Learning from Medical

Errors: Clinical Problems

Anh Vu T Nguyen
MD, FAAFP

and

Dung A Nguyen
MD, FAAFP

Foreword by Frank J Edwards

Radcliffe Publishing
Oxford • Seattle
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2005 by Anh Vu T Nguyen and Dung A Nguyen
CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Version Date: 20160525

International Standard Book Number-13: 978-1-4987-9986-7 (eBook - PDF)

This book contains information obtained from authentic and highly regarded sources. While all reasonable efforts have been made to publish
reliable data and information, neither the author[s] nor the publisher can accept any legal responsibility or liability for any errors or omissions
that may be made. The publishers wish to make clear that any views or opinions expressed in this book by individual editors, authors or con-
tributors are personal to them and do not necessarily reflect the views/opinions of the publishers. The information or guidance contained in
this book is intended for use by medical, scientific or health-care professionals and is provided strictly as a supplement to the medical or other
professional’s own judgement, their knowledge of the patient’s medical history, relevant manufacturer’s instructions and the appropriate best
practice guidelines. Because of the rapid advances in medical science, any information or advice on dosages, procedures or diagnoses should
be independently verified. The reader is strongly urged to consult the relevant national drug formulary and the drug companies’ and device
or material manufacturers’ printed instructions, and their websites, before administering or utilizing any of the drugs, devices or materials
mentioned in this book. This book does not indicate whether a particular treatment is appropriate or suitable for a particular individual. Ulti-
mately it is the sole responsibility of the medical professional to make his or her own professional judgements, so as to advise and treat patients
appropriately. The authors and publishers have also attempted to trace the copyright holders of all material reproduced in this publication and
apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged
please write and let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any
electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any
information storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or
contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organiza-
tion that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a
separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explana-
tion without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Contents

Foreword iv
List of cases v
About the authors ix

Introduction 1
1 Abdominal pain 5
2 Chest pain 33
3 Fever 67
4 Flank pain 87
5 Headache 97
6 Leg pain 111
7 Low back pain 119
8 Musculoskeletal injuries 131
9 Shortness of breath 145
10 Syncope 161
11 Vaginal bleeding 171
12 Wounds and lacerations 181
13 Legal issues involving radiology 197

Index 215
Foreword

There probably wouldn’t be much need for a book like this if medicine could be
practiced at a leisurely pace with patients well known to the physician in a setting
of abundant resources and willing consultants and if all diseases presented the
way they’re supposed to. But in the real world, of course – and especially in
emergency departments and urgent care centers – the exact opposite of this ideal
world holds true and the potential for good clinicians to commit diagnostic and
therapeutic misadventures is very real. Unfortunately, our standard medical
education doesn’t prepare us very well to avoid the traps and pitfalls. We are
grounded in pathophysiology and the recognition of symptom clusters and
classic diagnostic pathways. We are not, however, often shown how easily even
the best can be led astray.
One of the best ways to hone this aspect of the art of medicine, I believe, is by
studying negative examples – cases in which good doctors failed to take the right
steps, and more importantly, why this failure might have occurred. It is certainly
human to err, but as professionals we must let no mistake be wasted on us. This
book by the Nguyen brothers represents a major training contribution. It not
only reviews the basic steps involved in evaluating those major complaint
categories we face every day (such as chest pain and shortness of breath) but
supplements each section by relevant case examples that are always interesting
and enlightening.
There is something of value here for everyone from medical students interested
in the field to battle-scarred emergency physicians – veterans of a thousand shifts
facing similar dilemmas in the middle of the night.

Frank J Edwards MD, FACEP

President, Delphi Emergency Physicians, LLC
Assistant Professor of Medical Humanities, University of Rochester, NY
July 2005
List of cases

1.1 The importance of clinical suspicion 14

1.2 Did you bother to look? 15
1.3 Should we look for zebras? 16
1.4 What is the best test? 16
1.5 The elusive diagnosis 17
1.6 Already brewing 17
1.7 Not fitting the puzzle 18
1.8 Check the groin 19
1.9 Think of the obvious first 20
1.10 Did not read the text? 22
1.11 Get the necessary tests 22
1.12 Not always appendicitis 23
1.13 Remember the risk factors (‘F’s) 24
1.14 The interaction of organ systems 25
1.15 The answer is in front of you 25
1.16 Take ‘guarding’ literally 26
1.17 Getting to the bottom of things 27
1.18 It is not always gastrointestinal 28
1.19 Some information is less readily volunteered 28
1.20 Abdominal pain imitator 29
2.1 Do not forget the basics 42
2.2 Time is muscle 43
2.3 Do not be fooled by the age 43
2.4 What happens after the diagnosis is made? 44
2.5 Keep pulmonary embolus in the differential 45
2.6 Once with heart disease, always with heart disease 45
2.7 Drinks are not good for heart disease 46
2.8 A mouthful of problems 47
2.9 Once is not enough 48
2.10 Appearances can be deceiving 48
2.11 Common link between cardiac and neurologic diseases 49
2.12 Common link between cardiac and gastrointestinal diseases 50
2.13 Not the whole story 51
2.14 From one emergency to another 52
2.15 EKG first, questions later 53
2.16 What is up (or down) with the lungs? 53
2.17 Talk to me 54
2.18 Do not forget about pulmonary embolus 55
2.19 Two things at once 56
2.20 Check them again 56
2.21 Not likely to be usual 57
vi List of cases

2.22 Keep them coming 58

2.23 Treat it differently 60
2.24 It cannot wait 61
2.25 Treat them gingerly 61
2.26 Two test minimum 62
2.27 A result no-one expected 63
2.28 All in the family 63
3.1 Sometimes they fail 78
3.2 No time to waste 78
3.3 Fever and rash, double trouble 79
3.4 The picture did not fit 79
3.5 A crash course 80
3.6 One infection leads to another 80
3.7 Enter every orifice 81
3.8 Getting to the bottom of things 81
3.9 Every test counts 82
3.10 It takes time to grow 83
3.11 Pulmonary embolus, the great masquerader 83
3.12 Cannot live with food and water 84
3.13 A concern of great (or loss of) weight 85
4.1 It is not mandatory 91
4.2 A kidney attack 92
4.3 Coke is not it 92
4.4 Occasionally, one will go bad 93
4.5 A fresh approach 93
4.6 A good pick-up 94
5.1 But my pain was not better 103
5.2 Be careful if it hurts when they move 104
5.3 Is the brain working? 105
5.4 Not in pictures 106
5.5 Headaches and warning signs 106
5.6 Making a timely diagnosis 107
5.7 A real mind-bender 108
5.8 Upholding the standard of care 109
6.1 Think simple first 116
6.2 Not ‘just muscle’ 116
6.3 Not just growing pains 117
7.1 Put a hand on the belly 124
7.2 Red flags of back pain 125
7.3 Difficult diagnosis to make 125
7.4 More than pain in the back 126
7.5 Look both ways before passing 126
7.6 It pays to be lucky 127
7.7 Symptoms that cannot be ignored 127
7.8 Is that all that happened? 129
7.9 Is the Coumadin patient bleeding? 129
8.1 Thumbs down 137
8.2 Love yet known but lost 137
8.3 Look above and below 138
 List of cases vii

8.4 Kids need pictures 138

8.5 X-rays do lie 138
8.6 No injuries are minor 139
8.7 It is not broke, but fix it 140
8.8 Try to do no harm 141
8.9 (Hand)le with care 141
9.1 Never too young 152
9.2 Interpret with caution 153
9.3 That child will eat anything 153
9.4 X-rays are sometimes behind 154
9.5 Will travel, need airway 155
9.6 Old, fragile, and comorbidities 156
9.7 The masked acute coronary syndrome 157
9.8 Be careful of look-alikes 157
9.9 Easier to see the second time around 159
10.1 Search for syncopal injuries 166
10.2 Caution with multiple or crescendo patterns 167
10.3 Worth watching 168
11.1 I thought it was in . . . 176
11.2 Make sure it is over 177
11.3 Two for one 177
11.4 Do not wait for tests 178
11.5 What are we having? 179
12.1 Cannot read minds 187
12.2 Pollution is not the solution 188
12.3 They may not let you look 188
12.4 Check on your work 189
12.5 Guns can cause serious injuries 190
12.6 Be sharp with glass injuries 190
12.7 Cutting edge injuries 191
12.8 Do not seal the poison 191
12.9 How deep is it? 192
12.10 Neighboring injuries 192
12.11 Words of faulty wisdom 193
12.12 What was the culprit? 194
13.1 Passing the buck 199
13.2 Needed communication 200
13.3 Did you hear the report? 200
13.4 Choosing the right side 201
13.5 Taking the right pictures 202
13.6 Read it while it is still wet 203
13.7 That information would have been helpful 205
13.8 Hunt down reports 206
13.9 Treating something that is not there 207
13.10 Going on a search mission 207
13.11 Getting by with a little help from your friends 209
13.12 Can we see what we need to? 209
13.13 Only seeing what you want to see 210
13.14 More at stake 211
viii List of cases

13.15 Did not take much to tip over 212

13.16 Contrast dye is a prescription medicine 213
About the authors

Dr Anh Vu Nguyen was born in Saigon, Vietnam and grew up in Tampa, FL. He
attended the University of South Florida as a National Merit Scholar and gradu-
ated summa cum laude with a Bachelor of Science in Engineering degree. He then
attended the University of South Florida College of Medicine and graduated in
1996. Dr A Nguyen completed his family practice residency at Bayfront Medical
Center in Saint Petersburg, FL in 1999 and became board-certified through the
American Board of Family Practice. He then moved to Tallahassee, FL. Since then,
he has been a full-time emergency physician at Tallahassee Community Hospital
(now Capital Regional Medical Center), serving as associate director from 1999–
2001. From 1998–2003, Dr A Nguyen also worked part-time as an urgent care
provider in Saint Petersburg, FL and in Tallahassee, FL and also as a medical and
legal consultant for the Florida Department of Health. In 2003, Dr A Nguyen
began serving as a part-time emergency physician at the Bay Pines Veterans
Administration Hospital in Saint Petersburg, FL. He obtained board certifica-
tions in ambulatory medicine in 2003 through the American Board of Ambulatory
Care and in emergency medicine in 2004 through the American Association of
Physician Specialists. He has written numerous articles for Consultant and Patient
Care medical journals. Dr Nguyen became a clinical assistant professor at the
Florida State University College of Medicine in 2004.
His hobbies include reading, traveling, movies, and the martial arts. He is a
second degree black belt from the WTF in Taekwondo and is currently an assist-
ant instructor for the FSU Taekwondo club.

Dr Dung Nguyen is the older brother of Dr A Nguyen and was also born in
Saigon, Vietnam. He grew up in Tampa, FL and attended the University of South
Florida as a National Merit Scholar and graduated magna cum laude with a
Bachelor of Science in Engineering degree. He then attended the University of
South Florida College of Medicine and graduated in 1994. Dr D Nguyen com-
pleted his family practice residency at Tallahassee Memorial Hospital in
Tallahassee, FL in 1997 and became board-certified in family practice. He has
been in private practice in Tallahassee since finishing residency. His practice
includes both family medicine and urgent care medicine. Dr D Nguyen became
board-certified in ambulatory medicine in 2003. From 1999–2001, he served as a
medical consultant for the Office of Research and Practice at the University of
South Florida. Dr D Nguyen has also been a clinical instructor for the Florida
State University College of Medicine since 2002. He has been published in Patient
Care medical journal.
His hobbies include sports cars, music, traveling, and the martial arts. He is a
third degree black belt from the WTF in Taekwondo and is currently an assistant
instructor for the FSU Taekwondo club.
Dedications

Our parents, Lan T and Nham T Nguyen, for their tireless devotion to the welfare
and education of their children.
Our grandmother, Ca T Nguyen, for the things that she taught us that could
not be learned in school.

In appreciation of
We would like to give thanks to Frank J Edwards MD and Steven M Selbst MD.
They graciously allowed us to use their cases and gave us encouragement.
We are especially indebted to Jennifer Steimle MD and Shirley Swanson RN for
their help in reviewing the book.
Introduction

Why write a book on medical errors and medical

malpractice?
In today’s society, it is difficult to pick up a newspaper or watch a television
show without seeing an ad or a commercial concerning the current malpractice
crisis that we have in the United States. Indeed, medical malpractice has become
an issue of concern for more than the physicians and plaintiff attorneys.
Insurance companies are either ceasing to offer medical malpractice insurance
or charging exorbitant amounts in order to cover the enormous amounts awarded
by the courts. Politicians are receiving pressure from the medical community to
pass reforms limiting ‘pain and suffering’ damages. Hospital administrators
must deal with many physicians forfeiting their privileges and not providing
emergency room coverage. Furthermore, hospitals are being forced to revise
their by-laws to keep physicians who are now going ‘bare’ because they cannot
afford malpractice insurance. Finally, the most important part of our society –
the general public – has been and will continue to be affected, as access to
healthcare, and particularly specialists, becomes limited. Therefore it is easy to
understand why any literature that addresses this topic would be of paramount
importance.

Should we practice ‘defensive medicine’?

This is an extremely difficult question to answer. The general conception among
the public, and even among some healthcare providers, is that defensive
medicine is utilization of ‘every test and consultant available’ in order to ‘not
miss something.’ The term ‘CYA’ – ‘cover your ass’ – has sometimes become
synonymous with defensive medicine. In fact, some may even believe that
defensive medicine restricts the ‘thinking process’ and is ‘bad medicine.’ In this
sense, defensive medicine raises the cost burden of healthcare and should not
be practiced.
However, defensive medicine also means defending your patient from
potential harm to life and limb based on her complaints. This harm can come
acutely (acute appendicitis), subacutely (stable angina becoming unstable), or
even years from the initial visit (development of breast cancer). For the prudent
physician, this means getting a detailed history, performing a thorough exam,
formulating a differential diagnosis, ordering the appropriate tests, giving the
best treatment, and ensuring that proper instructions and follow-up are given.
It does not necessarily mean that more tests and increased healthcare costs are
going to occur with every patient. We give many examples in the book where
this is not the case. Utilizing telephone consultations with specialists, peer
2 Learning from medical errors: clinical problems

review of charts, and patient callbacks are some methods that are not costly or
regarded as ‘bad medicine.’
Furthermore, defensive medicine entails that the physician constantly
contemplates in her mind what she might be missing or what unexpected out-
come could result from each decision that she makes. This is not to say the
physician should practice medicine in paranoia of making a mistake; instead,
it is a reminder for the physician to stay alert and broad-minded during the
decision process. This concept also applies to and should be reflected in the chart
documentation. To give an example used by many other authors, the physician
should frequently imagine herself in the courtroom with the plaintiff lawyer
asking why she did or did not do something or if she had ever considered the
‘other’ possibilities of the problem. Alternatively, the physician can picture herself
in the courtroom staring at her progress note enlarged to a 24 by 36 inch poster
and having to explain it.
Once again, we are not advocating that physicians should practice in fear to
avoid the courtroom; we are only encouraging our colleagues to be aware of the
ramifications of their actions and decisions. This does not necessarily equate to
ordering more tests or consultations but it does mandate that examinations and
documentations are more thorough and complete in thought. In this manner,
defensive medicine not only benefits the physicians legally, but more
importantly, it serves the best interests of the patients by guiding physicians to
honor one of their Hippocratic oaths – ‘do no harm.’

Goals of this book

We believe that there has never been a better time for a book like this to be
published. It is written with the intention of providing physicians with a guide
for performing and documenting medicine that will decrease the chances of a
poor outcome for the patient and for the physician in the courtroom. Physicians
should take a proactive approach clinically to avoid contributing to the medical
malpractice dilemma. Although there will always be, and always should be,
medical malpractice cases in the United States, we believe that physicians will
be able to decrease the nontrivial and especially the trivial cases. It is our goal
to reach primary care providers, urgent care providers, and emergency
providers because their clinical scenarios are where we have based our
experiences. However, we believe other groups that may derive benefit and
foresight from this book are the senior medical students and young medical
residents. They are the ones that we can inform and influence early in their
careers about the realities of current medical practice with the goal of instilling
good practice habits.
Our hope is to provide the reader with a mindset to practice thorough and
sound medicine and minimize medical liability. Although it is not feasible or
recommended to use every concept that we have included in this book on each
patient encounter, it is left to the reader’s discretion as to what would be best
for each particular patient. However, remember that you will usually not have
the opportunity to make additions at a later time. We have included methods
and techniques that we have learned from years of clinical practice in various
different clinical settings (e.g. private office, private emergency department,
government emergency department, academic residency centers, and as legal
 Introduction 3

consultant for the State Board of Health). In addition, we have a great interest
in medical-legal literature that, as of the printing of this book, is still scarcely
authored by physicians. From our enthusiastic reading of medical-legal cases,
we have tried to incorporate interesting and educational cases into this book to
illustrate our concepts. We must also credit our colleagues through the years
with sharing their stories (some of which we have used in this book).
We hope that our readers enjoy reading this book as much as we enjoyed
writing it. Although the topic of medical malpractice is not a favorite subject of
discussion among physicians, we hope that this book will generate an increase
in physician interest and discussions concerning legal medicine. Therefore, we
encourage our readers to provide us with feedback on our concepts and send
us additional techniques that they may use in their own practice.
This book focuses on medical complaints that tend to have high medical
liability. We use a standard format for working through these complaints, begin-
ning with triage and ending with diagnosis and course of treatment. We conclude
each chapter with sections on pitfalls, errors, and interesting cases for each
complaint. Pitfalls are variants from the ‘normal’ that may lead to medical errors
if not suspected or detected. The ‘Errors and interesting cases’ section is
comprised of clinical summaries that we have witnessed through our personal
experience or read through medical malpractice literature. Details of the cases
may have been altered to ensure confidentiality.
We conclude with a chapter concerning legal issues involving radiology. Most
physicians do not have access to a radiologist’s expertise at all times in their
practice and it is difficult to consult with one via telephone concerning X-rays.
Therefore, we sometimes must act as radiologist and base critical treatment
decisions on our own interpretations. This creates a potentially high liability
situation. Furthermore, radiology studies are not innocuous and the risks
involved must be considered.
As a final note, the companion to this book, Learning from Medical Errors: legal
issues is also published by Radcliffe Publishing. This book discusses medical
errors and legal problems involved in medical documentation and medical
practice. A case presentation and discussion format is also used.
This page intentionally left blank
Chapter 1

Abdominal pain

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Abdominal pain is the single ‘most frequently mentioned’ complaint among

emergency department visits in 1996.1 It is also the fourth most common
diagnostic code for patients seen in emergency departments throughout the
United States in 2001.2 We prefer to think of the abdomen as ‘Pandora’s box’ due
to the multiple different organs within it and the numerous emergent conditions
in which they may present. In addition, abdominal emergencies may present
regardless of the age of the patient, although they are more frequent in the
elderly population. This is in contrast to, say, chest and head emergencies, which
occur largely in the adult population.
As expected from the anatomy, the etiologies for acute abdominal pain in the
emergency department is broad and ranges from benign, self-limiting conditions
to surgical, life-threatening emergencies. The most common diagnosis is
‘abdominal pain of unknown etiology’ and is made in 41% of patients.3 This is
followed by (in descending order of frequency): gastroenteritis, pelvic
inflammatory disease, urinary tract infection, ureteral stone, appendicitis,
acute cholecystitis, and intestinal obstruction. 3 However, in the geriatric
population, nonspecific abdominal pain (10%) becomes the fourth most common
behind acute cholecystitis (26%), malignant disease (13%), and bowel obstruction
(11%).3 In the pediatric population, acute abdominal pain is largely composed
of a dichotomy of nonspecific abdominal pain (62%) and acute appendicitis
(32%).3

Triage
Three types of patients with acute abdominal pain should be seen and treated
as soon as possible while another should be treated urgently. In general, patients
with surgical conditions (e.g. incarcerated hernias with strangulation, bowel
obstruction), infectious conditions (e.g. acute appendicitis, acute cholecystitis,
diverticulitis), and vascular conditions (e.g. ischemic colitis, ruptured
abdominal aortic aneurysm) should be seen emergently. Patients with
6 Learning from medical errors: clinical problems

intolerable pain (e.g. kidney stone, pancreatitis, urinary retention) should be

seen urgently. In addition, patients who have abnormal vital signs (e.g. hypo-
tension, fevers, tachycardia, or low pulse oximetry) or appear in moderate distress
(e.g. diaphoretic, cyanotic) should be seen immediately.

History
Obtaining the history for abdominal pain is much like obtaining the history for
pain in any other part of the body. The patient should be questioned on the onset
and duration of pain, the location and movement of pain, the quality and
severity of pain, and the associated symptoms.4 Sudden onset of pain is more
frequently associated with surgical conditions (e.g. perforated viscus) and
vascular conditions (e.g. aortic dissection) while an insidious onset of pain is
more commonly found in infectious conditions (e.g. appendicitis). The patient’s
description of the location of the pain must take into consideration the three
types of abdominal pain: visceral, somatic, and referred.
Visceral pain is generated from the abdominal organs, tends to be poorly
localized, and often precedes somatic pain. It is often accompanied by auto-
nomic responses such as nausea, vomiting, and diaphoresis.5 Although visceral
pain is difficult to attribute to a particular organ, several generalizations may be
made. Liver, stomach, gallbladder, and duodenum pain often manifest in the
epigastrium.5 The small intestine, appendix, and cecum have a tendency to refer
pain to the periumbilical area.5 Finally, the hypogastrium is the site of pain from
the colon, kidneys, ureters, bladder, and pelvic organs.5
Somatic pain is generated from the parietal peritoneum and the anterior
abdominal wall. It is conducted by nerve fibers that return to the spinal cord via
specific dermatomes that correspond to the anatomic site from which the pain
arises.5 This pain tends to occur later than visceral pain as the inflammation,
ischemia, or distention of the abdominal organ encroaches on the somatic parts
of the abdomen. The location of somatic pain more precisely reflects the organ
causing the pain.
Referred pain is due to the overlap of neural pathways from the cutaneous site
at which the pain is felt.5 Diseased abdominal organs may refer pain to extra-
abdominal sites and extra-abdominal organs may refer pain to abdominal sites.
The pain from cholecystitis and pancreatitis may be felt in the middle of the back
whereas the pain from salpingitis or cystitis may be felt in the mid-lower back.
Ureteral colic is often described in the inguinal or genital area; and biliary colic
is sometimes described in the shoulder area. Myocardial infarctions can present
as pain in the epigastric area while pneumonias and pulmonary emboli may
cause pain in the upper abdominal quadrants.
Associated symptoms often help with the diagnosis and evaluation of
abdominal pain. Nausea and vomiting that occur after the onset of abdominal
pain is more frequently associated with a surgical condition. In contrast, pain
after nausea and vomiting usually represents acute gastroenteritis.5
Past medical history is important because patients with abdominal disorders
often have recurrent bouts of similar pain (e.g. peptic ulcer disease, biliary colic,
diverticulitis, etc.). If they have a prior history of an abdominal disorder, ask
them what treatments have worked in the past, what complications they have
had from their disorders (e.g. bleeding from an ulcer), and how they have been
 Abdominal pain 7

treated (e.g. hospitalization, surgery). For patients with uncertain but recurrent
pain, ask them what tests they have had and obtain the results, if possible. Inquire
about medical disorders that may have manifestations in the abdomen such as
diabetes (diabetic ketoacidosis, gastroparesis), coronary artery disease (myocar-
dial infarction), and COPD (pneumonia). Medications often cause
abdominal discomfort (e.g. aspirin, anti-inflammatories, and steroids with gas-
tritis; numerous medications that may flare porphyria) and should be listed in
the patient’s chart.
Finally, we feel that the most important aspect of the past medical history
pertaining to abdominal pain is the surgical history. We have discussed
previously the association of previous abdominal surgeries with the occurrence
of small bowel obstruction. In addition, awareness of any organ previously
removed results in a narrowing of the differential diagnosis. In some cases,
additional possible diagnoses may exist in patients with prior surgeries
(e.g. adhesions with prior surgeries, aortic-intestinal fistula with abdominal
aortic aneurysm repair). Remember also to include any pelvic or gynecologic
surgeries.
Social history is also an integral component in the evaluation of abdominal
pain. Cigarette smoking is an important component of abdominal pathology
(e.g. pancreatic cancer, bladder cancer), although its effects are usually more
chronic. In contrast, alcohol abuse may present with both chronic manifestations
(e.g. liver cirrhosis, ascites) and acute manifestations (e.g. gastrointestinal bleed-
ing, pancreatitis). Travel history should also be obtained in certain situations
(e.g. suspected amoebiasis or hepatitis A).
For females with lower abdominal pain, we include a gynecologic history.
Frequent questions that we ask include last menstrual period (timing with pelvic
inflammatory disease, probability of pregnancy), history of ovarian cyst,
endometriosis, or fibroids (causes of lower abdominal pain), and history of
sexually transmitted diseases (frequency of pelvic complications).
Our standard inquiries in review of systems for acute abdominal pain include:
fevers, nausea, vomiting, diarrhea, dysuria, hematuria, vaginal bleeding,
vaginal discharge, and blood in the stool. We add weight loss, back pain, and
other constitutional symptoms for chronic abdominal pain because of the
increased incidence of malignancy.

Physical exam
The essential five vital signs (blood pressure, temperature, pulse, respiratory
rate, and pulse oximetry) must be obtained in every patient with abdominal
pain. New nurses commonly inquire about the necessity of pulse oximetry and
respiratory rate for abdominal pain. In addition to our previous discussion of
pneumonia as a common extra-abdominal cause of abdominal pain, intra-
abdominal processes may hinder the function of the respiratory system.
Common examples are found in ascites and small bowel obstruction where the
abdominal distention decreases the lung’s capacity. There also seems to be a
misconception among some nurses and medical assistants that temperature is
only important for upper respiratory symptoms and pulmonary symptoms.
Therefore, they may not obtain a temperature for the patient with abdominal
pain.
8 Learning from medical errors: clinical problems

In the Introduction, we emphasized the importance of the General Section on

the physical exam to future readers of the chart. The description of the patient
with abdominal pain often clues the reader to the presence of a serious
abdominal process. The colicky or screaming patient may be suffering from
ureterolithiasis or acute pancreatitis. In contrast, the patient who lies perfectly
still but is grimacing may have acute peritonitis. However, the patient who is
laughing with his friends and eating potato chips is very unlikely to have a
significant abdominal disease.
The gastrointestinal examination should be performed with the patient fully
undressed from the chest to the thighs. Make use of all of the basic skills that
you were taught in physical diagnosis class. Inspect, auscultate, palpate, and
percuss the abdomen if necessary. From our experience, there are certain find-
ings on the abdominal exam that should not be ignored without further diagnostic
testing or observation. These include guarding, rebound, rigidity, palpable
masses, abdominal distention, and positive provocative physical signs (Murphy’s,
iliopsoas, obturator, Rovsing’s, Cullen’s, and Grey-Turner’s).
We recommend that a genitourinary exam in males and a pelvic exam in females
be performed in all patients with lower abdominal pain. Please see Case 1.8. In
addition, a rectal exam should be performed in all patients with abdominal pain.
Although we realize that the yields from these exams are frequently low, they
can sometimes present additional useful information (hernia, pelvic mass, etc.).
The exams do not take much additional time, are safe to the patient, and do not
cost additional money. Therefore, save yourself the hassle of answering to the
court in a malpractice suit why you did not perform these simple tests, which
might have prevented an incorrect diagnosis. Please see Case 1.2. Furthermore,
you will not look like a cost-effective physician if you order expensive tests such
as CT scans, ultrasounds, and colonoscopies before performing a simple com-
plete physical examination. In the current managed care environment, being
cost-ineffective is often associated with not being a good physician.

Differential diagnosis
We approach the differential diagnosis systematically by considering the local
disease processes. Then we expand our consideration to distant disease
processes. We also start with more emergent conditions. Then, we proceed to
more benign conditions. We feel that this systematic approach is consistent
because, as we discussed in the ‘History’ section, a disease process usually
becomes more localized as its severity progresses. Our approach also forces us
to consider the more serious etiologies before entertaining the benign ones. To
reiterate, there is little harm in missing a benign diagnosis. However, there are
grave consequences for missing an emergent diagnosis.
For females with lower abdominal pain, remember to include gynecologic
etiologies. Extra-abdominal sources should be considered such as pneumonia,
diabetic ketoacidosis, and porphyria. In those with coronary artery disease or
peripheral vascular disease, keep vascular etiologies in mind. These may include
aortic aneurysm, aortic dissection, mesenteric ischemia, and superior mesenteric
venous thrombosis. These latter diseases are frequently not considered in the
evaluation of abdominal pain because they are not regarded as part of the
gastrointestinal organs.
 Abdominal pain 9

Diagnostic tests
For patients with no alarming symptoms or significant findings on the history
and physical exam and normal vital signs, laboratory testing and radiology
imaging are probably not needed. The majority of these patients will harbor a
benign, self-limited diagnosis such as gastritis, gastroenteritis, or functional
abdominal pain. Similarly, the patient who presents with a surgical abdomen
(e.g. rebound tenderness, tender and rigid abdomen) or the patient with
abdominal pain and hemodynamic instability (e.g. hypotension) may not require
any testing. Immediate surgical consultation is more appropriate in these
situations. However, many patients with abdominal pain will not satisfy all of
the above criteria and will require additional testing.
We recommend a complete blood count in any patient with a suspected
bacterial infection (e.g. appendicitis, cholecystitis), blood loss (e.g. peptic ulcer
disease), or vascular injury (e.g. abdominal aortic aneurysm rupture). Although
a normal white blood cell count does not rule out serious infection, it is often
used as reassurance in combination with other reassuring factors on the evalu-
ation (e.g. nontender abdomen, no fever). A mildly elevated white blood cell
count (10,000 cells/microliter to 12,500 cells/microliter) creates a gray zone for
the physician. This could result in high liability situations (see Case 1.3). A
moderately elevated white blood cell count (greater than 12,500/microliter) is
a marker for alarm that future action is usually needed (e.g. observation, radio-
logical imaging, repeating white cell count after fluids). A white blood cell count
differential that shows a left shift (i.e. elevated percentage of bands), however,
is much more specific for a bacterial infection. This should be alarming to the
physician for the possibility of a bacterial infection regardless of the absolute
white blood cell count.
A basic metabolic panel is required for patients with moderate vomiting or
diarrhea and suspected electrolyte imbalance. Drug levels should be checked
due to the inclination of drug toxicities to affect the gastrointestinal system (e.g.
Dilantin toxicity with nausea and vomiting). Consider obtaining liver function
tests and lipase for patients with upper abdominal pain and periumbilical
abdominal pain. A urinalysis should be obtained in patients with lower abdomi-
nal pain or urinary symptoms.
Pregnancy testing should be considered for all females of child-bearing
capacity. These include those who have had a bilateral tubal ligation or those
who are using a form of contraception. This is because of the known failure rates
of these methods of contraception. We have seen females that are ‘not sexually
active’ become pregnant. A common misconception we hear is that ‘I can’t be
pregnant because I’m on my period.’ In general, the only exceptions where
pregnancy is absolutely impossible are the premenopausal and post-
menopausal patients and the patients who have had a hysterectomy (although
we present an exception, Case 1.6).
Additional testing for extra-abdominal causes of abdominal pain is warranted
in certain patients. It is recommended in Tintinalli’s Emergency Medicine: a
comprehensive study guide that an electrocardiogram (EKG) be performed
for patients over 50 years of age and for those with a history of coronary
artery disease.6 Please see Case 1.14. Consider a chest X-ray for patients
with a low pulse oximetry or concurrent respiratory symptoms. Pelvic
10 Learning from medical errors: clinical problems

cultures may be appropriate in sexually active females with lower abdominal

pain.
Unlike chest pain where radiology studies rarely contribute to the diagnosis,
radiology studies can be an integral part in the evaluation of abdominal pain.
Although some authors feel that plain films do not contribute much in the
evaluation of abdominal pain, we find them useful for two particular situations:
bowel obstruction and perforated viscus. Therefore, we typically will obtain
abdominal films in patients with abdominal pain who have had previous
abdominal surgeries, history of hernias or malignancies, abdominal distention,
or peritoneal signs.
The CT scan has made a great contribution in the evaluation of acute abdomi-
nal pain. It has become the test of choice for certain acute abdominal conditions.
These include appendicitis, diverticulitis with or without abscess, abdominal
aortic dissection, and colitis. It can also be used for bowel obstruction (when
plain films are nondiagnostic), abdominal trauma (for the stable patient with
possible viscus hematoma or laceration), cholecystitis (although ultrasound is
better), and complications of pancreatitis (e.g. pseudocyst). We recommend the
consideration of CT scanning for any of the above conditions (except if ultra-
sound is available for cholecystitis). These conditions usually are associated with
either abdominal tenderness or an elevated white blood cell count. Therefore,
we have a very low threshold for ordering a CT scan for patients with either of
these two criteria.
Ultrasound scanning has also developed an increasing role in the evaluation
of acute abdominal pain. Ultrasound machines are found with increasing
frequency in emergency departments across the country and emergency
physicians are receiving more ultrasound training through their residencies.
There are many ultrasound courses and some organizations that will come to
your facility to deliver hands-on training. Ultrasound has a few distinct
advantages over CT scanning in the emergency department. It is quicker
and can be performed in the emergency department. Hence, it is safer for
the unstable patient. It does not involve the radiation exposure that CT does,
which makes it particularly useful for young children and pregnant females.
Finally, it provides enhanced sensitivity for acute cholecystitis and pelvic
pathology.
Physicians must also be aware of the limitations of ultrasound before utilizing
its results in their treatment decisions. Results are partly dependent on the ability
of the ultrasound technician. They also may be dependent on the patient’s size
(heavier patients are more difficult to scan) and when the patient’s last oral intake
was (distended bowels create artifact on the scan). In addition, the physician
must be clear what he is looking for because the technician is likely to perform
a much more limited exam than with a CT scan. For example, if the physician
was looking for ovarian torsion and simply ordered an ultrasound without color
Doppler flow, he would likely get an ultrasound report that cannot comment on
whether there is torsion of the ovaries.

Hospital/office course
Many patients who present with complaints of abdominal pain will receive
intravenous fluids and/or medications for pain and nausea. The traditional belief
 Abdominal pain 11

that pain medicines should be withheld to prevent the masking of the pain is
no longer popular among surgeons and emergency physicians. Patients with
possible surgical conditions should be kept NPO (nothing by mouth) until surgical
conditions are ruled out. Patients should be reassessed after interventions and
their subsequent physical status and exams should be documented. As we
discussed earlier, it is very important to perform multiple exams on the patient
with abdominal pain.
Disposition decisions for patients with abdominal pain can sometimes be
difficult. This is due to the large number of abdominal pains that are of
‘uncertain etiology’ despite a complete evaluation. We find abdominal pain cases
challenging because patients of any age can harbor an abdominal emergency. It
is the extremes of age that usually have atypical manifestations that tend to
result in misdiagnoses and malpractice cases. This brings us to an old but wise
adage in emergency medicine, ‘If in doubt, don’t send them out.’ We add our
own version of this adage, ‘If unsure, don’t show them the door.’
From our experience, we can offer a few general statements regarding the
disposition of patients with abdominal pain. Never discharge a patient with
abdominal pain that is not better after your interventions. Never discharge a
patient that has persistent vomiting despite medications. Be extremely cautious
about discharging any patient with a tender abdomen despite normal
diagnostic tests. Finally, do not hesitate to call the patient’s primary physician
or a surgeon to admit a patient with an uncertain diagnosis that you are
concerned may have an evolving surgical condition. Although the patient may
be given good discharge instructions on when to return, the physician may still
be sued for a delay in diagnosis if the patient suffers a complication (e.g.
appendiceal abscess) (see Case 1.3).

Pitfalls
As mentioned above, be aware of abdominal pain arising from extra-abdominal
disease processes. Common cases such as acute myocardial infarction present-
ing with epigastric pain are not usually missed, although please see Case 1.20
below. Unusual cases such as pneumonia may be missed if not suspected. We
had a patient who presented with right lower quadrant pain and tenderness
with an elevated white blood cell count. The patient had distended loops of
small bowel on an abdominal X-ray. The abdominal X-ray also managed to show
a left lower lung infiltrate. The subsequent chest X-ray confirmed the
pneumonia and the patient’s abdominal pain was a result of an ileus secondary
to the pneumonia.
Do not forget that the pelvis is considered by some to be extra-abdominal. It
is not uncommon for pelvic diseases to present as abdominal pain. For example,
the presentation of appendicitis is sometimes similar with that of pelvic inflam-
matory disease. Please see Case 1.18 below. Therefore, the management of
abdominal pain will sometimes require a multi-specialty approach.
There is overuse of acute gastroenteritis and urinary tract infection as
diagnoses of acute abdominal pain. Although these two entities are fairly
common, they should not prevent a more extensive work-up if there are red
flags present. Red flags include fever, hypotension, blood in the stool, weight
loss, and any significant abdominal tenderness on physical exam. Our
12 Learning from medical errors: clinical problems

experience has convinced us that acute gastroenteritis and urinary tract

infections rarely cause abdominal tenderness. It is written in Rosen’s Emergency
Medicine: concepts and clinical practice: ‘the EP [emergency physician] should resist
the temptation to assign a diagnosis of gastroenteritis as a catchall to patients
with unexplained abdominal pain. To do so may divert the attention of the patient,
and of other physicians, from a more serious, possibly surgical, disorder.’7 Please
see Case 1.7 and Case 1.16 below.
As a similar thought, remember that if it looks and smells like appendicitis,
it probably is. Do not try to convince yourself that it is gastroenteritis or urinary
tract infection. This common mistake has led to many medical errors,
malpractice lawsuits, and large financial settlements. Please see Case 1.15 below.
Similarly, not all dysuria and urinary frequency represent the presence of a
urinary tract infection. These symptoms are sometimes a result of irritation of
the urinary system from the inflammation of abdominal or pelvic organs.
For example, irritation of the ureters may occur with a retroileal appendix
while a pelvic appendix may cause irritation of the bladder.8 Furthermore, the
mere presence of a urinary tract infection does not preclude the coexistence of
a more serious abdominal disorder. Finally, many urine collections are not
performed under sterile conditions and may be contaminated with epithelial
cells.
Excessive reliance on diagnostic test results can sometimes lead to poor
disposition decisions if they are not consistent with the history or physical exam.
Selbst and Korin offer a tip in Preventing Malpractice Lawsuits in Pediatric
Emergency Medicine for cases of suspected acute appendicitis. ‘Consider
admission or surgical consultation if the patient exhibits two of these three
characteristics: classic history, suspicious physical examination, or abnormal
diagnostic test results.’9 This advice is based on the fundamental of looking at
the overall picture of the evaluation and not just on one aspect of it. Please see
Case 1.1 in the following section for a good example of this fundamental.
As discussed in the ‘Diagnostic tests’ section above, physicians must be aware
of the limitations of the tests that they order. Without knowledge of this, they
may make disposition errors based on ‘normal results’ from a suboptimal test.
An excellent example is found below in Case 1.4.
In children and the elderly, abdominal pain can present with very atypical
symptoms. Likewise, their physical exam signs are also frequently not of a
classic nature and diagnostic testing in them is often difficult to interpret or
perform (e.g. ultrasounds have relatively high false negative rates and it is
difficult to administer intravenous and rectal contrast in children).
‘Appendicitis remains a commonly misdiagnosed entity in children. It is
initially missed in up to 60% of children younger than six years old. Acute gastro-
enteritis is the most common misdiagnosis.’10
Abdominal pain in the elderly deserves special consideration. The elderly are
more susceptible to many more age-related disease processes in the abdomen.
These include abdominal aortic aneurysm, aortic dissection, mesenteric ischemia,
mesenteric venous thrombosis, cholecystitis, diverticulitis, and ischemic colitis.
In addition, they have other chronic illnesses such as diabetes, coronary artery
disease, and peripheral vascular disease that hamper their ability to tolerate
even minor disease processes in the abdomen. Therefore, the prudent physician
will maintain a lower threshold for ordering tests (including imaging), observ-
 Abdominal pain 13

ing, and admitting elderly patients with abdominal pain. Our emergency phy-
sician group recommends the consideration of radiological imaging and
admission for any elderly patient with unexplained abdominal pain. We also
require that all elderly patients with abdominal pain receive a repeat examination
prior to disposition. Please see Case 1.9 below.
Ischemic colitis and mesenteric ischemia are two of the most difficult and
elusive diagnoses to make in patients with abdominal pain. Both can produce
abdominal pain that is diffuse and poorly localized. The pains are frequently out
of proportion to the physical exam and can be intermittent. Finally, abdominal
CT (the usual best imaging modality for abdominal pain) has relatively poor
sensitivity for both processes. In the case of mesenteric ischemia, ‘timely diag-
nosis requires that an angiogram be obtained very early in the evolution of the
pathologic process – so early, in fact, that it may seem clinically premature to
order such an invasive test on an elderly patient who may not appear ill.’11 For
ischemic colitis, the sensitivity of a contrast-enhanced CT is 71%, whereas the
sensitivity of angiography is still only 88%.12 We present our personal difficulty
with these diagnoses in Case 1.5 below.
Patients with abdominal pain and fever who have foreign bodies in their
abdomen or pelvis (e.g. tampons, intrauterine devices (IUD), pacemakers, etc.)
should be considered as having sepsis until proven otherwise. Intravaginal and
intrauterine devices are common and are good harbors for bacteria and
infection in the proper setting (i.e. incorrectly placed, left in too long, etc.). Toxic
shock syndrome and pelvic inflammatory disease from foreign bodies are
two well-described phenomenon in the medical literature. These patients
require a lower threshold for admission and intravenous antibiotics. Please see
Case 1.3.
Presentations are not always ‘textbook’ ones and histories are not always
‘classic’ ones. An old but very true saying in medicine is that ‘patients don’t
always read the book.’ Therefore, keep an open mind and be alert when
the patients present with atypical symptoms. Disease processes that are
common to certain patient populations should still be entertained even if the
presentation is not as expected or test results indeterminate. Please see Case 1.10
below.
Mallett, in Emergency Care of the Woman, makes a very true and profound
statement: ‘Any woman of reproductive age presenting with pelvic pain or
bleeding should have ectopic pregnancy ruled out.’13 We have heard many
emergency department (ED) physicians and many radiologists say that a patient
cannot have an ectopic pregnancy because she ‘is not having abdominal pain or
vaginal bleeding.’ The problem lies in the fact that these two characteristics are
typical for a ruptured ectopic pregnancy and does not apply for ectopic preg-
nancies that have not ruptured. This misconception may be reflected in Brennen’s
comment: ‘even with current diagnostic methods available, the diagnosis is
missed in 50 percent of cases at the first office visit and in 36 percent at the time
of the first emergency department visit.’13 Mallett also states that ‘lack of pain
is not useful if the goal is early diagnosis and tubal preservation prior to tubal
rupture.’14 Please see Case 1.11 below.
Pain in the right lower quadrant is not always appendicitis. Most physicians
are aware of this fact and also aware of the significant number of false negative
diagnoses of appendicitis that are made. There are less frequent causes of
14 Learning from medical errors: clinical problems

abdominal pain in this area including Crohn’s disease, pelvic inflammatory

disease, and ovarian cyst. In addition, there are some rare disease processes
that also tend to produce pain in the right lower abdomen. Please see Case 1.12
below.
Acute cholecystitis is described in almost every medical text as consisting of
fevers, nausea, vomiting, and right upper quadrant pain. Although this descrip-
tion is very accurate in most cases, we have seen a number of cases where the
pain is in the epigastric area. Our opinion is that epigastric tenderness in a patient
that is fat, fertile, forty, and female (4 ‘F’s) should be considered as cholecystitis
until proven otherwise. Please see Case 1.13 below.
Abdominal pain in children should be evaluated with caution because they
often have atypical symptoms. Children have poor ability to localize pain and
this makes the assessment difficult. Furthermore, children will often describe
testicular processes as abdominal pain. Testicular disease processes are common
in the second decade of life. Children and adolescents are less likely to volunteer
that there is pain or swelling in their scrotum. This may be due to the child’s fear
of having a genitourinary exam. Therefore, we agree with Selbst’s comment in
Preventing Malpractice Lawsuits in Pediatric Emergency Medicine that ‘the physical
examination of a boy with abdominal pain must include careful inspection of
the genitalia.’15 Please see Case 1.17 and Case 1.19 below.

