity should be considered in all organ transplant patients who present to the ED despite subtle presentations. ■ A patient’s inability to take oral immunosuppressants for even a single day should be considered an emergency condition.
The references for this chapter can be found online by accessing the accompanying Expert Consult website.
■
When prescribing care in the ED, the physician must be careful to avoid drug interactions and toxicity. ■ Infections that occur from 1 to 6 months after transplantation are generally immunomodulating viruses such as CMV or opportunistic infections. ■ Close family members of transplant patients should not receive live viral antigen vaccines due to the risk of transmission.
•
Section Six The Alcoholic and Substance Abuse Patient Chapter 183
Alcohol-Related Disease
John T. Finnell and David B. McMicken
■ PERSPECTIVE Epidemiology The disastrous effects and widespread incidence of alcoholism are well known to the emergency physician. Motor vehicle collisions, drowning, suicides, homicides, divorce, violent crime, child abuse, unemployment, and disruption of the family are often either directly or indirectly associated with excessive alcohol consumption. The tragic effects of alcohol not only affect the individual drinker but also have far-reaching implications for the family, community, and workplace. There are an estimated 68.8 million emergency department (ED) visits, with a rate of 28.7 per 1000 U.S. population.1 The 5-year mortality rate among alcohol-intoxicated ED patients was 2.4 times that of a comparison group in one study.2 A simple, rapid, and respectful screening test for alcoholism is the four CAGE questions: Have you ever felt ■ ■ ■ ■
The need to Cut down on your drinking? Annoyed by criticism of your drinking? Guilty about your drinking? The need to drink an Eye opener in the morning?
Positive answers to two or more of these questions are sufficient to identify individuals who require more intensive evaluation.3 Also, a positive answer to the question, “Have you ever had a drinking problem?” plus evidence of alcohol consumption in the previous 24 hours provides greater than 90% sensitivity and specificity as a screening tool for identifying alcoholism. Alcohol is the most common recreational drug taken by Americans, and per capita consumption is increasing. Alcoholism affects approximately 12% of individuals in the United States during a lifetime.4 Alcoholism permeates all levels of society and is the leading cause of preventable mortality and morbidity, with a cost to the nation estimated to be greater than $185 billion annually.5 An estimated 18 million alcoholics live in the United States. With more than 100,000 alcoholrelated deaths occurring each year, alcohol is the third leading cause of preventable death in the United States.6
Definition and Natural History A precise definition of alcoholism is difficult. A proposed definition encompassing the features of alcoholism is “a primary chronic disease with genetic, psychosocial, and environmental factors influencing its development and manifestations.” The disease is often progressive and fatal. It is characterized by
impaired control over drinking, preoccupation with and use of alcohol despite adverse consequences, and distortions in thinking, most notably denial. Each of these symptoms may be periodic or continuous.7 Alcoholism is present when drinking adversely affects an individual’s physical health, ability to function in society, or interpersonal relationships. Certainly, the patient who has a dependence on ethanol can be labeled “alcoholic.” Hazardous or “at-risk” drinking is defined by the National Institute on Alcohol Abuse and Alcoholism as follows: Men: more than 14 drinks per week or more than 4 drinks per occasion Women: more than 7 drinks per week or more than 3 drinks per occasion Age older than 65: more than 7 drinks per week or more than 1 drink per occasion Harmful drinkers present with negative consequences related to alcohol. The cause of alcoholism is not completely understood but appears to be a complex interaction between biologic and environmental factors. Genetic variability of enzymes for alcohol metabolism may be a risk factor supported by family, twin, and adoption studies.8 A large study, entitled the Collaborative Studies on Genetics of Alcoholism (COGA) is now identifying additional gene loci. The natural history of alcoholism is variable, and it may appear in any patient despite age or social status. The age of onset of alcoholism continues to decrease. Up to 6% of high school seniors drink daily, and it is not unusual to see children younger than 16 years of age who have already graduated from an alcohol detoxification program.9 Many individuals also begin drinking heavily after age 60. The fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) has two categories for substance disorders that include alcohol abuse. It lists criteria for substance abuse and substance dependence.10 The chronic substance abuse of alcohol eventually leads to acquired tolerance, a condition in which increasingly larger doses of alcohol are required for the same effect. An inborn tolerance also exists. There is a wide variance in abnormal behavior independent of the patient’s drinking experience. Continued alcohol abuse progresses to substance dependence, defined in DSM-IV as a maladaptive pattern of substance use leading to clinically significant impairment, as manifested by three or more of the following occurring in the same 12-month period: 2375
PART V ■ Special Populations / Section Six • The Alcoholic and Substance Abuse Patient
2376
1. Physiologic dependence, as evidenced by tolerance or withdrawal 2. Alcohol taken in larger amounts or over a longer period than was intended 3. Persistent desire or unsuccessful efforts to control alcohol consumption 4. Great amount of time spent in activities necessary to obtain alcohol or recover from its effects 5. Important social, occupational, or recreational activities forsaken for sustained alcohol use 6. Continued alcohol use despite knowledge of having a persistent or recurrent physical or psychological problem that is likely to have been caused or exacerbated by alcohol
■ PRINCIPLES OF DISEASE: METABOLISM OF ALCOHOL Ethanol is rapidly absorbed from the stomach and small intestine. It is distributed uniformly to all organ systems, including the placenta. The oxidation of alcohol is a complex process involving three enzyme systems, all contained in the hepatocyte. The pharmacokinetic properties of alcohol metabolism are well-known. The class I alcohol dehydrogenase (ADH) isoenzymes, ADH1A, ADH1B, and ADH1C, oxidize ethanol. ADH1B and ADH1C have polymorphic properties with distinct kinetic properties. At the ADH3 locus, two alleles account for pharmacokinetic differences of 2.5-fold in maximum velocity of ethanol oxidation.11 dehydrogenase Ethanol Alcohol → NAD→ NADH dehydrogenase Acetaldehyde Alcohol → NAD→ NADH acid Acetyl coenzyme A Citric → CO2 + H2O Cycle
An alternative pathway, the microsomal ethanol-oxidizing system (MEOS), can be induced by chronic alcohol exposure. The primary component of MEOS is the molecule cytochrome P450, which exists in several variants. The variant most important for alcohol metabolism is cytochrome P450 2E1 (CYP2E1). Many effects of alcoholism are produced by the toxic byproducts (hydrogen and acetaldehyde), the acceleration of metabolism of other drugs, and activation of hepatotoxic compounds by these metabolic pathways. Although the liver is the major site of ethanol metabolism, other tissues contribute to its metabolism. ADH is found in the gastric mucosa, but the gastric metabolism of alcohol is decreased in women and those of Asian decent. This increased bioavailability of ethanol or decreased first-pass metabolism may explain the enhanced vulnerability of women to acute and chronic complications of alcohol. Studies have shown two alcohol elimination curves. The alcohol elimination rate approximates zero-order kinetics (constant rate) for lower ethanol levels and first-order kinetics (the amount of drug removed over time is proportional to the concentration of the drug) for higher levels, especially in chronic alcoholics. The MEOS pathway may account for the increased elimination rate at higher blood levels. The absorption and elimination rates of alcohol vary by individual and depend on many factors: diet, gender, body weight and habitus, speed of consumption, gastric motility, the presence of food in the stomach, smoking history, age, whether the person is a chronic alcohol consumer with enzyme induction and high-activity MEOS, advanced cirrhosis, the presence of ascites, and the state of nourishment.12 There is enormous variation among patients in the rate of disappearance of ethanol from the blood, ranging from 9 to 36 mg/dL/hr in published data.
Table 183-1 Physiologic Effects and Blood Alcohol Levels BLOOD ALCOHOL CONCENTRATION (MG/DL)
20–50 50–100 100–150 150–250 300 400
EFFECTS*
Diminished fine motor control Impaired judgment; impaired coordination Difficulty with gait and balance Lethargy; difficulty sitting upright without assistance Coma in the novice drinker Respiratory depression
*These effects are for the occasional drinker. Chronic drinkers can function at much higher alcohol concentrations because of tolerance. On the other hand, patients may become comatose with low levels of alcohol in mixed alcohol-drug overdose.
Although the clearance rate may be as high as 36 mg/dL/hr in some chronic drinkers, 20 mg/dL/hr is a reasonable rate to assume in a typical intoxicated ED patient. This holds true for adults, adolescents, and children. In the unusual circumstance that an accurate prediction of the rate of clearance is required, a second measurement should be obtained several hours after the initial value.13 Physiologic effects vary directly with the blood alcohol (Table 183-1). Diminished fine motor control and impaired judgment appear with alcohol concentrations as low as 20 mg/dL (0.02 mg%), but wide individual variability exists. Chronic alcoholics can exhibit impressive tolerance. The blood alcohol concentration (BAC) of a person cannot be accurately determined without quantitative testing. More than 50% of the adult population are obviously intoxicated with a level of 150 mg/dL (0.15 mg%). As the ethanol level rises, the patient’s level of consciousness declines, eventually ending in coma. Death is caused by aspiration or respiratory depression. In most states, the legal level of intoxication while driving is 80 mg/dL (0.08 mg%). Many states now allow administrative driver’s license revocation at BACs as low as 20 mg/dL (0.02 mg%). There is not an illegal BAC for activities other than driving. Alcohol via passive diffusion will be present anywhere there is water in the body. Hence, expired breath alcohol or saliva can be used to obtain a reliable approximation of BAC in a cooperative patient. This value can be used as a rapid screen for alcohol intoxication.14,15 The breath alcohol level can be falsely low with uncooperative patients.
■ DIFFERENTIAL CONSIDERATIONS Acute alcohol intoxication is a diagnosis of exclusion. Before assuming that a patient’s behavior is caused only by alcohol, other conditions should be considered. Hypoglycemia, hypoxia, carbon dioxide narcosis, mixed alcohol-drug overdose, ethylene glycol or methanol poisoning, hepatic encephalopathy, psychosis, severe vertigo, and psychomotor seizures can manifest in a manner similar to ethanol intoxication. The possibility of occult head trauma and the presence of associated metabolic disorders should be considered after alcohol intoxication has been established. Adequate history from paramedics and family, repeated physical examinations by the same clinician, and diagnostic adjuncts can help resolve this dilemma.
■ MANAGEMENT Comatose or stuporous patients need to have their airway and ventilation evaluated, with endotracheal intubation performed as necessary. Gastric lavage and activated charcoal are of little
■ ALCOHOL WITHDRAWAL SYNDROME Principles of Disease The neurophysiology of alcohol withdrawal is complex and not fully understood. Chronic alcohol consumption has a depressant effect on the central nervous system (CNS). The hallmark of alcohol withdrawal is CNS excitation with increased cerebrospinal fluid (CSF), plasma, and urinary catecholamine levels. Chronic alcohol consumption affects central adrenergic alpha-receptors, glutamate, central adrenergic beta-receptors, the inhibitory neurotransmitter γ-aminobutyric acid (GABA), and dopamine turnover. The effectiveness of lofexidine and clonidine (alpha2-adrenergic agonists), propranolol and atenolol (beta-blockers), haloperidol (dopamine blocker), and benzodiazepines and propofol (GABA transmission blockers) in suppressing the signs and symptoms of alcohol withdrawal supports this concept.17,18
Differential Considerations Alcohol withdrawal syndrome can initially be confused with acute schizophrenia, encephalitis, drug-induced psychosis, thyrotoxicosis, anticholinergic poisoning, and withdrawal from other sedative-hypnotic–type drugs. It may be difficult to differentiate between alcohol withdrawal and alcohol-induced hypoglycemia. Acute schizophrenia usually has its onset in adolescence or early adulthood. Manifestations include multiple bizarre delusions and a flat affect with the patient otherwise oriented.
2377
The patient in alcohol withdrawal is usually older (20s or 30s), hyperactive, and often disoriented. An important distinction: the schizophrenic patient will typically describe auditory hallucinations, whereas the alcoholic patient in withdrawal will more often describe visual hallucinations. Encephalitis can produce headache, confusion, fever, and seizures. Thyrotoxicosis is more common in women, and its features include irritability, insomnia, tremor, weight loss despite a good appetite, palpitations, and frequent stools. Physical examination may reveal lid lag, tachycardia, and a bruit over the thyroid. No relationship exists between the onset of encephalitis or thyrotoxicosis and alcohol consumption. Anticholinergic poisoning can occur with several different drugs or plant ingestion. The classic clinical picture is a patient with dry mouth, dry eyes, dry skin, hypoactive bowel sounds, urinary retention, and delirium. Amphetamine and cocaine intoxications produce anorexia, insomnia, and physical signs of CNS sympathetic overactivity. In opioid withdrawal, patients complain of abdominal pain and diarrhea, the mental status is usually normal, the patient is afebrile, and seizures are uncommon (with the exception of meperidine). In contrast, patients with major alcohol withdrawal are usually disoriented and febrile and may have seizures. Signs of alcohol withdrawal usually begin 6 to 24 hours after a decrease in the patient’s usual intake of alcohol. If patients manifest withdrawal 3 days or more after their last drink, drugs with a longer half-life should be considered. The barbiturate and benzodiazepine withdrawal syndromes usually progress more slowly, with a higher frequency of seizures later (7 vs. 2 days), and status epilepticus is more common than with alcohol withdrawal.
Clinical Features Isbell and colleagues’ classic 1955 study confirmed the relationship between alcohol and the withdrawal syndrome.19 They documented that the severity of signs and symptoms depends on both the dose and the duration of ethanol consumption. The withdrawal syndrome may occur any time after the blood alcohol level starts to fall. Therefore, only a reduction, not the abrupt cessation, of ethanol intake may result in withdrawal. The withdrawal syndrome usually develops 6 to 24 hours after the reduction of ethanol intake and lasts 2 to 7 days. The alcohol withdrawal state ranges from mild withdrawal with insomnia and irritability to major withdrawal with diaphoresis, fever, disorientation, and hallucinations. Minor alcohol withdrawal occurs as early as 6 hours and usually peaks at 24 to 36 hours after cessation of or significant decrease in alcohol intake. It is characterized by mild autonomic hyperactivity: nausea, anorexia, coarse tremor, tachycardia, hypertension, hyper-reflexia, sleep disturbances (e.g., insomnia and vivid dreams), and anxiety.20 Major alcohol withdrawal occurs after more than 24 hours and usually peaks at 50 hours but occasionally takes up to 5 days to manifest after the decline or termination of drinking. The syndrome is characterized by pronounced anxiety, insomnia, irritability, tremor, anorexia, tachycardia, hyper-reflexia, hypertension, fever, decreased seizure threshold, auditory and visual hallucinations, and, finally, delirium.21 Delirium tremens is the extreme end of the spectrum and consists of gross tremor, frightening visual hallucinations, profound confusion, agitation, and a hyperadrenergic syndrome characterized by a temperature greater than 101ºF, a blood pressure greater than 140/99 mm Hg, and tachycardia. It seldom appears before the third postabstinence day. Only 5% of patients hospitalized for alcohol withdrawal develop
Chapter 183 / Alcohol-Related Disease
value in ethanol overdose because of the rapid absorption of alcohol but may be appropriate in a suspected mixed drugalcohol overdose. Ethanol is similar to general anesthetics that act on the lipid moiety of cell membranes. Because there are no specialized receptors for alcohol, a specific antagonist does not exist. Thiamine (100 mg intravenously [IV]) to prevent or treat Wernicke-Korsakoff syndrome, glucose (dextrose, 25–50 g IV) for hypoglycemia, and naloxone (0.8 mg IV) for possible opioid ingestion are considered in comatose patients. Whenever possible, hypoglycemia should be documented before the empiric administration of glucose. With the airway maintained and respirations supported, the patient’s liver eventually metabolizes the alcohol, and the patient should recover. Glucose (dextrose, 25 g IV) produces a dramatic response in alcohol-induced hypoglycemic patients. Unlike hypoglycemia of other causes, alcohol-induced hypoglycemia may be unresponsive to glucagon because of depleted liver glycogen stores. Although Wernicke’s encephalopathy is a medical emergency, alcohol-induced hypoglycemia is a much more common condition with serious and permanent morbidity if left untreated. Therefore, thiamine can be given in a timely fashion, but glucose therapy should not be delayed.16 Intoxicated patients require evaluation and treatment in the ED regardless of their obstreperousness. Inappropriate discharge and failure to diagnose are two common areas of liability when treating the alcohol-dependent patient. The theoretical liability for detention by reasonable restraint is less than the potential liability for injury sustained by the alcoholdependent patient or an innocent bystander after premature discharge. Discharge (after excluding significant abnormal laboratory values or suspected head injury) can be considered when a patient is clinically sober and able to dress, walk, and function independently. In ideal circumstances, a concerned, sober adult is available and willing to take responsibility for and remain with the patient for the next 24 to 48 hours.
PART V ■ Special Populations / Section Six • The Alcoholic and Substance Abuse Patient
2378
delirium tremens. Other causes of delirium to be considered in the alcoholic patient include sepsis, meningitis, hypoxia, hypoglycemia, hepatic failure, and intracranial bleeding. True delirium tremens is rare and is not synonymous with alcohol withdrawal.
Management Out-of-Hospital Care The alcohol-dependent patient in withdrawal may also have a mixed alcohol-drug ingestion, occult head trauma, or cervical spine injury. Patients who are unable to sit without assistance or have an altered mental status require IV access. Naloxone (0.8 mg) and glucose (dextrose, 25 g) may be given in an IV bolus. Rapid blood glucose testing is preferable, but glucose may be given for altered mental status if this testing is not readily available. Thiamine should be given as soon as possible after glucose. The airway should be maintained and respirations supported. Emergency medical service personnel should monitor the patient’s vital signs and neurologic status. The cervical spine should be immobilized if trauma is suspected. It is usually best to withhold additional treatment until the patient can be evaluated in the ED. Emergency medical service personnel should be alert for other medical disorders that accompany alcoholism, such as pneumonia, sepsis, gastrointestinal bleeding, pancreatitis, hepatic failure, hypoglycemia, and intracranial hemorrhage.
Hospital Care Initial Assessment. Family, friends, bystanders, or paramedics may give more reliable historical data than the patient. Accurate vital signs are essential. This may require a rectal temperature. Hyperthermia, hypothermia, tachypnea, or tachycardia may suggest serious disorders that often accompany the alcoholdependent patient. These disorders should be considered during this first assessment. A rapid, thorough examination should be done with attention to the level of consciousness, signs of hepatic failure, or coagulopathy. Signs of trauma are sought, such as subcutaneous emphysema, ecchymosis, subconjunctival hemorrhage, hemotympanum, Battle’s sign, or palpation for fractures. The neurologic examination should search for focal findings, including central facial nerve palsy, hemiparesis, asymmetry of reflexes, or asymmetry of pupillary response. Treatment Plan. The alcohol withdrawal syndrome should be promptly recognized and treated. Treatment is necessary (1) to provide relief from anxiety and hallucinations; (2) to halt progression to major withdrawal and withdrawal seizures; (3) to allow detection of a treatable primary psychiatric illness; (4) to prepare the patient for long-term alcohol abstinence with the lowest risk of new drug dependence; and (5) to calm the patient and allow adequate examination for the detection of medical illnesses that typically accompany alcoholism, such as gastritis, dehydration, pancreatitis, pneumonia, electrolyte disorders, and hepatitis. In combination with appropriate chemical sedation, detention by reasonable restraint may be an option to prevent potential injury that patients may inflict on themselves or the hospital staff. Appropriate restraints are preferable to allowing decision-challenged patients to sign an “against medical advice” form and be discharged. Pharmacologic Intervention. Patients suffering from alcohol withdrawal should receive pharmacologic intervention along with supportive care. The ideal drug for alcohol withdrawal would have a rapid onset, a wide margin of safety, a metabolism not
dependent on liver function, and limited abuse potential. Although no one drug class fits all these requirements, benzodiazepines are clearly the mainstay of treatment. Benzodiazepines. The benzodiazepines have superior anticonvulsant activity, have the least respiratory and cardiac depressive effect of all the CNS depressants, and can be given parenterally in the uncooperative patient. By interacting with receptors linked to the GABA-associated chloride ion channel, benzodiazepines substitute for the withdrawal of the GABApotentiating effect of alcohol and abate withdrawal signs and symptoms.22 Numerous benzodiazepines have been studied. No evidence of clear superiority of any one benzodiazepine exists. Lorazepam has good bioavailability with oral, intramuscular (IM), and IV routes. It is rapidly and completely absorbed from IM sites in agitated patients with no IV access. Lorazepam’s half-life is intermediate (7–14 hours), and it reaches a steady state in 36 to 48 hours without active metabolites. Excessive sedation, confusion, and ataxia are potential complications of all benzodiazepines with prolonged half-lives. Lorazepam is metabolized (conjugated) in the liver, yielding inactive products. Although lorazepam’s half-life increases in patients with cirrhosis or liver failure, it is much less than the increase with chlordiazepoxide. Lorazepam’s elimination is only minimally altered in patients with renal failure and in the elderly. Lorazepam may be given IV in a dose of 1 to 4 mg, depending on the severity of the withdrawal. Dosing can be repeated at 5- to 15-minute intervals for patients in severe withdrawal. An IM dose of 1 to 4 mg can be used every 30 to 60 minutes until calm, then every hour as needed for light somnolence. The oral schedule for moderate withdrawal is 6 mg/day in three divided doses, tapering the amount by 1 or 2 mg/day over 4 to 6 days. Diazepam can be given in a dose of 5 mg IV every 5 to 10 minutes (2.5 mg/min) in major withdrawal until the patient is calm. The dose can be repeated in 5 to 10 minutes. If the second dose of 5 mg is not working, consider 10 mg for the third and fourth doses every 5 to 10 minutes. If this is not effective, consider 20 mg for the fifth and subsequent doses until adequate sedation is obtained.23,24 Because of erratic absorption, diazepam should not be given IM. The dosage of benzodiazepines required for alcohol withdrawal is highly variable. Practically, the dose is titrated to the patient’s agitation. Massive IV drug doses have been required in patients with delirium tremens, including a recorded 2640 mg of diazepam and 35 mg of haloperidol over 48 hours, 75 mg of midazolam in 1 hour, and 2850 mg of midazolam over 5 days.25 Butyrophenones. Haloperidol, a dopamine antagonist, can be considered in patients with major alcohol withdrawal or delirium tremens not responding to IV benzodiazepines. Haloperidol is more potent than chlorpromazine, has lower anticholinergic properties, and has less propensity to cause cardiovascular side effects or lower the seizure threshold. Haloperidol has little effect on myocardial function or respiratory drive, and its safety and efficacy by the IV, IM, or oral route in the ED have been established. Haloperidol has no anticonvulsant properties, however extrapyramidal effects may be seen. In 2008, Ortho-McNeil placed a black box warning on haloperidol for elderly patients with dementia-related psychosis. Elderly patients with dementia-related psychosis treated with antipsychotic drugs such as haloperidol have 1.6 to 1.7 times the risk of death compared with placebo-treated patients. Caution should be used in patients who may be susceptible to a prolonged QTc. Droperidol has similar effects as haloperidol. In 2001, the Food and Drug Administration issued a black box warning regarding QTc prolongation and torsades de pointes following droperidol use; nevertheless, droperidol remains a
Emergency Department and Outpatient Approaches Rapid, aggressive control of alcohol withdrawal is crucial. The cornerstone of treatment is a benzodiazepine. Lorazepam is preferable because of its previously discussed qualities. An initial test dose of lorazepam may be given IV in a dose of 1 to 4 mg, depending on the severity of the withdrawal. Dosing can be repeated at 5- to 15-minute intervals for patients in severe withdrawal. An IM dose of 1 to 4 mg can be used every 30 to 60 minutes until a patient is calm and then every hour as needed for light somnolence. Patients remain under observation or are admitted until the manifestations of withdrawal do not progress after the effects of the benzodiazepine have dissipated.22 Outpatient treatment consists of lorazepam, 1 or 2 mg three times a day in a tapering dose for 3 to 6 days; chlordiazepoxide, 25 to 100 mg three times a day in a tapering dose for 3 to 6 days; or diazepam, 30 mg once a day tapered over 5 days depending on the severity of symptoms. Adequate diet, abstinence, and participation in a rehabilitation program in the community are also desirable. Any patient requiring 300 mg of chlordiazepoxide or 60 mg of diazepam per day to control withdrawal should be considered for admission. Patients with major alcohol withdrawal (disorientation, hallucinations, diaphoresis, or fever) are admitted. Doses approximately equivalent to 100 mg of chlordiazepoxide are 20 mg of diazepam and 5 mg of lorazepam.22
Adjunctive Therapy
2379
Differential Diagnosis of
BOX 183-1 Alcohol-Related Seizures Withdrawal (alcohol or drugs) Exacerbation of idiopathic or post-traumatic seizures Acute intoxication (amphetamines, anticholinergics, cocaine, isoniazid, organophosphates, phenothiazines, tricyclic antidepressants, salicylates, lithium) Metabolic (hypoglycemia, hyponatremia, hypernatremia, hypocalcemia, hepatic failure) Infectious (meningitis, encephalitis, brain abscess) Trauma (intracranial hemorrhage) Cerebrovascular accident Sleep deprivation Noncompliance with anticonvulsants
of the most challenging and controversial (Box 183-1). Patients presenting to the ED with seizures should be questioned about alcohol intake. In 20 to 40% of seizure patients presenting to an ED, the seizures are related to alcohol use or abuse.29 Alcohol is a causative factor in 11 to 24% of patients with status epilepticus.30,31 In states where alcohol sales are restricted on Sundays, EDs may see a spike in alcohol-related seizures on Mondays.32 The primary consideration in the initial care of seizure patients who use alcohol is the recognition of treatable, lifethreatening causes. These causes include, but are not limited to, CNS infection, metabolic disorders, and intracranial hemorrhage. Alcohol may act in one of several ways to produce seizures in patients with or without underlying foci: (1) by its partial or absolute withdrawal after a period of chronic intake; (2) by an acute alcohol-related metabolic disorder (e.g., hypoglycemia and hyponatremia); (3) by creating a situation leading to cerebral trauma; (4) by precipitating seizures in patients with idiopathic or post-traumatic epilepsy; (5) by lowering the seizure threshold in patients with prior existing intracerebral disease states; or (6) by persistent heavy ingestion of alcohol, independent of alcohol withdrawal. Moreover, alcoholics are more susceptible to other disorders associated with seizures, including neurosyphilis, acquired immunodeficiency syndrome, brain abscess, and meningitis.33-35
Patients being treated for major alcohol withdrawal should receive thiamine (100 mg IV) and magnesium (2 g IV). Although magnesium sulfate does not decrease the severity of withdrawal symptoms, the incidence of delirium, or seizures, it carries no significant risk (with adequate renal function) or cost. In the nonacute setting, oral magnesium supplementation in chronic alcoholics improves liver function tests, electrolyte balance, and muscle strength.28 Multivitamin preparations may be considered for chronic malnutrition. Although their clinical benefit is not proved, they carry no significant risk or cost.18 If present, volume depletion can be corrected with normal saline. Reversal of electrolyte and metabolic disorders (hypomagnesemia, hypophosphatemia, hypokalemia, and acidosis) benefits the patient, but it does little to abate the withdrawal syndrome. Phenothiazines are contraindicated because they can produce hypotension, lower seizure threshold, disturb central temperature regulation, and cause extrapyramidal effects in the dosages required to calm patients in alcohol withdrawal.