Errors and interesting cases

Case 1.1 The importance of clinical suspicion

• Our first case highlights a few of the above statements. A 30-year-old
female presented to us in the emergency department with four days of
right lower quadrant pain. She did not have any associated symptoms,
had never been sexually active, and had no history of gynecologic
diseases. She did not have a fever on presentation but had localized
tenderness in the right lower quadrant with guarding and no rebound.
Her white blood cell count was mildly elevated. Other emergency tests
included a negative pregnancy test and a CT of the appendix. The
radiologist read the CT scan as normal. She was observed in the emer-
gency department for four hours. Her pain did not improve during this
time and her abdominal exam was unchanged. We felt that the
combination of increased white blood cell count and remarkable exam
findings warranted an admission to surgery for observation. In addition,
she lived with her elderly mother who had just received a cardiac stent
two days ago and was told that she could not drive. Therefore, the patient
had a low likelihood of returning in a timely fashion if she was dis-
charged and her condition worsened.
• We placed a call to the on-call surgeon who was one of the older surgeons
on the medical staff and did not regularly take emergency call but was
covering for another surgeon that night. After we explained our find-
ings, concerns, and the patient’s social situation, the elderly surgeon

continued
 Abdominal pain 15

stated: ‘this patient does not have a surgical problem and I don’t admit
patients to observe them or for social reasons.’ He emphasized that the
CT scan was normal so we should send the patient home. We reiterated
that we have had patients with appendicitis and normal CT scan. The
surgeon boldly claimed, ‘I have been treating patients twice as long as
you and this patient does not have anything wrong with her.’ After we
held our ground, the surgeon finally acquiesced and admitted the
patient. The patient’s mother came to see us three days later and said
that her daughter had surgery for a ruptured retroileal appendix 24 hours
after admission.
• Patients with appendicitis will occasionally present with minimally
abnormal labs and negative CT scans. The ability to detect appendicitis
by CT scan is decreased if the appendix is not located in its usual location.
The disposition is then left to the physician’s clinical judgment. It is
certainly not feasible to admit every patient for observation. Likewise,
it is not prudent to discharge every patient with instructions to return
if her pain worsens. The physician should assess the social situation and
determine the likelihood that the patient will be able to return in a timely
fashion. Consultation with a surgeon is recommended. There will be
times when you must hold your ground and request that the surgeon
come evaluate the patient. Finally, as stated the ‘Pitfalls’ section, the
combination of an elevated white blood cell count and an impressive
abdominal exam made our decision to admit this patient.

Case 1.2 Did you bother to look?

• Another case from the emergency department demonstrates the
importance of performing a pelvic exam in females with lower abdomi-
nal pain. A 40-year-old female with a history of uterine fibroids for several
years presented to the emergency department complaining that the same
lower abdominal pain had become worse. She had repeatedly refused
a hysterectomy because of her desire for childbirth. However, on this
visit, she was determined that she wanted her uterus removed because
the pain was unbearable. The physician on duty offered her pain
medicine and a gynecologic follow-up for her request. The physician did
not feel it was necessary for any further work-up for this chronic
problem.
• The patient went into cardiac arrest at home two days later and died in
the ambulance before arriving at the hospital. An autopsy concluded
that she died of sepsis from a pelvic infection. The treating physician was
faulted for failing to perform a pelvic examination and blood tests.
Although it is questionable whether these tests would have resulted in
a difference in treatment, it was easy to retrospectively blame the
physician for not performing simple tests. However, it is, likewise, easy
to criticize a physician for ordering an evaluation for a ‘chronic medical
problem.’
16 Learning from medical errors: clinical problems

Case 1.3 Should we look for zebras?

• A 29-year-old female went to her physician’s office for left lower quadrant
pain. The pain had been present for 10 hours. Her history was significant
for an intrauterine device placed eight months ago. She had nausea but
no fevers, vomiting, or diarrhea. Her vital signs were within normal limits
with the exception of a blood pressure of 90/60. The physician performed
a physical examination and found tenderness in the left lower quadrant
without guarding or rebound. There was also tenderness in her left
adnexa. A stat complete white blood cell count was slightly elevated at
11 500 cells/microliter. The patient’s urinalysis and urine pregnancy test
were negative.
• The physician then arranged for the patient to have an outpatient CT
scan of the abdomen and pelvis. The radiologist read the CT scan as
negative. The physician called in some antibiotics for the patient and
treated her empirically for diverticulitis. He instructed her to return the
following day for a recheck. The patient did not return the next day.
Instead, she went to an emergency department three days later. She now
had a temperature of 103.4ºF and was found to be in sepsis. A repeat CT
scan in the hospital showed that she had a thrombosis in her ovarian
vein. The patient endured a long and stormy hospital course.
• A lawsuit was filed against the clinic physician and the radiologist who
read the first CT. Although the case is pending, we felt that the clinic
physician’s care was meticulous and within the standard of care. His
disposition decision was based on a CT report that was inadequate. In
retrospect, the thrombosis was also present on the first CT. With the
exception of mild hypotension, she did not clinically appear septic.
Furthermore, her abdomen did not feel ‘surgical.’ He had also given the
patient strict instructions on when to return for re-evaluation. These
instructions were not followed. Septic thrombophlebitis is an unfortu-
nate complication of an intrauterine device.

Case 1.4 What is the best test?

• The next case was included to demonstrate the benefits of realizing the
limitations of diagnostic tests. An emergency physician colleague treated
an obese young woman with right upper quadrant tenderness, low-grade
fever, and a slightly elevated white cell count in the emergency depart-
ment at 3 am. Since the ultrasound technician was not in the hospital at
this hour while the CT technician was, the emergency physician ordered a
CT scan for acute cholecystitis without consulting the radiologist. The
scan was read as normal and the patient was discharged. The patient
returned two days later with worsening pain and had an ultrasound,
which showed acute cholecystitis. She was found to have a gallbladder
perforation during surgery. This complication might have been
prevented if a gallbladder ultrasound was ordered as the initial
continued
 Abdominal pain 17

test. With regard to gallbladder diseases, it is stated in Tintinalli’s

Emergency Medicine: a comprehensive study guide that ‘the sensitivity of CT
scanning is insufficient (as low as fifty percent) for it to replace ultra-
sonography as the diagnostic procedure of choice.’16

Case 1.5 The elusive diagnosis

• An elderly lady presented to us five years ago complaining of diffuse
abdominal pain that had been present for over one month. She had not
been to see her primary care physician and denied any fevers, nausea,
vomiting, blood in the stool, or weight loss. There were no cardiac risk
factors and no previous abdominal surgeries. Her physical exam revealed
her to be afebrile and she was noted to have very minimal diffuse
tenderness but no guarding or rebound. Her stool guiac was hemoccult
negative. The emergency department evaluation revealed an elevated
white blood cell count at 16,000/microliter and 3+ bacteria in her urine.
She had an abdomen/pelvic CT scan with intravenous, oral, and rectal
contrast, which was read as normal by the radiologist. Her abdominal
exam did not change while in the emergency department and she was
given the diagnosis of abdominal pain – uncertain etiology. She was
offered admission but chose to go home with the stipulation that her
daughter would bring her immediately back if the pain worsened (along
with treatment for the urinary tract infection). Notice that we were hesitant
to give her the diagnosis of abdominal pain – urinary tract infection. This
was because we did not feel that a urinary tract infection would cause
diffuse abdominal pain for one month. The patient returned two days
later with worsening pain, a surgical abdomen, and was found to have
a segment of necrotic bowel during surgery. We felt extremely bad about
this outcome but this case was a very valuable teaching case about the
limitations of the CT scan, the atypical presentation of elderly patients,
and the fragility of the elderly.

Case 1.6 Already brewing

• The next case is extremely esoteric and given to us by our friend Ted
Koutouzis, MD. A woman came to the emergency department with lower
abdominal pain that had been present for almost three months. She had
a hysterectomy three months ago but had not seen a physician since.
There were no fevers and her abdomen was mildly tender in the lower
quadrants. The emergency physician did not perform a pregnancy test
because of the hysterectomy. An obstructive abdominal X-ray showed a
non-obstructive gas pattern. A complete blood count showed a normal
white count. The patient was sent home with the diagnosis of chronic
abdominal pain secondary to adhesions. The lady returned to the
emergency department two days later and underwent emergency

continued
18 Learning from medical errors: clinical problems

laparotomy for a ruptured ectopic pregnancy. This pregnancy had

apparently been present prior to the hysterectomy. The gynecologist did
not order a pregnancy test prior to performing the procedure. This case
reminds us to consider the possibility of pregnancy in any female patient
with abdominal pain.

Case 1.7 Not fitting the puzzle

• Edwards presents a case, in The M & M Files, where the physician made
a diagnosis of urinary tract infection in a clinical setting where the
symptoms were atypical. 17 A 29-year-old female with a 22-week
pregnancy came to the emergency department for upper abdominal pain,
mid-back pain, and nausea. The patient did not appear in any distress
and had normal vital signs. She had mild tenderness in the upper abdo-
men but had no costovertebral angle tenderness.
• The patient had a slightly elevated white cell count. The rest of her blood
work was unremarkable based on her stage of pregnancy. A urinalysis
was collected by clean catch and showed 15–20 white blood cells, trace
leukesterase, a few red cells, 5–10 squamous epithelial cells, no nitrites,
and 1+ bacteria. She was then discharged by the ED physician with a
‘possible early pyelonephritis’ and given a prescription for the antibiotic
cephalexin.
• Throughout the night, the patient continued to experience intermittent
episodes of burning epigastric pain. This led her to return to the ED the
following morning. A different physician evaluated her and gave her a
dose of Mylanta. This relieved her pain and she was discharged with
‘esophageal reflux of pregnancy.’ No further tests were ordered. She was
instructed to continue the antibiotics for the infection in her urine.
• Nine days later, the patient saw her obstetrician with increasing upper
abdominal pain. She now also had two days of diarrhea and low-grade
fevers. Her obstetrician admitted her to the hospital. An ultrasound
showed gallstones and a stool culture was positive for Clostridium difficile.
She remained in the hospital for one week for the treatment of pseudo-
membranous colitis. The obstetrician discovered that the patient’s urine
culture was negative. He complained to the ED director of the
misdiagnosis and the harmful iatrogenic treatment. The patient went on
to have no further complications during her pregnancy. She had a chole-
cystectomy seven months after her delivery.
• The first ED physician made a few clinical decisions that did not
represent good medical judgment. The patient’s complaint of upper
abdominal pain and the lack of urinary complaints were both inconsist-
ent with an infection of the urinary tract. Upper abdominal pain, especially
in a pregnant female, is more likely to be related to gallbladder disease
or gastroesophageal reflux disease. Both conditions are worsened
during pregnancy. He noted that her abdomen had mild tenderness in

continued
 Abdominal pain 19

the upper abdomen. Again, both of these findings are not consistent with
a urinary tract infection. Our experience has been that a urinary tract
infection very rarely makes an abdomen tender. When tenderness is
present, it is usually located in the suprapubic area or flanks (in the case
of pyelonephritis). Finally, accepting a urine result that is contaminated
and committing a patient to antibiotic treatment for a false positive result
led to an iatrogenic complication. This error could have easily been
avoided by collecting a urine sample by sterile catheterization.
• The second ED physician also made errors in this case. He did not
recognize the increased prevalence of cholelithiasis during pregnancy.
He also did not bother to review that patient’s first evaluation for
appropriateness before telling the patient to continue the antibiotics. By
accepting the first physician’s diagnosis, he also persisted down the wrong
diagnostic path.

Case 1.8 Check the groin

• Edwards presents a case, in The M & M Files, where the failure to perform
a genitourinary examination in a patient with abdominal pain resulted
in an adverse outcome.18 A 22-year-old man came to the ED complaining
of sudden, severe pain in the right suprapubic area and radiated to his
groin. He appeared colicky and was tachycardic on examination.
• The physician performed a quick abdominal examination and ordered
an evaluation for renal calculi along with intravenous hydration and
pain medications. There were 1–2 red blood cells in the patient’s urine
and the pain improved greatly with the pain medications. An
intravenous pyelogram (IVP) was performed at this point and read by
the radiologist as negative for stones. There were also no obvious signs
of obstruction or hydroureter. The patient was told that he had passed
a kidney stone and discharged with pain medication and instructed to
follow up with his physician in one week.
• The patient continued to have pain and followed up with his primary
physician one week later. The groin pain was now more localized to the
right testicle and there was accompanying swelling and discoloration of
the testicle. An ultrasound was ordered and confirmed the suspicion of
testicular torsion. The right testicle could not be saved and the left
testicle’s viability was uncertain. A lawsuit was settled for a significant
sum because of the patient’s loss of ability to conceive.
• This case serves as a reminder that patients with abdominal pain may
have extra-abdominal disease processes and require more than just an
abdominal exam. Although it is not unusual for the pain of a kidney
stone to radiate to the scrotum, any patient that complains of groin or
scrotal pain should have a genitourinary examination. The ED physician
in this case quickly concluded that the patient’s groin pain was due to
a kidney stone and did not perform a genitourinary examination. Edwards

continued
20 Learning from medical errors: clinical problems

states that ‘the emergency physician actually had a second chance to

make the right diagnosis. Ordinarily, even after a stone has passed, for
some hours there is some degree of residual hydroureter and
hydronephrosis.’19 Therefore, the physician should have embarked on
a search for another etiology besides ureterolithiasis at this point.

Case 1.9 Think of the obvious first

• This case by Edwards, in The M & M Files, shows how abdominal disease
in the elderly can lead to a complicated course if conservative measures
are not taken.20 A 75-year-old man with type 2 diabetes presented to the
ED with epigastric pain that radiated to his back. He had one bout of
nausea and vomiting (vomitus was dark) and failed to have a bowel
movement for a few days despite the use of laxatives. There was a past
history of hypertension and peptic ulcer disease and his current
medications were glipizide, hydrochlorothiazide, enteric-coated aspirin,
and occasional ibuprofen for arthritis pain.
• The patient appeared uncomfortable with a blood pressure of 175/95
mm Hg and a pulse of 110 beats/min. The ED physician wrote that the
patient had ‘nonspecific’ epigastric and upper quadrant tenderness and
his rectal exam was hemoccult negative. A nasogastric tube was also
placed and showed no evidence of gastrointestinal hemorrhage. The
complete blood cell count was normal along with the serum electrolytes
with the exception of an elevated glucose at 190 mg/dl. The electro-
cardiogram did not show any acute change from a previous one. The
abdominal X-ray did not show any acute disease but did show a lot of
stool. The patient was given intravenous ranitidine and a bottle of
magnesium citrate and sent home with the diagnosis, ‘constipation versus
gastroenteritis.’
• The patient’s symptoms did not improve and he returned to see a
different ED physician two days later. Although his bowel movements
were now fine, he had increasing pain after he ate and the vomiting had
become more frequent. He also had some new respiratory complaints
with a cough and was slightly short of breath. His vital signs were
normal and his abdomen, once again, had nonspecific upper abdominal
tenderness without guarding or rebound. The patient was given similar
medications and underwent the same tests as the prior visit. His white
blood cell count had increased from 8700/microliter to 11 700/microliter
with a slight left shift on the differential. The rest of the labs were
essentially unchanged from the previous visit. The second ED physician
did, however, order a chest X-ray because of the respiratory complaints.
He read it as a small infiltrate at the right lung base. Because the patient
was feeling better and was tolerating oral intake, the physician discharged
the patient with the diagnosis of ‘pneumonia’ and believed that it was

continued
 Abdominal pain 21

the cause of his abdominal pain. She gave him a dose of azithromycin
and also wrote him a prescription for it.
• The patient’s symptoms progressively worsened when he returned the
following day. He not only had more vomiting and pain, but also now
had pronounced chills. He did not look well on presentation and his
blood pressure was 110/54 mm Hg and his pulse was 115 beats/min. His
temperature remained normal but the third ED physician thought that
the patient felt warm and requested a rectal temperature. This was
recorded as 102.3ºF.
• On review of the prior charts, it was discovered that the radiologist’s
interpretation of the chest X-ray was normal. The patient’s abdomen now
seemed to be most tender in the right upper quadrant. Abdominal
guarding or rebound was still absent. An attempt to elicit a Murphy’s
sign was negative. An abdominal ultrasound confirmed the suspicion of
cholelithiasis and acute cholecystitis. The patient was admitted and
started on intravenous antibiotics immediately. During surgery, it was
discovered that his gallbladder was full of pus and contained necrotic
material. His post-operative course was complicated by acute respiratory
distress syndrome. The patient’s family filed a complaint for the two
prior visits and misdiagnoses.
• Let us now review the first two visits and find where the clues to the
correct diagnosis were missed. Although we are being critical with the
benefit of hindsight, it is nevertheless a good way to learn for future
practice. The patient presented with upper abdominal pain. The number
one cause of surgical abdominal pain in the elderly patient is acute
cholecystitis.21 Therefore, both ED physicians should have immediately
considered this as the diagnosis of exclusion despite a normal temperature
and white blood cell count. In Emergency Medicine: a comprehensive study
guide, McNamara points out that half of the patients with acute
cholecystitis may present without a fever and 30–40% of them will have
a normal white blood cell count.21
• Both physicians also seemed to ignore the more serious diagnoses that
were possible given the patient’s complaints. Namely, these are abdominal
aortic aneurysm, mesenteric ischemia, and acute cholecystitis. Instead,
they seemed more concerned with focusing on benign diagnoses such
as constipation and gastroenteritis and an obscure diagnosis (pneumonia)
that did not seem to fit with the clinical picture. The second physician
should have been alarmed with the increasing white blood cell count,
particularly the left shift. Furthermore, a repeat visit for the same problem
requires a step up in care, which in this case would have probably
consisted of either radiological imaging or hospital admission.
22 Learning from medical errors: clinical problems

Case 1.10 Did not read the text?

• This is a case about a gentleman with abdominal pain who ‘forgot to read
the medical texts.’ A 54-year-old man with no prior history of
gastrointestinal diseases presented to the ED with diffuse abdominal
pain that had been present for one week. He had no chronic medical
problems and did not have any prior abdominal surgeries. He denied
any associated fevers, nausea, vomiting, diarrhea, dysuria, hematuria,
stool changes, or weight loss. He also claimed that the pain had not gotten
worse but that he just wanted to find out what the cause was. He walked
in slightly bent over and appeared in mild pain. Vital signs were
unremarkable and temperature was normal. Abdominal exam was noted
for mild tenderness along the lower abdomen without guarding or
rebound. His prostate was nontender and his stool was hemoccult
negative. He did not have any inguinal hernias.
• A supine and upright abdominal X-ray was performed and showed a
nonobstructive gas pattern without evidence of free air. The patient’s
EKG, urinalysis, hemoglobin, hematocrit, liver function, lipase, and serum
electrolytes were all unremarkable. His white blood cell count, however,
was elevated at 21,000/microliter. An abdominal CT was then performed
for the evaluation of possible diverticulitis. This diagnosis was confirmed
by the scan along with the additional findings of a perforation of the
diverticulum and free air within the peritoneal cavity. The patient was
immediately taken to the operating room where extensive peritonitis
from leakage of bowel contents from the perforation was found. The
patient did well post-operatively.
• This patient did not present with the classic presentation of bowel
perforation. He did not appear toxic, did not have abnormal vital signs
or show evidence of hemodynamic compromise, and did not have the
board-like rigid abdomen that is expected from someone with extensive
peritonitis. The only giveaway was the elevated white blood cell count.
Consideration, however, must be placed on the most likely diagnosis
and appropriate tests should be initiated to confirm it. In this particular
case, the suspicion was diverticulitis and was confirmed with a CT scan.

Case 1.11 Get the necessary tests

• A 19-year-old woman presented to the ED with intermittent lower
abdominal pain for three days. She had a positive home pregnancy test
and her last menstrual period was six weeks ago. She also complained
of some vaginal spotting but had no current bleeding. This was her first
pregnancy and she had not had any prenatal care due to lack of medical
insurance.
• Initial examination of the patient revealed normal vital signs and no acute
distress. There was no vaginal bleeding or abdominal tenderness. A
complete blood count was unremarkable and the serum human
continued
 Abdominal pain 23

chorionic gonadotropin (HCG) was 7,200 mIU/mL. The ED physician

followed hospital protocol and called the radiologist to obtain approval
for an ultrasound to rule out ectopic pregnancy. The radiologist asked
if the patient was having abdominal pain or vaginal bleeding currently
and received an affirmative ‘no’ to both questions from the ED
physician. He then denied approval for the ultrasound and told the ED
physician that the patient could get one done as an outpatient. The patient
was discharged and told to schedule an ultrasound within a couple of
days.
• The patient returned by ambulance two days later with increasing lower
abdominal pain and hypotension. An ultrasound revealed a ruptured
ectopic pregnancy on the right. The patient was taken to surgery where
the ectopic pregnancy and her right fallopian tube were removed. She
and her husband filed a lawsuit against the ED physician and the
radiologist for the delayed diagnosis. A settlement was made against the
radiologist. The ED physician, however, was acquitted. Subsequent
requests for obtaining pregnancy ultrasounds at this hospital were much
more lenient.

Case 1.12 Not always appendicitis

• Perri Klass, MD, presents a case, in Diversion, of an unusual cause of
right lower quadrant pain.22 Her eight-year-old son woke her up one
night. He was screaming of pain in his abdomen. When she asked him
to localize the pain with one finger, he pointed directly to McBurney’s
point. Concerned that it might be his appendix, she drove him directly
to the emergency department.
• Along the way, he stated that his pain had improved, which made her
wonder whether his appendix had just burst. On arrival, the ED physician
agreed with the mother’s assessment and called the surgeons to see the
patient. They wanted the patient to undergo a CT scan and an air contrast
enema. While being transported for these tests, the patient’s pain started
to increase.
• After completing the tests, the patient was taken back to the emergency
department where a different ED physician was on duty. He informed
the mother that the white count, the urinalysis, and the CT scan were all
normal. Since the pain had once again gotten better, the patient was
discharged with the diagnosis of acute gastroenteritis.
• Late in the afternoon, the pain had recurred and was more severe. The
mother was confused because she had never seen acute gastroenteritis
with a presentation like this in her own pediatric practice. She headed
back to the ED with her son and paged the senior surgeon. After explaining
her dilemma, the surgeon decided to review the patient’s CT scan with
several of the radiologists. They concluded that the patient’s pain was
caused by torsion of the omentum.
• Although this patient’s pain resolved over the next several days without
continued
24 Learning from medical errors: clinical problems

surgical treatment, most patients with torsion of the omentum require

resection of the involved part. Before the arrival of the CT scan, torsion
of the omentum was a diagnosis that was usually made during surgery
for ‘appendicitis’ in which the appendix was normal in appearance. In
fact, Daly, Adams, Fantini, and Fischer in Principles of Surgery write: ‘the
finding of free serosanguineous fluid at the time of laparotomy in the
absence of a pathologic condition in the appendix, gallbladder, or pelvic
organs should suggest the possibility of omental torsion.’23
• There are several clinical pearls to remember about this case. Right lower
quadrant pain does not always represent appendicitis. Appendicitis is
not the only surgical emergency that presents as pain in the right lower
quadrant (e.g. testicular torsion, tubo-ovarian abscess). Surgical
emergencies may present with waxing and waning symptoms. Finally,
the limitations of the CT scan must be realized.

Case 1.13 Remember the risk factors (‘F’s)

• Acute cholecystitis does not always manifest as right upper quadrant
pain. An obese 41-year-old woman presented to our emergency depart-
ment with three days of epigastric pain accompanied by a few bouts of
nausea and vomiting. The pain was worse after she ate and she also had
low-grade fevers. She denied any previous history of similar symptoms.
• The ED physician noted that her abdominal exam had tenderness in the
epigastric area but no guarding or rebound. He ordered liver function
tests, amylase, and lipase and they were all unremarkable. He also
ordered a white blood cell count that was elevated at 12,000 cells/
microliter. She was given a gastrointestinal cocktail, which gave her some
relief. The physician subsequently discharged the patient with the
diagnosis of gastrointestinal esophageal reflux disease.
• The patient returned two days later with worsening of the same
symptoms. Her temperature was now 101ºF. There was exquisite tender-
ness and guarding in her epigastric area. The liver function tests, amylase,
and lipase were normal again. The white blood cell count, however, had
increased to 19,000 cells/microliter. A CT scan of the abdomen showed
acute cholecystitis with evidence to suggest perforation. She was found
to have a necrotic gallbladder during surgery. Postoperative course,
however, was complicated with a longer hospital stay.
• The empiric diagnosis of peptic ulcer disease (PUD) or gastroesophageal
disease (GERD) should be made with caution. This is especially true in
patients with other ‘red flags’ of abdominal pain. Fever, elevated white
blood cell count, significant abdominal tenderness, abnormal liver
function tests, or elevated lipase, all suggest an alternative diagnosis to
PUD or GERD. Therefore, these diagnoses should be made only in the
presence of normal blood tests, normal vital signs, and the absence of
significant findings on the physical exam.
 Abdominal pain 25

Case 1.14 The interaction of organ systems

• As we have discussed, we try to follow protocols as strictly as possible.
In particular, we order EKGs in almost every patient with abdominal
pain who is over 50 years of age or has a cardiac history. It is written in
the literature that inferior myocardial infarctions can produce vague
abdominal pain. We had a case recently that reinforced our use and
reliance on protocols. Fortunately, it is not really an error case but we
wanted to share it.
• A 44-year-old man who had a history of coronary artery disease
presented to the ED with periumbilical pain for 24 hours. He had a myo-
cardial infarction five weeks ago and underwent a two-vessel coronary
artery bypass. There were no associated fevers, nausea, vomiting, or
diarrhea with this pain. The patient’s vital signs were all unremarkable.
His abdominal exam also did not reveal any significant tenderness,
guarding, or rebound.
• We ordered liver function tests, amylase, lipase, an abdominal X-ray, and
a urinalysis. All of these tests were unremarkable. An EKG, however,
showed T inversions in the inferior leads, which were new from his
previous EKG. Cardiac enzymes were added and revealed a troponin I
of 2.7 NG/ml (range 0–1.4 NG/ml) along with a normal CPK-MB. We
consulted the cardiologist because we believed that the patient’s troponin
from his myocardial infarction should have normalized five weeks later.
This suspicion, along with his EKG changes, led the cardiologist to admit
him to the hospital. His troponin and CPK-MB continued to increase
during the next two days and he was found to have another myocardial
infarction.
• Although an EKG is recommended in all patients over 50 years of age
with abdominal pain as discussed in the ‘Diagnostic tests’ section, this
practice is not followed in the majority of patients. We have not seen any
studies to show that this practice is cost-effective. Indeed, we have heard
a number of patients question us why an EKG was performed for their
abdominal pain. However, an EKG is such a quick and inexpensive test
that we find it hard to not follow this recommendation. Furthermore,
EKGs may be helpful in the diagnosis of abdominal pain in the presence
of other findings. Atrial fibrillation may be the cause of mesenteric
ischemia, while heart block may be a sign of digoxin toxicity and its
concomitant abdominal pain.

Case 1.15 The answer is in front of you

• Selbst and Korin, in Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine, tell of a 13-year-old boy who had classic findings of appendi-
citis but was given an alternative diagnosis.24 The boy came to the ED
complaining of lower abdominal pain, mainly on the right side for one
day. The pain was associated with anorexia, nausea, and vomiting. He

continued
26 Learning from medical errors: clinical problems

had not had a bowel movement in 2 days and had no urinary symptoms.
• On physical examination, he had a temperature of 103.9ºF. There was
tenderness in the right and left lower quadrants with moderate
involuntary guarding. His blood work revealed a white blood cell count
of 9,760 cells/microliter with 14% bands on the differential. The urinalysis
was normal. The ED physician and the radiologist read the abdominal
X-ray as a possible appendicolith.
• The patient was given Tylenol, intravenous fluids, and observed in the
ED. He pain had improved and he was able to tolerate oral fluids. Sub-
sequently, he was discharged with a diagnosis of ‘renal colic versus acute
gastroenteritis.’25 Instructions were given for supportive treatment and
to strain all urination for stones. His pain became much worse three days
later and he was brought back to the ED. A pelvic ultrasound confirmed
the suspicion of appendicitis. The family filed a lawsuit against the ED
physician for a delay in diagnosis and, subsequently, received an out of
court settlement.
• The poor outcome to this story could have been avoided if the clinician
had followed his clinical and laboratory findings. There was no history
of diarrhea to warrant a diagnosis of acute gastroenteritis. Furthermore,
there was no blood in the urine to suggest ureterolithiasis. Instead, there
were classic symptoms of right lower quadrant pain, abdominal tender-
ness and guarding, anorexia, nausea, vomiting, left shift on the differential,
fever, and possible appendicolith. Ignoring classic findings of a disease
is a very easy way to make errors and be found liable for them.

Case 1.16 Take ‘guarding’ literally

• As we discussed above, with very few exceptions, the patient with
abdominal pain and abdominal guarding on exam should be examined
meticulously. We believe that the term ‘guarding’ has a dual purpose. It
is used to describe the patient’s response to the physician’s palpation of
the abdomen. It is also used to warn the physician to guard against a
serious etiology. In our experiences, guarding or rebound on an
abdominal exam necessitates a work-up and should not be routinely
regarded as urinary tract infection or gastroenteritis. The following case
illustrates this point.
• Selbst and Korin, in Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine, present a case of an 11-year-old girl who was sent by her
pediatrician to the ED. 26 She had diarrhea, vomiting, and lower
abdominal pain for five days. She had pain on urination, fevers at home,
and vaginal discharge. On physical examination, her temperature was
37.5 C and her abdomen had mild guarding with diffuse tenderness. The
ED physician was extremely thorough with his physical exam, which
included a pelvic, vaginal, and rectal examination. He decided, however,
to not order any laboratory tests and discharged the patient with
‘gastroenteritis.’
continued
 Abdominal pain 27

• She felt better for a few days, but then the pain returned and actually
worsened. She was brought back to the ED five days after the first visit
with bilious vomiting. Her abdomen was distended with diffuse
voluntary guarding and questionable rigidity. Laboratory tests revealed
a white blood cell count of 20,000 cells/microliter with a left shift on the
differential. An X-ray showed a small bowel obstruction. Eight hours
after her arrival, she was found to have a perforated appendix and
contaminated peritoneal fluid. Her postoperative course was complicated
by pleural effusion, abdominal wall infection, wound dehiscence, and a
three-week hospital stay. A lawsuit was filed based on delayed
diagnosis, excessive weight gain as a result of her surgical complications,
and questionable future fertility. An expert witness faulted the physician
for not ordering an X-ray, a complete blood count, and surgical
consultation. The case was eventually dropped because of the difficulty
in proving that appendicitis was present during the initial visit. If the
patient had returned within 24–48 hours of the first visit, however, the
outcome of the suit would probably have been different.

Case 1.17 Getting to the bottom of things

• Another case from Selbst and Korin from Preventing Malpractice Lawsuits
in Pediatric Emergency Medicine highlights the necessity of a genitourinary
exam for abdominal pain in the pediatric patient.27 After getting hit by
a baseball in the upper leg, a 15-year-old boy presented to the ED with
upper abdominal pain. Although the physician indicated that there were
no hernias on physical examination, the patient and his mother contend
that the patient was never asked to remove his underwear. After the
exam, the patient was discharged and told to return if there was no
improvement.
• The mother called the following day to say that her son was worse. This
was contradictory to what the records indicated. Four days later, he was
taken to see his family physician for worsening pain and was diagnosed
with testicular torsion. He received immediate surgery but the testicle
could not be saved. Although the family filed a lawsuit against the ED
physician for misdiagnosis and failure to perform a genital exam, the
defense claimed that the testicular torsion occurred after the visit to the
ED. Support for the defense’s argument came when it was discovered
that the patient had told his own physician that his testicular pain had
not begun until 3 to 4 days after being in the ED. Subsequently, the verdict
was in favor of the defense.
• The ED physician in this case was saved by the patient’s admission as
to the onset of testicular pain. Would the physician have lost the case if
this admission had not been given? It is difficult to answer this question;
however, the defense would have the difficult burden of proving that
testicular symptoms were not present on the first visit because an exam-
ination was not performed. Hence, a quick examination is a brief amount
of time well spent and a huge amount of time saved in legal proceedings.
28 Learning from medical errors: clinical problems

Case 1.18 It is not always gastrointestinal

• A 40-year-old female came to our ED with severe abdominal pain. The
pain was on the right side and started approximately eight hours ago.
It was accompanied by fevers and nausea. Although the pain started in
the right lower quadrant it was now worse in the right upper quadrant
and also radiated to the shoulder. She denied any genitourinary
symptoms. During the past three days, she had been celebrating her 40th
birthday and partying with her friends.
• Her past medical, surgical, and gynecologic histories were unremark-
able. On physical examinations, she was in moderate discomfort and
had a temperature of 102ºF. Her abdomen was extremely tender with
involuntary guarding on the entire right side. The right upper quadrant
was the area of the most tenderness.
• A complete blood count and blood cultures were immediately drawn.
Intravenous antibiotics were then given. Afterwards, the patient was taken
for an ultrasound of the gallbladder. This showed no evidence of gall-
bladder disease. A pelvic examination was then performed which showed
tenderness in the right adnexa. It was difficult to appreciate a mass because
the patient could not tolerate a full exam. Cervical motion tenderness,
however, was not present (the patient was medicated with narcotics by
this time). There was no bleeding or tenderness in her rectum.
• The on-call surgeon was called to evaluate the patient. Her white blood
cell count was 20,000 cells/microliter. The surgeon met the patient in
radiology while she was getting a CT scan of her appendix. The
appendix was not visualized but there was no indication of
inflammatory changes in its expected area. Edema was seen surround-
ing the liver and moderate free fluid was seen in the pelvis. These findings
suggested pelvic inflammatory disease and the Fitz-Hugh-Curtis
Syndrome. She was taken to the operating room by the surgeon and the
gynecologist and found to have a tubo-ovarian abscess. This case was an
excellent example of how a pelvic disease process can present as
symptoms of an upper abdominal organ disease process (e.g.
cholecystitis). It also shows that a step-wise approach to evaluate for
multiple possible surgical emergencies may be required in some patients.
Her urine pregnancy test was negative while urine drug screen was
positive for cocaine.

Case 1.19 Some information is less readily volunteered

• A colleague shared a case that is similar to Case 1.17 with a different
outcome. A 16-year-old boy was taken to the ED with four hours of right
lower quadrant pain. The boy had told his nurse that he felt tightness in
his scrotum but this conversation was not relayed to the ED physician.
The physician noted that the pain had been present for four hours and

continued
 Abdominal pain 29

was accompanied by nausea. There were no fevers, vomiting, or diarrhea.

He also wrote that there was no hematuria or dysuria.
• Vital signs were all within normal limits. The boy’s abdomen was tender
in the right lower quadrant with mild guarding and no rebound. Genital
and rectal exams were not performed. A white blood cell count and
urinalysis were within the normal ranges. A radiologist read a CT of the
patient’s appendix as normal. The boy was discharged with the
diagnosis of ‘acute gastroenteritis’ and given symptomatic treatment.
• He returned to the ED four hours later complaining of severe pain and
swelling in his right scrotum. The same physician was on duty and
ordered a color flow ultrasound of the testicles. This showed no blood
flow to the right testicle and was consistent with testicular torsion. The
patient was taken to surgery but his testicle was unsalvageable. A law-
suit for delay in diagnosis was filed against the ED physician. We are not
aware of the outcome.
• Children and adolescents frequently give different histories to different
healthcare providers. They may not inform the physician of pain in their
‘private parts’ in fear of an examination. In contrast, they may be more
willing to share this information with a nurse or medical technician. Some
physicians may also feel uncomfortable at performing genitourinary and
rectal exams in children. The physician, however, must be as thorough
with his history and physical exam as needed in order to make a correct
diagnosis.

Case 1.20 Abdominal pain imitator

• An 86-year-old man presented to the ED complaining of epigastric pain
for three hours. He had a history of coronary artery disease in the past
but stated that this did not feel like his prior heart problems. Instead, he
said it felt like indigestion and he had a history of GERD. He had nausea
but denied any chest pain, sweating, or vomiting.
• His vital signs and physical exam were unremarkable. In particular, his
abdomen was soft and nontender. Electrocardiogram showed sinus
rhythm with occasional premature ventricular contraction and no acute
ST segment changes. His cardiac enzymes, liver function tests, amylase,
and lipase were all within normal limits. He was given a gastrointestinal
cocktail with some relief. The physician wrote him a prescription for
lansoprazole and discharged him with the diagnosis of GERD.
• He returned four hours later with worsening epigastric pain. It was now
accompanied by chest pain. A second EKG showed 2 mm of ST elevation
in leads V2 to V4 with T inversions in the inferior leads. He was taken
to the cardiac laboratory and had a cardiac stent placed. Acute
myocardial infarction was confirmed with a troponin level that was three
times the upper limit of normal.
• The first ED physician made some preventable mistakes in this case. He

continued
30 Learning from medical errors: clinical problems

correctly acknowledged the patient’s history of coronary artery disease

and ordered a cardiac work-up. However, he should have been more
suspicious for a cardiac etiology when the abdominal exam was
unimpressive. He also failed to notice the subtle changes on the patient’s
EKG (no PVCs were present on the patient’s prior EKG). Finally, he did
not consider the limitations of one set of EKG and cardiac enzymes in
ruling out myocardial infarction.

References
1 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 497.
2 (2003) ACEP, Reference+Resource Guide. American College of Emergency Physicians,
Dallas, Texas, 19.
3 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St. Louis, Missouri, 1888.
4 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St. Louis, Missouri, 1890–91.
5 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St. Louis, Missouri, 1890.
6 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 343.
7 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St. Louis, Missouri, 1892.
8 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St. Louis, Missouri, 2006.
9 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 67.
10 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem solving (2e). Saunders, Philadelphia, Pennsylvania, 92.
11 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 508.
12 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 571.
13 Pearlman MD and Tintinalli JE (1998) Emergency Care of the Woman. McGraw-Hill,
New York, New York, 22.
14 Pearlman MD and Tintinalli JE (1998) Emergency Care of the Woman. McGraw-Hill,
New York, New York, 23.
15 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 83.
16 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 578.
17 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 102–4.
18 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 111–12.
19 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 112.
20 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 124–7.
 Abdominal pain 31

21 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive

study guide (5e). McGraw-Hill, New York, New York, 517.
22 Klass P (2004) A Doctor ’s Life, Physician as Parent, When your own child is
mysteriously ill. Diversion. 32(4): 31–3.
23 Schwartz SI, Shires GT, Spencer FC et al. (1999) Principles of Surgery (7e). McGraw-Hill,
New York, New York.
24 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 64–5.
25 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 65.
26 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 65–6.
27 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 81–2.
This page intentionally left blank
Chapter 2

Chest pain

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Chest pain is the fifth most common diagnosis code for patients seen in
emergency departments throughout the United States in 2001.1 It represents
approximately 5% of all emergency department visits.2 The etiologies of chest
pain can be of various organ systems including cardiac (acute coronary
syndrome, pericarditis), pulmonary (pneumonia, pneumothorax, pleurisy),
gastrointestinal (cholelithiasis, esophagitis), musculoskeletal (costochondritis,
rib fracture) and dermatologic (herpes zoster). They may also range from an
acute life-threatening process such as myocardial infarction to a relatively
benign process such as costochondritis.
Chest pain patient encounters also represent a significant percentage of
malpractice cases. Specifically, patients with acute myocardial infarctions (AMI)
represent the greatest legal risk. This is due to the combination of the high
immediate mortality and the increase subsequent morbidity of AMI. In fact,
‘AMI was the most frequent and expensive medical condition seen in claims
against emergency room physicians.’3 Furthermore, AMI is also ‘the most
common condition associated with malpractice claims against family
physicians and internists.’3 Hence, a physician can conclude that if he sees
patients with chest pain, these patients are the ones likely to represent his
greatest liability.

Triage
The general rule of thumb is that all patients complaining of chest pain should
be seen by the physician as soon as possible. This is because of the potential of
a life-threatening condition that may warrant immediate treatment. We have a
standing protocol in our emergency department where all adult chest pain
patients are brought back immediately and an electrocardiogram (EKG) is
performed as the physician is summoned. It is recommended in Emergency
Medicine: a comprehensive study guide that a 12-lead EKG be performed within 10
minutes of arrival in the emergency department for all adult patients.4 Most
34 Learning from medical errors: clinical problems

emergency departments have similar protocols. Furthermore, make sure that

the triage nurse follows the protocol adamantly because leniency may lead to
disastrous results (please see Case 2.15).
Having stated our ED protocol for chest pain, it is worth mentioning that it
was written under ideal conditions and not under practical conditions. In a busy
ED, there will frequently be more than one patient with an emergency at one
time and with limited healthcare provider resources (i.e. physicians, nurses)
some patients with true emergencies may still have to wait to get seen. There is
no easy solution here except to see the most critical first and move on to the next
as soon as the first is stabilized. Although we strive to see all chest pain patients
as soon as they come, this goal is simply not feasible and the triage nurse often
has to use his or her discretion on who sees the physician first. All patients with
cardiac histories, risk factors for coronary artery disease, or a history of prior
serious etiologies of chest pain (e.g. pneumothorax, aortic dissection, etc.) should
be seen immediately. In addition, all patients who are in distress or have
abnormal vital signs should be seen immediately. In contrast, the younger
patient (less than 30 years of age), who has no cardiac history with normal vital
signs and no obvious distress, can probably be seen after the patient with
refractory vomiting.
Of course, you will have the occasional patient who will sit in your waiting
room with an acute myocardial infarction or a pneumothorax and appears in no
acute distress and is not seen immediately. These patients do not necessarily
represent a failure of the triage system as long as emergent patients who need
to be treated are being treated. They do, however, represent an increased liability
to the physician due to the delayed treatment. We cannot offer any solution to
this inevitable problem except to reiterate that you should try to see patients
who complain of chest pain as soon as possible.
The same triage concepts apply to the chest pain patient presenting to the
office. However, the office staff may not be accustomed to performing a bedside
triage and EKG within 10 minutes. We have seen some offices advise every chest
pain patient to go to the emergency room and not offer a physician (or nurse)
evaluation. We feel that this practice is bad medicine and represents high legal
liability. A patient with a possible acute coronary syndrome should not be
instructed to drive himself anywhere. Furthermore, acute coronary syndrome
represents a medical emergency that should be stabilized by the best
personnel available. In this scenario, this would mean the office physician and
staff and, subsequently, the paramedics who will transport the patient to
the hospital in a monitored setting. Finally, obtaining a diagnostic EKG and
relaying this information to the cardiologist or emergency physician may
expedite the time that it takes for the patient to get to the heart catheterization
laboratory.

History
The history obtained from a patient with chest pain should elicit the same
characteristics as that of other pain complaints (i.e. quality, severity, onset,
modifying factors, precipitating events, etc.). In addition, classic symptoms of
the suspected causes should be ascertained. For example, patients with
suspected angina should be asked about pain that radiates to the jaws or arms
 Chest pain 35

or is associated with diaphoresis, nausea, and vomiting. Patients with suspected

pericarditis should be asked about their pain in relation to body positions.
Finally, patients with suspected aortic dissection should be asked if the pain
goes to their back.
In the adult patient, the most important assessment is whether there are risk
factors for coronary artery disease. These include male age over 40, post-
menopausal female, hypertension, cigarette smoking, hypercholesterolemia,
diabetes, truncal obesity, family history, and sedentary lifestyle.5 In younger
patients, cocaine use should be ascertained because of its association with
myocardial infarction.
Another important assessment for patients with suspected acute coronary
syndrome is to ask if they have ever had a cardiac evaluation (i.e. heart catheteri-
zation, stress test). If the answer is yes, then inquire about the results and try to
obtain the reports. We have discussed previously how helpful previous reports
are to the emergency or primary care physician. They are also helpful for the
cardiologist. In addition, they may save valuable time and money spent on tests
that have recently been performed. We see colleagues treat chest pain without
inquiring about previous work-ups and realize that sometimes their decision
making would have been altered had the information been known. Also,
requesting previous medical records, even though you may often not be able to
obtain them, will emphasize that you are a meticulous and thorough physician
(characteristics that will help you in the courtroom).

Physical exam
Complete vital signs are extremely important in the patient with chest pain.
Blood pressure has obvious importance in cardiac causes of chest pain. In the
particular case of a suspected aortic dissection, the physician should request that
blood pressure be performed on both arms. Although not all dissections are
going to give you a significant discrepancy, you may be able to pick up a few.
In the unfortunate event that you miss one, at least you can defend yourself in
court by stating that you considered it and felt it less likely because there was
no discrepancy.
As mentioned in the beginning of this chapter, chest pain can originate from
many organ systems. Therefore, make sure that you perform a thorough exam
of all of these organ systems. Have the patient fully undressed and use repeated
exams to assess for response to medications and for clinical deterioration. In
particular, remember that patients with potential acute coronary syndromes
should have repeat EKGs for increases in pain, changes in vital signs, or after
acute intervention (e.g. thrombolytics).
Patients with acute coronary syndrome need rapid therapeutic interventions.
Therefore, the exam must be focused and expedient. Since treatment frequently
includes anti-thrombin and anti-platelet therapy, an assessment for possible
bleeding complications must be included in the rapid initial assessment (i.e.
uncontrolled blood pressure, recent surgeries, blood in stool). Remember that
if pericardial effusion is suspected or the patient had recent chest trauma, the
use of heparin may be contraindicated.
36 Learning from medical errors: clinical problems

Differential diagnosis
As we mentioned earlier, there are myriad causes of chest pain. For the adult
patient, the most important is acute coronary syndrome (myocardial infarction,
unstable angina). For the pediatric patient, respiratory emergencies such as pneu-
mothorax and asthma exacerbation are more common. However,
physicians should always be alert for other emergencies that may pose as chest
pain and require immediate treatment. These include pulmonary embolus,
pericarditis, pneumonia, aortic dissection, perforated peptic ulcer, and esophageal
rupture.6
After the serious causes mentioned above and coronary artery disease is ruled
out, the physician can then entertain less serious etiologies. These include, but
are not limited to, cholelithiasis, gastroesophageal reflux disease, costochondritis,
pleurisy, anxiety, and mitral valve prolapse. As we stress over and over in this
book, it is a very low legal risk to misdiagnose any of these relatively benign
diagnoses. However, it is a very high legal risk to misdiagnose any of the
conditions in the above paragraph. Furthermore, it is probably even worse to
misdiagnose a serious condition with a less serious diagnosis. These latter cases
are better labeled as ‘chest pain, uncertain etiology.’