Descriptions of alcohol withdrawal seizures (AWSs) were based on data collected by Victor and Brausch on 241 alcohol abusers with seizures or an alcohol-related illness complicated by seizures.36 Seizures occurred 6 to 48 hours after the cessation of drinking. Ninety percent had one to six generalized tonic-clonic seizures. Sixty percent experienced multiple seizures within a 6-hour period. However, data suggest seizure recurrence can be reduced to 3% with lorazepam administration following the initial seizure.37 The incidence of partial seizures, common with post-traumatic epilepsy, is increased in alcohol withdrawal. Regardless, first-time seizures and partial seizures warrant an evaluation for intracranial pathology. The term alcohol withdrawal seizure is reserved for seizures with the characteristics described by Victor and Brausch.36 The term alcohol-related seizure (ARS) is used to refer to all seizures in the aggregate associated with alcohol use, including this subset of AWS.
■ ALCOHOL-RELATED SEIZURES
Management
Among the many medical problems related to alcohol abuse, the differential diagnosis and management of seizures are one
After ensuring airway, breathing, and circulation (ABC), treatment should start by quickly establishing intravenous access.
Alcohol Withdrawal Seizures
Chapter 183 / Alcohol-Related Disease
safe and effective treatment for agitated patients.26 Dystonic reactions occur with both haloperidol and droperidol. Haloperidol and lorazepam in combination are safe and may be synergistic, with haloperidol for the hallucinations and lorazepam for the hyperadrenergic effects. Use in extremely high doses in patients with serious underlying medical illness has been documented in several studies: 240 mg of haloperidol and 480 mg of lorazepam given over 24 hours in a patient and 485 mg of haloperidol over 8 hours in another patient without significant adverse effects. Some advocate an aggressive protocol with escalating doses of the combination of haloperidol and lorazepam.27
PART V ■ Special Populations / Section Six • The Alcoholic and Substance Abuse Patient
2380
If the patient has altered mental status, thiamine, magnesium, dextrose, and naloxone should be considered. An empirical glucose bolus should not be used if an accurate determination of blood glucose is quickly possible.38 Although magnesium administration does not decrease the severity of withdrawal symptoms, the incidence of delirium, or seizures, it carries no significant risk or cost. In the nonacute setting, oral magnesium supplementation in chronic alcoholics improves liver function tests, electrolyte balance, and muscle strength.28 Multivitamin preparations may be considered for the patient’s IV fluid for chronic malnutrition. Their clinical benefit is not proven. Treatment with 2 mg of lorazepam IV reduces the risk of recurrent seizure in chronic alcoholics from 3 to 24%.37 It is often difficult to rule out a CNS infection in alcoholics because of concomitant hyperthermia, serum leukocytosis, and CSF pleocytosis. Although fever may suggest meningitis, it may be found in intracranial hemorrhage, brain abscess, alcohol withdrawal, toxic ingestions, and infections outside the CNS. Temperature can rise as a result of tonic-clonic seizure. Although there were few infections found in a retrospective series of 140 patients presenting with ARS, if CSF infection is a possibility, IV antibiotics should be started, blood cultures obtained, and a lumbar puncture (LP) undertaken.32 An LP may be delayed until a space-occupying lesion is ruled out with computed tomography (CT) scanning, but antibiotics should not be delayed if meningitis is a possibility.
Patients Presenting with a Normal Neurologic Exam New-Onset Alcohol-Related Seizures Patients with new-onset ARS should be thoroughly evaluated. This includes alcoholics who claim to have had seizures but for whom no documentation or an appropriate workup is available.39,40 Metabolic disorders, toxic ingestion, infection, and structural abnormalities should be considered. Laboratory and radiographic testing including electrolytes, blood urea nitrogen, creatinine, glucose, anticonvulsant levels, and brain CT scan may be necessary. Of 259 patients presenting with their first ARS, clinical management was changed in 3.9% based on head CT results.41 If the initial physical exam, imaging studies, and laboratory tests are within normal limits, patients who remain seizure free and symptom free with no sign of withdrawal after 4 to 6 hours of observation may be discharged. These criteria may be difficult to meet; therefore, admission may be considered. It may be unclear whether the patient has had a pure AWS or a new-onset seizure disorder in the setting of alcohol ingestion. Long-term treatment with antiepileptic drugs (AEDs) is not necessary in unprovoked new-onset seizures that have resolved or when a clear cause of seizures, such as alcohol consumption, can be identified. Optimal outpatient treatment includes follow-up and referral to a detoxification/ rehabilitation program. Ideally, the help of a concerned family member or friend, who is not a drinking partner and can remain with the patient for at least 1 or 2 days, is helpful.
the first 24 hours after alcohol withdrawal. An initial dose of 2 mg of lorazepam or 5 mg of diazepam can be given intravenously. These doses frequently need to be repeated.45 The patient is observed for 4 to 6 hours before he or she is considered for discharge. Prescribing benzodiazepines or AEDs upon discharge carries its own hazards. Prescribing benzodiazepines (other than a short 3- to 6-day tapering dose for withdrawal) may increase the potential risk of addiction. Using AEDs, such as phenytoin, in a potentially noncompliant patient may paradoxically increase the number of seizures by having large fluctuations in blood levels. The poorly compliant alcoholic patient may do better without outpatient anticonvulsants for a concurrent seizure disorder.43 The ideal disposition is referral to a detoxification/rehabilitation unit.
Alert Patient with a Seizure before or after Presentation The alcoholic patient with a known history of ARSs who experiences a single seizure or a short burst of seizures should be treated with benzodiazepines. These patients can usually be discharged after monitoring of neurologic status for 4 to 6 hours.37
Patients with an Abnormal Neurologic Presentation New-Onset Partial Seizures Partial seizures are reported to account for 24 to 51% of ARS.46 Conversely, studies have shown that 17 to 21% of partial ARS patients have structural lesions (hematomas, tumors, vascular abnormalities, or stroke).47 These primary causes of partial ARS, such as prior head trauma, may be easily missed in the history taking. As a result, an emergent CT scan is indicated to evaluate new-onset partial seizures. The patient with a history of a focal ARS who has been previously evaluated does not require an emergency CT scan, provided a return to baseline occurs promptly. A patient presenting with a focal ARS with subsequent normal neuroimaging can be managed with supportive care, observation for 4 to 6 hours, and a benzodiazepine for withdrawal signs or symptoms. Appropriate follow-up should be arranged.
Status Epilepticus Although fewer than 8% of ARS patients go into status epilepticus, alcohol is implicated in 15 to 24% of status epilepticus cases.48 Status epilepticus may also be the first presentation of ARS. The most common cause of status epilepticus is discontinuation or erratic compliance with an anticonvulsant drug regimen, followed by ARS.30,31 However, status epilepticus may arise for a variety of reasons and is often multifactorial. Initial interventions for the alcoholic in status epilepticus include stabilization of the ABCs, administration of thiamine and glucose as indicated, and treatment with a benzodiazepine. Lorazepam and diazepam are both effective in terminating seizures in status epilepticus. Lorazepam is preferable because its anticonvulsant effect lasts several hours, whereas diazepam’s anticonvulsant effect lasts only 20 to 30 minutes.41,49-52
Seizures in the Alert Patient with a History of Seizures during Prior Withdrawal
Obtundation
The risk of seizure increases significantly in alcoholic patients with manifestations of alcohol withdrawal who relate a history of AWS.43,44 Detoxification with benzodiazepines reduces AWS and should be initiated early because most AWSs occur within
The obtunded or stuporous alcohol-dependent patient with a history of seizure activity poses a diagnostic challenge. The patient’s decreased level of consciousness (LOC) may be the result of a postictal state, occult head trauma, unrecognized
No History of Seizures, No Current Seizure In the alcoholic patient in withdrawal who lacks a history of seizures, benzodiazepines generally have sufficient anticonvulsant activity to prevent withdrawal seizures.
Phenytoin-Anticonvulsant Conundrum Phenytoin has no significant benefit over placebo in preventing recurrence of AWS. Considering the risks of phenytoin and because it has no demonstrated benefit in the setting of AWS, it is not indicated for the treatment of AWS. The sudden withdrawal of phenytoin may potentiate the convulsive effects of alcohol withdrawal. Withdrawal seizures may occur in epileptic patients withdrawn from phenytoin.29,53 Alcoholic patients with preexisting seizure disorders pose a dilemma when they are supposed to be taking antiepileptic drugs but their blood levels suggest noncompliance. This is especially problematic when their epileptic attacks are uncommon and appear to occur exclusively in the context of alcohol withdrawal. Some of these patients may have AWS and may have been misdiagnosed. Others may have a seizure disorder that appears to be confined to the setting of alcohol withdrawal. Such patients have demonstrated that they cannot maintain compliance with their treatment. A patient currently taking AEDs for an antecedent seizure disorder who presents with a seizure while intoxicated falls into a different category. Such an episode could be an isolated
2381
event in a usually compliant patient without a history of chronic alcohol abuse. In this patient, a seizure in the setting of a subtherapeutic AED level may represent the consequences of noncompliance with AED or sleep deprivation versus AWS.43
■ OTHER CLINICAL FEATURES AND MANAGEMENT Cardiovascular Effects Acute and chronic ethanol consumption can affect the mechanical function of the heart, produce dysrhythmias, and exacerbate coronary artery disease (CAD). It may alter myocardial function by direct toxic effects, associated hypertension, or indirectly by altering specific electrolytes. Acute intoxication can decrease cardiac output in both alcoholic and nonalcoholic patients with preexisting cardiac disease.53 Studies have linked moderate alcohol consumption (up to two drinks per day in men and one in women) to a protective effect from CAD. As much as 50% of the relative risk reduction of CAD can be explained by increased levels of highdensity lipoprotein (HDL) and its subfractions HDL-2 and HDL-3. Genetic variations in the ADH allele may account for these changes. Low to moderate alcohol consumption decreases platelet aggregation, raises plasma levels of endogenous tissue plasminogen activator,54 and lowers insulin resistance. Experimental data suggest that alcohol may have antioxidant properties, produce effects on smooth muscles through interactions with nitric oxide, and alter plasma total homocysteine levels.55,56 Studies suggest that moderate alcohol consumption, through a reduced risk of CAD, may also protect individuals from congestive heart failure (Box 183-2).57 All of these beneficial effects are lost in heavy drinkers, in whom chronic alcoholism is associated with hypertension and congestive cardiomyopathy.
BOX 183-2 Risks and Benefits of Light, Moderate, and Heavy Drinking Light/Moderate Drinking Risks Established Unresolved Unlikely
Heavy Drinking Risks Noncardiovascular
Cardiovascular
Heavy drinking Breast cancer Fetal damage Bowel cancer Hemorrhagic stroke High blood pressure
Liver cirrhosis Pancreatitis Certain cancers Accidents Homicides Suicides Fetal damage Degenerative central nervous system disorders High blood pressure Arrhythmia Hemorrhagic stroke Cardiomyopathy
Benefits Probable Possible
Benefits None
From Klatsky A: Drink to your health? Sci Am 288:74, 2003. Copyright © Scientific American, Inc.
Decreased risk of coronary heart disease Decreased risk of ischemic stroke Decreased risk of gallstones Decreased risk of diabetes Decreased risk of peripheral vascular disease
Chapter 183 / Alcohol-Related Disease
metabolic disorder, or poisoning. The first diagnostic task is to quickly determine the possibility of hypoglycemia (diagnosed and reversed at the bedside in minutes) and the evaluation of other metabolic and toxic causes of altered mental status. Patients with an acute alteration in mental status should undergo an emergent head CT if they are not demonstrating expected and obvious improvement.
PART V ■ Special Populations / Section Six • The Alcoholic and Substance Abuse Patient
2382
The typical patient with alcoholic cardiomyopathy is a man older than 30 years with a greater than 10-year history of chronic alcohol intake. The signs and symptoms are no different from those of low-output congestive heart failure of other causes: dyspnea, palpitations, weakness and fatigue, jugular venous distention, poor R wave progression, nonspecific electrocardiographic abnormalities, and biventricular enlargement on a chest radiograph. Echocardiography shows four-chamber enlargement with decreased left and right ventricular contractile function. Up to one third of chronic alcoholic patients have left ventricular dysfunction demonstrated by a radionuclide ventriculogram, usually coexisting with skeletal muscle disease. Women appear to be more sensitive to the toxic effects of alcohol on striated muscle and are at greater risk for cardiomyopathy and myopathy. The diagnosis is made by obtaining a history of prolonged alcohol use and excluding hypertensive, coronary, valvular, and congenital disease. Heavy alcohol consumption (more than 2 ounces a day) has a detrimental effect on those with preexisting CAD. It can reduce exercise tolerance, induce coronary vasoconstriction, and raise heart rate and blood pressure.58 Additive cardiovascular effects of ethanol and nicotine contribute to dysrhythmias and sudden death in patients with CAD. In one study, nearly half the patients with alcohol withdrawal had prolongation of the QT interval. Prolonged QT can precipitate a dysrhythmia, resulting in sudden death.59 There is an increased incidence of sudden death among heavy drinkers regardless of concomitant CAD or smoking. Supraventricular (usually atrial fibrillation) and ventricular (usually transitory ventricular tachycardia) dysrhythmias, labeled “holiday heart,” have been documented in alcoholic patients who have been drinking heavily. One study reported that alcohol contributes to or causes new-onset atrial fibrillation in approximately two thirds of patients younger than 65 years. Alcohol also affects cardiac function indirectly by lowering potassium and magnesium levels. Data from the Framingham Heart Study indicate that patients with lower levels of potassium and magnesium have higher rates of dysrhythmias.60 Left ventricular ejection fractions improve with either abstinence or sustained decrease of alcohol consumption starting at 1 year and continue to improve over at least 4 years. Tachydysrhythmias as a result of episodic drinking commonly revert to sinus rhythm with abstinence and do not require immediate intervention if the patient is hemodynamically stable. Nevertheless, correction of electrolyte abnormalities is prudent.
Pulmonary Effects Alcohol reduces the mobilization of alveolar macrophages and their bactericidal capacity. Their impairment is greatest in alcoholics with hepatic cirrhosis. These effects, along with aspiration, decreased airway sensitivity, concomitant smoking, and malnutrition, probably account for the increased incidence of pneumonia, particularly lobar pneumonia, among alcoholic patients.61 At least 80% of alcoholics are smokers, making it difficult to distinguish between alcohol-induced and tobacco-induced injury to the lungs. The high prevalence of respiratory disease in alcoholics is largely caused by smoking. Chronic alcohol abuse has been shown to increase the risk of developing adult respiratory distress syndrome.62,63 Alcohol induces bronchospasm in some asthmatics and increases ventricular ectopy and sleep apnea in patients with chronic obstructive pulmonary disease. Alcoholic patients with hepatic cirrhosis can have hypoxemia as a result of precapillary shunting in their lungs. Hyperventilation and respiratory alka-
losis are also seen with hepatic cirrhosis. One or two drinks per day decrease the risk of pulmonary embolus and deep vein thrombosis in elderly patients.64
Gastrointestinal and Hepatic Effects Esophagus and Stomach Alcoholic patients have a higher incidence of esophagitis, gastric cancer, and esophageal carcinoma than the general population. Acute alcohol ingestion also decreases lower esophageal sphincter pressure, delays gastric emptying, and disrupts the normal gastric mucosal barrier. Vomiting is common among drinkers. Forceful or persistent emesis can lead to a Mallory-Weiss tear or Boerhaave’s syndrome.
Gastrointestinal Bleeding Alcohol is closely associated with gastrointestinal bleeding. Causes include Mallory-Weiss tears, esophagitis, esophageal varices, acute and chronic gastritis, thrombocytopenia, portal hypertensive gastropathy, qualitative and quantitative platelet disorders, and prolonged clotting times. Alcohol may exacerbate gastric mucosal damage when combined with nonsteroidal anti-inflammatory drugs (NSAIDs), but ethanol is not a risk factor for peptic ulcer disease. An inverse relationship exists between consumption of alcohol, particularly wine, and active Helicobacter pylori infection. Peptic ulcer disease is the most common cause of bleeding in alcoholic patients with upper gastrointestinal hemorrhage as well as in those who do not drink.65
Liver The liver is the primary site of ethanol metabolism. Hepatic damage has been recognized for centuries as the hallmark of chronic alcohol abuse. Obesity potentiates the severity of alcohol-induced liver damage.66 The production of cytokines such as tumor necrosis factor alpha is one of the earliest events in many types of liver injury. This cascade may trigger the production of other cytokines that together enlist inflammatory cells, kill hepatocytes, and initiate healing through fibrogenesis. There is no one test that can be used to diagnose alcoholic liver disease (ALD) reliably. However, the ratio of aspartate transaminase (AST) to alanine transaminase (ALT) greater than 1.5 suggests that alcohol is the cause of liver injury.67 ALD is the most common liver disorder in the Western world and, along with hepatitis C, is a leading cause of liver transplantation. The earliest, mildest, and most common liver change seen in alcoholism is the accumulation of macrovesicular fat in the hepatocytes, predominantly involving triglycerides. Alcoholic fatty liver is usually asymptomatic, associated with mild elevations of AST and ALT. It is usually detected by the finding of hepatomegaly on physical examination or abnormalities on ultrasonography or CT but is confirmed by liver biopsy. Fatty liver is a reversible disorder if the patient can refrain from drinking.68
Alcoholic Hepatitis Alcoholic hepatitis is more serious than fatty infiltration and develops in up to 35% of heavy drinkers.69 These individuals usually have right upper quadrant pain, a tender enlarged liver, fever, jaundice, leukocytosis, and altered liver function tests. AST levels are usually less than 500 IU/L, and ALT levels are typically less than one half the AST levels. Alcoholic hepatitis
2383
other causes include biliary tract disease, hypercalcemia, hypertriglyceridemia, penetrating peptic ulcer, abdominal trauma, and reactions to various drugs. Alcohol is the leading cause of chronic pancreatitis. Diarrhea and impaired intestinal absorption are common problems of the chronic alcoholic. Alcohol increases small intestine transit time and decreases brush border enzyme activity. Thiamine, vitamin B12, amino acids, folic acid, and glucose all have impaired absorption in alcoholics. Dietary deficiencies in folic acid and protein, pancreatic insufficiency, abnormal biliary secretion, and direct toxic effects of ethanol on the gastrointestinal tract all contribute to malabsorption. Abstinence and adequate nutrition reverse the diarrhea and much of the malabsorption.68
Alcoholic Cirrhosis
A potential pitfall is ascribing the cause of the patient’s altered mental state to acute intoxication without considering other conditions. Coma or an altered mental state may be caused by acute intoxication, mixed alcohol-drug overdose, postictal states, head trauma, hypoglycemia, shock from gastrointestinal bleeding or sepsis, hypothermia, hyperthermia, hepatic encephalopathy, methanol-isopropyl alcohol-ethylene glycol poisoning, or Wernicke-Korsakoff syndrome. These potentially catastrophic diagnoses are usually detected by a thorough history and physical examination, a blood alcohol level (coma is rare in patients with blood alcohol levels 100 fL) by an unknown mechanism. Macrocytosis is the most common hematologic manifestation of the chronic alcoholic. It may be caused by folate deficiency, reticulocytosis (the younger reticulocytes are larger), liver disease (producing an abnormal lipid coating of RBC membrane), or vitamin B12 deficiency. The most common condition is idiopathic macrocytosis of alcoholism. Iron deficiency anemia is common among alcoholic patients and usually is a result of blood loss from the gastrointestinal tract. Alcoholics are subject to chronic inflammatory diseases such as endocarditis, tuberculosis, empyema, lung abscess, malignancy, and hepatic disease. These chronic inflammatory illnesses can produce the anemia of chronic disease, a mild microcytic or normocytic anemia in which the serum iron is low, the total serum iron–binding capacity is low or low-normal, and serum ferritin is increased. With iron deficiency anemia, the serum iron is decreased, the total serum iron–binding capacity is elevated, and serum ferritin is decreased. Ethanol also has a direct toxic effect on erythropoiesis. Bone marrow biopsies reveal vacuolization of erythroid precursors, resulting in decreased reticulocytosis and a reversible sideroblastic anemia. Sideroblastic anemia, usually in the presence of malnutrition with pyridoxine deficiency and folate deficiency, occurs in 25 to 30% of anemic alcoholics.