Diagnostic tests
We strongly believe that every patient who complains of chest pain should have
an electrocardiogram and a chest X-ray ordered. Please see Case 2.1 on why we
feel this way. These two tests are the most sensitive tests for determining if
immediate therapy is needed (e.g. fibrinolytics for acute myocardial infarction,
chest tube thoracostomy for pneumothorax). In fact, although we do not like to
order tests without seeing the patient first, we sometimes order these two in
patients with chest pain even if we cannot see the patient immediately. These
two simple and quick tests may give us clues to all of the emergent diagnoses
listed in the previous section. When patients are seen on return visits with chest
pain, many physicians will defer a repeat chest X-ray. This is not always prudent
as shown in Case 2.20.
Further testing is dictated by the suspected condition and we refer the reader
to other medical texts of reference. However, we will comment further on the
patient with suspected acute coronary syndrome. Any adult patient with as least
one of the risk factors for coronary artery disease listed above should, for the
most part, have blood work drawn for cardiac enzymes. The reason for this is
that only about half of patients with an acute myocardial infarction presenting
to the emergency department will have diagnostic changes on the initial EKG.4
Likewise, patients with unstable angina may also have normal initial EKGs.
The few exceptions may be when an obvious other cause for the chest pain
is present such as pneumothorax, varicella zoster, or rib fracture. Also, in
the office setting where blood tests for cardiac enzymes are not possible,
patients presenting with chest pain and cardiac risk factors may be evaluated
without blood work if they have normal or unchanged EKGs and a recent
unremarkable cardiac evaluation (i.e. heart catheterization, stress test). A heart
catheterization is preferable because it is the most definitive evaluation
for coronary artery disease and a normal catheterization report from within
 Chest pain 37

the previous two years would be fairly accurate for the current state of the
coronaries.5
Patients with consideration for pulmonary embolism as the cause of their
chest pain usually receive two to three of the following tests. Due to improved
technology and the development of the spiral CT scanner, CT pulmonary scans
of the chest are increasingly used in the diagnosis of pulmonary embolism.
Although these scans have the potential to miss small, peripheral embolisms, its
sensitivity for detecting significant emboli is extremely high. Consequently, the
traditional gold standard of ‘pulmonary angiography’ is rarely used in clinical
practice today due to its invasive nature and its complications. The other imaging
modality that is also sensitive and noninvasive for pulmonary embolism is the
lung ventilation/perfusion scan (V/Q). This test will usually be helpful when
the chest X-ray is fairly normal. Otherwise, there will be a high false positive rate
and the CT scan would be the better choice.
For patients with suspected pulmonary embolism (PE), we usually start the
evaluation with an arterial blood gas (ABG) and a d-dimer. We consider these
as quick and relatively inexpensive screening tests for PE. It is not practical for
every patient with shortness of breath or pleuritic chest pain to receive an imaging
modality for pulmonary embolism based on the costs, radiation, and time of the
procedures (not to mention the consequences of having to treat the false posi-
tives). If the patient is deemed at low risk for PE and the arterial-alveolar gradient
on the ABG and the d-dimer are normal while the pO2 is greater than 80 mm Hg
on the ABG, this is adequate assurance for us to not pursue the diagnosis of PE
any further. The caveats to this rule are that the patient’s respiratory rate must
be less than 20 breaths per minute (the pO 2 could be elevated with
compensatory hyperventilation) and that the normal range of the alveolar-
arterial gradient in ‘people without lung disease is 5 to 20 mm Hg.’7 Hamilton,
in Emergency Medicine: an approach to clinical problem-solving, writes: ‘the presence
of a negative d-dimer test, with a pO2 greater than 80 mm Hg and a respiratory
rate less than 20 breaths per minute helps rule out pulmonary embolism.’7
However, if one of the above tests is elevated, the patient’s respiratory rate is
above 20 breaths per minute or the pO2 is less than 80 mm Hg on the ABG, or
if the patient is at high risk for PE, then an imaging modality should be ordered.
In our practice, when the CT scan or the V/Q scan is normal and the suspicion
is still high, we will sometimes order the other test. In the event that the other
test is also normal and a high suspicion persists, then pulmonary angiography
should be considered.

Hospital/office course
This section is extremely important to document and document carefully in the
patient with chest pain – in particular, acute coronary syndrome. Heart muscle
could be lost or saved with each minute depending on the physician’s actions.
Therefore, every medication given must be written on the chart along with the
time that it was given. The documentation of the patient’s response to the
medicines (e.g. ‘pain improved with nitroglycerin’) will have huge legal
implications. When the patient’s condition deteriorates, plaintiff attorneys care-
fully review the immediate response of the physician. Hence, every change in
the condition of the patient should be recorded on the chart along with the
38 Learning from medical errors: clinical problems

physician’s response (e.g. ‘the patient’s pain improved with nitroglycerin but
then recurred – repeat EKG ordered, cardiologist notified’). Remember acute
coronary syndrome is a rapidly dynamic process and that these patients need
frequent and repetitive evaluations. Consultants should be contacted as soon as
an emergent condition is diagnosed (or suspected), especially in acute coronary
syndrome. Finally, we feel that it is always valuable to remember a line written
by Diercks in Cardiovascular and Neurovascular Emergencies: implications for
clinical practice, ‘a recent multicenter study demonstrated that approximately
2% of patients with myocardial infarction and 2% of patients with unstable angina
are inadvertently discharged from the ED.’8
In the outpatient setting, it is important to supervise and instruct nurses and
medical assistants to ensure that treatment is done in a prompt fashion. They
may not be accustomed to bringing a patient immediately back, obtaining a stat
electrocardiogram (EKG), and calling the paramedics. Your treatment options
are extremely limited in the office but at the very minimum, you should give any
patient with an acute myocardial infarction an aspirin (if there is no contrain-
dication) and have the patient transported by ambulance. There is much danger
in not transporting the patient by emergency medical services. Please see Case
2.2. Finally, it is helpful to send a copy of the EKG with the patient.

Pitfalls
For acute coronary syndrome, we have already discussed the limitations of the
initial EKG. Heavy reassurance should not be placed on a normal initial EKG
in a patient who has a high probability of ischemia.9 Please see Case 2.9 below.
In addition, atypical presentations of chest pain are common. Tintinalli, in
Emergency Medicine: a comprehensive study guide, writes that ‘up to 22 percent of
patients with acute myocardial infarction describe their symptoms as being sharp
or stabbing in character and up to 6 percent describe a pleuritic component of
their pain.’5 Up to 15% of patients with myocardial infarction have tenderness
to palpation of the chest wall.4 Also, you should remember that up to 33% of
myocardial infarctions may be silent.5 This is more common in diabetics and the
elderly. Another situation is the cardiac transplant patient. These patients have
denervated hearts where cardiac ischemia is more likely to present as heart failure
or sudden death rather than as angina.10 Atypical chest pain commonly occurs
in perimenopausal women, diabetics, and the elderly.
Similarly, there are also limitations of the cardiac enzymes, creatine kinase
(CK) and troponin I, in detecting early myocardial infarction. Since CK values
typically do not rise until four to eight hours after coronary occlusion and troponin
I values six hours after occlusion, the patient presenting earlier than these times
may have normal values. Furthermore, some hospitals only compute a CK value
and do not fractionate this value to a CK-MB value. CK-MB is the isoenzyme
predominantly found in cardiac muscle and is much more specific of cardiac
damage than CK which is found in brain, skeletal muscle, and cardiac muscle.
Therefore, elevated CK values could be found in other conditions such as trauma
and myositis. At these hospitals, how does the clinician interpret an elevated CK
(without a CK-MB) fraction and a normal troponin? Please see the case of the
71-year-old diabetic (Case 2.9) in the following section.
In some cases, there may be multiple causes of chest pain present
 Chest pain 39

simultaneously. We had a patient who complained of right-sided chest pain after

falling off a motorcycle on his right side. He was found to have rib fractures on
that side but started vomiting after returning from X-rays. Electrocardiogram
showed an acute anterior infarction and subsequent catheterization showed
complete occlusion of his left anterior descending artery.
We alert clinicians to symptoms occurring in patients with risk factors for
coronary artery disease that we consider as ‘chest pain equivalent.’ These symp-
toms include shortness of breath, nonexertional diaphoresis, syncope, and
weakness. Our diagnostic approach to patients at high cardiac risk with these
symptoms is very similar to high cardiac risk patients with chest pain. It is the
rare exception, that we would not consider a cardiac work-up for a patient with
chest complaints and multiple risk factors for coronary artery disease. Please see
the case of the 49-year-old woman with atypical chest pain below, Case 2.6.
A ‘chest pain equivalent’ symptom that we see occasionally is left shoulder or
arm pain. For patients with risk factors for coronary artery disease, look for the
presence of trauma or reproducible pain on physical examination. In the absence
of such, evaluation for acute coronary syndrome is mandatory. Insist on the
evaluation even though the patient (or the nurse) may question the necessity of
it. Please see Case 2.21 below.
Patients with myocarditis can present as a diagnostic nightmare. Their
presentations can be extremely varied from simulating musculoskeletal
symptoms to imitating symptoms of myocardial infarction or pleurisy. Instigat-
ing factors are likewise extremely varied. The cause may be of viral, traumatic,
drug-induced, autoimmune, or idiopathic etiology. It may affect patients of any
age and is not necessarily related to coronary artery disease risk factors. Finally,
there are no specific diagnostic tests to detect the condition and there is no
consensus on the disposition for patients with myocarditis. However, Tintinalli,
in Emergency Medicine: a comprehensive study guide, does give a recommendation,
‘Because most patients present with rapidly progressive CHF (congestive heart
failure), admission is usually indicated.’11 We present a case of myocarditis below,
Case 2.3.
Similarly, much of what was said about myocarditis also pertains to
pericarditis. The only exception may be that EKG changes and an elevated
erythrocyte sedimentation rate are usually diagnostic of pericarditis. However,
the decisions to hospitalize and obtain an emergent echocardiogram to evaluate
for pericardial effusion are variable among cardiologists and emergency
physicians. We present a case of pericarditis below, Case 2.4.
Young patients with chest pain rarely have coronary artery disease. However,
be alert to some predisposing conditions that may present as serious cardiac
diseases. These include: history of Kawasaki disease (small risk of myocardial
infarction), history of Marfan’s syndrome (increased risk for aortic dissection
and pneumothorax), and history of cocaine use (risk of ischemic cardiac injury).
Another diagnosis to keep in mind for the young patient with chest pain is
rhabdomyolysis. This can occur when patients perform strenuous activities, use
certain medications, or abuse cocaine. Be suspicious when the patient complains
of pain in other muscles along with the chest pain. Please see Case 2.26 below.
Pleuritic chest pain in a person with any risk factor for deep venous
thromboembolism (e.g. birth control pills, hypercoagulable state, recent prolong
immobilization, malignancy, etc.) should be regarded as a pulmonary embolus
40 Learning from medical errors: clinical problems

until proven otherwise. When we state ‘proven’ in the previous sentence, we

usually grade the likelihood of a pulmonary embolus based on the results of
three tests in our clinical practice (see ‘Diagnostic tests’ section above). Please see
Case 2.5 below.
Gastrointestinal (GI) cocktails are great therapeutic modalities in the
emergency department. However, they can be legally dangerous to use in a
diagnostic fashion. Emergency physicians have sometimes discharged patients
with chest pain after noting improvement upon the administration of a GI cock-
tail. We feel that there is only one situation where this may be appropriate. This
is the relatively young patient (less than 35 years of age) with no cardiac risk
factors. It has been written in many journals and texts that a significant percent-
age of patients with ischemic cardiac pain will receive pain improvement with
a GI cocktail. Furthermore, there are patients who may have concurrent
coronary artery disease and gastrointestinal esophageal reflux disease. Marx
writes in the latest edition of Rosen’s Emergency Medicine: concepts and clinical
practice, ‘caution is also advised in the patient with chest pain who appears to
respond to antacid administration; over-reliance on this response as a major
reason in “ruling out” acute ischemic coronary syndrome (AICS) is not
encouraged.’12 Please see the case of the 42-year-old male with chest pain below,
Case 2.7.
Keep in your mind that ‘zebras’ in medicine do occur. We refer to clinical
presentations as zebras when they are so atypical that most, if not all, physicians
would not consider the underlying etiology. There is really very little that a
physician can do to defend herself against these rare cases except read about
them and know that that exist. Please see the strange case of a 72-year-old woman
with tooth pain below, Case 2.8.
In patients with possible acute coronary syndromes, maintain a low threshold
for ordering multiple EKGs. Remember that an EKG is a static recording of the
patient’s electrical activity for a specific moment in time. Acute coronary
syndrome, however, is a dynamic process. It is also a process where interven-
tions may change the EKG findings. Please see Case 2.10 below. Therefore, reliance
on a single tracing in a patient with possible acute coronary syndrome may
sometimes lead to inappropriate therapy. Ordering serial EKGs also minimizes
potential delays in treatment. Case 2.22 below shows how helpful this could be
in defending claims.
Be careful about patients who may be stoic on the degree of their chest pain.
For some uncertain reason, we have experienced that it is uncommon for
patients to dramatize their chest pain and not uncommon for them to
underestimate their chest pain. This observation is in contrast to that of
abdominal pain. We have seen many patients act out a benign cause of
abdominal pain. Very few patients are stoic about the degree of their abdominal
pain. Perhaps some patients have a greater fear of a heart attack than of
abdominal disease processes; hence, this leads to a defensive denial of severe
pain in the chest. Most ED physicians can probably admit that they have seen
patients with an acute myocardial infarction who did not appear in great
distress. On the contrary, we have never seen a patient with acute appendicitis
who appeared comfortable and ‘pain-free.’ Therefore, develop a sense of whether
your patient is the stoic type. Keep asking the patient and ask in different ways
whether the patient is having pain. It also helps to have multiple people such
 Chest pain 41

as nurses and medical assistants assess the patient’s pain. Please see Case 2.10
below.
Chest pain and neurologic symptoms is an aortic dissection until proven
otherwise. The tear and subsequent disturbance in blood flow causes vascular
insufficiency to certain areas. This can produce symptoms similar to those of a
cerebrovascular accident. Patients should be evaluated for acute coronary
syndrome and cerebrovascular event in the normal fashion because they
certainly could occur concurrently. In addition, remember to consider an
evaluation for aortic dissection. Please see Case 2.11 below. In addition, as shown
in Case 2.19 below, failure to recognize an aortic dissection may lead to therapy
that is contraindicated (i.e. anticoagulation for cerebrovascular accidents and
unstable angina).
Physicians should also remember that there are extrathoracic causes of chest
pain. These are less common, but two common examples would be biliary
disease and gastroesophageal reflux disease. Be alert whenever there are accom-
panying gastrointestinal symptoms such as nausea and vomiting (although both
of these may be present in acute coronary syndrome). Please see Case 2.12 below.
In the simultaneous presence of both chest and abdominal pain, the evaluation
must be extended for etiologies in both areas. Please see Case 2.13 below.
Be careful with patients with recent heart catheterizations showing ‘insignifi-
cant lesions.’ This term can be used very loosely by patients and by clinicians.
Its use can mean small 20% lesions to lesions that are much larger but not enough
for angioplasty or coronary stenting. In addition, although these lesions may be
too small for acute intervention, it does not mean that they are not at risk for
myocardial infarction. Early in our careers, we wanted to send patients with 30%
lesions of one coronary artery home. Fortunately, a cardiologist advised us not
to do so and stated that these patients ‘are the ones that tend to have myocardial
infarctions.’ In fact, Brian Holroyd, MD writes (from a study by PK Shah) in
Cardiovascular and Neurovascular Emergencies: implications for clinical practice,
‘retrospective analysis of serial angiograms, as well as prospective serial
angiographic observations, have suggested that in nearly two-thirds of all
patients presenting with acute ischemic syndromes, a coronary angiogram
performed weeks or months before the acute event had shown the culprit lesion
site to have <70% (often <50%) diameter narrowing.’13 To illustrate this point
further, it is written in Emergency Medicine Reports that ‘it now is understood that
coronary occlusion often develops in arteries that have minimal (10–40%)
stenosis at baseline.14
Extracardiac diseases reinforce our belief that just about all adults with chest
pain require an EKG and a chest X-ray. These are the two quickest tests to screen
for thoracic emergencies. They are also within the capabilities of most medical
offices. Therefore, get into the habit of ordering them routinely to avoid missing
diagnoses like Case 2.1 and Case 2.16.
We believe that the evaluation of chest pain relies more on the history than
most other complaints. This is because many etiologies such as acute coronary
syndrome, aortic dissection, pericarditis, and pulmonary embolism often have
normal physical exams. One cannot rely on the physical exam in the same
fashion as other complaints (e.g. abdominal tenderness for significant
pathology to abdominal pain). Therefore, in patients where obtaining the
history is difficult, extreme caution must be taken. Please see Case 2.17.
42 Learning from medical errors: clinical problems

Reproducible chest tenderness does not rule out serious heart or lung
problems. There have been studies showing that a significant number of
patients with acute coronary syndrome also had reproducible chest tenderness.
In addition, as Case 2.18 shows, pulmonary pathology may also occur with
chest tenderness.
There has been an abundance of literature produced recently concerning the
management of acute myocardial infarction. While the goals of these literatures
are to emphasize that protocols are followed in acute myocardial infarction, the
physician must also always remember to treat the individual patient at hand.
This means avoiding beta-blockers and nitroglycerin in hypotensive patients,
avoiding beta-blockers in bradycardic patients, and using these medications
cautiously in patients with inferior MIs. Please see Case 2.23 below.
In the outpatient setting, the diagnostic evaluation of chest pain is limited.
Cardiac enzymes are not immediately available. In addition, cardiology consul-
tation and/or cardiology reports may not be obtained. Hence the approach to
these patients must be more conservative. Patients with chest pain and an
abnormal EKG that cannot be shown to be old will probably need further
evaluation in the emergency department. In addition, patients with diabetes
and abnormal EKGs must not necessarily have chest pain to require further
evaluation. Please see Cases 2.24 and 2.25 below.
Coronary stents may develop restenosis, especially within the first few months
after placement. It is important to ask patients who present with chest pain after
having stents placed recently if they are taking their blood thinners (e.g. Plavix). If
they are not, then they are at extremely high risk for restenosis. In Case 2.27 below,
we present an unusual case of coronary stenosis in a patient despite Plavix use.
The final point is to not ignore the importance of family history in the
evaluation of acute coronary syndrome. Coronary artery disease has a strong
familial tendency. We have seen a number of patients with acute coronary
syndrome who had family history as the only risk factor. Please see the unfor-
tunate Case 2.28 below.

Errors and interesting cases

Case 2.1 Do not forget the basics
• An ED colleague once treated a patient with chest pain in a small emer-
gency department. He kept the patient in the ED for 12 hours while doing
serial EKGs and cardiac enzymes. The patient’s pain improved with
morphine. Since the tests were all normal, he ruled out acute myocardial
infarction and discharged the patient. The patient died two days later
due to an aortic dissection. The physician was faulted for not performing
a chest X-ray to look for a widened mediastinum.
• Although the sensitivity of chest X-rays for aortic dissection is extremely
low, it is occasionally helpful. The chest X-ray is a simple, common, and
fairly standard test in the evaluation of chest pain. Therefore, the test’s
sensitivity for aortic dissection becomes secondary when a diagnosis is
missed. The defendant now has the burden of explaining why the test
was not ordered and that evidence of an aortic dissection would not have
shown up on the X-ray (an almost impossible task).
 Chest pain 43

Case 2.2 Time is muscle

• Another incident involving an urgent care physician serves as an example
of mishandling acute chest pain in the outpatient setting. A 48-year-old
male with a family history of coronary artery disease presented to the
clinic at 8 pm complaining of 24 hours of substernal chest pain. He was
treated by the nurse practitioner at the clinic for gastroesophageal reflux
without any consideration for acute coronary syndrome or an EKG
(according to the clinic note). The patient returned the following morn-
ing stating that his chest pain had not improved with antacids. The urgent
care physician then performed an EKG, which showed 2 mm of ST
segment elevation in the inferior leads with reciprocal changes (classic
manifestations of acute myocardial infarction). The physician then
instructed the patient to go to the emergency department. An aspirin
was not given and an ambulance was not called. Because of the increased
risk of cardiogenic shock, arrhythmias, and syncope, the physician placed
his patient’s health and the public’s safety at risk by having a patient
with an ongoing heart attack drive a motorvehicle. He also delayed
important antiplatelet therapy with aspirin. The patient’s EKG on arrival
in the emergency department showed 3 mm of ST segment elevation in
the inferior leads and his troponin I was extremely elevated. He was
taken directly for cardiac catheterization and percutaneous coronary
intervention. However, a significant part of his cardiac damage might
have been prevented had an EKG been done or the possibility of acute
coronary syndrome been considered during the first visit in this patient
with cardiac risk factors. Remember that ‘time is muscle!’

Case 2.3 Do not be fooled by the age

• The following cases highlight the difficulties presented to the emergency
physician in cases of myocarditis and pericarditis. Selbst and Korin
present, in Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine, the case of a 15-year-old boy. He presented to the emergency
department with stabbing chest pain that was worse with minimal
exertion.15 He was hit in his right ribs one week earlier during a football
game. He stated that he was dizzy when he stood and had intermittent
episodes of diaphoresis, nausea, vomiting, and shortness of breath. His
vital signs were stable but he was febrile. He was borderline orthostatic.
There was tenderness with palpation of his chest but his heart and lung
exams were normal. His chest X-ray showed a ‘top’ normal heart size
and his EKG showed ST elevation in the inferolateral leads. Intravenous
fluids were given but this did not correct the patient’s orthostasis. The
case was discussed with the on-call cardiologist, who recommended
discharge of the patient with the diagnosis of musculoskeletal chest pain.
• The patient went into cardiac arrest and was taken to another hospital
six hours later, where he died. An autopsy revealed myocarditis. The
continued
44 Learning from medical errors: clinical problems

family sued the hospital and the emergency physician for failure to
diagnose myocarditis and for discharging the patient in an unstable
condition. The case concluded with an out-of-court settlement. The
cardiologist was not included in the suit because she claimed that she
never saw the patient.
• Age is an extremely important factor in the evaluation of chest pain.
Diseases of the heart, lungs, and aorta all tend to increase in prevalence
with increasing age. They are also rare in patients who are under 30 years
of age. Unfortunately, exceptions to these rules do occur and clinicians
should be alert to ‘red flags’ that may accompany them. This patient had
traumatic chest pain along with an abnormal chest X-ray and EKG. These
are not consistent with musculoskeletal chest pain. In addition, he had
symptoms (dizziness, nausea, vomiting, shortness of breath) and signs
(orthostasis, fever) to suggest disturbances in his hemodynamic regula-
tory systems. Therefore, by not recognizing the ‘red flags’, the two
physicians made a critical error in discharging this patient.

Case 2.4 What happens after the diagnosis is made?

• Our pericarditis case involves a 48-year-old male who presented to the
emergency department with two days of substernal chest pain without
radiation, diaphoresis, nausea, or vomiting. He had no prior history of
coronary artery disease and had no risk factors with the exception of
being a heavy smoker. He noted that the pain was worse when lying
down and better when he sat forward. His vitals were stable with the
exception of a mild tachycardia. An EKG showed diffuse ST segment
elevation and his cardiac enzymes were normal. The patient’s pain
improved after an intravenous injection of Toradol and he was discharged
with the diagnosis of acute pericarditis and given a prescription of
ibuprofen.
• The patient returned to the emergency department two days later with
worsening of his chest pain. As he was waiting to see the emergency
physician, he suddenly became hypotensive and went into cardiac arrest
shortly afterwards. An emergency pericardiocentesis and resuscitative
efforts were unsuccessful. The patient died from complications of
cardiac tamponade due to an increase in pericardial effusion from his
pericarditis.
• This case initiated a discussion among the emergency physicians and the
cardiologists in the hospital concerning the management of acute peri-
carditis. Should echocardiograms be performed on every patient with
pericarditis to look for a pericardial effusion? It certainly should be
ordered for patients with hypotension, respiratory distress, syncope, and
other signs and symptoms of hemodynamic compromise. However, a
consensus was not reached whether the test would be practical or cost-
effective in a patient with a normal hemodynamic state. This is because
the majority of patients with pericarditis will not have complications
and because of the decreased availability of echocardiograms.
 Chest pain 45

Case 2.5 Keep pulmonary embolus in the differential

• In a case from Edwards’ The M & M Files16, a 49-year-old obese women
with a history of tobacco and Premarin use presented to the emergency
department with weakness and pleuritic chest pain. Vital signs taken in
the department showed a pulse oximetery of 97% on room air, a pulse
of 100 beats/minute, and a respiratory rate of 20 breaths/minute. The
patient appeared comfortable and in no significant distress during the
examination. The ED physician ordered an EKG, which showed a slight
sinus tachycardia and nonspecific ST segment changes. Her blood work
revealed a slightly low potassium level but was otherwise normal, as
was her chest X-ray. The patient was discharged with the diagnoses of
possible viral syndrome and hypokalemia.
• The patient was brought back to the emergency department the follow-
ing day in cardiac-pulmonary arrest and was not able to be resuscitated.
Her cause of death at autopsy was a massive pulmonary embolus.
Emergency physicians treat many patients with similar symptoms who
do not have pulmonary embolus. After all, pleuritic chest pain is extremely
common and the patient’s oxygen saturation was relatively normal.
However, this particular patient had pleuritic chest pain along with factors
for deep venous thromboembolism and pulmonary embolism. Namely,
she was an obese smoker who was also taking hormones. Furthermore,
pulse oximetry can sometimes be falsely misleading, as in the case of this
patient. Patients can maintain high oxygen levels by increasing their
respiratory rate. As we discussed earlier, this patient’s respiratory rate
was not less than 20, thereby necessitating an increased suspicion of
further testing to rule out PE. The final important point is that this
patient never complained of shortness of breath. This further distracted
the emergency physician from considering an ongoing pulmonary
pathologic process.

Case 2.6 Once with heart disease, always with heart disease
• The following case from Edwards’ The M & M Files is of a 49-year-old
woman who presents with atypical chest complaints.17 She presented to
the ED complaining of productive cough and chills for several weeks.
She told the nurse but not the ED physician that she had a burning
sensation in her right chest. Her past history was significant for coronary
artery angioplasty during the prior year and the fact that she was still
smoking cigarettes.
• Since her vital signs in the ED were within normal limits and her
physical exam were not impressive for pneumonia or congestive heart
failure, the emergency physician elected not to order any diagnostic tests.
Instead, he discharged her with the diagnosis of acute bronchitis and
prescribed a course of antibiotics.
• Three days later, she returned by ambulance with classic symptoms of

continued
46 Learning from medical errors: clinical problems

myocardial infarction. She had crushing substernal chest pain with nausea
and diaphoresis and her electrocardiogram revealed an acute anterior
wall myocardial infarction. Although fibrinolytic therapy was started
immediately, the patient had a complicated course and was left with
poor residual left ventricular function. She is considering consulting an
attorney about the misdiagnosis of the first emergency physician.
• Because this patient did not complain to the physician of chest pain and
presented with atypical symptoms, is the physician following the stand-
ard of care for not pursuing a cardiac work-up? We feel that we would
be hard-pressed to say that the standard of care was followed because
of the patient’s known history of coronary artery disease. Furthermore,
it is well publicized that women will frequently present with atypical
signs and symptoms of cardiac ischemia. There is enough evidence here
that warrants a cardiac work-up. Edwards writes that ‘one point is
certain: when a complaint is missed or misunderstood, the onus lies with
the physician, not the patient.’18

Case 2.7 Drinks are not good for heart disease

• The next case by Edwards, in The M & M Files, is of a 42-year-old man
with chest pain. It illustrates how distracting the use of a GI cocktail
could be in a patient with cardiac risk factors.19 The man came to the ED
by ambulance with epigastric and substernal pain that radiated to the
left shoulder blade. The pain started after he had had lunch and was only
minimally relieved with nitroglycerin that was administered by the
paramedics. He did not have any diaphoresis, nausea, or shortness of
breath. He also had a history of gastrointestinal esophageal reflux
disease (GERD). Up to this point, this patient almost fits the profile of the
patient that we described above who may be given a GI cocktail
diagnostically. The exception was that this patient was 42 years old and
had a significant medical history.
• Upon further history, it was also discovered that he had a personal
history of hypertension and coronary artery disease (he had angioplasty
two years ago). He also had a family history of coronary artery disease
and was a one and one-half pack per day smoker. Therefore, he had
multiple risk factors and a GI cocktail would only be helpful as a
therapeutic and not diagnostic maneuver in the emergency department.
• The patient did not appear in any distress, had unremarkable vital signs,
and had a normal physical exam with the exception of some epigastric
tenderness. The electrocardiogram showed a normal sinus rhythm with
nonspecific findings. He was given a GI cocktail before consideration of
any nitrates. This medication made the patient pain-free and his cardiac
enzymes were all normal. An old electrocardiogram was obtained and
was without significant change from the current one. The ED physician
called the patient’s primary physician to discuss the case. Both
physicians agreed that the etiology of the pain was probably
continued
 Chest pain 47

gastrointestinal and the patient was discharged after a three-hour stay

in the department.
• The patient developed a recurrence of the pain on his drive home. This
pain was more severe and accompanied by sweating and nausea. Because
his family was upset with his discharge, they recommended that he drive
to another hospital 20 miles away. Here, he was hospitalized for unstable
angina and had triple vessel bypass surgery several days later. The
primary physician responded to the patient’s wife complaint by saying
that he was not given an accurate account of the presentation.
• Edwards gives some additional opinions on the use of GI cocktails in the
emergency department from the ones that we stated above: ‘It has long
been a truism that subjective relief after a GI cocktail is neither sensitive
nor specific for excluding cardiac causes for pain.’20 He also adds that
physicians should ‘consider banishing the GI cocktail’ from the diagnos-
tic evaluation of chest pain.20
• With this patient’s risk factors and cardiac history, acute coronary
syndrome must be excluded first. Nitroglycerin is more appropriate to
try for his pain before the administration of a GI Cocktail. In addition,
consultation with a cardiologist to discuss hospital admission or further
testing is warranted.

Case 2.8 A mouthful of problems

• Here is a strange case from Edwards’ The M & M Files21 that we added
for the interest of our readers. We will admit that this case would have
stumped us the first time that we saw it. Additionally, it would probably
continue to stump us if we had to face it again because we are not sure
we would change our practice based on one bizarre case.
• A 72-year-old woman came to the emergency room complaining of
intermittent tooth pain for two days that was getting worse. She could
not see her dentist until the following week. She did have a history of
hypertension and diabetes. She was seen in the fast track area of the ED.
Her examination was significant for tenderness over the temporoman-
dibular joint (TMJ) and the pain seemed to be worse when she opened
or closed her mouth. Her teeth were not tender. The patient was
diagnosed with TMJ syndrome versus early dental infection and was
prescribed ibuprofen.
• The patient went into full cardiac arrest three hours later and could not
be resuscitated once she was brought to the ED. Her autopsy revealed
severe blockage of her left anterior descending artery with evidence of
a recent myocardial infarction. The patient’s family contacted an
attorney and eventually received a settlement.
• As we mentioned above, we would not necessarily change our decision
process by reading this case with the exception of maybe asking the patient
if she had any associated symptoms of a myocardial infarction.
Performing an electrocardiogram or blood work in everyone with
continued
48 Learning from medical errors: clinical problems

dental pain and cardiac risk factors is impractical. Furthermore, this

patient’s defense could have argued that that patient was a diabetic and
she probably had a silent myocardial infarction and that her dental pain
was from TMJ. The defense would have benefited from hiring a dentist
to do a post-mortem examination of this lady’s TMJ to confirm this. Finally,
the defense could have taken the stand that misdiagnosing acute
coronary syndrome based on a complaint of toothache is within the
standard of care. Although there may be one or two medical texts that
list dental pain as a symptom of ACS, the majority of texts do not.

Case 2.9 Once is not enough

• Edwards presents in The M & M Files a 71-year-old diabetic who came
to the ED with flu-like symptoms for several days.22 She complained of
a cough with yellow sputum, sore throat, loss of appetite, and some
intermittent pain in her upper chest. The pain started a few hours before
her presentation, radiated to her left shoulder, and resolved in one hour.
Her cardiac risk factors included a history of hypertension, angina, and
diabetes.
• The ED physician ordered a chest X-ray, electrocardiogram, and a troponin
level. All three tests were interpreted as normal. The patient did not have
any further pain. She was then discharged with the diagnosis of
bronchitis and flu-like syndrome and given a prescription for
antibiotics. Her chest pain recurred the following day and she went to
see her primary physician. The electrocardiogram now showed changes
consistent with an inferior myocardial infarction. The primary physician
admitted the patient to the hospital and called the ED to ask why the ED
physician discharged the patient after a single troponin level. It is
difficult to justify the actions of this emergency physician based on the
patient’s risk factors and the limitations of a single troponin level in the
evaluation of hyperacute chest pain.

Case 2.10 Appearances can be deceiving

• A 48-year-old man came to the ED complaining of intermittent left upper
chest pain for 3–4 weeks. The pain was felt in his left shoulder but was
not associated with diaphoresis, shortness of breath, nausea, or
vomiting. His current episode of pain had been present for two hours
but he denied that the episodes were occurring more frequently or
becoming more severe. He had never seen a physician for the evaluation
of chest pain. He added that he had been congested for a few days and
the pain was worse when he moves.
• The patient did not appear to be in distress when we saw him. We
assessed his cardiac risk factors as being a heavy smoker and having a
family history of coronary artery disease. We obtained an EKG, which

continued
 Chest pain 49

showed a normal sinus rhythm and very subtle (less than ½ mm) ST
segment depression in leads V2-V4. This did not impress us much
initially, but we ordered cardiac enzymes and asked the nurse to give
him aspirin and nitroglycerin for the pain.
• When his cardiac enzymes returned as normal, we returned to ask him
if the nitroglycerin had helped. He stated that he had not been given any.
His nurse informed us that he had told her that he was not having pain
so she did not give him any. At this point, we got the impression that this
man was stoic in his pain response and would be difficult to evaluate
subjectively. The patient responded that he ‘had a little pressure’ but no
pain when we questioned him again. Nitroglycerin was then
administered and the patient was uncertain whether there was improve-
ment in his chest pressure. A repeat EKG was performed which showed
normalization of his ST depression after the nitroglycerin. This convinced
us that this patient needed a cardiology consultation despite his
symptoms being atypical in nature.
• When the cardiologist reviewed the patient’s initial EKG, she was
similarly not impressed with the tiny ST segment depressions. After she
was given the second EKG and given a few minutes to review both
tracings, she agreed that the ST segment depression was significant and
immediately took the patient for cardiac catheterization. During the
procedure, the patient was found to have severe triple vessel blockage
and was scheduled for coronary artery bypass graft surgery the next day.
• This case highlights the importance of multiple EKGs in the evaluation
of chest pain. The first EKG, by itself, was not convincing of an acute
coronary syndrome. Comparison of the first EKG with the second EKG,
in combination with the clinical setting, made the diagnosis of acute
coronary syndrome much more likely. The case also demonstrates the
difficulty of evaluating patients who may be stoic or may use different
terms to describe their chest discomfort.

Case 2.11 Common link between cardiac and neurologic diseases

• A 70-year-old man with a history of coronary artery disease and previous
coronary artery bypass graft surgery presented to our ED with the
complaint of six hours of left-sided chest pain and numbness and
weakness on his left side. His exam revealed no slurring of speech or
facial droop but he was weaker in his left extremities in comparison to
his right extremities. There was no substantial difference in his right and
left arm blood pressures. His EKG did not show any acute changes and
he stated that nitroglycerin did not help his chest pain. He repeatedly
requested that we give him morphine for the pain because he said that
it is the only medicine that helped him.
• His cardiac enzymes were negative and a repeat EKG did not show any
changes. His head CT, however, showed two infarcts on the right side

continued
50 Learning from medical errors: clinical problems

of his brain that was of uncertain age because he did not have any prior
scans for comparison. We considered the diagnosis of aortic dissection
but decided to have the patient admitted for possible stroke and hold the
CT scan because the patient’s creatinine was 2.2 mg/dl. Although we
were taking a risk (in retrospect, a risk that we should not have taken)
by not investigating for a dissection, there were three reasons why we
did not do so. The first was the patient’s creatinine. The radiologists at
our hospital set a creatinine value of 2.0 mg/dl as the cutoff for doing
a contrast study CT. They recommend that anyone with a creatinine
greater than 2.0 mg/dl be admitted and hydrated before the scan is
performed. This is to avoid causing iatrogenic renal failure to a patient.
It is also extremely difficult to obtain a transesophageal echocardiogram
at night due to the paucity of qualified technicians. Finally, the patient
did not have a widened mediastinum on chest X-ray or a significant
blood pressure discrepancy among the arms. He also had a possible
etiology for the left-sided neurologic symptoms with his abnormal head
CT.
• The admitting physician, however, was not as satisfied with the
evaluation as we were. He insisted on obtaining a CT of the chest with
intravenous contrast. After an extensive discussion with two different
radiologists who were both hesitant about performing the procedure, he
finally convinced them to proceed with it. The CT scan showed edema
surrounding the aortic arch without a definite intimal flap. The
diagnosis of aortic dissection was felt to be possible but not confirmed
with this study. We lost the patient to follow up and never learned whether
he had an aortic dissection.

Case 2.12 Common link between cardiac and gastrointestinal

diseases
• An 88-year-old woman with a history of coronary artery disease
presented to our ED at 3 am with severe substernal chest pain that started
one hour earlier. There was no radiation of the pain but it was associated
with two bouts of nausea and vomiting. She had a cardiac catheteriza-
tion less than one year ago that showed severe coronary artery disease
and was felt to be a poor candidate for percutaneous coronary interven-
tion. She also had a history of gallstones. Her EKG revealed her usual
chronic atrial fibrillation with T wave inversions in the lateral precordial
leads that were more prominent than those of her previous EKGs.
• With the suspicion of an acute coronary syndrome in progress, she was
given an aspirin and a sublingual nitroglycerin. After the nitroglycerin,
she became extremely nauseated and vomited. Intravenous morphine
was then attempted for the chest pain along with ondansetron for
the nausea. This seemed to make the patient’s pain worse and she
vomited again. When her repeat EKG did not show any difference,
we became suspicious that another etiologic process was present.
continued
 Chest pain 51

• We ordered liver function tests along with serum amylase and lipase. In
the meantime, we continued to treat her symptoms with a nitroglycerin
paste and intravenous Phenergan for nausea. We also administered
intravenous Demerol for pain and this seemed to help her the most. The
persistent vomiting, intolerance to morphine, and improvement with
Demerol made us suspicious of an acute biliary attack. Although
morphine is very good for the pain with cardiac ischemia, it is believed
to be a poor choice of narcotic for the pain with acute pancreatitis. This
is due to the belief that morphine causes spasms of the sphincter of Oddi.23
We asked the laboratory to add on liver function tests and serum
amylase and lipase to the patient’s blood work.
• Her cardiac enzymes returned as normal but her liver transaminases
were slightly elevated. The serum amylase and lipase, however,
confirmed our clinical suspicion as they were both well over 1,000 units/
L. The patient was admitted to the hospital and was ruled out for acute
myocardial infarction with serial cardiac enzymes. She was subsequently
treated medically for gallstone pancreatitis.
• Although we were able to pinpoint the correct diagnosis in this case,
there have been a few instances where we were not as fortunate. We can
remember three instances where we admitted a patient to the cardiology
service for acute coronary syndrome only to find out later that biliary
disease was the cause of their chest pain. All three patients presented
with crushing chest pain accompanied by diaphoresis, nausea, and
vomiting. None complained of abdominal pain. Multiple risk factors for
coronary artery disease were present in all three. Embarrassingly, two of
these patients were seen in the same night. However, we emphasize that
we still believe that these patients received appropriate care because a
more serious diagnosis was ruled out. This is in contrast to the situation
where a patient with acute coronary syndrome is discharged with the
diagnosis of cholelithiasis.

Case 2.13 Not the whole story

• A 40-year-old male presented to our ED complaining of pains in the left
abdomen, left chest and left shoulder since 10 am. He denied any fevers,
nausea, or vomiting but did appear to be in moderate distress from the
pain and was slightly diaphoretic. He could not lie flat because this made
the pain worse. He denied any recent traumatic event and had no chronic
medical problems. He admitted to heavy use of cigarettes, alcohol, and
marijuana.
• Physical exam was limited because the patient would not lie flat.
Otherwise, it was essentially unremarkable with the exception of some
nonspecific tenderness in his left upper abdomen. An entire cardiac work-
up (EKG, chest X-ray, cardiac enzymes, and complete blood count) was
unremarkable except for a slightly elevated white blood cell count of
14,900 cells/microliter. Liver function tests along with serum amylase
continued
52 Learning from medical errors: clinical problems

and lipase were also unremarkable. His drug screen, however, was
positive for marijuana and cocaine. An abdominal X-ray did not show
any evidence of free air or obstructive bowel gas pattern.
• A CT scan was ordered to evaluate the patient’s abdomen. It showed
multiple linear lacerations within the spleen, blood surrounding the
spleen, and free fluid in the pelvis. The patient was questioned again by
us and later by the surgeon regarding any recent abdominal trauma. He
repeatedly denied any traumatic event to his abdomen and was treated
conservatively in the hospital by the surgeon without any complications.
• Although the diagnostic evaluation for ACS is usually a focused one, the
physician must not forget that chest pain can be caused by other organ
systems. In this particular patient, his left chest and shoulder pains were
believed to arise from diaphragmatic irritation caused by the blood
surrounding the spleen. If we had not taken noticed of the tenderness in
his left upper abdomen and his elevated white blood cell count, the correct
diagnosis would probably have been missed.

Case 2.14 From one emergency to another

• We recently had an interesting case of chest pain in our ED. A 36-year-
old female with no medical problems presented with 12 hours of severe
chest pain and shortness of breath. She did not smoke or drink alcohol
but did have a significant family history for coronary artery disease. She
had no prior cardiac evaluation and no prior EKGs.
• On exam, her pulse was in the 130s (beats/minute). The rest of her vitals
were in the normal range. She had some reproducible tenderness in the
epigastric area. She had taken her father’s nitroglycerin and noted
improvement in the pain. An EKG revealed a left bundle branch block.
We ordered a cardiac work-up, liver function tests, amylase, and lipase
because of her abdominal tenderness.
• Because of her EKG, we were concerned about an acute MI and called
the cardiologist on call. We started the patient on aspirin, nitroglycerin,
heparin, and Lopressor. The cardiologist came to see the patient and felt
that it was gastrointestinal and asked us to stop the cardiac medicines
and discharge her. The patient was ready to go home at that point and
did not want any bloodwork done. We insisted that she stay for the
bloodwork.
• Her white blood cell count was 23,000 cells/microliter with 23% bands on
the differential. The liver function tests, amylase, and lipase, however,
were all normal. An ultrasound of the gallbladder was performed and
showed multiple gallstones with mild gallbladder wall thickening and
pericholecystic fluid. The patient was admitted to the surgical service and
had a cholecystectomy the following day. She was scheduled to follow up
with the cardiologist for outpatient evaluation of her abnormal EKG. We
thought that it was interesting how a potential emergency of one organ
(acute MI) turned out be an emergency of another organ.
continued
 Chest pain 53

• It is important to review all ordered tests before discharging a patient

with chest pain. We have seen some patients discharged after the CPK
returned but the troponin was still pending. In some instances, patients
were discharged when their cardiac enzymes were normal and the
patients were later discovered to have chest pain due to severe anemia.
In one case, the physician reviewed the patient’s blood work but did not
look at the patient’s chest X-ray. The patient returned and a pneumo-
thorax was discovered on the chest X-ray from the first visit. With regards
to this patient, her extremely elevated white blood cell count generated
an investigation for an infectious etiology.

Case 2.15 EKG first, questions later

• Regardless of the preconception, an adult patient complaining of chest
pain should always get an expedient EKG. This case from Selbst and
Korin’s Preventing Malpractice Lawsuits in Pediatric Emergency Medicine is
an example where the EKG was delayed on a patient based on her prior
visits.24 A 65-year-old woman had two ED visits within a span of two
weeks for chest and epigastric pain with arm tingling and sweating.
Electrocardiograms were performed and she was discharged on both
occasions with the diagnosis of ‘reflux esophagitis.’ When she returned
for a third visit three days later, the chest pain was radiating to her arms
and associated with nausea. The triage nurse, however, made the patient
wait based on her prior visits. The patient left without being seen after
waiting for three hours.
• The following morning, she presented with similar symptoms and an
EKG showed an anterior myocardial infarction. The delayed treatment
for it left her with significant unrecoverable heart muscle and decreased
her life expectancy. The patient filed a lawsuit against the emergency
department and the case was settled for $100,000.
• Triage nurses can develop opinions of patients based on the patient’s
prior visits. This bias may prevent the patient from getting optimal care.
Chest pain must always be taken seriously to avoid missing an
inexcusable mistake. In our facility, there is a sign in the waiting room
that states that you must tell the nurse at once if you are having chest
pain. We will occasionally hear a triage nurse state that a patient is just
saying that they have chest pain so that they will be seen earlier. With
few exceptions, we still tell the nurse to follow standard protocol.

Case 2.16 What is up (or down) with the lungs?

• A 45-year-old man with a ‘coronary artery disease’ look came by
ambulance to our ED for chest pain. He was a three pack per day
cigarette smoker and looked every bit of it. He had a thin, emphysematous
chest, hoarse voice, and prolonged expiratory phase breathing. Other
risk factors included hypertension and a family history. Although his
continued
54 Learning from medical errors: clinical problems

chest pain was improved when the paramedics gave him nitroglycerin,
his description of the pain was very atypical.
• He stated that the pain started two hours ago in the right parasternal area
and radiated to his right back and right upper quadrant. After the nitro-
glycerin, the pain was located substernally. The pain was associated with
profuse sweating and nausea. He had no prior cardiac studies.
• On examination, the patient did not have any heart murmurs or
evidence of fluid overload. His lung exam was significant for diffuse
expiratory wheezes. An EKG revealed normal sinus rhythm with no acute
ST segment changes. His chest X-ray, however, showed a 50% pneu-
mothorax on the right side without evidence of tension.
• Our medical student readers are probably wondering why this pneu-
mothorax was not picked up clinically. In the actual emergency setting,
acute coronary syndrome is much more common than pneumothorax.
Hence, the initial history is often focused on the details associated with
coronary artery disease. When we took a second history from this
patient, he admitted that he was having violent coughing spells prior to
the onset of the pain. How did we miss the absence of breath sounds
from a pneumothorax? It is a rare moment in the ED when there is no
shouting, crying, moaning, or mechanical sounds. Therefore, ausculta-
tion is not performed under the most optimal circumstances.
• In the perfect environment with ample time to take a good history, a
chest X-ray may not be needed for a pneumothorax of this size. How-
ever, most are much smaller and are difficult to pick up clinically.
Therefore, the results of a simple chest X-ray can prove to be extremely
fruitful.

Case 2.17 Talk to me

• Medical facilities have obligations by law to provide a translator for
patients who do not speak English. There are times, however, when a
translator is not available. A 46-year-old migrant worker came to our ED
complaining of chest pain. He did not speak any English and came alone.
Although the hospital usually had employees that spoke Spanish that
could translate for us, there was no one working at 3 am. We called the
24-hour translator service for a telephone translator. The Spanish trans-
lator was on sick leave for the day.
• We could not assess the patient’s pain duration, characteristics, associ-
ated factors, etc. He did not comprehend when we asked about
hypertension, diabetes, hyperlipidemia, or family history of coronary
artery disease. The only information we were able to obtain was that the
patient smoked (there was a pack of cigarettes in his shirt pocket) and
that he was seen at the other ED in town (he presented a discharge sheet
from the hospital). It was even difficult for us to determine if the patient’s
pain improved when we gave him nitroglycerin.
• We ordered a cardiac work-up and tried to obtain any medical records
continued
 Chest pain 55

from the other hospital. His EKG and blood work were unremarkable.
Repeated attempts at obtaining medical records were unsuccessful.
Although it was hard to convince the cardiologist that his pain
represented acute coronary syndrome, we felt that we had to admit the
patient until further information could be obtained or further testing
performed.
• The dilemma of evaluating a patient who is not able to give a history is
also present in other situations. Patients with dementia, psychiatric
histories, or alcohol intoxication are other examples. Physicians must
use caution to prevent making errors in the evaluation. In addition,
attempts should be made to obtain information from alternate sources
(e.g. medical records, family members, etc.). If there is inadequate
information to comfortably discharge a patient, then the patient should
probably be admitted.

Case 2.18 Do not forget about pulmonary embolus

• Wilderson and Wagner present a case, in Foresight, where a physician let
a clinical finding distract him from further evaluation.25 A 23-year-old
presented to the ED with ‘stabbing’ right-sided chest pain for two days.
The pain was worse with movement and prevented him from taking
deep breaths. He denied any shortness of breath or chest trauma. He was
confined to a wheelchair due to severe paresis of his lower extremities.
The weakness resulted from a hypoxic injury to his spinal cord from a
motor vehicle accident.
• Recorded vital signs in the chart were: blood pressure 135/89 mm Hg,
pulse 105 beats/minute, respiratory rate 26 breaths/minute, tempera-
ture 97.8ºF, and pulse oximetry 95% on room air. The ED physician found
localized tenderness over his right pectoralis muscle. The pain was re-
produced when he moved the patient’s right shoulder and trunk. Heart
and lung evaluations were unremarkable. There was atrophy and no
swelling of the legs. The chest X-ray was read as normal. The patient was
sent home with the diagnosis of chest muscle strain and told to take
ibuprofen.
• The patient returned to the ED the following day with worsening pain
and dizziness. The pain was especially worse with breathing. Although
blood pressure and respiratory rate were essentially unchanged, his pulse
and temperature had increased to 120 beats/minute and 100.4ºF
respectively. In addition, his pulse oximetry had decreased to 92% on
room air. This later decreased to 87% on room air during the physical
examination. His breathing became more agonal and he was intubated.
A pulmonary CT scan revealed multiple large pulmonary emboli. The
patient subsequently had a prolonged and complicated hospital course.
A lawsuit was filed against the ED physician and the hospital for a delay
in diagnosis.
• The ED physician in this case was distracted by the musculoskeletal
continued
56 Learning from medical errors: clinical problems

findings on physical examination. As the authors of the case point out,

‘chest wall tenderness may be a nonspecific finding, but it is present in
a significant proportion of patients with pulmonary embolism.’26 This
distraction prevented the physician from recognizing the red flags of
pulmonary embolism. These included: immobilization, tachycardia,
tachypnea, and low pulse oximetry. Further diagnostic tests were not
initiated due to the failure to acknowledge these red flags. These factors
led the defense team to conclude that the case would be extremely
difficult to defend. Therefore, the hospital and the physician agreed to
a settlement.

Case 2.19 Two things at once

• A second case from Wilderson and Wagner in Foresight illustrates the
problems with a delay in diagnosis of an aortic dissection.27 A 42-year-
old man presented to the ED with a right-sided headache and dizziness.
He stated that he ‘can’t stand this light’ in triage ‘and my arm feels weird.’
He also added that he had about 15 minutes of pain in his chest and back
a few hours ago.
• On physical examination, he was mildly hypertensive with weakness of
the left arm. An EKG was unremarkable. His chest X-ray showed a normal
sized mediastinum. The preliminary reading on the head CT was nega-
tive. The ED physician made a diagnosis of acute CVA and consulted the
on-call neurologist. The patient is admitted to the neurology service and
heparin therapy is initiated. Eight hours later, the patient developed a
worsening left hemiparesis. He went into cardiac arrest and died. The
cause of death at autopsy was an aortic arch dissection that had
progressed into the right carotid artery. The final head CT report noted
early ischemic changes in the right hemisphere.
• A few months later, a wrongful death lawsuit was filed against the ED
physician and the neurologist. The hospital attorney was concerned with
the iatrogenic damage from the administration of heparin and the
incorrect preliminary head CT reading. He did not feel that the case was
defendable and recommended that both defendants settle out of court.
• The prevalence of aortic dissection is much less than that of either
cerebrovascular accident (CVA) or acute coronary syndrome (ACS). How-
ever, vigilance to its presence is mandated in order to make an early
diagnosis. Timely treatment significantly decreases the morbidity
associated with this disease. In addition, treatment that is standard for
CVAs or ACS may be contraindicated in aortic dissection (i.e. heparin
therapy).

Case 2.20 Check them again

• This case by Wilderson and Wagner from Foresight illustrates the benefits
of obtaining a repeat chest X-ray in the patient complaining of chest pain.28
continued
 Chest pain 57

It also is a good example of how a skilled clinician can use two simple
tests to make a difficult diagnosis. A 38-year-old woman had been having
chest pain for five weeks for which she had already made two ED visits.
When the pain became worse, she made her third visit to the ED. The
pain was described as ‘stabbing’, substernal, and associated with
exertional shortness of breath. There was also some pain in her left
shoulder.
• In the emergency department, her vital signs were normal and her
physical examination was unremarkable. The ED physician reviewed
the patient’s old charts. She was diagnosed with musculoskeletal chest
wall pain on both occasions and given pain medications. The second
physician documented that she did not follow instructions to see her
primary physician and repeated this recommendation. Neither
physician had ordered an EKG on the patient.
• The physician ordered an EKG and noticed that there was variability in
the size of the R wave with every other beat. This was consistent with
a phenomenon known as electrical alternans. He also ordered a chest X-
ray, which revealed a mildly prominent heart. The relative size of the
heart, however, was much larger than on a chest X-ray one month earlier.
This led the physician to order a CT scan of the chest. Computed
Tomography findings were a lung mass and a large pericardial effusion.
The patient was admitted and subsequently diagnosed with lung
carcinoma.
• We see how important the chest X-ray and the EKG are in the evaluation
of chest pain. Furthermore, maintain a low threshold for repeating them
if the patient returns and the diagnosis remains uncertain. These two
tests not only provide a wealth of information, they are also helpful when
used in comparison to prior studies.

Case 2.21 Not likely to be usual

• Joseph presents a case, in Foresight, of atypical symptoms of ACS.29 A 37-
year-old female recently joined a health club. After each class, she noticed
pain in her left shoulder. By the fourth class, the pain started radiating
to her left arm and was associated with nausea. This led her to the ED
for evaluation.
• Her chief complaint was ‘left shoulder pain after aerobics.’ She was triaged
to the fast track section of the ED to receive X-rays. An intern evaluated
the patient initially and then presented to the ED attending. They both
examined her shoulder and reviewed her X-rays. After not finding any
significant findings on her evaluation, they sent her home and told her
to spend more time warming up and take ibuprofen for the pain.
• Later that morning, the patient’s daughter called the paramedics after
her mother had developed severe indigestion. When the paramedics
arrived, the patient complained of chest pain. An on-site EKG revealed
5 mm of ST elevations in lead I, lead aVL, and the anterior leads with
continued
58 Learning from medical errors: clinical problems

reciprocal changes in the inferior leads. During transportation, the

patient developed ventricular fibrillation and cardiac arrest. Resuscita-
tion attempts were unsuccessful and the patient was pronounced dead
in the ED. It was discovered by the paramedics that the patient had a
history of diabetes.
• The patient’s family filed a lawsuit of medical negligence and wrongful
death against the ED physician six months later. The plaintiff ’s
arguments consisted of: inappropriate triage of the patient, inadequate
history from the ED physician, and negligent evaluation of the patient’s
symptoms. The physician realized that these arguments were extremely
difficult to defend and agreed to settle for a significant amount.
• Risk management issues in this case extend beyond the evaluation of the
ED physician. The ED physician has a responsibility to be aware of
atypical presentations of ACS. They are not uncommon in the premeno-
pausal patient. He should also always obtain a thorough history from
the patient. Be skeptical if the history or physical does not fit the
symptoms (there was no trauma or reproducible pain). Furthermore, it
is the attending physician’s responsibility to be the most vigilant
provider in the department. If an intern provides an inadequate
evaluation or a nurse performs an inappropriate triage, the attending
physician must correct these errors in a constructive manner.

Case 2.22 Keep them coming

• The second case from Joseph, in Foresight, shows the usefulness of serial
EKGs in ensuring that treatment is initiated timely.30 A 58-year-old woman
woke up with chest pain and shortness of breath. When she arrived by
ambulance to the ED, she is pain-free and had a normal EKG. Her
husband was extremely worried that his wife had a heart attack. The ED
physician assured him that he will take good care of his wife and
suggested that he go get a cup of coffee. The man agreed with some
reluctance.
• The physician then talked to the patient and discovered that she was a
smoker and that her mother had died of a heart attack. The patient’s pain
lasted less than 30 minutes and was relieved with oxygen given by the
paramedics. Concerned that the patient may have coronary artery
disease, the physician told her that he is going to run some tests and call
her physician to discuss admission. As he walked out of the room, the
patient’s chest pressure started to return. The physician ordered a stat
repeat EKG, one every 30 minutes times three, and as needed for pain.
He also asked for a cardiac profile, an aspirin for the patient, and a phone
call to the patient’s primary physician.
• The second EKG was performed 15 minutes later, shortly before the ED
physicians’ shifts changed. When it was time for the third EKG to be
performed, the on-coming ED physician was informed that the patient’s
pressure was becoming worse. The new physician looked at the second
continued
 Chest pain 59

EKG, which showed <1 mm ST elevation in leads I and aVL with recip-
rocal changes in the inferior leads. He entered her room and found that
the third EKG was being done and the patient’s husband was in the
room. The husband questioned, ‘What’s going on and what happened
to the other doctor and nurse?’31 As the physician started to explain, he
noticed that the third EKG shows >4 mm ST elevation in leads I and aVL
with reciprocal changes.
• The physician contacted the cardiologist immediately concerning the
patient’s acute myocardial infarction. The patient was taken for cardiac
catheterization and angioplasty. The procedure was complicated by the
patient’s worsening hypotension and arrhythmias. She died about 3½
hours after she was first seen in the ED.
• A lawsuit was filed against all of the patient’s providers except the
paramedics. The first ED physician was charged by the plaintiff as
demonstrating a ‘lackadaisical attitude’ towards the patient and for
‘abandonment’ of the patient.31 The paramedics testified that they
noticed a ‘carefree attitude’ in the ED that day. They also stated that a
nurse had proclaimed that the patient was a ‘false alarm.’ The plaintiff
attorney pointed out the fact that the physician sent the husband off
while his wife was dying.31
• The defense arguments were that the patient was treated appropriately
and timely. It was the timely order of tests that the heart attack was
detected early in evolution. They contended that the first physician was
appropriate in administering aspirin, calling the patient’s physician for
admission, and ordering serial tests. They also demonstrated that the
second physician and the cardiologist responded quickly to the patient’s
condition. They insisted that the patient’s outcome was dictated by the
nature of the event and was unpreventable. The jury returned a verdict
in favor of the defense.
• Although the first ED physician laid the correct groundwork in treating
this patient, there are several steps that he could have taken to decrease
the husband’s perception of medical malpractice. The patient started to
develop the chest pressure approximately 30 minutes before his shift
ended. Because the 5–10 minutes before and after shift change is usually
spent standing around, he should have made sure that the second EKG
was done prior to his departure. In terms of risk management, it is
usually not good to sign over a patient whose condition is deteriorating.
By identifying the subtle but definite changes on the second EKG, the
physician would have acknowledged ACS sooner and called the
cardiologist sooner. In addition, after a physician turns over a high-
risk patient, it is prudent to give the patient an update before he leaves.
This gives the patient a sense of closure and avoids a sense of abandon-
ment.
60 Learning from medical errors: clinical problems

Case 2.23 Treat it differently

• The final case from Joseph in Foresight highlights the difficulty with
managing acute inferior myocardial infarctions.32 These infarctions are
occasionally accompanied by bradycardia and hypotension. Hence,
standard medications such as nitroglycerin and beta-blockers must be
used with caution because they may be detrimental. Blood pressures
and pulses must be monitored frequently in order to guide therapy.
• A female patient developed epigastric pain while having dinner. She
believed that this was a gallbladder attack and went back to her hotel.
A few hours later, the pain became worse and was associated with
nausea. She called her husband and he consulted their family physician.
Their physician recommended that she either go to the ED, or alterna-
tively, try and wait for the pain to abate. She waited a few hours, then
started to vomit and called a taxi.
• When she arrived at the ED, she was too weak to get out of the cab. Initial
vital signs were: blood pressure 110/62 mm Hg, pulse 52 beats/minute,
respiratory rate 16 breaths/minute, and temperature of 98.4ºF. She
complained of light-headedness upon sitting. The physician did not elicit
any significant findings on physical exam and ordered an EKG,
complete blood count, amylase, urinalysis, abdominal X-rays, and a ‘GI
cocktail.’
• Thirty-five minutes later, the EKG was performed and showed a sinus
rhythm at 52 beats/minute, an occasional premature atrial contraction,
and ST elevation in leads II, III, and aVF. Upon seeing this EKG, the ED
physician diagnosed acute inferior myocardial infarction and asked the
nurse to start a nitroglycerin drip and titrate it for the patient’s pain. He
also requested fibrinolytic therapy and morphine.
• After these medications were administered, the patient became
unresponsive. The patient was now in pulseless electrical activity and
resuscitation attempts were unsuccessful. Her husband filed a lawsuit
against the hospital and the ED physician two months later. The basis of
the lawsuit was that the delay in diagnosis prevented effective
treatment.
• The plaintiff attorney made insightful arguments that led the defense to
settle the case. He emphasized that there was a 35-minute delay in get-
ting the EKG. He believed that the reason for this delay was that the
physician was focused on an abdominal etiology. This belief was
supported by the physician’s order of a GI cocktail, blood work for
abdominal processes, and the absence of a cardiac panel. The attorney
was also clever to realize that repeat vital signs were not performed before
the nitroglycerin and morphine were given. Other deficiencies in care
that he found included: the omission of aspirin, the absence of a
cardiology consultation, and the lack of orthostatic vital signs. In
addition, he summoned the cardiologist that was on call that morning
to testify. The cardiologist addressed his hesitation with using morphine
and nitroglycerin in this patient.
continued
 Chest pain 61

• We can see from this case that all acute myocardial infarctions cannot be
treated similarly. We must be cognizant of the patient’s hemodynamic
status and take this into consideration. We must also be alert to the
potential complications of MIs to specific areas of the heart.

Case 2.24 It cannot wait

• Rice presents a case, in Medical Economics, where an abnormal EKG from
the office was not addressed emergently.33 This resulted in a preventable
complication. A 50-year-old man told his family physician that he was
having chest pain and other symptoms of coronary artery disease. The
man was a smoker and had a history of hypertension. The physician
ordered an EKG, which was abnormal. Based on this test and the
patient’s history, the physician diagnosed chronic obstructive
pulmonary disease, CAD, CHF, and ischemia. The patient was started
on a diuretic and an ace inhibitor. His physician also said that he was
going to refer him to a cardiologist.
• The physician, however, felt ‘no particular urgency’ to call the cardiolo-
gist until two days later.34 After speaking with the cardiologist and faxing
the EKG to him, the physician wrote that the cardiologist’s office would
arrange an appointment for the patient. He anticipated that an
angiogram would be scheduled. The cardiologist later testified that he
was willing to see the patient that day but the family physician did not
feel that it was necessary. The appointment was made four days from the
telephone conversation.
• The patient returned to his physician for a follow-up the next day. He
had some improvement in his symptoms. Therefore, his physician made
no further attempt to contact the cardiologist. On the evening before his
scheduled referral, the patient had an acute cardiac arrest and died. The
patient’s family filed a lawsuit against the family practitioner and the
cardiologist. The family practitioner settled the case while the
cardiologist was dropped from the case.
• This case reminds us of the difficulty in evaluating acute coronary
syndrome in the office. Tests and consultations are limited. Therefore, be
extremely conservative and refer the patient to the emergency
department (or to the direct care of a cardiologist) if there is a concern
for acute coronary syndrome. It certainly is not prudent to diagnose
a patient with ‘ischemia’ and not arrange an immediate cardiac
evaluation.

Case 2.25 Treat them gingerly

• This case illustrates the importance of conservative treatment of diabetics
with suspected coronary artery disease. We received a phone call from
a cardiologist one morning. He said he had a diabetic patient with an
abnormal EKG in his clinic. The patient was not his regular patient and
continued
62 Learning from medical errors: clinical problems

had come for a routine scheduled EKG. He did not have any symptoms
and had no prior EKGs in the clinic or the hospital (the clinic was located
in the hospital). There was slight upcaving ST segment elevation of less
than 1 cm in the inferior leads and T wave inversions throughout the
precordial leads. The cardiologist said that the patient was not having
any symptoms and did not need any immediate treatment but needed
to have some blood drawn. We agreed and he sent the patient to us.
• When the patient arrived, he was wondering why he was sent to us. He
said that he was going to miss his golf game in 30 minutes. He also denied
that he could be having a heart attack and said that he was not going to
stay. We ordered a cardiac panel and expected it to be normal. We were
probably still going to recommend that the patient be admitted, how-
ever, for the abnormal EKG. He made it clear though that he would not
stay in the hospital if his labs were normal. His CPK was within the
normal range. His troponin I, however, was three times the upper limit
of normal. We started medications for acute myocardial infarction and
admitted the patient to the cardiology service. This case taught us that
for diabetics, appearances can certainly be deceiving.

Case 2.26 Two test minimum

• For most young patients with chest pain, a chest X-ray and an EKG are
usually sufficient. We talked about some exceptions in the prior cases.
Another exception would be that of rhabdomyolysis. Patients who
perform strenuous physical activity or use cocaine are the groups that
are at the most risk.
• An 18-year-old athletic male came to the ED by ambulance for chest pain.
The pain had started the night before. It was not associated with
sweating, nausea, or vomiting. He also complained of pain in both arms.
He had no medical history except for a recent diagnosis of attention deficit
disorder. His physician had started him on Adderall (amphetamine) two
days ago. Although he denied tobacco use, he admitted to smoking
marijuana.
• In the ED, his vital signs and physical exam were unremarkable. His
EKG showed sinus tachycardia with a rate of 111 beats/minute. A chest
X-ray did not reveal any acute disease. Urine drug screen was positive
for marijuana but negative for cocaine. The ED physician discharged the
patient with the diagnosis of stimulant induced chest pain. He instructed
the patient to stop the Adderall and follow up with his physician in two
days.
• Later that evening, the patient returned because his pain was worse. A
second ED physician obtained the additional history that the patient had
not urinated in eight hours. The physician ordered cardiac enzymes. His
CPK was 50,000 U/L and his troponin I was zero. The patient also stated
that he had recently increased his exercise regimen. He was admitted to
the hospital for rhabdomyolysis and intravenous hydration. The cause
continued
 Chest pain 63

was believed to be a combination of stimulant use and strenuous

exercise. His hospital course was uncomplicated.
• Although bloodwork is usually not required in a young patient with
chest pain, a CPK level is helpful if rhabdomyolysis is suspected. An
elevated level may give the clinician a suspicion of the diagnosis until
further confirmatory tests (e.g. urine myoglobin) are completed. In
addition, it may serve as a marker of disease progression or improve-
ment during therapy.

Case 2.27 A result no-one expected

• This is an interesting case that is also very personal to us. A 60-year-old
man was referred to a cardiologist after having an abnormal screening
exercise stress test. The patient had developed some small ST segment
depression during the test. He had no risk factors for coronary artery
disease except a history of tobacco use in the past. The cardiologist decided
to perform a cardiac catheterization.
• During the catheterization, the cardiologist found 90% blockages in the
right coronary artery and the circumflex artery. He opened both block-
ages with angioplasty and placed stents in both vessels. The patient was
placed on Plavix. He took the medicine as instructed and had no
problems until a year and a half later.
• He then developed chest pain with minimal exertion. His primary
physician and his cardiologist were both concerned a stent restenosis
despite the elapsed time and his continued use of Plavix. Since he was
now at a city four hours from the city of his cardiologist, he was referred
to a different cardiologist for a heart catheterization. During the
catheterization, his stents were found to be widely patent. There was
severe stenosis, however, just proximal to both stents. These were
believed to have resulted from the iatrogenic vessel injury during the
angioplasty. The stents placed did not cover the entire length of the
angioplasty. The cardiologist opened these areas with angioplasty and
placed longer drug-eluted stents in the vessels. He has done well since.
This patient was our father.

Case 2.28 All in the family

• During a shift in the ED, we received a call from the paramedics
concerning a patient that they were bringing in for chest pain. The report
was a 34-year-old male with chest pain and a heart rate of 160 beats/
minute. Our immediate thought was a probable case of supraventricular
tachycardia and unlikely coronary artery disease based on his age.
• Upon arrival to the ED, the patient was profusely diaphoretic and had
an oxygen saturation of 82% on a non-rebreather mask. He was in severe
clinical heart failure and having severe chest pain. His EKG showed
continued
64 Learning from medical errors: clinical problems

ventricular bigeminy at a rate of 140 beats/minute and 4 mm of ST

segment elevation in leads V1 to V4. The cardiologist was called
immediately and medical therapies for acute myocardial infarction and
heart failure were started.
• The patient was able to give us some history. He was a Christian Scientist
and did not believe in doctors. He had not been to a doctor for many
years. He denied tobacco or illicit drug use. His only apparent coronary
artery disease risk factor was a family history. The chest pain had started
12 hours prior to his arrival. He did not believe that it could be a heart
attack and tried to avoid the hospital. The pain then became unbearable
and he asked us to do whatever was necessary. When the cardiologist
arrived, we were ready to move him to the catheterization laboratory. He
went into cardiac and respiratory arrest in the ED. There was copious
frothy sputum seen with oral intubation. Resuscitative efforts were
unsuccessful.
• This patient was unfortunate in having a family history of coronary artery
disease. Most males of this age would not be expected to have coronary
artery disease. His presentation was certainly the most dramatic of acute
myocardial infarction that we have seen in someone less than 40 years
of age. In thinking about this case, we wonder what our disposition of
this patient would have been had his EKG and cardiac enzymes been
normal and he had been in no hemodynamic distress.

References
1 (2003) ACEP, Reference+Resource Guide. American College of Emergency Physicians,
Dallas, Texas, 19.
2 Tintinalli JE, Kelen GD and Stapczynski JS Emergency Medicine: a comprehensive study
guide (5e). McGraw-Hill, New York, New York, 341.
3 Lapp T (2003) Chest pain a high-stakes diagnostic challenge. FPReport. November: 12.
4 Tintinalli JE, Kelen GD and Stapczynski JS Emergency Medicine: a comprehensive study
guide (5e). McGraw-Hill, New York, New York, 343.
5 Tintinalli JE, Kelen GD and Stapczynski JS Emergency Medicine: a comprehensive study
guide (5e). McGraw-Hill, New York, New York, 342.
6 Tintinalli JE, Kelen GD and Stapczynski JS Emergency Medicine: a comprehensive study
guide (5e). McGraw-Hill, New York, New York, 341.
7 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 628.
8 Diercks D (2004) Symptoms, Markers, and ECGs. Cardiovascular and Neurovascular
Emergencies: implications for clinical practice. EMCREG International, Cincinnati, Ohio,
7.
9 Tintinalli JE, Kelen GD and Stapczynski JS Emergency Medicine: a comprehensive study
guide (5e). McGraw-Hill, New York, New York, 347.
10 Cline DM, Ma OJ, Tintinalli JE et al. (2001) Just the Facts in Emergency Medicine. McGraw-
Hill, New York, New York, 310.
11 Tintinalli JE, Kelen GD and Stapczynski JS Emergency Medicine: a comprehensive study
guide (5e). McGraw-Hill, New York, New York, 391.
12 Marx JA, Hockberger RS and Walls RM (2002) Rosen’s Emergency Medicine: concepts and
clinical practice (5e), Vol 2. Mosby, St. Louis, Missouri, 1015.
 Chest pain 65

13 Holroyd BR (2004) Pathophysiology of Acute Coronary Syndromes. Cardiovascular

and Neurovascular Emergencies: implications for clinical practice. EMCREG International,
Cincinnati, Ohio, 4.
14 (2004) Reperfusion Strategies for ST-Segment Elevation Myocardial Infarction: An
Overview of Current Therapeutic Options, Part I: Pharmacologic Reperfusion.
Emergency Medicine Reports. 25(7): 82.
15 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 69–70.
16 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 37–8.
17 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 55.
18 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 56.
19 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 59.
20 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 60.
21 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 61.
22 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 63–4.
23 Tierney LM, McPhee SJ and Papadakis MA (2004) Current Medical Diagnosis &
Treatment (43e). Lange Medical Books/McGraw-Hill, New York, New York, 663.
24 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 145.
25 Wilderson D and Wagner MJ (2003) Noncardiac Chest Pain. Foresight: Risk Manage-
ment for Emergency Physicians. 56(February): 2–3.
26 Wilderson D and Wagner MJ (2003) Noncardiac Chest Pain. Foresight: Risk Manage-
ment for Emergency Physicians. 56(February): 2.
27 Wilderson D and Wagner MJ (2003) Noncardiac Chest Pain. Foresight: Risk Manage-
ment for Emergency Physicians. 56(February): 4–6.
28 Wilderson D and Wagner MJ (2003) Noncardiac Chest Pain. Foresight: Risk Manage-
ment for Emergency Physicians. 56(February): 6.
29 Joseph AJ (2002) Acute Coronary Syndromes. Foresight: Risk Management for Emergency
Physicians. 54(June): 2–3.
30 Joseph AJ (2002) Acute Coronary Syndromes. Foresight: Risk Management for Emergency
Physicians. 54(June): 3–5.
31 Joseph AJ (2002) Acute Coronary Syndromes. Foresight: Risk Management for Emergency
Physicians. 54(June): 4.
32 Joseph AJ (2002) Acute Coronary Syndromes. Foresight: Risk Management for Emergency
Physicians. 54(June): 5–7.
33 Rice B (2004) Protect yourself when you refer. Medical Economics. 81(9): 17–19.
34 Rice B (2004) Protect yourself when you refer. Medical Economics. 81(9): 17.
This page intentionally left blank
Chapter 3

Fever

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Fever is probably the most alarming vital sign to the general public. It certainly
draws more physician visits than any other abnormal vital sign. Temperature is
the one vital sign that most people know how to take. Therefore, people are more
aware of the presence of fever than any other abnormal vital sign. Most authors
regard an oral temperature of greater than 100.4ºF as a fever. The majority of
fevers arise from infectious etiologies. Some noninfectious causes include:
allergic reactions, connective tissue diseases, drug reactions, heat illnesses, thy-
roid disorders, autoimmune disorders, and malignancies.1
Fever is the most common chief complaint among ambulatory children.2
Children are more likely to have an infectious cause to their fevers. In addition,
children are more susceptible to complications from fever such as seizure,
irritability, anorexia, and vomiting. ‘Failure to recognize bacterial illness is
perhaps the most costly of all misdiagnoses in pediatric care.’3
A special group of fevers is classified as ‘fever of unknown origin’ (FUO).
Although emergency physicians often use the term FUO as their diagnosis after
the work-up of potential sources in the emergency department is negative, the
diagnosis of FUO in the emergency department is used quite often, incorrectly.
The technical definition of FUO consists of documented fevers of greater than
101ºF for more than three weeks and no established diagnosis despite
appropriate investigation for one week.4

Triage
A temperature should be recorded in triage and a fever should be immediately
treated with antipyretics. Treatment is especially important in young children
because of the risk of febrile seizures but treating fevers in patients of all ages
enables the patient to be more comfortable and amenable to a good examination
by the physician. The only exception may be patients who have liver disease and
cannot take acetaminophen (an alternative is ibuprofen) or patients who have
adverse drug reactions to anti-inflammatory medicines (an alternative is
68 Learning from medical errors: clinical problems

acetaminophen). Smaller children should have temperatures recorded by rectal

thermometers because their lack of compliance will give inaccurate oral
temperatures. An oral temperature usually suffices for adults.
All febrile patients who have other abnormal vital signs should be brought to
the physician’s attention immediately. Fever in combination with another
abnormal vital sign increases the likelihood of a serious infection or a state of
hemodynamic compromise. In addition, patients who have an ill appearance
should see the physician emergently. This includes any patient that is lethargic,
in respiratory distress, is orthostatic, is vomiting, etc. Other patients who appear
more comfortable can be seen on a less urgent basis. The exception to this rule
would be the patient that has a potentially serious communicable infection (e.g.
meningitis, tuberculosis). These patients need to be brought into an isolated area
immediately.
Some triage systems allow the triage nurse to initiate specific tests whenever
there appears to be an obvious focus of infection. Examples include the patient
with urinary symptoms having a urinalysis ordered or a patient with
respiratory symptoms having a chest X-ray ordered. This practice expedites the
diagnostic process in a busy practice and enables the patient to be treated in a
quicker fashion. The disadvantage of this practice is the occasional unnecessary
order of a test and its subsequent cost and side effects (e.g. radiation from a chest
X-ray).

History
Most patients with fevers will have specific symptoms that can be elicited
during the history. Upper respiratory symptoms will require an assessment of
the ability to maintain an airway and the potential of developing airway
compromise. We often take upper respiratory infections for granted because of
the rarity of serious complications. We have seen, from our practice, meningitis
from untreated sinusitis (please see Case 3.5), orbital cellulitis from poorly treated
streptococcal A pharyngitis (please see Case 3.6), and peritonsillar abscesses
from untreated streptococcal A pharyngitis.
Patients with meningitis may present with extremely variable presentations.
Although the classic symptoms of meningitis are fever, headache, stiff neck,
photophobia, and altered mental status, these symptoms are present in only 25%
of adult cases.5 Nevertheless, patients presenting with fever should be asked
about these symptoms. Spach comments that (and we concur from our own
clinical experiences), ‘the absence of all three features of fever, neck stiffness, and
altered mental status virtually eliminated the diagnosis of acute bacterial
meningitis.’6 These results were from a study published in the Journal of the
American Medical Association (JAMA).7 Viral meningitis, however, is usually less
fulminant and patients will often not appear toxic. A heightened suspicion must
be maintained. Suspected meningitis warrants questioning about living
conditions such as when people live in close quarters (e.g. army barracks,
college dormitories) that are prone to meningitis cases due to N meningitidis.8
Immunization history of H influenzae type b or N meningitidis should be
obtained.
Some fevers are associated with occult infections and may be missed if clinical
suspicion is not maintained at a high level. Endocarditis, osteomyelitis, epidural
 Fever 69

abscesses, and necrotizing fasciitis are a few examples. Specifically directed

questioning is needed to detect these occult infections. Is the patient an intra-
venous illicit drug user? Does the patient have a history of heart valvular
problems? If so, have they had recent surgery? Does the patient have sickle cell
disease? We feel that all patients with potential infections should be asked about
a personal and a family history of diabetes.
The patient’s current medications and allergies are important for several
reasons. If the patient is on steroids, the physician can rely less on an elevated
white blood cell count as a marker for infection. Certain drugs may cause
leukopenia including: cyclophosphamide, methotrexate, penicillins,
sulfonamides, carbamazepine, and clozapine.9 Ascertaining allergies is always
paramount because of the numerous allergic reactions and intolerance to anti-
biotics.
Social habits should also be included in the history. There is a substantial
increased incidence of upper and lower respiratory infections with tobacco use.
Intravenous illicit drug use can contribute to cellulitis, endocarditis, and
epidural abscesses. Promiscuous sexual activity can lead to sexually
transmitted diseases. Finally, recent travelers may become stricken with acute
gastroenteritis or other diseases that are prevalent in the areas of visit.
For females, the physician should add a gynecologic history. Pelvic inflamma-
tory disease has an inclination to occur around the menses. Does the patient
have a retained tampon or intrauterine device as a possible source of the fever?
Vaginal bleeding and fever in pregnancy could represent a septic abortion.
The final component of the history is the surgical history. There are several
causes of post-surgical fever. These include infectious causes such as surgical
abscesses, retained foreign bodies, and endocarditis. Fevers may also be from
noninfectious causes such as atelectasis or deep venous thrombosis.

Physical exam
Fever is a systemic response of the human body. Therefore, it is often accompan-
ied by other abnormalities in vital signs. Furthermore, abnormalities in the other
vital signs may provide clues to the source of the fever or the amount of
hemodynamic compromise of the individual. The classic example would be in
the setting of sepsis. Here, a high fever is likely to be accompanied by
tachycardia, tachypnea, and hypotension. As a consequence, all five vital signs
should be recorded in the patient who presents with a fever. We commented
earlier that oral temperatures are usually sufficient for adults. However, this is
not always true due to the poor peripheral circulation of some adult patients.
Edwards, in The M & M Files, presents a few cases where an oral temperature
was used and later found to be incorrect (please see Case 9.6). We have also
experienced a few similar cases. If the patient feels warm to the touch, ask the
nurses to get a rectal temperature. You will be surprised how many times you
will be glad that you did.
Patients who are hemodynamically stable and have localized symptoms may
be candidates for a focused examination. Uncomplicated upper respiratory
infections (otitis media and sinusitis) and cellulitis are examples of such. How-
ever, whenever these ‘simple’ infections are accompanied by other abnormal
vital signs (besides fever) or when they occur in immunocompromised patients
70 Learning from medical errors: clinical problems

(i.e. diabetics, elderly, etc.), a more extensive and thorough exam is warranted
due to the increased risk of complications.
For all other patients, a general head to toe examination should be performed
to search for the etiology of the fever. The general appearance should always be
noted first. Does the patient appear ‘toxic’ or ‘septic’? These patients usually
appear extremely fatigued, listless, and pale. They have difficulty giving you
their full attention. Is the patient moving his or her neck freely or experiencing
any respiratory difficulties? Does the patient appear to have localized pain that
may suggest a source?
The head is probably the source of the majority of fevers. Most are going to
be routine upper respiratory tract infections such as otitis media, sinusitis, viral
‘colds’, pharyngitis, tonsillitis, and conjunctivitis. A thorough exam of the eyes,
ears, and throat is required when there is suspicion of complicated infections
such as orbital cellulitis, malignant otitis externa, and peritonsillar abscess. Finally,
any patient with a headache and a fever should be checked for meningeal signs.
The neck is the source of fever occasionally. More often, the neck contributes
to the examination of fever in the presence of adenopathy. Enlarged cervical
nodes suggest to the physician the presence of an upper respiratory infection
and/or mononucleosis. Infections in the neck, however, do occur and are more
common in children than in adults. These include epiglottitis, croup, Ludwig’s
angina, retropharyngeal abscess, and neck abscess. With the exception of
Ludwig’s angina, where there may be elevation of the tongue, the rest of these
diagnoses cannot be visualized directly by the clinical exam. We must rely on
the presence of neck swelling, stridor, tracheal deviation, and patient position-
ing. With regard to the last finding, the patient with epiglottitis prefers to lean
forward while the patient with a retropharyngeal abscess may have his or her
neck turned to the side of the abscess (due to unilateral spasms of the sterno-
cleidomastoid muscle). The neck should also be examined for meningeal signs.
The cardiopulmonary examination should be routine in the patient with fever.
It is obviously important for the patient with symptoms localized to the thoracic
area. The thoracic area is also a final destination for many upper respiratory
infections so it is not unusual to have concurrent infections of the upper and
lower respiratory tracts. In addition, it is important even for patients with
extrathoracic symptoms because it gives the physician a quick assessment of the
hemodynamic status of the patient.
The abdomen is filled with organs that have the potential to become infected
and cause fevers. Fortunately, the dictum is that with the exception of a urinary
tract infection, abdominal organs are very unlikely to be infected in the absence
of pain in the abdomen. Having said this, it is imperative that any patient with
abdominal pain and fever receives a thorough examination. This includes a rectal
examination to look for blood in the stool for all patients and a pelvic
examination in all females with lower abdominal pain.
Examination of other parts of the body should be carried out as dictated by
the clinical symptoms. The patient with urinary tract symptoms should have the
flanks palpated for tenderness. Although we do not see it performed frequently,
consider a prostate exam for males with a urinary tract infection because of
the difference in the treatment regimen for prostatitis. Skin examinations
are important for those with cellulitis, superficial abscesses, varicella,
thrombophlebitis, etc. In addition, fever and a rash should draw consideration
 Fever 71

for sepsis or bacteremia. Musculoskeletal examinations are necessary for

those with possible septic arthritis, tenosynovitis, bursitis, and deep venous
thrombosis.
We will now discuss the physical examinations for specific cases of fevers.
Physical exam findings in meningitis, like its history, are also extremely variable.
The ‘classic bedside physical examination findings of meningitis (i.e. Kernig’s
sign, Brudzinski’s sign, and nuchal rigidity) all have low diagnostic values.’10
From our experience with many cases of meningitis through the years, we would
agree with this statement. Most patients will complain of neck pain (although
a couple did not), however, it is rare that we see actual neck stiffness. The majority
of those with neck stiffness turned out to have bacterial meningitis and only in
a few instances was it a viral infection. We have also noticed that patients with
viral meningitis may not look toxic at all and may appear to have cold symptoms
with a headache. Therefore, it is not surprising that viral meningitis is often a
‘second visit’ diagnosis.
One of the toughest diagnoses to make clinically is osteomyelitis. Whenever
there is bone exposed through an infected wound, the infection is osteomyelitis
until proven otherwise. Although the diagnosis is easy in this case, we have seen
many physicians fail to make it because of the infrequent occurrence of osteo-
myelitis.
The final comment worth mentioning is that infections sometimes like to hide
in orifices. We have already talked about oral infections. Likewise, infections are
also common at the other end of the body. Fever and rectal complaints mandates
a rectal examination for perirectal abscess or prostatitis. We have seen cases where
perirectal abscesses were not clearly evident on external exam. Please see Case
3.7 below. Similarly, the genitourinary tract system is also a common harbor for
infections. Epididymitis and urethritis are the two most common sexually
transmitted diseases in males. Sexually transmitted diseases (STDs) in females
may present as vaginitis, cervicitis, pelvic inflammatory disease, and tubo-
ovarian abscess. In addition, non-STD causes of fevers in the female genitourinary
system include infected Bartholin cysts, retained tampons, and septic
thrombophlebitis from intrauterine devices. Finally, remember that the uterus
and the rectum are potential sites for the introduction of foreign bodies. Please
see Case 3.8 below.

Differential diagnosis
As discussed in the introduction, most fevers are caused by infections. Most
infections are due to viruses and tend to have benign and self-limiting courses.
There are two exceptions to this rule, however, where viral infections can lead
to high morbidity. This is the dehydration associated with pediatric viral
gastroenteritis and the general ill appearance of patients with viral meningitis.
Please see Case 3.4. Be alert for viral infections that may be occult such as diskitis,
myocarditis, and endocarditis.
Although less common, bacterial infections are associated with more
complications than viral infections. Tonsillitis, pyelonephritis, meningitis, chole-
cystitis, appendicitis, and diverticulitis are bacterial infections with frequent
complications. Pneumonia, urinary tract infection, and cellulitis are bacterial
infections that are usually easy to treat but have a greater tendency to produce
72 Learning from medical errors: clinical problems

complications in those who are immunosuppressed. Be aware of bacterial infec-

tions that may be occult such as prostatitis, necrotizing fasciitis, epiglottitis in
adults, septic arthritis, endocarditis and abscesses.
Fevers may also be due to noninfectious causes. Teething in children or
dehydration are two common benign examples. Some medications that may
produce fevers include: stimulants, antihistamines, antipsychotics, seizure medi-
cations, and a few antibiotics.11 Atelectasis, allergic reactions, malignancies and
endocrine disorders may also cause fevers. Finally, fever of unknown origin
may not have an underlying infectious process.