Thrombocytopenia can occur with folate deficiency, sepsis, disseminated intravascular coagulation, or splenic sequestration. The direct toxic effects of alcohol decrease measured survival time and impair production of platelets in the bone marrow, but marrow toxicity will rarely reduce the platelet count below 30,000. Qualitative platelet function is also impaired. Binge drinking is associated with a reactive thrombocytosis potentially responsible for acute stroke and sudden death.99
Hemolytic Syndromes and Erythrocyte Abnormalities A variety of hemolytic syndromes have been associated with alcoholism. Zieve syndrome is a transient hemolytic anemia with hyperlipidemia and fatty infiltration of the liver. Acquired stomatocytosis, a condition characterized by abnormally shaped RBCs that are susceptible to hemolysis and acanthocytosis (spur cell anemia), has been associated with alcohol abuse. Zieve syndrome and stomatocytosis may be reversed with abstinence. Spur cells are RBCs with spicules. Spur cell hemolytic anemia is frequently linked to alcoholics with cirrhosis. With jaundice and splenomegaly, spur cell hemolytic anemia is usually fatal. When severe hemolysis is present, remission is rare.98 These syndromes are associated with liver disease, which alters the lipid composition of the RBC membrane, and congestive splenomegaly, which produces hemolysis. Severe hypophosphatemia (5 years old) for 5 minutes continuously or two or more discrete seizures without full recovery of consciousness.18 Early intervention is important. Treatment initially includes benzodiazepines, which are often effective; if two doses of a benzodiazepine are not effective, subsequent doses are not likely to work, and phenytoin (Dilantin) or fosphenytoin (Cerebyx) is considered the next line of medication.19 (See Chapter 173 for a complete discussion of the management of childhood seizures.)
Vagal Nerve Stimulators Vagal nerve stimulators (VNS) are implantable devices used to prevent seizures. Intractable epilepsy affects approximately 20 to 30% of patients with seizures. Newer antiepileptic drugs and surgical procedures decrease seizure frequency in a significant number of patients. Even among this group, up to 10% continue to have disabling seizures. For these patients, VNS provides hope for better seizure control. The VNS is an implantable device that looks like a pacemaker. It is implanted by a neurosurgeon just under the skin in the chest. In the small number of children who have had VNS placed, most have seen a reduction in seizure frequency of 50%.20 How the VNS works on the brain is not known. The vagus nerve is a peripheral nerve that leads directly to the brain and it is there that the VNS has an effect.
Chapter 185 / Evaluation and Management of Children with Special Health Care Needs
PROCEDURE
PART V ■ Special Populations / Section Seven • The Developmentally or Physically Disabled Patient
2406
The VNS is programmed to provide baseline intermittent stimulation of the left vagus nerve. Although the VNS system delivers stimulation automatically in regular pulses all the time, a magnet can be used to deliver extra electronic stimulation in between cycles. The patient or caregiver activates the device by placing a handheld magnet over the device implanted in the chest.21 Patients who can sense that they are about to experience a seizure can activate the VNS themselves by passing the magnet over the device. Out-of-hospital providers who encounter a seizing patient with a VNS should assist the caregiver with its activation or seek advice from medical control. Children with VNS devices should otherwise be treated as other seizing patient, with careful attention to airway, breathing, and circulation. A magnet can also be used to stop the stimulation by holding the magnet over the device. Sometimes the patient may want to shut the stimulation off. Some of these include when eating if the stimulation causes problems with swallowing, when the patient plans to speak or sing in public as the stimulation can change one’s voice, and if the stimulation causes discomfort.22,23
Myelomeningocele and Hydrocephalus Myelomeningocele (or spina bifida) refers to a gap in the vertebral arches. The most common types of spina bifida are spina bifida occulta, the milder form, and spina bifida associated with malformation of the spine and attachment of the sac—a condition known as myelomeningocele. The rate of spina bifida in the United States has decreased markedly since the introduction of folic acid to the diets of women of childbearing age. Clinical presentation depends on the level of the defect. The condition results in partial or complete paralysis and loss of sensory function (not always symmetrical) with motor loss. In addition, the child may present with loss of bladder or bowel control, cognitive impairments, visual deficits, and seizure disorders. Bladder and bowel dysfunction is very common among children with spina bifida. In such cases, there is incomplete emptying of the bladder, predisposing the child to urinary tract infections. Management of such cases requires emptying of the bladder by urinary catheterization several times a day. Hydrocephalus associated with Chiari malformation (displacement of the brainstem and part of cerebellum downward through the skull) occurs in greater than 68% of children with spina bifida.24 One should always assume that children with spina bifida are allergic to latex.25
Ventriculoperitoneal Shunts Ventriculoperitoneal (VP) shunts are catheters that are inserted into the ventricles within the brain and then threaded under the skin from the skull to the peritoneum where excess cerebral spinal fluid (CSF) is drained (Fig. 185-6). Hydrocephalus occurs when an obstruction occurs somewhere in the CSF circulation system. If the obstruction occurs beyond the lateral ventricles, then they enlarge with CSF, increasing intracranial pressure. Hydrocephalus can be seen in formerly premature infants who sustained an intracranial hemorrhage during the neonatal period, in children with brain tumors, and posttraumatically. Also, children with spina bifida (myelomeningocele) often have hydrocephalus if they have an Arnold-Chiari malformation that obstructs CSF flow. Within the first 3 postoperative months, although rare, a child with a VP shunt can develop an infection of the shunt track. Symptoms include fever, ill appearance, erythema over the shunt site or tubing, tenderness over the tubing, abdominal pain and tenderness, vomiting, and altered mental status. If a VP shunt infection is suggested, CSF should be drawn
Point where shunt dips into ventricles
Point where shunt dips into abdomen
Figure 185-6. Ventriculoperitoneal shunt. (Adapted from Susan Gilbert in Teaching Resource for Instructors in Prehospital Pediatrics [TRIPP]; www.cpem.org.)
from the shunt by a neurosurgeon and sent for cell count and culture. Broad-spectrum antibiotics should be administered pending cell count and culture results; however, likely organisms include Staphylococcus epidermidis, Staphylococcus aureus, and, rarely, Haemophilus influenzae. Neurosurgeons should manage all shunt infections. Peritonitis is another complication of VP shunts. The tip of the shunt empties into the peritoneal cavity and as a foreign body can serve as a nidus for infection. Signs and symptoms include fever, vomiting, abdominal pain and tenderness, and abdominal distention. Management is similar to that for VP shunt infections and should include diagnostic lab work, broad-spectrum antibiotics, and consultation with the neurosurgeon. The more common complication of a VP shunt is shunt obstruction or malfunction. This can occur with an increase in protein in the CSF that causes a blockage in the tubing or as a result of a mechanical disruption in the shunt tubing. This can cause a buildup of CSF in the ventricles and an increase in intracranial pressure. Signs and symptoms include headache, nausea, vomiting, irritability, altered mental status, ataxia, change in vital signs, and a bulging fontanel in an infant. Initial management should include elevating the child’s head and managing the patient’s airway, breathing, and circulation. Stable patients should have head computed tomography performed, and a neurosurgeon should be consulted as soon as possible. The definitive treatment is surgical shunt revision.6
Gastrostomy Tubes Gastrostomy tubes are placed in children who need long-term nutritional supplementation or cannot take food by mouth. There are many conditions that necessitate placement of an artificial feeding tube including severe developmental delay, coma, short bowel syndrome, swallowing difficulties, burns to mouth and esophagus, failure to thrive, and chronic diseases that affect nutrition such as cystic fibrosis. Gastrostomy tubes are surgically or endoscopically placed feeding catheters. These tubes include gastrostomy tubes
2407
Coca-Cola or a proteolytic enzyme solution; otherwise, they need to be replaced. G tubes that have fallen completely out of the patient need to be replaced as soon as possible, so the stoma does not constrict and make replacement difficult. Gastrostomyjejunostomy tube placement is more problematic because after placing a tube in the stomach, passing the tube to the jejunum is usually done under fluoroscopy by a radiologist familiar with this procedure. For G tube replacement, the physician should ask when the tube was first inserted. For tubes that are less than 3 months old, consultation with the person or service that performed the procedure is necessary because the track may not yet be fully formed and insertion by a practitioner not adept at replacing these tubes may create a false track. When performing the reinsertion procedure, a similar size gastrostomy tube should be used. Parents often have an extra tube with them. If cases when the ED does not have access to G tubes (the Mic-Key type of tube is easiest to use), using a Foley catheter to temporarily keep the stoma open is acceptable until definitive placement can occur. Once the G tube is replaced, tube placement can be confirmed by instilling a small amount of diatrizoate meglumine (Gastrograffin) dye into the tube (15–30 mL) and obtaining two radiographic views to observe the tube and dye in the stomach. Confirma-
Table 185-3 Troubleshooting Gastrostomy Tubes/Buttons PROBLEM
POSSIBLE CAUSATION
ACTIONS
Nausea, vomiting, cramping, and/or diarrhea
Too rapid feeding Feeding too cold Spoiled formula or a change in formula Mechanical Broken or sticking valve Malpositioned tube wedged into the gastric mucosa Balloon not inflated properly Organic Increased volume of feeding (too much gas or formula) Constipation Ileus Pneumonia (from coughing), seizures Feeding or medications too thick or remain in shaft too long
Increase feeding time. Ensure feeding is at room temperature.
Leakage of stomach contents
Blockage of button
Accidental removal of button
Vigorous pulling on the button Spontaneous deflation of the gastrostomy balloon
Stoma site irritation Maceration due to moisture Gastric acid burn from leakage of gastric contents Purulent mucus
Regular cleaning not effective Incomplete drying Feeding spillage with incomplete cleaning Leakage of gastric contents from or around tube Excessive granulation tissue
Balloon leaks or ruptures
Silicone balloons generally last about 3 months; however, life of balloon may vary
Place insertion obturator, decompression tube, or 8to 10-Fr suction catheter into shaft of button so it moves the flex valve. Anti-reflux valve makes popping sound when it moves back into closed position. Ensure balloon is fully inflated. Check stomach residual, readjust method of feedings as appropriate. Evaluate for other organic causes and treat accordingly. Flush button with 5–10 mL of tap water after using. Use thin solutions with well-crushed meds. Use insertion obturator or 8- to 10-Fr suction catheter to gently try to push the plug through. Insert gastrostomy tube to keep stoma open. As a last resort, a Foley catheter with an anti-migration device may be inserted into the stoma. Tube must be securely taped in place. Immediately transport to tertiary hospital for button replacement. Button length may be too long. Button length may be too short. Do not use occlusive dressing, gauze, tape, or creams. Rotate button 360 ° once a day. Clean site daily with soap and water. Clean site with 1/3 strength hydrogen peroxide if irritated. Consider using a skin barrier cream and H2 blocker for gastric burns. Replace gastrostomy button.
From Adirim T, Smith E: Special Children’s Outreach and Prehospital Education. London, Jones and Bartlett, 2005.