Diagnostic tests
Patients with suspected meningitis should have blood cultures drawn
immediately before the initiation of antimicrobial therapy. This is done for two
reasons. Blood cultures will be positive in up to 50% of cases of bacterial
meningitis and the epidemiology is of great importance in treating and report-
ing bacterial meningitis. Cultures may become obscured once antimicrobial
therapy is given. Other diagnostic tests, such as CT scan of the head and lumbar
puncture, may be performed after antibiotics have been given. Antibiotic treat-
ment of meningitis is often done empirically and should be adjusted to the age
range of the patient and the clinical environment of the patient. In our training,
we were taught to obtain a head CT in every adult to check for increased
intracranial pressure before performing a lumbar puncture. We have adjusted
our practice, however, because of the occasional uncontrollable delays (i.e. CT
scanner not working, multiple patients waiting for CT scans, etc.) with getting
a CT report. Increased intracranial pressure is extremely unlikely in a patient
with a normal neurologic examination. Hence, for these patients, we are more
afraid of having them ‘crash’ from a bad infection than the occurrence of brain
herniation. In addition, the majority of the neurology and emergency medicine
literature suggests that a lumbar puncture can be performed safely in patients
with a normal mental status and no focal neurologic signs or symptoms.12
In the outpatient setting, most patients with suspected pneumonia will simply
require a chest X-ray and a pulse oximetry. Remember that chest X-rays are
sometimes not helpful in patients who are dehydrated or patients who have
Pneumocystis carinii pneumonia. In addition, a significant amount of
pneumonias are not seen or are difficult to see on an anterior/posterior
projection film and a lateral film should always be obtained if possible. If the
chest X-ray is normal and the patient has respiratory symptoms, pulmonary
embolus should be considered as the possible cause of the fever. Please see
Case 3.11. Appropriate testing for pulmonary embolus is discussed in Chapter
9. The final comment about fever and infections in the chest is the insensitivity
of the chest X-ray for pericarditis. If suspected, an echocardiogram should be
ordered.
In patients with fever and other vital signs that are abnormal (e.g. hypo-
tension, tachycardia, tachypnea), a toxic appearance, or the clinical suspicion of
sepsis and/or bacteremia, blood should be collected as soon as possible. These
patients will require a complete blood count, blood cultures, and appropriate
tests directed at the source of the fever. The timing of blood collection is
important in order for antibiotics to be started as soon as possible.
 Fever 73

Testing for patients with fever and urinary tract symptoms is dictated by the
clinical presentation. For the patient with normal vital signs and a benign physical
examination, a urinalysis and/or pregnancy test are probably all that are needed.
If they have a history of complicated urinary tract infections (UTIs) or pregnancy,
a urine culture should be ordered. Nursing home patients and pediatric patients
also require a urine culture. In contrast, patients with abnormal vital signs,
significant abdominal or flank tenderness, or significant vomiting should
probably have a complete blood count and a serum electrolyte ordered along
with a urine culture.
Most infections of the head are diagnosed clinically and do not require any
testing. Some, however, are clinically occult and are best detected by CT scan.
These include: orbital cellulitis, facial abscesses, Ludwig’s angina, mastoiditis,
and retropharyngeal abscesses.
Various modalities may be used for confirming infections in the neck. Epiglot-
titis is suggested on a soft tissue neck X-ray by a ‘thumbprint’ sign. The diagnosis,
however, is made by direct visualization in the operating room. We have seen
neck abscesses commonly in patients with head or neck cancer and necrotic
lymph nodes. These are usually diagnosed and delineated by CT scan. Keep in
mind that a CT scan is only appropriate with patients who are unlikely to have
airway involvement (to avoid the loss of the airway while outside of the
physician’s supervision).
Fever and infections in the abdomen are extremely common. They may occur
in any age group and without any predisposing factors. They also represent a
large percentage of lawsuits involving patients with fever. Their management
is discussed in detail in Chapter 1.
Infections in the groin, perineum, and rectum are not common but can lead
to great morbidity if not detected and properly treated. Epididymitis, urethritis,
orchitis, and scrotal abscesses are all genitourinary infections that cause fever.
Color flow ultrasound is usually the test of choice because it is not only helpful
to determine the area involved but it also can rule out testicular torsion, which
sometimes cannot be differentiated clinically. Other tests that should be ordered
are a genital swab for gonorrhea and chlamydia and a urinalysis. Perianal
infections include perirectal abscesses and infected pilonidal cysts. These are
almost always diagnosed clinically. However, a CT scan can be used for
perirectal abscesses if the diagnosis is in doubt. Finally, the two most severe
infections in this area are necrotizing fasciitis and Fournier’s gangrene. They
should be suspected when the patient’s pain is out of proportion to the clinical
findings and the patient is immunocompromised. This is particularly true in
patients with diabetes. A CT scan may be helpful if these conditions are
suspected but specialty consultation should be obtained immediately.
The rest of this section is devoted to the diagnostic evaluation of fever in the
pediatric patient.13 The group that poses the most problems is the newborn to
three-month group. In addition to being more prone to serious bacterial
infections because of their underdeveloped immune system, these infants lack
the social skills to make the physical exam reliable. ‘Infants who appear
generally well; who have been previously healthy; who have no evidence of
skin, soft tissue, bone, joint, or ear infection; and who have a total white blood
cell count from 5,000–15,000 cells/microliter, an absolute band count of less than
1,500 cells/microliter, and normal urinalysis results are unlikely to have a
74 Learning from medical errors: clinical problems

serious bacterial infection.’14 Therefore, for infants meeting these guidelines, a

complete blood count with differential and a urinalysis is recommended.
Although the collection of urine in these patients is recommended by catheteri-
zation, a specimen collected by an adhesive, sealed, sterile collection bag after
disinfection of the skin may also be helpful. However, in girls and
uncircumcised boys, the contamination rate is extremely high with this
technique. Furthermore, if the urine culture does not grow a single organism
with a colony count of greater than 100,000, confirmation of a true infection is
needed with a catheter specimen. In the emergency department, for all practical
purposes, we do not recommend a bag specimen because we may not have follow-
up with the patient.15 When the infant appears toxic, the total white blood cell
count is greater than 15,000 cells/microliter, the absolute band count is greater
than or equal to 1,500 cells/microliter, or there is evidence of a urinary tract
infection, cultures of the blood, urine, and cerebrospinal fluid should be obtained.
For children in the three months to three years age group, diagnostic testing
is also based on the toxicity of the child’s appearance.14 Children who appear
toxic require a complete blood count with differential, a urinalysis, and cultures
of blood, urine, and cerebrospinal fluid. Children who have a temperature of
less than 39 C and do not appear toxic, however, do not require any diagnostic
tests.16 If the non-toxic appearing child has a temperature greater than 39 C, the
physician has two treatment options. The first is to administer empiric
antibiotics after drawing a blood culture. The second is to draw a white blood
cell count and do the first only if the count is greater than or equal to 15,000 cells/
microliter.16 A third alternative that is used in clinical practice is to draw the
blood culture but withhold the antibiotics.14 The child will return for further
evaluation if the culture is positive. If the child is well and the exam is normal,
a second blood culture is taken and the child is discharged. In the event that the
child still has fevers or appears ill without a specific focus, or if H influenzae or
N meningitidis is found on the first culture, the child will need hospitalization,
antibiotics, and a complete septic work-up with cultures of blood and cerebro-
spinal fluids.14

Hospital/office course
Because fever can be a manifestation of a serious infectious process, the timing
of evaluation and intervention is extremely important. The accurate recording
and timing of events in the medical evaluation should be documented in the
chart. Serious infections such as sepsis, pneumonia, and meningitis can produce
devastating outcomes if treatment is delayed by as little as one hour (one of the
quality and assurance checks at our hospital is the rapidity that antibiotics are
given once the diagnosis of pneumonia is made). If there is a substantial delay
in obtaining blood studies that cannot be overcome (e.g. difficult blood draw,
parent indecision), then antibiotics should not be delayed. Remember that treat-
ment is generally more important than diagnosis. Without accurate and precise
documentation, an infection with a bad outcome will be difficult to defend in
court, even if everything was performed correctly.
In the case of suspected meningitis, the timing of interventions and
medications becomes paramount. Please see Case 3.5 of the 60-year-old
attorney in the ‘Errors and interesting cases’ section below. Administering
 Fever 75

antibiotics for suspected bacterial meningitis immediately after drawing blood

cultures and not waiting for CT scan or lumbar puncture results is the
recommended standard of care because of the potential of rapid deterioration
in these patients.
In addition, the traditional therapeutic approach towards using steroids in
bacterial meningitis was that it was recommended for children with H influenzae
meningitis. There were studies showing a reduced incidence of hearing impair-
ment in children with H influenzae meningitis when dexamethasone was
administered before antibiotics.17 There are now studies suggesting that steroids
may also be beneficial for adults with bacterial meningitis. A European trial
showed that dexamethasone ‘reduced the risk of death and unfavorable neuro-
logic outcomes.’17 ‘Benefits were most marked in patients with S pneumoniae
meningitis.’ ‘This study has resulted in the first general recommendation for
using dexamethasone before or at the time of first antibiotic administration for
S pneumoniae meningitis in patients who are not in septic shock.’17
The final comment concerning meningitis is to make sure that your staff and
others that have been exposed to the patient receive appropriate prophylaxis
medication. Chemoprophylaxis with ciprofloxacin or rifampin is available for
prevention of secondary cases of N meningitidis and H influenzae. The physician or
another member of the medical staff should be certain to contact everyone that
has been exposed to the patient and also ascertain that the proper health authori-
ties are notified of the case.
The most challenging question concerning the treatment of patients with
pneumonia is whether to admit them. Pneumonia is different from most other
infections because of its wide range of virulence. We have seen patients that were
not diagnosed with pneumonia clinically. Their chest X-rays, however, were
read by radiologists as positive for pneumonia. We found out later that these
patients had spontaneous improvement without medications. On the other hand,
we have seen patients decompensate and die within an hour of being medically
stable. Therefore, an experienced clinician must consider all the intangibles of
the patient (e.g. age, present medical conditions, home situation, ability to
obtain medications, tobacco use) before making a patient disposition.
Discussion of disposition with the patient’s primary is often helpful because she
will usually have a good feel on whether the patient will do well with outpatient
therapy.
Patients with suspected sepsis should have antibiotics and intravenous fluids
(if they are not fluid overloaded) started as soon as possible. The degree of illness
with most of these patients will be so obvious that a disposition decision is rarely
problematic. Nevertheless, the poor outcome associated with sepsis has led to
a number of lawsuits of medical negligence. Therefore, minimize the time needed
to obtain diagnostic studies and initiate medications and fluids. On some
occasions, however, sepsis or bacteremia can be extremely early and occult. Please
see Case 3.10. This presents a problem because blood culture results will not be
available for at least one day. For these situations, we recommend two clinical
pearls to decrease your medical liability. If you suspect that a patient may have
sepsis, order blood cultures and observe that patient for at least four hours. It
will also be of great benefit if the laboratory can do a gram stain of the blood
cultures. Make sure that the patient’s vitals are recorded frequently and remain
stable. The second recommendation is to ensure that blood culture results are
76 Learning from medical errors: clinical problems

reviewed as soon as possible. This can be done with laboratory review by the
ED physician daily or by asking the laboratory to call the ED physician with any
positive blood cultures.
With the exception of the three previous infections (meningitis, pneumonia,
and sepsis) and surgical causes of fevers (e.g. appendicitis, cholecystitis, etc.),
the majority of other causes of fevers are treated on an outpatient basis. Patients
should be given instructions on fever control, signs and symptoms to seek, when
to return, and with whom to follow up. One of the most common scenarios
where physicians get faulted is when an infection worsens because the patient
could not afford the antibiotics that were prescribed. Therefore, be extremely
sensitive to this and try to write for something that they will be able to get.
Finally, we leave one last caution with the disposition of patients with fevers
and/or infections. Always be conservative about treating patients with diabetes
or peripheral vascular disease because this is the main population where we
have seen treatment failures. Helpful techniques include using a broader
spectrum antibiotic, closer follow-up, and hospitalization for questionable or
borderline cases.
The disposition of infants and young children with fever is not standardized
among clinicians. It is generally agreed that toxic-appearing patients be
admitted for a complete septic evaluation and administration of antibiotics. The
disparity, however, arises in well-appearing patients with abnormal laboratory
tests. In addition, there is no consensus on the minimum age when all febrile
children should be admitted.
For children who are less than three months of age, there are still some
physicians who uniformly hospitalize those with fevers. More commonly, most
clinicians will use one month as the cut-off and treat children between one and
three months of age on an individual case basis. Empiric antibiotics are
generally recommended and patients who are candidates for outpatient therapy
must have follow-up arranged and reliable parents. Management of children
between three months of age and three years of age is discussed above in the
‘Diagnostic tests’ section.

Pitfalls
Be cautious with patients presenting with low-grade fevers (i.e. greater than
99ºF but less than 100.4ºF). Low-grade fevers in combination with certain clinical
scenarios should always raise suspicion for serious pathologies. These findings
include abdominal pain, immunocompromised states, and when there is a
possibility of occult infections (e.g. pelvic abscess, epidural abscess, foreign
bodies). When infections with low-grade fevers lead to unfavorable
outcomes, plaintiff attorneys may use this gray zone of fever as evidence for
their case.
Patients who state that they have fevers at home (especially if they use a
thermometer) should be regarded as being febrile even if their recorded
temperature at the visit is in the normal range. The use of antipyretics prior to
the visit may mask the actual temperature. In addition, the patient may have a
condition manifested by intermittent fevers.
Patients with persistent fevers while on antibiotics should be considered to
have a serious bacterial infection (e.g. resistant organism) until proven otherwise.
 Fever 77

Although a viral illness may be present in these cases, you do not want to miss
a serious bacterial infection and the general rule is that viral illnesses get better
with time. We present a case in the ‘Errors and interesting cases’ section below
that highlights this dictum (Case 3.1).
Some infections can cause rapid deterioration and warrant prompt diagnostic
efforts and treatment (e.g. pneumonia, meningitis). At times, empiric antibiotic
therapy is warranted while diagnostic efforts are underway. Time of therapy
should be documented in the chart as well as an explanation for any delays (e.g.
antibiotics had to be obtained from pharmacy). We present a case of delayed
treatment and its subsequent adverse outcome in the following section (Case
3.2).
The combination of fever and rash should be considered as bacteremia and/
or sepsis until proven otherwise. Serious bacterial infections commonly
associated with rashes include meningococcemia and gonococcal sepsis. Less
common infections with rash are Lyme disease and parasitic infections. The
unfortunate death of an infant who presented with a fever and a rash is pre-
sented in the following section (Case 3.3).
In patients presenting with fever without a specific source on physical exam
or diagnostic testing, meningitis should always be considered. We talked about
the classic symptoms of meningitis above. We can evaluate for pneumonia and
urinary tract infection relatively easily in the emergency department with a chest
X-ray and a urinalysis. However, use caution at making these diagnoses when
the tests are positive but the clinical scenario does not fit. Please see the case of
the 19-year-old female with fevers and vomiting below (Case 3.4).
Neurologic symptoms presenting with fever is meningitis until proven
otherwise. Although other etiologies such as encephalitis or spinal abscess may
be present, these patients will usually require a lumbar puncture to rule out
meningitis. Please see Case 3.5 below.
Fever and sore throat may be problematic in two ways. The source may be
obvious as in the case of swollen tonsils with white exudates or in the case of
a huge peritonsillar abscess. The inflammation and swelling, however, may
obscure further visualization of the posterior pharynx where a retropharyngeal
abscess may be hiding (although this occurrence is rare except in very young
children). On the other hand, when the pharynx looks relatively normal in the
presence of a fever and a sore throat, the diagnosis becomes more difficult. Make
sure to examine the ears because they share some of the same pain receptors as
the throat. In addition, consider etiologies that are beyond your clinical
visualization of the pharynx. Namely, consider the possibility of epiglottitis or
esophageal/tracheal foreign body.
Physicians commonly rely on a white blood cell count as an indicator of the
severity of an infection. Although the white blood cell count is often helpful if
it is elevated, a normal white blood cell count has very little specificity for ruling
out a serious infection. We have found that the differential on the white blood
cell count has increased but still limited specificity for serious infections. Please
see Case 3.9 below. In a recent study published in the Annals of Emergency Medicine,
normal white blood cell counts were found in one third of children with
bacteremia.18
Fever with accompanying nausea and vomiting in children deserve special
attention. Acute gastroenteritis (AGE) is one of the top causes of mortality in
78 Learning from medical errors: clinical problems

children of third world countries. Although deaths from AGE are not common
in the United States, they do occur, and hospitalizations for AGE are frequent.
We present two such examples in Case 3.12 and Case 3.13 below.

Errors and interesting cases

Case 3.1 Sometimes they fail
• Selbst and Korin, in Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine, present a case of an eight-month-old girl who was seen at a
clinic for a fever and was treated for an otitis media. The child did not
get better and returned to see the same physician and was prescribed
ampicillin without having any tests ordered. The child was brought back
to the clinic twelve days later after showing no improvement. A different
physician treated the patient on this visit and ordered a complete blood
count and other tests but decided to cancel these tests after examining
the patient. The patient failed to improve and was finally diagnosed with
meningitis at a hospital two days later and died shortly after that.19 The
family was awarded a $26 million settlement at the fault of the second
doctor. The first doctor was acquitted.
• This case illustrates the importance of increased vigilance for any patient
with a fever who does not improve on antibiotics. This is especially true
in today’s medical environment with the increased presence of drug-
resistant bacteria and indiscriminate use of antibiotics for viral infections.
The first physician did not have the benefit of this knowledge and was
subsequently not found negligent. Returning patients need the ‘step up’
in care that we discussed previously.

Case 3.2 No time to waste

• Selbst and Korin offer another case, in Preventing Malpractice Lawsuits in
Pediatric Emergency Medicine, that describes the consequences of not
responding in a timely fashion.20 This case involved an eight-week-old
girl that presented to the emergency department with a fever and had
a chest X-ray that showed a probable pneumonia. Although the patient
was admitted and intravenous antibiotics orders were written, the
medicine was not given until the pediatrician arrived. A lumbar puncture
by the pediatrician showed meningitis and she suffered severe brain
damage. The hospital and the emergency physician were found at fault
for $140,000.
• Many emergency departments have developed protocols for the timely
initiation of antibiotics for serious infections. These infections include
pneumonia, sepsis, and meningitis. In general, antibiotics for these
infections should be started as soon as there is good evidence of their
presence (e.g. infiltrate on chest X-ray for pneumonia, fever and stiff
neck for meningitis). Antibiotics should usually be given in the emergency
department and before the admitting physician sees the patient.
 Fever 79

Case 3.3 Fever and rash, double trouble

• The next case involves the death of an infant who presented with a fever
and a rash. In Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine, Selbst and Korin write about a five-month-old girl who was
brought to a clinic with a fever and was prescribed a medicine for the
‘flu.’21 The infant was brought back to the clinic to see the same physician
that night after she had developed a rash. She was given an additional
medication on the second visit and no testing was initiated. Shortly after
the second visit, the child developed respiratory distress and cyanosis
in the extremities. Her parents brought her back to the clinic but the
infant could not be saved and died. An autopsy revealed
meningococcemia and the case was settled for $225,000 against the
treating physician.
• The presence of a rash with a fever is usually alarming for a serious
infection. This is especially true in the pediatric population. These situ-
ations often dictate further diagnostic tests to look for bacteremia or sepsis.
The threshold for diagnostic tests should certainly be lowered for return-
ing patients. There are few exceptions such as urticarial lesions in response
to an antibiotic or vesicles in the presence of a herpetic infection.

Case 3.4 The picture did not fit

• This case is of a 19-year-old college student who lived in a sorority house.
She came to the emergency department stating that she ‘was dehydrated.’
She had subjective fevers and numerous bouts of vomiting for two days.
She also complained of a headache but denied any other symptoms (stiff
neck, confusion, photophobia, cough or shortness of breath, abdominal
or flank pain, and urinary symptoms). Earlier, she went to a walk-in
clinic where they wanted to test her urine. She was unable to produce
any urine so the walk-in clinic diagnosed her with a viral syndrome.
• She was afebrile with stable vital signs on presentation to the ED. She
appeared tired but was not in distress. Her physical exam was relatively
unremarkable except that she complained of pain when her neck was
moved (it was not stiff). She was given intravenous fluids and medicine
for nausea and diagnostic tests were ordered. Her white blood cell count
was mild to moderately elevated at 15,000 cells/microliter and many
bacteria were present on her urine sample. The chest X-ray was normal.
The patient felt better and was written up for discharge with the
diagnosis of urinary tract infection. Before discharge, however, we felt
uneasy about this complaint of headache, which she still had, and the
lack of urinary tract symptoms. Therefore, we performed a lumbar
puncture that revealed viral meningitis. She was hospitalized for five
days due to the extreme fatigue, nausea, and vomiting.
• Be careful about attributing fevers to etiologies when the symptoms are
not consistent. In this case, the patient did not have convincing
symptoms of a urinary tract infection. When the diagnosis is in doubt,
consider evaluating multiple sources to exclude concomitant infections.
80 Learning from medical errors: clinical problems

Case 3.5 A crash course

• We had a very interesting case of bacterial meningitis at our facility a few
years ago. The case was interesting because it reminded us of how deadly
and how rapidly progressing bacterial meningitis can be. A 60-year-old
prominent attorney signed in to be seen in our emergency department
for leg pain. He was behaving normally and was alert in triage with no
other signs or symptoms. The triage nurse noted that he had a tempera-
ture of 100.6ºF. The patient continued to wait in the emergency department
for about 40 minutes when he suddenly became lethargic and developed
weakness on his right side. A different nurse saw the lethargy, noted
slurred speech, and brought the patient immediately back and summoned
us to examine the patient.
• The patient was sent immediately to get a CT scan of the head to evaluate
for a cerebrovascular accident. The immediate concern was for
intracranial hemorrhage or ischemic stroke, which would require
imminent treatment. The CT scan was given priority over drawing blood
work and starting antibiotics because meningitis was of less concern at
the moment. The patient returned from CT scan within 15 minutes with
the results being negative.
• While the patient was getting his CT scan, his wife was able to tell us that
he was battling a ‘sinus’ infection and came in because he had pulled a
muscle in his leg while playing tennis. With this history, the patient’s
fever, and a negative head CT, we decided to start antibiotics for
meningitis prophylaxis approximately 20 minutes after the patient was
brought back to the department. Ten minutes later, his white blood cell
count returned as 22,000 cells/microliter. We then proceeded with a
lumbar puncture and obtained cerebrospinal fluid that cultured
Streptococcus pneumoniae. The patient expired six hours later.
• This case illustrates that meningitis may produce symptoms that are
similar to those of a cerebrovascular accident. In the presence of a fever
and a nondiagnostic head CT, a lumbar puncture should be considered.
More importantly, the initiation of antibiotics can be life-saving
(unfortunately not in this case).

Case 3.6 One infection leads to another

• We saw a six-year-old boy in the office one day for eye swelling and
fever. The boy was in the office the day before with a sore throat and had
a swab performed on this throat that was positive for group A strepto-
coccus. The physician assistant who saw the child realized that he was
allergic to penicillin and decided to put him on Bactrim pediatric elixir.
• On the return visit, the child had a fever of 102ºF and appeared extremely
fatigued. He had prominent swelling and erythema of his right
periorbital area. Examination of his right eye revealed some proptosis
and a pupil that was more sluggish to light than the left pupil. His tonsils

continued
 Fever 81

were enlarged and erythematous with copious white exudates. These

findings prompted our concern of a right orbital cellulitis infection from
an extension of the streptococcal infection.
• We telephoned the child’s pediatrician and expressed our concern. We
recommended a direct admission to the hospital for the initiation of
intravenous antibiotics and a CT scan of the orbits. The pediatrician was
reluctant to admit the child and felt that it was probably a periorbital
cellulitis. He wanted us to switch the child to Zithromax and have the
child follow up the next day. We were not comfortable with this plan
because we felt that the child appeared toxic. Therefore, we spoke with
the emergency physician at the hospital and sent the child over for a
complete blood cell count and a CT of the right orbit. The child’s white
blood cell count was 28,000 cells/microliter and his CT showed right
orbital cellulitis with a developing abscess. He did well in the hospital
on intravenous antibiotics and required operative drainage of the
abscess by the Ear, Nose, and Throat specialist. Incidentally, the hospital
chaplain thanked us one month later and said that the child was his son.

Case 3.7 Enter every orifice

• A 32-year-old male presented to our ED complaining of rectal pain. He
was noted to have an elevated temperature of 99.6ºF. He did not have any
stool-related complaints such as bleeding, diarrhea, or constipation. He
was seen in the ED two days before for the same complaint and was
discharged with the diagnosis of internal hemorrhoids and instructed to
take sitz baths. A rectal examination was not performed. His external
exam revealed some tenderness around the perirectal area but there were
no definite abscesses or hemorrhoids. On internal exam, however, there
was marked tenderness over a significantly bulging area of the anal wall.
We ordered a white blood cell count, which was 32,000 cells/microliter.
We then ordered a CT scan of the rectal area which confirmed our
suspicion of a perirectal abscesses. The patient was taken to the
operating room to have the abscess drained.
• It is never pleasant to perform a rectal examination in a patient with
rectal pain. They may scream and be uncomfortable but the exam is
necessary. Do not accept the notion that all rectal pain is hemorrhoids.
Remember that a good physician leaves no stone unturned and no
crevice unentered.

Case 3.8 Getting to the bottom of things

• A 66-year-old retired professor presented to our ED with the complaint
of rectal pain for four days. He had a temperature of 99.5ºF and stated
that he had difficult bowel movements. The ED physician examined his
rectum exteriorly and did not see any swelling, redness or hemorrhoids.

continued
82 Learning from medical errors: clinical problems

He discharged the patient with the diagnosis of proctalgia and gave him
a stool softener.
• The patient returned two days later with the same complaint and a
temperature of 100.4ºF. The ED physician on this visit performed a rectal
examination and felt a hard substance at the tip of his finger. He ordered
an abdominal X-ray, which showed a mason jar impacted in the patient’s
rectum. While the patient denied having any knowledge of how the jar
got there, a review of his chart revealed that this was the third time that
he had been to the hospital for a rectal foreign body.
• The jar was removed in the operating room by a surgeon. His pain and
elevated temperature resolved after the procedure. This case reminds us
of a very valuable lesson from medical school. When searching for an
uncertain pain or elevated temperature, leave no orifice unexplored. In
addition, in dealing with rectal complaints, ask the patient about sexual
and/or social practices.

Case 3.9 Every test counts

• A five month-old infant was brought to a local ED in South Florida by
her parents because of fever and irritability for two days. The child had
a temperature of 102.5ºF but did not appear toxic. There was no focus of
infection on physical exam. The ED physician ordered a chest X-ray and
a urinalysis, which were both unremarkable. A white cell count was within
the normal limits at 10,900 cells/microliter. The differential, however,
had a slight left shift with a 12% bandemia. The patient was given a
Rocephin injection intramuscularly and discharged with the diagnosis
of fever of unknown etiology.
• The child’s parents called the ED five hours later to report that their child
was developing a rash. The triage nurse told them that the rash was
probably a reaction to the antibiotics and that they should give her some
Benadryl. One hour later, the parents brought the child to the hospital
because she had become extremely pale and her temperature had risen
to 104.5ºF. The rash was a diffuse petechial rash and not urticarial as
expected from a drug reaction. Shortly later, the child went into
respiratory arrest and could not be resuscitated. The cause of death at
autopsy was bacterial meningitis.
• In the evaluation of many infants with fevers, the source will be
uncertain. This makes the discharge instructions and the follow-up
appointments extremely important. In addition, a callback concerning
a patient with an uncertain diagnosis must be handled with extra
caution. It is usually best to have these patients return immediately for
a recheck.
 Fever 83

Case 3.10 It takes time to grow

• A 26-year-old female came into our ED with fatigue and fever. She had
just had an abdominal hysterectomy two days prior. Her temperature
was 100.9ºF and her pulse was 105 beats/minute, but her blood pressure
was within normal range. She did not have any other associated
symptoms.
• The ED physician could not find any focus for her fever on physical
examination. In particular, her lungs were clear and her abdomen had
the ‘usual’ mild tenderness that is expected postoperatively. The wound
incision looked clean and did not appear to be the source. The physician
ordered two sets of blood cultures, a complete blood cell count with
differential, a urinalysis, and a chest X-ray. With the exception of a slightly
elevated white blood cell count at 11.8 cells/microliter, all of the other
tests were unremarkable. The blood culture results were not
immediately available. The physician called the on-call gynecologist and
discussed the case with him. They decided to send the patient home and
instructed her to take Tylenol for the fevers.
• Eight hours later, during our shift, the laboratory called us to inform us
that both of her blood cultures had gram-negative organisms on the gram
stain. We immediately called the patient and asked her how she was. She
said that her fever had broken but that she was feeling more tired. We
told her to return to the ED and also called the gynecologist to admit her
for sepsis on arrival. When she returned, her blood pressure was 90/
60 mm Hg and her pulse was 125 beats/minute. Antibiotics were started
in the ED and the patient was directly admitted for gram-negative
sepsis. She had a five-day hospitalization without any long-term
complications.
• The evaluation of fevers occasionally involves tests with a delay in
results. Specifically, cultures of blood, urine, cerebrospinal fluid, and
wounds will not be available immediately. It is, therefore, imperative for
physicians to respond to results promptly. In this case, delaying the treat-
ment of sepsis for a few hours may have led to a different outcome.

Case 3.11 Pulmonary embolus, the great masquerader

• We recently had an interesting case in our emergency department that
reminded us to always think of more than just the obvious. A 45-year-
old obese woman with a history of uterine cancer presented to us
complaining of fevers, shortness of breath, and productive cough. She
had temperatures of 101ºF for one week and was coughing up yellow
sputum. Four days prior, she had developed some swelling in her legs
and her primary physician ordered venous Doppler studies on both legs,
which were negative. Her physician had prescribed an antibiotic and the
patient was convinced that she had a pneumonia.

continued
84 Learning from medical errors: clinical problems

• Her temperature in the ED was 102ºF and she was slightly tachyneic. Her
lung exam did not reveal any evidence of wheezing or consolidation. A
white blood cell count and a chest X-ray were both unremarkable. We
became suspicious and ordered a CT pulmonary artery study, which
revealed large bilateral pulmonary emboli and a deep venous
thrombosis in one of her legs. There was no pneumonia seen on the CT.
The patient was extremely distraught at this finding because of her recent
ultrasound and because her father had died of a pulmonary embolus.
She was hospitalized and treated for this event without any problems.
• Although this patient had the clinical symptoms of pneumonia, her
diagnostic studies did not support this diagnosis. Hence, the differential
of fever and tachypnea must be expanded to include pulmonary
embolism. It would have been interesting if her chest X-ray did show a
pneumonia and she still had a pulmonary embolism. Would we still have
ordered the CT scan of the chest? Why was the deep venous thrombosis
not seen on the ultrasound? We are not sure of the answer since the test
was performed at another institution. It may have been an early clot that
could not have been detected. Alternatively, it may have been due to
technician or physician (radiologist) error.

Case 3.12 Cannot live with food and water

• Medical management in cases of AGE in kids is sometimes accompanied
by complications and has resulted in successful lawsuits. This is
evidenced by two cases from Selbst and Korin’s Preventing Malpractice
Lawsuits in Pediatric Emergency Medicine, this one and Case 3.13 below.
The underlying theme of both cases is that returning patients should
have a low threshold for hospital admission. Hospitalization is not always
easy, though, because of the increased emphasis on oral rehydration for
AGE and outpatient therapy.
• A physician prescribed Lomotil for a six-year-old boy who was brought
to his office for fever, vomiting, and diarrhea.22 The boy’s symptoms
worsened in the next few days and he was prescribed an antibiotic after
being seen by another physician. The second physician received a call
from the boy’s parents two hours later when there was no improvement,
but no additional therapy was recommended. Eight hours later, the boy
lost consciousness and died because of dehydration and electrolyte
imbalance. The family sued both physicians and eventually received a
settlement of $140,000.
• We believe that the physicians involved probably were fortunate with
the amount of this case settlement. Jury awards or settlements for pediatric
deaths usually result in substantial dollars based on loss years of
productivity and pain and suffering. In addition, there were many steps
along this boy’s treatment where the outcome may have been altered.
• At the first visit, it is questionable whether Lomotil was a good drug
choice to prescribe. A child with diarrhea and fever may have bacterial
continued
 Fever 85

gastroenteritis and many authorities would not recommend an anti-di-

arrheal because it prolongs the carrier state of the bacteria. An anti-emetic
might have been more appropriate. Laboratory testing during this first
visit is controversial among physicians, but we are proponents of them
whenever there is the presence of fever.
• Laboratory testing was probably warranted at the second visit. The child
was at a greater risk for dehydration and electrolyte imbalance at this
point. Secondly, in the absence of any laboratory tests (blood cultures,
stool cultures), it is most likely inappropriate to prescribe antibiotics.
Antibiotics will not quickly help with the vomiting and diarrhea and
may actually make the diarrhea worse. The biggest mistake during this
visit was that a returning patient with worsening symptoms did not get
a step up in care (e.g. tests, hospital admission). The physician was
offered a chance at redemption when the parents called back requesting
advice but he failed to take advantage of it.

Case 3.13 A concern of great (or loss of) weight

• After a one-year-old boy started to have vomiting and diarrhea, his family
practitioner prescribed Phenergan over the phone.23 The medicine did
not help and the child was brought to the clinic that evening. His weight
was ½ pound less than one that was recorded two months prior. He was
given an injection of Phenergan and discharged home along with
instructions for his parents to call if he was not better within 24 hours.
Over the next 24 hours, the vomiting decreased but persisted and the
child was becoming lethargic, anorexic, and disoriented.
• After his condition was reported to the clinic, the child was reevaluated
and given an injection of Compazine. He was discharged with mild
dehydration and was not weighed again. His symptoms persisted
throughout the night and his temperature rose to 103.5ºF with
accompanying rapid breathing. By the next morning, his hands were
cold and he was acting more disoriented. Phone advice from the
physician was to use Tylenol and a sponge bath but not to go to the
hospital at this point. During the bath, the child stopped breathing and
died 30 minutes later. An autopsy revealed that the child had a weight
loss of about 20% from the recent illness and died from dehydration. The
settlement in this case was $750,000.
• The discussion in the previous case would also have a lot of relevance
in this case. This child had objective evidence of dehydration that was
more than mild on the initial visit. His weight (which is sometimes as
important as a vital sign in the pediatric patient) loss from two months
ago was a signal that this child was severely ill. He also did not receive
a step up in care on the second visit and actually received a step down
because no weight was recorded. Finally, phone advice in the presence
of symptoms suggesting critical decompensation of the child’s
hemodynamic status is certainly not appropriate.
86 Learning from medical errors: clinical problems

References
1 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 235.
2 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 439.
3 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 59.
4 Amin K, Kauffman CA (2003) Fever of unknown origin, A strategic approach to this
diagnostic dilemma. Postgraduate Medicine. 114(3): 69.
5 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 1485.
6 Spach DH (2003) New issues in bacterial meningitis in adults. Postgraduate Medicine.
114(5): 45.
7 Attia J, Hatala R, Cook DJ et al. (1999) Does this adult patient have acute meningitis?
JAMA. 282(2): 175–81.
8 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 1486.
9 Braunwald E, Fauci AS, Kasper DL et al. (2001) Harrison’s Principles Of Internal
Medicine (15e). McGraw-Hill, New York, New York, 369.
10 Spach DH (2003) New issues in bacterial meningitis in adults. Postgraduate Medicine.
114(5): 45.
11 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 596.
12 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 1486.
13 Behrman RE, Kliegman RM and Jenson HB (2000) Nelson Textbook of Pediatrics (16e).
W. B. Saunders, Philadelphia, Pennsylvania, 742–3.
14 Behrman RE, Kliegman RM and Jenson HB (2000) Nelson Textbook of Pediatrics (16e).
W. B. Saunders, Philadelphia, Pennsylvania, 743.
15 Behrman RE, Kliegman RM and Jenson HB (2000) Nelson Textbook of Pediatrics (16e).
W. B. Saunders, Philadelphia, Pennsylvania, 1623.
16 Baraff LJ, Bass JW, Fleisher GR et al. (1993) Practice guidelines for the management of
infants and children 0–36 months of age with fever without a source. Annals
Emergency Medicine. 22: 1198.
17 Choi Chester (2003) Bacterial meningitis: Management for a changing disease profile.
Family Practice Recertification. 25(11): 45.
18 Bonsu BK et al. (2003) Identifying Febrile Young Infants with Bacteremia: Is the
peripheral white blood cell count an accurate screen? Annals of Emergency Medicine.
42(August): 216–24.
19 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 59–60.
20 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 60.
21 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 61.
22 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 78.
23 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 79.
Chapter 4

Flank pain

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

The flanks are a common area for pain complaints in the primary care office and
the emergency department. By far, the most common presentation will be
colicky pain and the most common diagnosis will be urolithiasis. The incidence
of this disease may be as high as 12%.1 It affects almost all ages and tends to be
recurrent. Most cases will have a benign nature and represent little medical
liability. However, adverse outcomes arise when physicians fail to seek and treat
other serious etiologies of flank pain or recognize the complications of
urolithiasis.

Triage
Patients with abnormal vital signs or appearance should be triaged to see the
physician emergently. Hypotensive or tachycardic patients may require imme-
diate resuscitation with intravenous fluids and simultaneous evaluation for
vascular collapse. Consider immediate consultation with a vascular surgeon for
hypotensive patients with risk factors for peripheral vascular disease or a
history of aortic disease. In addition, pregnant females with possible ectopic
pregnancy should be seen emergently.
Patients with moderate to severe discomfort but stable vital signs and color
should be brought urgently for evaluation. The majority of these patients will
only need pain relief while a few will be harboring a developing emergent
condition. Immediate diagnostic intervention is usually not required for these
patients and the physician can often wait for the results of the initial screening
tests (e.g urine) and/or response to pain medications.
The remainder of the patients with flank pain will have normal vital signs and
appear in no to mild discomfort. These patients can be seen in the order that they
arrive. Most of these patients will have lumbar strain or uncomplicated pyelone-
phritis. They may simply need pain medicines and a urine evaluation.
88 Learning from medical errors: clinical problems

History
The majority of patients that are seen with flank pain will fall into one of three
categories: urolithiasis, pyelonephritis, or lumbar strain. Therefore, initial
questioning should be directed at these three etiologies. To assess for pyelone-
phritis, ask about typical symptoms of urinary tract infection. Since pyelonephritis
can sometimes be recurrent or associated with a complicated course, ask about
prior episodes of such and how they were treated. Also, screen for symptoms
that may suggest systemic complications from pyelonephritis. We had a college
female who developed bacterial meningitis four days after beginning treatment
for pyelonephritis. Hemodynamic status (e.g. orthostasis) should be evaluated
based on the nausea and vomiting and the decreased fluid intake associated
with pyelonephritis.
Most patients with urolithiasis have a prior history of such and can usually
give you their diagnosis. Ask them if it feels like a kidney stone and be
suspicious for other etiologies if they tell you it feels different. Red flags that
may alert you that other etiologies are present or that complications of a kidney
stone are present include: abdominal pain (abdominal aortic aneurysm), scrotal
swelling (testicular torsion), neurologic symptoms (spinal abscess or hematoma),
fevers (spinal or renal abscess), and vaginal bleeding (ectopic pregnancy).
The flanks are well lined with musculature and are an area that is prone to
straining with heavy lifting. Hence, lumbar strains are common and the patient
should be questioned for any possible mechanism of injury. Strains, however,
are benign diagnoses and should be only considered in exclusion after other
serious etiologies have been investigated and determined unlikely.
In the presence of contact trauma to the mid or lower back, always consider
the presence of a renal/adrenal contusion or injury. Although the ‘kidneys are
well protected in the retroperitoneal location surrounded by bulky musculature,
fascia, and lower ribs’,2 the flank is a common site of injury during a fall because
of the inability of the hands to protect the area. Hence, renal contusions are not
uncommon (90% of renal injuries) but renal lacerations, renal pedicle injuries,
and renal ruptures are rare due to the considerable force that is necessary to
cause these injuries.2
Penetrating trauma to the flanks exposes the patient to a number of serious
injuries that may cause hemodynamic instability and is beyond the scope of this
chapter. These injuries may violate the peritoneum and cause the typical intra-
peritoneal injuries that are seen in abdominal trauma. Furthermore, they may
also cause retroperitoneal injuries that are often difficult to detect and contribute
to increased medical morbidity.
Does the patient have risk factors for peripheral vascular disease (PVD)? Is
there a history of tobacco use or risk factors for coronary artery disease (e.g.
hypertension, diabetes, hypercholesterolemia)? PVD is associated with aortic
disease as discussed below. It is also associated with renal artery embolism and
renal vein thrombosis, which may produce signs and symptoms that are similar
to kidney stones. Diabetes and sickle cell disease are risk factors for papillary
necrosis and flank pain.
What medications are the patients using? Non-steroidal anti-inflammatory
medications may cause papillary necrosis and acute renal insufficiency. Cocaine
use may lead to rhabdomyolysis and renal infarction. We treated a 21-year-old
 Flank pain 89

healthy male for flank pain during our residency. He abused intravenous
cocaine and developed acute renal failure and rhabdomyolysis with a CPK of
over 100,000 units/liter. Prior surgery history should be ascertained for patients
with a history of urolithiasis because it may affect one’s treatment. Patients with
a ureteral stent are difficult to image with a CT urogram to look for kidney stones
because the stent may obscure the images of the stones.

Physical exam
A complete set of vital signs is essential in the patient with flank pain. Fevers
may be a clue to urinary tract infection or renal abscess. Blood pressure and
pulse give clues regarding the hemodynamic stability in patients with possible
aortic dissection or abdominal aortic aneurysm leak or rupture. The pulse should
be especially noted if it is irregular (see ‘Diagnostic tests’ section). Finally, pulse
oximetry may be helpful for the patient with basilar pneumonia.
A detailed examination of the flanks for tenderness, masses, rashes, or warmth
is mandatory. Specific patient populations require examination of additional
areas. Patients with peripheral vascular disease should have an abdominal exam
for masses and palpation of lower extremity pulses. Abdominal examination is
important for patients with suspected gallstones and right flank pain due to the
proximity of the areas and the colicky nature of both types of pain. Females of
child-bearing age may require a pelvic examination if other symptoms or signs
of pregnancy or gynecologic disease are present. Consider a rectal exam for
prostate assessment in older males and also for rectal tone in patients with
suspected spinal disease. Patients with suspected spinal disease also require a
thorough neurologic examination. Finally, patients with pain radiating to the
groin or any groin symptoms require a genitourinary exam.

Differential diagnosis
As we have already discussed, a number of conditions can produce the
symptoms of renal colic. Aortic dissection and abdominal aortic aneurysm should
be ruled out in anyone with risk factors, due to the huge medical and legal
consequences of the conditions when undetected. Pyelonephritis may generate
symptoms that imitate or accompany those of urolithiasis and should usually
be evident on the urinalysis. We have discussed predisposing factors for
papillary necrosis and renal infarction. Pelvic or abdominal masses that
compress the ureters are other causes for colicky flank pain. Numerous
gynecologic disease processes can also manifest as renal colic. These include:
ruptured ectopic pregnancy, salpingitis, tubo-ovarian abscess, and ovarian cyst
or torsion.3 Likewise, a number of male genitourinary disease processes may
also produce renal colic. These are listed in Tintinalli’s Emergency Medicine: a
comprehensive study guide as: testicular torsion, epididymitis, prostatitis, Fournier’s
gangrene, and an incarcerated or strangulated hernia.3

Diagnostic tests
For younger patients with no risk factors for peripheral vascular disease and
normal vital signs, a urinalysis is probably all that is needed. Females capable
90 Learning from medical errors: clinical problems

of child bearing should also have a pregnancy test. Urine culture should be
requested for patients with possible pyelonephritis.
If complications of urolithiasis are suspected, a complete blood count (CBC)
may be ordered for suspected infection or excessive blood loss in patients with
hematuria and taking blood thinners. Basic metabolic panel is a good screening
test for renal function. Imaging with an intravenous pyelogram (IVP) or a CT
urogram may be performed to confirm the presence of a stone. The time required
to perform these tests has always been a concern for emergency physicians when
the department is busy because the results will seldom change the management
plan.
We quote a paragraph from Tintinalli’s Emergency Medicine: a comprehensive
study guide concerning this subject, ‘it is controversial whether all patients require
emergency department imaging for suspected renal colic. For young, healthy
patients…it may be appropriate to delay the work-up…on an outpatient basis.
For older patients, especially those in whom the differential diagnosis includes
aortic abdominal aneurysm (AAA), the diagnosis should be confirmed by some
imaging modality.’4
Patients who are noted to have an irregular pulse or a history of atrial fibril-
lation should receive an EKG to document any episodes of such. This rhythm,
if not properly anticoagulated, makes the patient at increased risk of developing
an embolus to the renal artery and a subsequent renal infarction. See the ‘Pitfalls’
section below for more on this subject.
Patients with trauma to the flanks and hematuria are usually best imaged with
an abdomen and pelvis CT with contrast to examine for renal vascular injury
and for injuries to other organs. Consider an ultrasound evaluation for renal
cystic disease or gynecologic etiologies. Angiography may be necessary for
suspected renal artery embolism or renal vein thrombosis.

Hospital/office course
Patients with flank pain usually receive therapeutic interventions in the office
or emergency department. The majority of interventions will be simple pain
management but can include fluid resuscitation and emergency surgery. A
patient’s hemodynamic status can change rapidly depending on the underlying
medical condition. Therefore, accurate and complete documentation must be
completed. Patients should not be discharged if their pain is not controlled, their
vital signs remain abnormal, or serious etiologies are still under consideration.
Furthermore, ensure that the patient is able to hold down oral fluids and is able
to urinate before discharge.

Pitfalls
Do not exclude the possibility of a kidney stone based on the absence of hematuria.
Although most kidney stones will cause a urinalysis to have red blood cells, this
is not always the case. Rosen et al. state in Emergency Medicine: concepts and
clinical practice that ‘almost 20% of patients with urolithiasis documented on
intravenous pyelogram (IVP) have no microscopic hematuria.’ ‘There is no
correlation between the degree of obstruction and the absence of hematuria.’5
We present a specific case in the following section (Case 4.1).
 Flank pain 91

Maintain vigilance for patients who have a prior history of urolithiasis but
also have risk factors for peripheral vascular disease. Since these patients are
prone to developing an AAA and the symptoms of such are difficult to distin-
guish from those of urolithiasis, their evaluation should be carried out in the
same fashion as patients without a history of urolithiasis (i.e. CT urogram to
screen for AAA). (Notice that we said screen and not rule out since the urogram
is performed without intravenous contrast.)
Remember that the kidneys are end organs just like the brain or the heart.
Therefore, they are susceptible to diseases such as peripheral vascular disease
and arterial emboli. Patients with peripheral vascular disease are also at risk for
aortic dissection and renal arterial embolism. Both conditions can be associated
with renal infarction and the production of renal colic. Please see Case 4.2 and
Case 4.3 below.
Although CT urogram (noncontrast CT) has recently become the test of choice
for renal stones, there are some physicians and some facilities that do not
routinely use it. We like the CT urogram over the IVP for several reasons. It has
greater sensitivity for detecting stones and is much better at determining the
presence of obstruction. This knowledge may make a difference in the treatment
plan of some patients. Please see Case 4.4 below. In addition, it has two advan-
tages since it is done without contrast. It can be followed up with an intravenous
pyelogram if needed (the reverse cannot be done). In addition, the procedure is
not dependent on the patient’s renal function.
Remember that the organ that is located in the flanks – the kidneys – is an
organ that encompasses the functions of several organ systems. These include
the cardiovascular, endocrine, genitourinary, and respiratory systems. There-
fore, patients with flank pain and constitutional signs or symptoms may need
an evaluation for systemic diseases. Please see Case 4.5 below.
Some physicians limit the diagnostic evaluation of flank pain to a urinalysis.
While CT scans, MRI, and ultrasounds are expensive and not practical for most
practice settings, a simple X-ray of the lumbar spine may sometimes be fruitful.
This is particularly true in the patient with risk factors for peripheral vascular
disease, a negative urinalysis, and no acute history of musculoskeletal trauma.
Please see Case 4.6 below.

Errors and interesting cases

Case 4.1 It is not mandatory

• This case illustrates the problem with over-reliance of hematuria in the
consideration of urolithiasis. A young man with no history of
genitourinary disease presented to the emergency department
complaining of acute onset of severe pain in his scrotum. He had no
scrotal swelling but did mention that his testicles were tender to palpation.
He also complained of intermittent low back pain but had no urinary
symptoms.
continued
92 Learning from medical errors: clinical problems

• The patient’s urine was completely negative and a scrotal ultrasound

did not show any testicular torsion. The patient received pain relief with
Toradol and was discharged with the diagnosis of groin strain. His wife
called the emergency department two hours after they returned home
because the patient’s pain had returned and he began to vomit. The patient
was instructed to return to the emergency department where he now had
blood in his urine and a CT urogram was positive for urolithiasis.

Case 4.2 A kidney attack

• A 44-year-old female with a history of uncontrolled hypertension
presented to our ED with right flank pain. The pain had been present for
24 hours and was not associated with fevers, nausea, vomiting, or any
urinary symptoms. She denied any prior history of kidney stones. She
had visited a walk-in clinic earlier in the day and was diagnosed with
a probable kidney stone because there was ‘blood in my urine.’ She told
us that no vitals were taken and that the physician did not perform an
examination.
• On presentation, her blood pressure was 180/110 mm Hg and her pulse
was 110 and irregular. She was in severe, colicky pain and had tender-
ness in the right flank area. Her urine had moderate red blood cells but
also contained moderate red blood cell casts. Her electrocardiogram
showed atrial fibrillation. Our suspicion of a renal artery embolism was
supported with the CT findings of an infarcted kidney. The test was
negative for kidney stones. The patient was admitted for anticoagula-
tion and further testing.
• The point of Case 4.1 was that not all kidney stones produce hematuria.
The corollary to this point is that not all hematuria represents a kidney
stone. The physician at the walk-in clinic should have become suspicious
when the patient did not have any urinary symptoms or prior history of
kidney stones. As we have discussed throughout this chapter, the kidney
has an integral function in maintaining an individual’s hemodynamic
status. Therefore, it is unacceptable to evaluate this patient without
obtaining vital signs or performing a physical examination.

Case 4.3 Coke is not it

• Another case of a similar nature came during our residency. A young
male presented to the ED with severe right flank pain. He was a healthy
man with no prior medical problems but had been on a cocaine binge
with intravenous injections the night before. His urinalysis result was
similar to that of the female in Case 4.2 and he was also found to have
an infarcted kidney on CT scan. His hospital stay was prolonged and
complicated for several weeks. He subsequently developed
rhabdomyolysis and required dialysis. The pearl from this case is to
question young individuals with flank pain about illicit drugs in the
same manner that you would a young person with chest pain.
 Flank pain 93

Case 4.4 Occasionally, one will go bad

• This next case highlights the rare morbidity and mortality that are some-
times found with ureterolithiasis. The story was relayed to us by a
urologist at a conference on complications of urinary tract infections. He
said that he once had a 24-year-old female with a history of insulin-
dependent diabetes who presented to the ED with left flank pain. She
had gone to another ED two days before and had an IVP and was told
that she had a kidney stone. She was also told that she had a ‘small amount
of bacteria in her urine’ and was started on oral antibiotics. Shortly
thereafter, she developed a low-grade temperature of 99ºF but the rest of
her vitals were within normal limits. Her urinalysis had greater than
60 red blood cells per high power field and 4+ bacteria. A CT
urogram showed a partially obstructing stone in the left ureter with
hydronephrosis. Her white blood cell count was slightly elevated at 13,000
cells/microliter.
• The ED physician gave the patient intravenous antibiotics and fluids. He
called the urologist on call and informed him of the findings. The
urologist decided to come and place an emergency stent in the patient’s
ureter. The hospital, however, was a rural hospital and the urology team
did not live within the vicinity of the hospital. After one hour, the patient
became septic and developed hemodynamic unstability. She arrested and
died before the stent could be placed.
• The intriguing question in this case is whether the outcome would have
been different had the presence of hydronephrosis been known on the
first visit. Certainly, hydronephrosis and evidence of a urinary tract
infection are criteria to consider inpatient treatment. The addition of a
diabetic history in this patient would probably further support
hospitalization. However, hydronephrosis could not have been
determined because an IVP was performed instead of a CT urogram. We
discovered later that this ED had the capabilities to do a CT scan but it
was the physician’s preference to order the IVP. Some facilities, although
they have the capabilities to do a CT scan, still prefer to use IVPs in the
evaluation of renal stones. We have even worked in some facilities where
the ‘kidney stone’ protocol consists simply of a kidneys, ureters, and
bladder (KUB) X-ray.

Case 4.5 A fresh approach

• Edwards presents a case, in The M & M Files, of an unusual cause of flank
pain.6 It is one of those that we call ‘zebras’ in medicine and included it
to stimulate the medical minds of our readers. A 35-year-old man came
to the ED complaining of three days of intermittent left flank pain. He
had no medical problems and no history of kidney stones but noted that
his urine was occasionally dark.
• His history was also notable for recent weight loss and fatigue. His
continued
94 Learning from medical errors: clinical problems

physical examination was normal with the exception of a slender and

tired appearance. A urinalysis and a kidney-ureter-bladder X-ray were
both normal. The radiologist, however, noticed that there was consider-
able stool in the colon. The ED physician explained to the patient that he
was not sure of the source of the patient’s pain but recommended a bottle
of magnesium citrate and to follow up with his primary physician if the
pain did not improve.
• The patient’s wife called the ED to complain the following day because
her husband did not appear to be any better. She was also not pleased
that he was told that constipation was causing his problems. She wanted
to file a complaint but was asked to bring the patient for re-evaluation.
On his return visit, he was interviewed and examined for an extended
amount of time by a nurse practitioner student.
• The student then presented the case to the ED director and the two
proceeded to see the patient together. The director noticed that the
patient had periorbital swelling and thinning of the skin and appeared
like a ‘renal dialysis’ patient. He suspected a renal pathology for the
patient’s problems. The student, however, suggested that the thyroid
function be tested.
• When the test results were obtained, the complete blood count, the se-
rum chemistries, and the urinalysis were all unremarkable. The
thyroid-stimulating hormone, however, was extremely high at 154
microunits/ml (normal 0.85–2.32). The patient was admitted for hypo-
thyroidism and given high doses of levothyroxine. His appearance was
dramatically improved and the endocrinologist believed that an
irreversible myxedema state had been prevented by the prudent
detection of his thyroid dysfunction.
• In his discussion, Edwards emphasizes that the diagnosis was missed
because the first clinician was focused on a common complaint (flank
pain). Indeed, even the medical director was concerned about a renal
etiology of the patient’s constitutional symptoms. In reality, the pain was
‘probably caused by intestinal colic secondary to hypothyroid-induced
constipation.’ 6 This case emphasizes our previous discussions
concerning the kidney’s interactions with other organ systems.

Case 4.6 A good pick-up

• We saw a patient in the ED who was referred by a chiropractor. The
patient was a 67-year-old army veteran with a history of tobacco use but
no medical problems. He had been seeing his chiropractor for pain in his
left flank for two weeks. After receiving several adjustments over this
time, the chiropractor performed a lumbosacral X-ray of the patient’s
back.
• The X-ray showed a large semicircular ‘dark bulge’ to the left of the
patient’s spine. Concerned that the bulge might be an aneurysm, the

continued
 Flank pain 95

chiropractor sent the patient to the ED. Although the patient was totally
asymptomatic, a CT scan confirmed the presence of a 6.7 cm abdominal
aortic aneurysm. The patient was scheduled to follow up with the
vascular clinic to schedule elective surgery and advised to stop
smoking. He was also instructed to follow up with his primary
physician to monitor his blood pressure.
• The chiropractor made a life-saving call in this case. He became
suspicious when the patient’s pain did not improve with the adjustments.
His clinical acumen was also displayed when he detected the aneurysm
on the X-ray. The diagnosis of AAA by X-ray is complicated by several
factors. X-ray imaging has very poor sensitivity for detecting AAAs,
particularly ones where there is little calcification of the aorta, as was the
case with this one. Secondly, even when evidence of an AAA is present
on X-rays, it is often not appreciated by clinicians and detected only by
radiologists. As discussed in Chapter 13, Case 13.13, physicians often
miss findings on X-rays when they are not specifically looking for them.
Finally, the patient in this case had no absolute indication for any radio-
logic imaging studies. There are no protocols that we are aware of
concerning low back imaging for nontraumatic pain of two weeks
duration. However, the chiropractor was keen in considering the patient’s
age, history of tobacco use, and lack of improvement with therapy.

References
1 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 640.
2 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 1714.
3 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 644.
4 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 642.
5 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St. Louis, Missouri, 2251.
6 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 226–28.
This page intentionally left blank
Chapter 5

Headache

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Headache is a common complaint in the primary care office and the emergency
setting. Emergency department visits for headache can be up to 4%.1 While some
have a history of chronic headaches and are only seeking pain relief, others have
not experienced similar pain previously and are concerned that they might have
a serious medical condition. The International Headache Society has classified
headaches into primary and secondary headaches. Primary headaches are mostly
benign and represent the majority of headaches. This group consists of migraines,
tension, and cluster headaches. Secondary headaches have underlying pathol-
ogy and represent approximately 4% of headaches seen in the ED.1 In certain
populations, it may be as high as 15%.2
Secondary headaches can originate from various different organ systems. Some
of the more common ones include: vascular (subarachnoid hemorrhage,
cerebrovascular accident, temporal arteritis), central nervous system
(meningitis, encephalitis, pseudotumor cerebri), immune (sinusitis, herpes
zoster), ophthalmic (glaucoma, iritis), drug-related (nitrates, chronic analgesic
use), metabolic (hypoglycemia, carbon monoxide poisoning), and miscellane-
ous (malignant hypertension, preeclampsia, post-lumbar puncture).1 It is the
primary task of the treating physician to determine whether there is a secondary
cause for the patient’s headache, followed by the secondary task of treating the
headache. More specifically, it is important to determine whether five emergent
causes of headaches are present. These consist of:
1 intracranial hemorrhage
2 infectious meningitis
3 hypertensive encephalopathy
4 hypoxic conditions and
5 mass lesions of the brain.
98 Learning from medical errors: clinical problems

Triage
As discussed in the introduction, any patient with signs or symptoms suggest-
ing any of the emergent five conditions above should be brought to the physician’s
attention immediately. For example, patients with headaches and any of the
following characteristics should be treated emergently:
1 a personal or family history of subarachnoid hemorrhage
2 fever, neck stiffness, or confusion (please see Case 5.5)
3 elevated blood pressure, confusion, shortness of breath, chest pain, nausea,
or vomiting
4 shortness of breath, low pulse oximetry, or exposure to carbon monoxide
5 history of cancer, human immunodeficiency virus (HIV) infection, or brain
trauma.
Similar to the triage of other patients, patients with headaches who have
abnormal vital signs or have focal neurologic complaints should be seen
immediately. In addition, headache patients with altered mental status or
seizures deserve top priority. New onset headache patients, older patients (greater
than 55 years), and patients with visual problems should be considered for
emergent evaluation and treatment.
Nonemergent headache patients include predominantly primary headaches
and the remainder of secondary headaches. These patients should be seen as
soon as possible for pain management and to ensure that an emergent diagnosis
was not missed during the triage. Triage nurses should be alert that patients
with primary headaches may have co-existing secondary headaches. This is
especially true if the patient does not usually seek medical attention for her
primary headache. Finally, there are subsets of headache patients who are
classified as ‘drug seekers.’ Again, triage nurses must be careful not to let prior
bias intervene with the current triage.

History
Soliciting a headache history should be as routine as obtaining a history for any
other pain complaint. Specifically, the onset and characteristic of the pain should
be elicited. Alleviating and precipitating factors are also important. In addition,
search for ‘red flags’ that may alarm you to a more serious cause for the head-
ache. These include headaches that occur suddenly and reach maximum intensity
within 30 seconds (thunderclap headaches) and headaches that are described as
‘the worst headache of my life.’ Both of these descriptions should spark the
physician to consider subarachnoid hemorrhage. Other red flags include fever,
neck stiffness, confusion, neurologic signs or symptoms, lethargy, history of head
trauma, no prior history of headaches, anticoagulant use, history of cancer or
HIV, headaches awakening the patient at night, position related headaches, and
elevated blood pressure. Associated symptoms to inquire about are
photophobia, confusion, visual problems, nausea, vomiting, neck stiffness,
shortness of breath, and weakness.
For the patients with prior histories of headaches, ascertain if this headache
is similar and, if so, document it. This may be your defense to the courts for not
pursuing further work-up. Inquire about previous evaluations for the past
 Headache 99

headaches (CT scans, MRI scans, neurologic evaluation, etc.). Record the
medications that they have taken for the pain (rebound headaches, potential
drug abuser), as headaches tend to respond to similar medicines. Family history
of headaches is important as it may suggest a secondary cause to the headaches.
Pertinent social history questions are tobacco use (risk factor for subarachnoid
hemorrhage), cocaine use (association with elevation of blood pressure,
subarachnoid hemorrhage, and seizures), alcohol use (association with subdural
hemorrhage), and HIV risk factors (association with intracranial infections).
Finally, a history of recent or remote head trauma is essential.

Physical exam
The importance of recording all five of the vital signs discussed in the other
chapters is again emphasized. Because many systemic processes can cause
headaches, these vital signs may serve as early clues. The febrile patient may
have sinusitis, encephalitis, meningitis, brain abscesses, or brain metastases
(paraneoplastic fevers) as the cause of their fever. Outside of the brain, gastro-
enteritis is another infection that is commonly associated with headaches.
Hypertensive patients may have headaches that are pressure-related. Patients
with hypoxia or tachypnea may have headaches that are caused by hypoxia.
Headaches in the presence of an irregular pulse suggest the possibility of a cerebral
arterial embolus from atrial fibrillation. Finally, any patient with a headache and
visual complaints outside of photophobia should have a visual acuity
documented and a measurement of intraocular pressure for glaucoma should
be considered in the population at risk. These patients should not be assumed
to be suffering from a migraine.
The physical assessment should always start with a general impression. How
does the patient look? How are they walking into the examination room? How
are they conversing with their family or the nurse? To what extent do they appear
distressed? Since headache patients are one of the most common types of
patients to be labeled as ‘frequent flyers,’ how do they appear in relation to their
past appearances?
The HEENT (head, eyes, ears, nose, and throat) examination is of extreme
importance in these patients because of its proximity to the pain. Examination
of the eyes is important for sclera color, pupil size and color, reactivity, and
intraocular pressure. In addition, the eyes should be checked for papilledema
or visual field deficits. Check for the presence of sinus tenderness to palpation.
Is there tenderness over the temporal arteries or are the arteries prominent? Is
the pain in the distribution of the trigeminal nerve? Can the pain be localized
to the temporomandibular joint or reproduced when the patient bites
down? Finally, patients with oral pain should be examined for dental
abscesses.
The neck is home for the extension of some brain components (e.g. meninges)
and should also be examined in every patient with a headache. Testing for nuchal
rigidity can be screened with active range of motion and if abnormal, passive
range of motion can be added. Kernig’s sign (with thigh flexed on abdomen
patient resists knee extension) and Brudzinski’s sign (attempt to flex the neck
results in reflex flexion of the knee and hip) are two familiar responses to
maneuvers designed to assess nuchal rigidity.3
100 Learning from medical errors: clinical problems

As we discussed above, headaches may be associated with many systemic

processes. Therefore, we feel that it is good practice to routinely perform lung,
heart, and abdomen examinations in these patients. The exams are quick and
serve as a good screen for the general health of the patient.
Certainly, the most important component of the physical examination is the
neurologic exam. We have already discussed the importance of having a mental
status examination for the patient with a headache. This does not have to be the
extensive mini-mental examination that we learned on our psychiatry and geri-
atric rotations. From a legal standpoint, it is probably sufficient to document in
the chart that the patient is alert and oriented and elaborate on that with a brief
statement describing the patient’s mental capacity. Examples could include:
‘patient has good understanding of our conversation, the patient’s mental status
seems appropriate, patient does not appear mentally impaired’, etc. The remain-
ing components of the neurologic exam are the cranial nerves, the motor and
sensory exams, the reflexes, and the testing of gait and cerebellar function. Any
new abnormal neurologic finding requires an emergent investigation because
‘focal or nonfocal neurologic findings in an ED headache patient have a 39 percent
positive predictive value for intracranial pathology.’4

Differential diagnosis
Most headaches are classified as primary headaches and consist of migraines,
tension, and cluster headaches. The diagnostic challenge for the physician is to
make sure that a secondary cause for the headache is not present because these
are ones that usually require emergent treatment. Schull provides an excellent
chart in Emergency Medicine: a comprehensive study guide of these secondary causes,
which he divides into seven broad categories.1
1 Vascular headaches consist of brain hemorrhages, ischemia, temporal
arteritis, and carotid or vertebral artery dissection.
2 Central Nervous System (CNS) infections include meningitis, encephalitis,
and cerebral abscesses.
3 Other infections that may cause headaches are sinusitis and herpes zoster
with cranial nerve involvement.
4 CNS headaches may also be caused by tumors and pseudotumor cerebri.
5 Eye diseases that may cause headaches are glaucoma, iritis, and optic
neuritis.
6 Common drug and metabolic causes of headaches include: nitrates,
hypoglycemia, monosodium glutamate, and carbon monoxide
poisoning.
7 Finally, malignant hypertension, preeclampsia, pheochromocytoma, and post-
lumbar puncture are miscellaneous etiologies of headaches.

Diagnostic tests
Usually, the most important decision concerning diagnostic testing is whether
to obtain a CT scan of the head. We recommend that a CT scan be strongly
considered in all patients with any of the ‘red flags’ mentioned above. In
addition, we have had colleagues who have ordered a CT scan when they just
 Headache 101

felt a ‘gut instinct’ that their patient was not having a benign headache and
found intracranial bleeds and masses. Hence, there are no rigid guidelines of
which we are aware concerning when to order a CT scan. Our motto is: if in
doubt, order the scan. Also, if the patient requests a CT scan, we usually comply
only because of the low sensitivity of the physical exam and the presence of
intracranial masses and bleeds in diverse age groups of patients. These factors
make it hard for us to convince a patient who thinks she needs a scan that she
does not need one.
In addition to providing reassurance for a primary headache, CT scans are
helpful in the evaluation of many disease processes that are associated with
secondary headaches. These include intracranial and extracranial blood or fluid
collections, intracranial masses, sinus inflammation, orbital infections, and
infections such as abscesses and encephalitis. Although CT scanning may not be
sensitive in the early detection of encephalitis, findings may include hypodense
involvement of the temporal or frontal lobes.8 Please see Case 5.4.
Other laboratory data that may be helpful in the evaluation of headaches
include erythrocyte sedimentation rate (temporal arteritis), cerebrospinal fluid
analysis and pressure measurement (more on this below), and intraocular
pressure (glaucoma).

Hospital/office course
The most important part of this section is the documentation of a neurologic
examination. This includes a mental status examination, a motor and sensory
examination, deep tendon reflexes, and observation of gait and cerebellar
function. Additional tests such as the ones described above for nuchal rigidity
should be used if appropriate. Serial neurologic exams are frequently required
to demonstrate stability or to look for changes. Obtain consent for any
procedures (e.g. lumbar puncture, angiogram, etc.) before administering
sedating medications.
In the event of suspected central nervous system infection and other
processes, the timing of your actions and documentation is crucial. Antibiotics
must be often started promptly (even before tests results are completed) with
suspected infections. In the event of intracranial bleeds or masses, increased
intracranial pressure must be addressed and seizure protection must be given.
Other specific emergent treatments include steroids for temporal arteritis and
beta-blocker drops for acute glaucoma.
In general, it is important to assure that the patient’s pain and neurologic
status have improved or remain stable before discharge. If these two conditions
are not met, it is usually prudent to perform more tests for secondary causes.
Please see Case 5.1. It is helpful to note comments such as ‘pain is improved’ or
‘neurologic exam is unchanged’ when the patient is discharged. Patients that
still have pain should be given strict instructions on follow-up and when to
return. It should be clearly understood that they are welcome to return to you
for re-evaluation if the condition does not improve. This understanding
is essential because a significant amount of diagnoses of secondary
headache diagnoses are made on repeat visits. Finally, any patient that
has worsening pain or altered mental status should not be a candidate for
discharge.
102 Learning from medical errors: clinical problems

Pitfalls
Emergency physicians have become lax about performing lumbar punctures for
headaches. In the past four years, we have seen two patients who were
discovered with subarachnoid hemorrhages on return visits and one patient was
found to have bacterial meningitis on a return visit. There are multiple reasons
that may contribute to hesitancy in performing lumbar punctures. The
complaint of headache is fairly common, with the majority stemming from non-
malignant causes. Lumbar punctures are extremely time-consuming (not so much
the actual procedure but the labs typically take hours to return) and require that
the patient remain in the emergency department for hours. The procedure is an
invasive one and is sometimes difficult in the obese individual due to the
distortion of the lumbar anatomy. Finally, the exam and head CT are often
normal, which make it difficult to justify the procedure to the patient.
Be careful about discharging any patient from your office or emergency
department if you cannot get their headache better or perform the necessary test
to comfortably exclude serious etiologies. This is sometimes difficult in the
primary care office because of the lack of availability of tests. However, it is also
problematic in the emergency department when the department is busy or lacks
the necessary equipment (CT scanner). Please see Case 5.1 below.
We have reviewed many charts of patients presenting with the complaint of
headache that we have found to be deficient. Namely, these charts lack a
neurologic exam. This is probably one of the most common documentation
errors that we have seen. It is frequently seen in young and inexperienced
physicians but also seen in older physicians. Failure to document a neurologic
examination in a patient with a headache will make the defense of an adverse
outcome almost impossible. Please see Case 5.3 below.
Headaches that are position-related should be of great concern. This may be
a subtle sign of an intracranial mass or pressure effect. The etiology may range
from benign processes such as Arnold-Chiari syndrome or sinusitis to emergent
processes such as intracranial hemorrhage or tumor. Raskin writes in Merritt’s
Neurology, ‘head pain appearing abruptly after bending, lifting, or coughing can
be a clue to a posterior fossa mass or the Arnold Chiari Malformation . . . Ortho-
static headache arises after lumbar puncture and also occurs with subdural
hematoma and benign intracranial hypertension.’6 If these headaches are
further complicated by symptoms in the spine or extremities, there should be an
even greater concern that there is central nervous system irritation or
inflammation. Please see Case 5.2 in the following section.
Some patients with traumatic headaches may have a delayed traumatic
intracranial hemorrhage (DTICH). According to Trauma Management: an
emergency medicine approach, there are two common scenarios of DTICH.7 The
first is when an intraparenchymal hemorrhage develops within hours to days
in an area of contused brain. The second occurs when an epidural hemorrhage
or a subdural hemorrhage occurs in a patient with a normal initial head CT scan.
The incidence of DTICH has been reported to be as high as 8.5%, with most
injuries occurring within 36 hours of the injury.7 Therefore, maintain a low
threshold for repeat head CT imaging in the trauma patient with a normal initial
head CT and progressive neurologic deterioration.
Although sinusitis is a frequent cause of headaches, headaches by themselves
 Headache 103

almost never represent sinusitis. This was the statement made by Dr Howard
Levine, otolaryngologist and director of the Cleveland Nasal Sinus and Sleep
Center, at a recent meeting.8 Dr Levine noted that there should be other
symptoms to suggest sinusitis such as facial pain or pressure, nasal obstruction,
purulent nasal discharge, or changes in taste or smell. Furthermore, he adds that
headache is a minor criterion for sinusitis. Therefore, for patients with the
lone complaint of headache, you should look for an etiology other than
sinusitis.
We have performed many lumbar punctures throughout our careers. We must
admit, however, that we rarely measure the cerebrospinal opening pressure. The
goal of most lumbar punctures is to examine the fluid for infection, blood, and
cell counts. Because the diagnosis of pseudotumor cerebri is not common and
it is not considered a ‘life-threatening’ disease, physicians may not measure the
opening pressure to look for it. However, as Case 5.6 demonstrates, it is helpful
to include it in your work-up for certain patients who are at risk.
Cerebral venous thrombosis is a rare cause of headaches. It is extremely
difficult to diagnose because the clinical manifestations can be nonspecific and
variable. Furthermore, diagnostic tests that are available in most EDs (CT scan,
lumbar puncture) are not useful. MRI (magnetic resonance imaging) and/or
MRA (magnetic resonance angiography) are frequently needed. Suspicion should
be maintained in those with risk factors. These factors are similar to those of
deep venous thrombosis and pulmonary embolism discussed in Chapter 9. Please
see Case 5.7 below.
What is the standard of care for the initial evaluation of a vascular headache?
From our experience, most neurologists and emergency physicians agree that a
noncontrast head CT and a lumbar puncture are adequate for ruling out a
subarachnoid hemorrhage. Further diagnostic tests are not usually recommended
for the initial evaluation of a headache. Are these tests, however, sufficient for
the patient who has a headache from an unruptured intracranial aneurysm?
Please see Case 5.8 below. Case 5.9 is an interesting follow-up case.

Errors and interesting cases

Case 5.1 But my pain was not better

• A 32-year-old man with no history of headaches or any other medical
problems presented to the emergency department with a headache. He
stated that the headache began suddenly and violently at 2:30 pm. The
pain was located in the back of his head with radiation to the front of his
head. He saw his primary care physician and was given an intramuscular
injection of Stadol and Phenergan and discharged. The patient stated
that no tests were ordered and his pain was not improved at discharge.
After no improvement for one hour at home, he decided to come to the
emergency department, where a head CT showed a subarachnoid
hemorrhage.
• This man had a couple of red flags associated with his headache. He had
no prior history of headaches. The onset of the pain was sudden and
continued
104 Learning from medical errors: clinical problems

severe (thunderclap), suggesting subarachnoid hemorrhage. His primary

physician, in addition, discharged him while his condition was not
improved. This practice is high medical risk in patients with headaches.
In patients without a prior diagnosis of headaches, it represents an even
greater risk.

Case 5.2 Be careful if it hurts when they move

• This is a case from Edwards’ The M & M Files.9 A 69-year-old woman
came to the ED complaining of right shoulder and arm pain for several
days. There was no history of trauma and she had no other associated
symptoms with the exception of a headache that was worse when she
bent forward and an unusual discomfort in the back of both thighs. The
ED physician that saw her performed a neurologic exam that was
unremarkable; he then reassured her family’s concerns that the patient
was not having a stroke. He also felt that no diagnostic tests were in order
and discharged the patient with the diagnosis of ‘transient arthralgias.’
• The patient returned to the ED three days later with the same discomfort
in both thighs and a position-related headache. She also now stated that
she had an aching sensation in her lower back but denied any other new
symptoms. A second ED physician evaluated her on this occasion and
ordered X-rays of the lumbosacral spine, which were negative. He also
ordered a complete blood count and a chemistry panel, both of which
were normal. The patient was given ibuprofen, which seemed to help her
pain. She was then discharged with the diagnosis of ‘possible lumbar
disc disease and flu-like syndrome.’
• There was no improvement in the patient’s symptoms and she presented
to a different ED three days later. The third ED physician to see her sensed
the frustration and anger that the patient’s family had with her previous
evaluations. He, therefore, took a detailed history of the patient’s illness.
She was ‘sitting at a church meeting when a sudden and strange tingling
sort of pain moved up both thighs and, over a minute or so, sped up her
back and into her head.’ She felt like her ‘head was going to explode.’10
‘After 5 or 10 minutes, the headache and other sensations calmed down’
and she had felt like she had been ‘run over by a truck’ since.10
• Although her physical exam continued to be normal, the third ED
physician was concerned of a central nervous system spinal or intra-
cranial vascular accident. He contacted a neurosurgeon at a larger medical
center and transferred the patient to his service. The patient had an MRI
scan of the spine, which was normal but developed a severe headache
and became obtunded during a cerebral angiogram. The CT scan showed
a large aneurysm.
• Edwards’ analysis of the case is as follows. ‘In retrospect, the patient’s
initial symptoms obviously came from a “herald bleed” – a minor leak-
age from an aneurysm preceding a more catastrophic hemorrhage. The

continued
 Headache 105

extremity symptoms were paresthesias related to cerebral cortical

irritation.’10 He also recommends that ‘any abrupt onset of neurologic
symptoms – headache and/or extremity sensations of pain, paresthesias,
or weakness – should raise concern for a central nervous system vascular
accident.’11
• The first two physicians in this case seemed to be distracted by the patient’s
constitutional symptoms. They did not address the patient’s complaint
of headache. The patient’s headache should have raised concern for a
secondary cause because of its positional nature. A low threshold for
ordering a head CT should be maintained with these types of headaches.
In the event that the head CT is negative, a lumbar puncture should be
considered.

Case 5.3 Is the brain working?

• Failure to document a neurologic examination for a patient who presents
with a headache will leave future reviewers with many questions on
your thought processes. An ED physician treated a 79-year-old gentle-
man with a scalp laceration. The man had a history of atrial fibrillation
and was on Coumadin. The laceration occurred when he slipped and fell
on his head on the day of the visit. The physician’s note contained three
lines. ‘Patient with a non-syncopal fall.’ ‘Scalp laceration closed with
staples.’ ‘Return in 7 days for staple removal.’
• The patient’s daughter brought him back to the emergency department
one week later because of frequent episodes of loss of balance and
confusion since the fall. A subsequent head CT scan revealed a subacute
subdural hemorrhage from the previous fall. The patient’s daughter filed
a complaint with the hospital administration concerning the delayed
diagnosis.
• This physician had a case that no one would want to defend. Most
physicians would order a head CT for any patient who is on a blood
thinner and sustains a head injury. This practice, however, is not
universal and is also controversial in the medical literature. In this case,
however, the patient is also elderly and probably has brain atrophy. The
atrophy increases the likelihood of a traumatic subdural hemorrhage. In
addition, if the physician decides to not perform a CT scan, she must
realize that she is practicing within the guidelines of the minority of
physicians and should document the reasoning of her thought process.
Finally, head injury patients must be given clear and strict instructions
on worsening signs and symptoms and when to follow up.
106 Learning from medical errors: clinical problems

Case 5.4 Not in pictures

• Justin Ly, MD presents in American Family Physician the case of a young
lady with a headache and neurologic findings where initial testing was
nondiagnostic. 12 A 20-year-old woman with no prior history of
headaches presented to the ED complaining of one of the worst head-
aches of her life that was on the right side and associated with weakness
of the left arm. She had a low-grade temperature of 99.2ºF but had no
neck stiffness, nausea, or vomiting. Physical exam revealed mild weak-
ness of the left upper extremity. A CT scan was performed and interpreted
by the radiologist as normal.
• The patient returned several days later because the weakness had not
improved. On this occasion, a brain MRI was ordered with a T2 weighted
image, which showed abnormal asymmetric increased signal at the right
temporal-parietal gray matter. This apparent edema of the temporal lobe
was consistent with viral encephalitis. The patient had an uncomplicated
course on antiviral therapy. Herpes simplex virus was later confirmed
from the cerebrospinal fluid.
• This case is important because it highlights the limitations of a CT scan
in the evaluation of headaches. Many intracranial disease processes are
too small or too early to be detected with a CT scan. In these events,
maintain a very low threshold for performing a lumbar puncture. In the
case of this young lady, she had neurologic findings and a low-grade
temperature. This combination should lower the threshold for perform-
ing a lumbar puncture. In addition, she stated that this was one of the
worst headaches of her life, which should always raise concern for a
subarachnoid hemorrhage.

Case 5.5 Headaches and warning signs

• Drs Grassie, Henry, and Wagner present a case in Foresight where a patient
presented with a headache and an accompanying red flag and was
inappropriately triaged.13 The hospital, subsequently, had no defense
against a liability claim for a poor outcome. The patient was a 19-year-
old male with a history of illicit drug use and came to the ED complaining
of a headache and fever for the past two days. He admitted to the recent
activity of ‘shooting speedballs.’
• The triage nurse wrote down these complaints and noted that he had a
temperature of 101.5ºF. She then sent him back to the waiting room and
did not call for him again until 30 minutes later. At this time, the patient
did not respond to the call. The nurse made no further effort to find the
patient or inform the physician. She waited another 30 minutes before
calling him again. Again, the patient is absent and remains so during two
subsequent calls for him. Finally, the ED staff labels his chart as ‘left
without being seen’ and files the encounter with the rest of the charts.
• During all of this activity in the emergency department, the patient was
wandering through the hospital. When he did return home, he appeared
continued
 Headache 107

confused and went to sleep in his room. These events, however, did not
alarm his mother until two days later when she called the ambulance to
take him to the hospital. Here, the patient succumbed to a brain abscess
a short time later.
• The plaintiff attorney filed a lawsuit on behalf of the patient’s family
against the hospital for performing an inadequate screening examination
and not recognizing that the patient did not have the mental capacity to
deny treatment. The hospital could not muster a defense because the
triage nurse admitted that she did not remember the patient. Further-
more, there was no documentation of the patient’s mental status during
the encounter. The case went against the hospital for $1.5 million.
• What went wrong in this particular case? The triage nurse did not
recognize that the patient had a red flag with his headache. Fever with
a headache represents an intracranial infection until proven otherwise.
Due to the severity and rapid development of these infections, they
usually require immediate treatment. In addition, the infection’s location
gives it the potential to alter a person’s mental status and screening for
these changes is obligatory. Finally, once the patient was not found with
the first summons, a search around the hospital or a call to his home was
imperative from a legal standpoint to show that every effort was made
to treat the patient.

Case 5.6 Making a timely diagnosis

• This case from Wallenstein and Jagoda in Foresight demonstrate the
necessity of making a timely diagnosis of pseudotumor cerebri.14 A 29-
year-old obese female presented to the ED with headaches that had been
present for several weeks. She had four to five episodes during this time,
with the last one being unbearable. The pain occurred in a crescendo
pattern and lasted for hours with no relief from ibuprofen. There was no
association with fevers, nausea, vomiting, or neck stiffness.
• Her vital signs in the ED were all within normal limits. The physician
noted that head and neck, cardiopulmonary, abdominal, and neurologic
examinations were all unremarkable. He gave the patient a narcotic pain
pill and ordered a pregnancy test. When she received some relief and her
test was negative, he discharged her with instructions to take ibuprofen
as needed and to follow up with her physician in one week.
• Three weeks later, the patient was still having the headaches and her
earliest appointment with her physician was still three weeks away. She
returned to the same ED and saw a different physician. This physician
asked her if there were any visual symptoms. She responded that her
vision occasionally ‘goes out’ with the headaches. He then performed a
detailed eye examination and discovered that she had bilateral
papilledema and peripheral vision loss. A CT scan of the head was normal
and a lumbar puncture was performed. The opening pressure was elev-
ated and the patient was started on acetazolamide and a weight loss
program for pseudotumor cerebri.
continued
108 Learning from medical errors: clinical problems

• The patient continued to have visual problems and headaches six months
later. She filed a lawsuit against the first physician and the hospital for
a delay in diagnosis. The lawsuit claimed that an earlier diagnosis would
have prevented the permanent visual impairment. Although this claim
would have been difficult to prove in court, the defense had concerns
that prompted them to settle out of court. Namely, these concerns were
the absence of a detailed eye exam and questions about visual difficulties.
• This is an important learning case for all physicians who treat headaches.
The majority of patients like the one in this case will not have a serious
etiology and can be treated symptomatically. However, as we mentioned
in the ‘History’ section earlier, associated symptoms such as visual
changes should always be documented. In addition, a detailed eye exam-
ination is probably of equal importance as a neurologic examination in
the evaluation of headaches. Finally, the physician should recognize that
this patient fits the demographics for pseudotumor cerebri (i.e. obese,
child-bearing). In doing so, diagnostic tests may be performed earlier or
closer follow-up arranged.

Case 5.7 A real mind-bender

• Another case from Foresight by Wallenstein and Jagoda highlights the
difficulty with making the diagnosis of cerebral venous thrombosis.15 A
32-year-old lady presented to the ED with headaches. She was two weeks
postpartum and had migraines before her pregnancy. This pain, how-
ever, felt different. The headaches had been present for five days and
were not associated with fevers or vomiting.
• Her past medical history was significant for a blood clot in her leg 10
years ago. This occurred shortly after she was placed on oral contracep-
tives. She had to take blood thinners for a few months. Her doctor at the
time told her that she had ‘some blood disorder.’
• On physical examination, her vital signs were all unremarkable. The ED
physician documented normal eye, cardiopulmonary, abdominal, and
neurologic exams. He suspected that the patient was having a migraine
and gave her metoclopramide intravenously. This gave her some relief
and she was discharged and told to follow up with her primary
physician as needed.
• The patient returned to the ED one week later when the headaches
recurred. She was evaluated by the same ED physician and found to
have an expressive aphasia. A noncontrast CT scan of the head was
performed and read as normal by the radiologist. The patient was
admitted to the neurology service. A subsequent MRI showed a
thrombus in the left lateral sinus.
• Although her hospitalization was complicated by lethargy and seizures,
she made a complete recovery. She was discharged a week later and
followed up with the neurology clinic. At her follow-up visit three months
later, her neurologic exam was normal and her headaches were no longer
present. No legal action was taken against the physician or the hospital.
continued
 Headache 109

• The hospital risk management department reviewed the case with the
physician and addressed one concern to him. They felt that he did not
appropriately acknowledge the patient’s risk factors for venous
thrombosis. The patient’s post-partum state, history of deep venous
thrombosis, and history of ‘blood disorder’ should have all been red
flags to the diagnosis.
• It is difficult to find fault with the physician’s treatment in this case. A
young patient with a normal neurologic exam rarely requires any brain
imaging. Furthermore, the cost and limited availability of MRIs make it
an uncommon test for the initial evaluation of a headache. Finally, the
rare nature of cerebral venous thrombosis places it low on the
differential diagnosis of most physicians.

Case 5.8 Upholding the standard of care

• The final case from Wallenstein and Jagoda in Foresight highlight the
legal pitfalls in evaluating a patient for subarachnoid hemorrhage.16 A
38-year-old man came to the ED for a headache that had started two
hours earlier. He denied ever having a headache this severe before and
noted that the pain was intense, almost at the onset. It was associated
with nausea but no neurologic symptoms.
• Vital signs for the patient were unremarkable. A thorough physical exam-
ination, including eye and neurologic exams, was performed. Besides
photophobia, there were no significant findings. The ED physician
ordered a CT scan, which was normal. She then performed a lumbar
puncture and the cerebrospinal fluid showed no cells or xanthochromia.
The patient received some relief with a narcotic pain medicine. He was
then discharged and instructed to return if the pain worsened and to
follow up with his physician within five days.
• The patient was found in cardiac arrest three days later. He was taken
back to the same ED by ambulance. Resuscitative efforts, however, were
unsuccessful. The cause of death at autopsy was a subarachnoid
hemorrhage from a ruptured berry aneurysm. The patient’s family filed
a lawsuit against the ED physician and the hospital for failure to detect
the aneurysm before it ruptured. The plaintiff claimed that the physician
was negligent in not ordering angiography.
• Arguments from the plaintiff’s expert witness were also focused on the
reliability of the lumbar puncture. The expert claimed that he had
experienced ‘numerous’ instances where evidence of a subarachnoid
hemorrhage was not apparent on the initial lumbar puncture. The find-
ings showed up when a lumbar puncture was repeated in these patients.
The expert could not cite any sources to support his claim.
• The defense, on the other hand, was able to provide literature to counter
the plaintiff’s arguments. They were able to show data that supported
the disposition of patients with negative CT scans and lumbar
punctures. Furthermore, this data did not recommend emergent angio-
graphy. Instead, the defense showed that it was reasonable for the
continued
110 Learning from medical errors: clinical problems

physician to arrange follow-up for these patients. The jury, subsequently,

found in favor of the defendants.
• The emergency physician in this case upheld the standard of care in the
evaluation for subarachnoid hemorrhage. A statement of ‘the worst head-
ache of my life’ is classically associated with the actual hemorrhage of an
intracranial aneurysm. It is not stated in medical textbooks as the pain that
would occur with an unruptured aneurysm. Therefore, emergent angio-
graphy would not be mandatory. Do we send every patient with risk
factors for peripheral vascular disease and low back pain immediately to
the emergency department for a CT scan? The answer is no despite the fact
that abdominal aortic aneurysms are far more common than intracranial
aneurysm. Furthermore, the concept of repeat lumbar punctures to rule
out a subarachnoid hemorrhage is rare in clinical practice and certainly
does not represent the standard of care. The final caveat we would add
to the evaluation for subarachnoid hemorrhage concerns patients who
are at high risk. Specifically, these are patients who have had a prior
history or family history of intracranial aneurysm. These patients may
benefit from further testing that is arranged as soon as possible.

References
1 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 1422.
2 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 535.
3 Weiner HL and Levitt LP (1994) Neurology (5e). Williams & Wilkins, Baltimore, Mary-
land, 150.
4 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 1423.
5 Ly JQ (2003) Photo Quiz, Headache. American Family Physician. 68(12): 2434.
6 Rowland LP (2000) Merritt’s Neurology (10e). Lippincott Williams & Wilkins,
Philadelphia, Pennsylvania, 36.
7 Ferrera PC, Coluciello SA, Marx JA et al. (2001) Trauma Management: an emergency
medicine approach. Mosby, St. Louis, Missouri, 130–31.
8 (2004) Sinusitis Almost Never Presents as Headache Alone. Family Practice News. 34(7):
36.
9 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 88.
10 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 89.
11 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 90.
12 Ly JQ (2003) Photo Quiz, Headache. American Family Physician. 68(12): 2433–4.
13 Grassie C, Henry GL and Wagner MJ (2004) Foresight: Interruptions in Patient Care.
59(February): 6–7.
14 Wallenstein J and Jagoda AS (2003) Headache. Foresight: Risk Management for
Emergency Physicians. 57(June): 2–3.
15 Wallenstein J and Jagoda AS (2003) Headache. Foresight, Risk Management for
Emergency Physicians. 57(June): 3–5.
16 Wallenstein J and Jagoda AS (2003) Headache. Foresight, Risk Management for
Emergency Physicians. 57(June): 5–7.
Chapter 6

Leg pain

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Pain in the legs is a much more common complaint than pain in the arms. With
the exception of traumatic injuries, most disease processes that cause extremity
pain have a greater tendency to occur in the lower extremities. The muscular,
neurologic, and vascular organization of the lower extremities makes them more
susceptible to pathology. The greater muscle mass in the legs makes them more
susceptible to muscle spasms and heat cramps. The legs’ major function of
ambulation can be problematic when a patient becomes bedridden or
nonambulatory for an extended length of time (development of deep venous
thrombosis). Nerves that innervate and control the lower extremities are found
mainly in the lumbar area of the spine and in the lower abdomen and pelvis. This
anatomic placement makes them prone to compressive effects from herniated
disk, spinal stenosis, and spinal metastases. Furthermore, masses such as
abdominal aortic aneurysm or ovarian torsion may also cause leg pain from
compressive effects.
The greater vascular distance of the lower extremities from the heart and their
dependent positions also contribute to other causes of leg pain. Varicosities and
superficial thrombophlebitis may sometimes be painful. They can also lead to
stasis dermatitis, swelling, and cellulitis with their accompanying pain. Acute
and chronic arterial insufficiency is an important but often overlooked cause of
leg pain (please see Case 6.1). Stasis of the veins leads to deep venous thrombosis
(DVT).
There are some other nontraumatic diseases that have a predilection to cause
pain in the legs (see the ‘Differential diagnosis’ section). Traumatic injuries are
discussed in Chapter 8 of this book and not included here.

Triage
The triage of leg pain can sometimes be difficult. Most nontraumatic causes of
leg pain either do not represent true emergencies or are very occult when they
are emergencies. The conception frequently is that there are no vital organs in
112 Learning from medical errors: clinical problems

the legs and these patients are frequently sent to a step down area of care (i.e.
express care) or kept in the waiting room. This is certainly not true for leg
pain arising from spinal cord compression or from the ischemia caused by an
aortic dissection or an aortic aneurysm rupture. Even true emergencies that
arise from the legs can sometimes be very subtle, such as DVT.
As with the triage of other patients discussed in this book, any patient with
abnormal vital signs or an appearance suggesting severe distress should be seen
immediately. In addition, patients with visible deformities or deficits should be
seen expeditiously. Examples include those with leg swelling or discoloration
and those with leg weakness. Finally, patients with other associated symptoms
such as abdominal pain or neurologic symptoms need to be seen as soon as
possible.