Chapter 185 / Evaluation and Management of Children with Special Health Care Needs
(G tubes), jejunostomy tubes (J tubes), and percutaneous endoscopic gastrostomy tubes (PEGs or buttons). J tubes are placed in children with GERD. The tube is placed in the stomach and then passed through the junction between the stomach and jejunum, bypassing the stomach. PEGs are placed by gastroenterologists either at the bedside or in the outpatient procedure area. There are three potential gastrostomy tube emergencies. (1) The tube can leak gastric contents, (2) become obstructed, or (3) come completely out. When presented with a child who has any of these problems, assessment of hydration status is important, especially in those children who are totally dependent on their feeding tubes for hydration and nutrition. One should inquire about prescribed medications and whether any doses were missed. In children experiencing leaking around the catheter, management focuses on delineating and fixing the cause of the leak. Possible causes of leakage include balloon deflation, coughing, constipation, bowel obstruction, and seizure. Addressing the cause may solve the problem; otherwise, consultation with the subspecialty service that manages the child’s tube may be necessary (Table 185-3). Sometimes the child’s G tube becomes clogged with medication or food. Obstructed tubes can be cleared with either
2408
PART V ■ Special Populations / Section Seven • The Developmentally or Physically Disabled Patient
Table 185-4 Potential Central Line Emergencies and Physician Management/Considerations External catheter dislodgement
Complete catheter dislodgement Damaged catheter
Bleeding at catheter entry site
Internal bleeding
Catheter occlusion
Dislodgement of blood clot
Air embolus Fever
Reactions from home nutrient or medication infusions
Direct pressure to skin site Clamp line near the exit site Place a peripheral IV to continue necessary fluids and medications Do chest radiograph to locate distal end of catheter Consider a dye study to determine location of catheter tip Obtain surgical consultation Direct pressure to skin site Place a peripheral IV to continue necessary fluids and medications Obtain surgical consultation to determine disposition Clamp catheter proximal to the break with a hemostat wrapped in gauze Estimate blood loss Observe for symptoms of air embolus If PICC line in place, remove it For other partially implanted lines, repair catheter if >2 in. of the line exit from the skin Direct pressure to entry site Estimate blood loss Place a peripheral IV to continue necessary fluids and medications Do chest radiograph to locate distal end of catheter Consider a dye study to determine location of catheter tip Obtain surgical consultation Direct pressure to site Observe for symptoms of hemopneumothorax Place a peripheral IV to continue necessary fluids and medications Do chest radiograph to locate distal end of catheter Consider a dye study to determine location of catheter tip Obtain surgical consultation For problems withdrawing blood, try placing the child in various positions (arms high, laying supine, etc.) Attempt to gently aspirate the clot using a 10-mL syringe half filed with NS placed directly onto the male Leur lock of the catheter Consider using a fibrinolytic agent Never force fluids through the catheter! Pulmonary embolism may result Observe for tachycardia, tachypnea, hypoxemia, and chest pain Immediately stop infusion of fluids Clamp catheter Observe for sudden changes: tachypnea, chest pain, shortness of breath, or loss of consciousness Clamp system, place child on left side in a head down position, give oxygen Consider patient bacteremic or septicemic until proven otherwise Sepsis can quickly develop in immunosuppressed children Evaluate for source of infection Begin broad-spectrum antibiotics Admit to the hospital for antibiotic therapy Immediately stop infusion of medications and begin infusion of NS Treat signs of allergic reaction with Benadryl, steroids, and subcutaneous injections of 1 : 1000 epinephrine if symptoms are severe
IV, intravenous; NS; normal saline; PICC, peripherally inserted central venous catheter. From Adirim T, Smith E: Special Children’s Outreach and Prehospital Education. London, Jones and Bartlett, 2005.
tion of replacement of the G tube is indicated when the track is relatively new (placed within 1 month of ED visit), when the tube is difficult to replace, or when there is any doubt that the tube is in the stomach. Patients should not be discharged home without a definitive G tube placement. Discussion with the subspecialty service that manages the tube is important for further advice. If there is a delay in G tube replacement, then assessment of hydration is important. For the child who cannot take fluids orally, intravenous hydration may need to be considered.6
Central Venous Catheters Central venous catheters, or central lines, are used to deliver medications, blood products, and nutrition directly into a central vein. There are a number of circumstances when a child would need a central venous line, including children
with cancer who need chemotherapy, children with sickle cell disease who need frequent transfusions, children with infections who need long-term antibiotic therapy, and various conditions in which nutritional supplementation is needed, such as in short bowel syndrome. There are three types of catheters. Peripherally inserted central venous catheters (PICC) are long catheters inserted into the cephalic vein via the antecubital fossa. The catheter is advanced into the subclavian vein. These tubes are placed in children who need temporary venous access, such as for antibiotic therapy. These tubes can be placed by nonsurgeons. Potential complications include infection, obstruction, and easy dislodgement because they are not sutured in place (Table 185-4). Tunneled central venous catheters are placed surgically. They are inserted directly into a central vein, most commonly the subclavian, cephalic, or external jugular. There are three
2409
known. Neutropenic patients (absolute neutrophil count 9100 mg/ week).144,145 Chronic use of acetaminophen should be avoided in patients with hepatic or renal disease. Renal failure report-
edly can worsen with acetaminophen use, but the mechanism is unknown.146 Patients with a history of salicylate hypersensitivity characterized by urticaria have an 11% cross-reactivity to acetaminophen, and the agent should be used with caution in this group.147 For mild analgesia and fever reduction, acetaminophen is the first-line agent and is a first choice for use in combination with other agents, usually opioids, in the treatment of patients with more severe pain. Acetaminophen is available in many formulations, including caplets, tablets, and liquid and PR formulations. The recommended dose of acetaminophen for an adult is 650 to 1000 mg every 4 to 6 hours, not to exceed 4000 mg/day.
Nonsteroidal Anti-inflammatory Agents NSAIDs inhibit cyclooxygenase (COX) and, as a result, the synthesis of prostaglandin, a key mediator of inflammation. The analgesic effect of NSAIDs is peripherally mediated by decreasing prostaglandin and effectively raising the threshold of activation of nociceptors. NSAIDs have synergistic effects with opioids and can reduce the amount of opioids needed to achieve pain relief in a patient. Two COX isoenzymes mediate prostaglandin synthesis. COX-1 is present in all cells and plays an important role in homeostatic functions. COX-2 is induced by injury or inflammation and generates prostaglandins as part of the inflammatory process. Nonselective NSAIDs inhibit both COX-1 and COX-2, which results in multiple beneficial effects (reduction of inflammation, pain, and fever) but also some important undesirable effects. As a group, and because of their wide use, NSAIDs are responsible for more serious drug-related side effects than any other class of analgesic drugs.148 The major side effects of NSAID analgesic agents are gastrointestinal (GI) bleeding, renal failure, anaphylaxis, and platelet dysfunction. The majority of these side effects occur in patients who are taking NSAIDs for chronic conditions. It is estimated that more than 100,000 hospital admissions and approximately 16,500 deaths each year from GI bleeding are related to NSAID use for osteoarthritis and rheumatoid arthritis.149 One survey estimated that for every 100,000 people taking NSAIDs each year, there are 300 GI-related deaths, 5 liver-related deaths, 4 renalrelated deaths, and some congestive heart failure–related deaths.150 Limited evidence suggests that prostaglandins promote bone formation and that NSAIDs might inhibit the process, but this has not been established through properly conducted studies.150,151 There is no evidence that short term use of NSAIDs for analgesia after fracture is deleterious to healing. In addition to prostaglandin, cyclooxygenase helps generate prostacyclin, a vasodilator that increases GI mucosal perfusion. In the stomach, COX-1 increases bicarbonate and mucus production, important for protecting the mucosal lining. Inhibition of COX-1 compromises these protections, predisposing patients to ulcerations and bleeding, which are then exacerbated by concomitant NSAID-induced platelet dysfunction.152 COX-1 and COX-2 affect the cardiovascular system through the production of endothelial prostacyclin (vasodilatory) and thromboxane (platelet aggregation). Inhibition of COX-1 produces antiplatelet activity that may be cardioprotective by inhibiting thromboxane production more than prostacyclin. Inhibition of COX-2 inhibits prostacyclin production more than thromboxane and may produce prothrombotic effects, increasing the risk of cardiovascular events. In the case of nonselective COX inhibitors, these two effects appear to balance each other out, resulting in few changes in cardiovas-
Drug Interactions Aspirin. NSAIDs may impair the cardioprotective effect of aspirin, although the available evidence is unclear and the use of daily aspirin for cardiac prophylaxis should not deter the prescribing of an NSAID for acute pain or inflammation.158,159 Oral Anticoagulants. The antiplatelet effects of NSAIDs add to the anticoagulant properties of warfarin, compounding the risk of significant bleeding complications, especially from GI ulcers. Furthermore, NSAIDs displace protein-bound warfarin and cause subsequent increases in prothombin times at a constant warfarin dose.160 NSAID use is generally avoided in patients who are taking warfarin. ACE Inhibitors. Concurrent use of NSAIDs with angiotensinconverting enzyme (ACE) inhibitors may impair renal function and impair the antihypertensive effects of ACE inhibitors. Diuretics. Patients who are taking diuretics have a greater risk of developing renal failure because of NSAID-mediated decreased renal blood flow. Also, the natriuretic response to diuretics depends in part on prostaglandin-mediated vasodilatation.
of Serious Gastrointestinal Effects of Table 186-5 Risk Nonselective NSAIDs149,150 NSAID
COX-2 inhibitor Ibuprofen Diclofenac Sulindac Naproxen Indomethacin Tolmetin Piroxicam Ketoprofen Ketorolac Risk Reduction When Added to Ibuprofen164 Proton pump inhibitor Misoprostol
RELATIVE RISK OF SERIOUS GI TOXICITY
0.6 1.0 1.8 2.1 2.2 2.4 3.0 3.8 4.2 24.7 0.09 0.57
GI, gastrointestinal; NSAIDs, nonsteroidal anti-inflammatory drugs.
2423
Glucocorticoids. Patients on corticosteroids have an increased risk of peptic ulcer disease. NSAIDs should generally be avoided in patients concurrently taking glucocorticoids unless closely supervised by an ambulatory care physician. Lithium. NSAIDs enhance lithium reabsorption and may directly reduce lithium excretion, leading to increased lithium levels. CNS symptoms (drowsiness, confusion, vertigo, convulsions, or tremors), cardiac dysrhythmias, and QRS widening are warnings of lithium toxicity. The lithium dosage should be reduced when an NSAID is prescribed. Methotrexate. Chronic coadministration of NSAIDs and methotrexate have resulted in prolonged, elevated blood levels of methotrexate, resulting in severe toxicity. A possible mechanism may be the decreased renal perfusion caused by NSAIDS, decreasing the elimination of methotrexate.
Nonselective Cyclooxygenase Inhibitors NSAIDs combine analgesia and anti-inflammatory effects with low abuse potential and much different side effects from those for the opioids. Oral NSAIDs can be as effective as PO opioids for mild to moderate pain. Parenteral NSAIDs are available in many countries (only ketorolac is available in the United States) but offer little advantage over their PO forms.157 Different patients respond differently to both the effects and the side effects of different NSAIDs; some experimentation may be necessary to determine the best choice for a particular patient. No particular NSAID is proven superior for any indication. Drug selection should depend on availability, side effect profile, convenience, and cost. Patients at risk for adverse events using NSAIDs are listed in Box 186-4. Ketorolac Tromethamine. Ketorolac is the first nonopioid analgesic agent available for parenteral use in the United States, but it is rarely indicated because 60 mg of ketorolac administered IM is not superior to 800 mg of PO ibuprofen, which is easier to administer at a fraction of the cost.161-163 Ketorolac’s main use is in the early treatment of renal colic (accompanied by a loading dose of IV morphine), a pain mechanism for which NSAIDs are particularly effective. If the patient can tolerate PO medication, 800 mg of PO ibuprofen is given instead. Ibuprofen. Ibuprofen is the most widely used drug in the NSAID class. It is available over the counter in a variety of preparations, including tablets, liquid suspension, and sup-
Patients at Risk for Adverse Events
BOX 186-4 Using NSAIDs
1. Patients who are dehydrated or hypovolemic or who have impaired renal function are at increased risk for decreasing renal function or renal failure. 2. Patients with liver disease or congestive heart failure, in particular, those already taking ACE-inhibitors, ABRs, or diuretics, in whom liver or heart conditions may worsen. 3. Elder patients, particularly those at risk for GI and renal events. 4. Patients with asthma and known aspirin hypersensitivity are at an increased risk of bronchospasm. 5. Women in the third trimester of pregnancy. NSAIDs may prolong gestation or prematurely close the ductus arteriosus. 6. Patients who use tobacco or ethanol or who have a history of gastritis or peptic ulcer disease are at risk for peptic ulcer and GI bleed. ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; GI, gastrointestinal; NSAIDs, nonsteroidal anti-inflammatory drugs.
Chapter 186 / Pain Management
cular risk in studies of these drugs. In the case of selective COX-2 inhibitors, this may result in an increase in cardiovascular risk.153-155 Prostaglandin produced by COX-1 causes renal vasodilation that maintains renal blood flow and the glomerular filtration rate (GFR). Inhibition of COX-1, especially in volumedepleted patients, can result in decreased GFR and even acute renal insufficiency. Sodium and water retention, hypertension, hyperkalemia, and acute renal failure may also ensue, especially in patients with congestive heart failure. The most common adverse effect of NSAIDs is GI mucosal injury. In patients taking NSAIDs continuously for 1 year, 10 to 60% will develop abdominal pain, dyspepsia, or nausea and 2 to 4% will develop symptomatic ulcers.156 Risk factors include age, concomitant use of warfarin or corticosteroids, congestive heart failure, diabetes, and coronary artery disease. There is evidence that cytoprotective agents such as misoprostol and proton pump inhibitors reduce this risk.157 The relative risk for causing GI effects of the various NSAIDs are listed in Table 186-5.