History
Vascular disease is not an uncommon cause of leg pain in two groups of patients.
These include patients who are over 50 years of age with risk factors for peripheral
arterial disease or patients who are over 70 years of age. The Edinburgh Artery
Study showed that the incidence of intermittent claudication (the principal
symptom of peripheral arterial disease) was 2.2% in the 50–59-year-old age group
and 7.7% in the 70–74-year-old age group.1 Although venous disease such as
thrombosis is usually fairly evident from the concurrent swelling, arterial disease
such as insufficiency is often more occult. Remember that pain is one the six
classic signs of arterial insufficiency (others being pallor, paresthesias,
pulselessness, poikilothermy, and paralysis). According to Hirsch and Money,
‘peripheral arterial disease (PAD) and claudication should be suspected in any
patient complaining of ambulatory leg pain.’2
Claudication pain is frequently described as ‘aching’, ‘cramping’, ‘tightness,’
or ‘tiredness’ that occurs in leg muscles but not in joints.2 It is important to localize
where the claudication pain is because it tends to occur below the level of the
stenosis. Femoral, popliteal, or tibial arterial lesions usually produce pain in the
calf while lesions in the common femoral artery may produce pain in the thigh,
the calf, or both.3 With more proximal lesions in the aortoiliac vessels (Leriche’s
disease), there will be pain and weakness in the hip and buttocks along with
erectile dysfunction in men.3 The pain will tend to occur with ambulation and
will be relieved with rest. Therefore, ask specifically about the amount of exer-
tion before the onset of pain and the amount of rest before the resolution of the
pain. However, in cases of severe ischemia, pain occurring at rest may be present.
Is there back pain associated with the leg pain? If so, a spinal process such as
herniated nucleous pulposus or spinal stenosis may be present. Spinal stenosis
is suggested if there is pain in the legs with walking and relieved at rest. It is also
worse with extension of the back such as with walking and improves when the
patient leans forward. Symptoms of disk herniation are usually those of sciatica
(radicular pain caused by the compression of nerve roots).
Is there swelling associated with the leg pain? Cellulitis, deep venous
thrombosis, thrombophlebitis, gout, and septic arthritis may all cause leg swell-
ing. The addition of fever is more suggestive of cellulitis, thrombophlebitis, or
septic arthritis. Ask whether the symptoms are localized to the joints.
Specific questions may point to the etiology of the leg pain. Did the patient
 Leg pain 113

have recent surgery or prolonged immobilization (e.g. DVT)? Please see Chapter
9 for other risk factors for DVT. Was there recent use of alcohol to instigate a
gout attack? Are there symptoms of a sexually transmitted disease to
suggest gonococcal arthritis? Has the patient increased his exercise regimen
recently?
Past medical histories of diabetes or alcohol use are important if the pain is
felt to be a neuropathy. Patients with a prior history of DVT are at increased risk
of a recurrence. Pain is occasionally problematic for patients with subacute DVTs
as the thrombus resolves. The importance of assessing for peripheral vascular
disease and spinal disease has been discussed. Females with a history of
significant size ovarian cysts are at risk of torsion with accompanying leg pain.
For patients with a history of coronary artery disease, studies have shown a
strong correlation with concurrent peripheral vascular disease. In a study of
almost 7,000 high-risk patients screened in primary care practices, 16% were
found to have both cardiovascular disease and peripheral vascular disease.1
Cigarette smoking is a major contributor towards peripheral vascular disease,
while intravenous illicit drug use is a risk factor for septic arthritis. Social history
is important for type of occupation or recreational physical activities. Many of
the diseases discussed in this chapter have familial tendencies and family
history needs to be assessed. Finally, obtain a review of systems for
constitutional symptoms such as fever (septic arthritis, cellulitis), shortness of
breath (DVT), or weight loss (malignancy with spinal metastases).

Physical exam
All vital signs should be obtained in the same manner as in the other sections.
Vital signs may give clues to leg pain if they are abnormal. For example, a fever
may suggest cellulitis while hypotension may suggest an aortic dissection. Calf
or thigh circumferences should be measured if asymmetric swelling is present.
ankle/brachial indices (ABIs) are helpful if peripheral vascular disease is
suspected. It is defined as the higher systolic pressure of either the dorsalis pedis
or posterior tibial artery divided by the systolic pressure of the higher of the two
brachial pressures.4
Abdominal and pelvic examinations should focus on disease processes such
as abdominal aneurysm or ovarian torsion that may cause pain in the legs by
compromising the vascular supply through compression. These diseases may
also cause nerve irritation with accompanying leg pain. In addition, examination
of the inguinal areas may reveal an inguinal or a femoral hernia as the cause of
upper leg pain.
Examination of the legs should be focus on the area of pain. Look for swelling,
warmth, erythema, tenderness, fluctulance, and discoloration. Are there
varicosities present or palpable cords? Could the pain be an intra-articular
process? Is the pain in the distribution of a stocking as that typically found in
diabetic neuropathy? Pain radiating down the posterior leg may suggest
sciatica.
A thorough neurovascular examination is mandatory to assess for deficits.
Signs suggestive of arterial insufficiency include: dependent rubor, cyanosis,
loss of sensation or movement, atrophy of skin, nail changes, and lack of hair
growth. If a vascular deficit is detected distally, check for more proximal deficits
114 Learning from medical errors: clinical problems

(e.g. popliteal or femoral vessels). Neurologic testing should consist of motor,

sensory, and deep tendon reflex assessments. Focal neurologic deficits may
provide a clue to the area of the spinal cord involved. Remember to use
provocative maneuvers such as having the patient walk to assess gait,
performing straight leg raises, and squeezing the calf or dorsiflexing the foot
(Homan’s test, although this is not very sensitive).

Differential diagnosis
We discussed vascular claudication as an etiology of leg pain in the Introduc-
tion. For the most part, this vascular claudication is used to mean
atherosclerosis-induced arterial stenosis of the lower extremities. McKenna,
Boden, and DeEugenio, however, in Diagnosis and Management of Peripheral
Vascular Disease, refer to other causes of vascular claudication.5 These include
Buerger’s disease, peripheral emboli, and popliteal aneurysms.
Besides the diagnoses discussed in the ‘Introduction’ section, there are other
nontraumatic etiologies of leg pain. Two causes that have an almost exclusive
inclination for the lower extremities are diabetic neuropathy and restless leg
syndrome. In addition, Seller lists some others in Differential Diagnosis of
Common Complaints. 6 These include: muscle strain, ligamentous sprain,
degenerative joint disease, anterior compartment syndrome (shin splint), and
gout. Finally, a diagnosis of exclusion may be malingering for secondary gain
or pain medications.
Nontraumatic leg pain that is believed to be of muscular origin should raise
consideration for certain instigating events. Is the patient suffering from
myositis induced by cholesterol medications? Did the patient have a recent
prolonged immobilization state or an excessive heat exposure state to engender
an episode of rhabdomyolysis? Another frequent etiology of rhabdomyolysis is
cocaine use. Please see Case 6.2. Finally, could there be an immune-mediated
disease process present such as polymyositis?

Diagnostic tests
The choice of diagnostic test should be based on the disease of clinical suspicion.
Doppler venous ultrasound is the test of choice in our practice for venous throm-
bosis. Other testing modalities exist in the literature. Arteriogram of the lower
extremities is the definitive test if ankle/brachial indices suggest arterial
insufficiency. Magnetic resonance angiography (MRA), however, has produced
results that may be comparable to angiography and is a much safer test due to
the greater safety of the procedure. If ABIs cannot be obtained clinically, a color
flow Doppler study may be an alternative.
Swollen joints that are warm or erythematous should be aspirated and analyzed
for infection or the presence of gout or pseudogout crystals. Some physicians
use the erythrocyte sedimentation rate or the c-reactive protein as screening tests
for septic arthritis. Our opinion is that normal values do not exclude a septic
joint.
A complete blood cell count may be helpful if infection is suspected but the
test is very nonspecific and not tremendously sensitive. As discussed in ‘Pitfalls’
below, consider X-rays for nontraumatic leg pain that has been present for
 Leg pain 115

several weeks or longer. This practice has yielded the detection of occult
fractures, bone metastases, foreign bodies, and other diseases of the bone for us
through the years.
Leg pain with other constitutional symptoms will require specific testing based
on the particular symptom. Abdominal or pelvic pain may require a CT scan or
ultrasonography. MRI is the best choice for imaging of the spinal cord and the
intervertebral disks. The presence of fever with leg pain and no apparent clinical
source may require a bone scan to evaluate for osteomyelitis.

Hospital/office course
Emergent causes of leg pain need to be addressed and treated as expeditiously
as possible. Suspected cauda equina requires neurosurgical consultation. Anti-
coagulation must be started immediately for deep venous thrombosis if there
are no contraindications. Acute arterial insufficiency mandates a consultation
with the vascular surgeon. Lastly, septic joints need immediate aspiration and
the initiation of antibiotics.
Other nonemergent causes of leg pain may need specific treatment on an urgent
basis. Examples are: antibiotics for cellulitis, compression stockings for venous
stasis, colchicine or anti-inflammatories for gout, and insulin for diabetic
neuropathy. Most patients, however, will have leg pain from benign
musculoskeletal conditions or pain from unknown etiology and will simply
require pain management. Pain from an unknown etiology should be a
diagnosis made only after the exclusion of serious causes.

Pitfalls
Despite the numerous causes of nontraumatic leg pain discussed in this chapter,
the basics of musculoskeletal evaluation must be remembered. This includes X-
rays for chronic pain (greater than one month) even if there is no history of
trauma. Stress fractures from overuse and bony cancers have a tendency to occur
in the lower extremities. X-rays should always be considered before other more
sophisticated radiology tests. Please see Case 6.1 below.
We talked about potential causes of muscle pain in the ‘Differential diagnosis’
section above. These should be strongly considered for the patient with leg pain
who also complains of other muscle aches. The presence of pain elsewhere serves
as a clue for a more systemic process. Please see Case 6.2 below.
Complaints of leg pain in the pediatric population deserve special mention.
Kids are prone to disease processes that affect the lower extremity without
any significant acute trauma. These diseases tend to have a more insidious
onset and may be missed in the urgent care visit. Please see Case 6.3 below. For
that reason, we have chosen to include this topic in this chapter and not in
the chapter on musculoskeletal injuries (Chapter 8). Namely, some of these
diseases are Osgood-Schlatter, slipped capital femoral epiphysis, and Legg-Calve-
Perthes.
116 Learning from medical errors: clinical problems

Errors and interesting cases

Case 6.1 Think simple first

• A 62-year-old man was brought to our ED by ambulance with pain in the
left femur. The pain had been present for four months and there was no
history of trauma. It had worsened during the past week and he was no
longer able to bear weight on it. It had also developed some swelling. He
saw his primary care physician who ordered an arterial Doppler flow
study for arterial insufficiency because the patient was a heavy smoker.
They study was negative. She also ordered a venous Doppler study for
deep venous thrombosis which was also negative. The patient was then
arranged for physical therapy, which he did not attend because of the
pain.
• On examination, the thigh had mild swelling diffusely with no focal
bony tenderness. His distal pulses were adequate to palpation and there
was no skin discoloration. He was not able to lift the left thigh. An X-ray
of the left hip showed a comminuted intertrochanteric fracture. The patient
was admitted to the orthopedist for an open reduction internal fixation
and it was never determined the exact cause or time of the fracture.
• The primary care physician in this case missed a simple diagnosis. She
did not feel that an X-ray was needed because of the absence of trauma.
However, as we mentioned earlier in the chapter, the initial radiology
study for chronic pain is usually an X-ray regardless of a history of trauma.
The test is relatively inexpensive and should usually be obtained before
other more expensive radiology studies (with the exception of ultrasound
for suspected deep venous thrombosis). Fractures of bones may occur in
certain states that do not require trauma. Osteoporosis, cancer metastases,
and overuse are some examples.

Case 6.2 Not ‘just muscle’

• Edwards presents a case, in The M & M Files, of a young lady with leg
pain from a muscular origin.7 A 22-year-old lady came to the ED with
diffuse muscle aches that started after she was moving into another
apartment. The pain was especially worse in her legs, back, and arms.
She had no prior history of such symptoms, was only taking birth control
pills, and denied any precipitating trauma. Her vital signs were
unremarkable.
• The ED physician documented a normal physical examination but did
note whether there was any tenderness of the muscles. He ordered an
intramuscular injection of an anti-inflammatory medication and asked
for a urinalysis because the patient had a recent urinary tract infection.
There were 10–15 white blood cells and 5–10 epithelial cells with no red
blood cells on the microscopic exam. The urine dipstick, however, was
positive for blood. He then discharged the patient with the diagnosis of
‘lame muscles and UTI’ (urinary tract infection) and prescribed her
continued
 Leg pain 117

antibiotics and a pain medicine. A nurse noted that she left with an antalgic
gait.
• Another physician reviewed her chart the following day and was con-
cerned that she might have had rhabdomyolysis. When an attempt to
call her back was unsuccessful, a police officer was sent to the patient’s
home. The officer learned that the patient had visited another ED because
of worsening pain and dark urine. She was admitted at this facility for
intravenous hydration to treat rhabdomyolysis. This was brought on by
her use of cocaine and was complicated by a short course of renal failure
when her creatine kinase had reached over 20,000 IU/L. She had a full
recovery.
• The ED physician missed some potential clues in this case. The patient’s
pain developed after the physical exertion of moving. This should have
brought the possibility of rhabdomyolysis into the differential. This
diagnosis was further supported with the urine findings of blood on the
macroscopic exam but its absence on the microscopic exam (actually
myoglobin and not blood). The confirmatory test is usually a urine
myoglobin. This test, however, is usually a send out lab in most centers
and is not immediately helpful. However, a creatine kinase (CK) level
can be check as a relatively sensitive marker for rhabdomyolysis.

Case 6.3 Not just growing pains

• Selbst and Korin present a case in Preventing Malpractice Lawsuits in
Pediatric Emergency Medicine that shows how serious a relatively benign
complaint of leg pain can be in the pediatric patient.8 When a 14-year-
old boy saw his pediatrician for knee pain, he was referred to an
orthopedist. The orthopedist ordered an X-ray and an MRI of the knee,
both of which were negative.
• Four months later, the boy was brought back to his pediatrician because
of pain in his right hip and a limp. The pediatrician told him that it was
growing pains and did not order X-rays or give him a referral. When he
received further medical care six months later, he was found to have a
slipped capital femoral epiphysis (SCFE) and had immediate surgery for
femoral pinning. The delay in diagnosis resulted in articular cartilage
damage and permanent loss of range of motion.
• Although the pediatrician argued that the boy had developed the slipped
capital femoral epiphysis after his initial evaluation, the jury did not
agree and awarded the patient $1.4 million. The initial orthopedist was
dismissed from the case. The possibility remains that the boy had SCFE
when he saw the first orthopedist. After all, it is well published in
musculoskeletal literature that knee pain may be referred pain from the
hips. The case against the orthopedist, however, was less convincing
because the patient did not complain directly of hip pain, did not have
a limp, and did receive X-rays of the painful area. This case serves as a
reminder that the threshold for ordering X-rays in children should be
lower than that of adults.
118 Learning from medical errors: clinical problems

References
1 McKenna MW, Boden WE and DeEugenio D (2004) Diagnosis and Management of
Peripheral Vascular Disease. The Albert J Finestone, MD, Office for Continuing Medical
Education, Philadelphia, Pennsylvania, 3. See also Fowkes FG (1997) Epidemiology of
peripheral vascular disease. Atherosclerosis. 131 (Suppl.): 829–31.
2 Hirsch AT, Money S (2003) A Primary Care Approach to the Management of PAD and
Intermittent Claudication. Symposia Highlights for the Primary-Care Physician. Fall
edition.
3 McKenna MW, Boden WE and DeEugenio D (2004) Diagnosis and Management of Pe-
ripheral Vascular Disease. The Albert J Finestone, MD, Office for Continuing Medical
Education, Philadelphia, Pennsylvania, 4.
4 McKenna MW, Boden WE and DeEugenio D (2004) Diagnosis and Management of
Peripheral Vascular Disease. The Albert J Finestone, MD, Office for Continuing Medical
Education, Philadelphia, Pennsylvania, 6.
5 McKenna MW, Boden WE and DeEugenio D (2004) Diagnosis and Management of
Peripheral Vascular Disease. The Albert J Finestone, MD, Office for Continuing Medical
Education, Philadelphia, Pennsylvania, 5. See also Hirsch AT, Criqui MH and Treat-
Jacobson D (2001) Peripheral arterial disease detection, awareness, and treatment in
primary care. JAMA. 286: 1317–24.
6 Seller RH (2000) Differential Diagnosis of Common Complaints. WB Saunders,
Philadelphia, Pennsylvania, 255–7.
7 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 229–31.
8 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 84–5.
Chapter 7

Low back pain

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Low back pain is a common complaint during visits to both the primary care
office and the emergency department. A significant percentage of these patients
have low back pain that may be work-related with its inherent legal
implications. In fact, low back pain is the most common cause of disability among
adults less than 45 years of age.1 Most cases of low back pain represent simple
mechanical pain and require only supportive treatment and lifting instructions.
However, the physician must not take every low back pain for granted because
there are also some potential serious causes. Our friend, Jeff Abraham, MD, who
is the emergency director at Baypines Veterans Administration hospital, states
it best: ‘If you have never had a bad outcome from a low back pain, you have
not treated enough low back pains.’

Triage
Any patient with low back pain and ‘red flags’ should be seen emergently. ‘Red
flags’ for low back pain emergencies include fever (epidural abscess), new
neurologic symptoms (spinal cord compression), bowel or bladder problems
(cauda equina syndrome), abdominal pain (aortic aneurysm or dissection), and
other abnormal vital signs (hypotension with ectopic pregnancy). Other
possible warning signs or symptoms are hematuria (renal infarct), weight loss
(malignancy with bone metastases), trauma history, and vaginal bleeding
(ectopic pregnancy). Patients without any red flags but appearing in moderate
distress should be seen urgently for evaluation and pain control. Patients who
appear comfortable and have no ‘red flags’ can be triaged in the usual order of
urgency. The exceptions to this rule are the older patient (greater than 65 years
of age), the patient with vascular disease, and the possible pregnant patient who
may have more serious etiologies.
120 Learning from medical errors: clinical problems

History
A detailed history of the patient’s low back pain should be obtained in the usual
fashion as with any other pain complaint. In addition, inquire about any previous
history of low back pain and whether imaging has been performed and if there
are new symptoms. Helpful details of the patient’s back pain include trauma
history, radicular symptoms, perianal or saddle numbness/paresthesia, loss of
bowel or bladder function, or new neurologic deficits.2 Assess if the patient has
a history of cancer (spinal cord compression), blood thinner use (epidural
hematoma), peripheral vascular disease (abdominal aortic aneurysm), and
intravenous drug use or tuberculosis history (epidural abscess).
Ascertain whether there are associated symptoms with the back pain such as
abdominal pain, dysuria, hematuria, or fever. Is there a prior history of
genitourinary diseases such as kidney stones, urinary tract infections, or pyelone-
phritis? These diseases have inclinations to cause low back pain. For the female
patient, is there a history of ovarian cyst or can the patient have an ectopic
pregnancy? An occupational history is frequently necessary if there was a
precipitating injury at work. Knowledge of the patient’s occupation and duties
is mandatory in order to devise an appropriate treatment plan. Additionally,
documentation needs to be thorough because of the increased predilection for
these injuries to result in legal action.

Physical exam
As with any other complaint, we prefer to have recorded all five of the vital signs
in the patient with low back pain. However, since this complaint is so
common, vital signs are frequently not complete in these patients. We suggest
that a minimum of temperature and blood pressure recordings be made. These
two vital signs may direct us to more serious etiologies of low back pain such
as infection and vascular disease.
An abdominal exam should be performed and documented for all patients
with low back pain who have hypertension or risk factors for peripheral
vascular disease. Please see Case 7.1. Note whether there is a pulsatile mass in
the abdomen or whether there are unequal pulses in the groin. Although these
findings do not have great sensitivity in detecting abdominal aortic aneurysm
or aortic dissection, they do demonstrate that you considered these diagnoses.
Furthermore, the abdominal exam represents no additional cost to the patient
and poses no medical risks and therefore its exclusion would be inexcusable to
the courts.
A thorough neurologic exam is essential in patients with low back pain. This
includes sensation and motor testing as well as testing of reflexes. The patient’s
gait or inability to walk should also be recorded. In suspected cases of cauda
equina syndrome, a rectal exam for rectal sphincter tone should be performed.
In fact, Grudem and Schwartz recommend in Principles and Practice of Emergency
Medicine: ‘for all patients with acute low-back pain, a rectal examination for
sphincter tone should be performed on all patients with objective neurologic
abnormalities.’3
The skin of the lower back should be palpated for localized tenderness,
erythema, or swelling. Consider pelvic examination (ovarian cyst) in females or
 Low back pain 121

rectal examination (prostatitis) in males if the etiology is still in question.

Examinations of other organ systems are dictated by the clinical history.

Differential diagnosis
The majority of low back pains arise from the musculoskeletal system (e.g. strain),
the neurologic system (e.g. herniated disc), or the genitourinary system (e.g.
kidney stone, pyelonephritis). These processes usually do not represent true
emergencies. However, there are rare occasions where emergencies can arise
from these conditions. Cauda equina syndrome, kidney stones with pyelone-
phritis, and pyelonephritis with refractory vomiting are examples.
As we discussed above, some causes of low back pain can lead to death or
permanent disability if not timely detected. Do not develop the habit of taking
low back pain for granted. Every patient over 50 years of age should have the
diagnosis of abdominal aortic aneurysm considered. Patients with recent back
surgery or procedure (e.g. lumbar puncture) or intravenous drug use could have
an epidural abscess. In addition, cocaine use may lead to renal infarction and
low back pain. Patients with trauma or those on Coumadin may have epidural
hematomas.
Other causes of low back pain may originate from the genitourinary system
or the gastrointestinal system. Gastrointestinal causes of low back pain are rare
and are discussed in the ‘Pitfalls’ section below. Genitourinary causes include:
ovarian cyst or torsion, ectopic pregnancy, uterine fibroids, testicular torsion,
and prostatitis.

Diagnostic tests
Most patients presenting with low back pain will not need any testing. This is
particularly true if there is no history of significant trauma and there are no red
flags (see the ‘Pitfalls’ section below) present. Most will give you a history of
some heavy lifting or some overexertion and will need to be treated
symptomatically for a muscle strain. Another common group will have a history
of herniated nucleus pulposus, degenerative disk disease, or spinal stenosis,
and have had extensive radiology imaging in the past. As long as there are no
new neurologic symptoms, repeat testing with magnetic resonance imaging is
usually not warranted.
Patients with trauma to the lower back will usually require a cross table lateral
X-ray of the lumbar spine while still under spinal immobilization. Subsequently,
the remaining views may be obtained. If there is either a fracture seen or a
fracture not seen in the setting of a high clinical suspicion of a fracture, then a
noncontrast CT scan of the lumbar spine may be obtained. It is often difficult to
assess the amount of retropulsion and involvement of the posterior column (i.e.
the spinal cord) in the setting of a compression fracture; therefore, a CT scan is
helpful to elicit further information. MRI scanning is useful for patients with
new neurologic findings and definitely required for those with new bowel or
bladder compression and saddle anesthesia.
A special mention has to be made concerning abdominal aortic aneurysm and
low back pain. The occurrence is not as rare as people believe. In fact, a friend
of ours once told us that he had worked at a facility where every emergency
122 Learning from medical errors: clinical problems

physician there had sent a patient with low back pain home who was later
discovered to have an aneurysm. Unfortunately, only one of those was caught
before the aneurysm had ruptured. We have seen two deaths from aortic
aneurysms in patients with low back pain in the past five years. This is evidence
that, like myocardial infarctions, you will miss a diagnosis eventually, if you see
enough cases. There are simply not enough financial or technological resources
to get a CT scan or MRI for every patient with low back pain.
We offer the following advice when treating all patients over 50 years of age
with unexplained low back pain. Ask them if they have had a recent CT scan of
the abdomen or MRI of the back within the previous year. If they did and there
was no mention of an abdominal aneurysm in the report, then it is unlikely that
an aneurysm is causing their low back pain. Document all of this in your chart.
If they have not had any recent imaging, then consider getting an abdominal CT
or a lumbar MRI to exclude this. Heavily consider these tests for patients over
50 years of age with concomitant risk factors for peripheral vascular disease.
Remember that this advice does not apply to aortic dissection (see Case 7.1).
Please also see Case 7.6 for our own case of misdiagnosing low back pain.
For patients with low back pain that is unlikely to be musculoskeletal or for
those with urinary symptoms, a urinalysis is usually a good test of choice. It is
helpful for detecting urinary tract infections, hematuria, or pregnancy. A
positive test may eliminate the need for further testing. Females with vaginal
bleeding or spotting should have a urinalysis collected by catheterization to
prevent false positives.
Blood work is helpful for a selective group of patients. A coagulation panel
is useful for patients who are taking Coumadin or heparin as excessive
anticoagulation may lead to epidural or renal hematomas. A reticulocyte count
is pertinent for patients with low back pain and sickle cell disease. Finally, a
complete blood cell count may assist in the diagnosis of infections (e.g. pyelone-
phritis, epidural abscess) or hemorrhagic ovarian cysts.
Other testing for low back pain should be ordered as dictated by the clinical
symptoms. A pelvic ultrasound should be ordered if ectopic pregnancy or
ovarian pathology is suspected. CT urogram is rapidly becoming the test of choice
for the detection of ureterolithiasis. It is much more sensitive and gives much
more information on the degree of renal obstruction when compared to an
intravenous pyelogram. Finally, patients with suspected testicular torsion require
a scrotal ultrasound with Doppler color flow.

Hospital/office course
As we mentioned above, most patients will present with musculoskeletal or
neurologic back pain. Although these conditions are usually stable and not true
medical emergencies, pain management can sometimes become problematic.
During our careers, we have had to admit more patients for refractory low back
pain than any other type of pain complaint (with the possible exception of sickle
cell disease patients). In the same manner, we have seen more drug seekers present
with the complaint of low back pain than with any other complaints. Therefore
the choice of analgesic treatment used and the careful documentation of this are
extremely important in these patients.
For patients with acute emergencies, the rapidity of diagnosis and treatment
 Low back pain 123

is essential to prevent an adverse outcome. Timely conversations with consult-

ants are sometimes needed before the results of diagnostic tests are available.
Similarly, obtaining and reviewing old records may save crucial time spent
waiting for current tests (e.g. prior history of abdominal aortic aneurysm,
history of large ovarian cysts).
A significant percentage of patients who present with low back pain will leave
your office or your emergency department with a diagnosis of uncertain etiology.
These patients require explicit and strict instructions to return if the pain
worsens, new neurologic symptoms develop (including bowel or bladder
problems), fever develops, or other symptoms such as light-headedness,
abdominal pain, or vaginal bleeding occur. Have the patient return or follow up
with their primary physician in a few days if there is no improvement for further
testing (e.g. CT scan, ultrasound, etc.).

Pitfalls
Low back pain accompanied by red flags represents serious etiologies and
warrants more extensive evaluation to prove otherwise. These red flags include:
fever, abdominal pain, weight loss, bowel or bladder problems, weight loss,
urinary symptoms, and new neurologic symptoms in the lower extremities. They
may also include historical clues such as intravenous drug use, recent back
surgeries, and history of cancer, tuberculosis, or sickle cell disease (spinal
involvement). Please see Case 7.2 below.
In the preceding paragraph, we mentioned new neurologic symptoms as a red
flag that requires further evaluation. This evaluation usually means MRI of the
spine since the majority of cases will be diseases of the disks and spinal canal.
MRI is problematic for most hospitals and offices when it is unavailable.
Furthermore, many centers that have outpatient MRI are not open 24 hours a
day. Therefore, a good percentage of the patients who have new neurologic
symptoms will not receive an emergent MRI. The most important determination
for the treating physician is whether there are symptoms of cauda equina
syndrome (please see Case 7.7 below). If symptoms are present, then emergent
treatment (MRI and/or neurosurgical consult) is warranted. If not, then an MRI
done as expediently as possible is a reasonable alternative. This is usually not
that difficult because of the growing popularity of outpatient MRI centers
(including open MRI) that are competing for business.
Disk space infection (diskitis) is frequently not considered in the differential
and, subsequently, the diagnosis will be delayed. This delay may result in grave
consequences and serious complications. 4 This infection can arise by the
hematogenous route (e.g. intravenous drug users) or by direct inoculation (e.g.
post-surgical patients).5 ‘It is characterized clinically by severe back pain and on
radiographs by disk space narrowing and involvement of the adjacent vertebral
end plates. A single vertebral space is usually involved.’5 Some cases may be
associated with a preceding viral illness. However, Wesolowski and Wang point
out in The Spine that ‘radiographic and CT findings in disk space infection may
be partially obscured by post-surgical changes or severe degenerative changes.’6
Therefore, they recommend MRI as the best diagnostic test for this disease. The
difficulty in making this diagnosis is seen in Case 7.3 below.
Although extremely rare, low back pain may sometimes be the clinical
124 Learning from medical errors: clinical problems

manifestation of abdominal pathology. Remember that some abdominal

organs are retroperitoneal and, therefore, may cause low back pain. These
disease processes may be accompanied by abdominal pain but they may some-
times be occult and not present with any abdominal pain. Please see Case 7.4
below.
In evaluating traumatic injuries to the lower back, remember to review all
components of the lumbar vertebrae carefully. A commonly missed fracture is
one of the transverse process. We believe that there is a twofold reason for this
error. The first is that the physician is focused on the central component of the
vertebrae (e.g. spinous process, facets, vertebral body) and often does not notice
the peripheral placement of the transverse processes. The second reason is that
these fractures are rare and usually require great traumatic force. Therefore, some
physicians may have never seen such a fracture while others have lost their
guard to search for them. Please see Case 7.5 below.
We used to believe that there must be a significant mechanism of trauma to
create back pain of neurogenic origin. We also believed that these injuries tend
to occur in older individuals with underlying degenerative disk disease. From
clinical experience, however, we have discovered that this is a misconception
and that neurogenic back pain can result from relatively minor trauma. Please
see Case 7.8 below.
Always consider whether a patient with low back pain is on any
anticoagulants, specifically Coumadin. If they have evidence of bleeding or easy
bruising, check their blood for a coagulopathy and consider the presence of an
epidural hematoma. In the same logic that we scan every patient with a head
injury who is taking anticoagulants, we feel strongly about scanning every spine
in patients on anticoagulants with traumatic low back pain. A missed diagnosis
can result in severe and permanent injury to the patient and a huge settlement
against the physician. Please see Case 7.9 below.

Errors and interesting cases

Case 7.1 Put a hand on the belly

• An elderly man with high blood pressure presented to the emergency
department with two weeks of low back pain that started after he had
bent over to pick up a heavy object. A colleague of ours examined the
patient’s back and performed an X-ray of the lumbar spine. He was noted
to be hypertensive and the X-ray was negative. The patient was discharged
and treated for lumbar strain with pain medicines. He died a few days
later from a ruptured abdominal aortic aneurysm. The physician was
found at fault for not detecting the aneurysm and, more specifically, not
performing an abdominal exam.
• As we discussed previously, it is probably not feasible to perform an
ultrasound or CT scan in every patient with low back pain. The
prevalence of abdominal aortic aneurysm is not high enough to warrant
this. However, it must be considered in all patients with risk factors

continued
 Low back pain 125

(discussed previously). Simple and inexpensive exams and procedures

should be performed. An abdominal and vascular exam should always
be documented in these patients. Consider an abdominal or lumbosacral
X-ray to look for an aneurysm. If the physician remains in doubt, then
testing for abdominal aortic aneurysm (AAA) should be instigated.

Case 7.2 Red flags of back pain

• A young man presented to the emergency department for low back pain
without trauma on two occasions. He had negative X-rays and was treated
for lumbar strain on both occasions. A temperature of 101ºF was recorded
on the second visit. The patient then returned to the ED two days later
with complete paralysis of his lower extremities from an epidural
abscess. The etiology was from intravenous drug use. This was not
suspected on the first two visits.
• Fever is one of the red flags of low back pain. In the patient with low back
pain and fever, it is prudent to ask about illicit drug use, recent back
surgery, or history of tuberculosis (Pott’s disease). The physician
should also note if there is any erythema or warmth over the involved
area.

Case 7.3 Difficult diagnosis to make

• A 30-year-old restrained male driver came to the outpatient clinic with
low back pain after he had been rear-ended from a stop position. He
complained of severe pain in his lower back that did not radiate to his
legs and was not associated with any neurologic deficits. He also had a
low-grade temperature of 99ºF and cold symptoms for the past week.
The treating physician ordered a lumbosacral series of X-rays and read
them as ‘no acute fracture or subluxation.’ He did not comment on the
size of the disk spaces. The patient was discharged on anti-inflammatories
and muscle relaxants. The patient did not receive any improvement in
his back pain during the next three weeks. He was then sent to a
neurologist who ordered an MRI of his lumbar spine, which confirmed
the diagnosis of diskitis. In retrospect, a narrowing of the disk space was
evident in the patient’s initial X-rays.
• Diskitis is rare and the diagnosis is difficult to make. The findings on
X-rays may be extremely subtle. Be suspicious of it in patients with
elevated temperatures, recent upper respiratory infections, and localized
tenderness over the disks. The presence of any of these factors should
alert the physician to look carefully for disk space narrowing on the
X-ray.
126 Learning from medical errors: clinical problems

Case 7.4 More than pain in the back

• We had an unfortunate case in our first year of practice. An 80-year-old
female came to our ED complaining of low back pain and diarrhea. She
denied any trauma to her back or any heavy lifting. She also had no
urinary symptoms and had no abdominal pain. Her family physician
was currently treating her for Clostridium difficile colitis with metronida-
zole. The patient believed that the colitis was the cause of her diarrhea.
She commented, however, that the diarrhea was improving on the
metronidazole. She denied any fevers but had a temperature of 99.5ºF.
• The rest of her vital signs were unremarkable as well as her physical
examination. This included an abdominal exam, which revealed no ten-
derness, and a rectal exam that was negative for blood. An X-ray of the
lumbar spine and a urinalysis were both normal. A white blood cell count,
however, was elevated at 16,000 cells/microliter. There was also the
presence of a left shift with a 20% bandemia. Our impression was that
the abnormal blood work was a result of the C difficile infection.
• Since the patient stated that the diarrhea was better, we discharged her
and instructed her to continue the metronidazole as prescribed. Two days
later, she returned to the ED with diffuse abdominal pain and was seen
by another physician. He noted a surgical abdomen on physical
examination and immediately called the surgeon to evaluate the patient.
She was taken to the operating room where a segment of her colon was
removed as a result of ischemic bowel. The segment happened to lay in
the retroperitoneal and was the cause of the patient’s low back pain.
• Although the complaint of low back pain usually does not elicit
attention towards an abdominal process, caution should be taken with
diabetics and elderly patients. These groups may have atypical
presentations and the threshold for ordering imaging studies should be
lowered.

Case 7.5 Look both ways before passing

• Edwards presents a case, in The M & M Files, where a traumatic injury
to the lower back was misdiagnosed.7 After slipping and falling down
a flight of stairs, a 28-year-old man went to the ED with low back pain.
He had no radicular symptoms. The ED physician noted localized
tenderness in the mid-lumbar area. The physician then medicated the
patient with pain medicine and ordered a lumbar spine X-ray to look for
a compression fracture.
• When the films were completed, the physician examined them
scrupulously for a compression fracture at the thoracolumbar junction.
After failing to see one, he discharged the patient with pain medicine
and bed rest for a few days. The radiologist looked at the X-rays two days
later and saw that there were transverse process fractures of L2 and L3.
He relayed this finding to the ED physician on duty (not the same one
that saw the patient). The patient was instructed to return for further
continued
 Low back pain 127

evaluation. His pain had not improved and a urinalysis detected micro-
scopic hematuria. A CT scan was performed to check the integrity of the
kidneys and was found to be normal. The patient was admitted for pain
management. He filed a complaint against the treating ED physician for
missing the fractures.
• As we mentioned above, fractures of the transverse processes are rare
and require a significant force of impact. They also extend to the flank
and therefore, can be associated with injuries to the kidneys. Consequently,
evaluation of patients with these types of fractures should include
consideration for renal or other intra-abdominal organ injuries.

Case 7.6 It pays to be lucky

• This is our own story of abdominal aortic aneurysm as the cause of low
back pain. Sometimes it pays more to be lucky than to be good. This
adage was true when we saw a 55-year-old man with low back pain. He
did not have any apparent risk factors for peripheral vascular disease
and had done some heavy lifting recently. He felt that he had pulled his
back. We obtained an X-ray of the lumbar spine and did not see any
abnormalities with the spine. Subsequently, we discharged the patient
with symptomatic treatment for a lumbar strain.
• Two days later, the radiologist reviewed the X-ray and noticed that there
was abnormal dilatation of his aortic calcifications. He read this as a
probable abdominal aortic aneurysm. We called the patient and had him
come in for a CT scan, which confirmed a 5.3 cm abdominal aortic
aneurysm. He was referred to a vascular surgeon and had elective repair
of his aneurysm in time to prevent a serious complication.
• The AAA may have been an incidental finding and not related to the
patient’s low back pain. However, it was beneficial for us that a radiolo-
gist read this X-ray. Aortic aneurysms are frequently not seen on X-rays
and most physicians (excluding radiologists) rarely look for them. As a
lesson for us, we now try to include a search for them on every lumbar
spinal film.

Case 7.7 Symptoms that cannot be ignored

• Edwards presents another case, in The M & M Files, where a patient with
progressive neurologic symptoms was ignored on multiple visits and
subsequently developed a cauda equina syndrome.8 A 41-year-old man
with no history of back problems went to see his family physician for low
back pain that radiated down his right leg occasionally. He did heavy
lifting with his job but denied any other trauma. The physician gave him
pain medicine and a muscle relaxant and instructed him to undergo two
days of bed rest.
• The patient returned 10 days later and stated that he could now hardly
get out of bed. The physician ordered X-rays of the lumbosacral spine,
which were negative. He then discharged the patient with more muscle
continued
128 Learning from medical errors: clinical problems

relaxants and pain medicines. The pain became worse that night, prompt-
ing his first visit to the emergency department. Physical exam revealed
normal leg strength but also revealed decreased pinprick sensation in
the right lateral thigh. His back went into spasms with every attempt to
lift his legs. It was documented in the chart that the patient did not have
any bowel or bladder dysfunction and his sphincter tone was normal.
He was given stronger pain medicine and instructed to see his primary
doctor.
• Two days later, when the patient saw his primary physician, his leg pain
was so severe that he required a cane to ambulate. In the waiting room,
he became pale and had a bout of near syncope. An ambulance was called
to take the patient to the emergency department. On this evaluation,
chart documentation showed that the patient denied any bowel or
bladder problems but complained of numbness in his right buttocks that
extended to the groin area. The patient was given a shot of pain medicine
and discharged after some relief.
• Later that night, the pain became even worse and was now accompanied
by the onset of urinary incontinence. A third ED physician evaluated the
patient this time and noted that the patient had distention over his
bladder and a positive straight leg raise on the right. No further evidence
of a neurologic exam or a rectal exam was found in the chart. The
physician ordered catheterization drainage of the bladder and then
discharged the patient with the diagnosis of ‘urinary retention
secondary to cyclobenzaprine.’ He was instructed to follow up with his
family physician.
• On follow-up with his physician, an MRI was ordered, which showed
a large central disc herniation at the L5-S1 level. He was admitted by the
neurosurgeon for cauda equina syndrome and taken to surgery for
removal of the extruded disc segment and decompression of the neural
canal. As a complication of the delay, the patient developed neurogenic
bladder and sexual dysfunction and filed a lawsuit against the hospital
physicians.
• On each of the three emergency department visits, red flags were missed
or ignored by the ED physicians. Numerous errors were committed
including an incomplete evaluation of new neurologic symptoms, ignor-
ing worsening pain, failure to perform a neurologic exam (including one
for rectal tone), and incorrectly attributing autonomic dysfunction to a
medicine. The primary physician could also be held accountable for the
same errors. However, because the primary physician was the one who
ordered the MRI and because he is less likely to see a cauda equina
syndrome in his practice, the public’s perception (and the jury’s) is that
he is less accountable for the misdiagnosis.
• As we discussed above, diseases of the lumbar/sacral spine are frequently
not correctly diagnosed on the initial visit. This is largely due to the
difficulty of obtaining emergent MRI scans. However, this patient
certainly had reasons to raise the physician’s suspicion that a neurologic

continued
 Low back pain 129

process was ongoing and necessitated more prompt imaging of his back.
Ignoring the worsening of his symptoms and masking them with pain
medicines led to the delay in diagnosis.

Case 7.8 Is that all that happened?

• We can recall two cases of neurogenic low back pain from relatively benign
mechanisms. A fellow resident of ours in his late twenties was jogging
along the sidewalk when he stepped in a small ditch in the pavement.
He developed low back pain from this incident and did not make much
of it until he developed a foot drop. A subsequent MRI showed a large
disk herniation and he underwent back surgery.
• A 32-year-old nurse at our hospital who had no prior history of back
problems came to the ED with low back pain. The pain started immedi-
ately after she had bent over to pick up some clothes. She had weakness
and pain in her left leg and had numbness in her left fourth and fifth toes.
She did not have any bowel or bladder problems and had normal rectal
tone. An emergent MRI of the lumbosacral spine showed a large disk
herniation at L5-S1. The patient had emergent surgery and did well
subsequently.
• These two cases remind us that neurogenic low back pain may result
from minor trauma. The severity of spinal injury and pain is certainly
out of proportion to the actual trauma. Therefore, we must perform a
detailed neurologic examination and look for objective evidence of
neurogenic injury in every patient with low back pain.

Case 7.9 Is the Coumadin patient bleeding?

• Starr presents a case, in Cortlandt Forum, of low back pain that represents
one of an emergency physician’s worst nightmares.9 A 78-year-old woman
with a history of atrial fibrillation and Coumadin use saw her cardiolo-
gist. Although he performed an echocardiogram, which showed an atrial
thrombus, he was distracted before he could order a prothrombin time
(PT).
• Two weeks later, she telephoned him to tell him that she was having easy
bruising and nosebleeds. He told her to come to the office the following
day, at which time he ordered a PT. The patient, however, was distracted
this time and never got it done. A day later when the cardiologist was
out of town and his partner was covering for him, the patient called
again to say that she had a large bruise on her forehead. The covering
physician instructed her to stop the Coumadin and get a PT drawn the
next day. The PT was prolonged at 35 seconds, so he gave her further
instructions to go to the emergency department for a Vitamin K injection.
She received her shot in the ED and had another PT drawn, which was
30 seconds.

continued
130 Learning from medical errors: clinical problems

• Not long after returning home, she developed severe low back pain, which
prompted her to return to the ED. She was examined by the ED physician
and given a shot of Toradol and discharged with pain medicines. When
she awoke the next day, she was a paraplegic. A CT scan in the ED showed
a large epidural hematoma compressing her spinal cord at the second
lumbar vertebrae. Her neurologic injury did not improve with surgical
decompression.
• The patient filed a lawsuit a few weeks later against the cardiologist, his
partner, and the ED physicians. Although the first cardiologist had failed
to check the patient’s PT at the initial visit, he was not found at trial to
be negligent. This was partly due to the plaintiff expert witness who
faulted the other physicians for failing on multiple opportunities to correct
the patient’s coagulopathy. The other defendants, consequently, were
held responsible for a jury award of $12.2 million with an additional
$3 million in interest. The jury, in this case, did not consider that the
patient herself missed a golden opportunity to have her PT checked before
the development of the low back pain.
• We stressed earlier about using caution in patients taking anticoagulants
with traumatic low back pain. The patient in this case, however, had no
history of low back trauma. Instead, she had a coagulopathy along with
clinical manifestations of it (bruising, nosebleeds). Hence there was
evidence of spontaneous bleeding and a history of trauma is not required
to raise the suspicion of an epidural hematoma.

References
1 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 483.
2 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 485.
3 Schwartz GR (1999) Principles and Practice of Emergency Medicine (4e). Lippincott
Williams & Wilkins, Philadelphia, Pennsylvania, 1291.
4 Quiles M, Marchisello J and Tsairis P (1978) Lumbar adhesive arachnoiditis. Spine. 3:
45–50.
5 Mandell GL, Bennett JE and Dolin R (2000) Mandell, Douglas, and Bennett’s Principles
and Practice of Infectious Diseases (5e), Vol 1. Churchill Livingstone, New York, New
York, 1194.
6 Herkowitz HN, Garfin SR, Balderston RA et al. (1999) Rothman-Simeone, The Spine (4e),
Vol 1. WB Saunders, Philadelphia, Pennsylvania, 490. Reproduced with permission
from Elsevier.
7 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 151–2.
8 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 160–62.
9 Starr DS (2004) Have a patient on warfarin? You’d better read this. Cortlandt Forum.
17(3): 94–5.
Chapter 8

Musculoskeletal injuries

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Injuries to the musculoskeletal system are common to the emergency depart-

ment, urgent care offices, and primary care offices. They may involve muscle,
bone, joints, or associated ligaments or tendons. The injuries may be acute as in
the case of trauma or may be chronic as in the case of overuse or stress injuries.
Although there are many causes of insults to the musculoskeletal system such
as cocaine use, medicines (e.g. ‘statins’ for cholesterol), and dehydration, we will
limit the discussion in this chapter to the discussion of traumatic
musculoskeletal injuries.