PART V ■ Special Populations / Section Eight • The Patient in Pain
2424
pository. Ibuprofen is rapidly absorbed in the upper GI tract and has minimal interaction with other medications. The adult analgesic dose is 400 mg, and the anti-inflammatory dose is 600 to 800 mg. No NSAID has been shown to be more effective as an analgesic than ibuprofen 400 mg, including ibuprofen 800 mg.161,164,165
COX-2-Specific Inhibitors The discovery of two distinct cyclooxygenase isoenzymes (COX-1 and COX-2), one (COX-2) associated mostly with pain and inflammation, raised hope that an effective new class of analgesics could be developed. These would control pain and inflammation with fewer adverse effects (particularly GI mucosal injury) than traditional NSAIDs. Despite great initial promise, a dramatic pricing differential, and intensive marketing, the two classes of NSAIDs (nonselective COX inhibitors and selective COX-2 inhibitors) perform similarly in clinical use, with similar side effect profiles. Agents that selectively inhibit COX-2 are expected to cause less ulceration and have a lower risk of bleeding. COX-2 has been identified in normal gastric mucosa, however, and selective inhibitors may not confer any GI-protective advantage. COX-2 inhibitors may have prothrombotic effects from greater inhibition of prostacyclin than thromboxane, thus increasing the risk of cardiovascular events. COX-2 inhibitors also decrease renal perfusion, thereby decreasing renin activity and reducing sodium excretion by the same amount (approximately 20%) as NSAIDs. No studies on COX-2 inhibitors’ effectiveness in pain relief have been conducted in ED settings. COX-2 inhibitors are superior to placebo and equivalent to COX-1 inhibitors in treating acute postoperative dental pain, postoperative orthopedic pain, primary dysmenorrhea, and osteoarthritis. There are no studies comparing COX-2 inhibitors’ efficacy in renal colic, biliary colic, acute gout, headache syndromes, sickle cell crisis, or acute musculoskeletal or soft tissue injury. COX-2 inhibitors should not be combined with NSAIDs because of their similar pharmacologic effects. COX-2 inhibitors have similar interactions as NSAIDs with ACE inhibitors, antihypertensive agents, anticoagulants, and lithium. Given their price, potential for adverse cardiovascular events related to long-term use, and lack of superior safety or efficacy compared with nonselective NSAIDs, there is little or no role for COX-2 inhibitors in the ED or as a discharge prescription.
Skeletal Muscle Relaxants Skeletal muscle relaxants are advocated as an adjunct to analgesics in the management of musculoskeletal pain with a “spasm” component, principally back pain. Despite the common use of skeletal muscle relaxants, relatively little data exist on their role in the treatment of pain. Studies show that muscle relaxants, such as cyclobenzaprine, are indistinguishable from ibuprofen in analgesic effect but have an increased side effect profile. Although a Cochrane Systematic Review claims that skeletal muscle relaxants are more effective than placebo with respect to relieving acute low back pain, it is not possible to discern any differential or additive effect to that of NSAIDs when the primary trials are reviewed.166 These drugs have not been found to be of benefit in the treatment of chronic low back pain, which is their most common use.167 Skeletal muscle relaxants should not be used in the management of acute musculoskeletal pain as a substitute for proper doses of effective analgesics unless there is a high degree of anxiety accompanying the pain, and an anxiolytic is felt to be helpful. In that case, a benzodiazepine, such as diazepam,
Muscle Relaxants and Mechanisms Table 186-6 Selected of Action MEDICATION
MECHANISM
Baclofen Benzodiazepines Carisoprodol Chlorzoxazone Cyclobenzaprine Metaxalone Methocarbamol Orphenadrine Tizanidine
GABA agonist GABA agonist Sedative Sedative Tricyclic antidepressant (sedative) Sedative Sedative Antihistamine (sedative) Central alpha2-agonist
GABA, γ-aminobutyric acid.
5 mg three times daily, may be an effective adjunct for pain control. Based on (a lack of) outcome benefit, there is no indication for the prescription or use of the other muscle relaxants shown in Table 186-6. Benzodiazepines have hypnotic, anxiolytic, antiepileptic, and antispasmodic properties. Muscle relaxation is probably due to GABA-mediated presynaptic inhibition at the spinal cord level but has not been shown to be clinically relevant. Diazepam is the most commonly used benzodiazepine for muscle spasm.167
Nitrous Oxide/Oxygen Mixtures Nitrous oxide/oxygen mixtures can be used in the ED or the out-of-hospital care setting to reduce anxiety in patients and to manage mild to moderate pain states. The analgesic and anesthetic properties of nitrous oxide were discovered more than 200 years ago. Combined with oxygen, a mixture of nitrous oxide and oxygen in a 50 : 50 ratio is safe when selfadministered by the patient. This technique is one of the original forms of patient-controlled analgesia. Nitrous oxide and oxygen administered by nasal mask have long been used by dentists. Experience in emergency medicine with nitrous oxide/oxygen mixtures has been greatest in the ratio of 50 : 50 with self-administered hand-held masks.168,169 The actual mechanism of analgesia and anxiolysis is not fully delineated, but it is known to diffuse through tissue membranes and is poorly soluble in blood. In the two-tank self-administered system, a fixed-ratio nitrous oxide/oxygen mixture is delivered to the patient through a demand valve activated when the patient inhales through a face mask or mouthpiece. A pressure of 3 to 5 cm H2O must be produced within the mask or mouthpiece to activate the flow of gas. This element provides safety for the patient-controlled aspects of the system since patients must initiate a breath and hold the mask to their face to receive the medication. In 10 to 15% of patients, nitrous oxide is ineffective.168 It is much more potent as an anxiolytic than as an analgesic agent. As with all analgesic agents, its success should be determined by the patient’s subjective feedback. When necessary, nitrous oxide can be supplemented with other analgesics. Nitrous oxide/oxygen mixtures are relatively or absolutely contraindicated in patients with a decreased level of consciousness who are unable to follow instructions, patients with a head injury, or those with decompression sickness. Patients with severe chronic obstructive pulmonary disease who retain CO2 should be given nitrous oxide/oxygen mixtures carefully because the mixture contains 50% oxygen. Because nitrous
2425
Table 186-7 Characteristics of Common Local Anesthetic Agents POTENCY (LIPID SOLUBILITY)
DURATION (MIN)
ONSET
COMMENTS
Solutions of 0.5–2%, used in infiltration and blocks Topical in the eye Most commonly used agent, 1.5 times as toxic as procaine Less potent than lidocaine, less toxic Long acting, used in infiltration and blocks Twice as toxic as lidocaine, used mostly in epidurals
Procaine Tetracaine Lidocaine
1 8 3
60–90 180–600 90–200
Slow Slow Rapid
Mepivacine Bupivacaine Etidocaine
2.4 8 6
120–240 180–600 180–600
Very rapid Intermediate Rapid
Modified from Paris PM, Weiss LD: Narcotic analgesics: The pure agonists. In Paris PM, Stewart RD (eds): Pain Management in Emergency Medicine. Norwalk, Conn, Appleton & Lange, 1988.
oxide does diffuse into body cavities, it can worsen a pneumothorax or bowel obstruction.56 Minor side effects of the analgesic gas mixture are reported in 5 to 50% of patients. The most common adverse effect is light-headedness, with paresthesias and nausea reported less frequently. No documented adverse hemodynamic effects have occurred with the self-administered forms of this agent. This safety has also been confirmed in the out-of-hospital care setting.170,171 Side effects usually resolve within minutes of discontinuation. Chronic use or abuse of nitrous oxide is associated with vitamin B12 antagonism and secondary hematologic effects as well as development of a myelopathy.172,173 Nitrous oxide requires an effective scavenging and ventilation system to avoid accumulation and toxicity in health care workers, and this requirement has limited its use in EDs.
Local Anesthesia Mechanism of Action Peripheral nerves are responsible for transmitting pain information from the pain receptors to the spinal cord. Each fiber consists of an axon surrounded by a covering called the Schwann cell. A myelinated axon is one that is covered by the projection of a Schwann cell that wraps itself many times around the axon, which is called the myelin sheath. Local anesthetics are much more effective at penetrating unmyelinated or lightly myelinated fibers than heavily myelinated ones. This differential explains the fact that ordinarily local anesthetic agents provide sensory block without a motor block (see Table 186-1). Local anesthetic agents reversibly block the sodium channels of the lipid membrane and prevent the sudden influx of sodium ions into the axon, blocking depolarization and the action potential. After injection of a local anesthetic, tissue buffers increase the pH of the solution surrounding the agent, driving some of the water-soluble acidic form of the agent to its lipid-soluble nonionic form. The lipid-soluble phase of the drug is able to penetrate the lipid membrane of the axon where it then ionizes and enters the sodium channel, blocking the channel’s ability to allow sodium to enter the cell.
Classes of Local Anesthetic Agents Local anesthetic agents are chemical compounds that consist of an aromatic and an amine group separated by an intermediate chain. The class that has an ester link between the intermediate chain and aromatic portion is called amino esters and includes procaine, chloroprocaine, and tetracaine. Amides have an amide link and include lidocaine, mepivacaine, prilocaine, bupivacaine, and etidocaine. Esters are unstable in solu-
tion and are metabolized in the body by the plasma enzyme cholinesterase. The amides, after absorption into the body, are destroyed by enzymes in the liver.174
Specific Agents Each local anesthetic has a predictable effect when used in appropriate doses and by the appropriate route. The main considerations in the clinical use of these agents are potency, duration of anesthesia, and the speed of onset (Table 186-7). Potency. The ability of a local anesthetic drug to penetrate the lipid membrane of the axon determines its potency. Agents that have a high lipid solubility (e.g., tetracaine, etidocaine) are more potent than those with a low lipid solubility (e.g., procaine, mepivacaine). Less potent local anesthetics must be given in more concentrated forms and in larger doses to achieve an equivalent effect. Duration of Anesthesia. Agents that bind well to protein in the sodium channel are longer acting and provide anesthesia of long duration. Tetracaine and bupivacaine have a high affinity for protein and provide long-lasting anesthesia, whereas procaine, which is poorly bound, does not. Onset of Action. In most cases, it is helpful to have an anesthetic agent that acts quickly. The speed of onset of any local anesthetic agent is directly related to how quickly that agent, after injection, can diffuse through tissues to the nerve and through the nerve membrane. After injection, the agent is in two forms, ionized and nonionized. The amount of drug in the nonionized form is determined by its pKa (the pH at which 50% of the solution is nonionized and 50% is ionized). Because only the nonionized form of the drug diffuses into the nerve, solutions with a low pKa have a more rapid onset of anesthesia. Local anesthetic agents with higher pKas take effect more slowly. At a tissue pH of 7.4, 5% of tetracaine (pKa 8.5) is in the nonionized form compared with 35% of lidocaine (pKa 7.9) solution. Low tissue pH (5 or 6) in surrounding infected tissue delays the onset of local anesthesia in situations such as abscess incision and drainage, because the anesthetic primarily remains in an ionized state. The onset of action of a local anesthetic can be hastened by the alkalinization of the solution carrying the drug, which also decreases its irritant effect (pain) on injection. This can be done by adding sodium bicarbonate solution to the anesthetic at a ratio determined by the pKa of the agent (e.g., 1 : 10 for lidocaine).175,176 Several other factors influence the clinical performance of local anesthetic agents. In the clinical dosages used, these agents (except cocaine) are vasodilators, which tend to shorten the duration of anesthesia. Injection of the solutions into vascular tissues not only shortens the duration of anesthesia but also increases systemic absorption and the (small) chance of systemic toxicity. For these reasons, epinephrine is often added to local anesthetic solutions.
Chapter 186 / Pain Management
AGENT
PART V ■ Special Populations / Section Eight • The Patient in Pain
2426
Allergy. Patients sometimes provide a history of allergy to local anesthetic agents, but it is usually not an immunoglobulin-mediated true allergy. Allergy to ester agents is most probably caused by the preservative methylparaben and its breakdown products. True allergy to the amide group is exceedingly rare. When allergy is reported, it is often caused by one of the preservatives used. Because the two groups do not cross-react, if a patient gives a history of allergy to an agent of one group, an agent from the other group can be used. In one study, 236 patients with reported adverse reactions to local anesthetics were tested using commercial preparations of unrelated local anesthetics. No patient exhibited systemic reactions.177 In those patients who insist they are allergic to all “-caine” anesthetic agents, and the allergy is believed to be legitimate (a very rare circumstance), diphenhydramine can be used; 1 mL of a 50-mg/mL ampule can be diluted to 5 to 10 mL (1–0.5% solution) and can be used for local infiltration or nerve block. Diphenhydramine may cause direct tissue toxicity and should be avoided when possible in areas with poor collateral circulation.178,179
Local and Systemic Toxicity Local Toxicity. Local anesthetic agents are directly toxic to tissue, depending on the concentration. Some investigators have claimed that the use of a vasoconstrictor in the anesthetic solution produces a reduction in blood flow that increases the wound healing time and the vulnerability of the wound to infection, but this has never been demonstrated. Nerve blocks are preferable to local infiltration for wounds that are extensive or contaminated. Epinephrine-containing solutions have traditionally been avoided on digits, the penis, ears, or the nose. Literature, however, suggests that dilute epinephrine can be used safely on digits and possibly these other areas as well.180 A comprehensive review of the use of epinephrine in digits concludes that it is safe when diluted to 1 : 200,000 or less, but it should not be used in patients with vascular disease. Systemic Toxicity. Systemic toxicity of local anesthetics occurs when a sufficient quantity of the drug accumulates in the body so that sodium channel blockade occurs in the heart or the brain. There is a dose-related clinical progression of local anesthetic toxicity from subtle neurologic symptoms to seizures to cardiovascular collapse. All local anesthetics produce systemic toxicity at sufficiently high blood or CNS concentrations. Each local anesthetic has a range of therapeutic safety, beyond which systemic toxicity is more likely to occur. Table 186-8 provides a guideline to safe doses that can be administered for local anesthesia. Overdosage may occur more commonly in patients with large wounds and in very small patients. The more lipophilic agents (e.g., etidocaine, bupivacaine) are more cardiotoxic. Cardiac toxicity may also be increased for epinephrine-containing anesthetics when inadvertent IV injection occurs. Special care should be exercised in children and when performing certain blocks known to produce high blood levels of the anesthetic agent (e.g., intercostal). In pediatric patients, total dose guidelines are important, and the maximum dose should be calculated before administration. A wide variety of symptoms may be experienced from local anesthetic toxicity. These include light-headedness, headache, paresthesias, tinnitus, decreases in the level of awareness, and muscle spasm.181 The degree to which CNS symptoms are experienced is directly related to the blood level of the local anesthetic. At the extreme, CNS toxicity may result in seizures. The clinical progression usually begins with a circumoral paresthesias, dysarthria, and a report of tinnitus
for Maximum Doses of Commonly Table 186-8 Guidelines Used Local Anesthesia Agents* AGENT
WITHOUT EPINEPHRINE
WITH EPINEPHRINE
Lidocaine HCl† Mepivacaine HCl Bupivacaine HCl§
3–5 mg/kg 8 mg/kg 1.5 mg/kg
7 mg/kg 7 mg/kg‡ 3 mg/kg
*All maximum doses should be reduced 20–25% in very young, old, and very sick patients. † A lidocaine level of 0.5–2.0 µg/mL may be reached for every 100 mg of lidocaine infiltrated for blocks. ‡ Epinephrine adds to the potential cardiac toxicity of this drug. § Not to be used for pudendal blocks or intravenous regional anesthesia. Not recommended for children younger than 12 years old. Adapted from Stewart RD: Local anesthesia. In Paris PM, Stewart RD (eds): Pain Management in Emergency Medicine. Norwalk, Conn, Appleton & Lange, 1988.
BOX 186-5
Techniques That Can Be Used to Reduce the Pain of Injection
Buffering of local anesthetic agents Counterirritation Slowing rate of injection Use of topical anesthetics Warming solution Distraction techniques or a similar auditory phenomenon, followed by a decreased level of consciousness and progressing to confusion, seizures, and coma. Longer-acting, more potent agents (e.g., bupivacaine and etidocaine) are more likely than lidocaine to cause CNS symptoms at lower blood levels.181 Local anestheticinduced seizures should be treated with IV benzodiazepines. Local anesthetic agents also have direct effects on cardiac automaticity, conductivity, contractility, and vascular tone. In animal experiments, the cardiac toxicity of bupivacaine is commonly seen before the CNS toxicity, and this toxicity is exacerbated by epinephrine. Management of cardiovascular collapse caused by toxic levels of local anesthetic agents should follow standard advanced cardiac life support guidelines. Unless the overdose is massive, the toxicity should be relatively short-lived because of the redistribution of the lipophilic agents.
Reducing the Pain of Injection Counterirritation by scratching, jiggling, or repetitively pinching the skin during needle puncture or injection reduces discomfort182 (Box 186-5). The addition of sodium bicarbonate to lidocaine immediately before injection significantly reduces patient discomfort.175 A standard solution of sodium bicarbonate (8.4% in 50 mL) can be added to a syringe containing lidocaine in a ratio of 1 : 10 (e.g., 1 mL bicarbonate to 10 mL lidocaine, or 0.5 mL to 5 mL). Buffered lidocaine can be stocked in the ED and is effective for up to 1 week.183 Bupivacaine can also be buffered, but the ratio should be 1 : 50 (i.e., 0.1 mL bicarbonate to 5 mL bupivacaine). Slow injection attenuates pain of infiltration to a greater degree than buffering of the solution.184 Injection of local anesthetic into the edges of the laceration is less painful than injection through intact skin surrounding the wound.185 When time permits, warming the anesthetic or the application of a topical anesthetic agent can also greatly decrease the initial sensation associated with needle injection.176
2427
Topical Anesthesia
Topical Anesthetics Applied to Intact Skin Eutectic Mixture of Local Anesthetics. Eutectic mixture of local anesthetics (EMLA) is a mixture of lidocaine and prilocaine in an alkaline oil mixture in which the anesthetics occur primarily in their nonionized form, allowing them to diffuse through the skin. The term eutectic refers to mixtures that result in a melting point higher than that of either agent alone. The mixture should be applied on the desired area with an occlusive dressing 30 to 60 minutes before the desired procedure is performed. Heating the EMLA for 20 minutes improves analgesia but is less effective than a routine 60-minute application with or without heat.186 The duration of action after a 60-minute application is 1 to 5 hours. Indications for the use of EMLA include venipuncture, arterial puncture, lumbar puncture, or arthrocentesis when a 30- to 60-minute delay in performing the procedure is not an impediment. EMLA can be applied in triage, particularly for pediatric patients, whose IV can then be started later in the ED with little or no pain. Ethyl Chloride and Fluori-Methane sprays. Ethyl chloride and fluorimethane sprays are occasionally used for superficial analgesia. The agents evaporate quickly and cool the skin, providing brief (10,000 lux) for 2 hours after arising as an adjunct in adjusting to new shifts. 10. Exercise regularly. 11. Plan regular “quality time” with family and friends. 12. Do not try to live a day shift lifestyle while working night shifts.
2603
BOX 203-3 American Heart Association Dietary Guidelines 6. Limit sodium intake to ≤2400 mg/day (≤6.0 g/day of salt). 7. If alcohol is consumed, limit intake to two drinks/day for men, one drink/day for women. 8. Eat at least two servings of fish per week. 9. Eat five or more servings of vegetables and fruit per day. 10. Eat six or more servings of grain products per day. 11. Emphasize daily intake of low-fat or nonfat dairy products.
event in thoughts, dreams, or daily life; an avoidance of any stimuli associated with the event; a sense of numbness to one’s emotions; and a wide variety of other symptoms, including sleep disturbance, irritability, anxiety, and loss of emotional control.45 The concept of critical incident stress debriefing has evolved over the past 30 years from the combined experiences of a variety of emergency services and military personnel. The two main goals of debriefings are to lessen the short-term emotional impact of distressing events and to accelerate recovery from such events before stress reactions occur. As originally described, debriefings are structured group meetings that emphasize venting of emotion and discussion of other reactions to a critical event.46 Formal critical incident stress debriefing teams are available in many communities through local emergency medical services agencies. Debriefings also can occur on an informal basis in the ED. Such sessions can take the form of impromptu meetings of emergency and rescue personnel involved in specific incidents in which the details of a particular event and its associated emotional content are expressed and shared among members of the team.44 It is important for those involved in leadership roles to make some form of critical incident debriefing processes available when such events occur.
exercise regimens and healthy dietary habits. The impact of proper diet and exercise on a variety of wellness parameters is clearly established. Documented benefits of an appropriate diet include lower cholesterol levels, weight control, augmented blood pressure control, and a lower incidence of certain cancers. The American Heart Association (AHA) periodically provides dietary guidelines, as summarized in Box 203-3.50 The benefits of a regular and vigorous aerobic exercise program include enhanced exercise tolerance, cardiovascular fitness, lower blood pressure, increased high-density lipoproteins, decreased triglyceride levels, augmented weight reduction efforts, lessened anxiety and depression, improved glucose tolerance, and enhanced endurance, flexibility, and strength. According to current AHA guidelines, cardiovascular fitness is best maintained by a program of aerobic exercise (the repetitive use of large muscle groups) three to six times per week for at least 30 minutes per session. Strength-developing activities (resistance training) should be performed at least twice per week, with 8 to 10 exercises that use the major muscle groups of the legs, trunk, arms, and shoulders. A regimen of one or two sets of 8 to 12 repetitions of each exercise is recommended. The exercises should be vigorous, sufficient to raise heart rate to at least 50% of maximum (recommended average maximum heart rate is 220 minus age).51
Family and Social Relationships
Relaxation and Renewal
Many physicians are overly dedicated to their work and prone to delay sources of emotional gratification outside the work environment. These qualities, which may promote a high degree of professional success and satisfaction, also may interfere with the development of family and social relationships, a problem that may be compounded for emergency physicians by their irregular hours.47-49 The ability to develop and sustain intimate relationships may be the most critical element in any wellness formula. The simplest and most profound way to maintain mental health while actively absorbed in a demanding career is to establish and cultivate a viable and genuine relationship with a caring, emotionally expressive spouse or significant other. The promotion of a close family relationship involves, at the very least, a willingness to assign and prioritize protected time with spouse and other family members. For many physicians, it also involves a conscious effort to develop communication skills, particularly with respect to active listening and the open expression of feelings.49
In dealing with the innumerable stresses of emergency practice, emergency physicians must be equipped with considerable emotional resources, derived in part from strong peer and personal support systems. Methods of relaxation and renewal are important additional elements.6,34,36 Emergency physicians must continually cultivate the ability to remain functional and emotionally resilient when facing the turbulence of the typical ED. Physiologic correlates of stress— increased heart and respiratory rates and increased blood pressure, oxygen consumption, and serum lactate level—can be normalized by a variety of relaxation techniques. These include prayer, Zen, yoga, transcendental meditation, autogenic training, and progressive relaxation. Studies indicate that the regular use of such techniques can allow development of considerable control over the relaxation process.52-54 A final element in the promotion of wellness and lifestyle balance is the concept of renewal. Physicians typically develop a work ethic that makes it difficult to disengage from medical responsibilities. A single-minded devotion to career ultimately is impoverishing, and by neglecting restorative activities, physicians tend to lose their emotional resilience.34,55 The concept of renewal involves the ability to prioritize and establish time for rest and revitalization, as well as spiritual, emotional, and intellectual growth. Out of the maelstrom that constitutes a typical day in the ED, emergency physicians should look to sources of renewal as key elements in their formula for lifestyle balance.
Physical Fitness Physical fitness is an important concern for emergency physicians, given the rigors of a typical emergency medicine practice. Unfortunately, the obstacles to developing a comprehensive fitness program may be considerable. Factors such as irregular hours, hectic work pace, and the variable quality of hospital menus complicate the task of establishing and maintaining
Chapter 203 / Wellness, Stress, and the Impaired Physician
1. Restrict total fat to ≤30% of total calories. 2. Restrict saturated fat to