Triage
Most musculoskeletal injuries can be triaged as urgent or non-emergent. There
are certain groups, however, who require emergent care. These include:
amputations, injuries with extensive or arterial bleeding, injuries with heavily
contaminated wounds, and injuries with signs or symptoms of neurovascular
compromise. In addition, patients with abnormal vital signs such as fever,
hypotension, or tachycardia require attention as soon as possible.
Musculoskeletal injuries that involve breakage of the skin should be triaged
as open fractures or open joint injuries until proven otherwise. The time before
wound cleansing and antibiotic treatment is critical in preventing permanent
deformity. Therefore, these injuries must be seen and addressed emergently.
Musculoskeletal injuries that do not involve a breakage of the skin can be
divided into three categories. Those that do not involve any neurovascular
compromise or significant deformity represent the majority of these injuries.
These patients can usually be seen on an urgent basis. The second group
involves injuries where there is no compromise of the neurovascular status but
there is significant deformity. Deformity could consist of a fracture with
displacement or angulation, or it could represent a soft tissue injury with
significant swelling. Because there is significant disruption of the normal
132 Learning from medical errors: clinical problems

musculoskeletal structures, there is the potential for the rapid development of

neurovascular compromise. Therefore, these patients should be seen as soon as
possible and stabilized (e.g. reduction of dislocation, fasciotomy for compart-
ment syndrome) if necessary. Finally, the last group is the one with obvious
neurologic or vascular deficits. This group will likely have permanent loss or
injury to the limb if not treated immediately. These patients need expedient
stabilization and specialist consultation.

History
Ask the patient (or those who witnessed the incident) about the details of the
injury. The mechanism of injury is important to assess for expected and potential
injuries. In addition, inquire about the patient’s function of the involved part
since the injury. Significant injuries often impair normal musculoskeletal
functioning. The surroundings of the injury are important to assess for possible
foreign bodies and wound contamination (e.g. broken glass, dirty environment).
The time of injury is important for deciding the most appropriate wound care.
How the injury has been handled by the paramedics or by other prehospital
personnel is also important (e.g. splinting, wrapping, irrigation). The
immunization history and the possibility of intentional injury should be
addressed. Finally, check with your local laws because some injuries (e.g.
domestic abuse, assaults) require mandatory reporting to the local law enforce-
ment agencies.
Ask the patient about neurologic or vascular symptoms that may have occurred
from the injury. Any new numbness, weakness, limitations of motion, or
discoloration could be clues to a neurovascular compromise. The physician
should also inquire about associated injuries that may have been masked by the
main injury. Injuries proximal and distal to the focal point are not uncommon
and it is generally recommended to extend the evaluation to one joint above and
one joint below the involved site. Please see Case 8.3. Symptoms of neuro-
vascular compromise are extremely important with certain injuries because they
can produce deceptively normal exterior appearances. These injuries include
electrical injuries and high compression puncture wounds.
Assess for potential problems with wound care. Is the patient allergic to any
local anesthetic? Is the patient on a blood thinner? Does the patient have diabetes
or another immunosuppressive state? Could the patient’s current medications
interact with the pain medicine that will be prescribed? Does the patient smoke
cigarettes or have a history of peripheral vascular disease or poor wound
healing? Has the patient had problems with wounds becoming keloids in the
past?

Physical exam
Vital signs are often incomplete or not taken in patients with musculoskeletal
injuries. This is because of the conception that musculoskeletal injuries are local-
ized processes. For the most part, this misconception is true. However, because
of the potential of neurovascular compromise, a baseline set of vital signs is
extremely important to document. We also do not recommend the omission of
vital signs because any of them may be a marker of physical stress on the body
 Musculoskeletal injuries 133

and a more serious underlying injury. An extended vital sign that may be helpful
for certain situations is the ankle/brachial index (ABI). This is the ankle systolic
pressure divided by the brachial systolic pressure on the same side of the body.
The ratio then serves as an indicator for peripheral arterial disease in the
specified lower extremity.
All injuries should be examined for swelling, deformity, breakage of skin,
crepitance, bruising, and associated injuries. The characteristics of the injury
should be consistent with the mechanism and surroundings of the injury. If not,
the physician should become suspicious of the history. Use descriptive and
quantifiable terms if possible in your notes to avoid any confusion when some-
one else reviews your chart. Examples would include circumferential
measurements of swelling, degrees of angulation, etc.
Examine all musculoskeletal injuries closely for associated injuries to the deeper
structures. Are there injuries to nerves, tendons, vessels, etc? It is always
important to determine if there is exposed bone or impending exposed bone. If
exposure is present, then treatment for an open fracture must be undertaken.
Finally, all examinations of musculoskeletal injuries must be accompanied by
a neurovascular examination. Peripheral pulses and circulation must be
evaluated for compromise. Range of motion and muscle strength must be
determined for limitations. These exams must be repeated after any physical
manipulations such as reductions or splinting. Serial exams also help monitor
for the development of compartment syndrome.

Differential diagnosis
The differential diagnosis of musculoskeletal injuries is comprised of its
constituent anatomical parts. Consideration must be given to injuries to the skin,
muscles, bones, tendons, ligaments, nerves, or vessels. In addition, involvement
of areas such as joint spaces, bursas, or muscle compartments must be
determined in order to direct the proper therapy. Specific injury patterns or
associations should be expected or recognized. Examples of these are as follows.

1 Fractures with overlying skin violation should be assumed to be open

fractures.
2 Anterior shoulder dislocations with associated axillary nerve injuries occur
in approximately 12% of cases.1
3 Knee dislocations with associated popliteal artery injuries occur in approxi-
mately 33% of cases.2
4 Distal humeral fractures are frequently associated with neurovascular
injuries, typically the median nerve and brachial artery.3
5 Tenderness in the anatomical snuff box should be treated as a scaphoid
fracture regardless of the initial X-ray.
6 A posterior fat pad seen on a lateral elbow X-ray represents a fracture of the
radial head or proximal ulna in almost all cases.4
7 Because calcaneal fractures tend to occur in falls, there is a 10% association
of calcaneal fractures with compression fractures of the dorsolumbar
spine.5
8 Fractures of the pelvis or the femur may be associated with significant blood
loss.
134 Learning from medical errors: clinical problems

Musculoskeletal trauma to the anterior or posterior trunks may involve injuries

to the ribs, spine, sternum, clavicles, pelvis or underlying organs. There may be
significant injuries to internal organs such as hematomas, lacerations, or
perforation. These injuries may occur in the presence of a relatively normal
external appearance. Rib fractures may be associated with a pneumothorax,
sternal fracture may cause a myocardial contusion, and scapular or first rib injuries
may also injure the aorta.6 In addition, lower rib fractures may cause lacerations
to the liver or spleen. Finally, pelvic fractures are frequently associated with
injuries to the bladder.

Diagnostic tests
The standard test for evaluating most injuries to the musculoskeletal system is
the X-ray. X-rays are not only helpful for the determination of bony injuries, they
may also reveal foreign bodies and occasionally provide clues to other internal
injuries. These include: pneumothorax, widened mediastinum, free air under
the diaphragm, etc. As we mentioned briefly above, X-rays do have limitations
with bony injuries, however. Please see Case 8.5. Fractures of the scaphoid bone
or Salter Harris 1 fractures in children are commonly not apparent on X-rays.
Other imaging modalities such as CT scan, bone scans, or magnetic resonance
imaging (MRI) are reserved for occult fractures when the clinical suspicion is
high. They may also be used to further define a fracture such as a CT scan to
determine the extent of a pelvic fracture and when X-rays are difficult to perform
such as a cervical spine in a person with extremely broad shoulders. MRI is the
testing modality of choice for the spinal cord and the intervertebral disks. It is
also the imaging modality of choice to determine the extent of injury to the soft
tissues.
There are some legal aspects of ordering X-rays that are worth mentioning.
The possibility of pregnancy in a female of child-bearing ability must always be
considered. This, of course, also applies to CT scans. MRI has generally been
regarded as being safe in pregnancy. Please see Case 8.2 below. We remember
having to order an MRI once for a pregnant female who fell off a horse on her
right hip and could not bear weight on that hip. There are only a few standard
guidelines for ordering X-rays and this leaves a gap for potential medical
liability for missed fractures. Physicians should become familiar with the few
published guidelines for ordering X-rays such as the Ottawa Rules for the ankle
and the knee and the Pittsburgh Rules for the knee. We have found that three
important factors to consider in the ordering of X-rays are: the mechanism of
injury, the functional impairment to the patient, and the general health status of
the patient (the very young and the elderly deserve liberal consideration of
X-rays). We feel that X-rays should ideally be ordered before any attempt at
reducing a dislocation. The only exception is the dislocation requiring prompt
reduction to relieve a neurovascular compromise. We have seen fractures
associated with the dislocation seen on post-reduction films that were
subsequently blamed on the physician’s reduction technique. Without a pre-
reduction X-ray, there is no way to prove in court that the fracture had been there
before the manipulation. Finally, make sure that you have the adequate views
of an X-ray to make a treatment decision. Because some injuries will deter a
patient from moving into certain positions for an X-ray, it is not uncommon to
 Musculoskeletal injuries 135

receive a suboptimal amount of views or a suboptimal positioning for a

particular view. Again, we have seen missed fractures stemming from these
conditions.
In injuries with the possibility of neurovascular compromise, other diagnostic
tests may be ordered. Color Doppler studies and arteriograms are useful for the
determination of vascular injury. Ultrasound is helpful for assessing venous
stasis and thrombosis. Magnetic resonance angiography is becoming
increasingly popular as a potential replacement for angiograms. They are
particularly useful for traumatic neck injuries to rule out carotid dissection.
Finally, compartment pressures can be performed with pressure monitors.
Musculoskeletal injuries to the anterior and posterior trunk may require
additional testing. Examples would be a urinalysis for injuries to the flank or
pelvis, an electrocardiogram for injuries to the chest, and a complete blood cell
count for injuries to the left upper quadrant. Use the mechanism of injury and
the location of injury to guide your diagnostic testing.

Hospital/office course
Musculoskeletal injuries may require therapy ranging from symptomatic treat-
ment to emergent surgery to salvage a limb. A good number of lawsuits involving
musculoskeletal injuries arise from the subsequent finding that a significant
injury was missed on the initial exam. Also contributing to this is the fact that
these injuries sometimes will have delayed presentations and may become worse
if the patient is not given or does not follow the appropriate discharge
instructions. Therefore, it is important to have extremely thorough
documentation in your evaluation.
Document the initial exam and follow-up exams after any therapeutic
manipulations. Neurovascular status and functional impairment are the two
most important clinical assessments to note. This helps prevent accusations later
that injuries were a result of the physician’s manipulation. Please see Case 8.6.
If X-rays are not ordered, make a short justification in the chart on why (e.g. fell
on hip three days ago, has been ambulating without problems). All injuries should
be given a follow-up in case there is worsening, or new symptoms, or no
improvement. If X-rays are questionable, treat conservatively (e.g. splint) and
arrange for a specialist (e.g. radiologist, orthopedist) to over-read the film as
soon as possible. Similarly, if there is a discrepancy between your interpretation
of X-rays and that of the official report, inform the patient as soon as possible
and make the necessary adjustments to the therapy. Do not let the patient
discover the discrepancy from his primary care physician weeks or months later.
For injuries that require emergent treatment, do your best to treat the patient
as soon as possible. One of the biggest contributors to prolonged times in the ED
is the wait for X-rays to be completed. This is largely due to the large amount
of films that ED patients require. It is the treating physician’s responsibility to
inform the X-ray technician which films need to performed on a stat basis. There
are often times when the consulting orthopedist will be delayed in reaching to
the patient (e.g. physician is operating on another patient). Therefore it is import-
ant for emergency physicians to become familiar with some of the techniques
of dislocation and fracture reduction. Stable patients who are waiting to see the
orthopedist should be checked frequently to assess neurovascular status.
136 Learning from medical errors: clinical problems

The final discussion is the conservative treatment of certain injuries. With these
injuries, conservative treatment is necessary in order to prevent chronic pain or
disability, both of which are popular causes of medical malpractice cases. The
patient with a hand injury, for example, requires a thumb spica splint if there is
suspicion of a scaphoid fracture or a ligamentous injury to the thumb. These
types of injuries are not appropriately treated (although occasionally done) with
wrist splints and thumb splints, respectively. Please see Case 8.1. The adolescent
with pain over the tibial tubercle must avoid activities that reproduce pain despite
negative X-rays, in order to prevent chronic pain. Finally, the patient with a Jones
fracture needs a short leg cast for at least six weeks because of the increased risk
of malunion.

Pitfalls
Thumb injuries require a thorough examination and conservative therapy because
of the thumb’s integral function in almost all of the hand’s functions. Incorrect
evaluation or treatment can lead to severe disability and functional impairment.
Hence, there are specific splinting for thumb injuries (i.e. thumb spica) and
separate criteria for reimplantation of thumb amputations (versus the amputa-
tion of other fingers). There are also names given for a rupture of the ulnar
collateral ligament because of its frequent association with certain sports (e.g.
gamekeeper’s thumb, skier’s thumb). Please see Case 8.1 below.
Musculoskeletal injuries in young children can sometimes present with many
diagnostic challenges. They may not be able to verbalize (or verbalize poorly)
the mechanism of injury or the exact site of injury. Please see Case 8.4 below.
Physical testing of their neurovascular functioning is, likewise, limited due to
their narrow range of comprehension. Finally, intentional injuries of child abuse
should be considered.
Almost all musculoskeletal injuries of the extremities have the potential for
the development of compartment syndrome. Therefore, it is prudent to make it
clear in the discharge instructions when the patient should return. It may be
helpful to have a handout of the signs and symptoms of compartment syndrome
to give to the patient. Instruct the patient to read it and return immediately if
they notice any of the signs or symptoms. Also, instruct your nurses that any
callbacks from patients with these types of injuries who are doing worse should
be given to the physician or the patient should be instructed to return
immediately. From our experiences, nurses are not too familiar with
this medical complication and will tell a patient just to ride the pain out or
increase his pain medicine because the X-rays were negative and there cannot
be anything serious going on. We give a clinical example of this in Case 8.7
below.
Be aware that the treatment of musculoskeletal injuries may sometimes cause
iatrogenic injuries. Splints and casts may be placed too tightly and contribute
to compartment syndrome. They may also be placed over sensitive areas and
lead to pressure ulcers. In addition, pressure sores may result when a patient is
left on a spinal immobilization board. Most of the complications will occur after
the patient has been discharged so the physician has to be vigilant when the
patient calls back or returns due to increasing pain. The pain should not be
assumed to be from the primary injury. Please see Cases 8.8 and 8.9 below.
 Musculoskeletal injuries 137

Errors and interesting cases

Case 8.1 Thumbs down

• Edwards presents a case, in The M & M Files, where a misdiagnosis of a
thumb injury led to functional impairment.7 After jamming his right
thumb during a football game, a 20-year-old man presented to the ED
the following day. The ED physician did not document a functional
evaluation of the thumb. Instead, he ordered an X-ray, which was
negative, and discharged the patient with a splint of the thumb. He also
gave the patient instructions to keep the splint on for three days and
follow up with his family physician if he was not better.
• Five days after the patient removed the splint as instructed, he saw his
family physician because there was no improvement in the pain or the
swelling. He was then referred to an orthopedist, who felt that he had
ruptured his ulnar collateral ligament (also known as ‘gamekeeper’s
thumb’). The orthopedist felt that the thumb should have been splinted
more protectively and that the patient should have received a more
prompt referral. Although the patient had more conservative splinting
of the thumb for three weeks and then engaged in extensive physical
therapy, he was never able to function optimally as a cabinetmaker. The
case was settled without going to court because of the ED physician’s
poor documentation.
• Thumb injuries require a thorough search for functional deficits. They
also usually require more conservative treatment than those of the other
fingers. A simple finger splint is often insufficient for immobilizing the
thumb. A thumb spica splint provides improved immobilization.
Misdiagnoses and mistreatment of thumb injuries may lead to severe
and permanent impairments and large malpractice settlements.

Case 8.2 Love yet known but lost

• A friend of ours who works in our emergency department had a
traumatic evaluation that she will never forget. She was in her mid-
thirties when she was in a serious car accident. When she was taken to
the emergency department, she received numerous X-rays (chest, pelvis,
cervical spine) and a CT scan of her abdomen. She did not know at the
time that she was pregnant but it was discovered with her CT scan.
Although none of her injuries proved to be serious, she did lose her
pregnancy due to the radiation to which she was exposed. She did not
become pregnant again during her child-bearing years.
• If the patient is hemodynamically stable and has the potential to be preg-
nant, it is probably wise to hold radiology studies until the pregnancy
status is confirmed. The ideal situation is to have point of care pregnancy
tests in the emergency department. Results can then be obtained within
minutes. If the patient is not stable or the pregnancy test results take a long
time to get back (i.e. greater than 30 minutes), the physician has to
consider the patient’s risk/benefit ratio in getting radiology studies.
138 Learning from medical errors: clinical problems

Case 8.3 Look above and below

• In another case by Edwards in The M & M Files an injury was missed with
a limited exam.8 A 15-year-old boy injured his leg and was brought to the
emergency department. Although he could not describe the mechanism
of injury, he had tenderness and swelling over the distal tibia. The ED
physician ordered an X-ray that showed a nondisplaced medial
malleolar fracture. The patient was splinted with a short ankle splint and
discharged.
• When the patient saw the orthopedist two days later, he was also found
to have a proximal fibular fracture. The ED physician missed this
fracture because he had not followed the dictum of evaluating the joint
above and below an injury. He also failed to recognize a common injury
pattern where medial malleolar fractures are occasionally associated with
proximal fibular fractures.8 As a consequence, the patient was not
adequately splinted for his injury.

Case 8.4 Kids need pictures

• This case by Edwards in The M & M Files stresses the difficulty of evalu-
ating musculoskeletal injuries in young children.9 Because the history
and physical exam can sometimes be inaccurate in this population, a
more liberal approach towards diagnostic imaging is often warranted.
A 2½-year-old girl was brought to the emergency department after
colliding with her dog. When asked where she was hurting, the girl
pointed to her right chest. The physician palpated and auscultated her
chest and did not notice any significant findings. He also did a thorough
examination of the patient’s neck and right shoulder. The patient had
some minor discomfort on movement of her right shoulder. This
discomfort made the patient upset so the ED physician concluded his
examination and told the child’s parents that nothing was seriously wrong
and recommended Tylenol for pain.
• The parents were not satisfied with this evaluation and took the child to
her family doctor the following day. An X-ray in the office showed a right
clavicular fracture. This led to a parental complaint to the emergency
department. The complaint could have been avoided if the ED physician
had realized the limitations of the age of his patient. Small children are
not reliable at describing or localizing their injuries. Therefore, the slightest
suspicion of pain or discomfort usually warrants an X-ray or an
explanation to the parents of why one is not done.

Case 8.5 X-rays do lie

• Edwards presents another case, in The M & M Files, of a missed fracture
on the basis of a negative X-ray.10 A 43-year-old man with mental
retardation fell out of bed on his left hip and was taken by ambulance
continued
 Musculoskeletal injuries 139

to the emergency department. He stated that the pain was too severe to
stand but he was able to bear weight with some assistance. In the ED, he
was able to get up and use the commode. Physical exam revealed some
generalized tenderness of the left hip but there was no limitation of
motion. The radiologist read his hip and pelvis X-ray as normal. The ED
physician helped the patient walk around the room before discharging
him.
• He returned five days later with no improvement in his symptoms. A
different ED physician noted similar findings on physical exam and
reviewed the X-rays himself. He then prescribed pain medicines and
instructed the patient to continue using the walker and follow up with
his primary care doctor. The patient collapsed while walking to the
bathroom that night. He returned to the ED where X-rays clearly showed
a femoral neck fracture.
• In Edwards’ discussion of the case, he notes that occult hip fractures are
not uncommon. They tend to occur in the elderly, however, where
degenerative disease and osteoporosis make the fracture difficult to
discern. When suspected, especially in patients with return visits, a more
sensitive test should be ordered. Computed Tomography scan, bone scan,
and MRI are diagnostic modalities with greater sensitivity for detecting
fractures.

Case 8.6 No injuries are minor

• Musculoskeletal injuries sometimes occur in association with many other
injuries due to their traumatic nature. In the setting of multiple injuries,
evaluation and documentation of less important injuries tend to be less
complete than in the more severe injuries. This practice must be avoided
because the injuries that may appear as less important are just as likely
(or more likely if not properly treated and charted) to result in a
malpractice claim. Edwards presents a case, in The M & M Files, where
the ED physician has his priorities correct but his smallest concern
resulted in a malpractice claim.12
• After a rollover accident in his truck while intoxicated, a 41-year-old
man walked into the ED seeking treatment. He was a restrained driver
and suffered a large laceration on the forehead and multiple facial
abrasions but denied any loss of consciousness or neck pain. He also
complained of pain in his left ankle, left chest and abdominal pain, and
right index finger pain and deformity. The ED physician decided to
perform an extensive evaluation based on the patient’s multiple organ
involvement and based on his alcohol intoxication.
• Computed Tomography scans were ordered for the head and the
abdomen, both of which were negative. Similarly, X-rays of the chest and
left ankle were also negative. X-rays of the right index finger, however,
showed a fracture-dislocation of the middle phalanx. After performing
a digital nerve block, the ED physician reduced the dislocation and
continued
140 Learning from medical errors: clinical problems

splinted the finger without post-reduction films. The patient’s blood work
was unremarkable and he was observed during the night in the ED. The
following morning, he was discharged with a family member and
instructed to follow up with his own physician.
• When the patient finally saw his physician three weeks later, the finger
was now locked in flexion. He filed a lawsuit against the ED physician
on the grounds that the reduction was not performed appropriately and
that the follow-up instructions were not clear. Although the ED
physician’s treatment was probably defendable, his insurance carrier
decided to settle the case. The physician took the extra time and effort
to ensure that the patient did not have any serious injuries. He forgot
however, that minor injuries also require meticulous care. He did not
order a post-reduction X-ray, which has become standard practice mainly
for legal reasons, as demonstrated by this case. In addition, he did not
document a post-reduction neurovascular exam and was very brief with
the details of the procedure – how the finger was splinted, and the
instructions that he gave the patient.

Case 8.7 It is not broke, but fix it

• Selbst and Korin present a case, in Preventing Malpractice Lawsuits in
Pediatric Emergency Medicine, where the injury treatment was nearly flaw-
less but nevertheless resulted in an out of court settlement. 12 A
nine-year-old boy was brought to the ED after falling off a fence and
hitting his leg on a rock. The boy was being chased by a dog but did not
receive any direct injury from the dog. Instead, he had a 3 cm laceration
on his left shin that extended to the superficial fascia, caused by the contact
with the rock. The ED physician examined the boy, closed his wound
with a three-layer closure, and ordered an X-ray of the boy’s leg, which
was negative. The boy’s mother was given instructions to return in nine
days for suture removal and to return ‘promptly’ for severe pain or any
other problems. Antibiotics were not prescribed for the child.
• The boy’s pain grew worse and he started to develop an elevated
temperature. The mother called the ED twice to report this and was told
by the triage nurse to give Tylenol. When the Tylenol did not help the
child’s pain, the mother brought him back to the ED 48 hours later. There
was now the presence of compartment syndrome in his left leg and he
was taken to the operating room for fasciotomies, debridement, and
muscle repair. He later developed a foot drop and required a second
operation for more debridement and skin grafting.
• The family sued the ED physician and the hospital for the delayed
diagnosis. An orthopedist examined the child two years later and found
a scar, subjective without objective leg weakness, and only a ‘5%
disability.’ Despite these minimal findings, an expert witness argued that
the child had suffered a ‘significant loss of function of the leg’ and the
case was eventually settled out of court. On review of the case, there was
continued
 Musculoskeletal injuries 141

not much more that the ED physician could have done for this child. He
treated the wound appropriately and gave the proper instructions about
when to return. In retrospect, it may have helped if he would have taken
an extra effort to make sure that the mother understood the possibility
of the development of a compartment syndrome (a handout works
extremely well). More specific instructions are: to return if increasing
pain, swelling, numbness, discoloration, or problems with motion. Triage
nurses also need to be reminded that patients who call with worsening
conditions need to be told to come back or have their call brought to the
physician’s attention.

Case 8.8 Try to do no harm

• The next case from Selbst and Korin’s Preventing Malpractice Lawsuits in
Pediatric Emergency Medicine shows the severity of an iatrogenic injury
from the treatment of a fracture.13 After an 11-year-old girl fell off a
bicycle, X-rays in the clinic revealed a metatarsal fracture of the right
foot. She was placed in a posterior splint that was wrapped with a tight
bandage. Anti-inflammatories were also given. Her pain became worse
when she returned home, prompting her parents to call the clinic. They
were told to give her acetaminophen. Several more calls were made to
the clinic the following day and a different physician prescribed narcotic
analgesics. A day later, there was greenish-gray discoloration of the child’s
foot. She had developed full thickness skin necrosis of the dorsum of her
foot. This required reconstructive surgery and multiple skin grafts. A
lawsuit was filed against the treating physician and the clinic employees
who gave phone advice to the child’s parents. Although the surgeries
were successful, the verdict was in favor of the plaintiff with an award
of $473,355.
• The duties of casting and splinting in medical offices or emergency
departments are often delegated to medical assistants and technicians.
The level of clinical expertise is extremely variable. Therefore, iatrogenic
complications should be considered when a patient’s symptoms have
become worse.

Case 8.9 (Hand)le with care

• Gardner and Mendelson present a case in Foresight of an iatrogenic
injury arising from an inappropriate treatment.14 The physician in this
case made the correct diagnosis and initial treatment of a musculo-
skeletal problem. His treatment before discharge, however, led to a
severe complication. A 43-year-old man came to the ED with pain in his
right middle finger sustained while rock climbing. The ED physician
noticed an obvious deformity of the patient’s proximal interphalangeal
joint. This was confirmed on X-ray as a dorsal dislocation of the joint.
continued
142 Learning from medical errors: clinical problems

The physician performed a digital block and reduced the dislocation.

Confirmation of the successful reduction was obtained with a repeat X-
ray. His orders on the chart were to ‘discharge patient with a splint.’ He
did not specify how he wanted the finger splinted. The nurse splinted
the joint in full extension and instructed the patient to follow up with the
orthopedist in one to two weeks.
• The patient was seen by the orthopedist 10 days later. He had significant
pain and decreased flexion at the joint. The orthopedist consulted with
and referred the patient to a hand surgeon. The hand surgeon performed
surgery to repair a volar plate injury. The functioning of the finger was
improved with the surgery but the patient continued to have pain and
swelling after strenuous activity. He filed a lawsuit against the ED
physician and the hand surgeon and contended that he could not climb
rocks any longer. The hand surgeon was eventually dropped from the
case and the ED physician decided to settle.
• The defense for the ED physician did not feel that he had a good case.
This is despite the fact that he made the correct diagnosis and initial
treatment. The finger was not splinted appropriately and the physician’s
poor documentation made the case difficult to defend. Most injuries to
the distal interphalangeal joints and the proximal interphalangeal joints
should be splinted in 30 to 35 degrees of flexion in a position similar to
that of one holding a soft ball. This prevents contractures of the collateral
ligaments, which may occur with prolonged immobilization in full
extension. Physicians should never assume that a nurse or technician
will splint an injury appropriately and must be specific with their
treatment instructions.

References
1 Simon RR and Koenigsknecht SJ (2001) Emergency Orthopedics: The Extremities (4e).
McGraw-Hill, New York, New York, 320.
2 Simon RR and Koenigsknecht SJ (2001) Emergency Orthopedics: The Extremities (4e).
McGraw-Hill, New York, New York, 477.
3 Simon RR and Koenigsknecht SJ (2001) Emergency Orthopedics: The Extremities (4e).
McGraw-Hill, New York, New York, 240.
4 Juhl JH, Crummy AB and Kuhlman JE (1998) Paul and Juhl’s Essentials of Radiologic
Imaging (7e). Lippincott-Raven, Philadelphia, Pennsylvania, 57.
5 Simon RR and Koenigsknecht SJ (2001) Emergency Orthopedics: The Extremities (4e).
McGraw-Hill, New York, New York, 535.
6 Ferrera PC, Colucciello SA, Marx JA et al. (2001) Trauma Management: An Emergency
Approach. Mosby, St. Louis, Missouri, 252.
7 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 147–8.
8 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 153.
9 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 156–7.
10 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 163–4.
 Musculoskeletal injuries 143

11 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 165–6.
12 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 74.
13 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 150.
14 Gardner AF, Mendelson DJ (2002) Avoidable Errors in Wound Management. Foresight:
Risk Management for Emergency Physicians. 55(October): 6–7.
This page intentionally left blank
Chapter 9

Shortness of breath

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Although shortness of breath represents only 2–3% of all emergency department

visits, it accounts for 15–25% of all hospital admissions.1 This disparity is related
to the seriousness of the complaint. As we are taught in Advanced Cardiac Life
Support, airway (the A in the ABCs of resuscitation) and breathing (the B) are
the first two requirements for survival. If there are problems with these two
functions, the patient is almost assured to have a poor outcome. Therefore,
admissions for shortness of breath tend to occur more frequently than that of
other complaints.
Depending on the source of reference, 66%2 to 75%1 of patients with shortness
of breath have cardiac or pulmonary etiologies to their complaints. However, we
will discuss in the ‘Differential diagnosis’ section that shortness of breath may
also arise from other organ systems. Unlike the complaints of pain discussed in
other chapters, the physiology of maintaining adequate ventilation and
oxygenation requires the interactions of multiple systems. The breathing
stimulus is controlled by the central nervous system. The anatomical airway
extends from the oropharynx to the lungs. Proper ventilation requires the
presence of adequate perfusion and functional hemoglobin through the lungs.
Therefore, there are many pathways where a disturbance in the normal
respiratory functioning of an individual may occur.

Triage
Because of the importance of breathing to maintaining the normal function of
all of the other organs in the body, a complaint of shortness of breath must be
taken seriously. There are very few exceptions to not have a patient with short-
ness of breath evaluated and treated emergently. Those with normal vital signs,
no significant lung history, and who appear in no distress are probably the only
reasonable candidates. Notice that we used the term and in the prior sentence,
not or. This is because we have all probably seen severe asthmatic patients who
are on the verge of respiratory collapse and have normal vital signs.
146 Learning from medical errors: clinical problems

The same can be said of a patient with a prior history of severe lung disease.
An asthmatic with a history of multiple intubations and shortness of breath
should not be left sitting in the waiting room even if her vital signs are normal
and she does not appear in any distress. These patients have the potential to
decompensate quickly and airway management is best when performed in a
monitored setting. Even if the patient does not have a prior history of lung disease,
a good triage nurse should be skilled at doing some detective work. Does the
patient have a cast on for a recent fracture? If so, the patient may need an emergent
evaluation for pulmonary embolism. How about the diabetic patient? Could he
or she be having a silent myocardial infarction manifested only by shortness of
breath? Is the patient in diabetic ketoacidosis? These are some of the situations
for which physicians should assure that their triage nurse is checking. At some
hospitals, nurses are taught to triage from seminars conducted by nurses. We
disagree with this policy. We believe that physicians should take the time to lend
our experience to our triage nurses. Physicians are more adept at instructing
nurses on how to detect the subtle findings of an occult emergency.
Any patient complaining of shortness of breath and with an appearance of
distress or abnormal vital signs requires a straight trip to an examination room
to see the physician. Even if the physician is occupied with another emergency,
the patient can be placed on a monitor and supportive treatment can be
instigated. Supportive treatment usually consists of oxygen, nebulizer treatments,
and/or electrocardiogram.

History
Patients with a history of asthma or chronic obstructive pulmonary disease
(COPD) are probably the most common groups to present to the emergency
department with shortness of breath. Asthmatics should be assessed according
to the severity of their asthma (i.e. hospitalizations, intubations, medications
used, normal peak expiratory flow, emergency room visits). For first time ‘wheez-
ers’, inquire about a family history of asthma. Events surrounding their current
attacks should also be ascertained. These include: triggers, fevers, colored
sputum, chest pain, medications used during the exacerbation, and peak
expiratory flow during the exacerbation. We make it a habit to ask every patient
with shortness of breath if there are any fevers, colored sputum (including
hemoptysis), chest pain, recent immobilization, and leg pain or swelling. If any
of these is present, be cautious for other etiologies. Some medications may
exacerbate asthma such as aspirin and beta-blockers, so the patient’s current
medications are always important to note in the chart.
Chronic obstructive pulmonary disease patients should receive most of the
same questions that are used for asthmatic patients. There are some aspects of the
history, however, that are distinct to the COPD patient. These patients usually have
a baseline production of sputum. Therefore, it is more helpful to ask if there has
been an increase in their sputum production and a change in their sputum char-
acteristics in order to assess the severity of their exacerbation. They also tend to
have other associated chronic medical problems such as coronary artery disease
and are more prone to have myocardial infarctions and congestive heart failure.
Finally, they are also frequently on medications for their lungs or their heart that
may be at subtherapeutic or toxic levels such as theophylline and digoxin.
 Shortness of breath 147

All older children and adults with shortness of breath should be asked if they
smoke cigarettes. This is important because of the numerous medical problems
(especially pulmonary problems) and the increased morbidity associated with
tobacco use. We have seen plaintiff attorneys base their arguments on whether
the patient was ever told to stop smoking. We have also found that questioning
about tobacco use is good medical practice because the majority of patients know
that it is detrimental to their health and will express some guilt if you simply
ask them about it. Adult patients should be asked about cardiac problems because
of their tendency to produce shortness of breath (e.g. congestive heart failure,
exertional angina). This includes asking the patient for the presence of chest
pain. Please see Case 9.7.
Selected individuals with shortness of breath may require certain additional
historical information. Those with night sweats and weight loss may have a
history of tuberculosis exposure. Tall, thin individuals may have had a
spontaneous pneumothorax in the past. Surgery history is important for we have
often seen pleural effusions as the cause of shortness of breath in post-operative
coronary artery bypass graft patients. Finally, nonsmokers with hemoptysis and
weight loss should be asked about asbestos exposure.
For patients with suspected pulmonary embolus (which is many patients with
the above complaint), inquire whether the patient has had a prior history of DVT
or PE because of the increased risk of subsequent events. If there is no prior
personal history of these events, suspicion should still remain high if a family
history is given. Risk factors for these events should also be assessed. We like
to use the acronym ‘Thrombosis’ in Tintinalli’s Emergency Medicine: a
comprehensive study guide to remember these factors.3
• T: Trauma
• H: Hypercoagulable, hormone replacement
• R: Recreational drugs (intravenous drugs)
• O: Old age (>40)
• M: Malignancy
• B: Birth control pill, blood group A
• O: Obesity, obstetrics
• S: Surgery, smoking
• I: Immobilization
• S: Sickness.
We would like to add some other conditions to the above acronym based on
listings in Rosen’s Emergency Medicine: concepts and clinical practice, Harrison’s
Principles of Internal Medicine, and some that we have created.4, 5
• T: Thrombocytosis (i.e. polycythemia vera)
• H: Human Immunodeficiency Virus, hemolytic anemia, heart failure
• R: Rest (bed)
• O: Orthopedic fractures
• M: Myocardial infarction, myeloproliferative diseases
• B: Burns
• O: Operations (especially pelvic and orthopedic)
• S: Spinal cord injuries
• I: Indwelling catheters
• S: SLE (systemic lupus erythematosus)
148 Learning from medical errors: clinical problems

Notice that we spent a rather lengthy discussion on assessing the possibility of

pulmonary embolism. This is because we feel that this is one of the most difficult
diagnoses in medicine to make and certainly one that carries extremely high
morbidity and mortality. Indeed, a missed pulmonary embolism is a
devastating case to the physician because it will usually mean a legal victory
and a large financial payout.

Physical exam
Probably no other complaint warrants the documentation of all five vital signs
as much as that of ‘shortness of breath.’ Vital signs provide a general overview
of the person’s health status and adequate respiratory function requires the
interaction of many organ systems. The caveat to this is that patients may be in
significant distress and have normal vital signs, as discussed above, and pulse
oximetry may be falsely abnormal in selected patients. For example, a patient
with poor peripheral perfusion may have a falsely low oxygen saturation while
a person with carboxyhemoglobin poisoning may have a falsely high oxygen
saturation.
The first part of the physical examination is the general appearance and
physiologic color of the patient. Does the patient have audible wheezing or
intercostal retractions? Is the patient sitting forward, having stridor, or having
problems with speaking? Is the patient pale or cyanotic? These are all signs of
severe respiratory distress and must be addressed immediately. Conversely, a
patient who is lying back and talking with normal speech is probably not in
significant distress.
The head, eyes, ears, nose, and throat (HEENT) examination should focus on
the oropharynx. Look for signs of obstruction that may explain the shortness of
breath. Are the tonsils or the uvula swollen and enlarged? Is there a peritonsillar
or retropharyngeal abscess present? Is the patient bringing up blood or foul
smelling secretions to suggest the presence of a foreign body? Is angioedema
present?
Continuing distally with the airway exam leads us to the neck exam. Is the
patient making stridorous sounds to suggest the presence of upper airway
obstruction? Check for deviation of the trachea or the presence of subcutaneous
emphysema. These conditions may occur with a spontaneous pneumothorax or
a pneumomediastinum. Prominent swelling or adenopathy in the neck may be
clues to an abscess or tumor that has eroded or infringed on the airway.
The most important components of the examination for the patient with short-
ness of breath are the pulmonary and cardiac exams. As we mentioned, in the
majority of cases, the etiology lies within these two systems. Just as important
as the documentation of accessory respiratory sounds such as wheezes, rhonchi,
or rales, the adequacy of air movement is also important to note. A person with
poor effort and air movement may not be moving enough air to create accessory
breath sounds. The absence of air movement may also suggest the presence of
a pneumothorax that requires immediate treatment before any diagnostic test.
Does the cardiac exam suggest the presence of congestive heart failure or
tachydysrhythmia?
Now that the assessment of the complete airway is complete, examination of
other organ systems is added as clinically indicated. Patients with a decreased
 Shortness of breath 149

respiratory drive should have a neurologic examination. Those who have a history
of liver disease or have had abdominal surgeries should have their abdomen
palpated for distention. A musculoskeletal examination is mandatory for those
with risk factors for deep venous thrombosis. Finally, a vascular examination is
required for those at risk for peripheral vascular disease to determine the
patient’s perfusion status.

Differential diagnosis
As we mentioned above, shortness of breath may result from pathology of
multiple organ systems. By far the two most common systems are the cardiac
and the pulmonary systems. Sources of respiratory distress from these systems
are discussed below. The physician must not forget, however, that dyspnea may
occur with diseases that do not originate from the thoracic area. Anemia, carbon
monoxide poisoning, and methemoglobinemia are common hematologic causes
of shortness of breath. Ascites and gastroesophageal reflux diseases are
gastrointestinal etiologies of dyspnea. Psychiatric causes of respiratory distress
include anxiety disorder and aspirin overdose. Finally, thyroid dysfunction and
diabetic ketoacidosis are endocrine disorders that may also cause shortness of
breath.
An extensive list of cardiac causes of shortness of breath is found in Emergency
Medicine: a comprehensive study guide by Judith Tintinalli, MD.6 Common ones in
our practice are congestive heart failure, myocardial ischemia or infarction,
arrhythmias, cardiomyopathy, and hypertensive crisis. Less common ones are
pericarditis, pericardial tamponade, and valvular disorders.
There is also an extensive list of respiratory causes of shortness of breath in
the same reference. 6 Common ones in our practice are asthma, COPD,
pneumonia, pulmonary edema, and atelectasis. Less common ones are pulmo-
nary contusion, pulmonary fibrosis, respiratory distress syndrome, sarcoidosis,
pneumothorax, and pulmonary embolism.

Diagnostic tests
We are supporters of arterial blood gases (ABG) for patients with shortness of
breath. We know that some physicians do not believe in them and there is no
literature that we can find that is clear on their use. However, we find that ABGs
are useful in certain situations. As mentioned above, in patients where pulse
oximetry is unreliable, ABGs can serve as an extension of the vital signs. We also
like to use the alveolar-arterial gradient on the ABG as a diagnostic indicator of
pulmonary embolism in the low risk patient. We feel that this test is especially
useful when the d-dimer test is not available. Arterial blood gases have also been
helpful for patients with hyperventilation or carbon dioxide retention. Finally,
a normal ABG is good legal documentation for a patient that decompensates
after being discharged.
The primary focus of diagnostic testing for patients with shortness of breath
should be on evaluating the pulmonary and cardiac systems. Chest X-ray and
electrocardiogram are the two most common tests ordered but do not need to
be ordered on every patient. Young patients (less than 40 years of age) who do
not have cardiac risk factors or a prior cardiac history probably do not need an
150 Learning from medical errors: clinical problems

electrocardiogram. We are, however, rather liberal with EKGs and usually will
order one in any patient with dyspnea who is over 50 years of age, has cardiac
risk factors, or has a cardiac history. Similarly, patients with dyspnea exacerbations
(asthma, COPD) that are similar to their prior episodes and improve with therapy
may not need a chest X-ray. In most other cases, chest X-rays are indicated because
of its ability to detect many of the diseases discussed above. It should be
remembered, however, that the initial chest X-ray may be normal in many
pulmonary disease processes. Please see Case 9.4.
Blood work can be useful in selected patients. A complete blood count is helpful
with patients who may be anemic. Serum levels of respiratory medicines such
as theophylline may give clues to the diagnosis. Finally, cardiac enzymes are
indicated if cardiac ischemia is suspected.
For patients with suspected pulmonary embolus, there are several clinical
tests that can suggest or confirm the diagnosis. The physician will often have to
combine the results of these tests along with her clinical suspicions and findings.
Please see the discussion concerning testing for pulmonary embolus in Chapter
2 in the ‘Diagnostic tests’ section. There are severe limitations to the
interpretation of these tests (with the exception of pulmonary angiography),
however, as pointed out in the ‘Pitfalls’ section below. In ordering diagnostic
tests in the evaluation for pulmonary embolus, the general theme is ‘with the
exception of angiography, a negative test in the presence of moderate or high
clinical probability tends to force the clinician toward further testing.’7
Echocardiography provides a quick and noninvasive modality to evaluate
cardiac origins of shortness of breath. The focused assessment with sonography
for trauma (FAST) evaluation includes a pericardial view to look for a
pericardial effusion. Other uses of ultrasound include valvular evaluation and
assessment of cardiac wall motion.

Hospital/office course
There will be many times when a patient with shortness of breath will require
emergency treatment. Specifically, a patient whose respiratory status is
decompensating rapidly may require an artificial airway and assisted
ventilation, needle decompression of a pneumothorax, or thrombolytics for a
pulmonary embolism. For the first two procedures, the general rule is: ‘if in
doubt, do them.’ Complications are extremely rare with both procedures and
hesitation may prove to be a fatal mistake for the patient. Please see Case 9.5
below.
Some patients will not need immediate life-saving procedures but will need
emergent treatment. These treatments must be ordered and timed appropriately
in the chart for possible further review. Examples would be nebulizer treatments
for bronchospasms and antibiotics for pneumonia. Our hospital set a goal of 30
minutes for the initialization of antibiotics once the diagnosis of pneumonia is
made. Chart follow-up exams after all therapeutic interventions (including peak
expiratory flows). The treatment for cardiac causes of shortness of breath is
discussed in Chapter 2.
Before discharging a patient who presented with dyspnea, document that the
patient’s breathing has improved and the patient has clear instructions
concerning when to return and who to follow up with. Get into the habit of
 Shortness of breath 151

recording oxygen saturation at discharge to show improvement or stability. Please

see Case 9.6. Make sure that the patient is able to get her medicines and also
knows how to use them. Antibiotics and inhalers can be fairly expensive and we
have all seen patients bounce back with shortness of breath because they
could not get their inhalers or they did not start the antibiotics. All smokers
should be encouraged to stop or decrease smoking. Finally, ask yourself if
you are comfortably convinced that the patient does not have a pulmonary
embolus.

Pitfalls
Be alert for angina masquerading as shortness of breath in patients with risk
factors for coronary artery disease. In these patients, we treat shortness of breath
as a chest pain equivalent as far as diagnostic testing. This conservative approach
has been rewarded with the detection of several silent myocardial infarctions.
Please see Case 9.7 in the section below.
Birth control pills can make an otherwise healthy, young female at risk for
deep venous thrombosis (DVT) and pulmonary embolism (PE). This is a well-
proven phenomenon and has been the foundation of many lawsuits. The
relationship of hormone replacement and DVT or PE has also been documented
but is not as strong as that of birth control pills. See the case of the young woman
with the PE below (Case 9.1). We feel that this is an important learning case
because PE is one of the rare diseases that may kill a young, healthy individual
with no preceding illness.
Interpretation of ventilation/perfusion (V/Q) scans can be tricky. In general,
V/Q scans are read as one of four results: normal, low probability, high
probability, or positive for pulmonary embolus. A scan that is read as normal or
positive is rarely accompanied by confusion unless the clinical suspicion is in
direct contradiction. However, V/Q scans are seldom read as normal due to the
abnormality that may be present from any other pulmonary process (e.g.
pneumonia, pulmonary edema). In addition, normal lung scans can still miss
pulmonary emboli involving subsegmental branches of the pulmonary arterial
circulation.8 Treatment decisions become more difficult when the scan is read as
low or high probability. Please see the case of an unfortunate 32-year-old lady
below. Finally, the use of spiral CT scan to rule out pulmonary embolism also
has similar limitations. They also have a ‘substantial rate of missing smaller,
peripheral pulmonary emboli.’7 Please see Case 9.2 below.
Persistent respiratory symptoms in young children such as wheezing and cough
without evidence of infection should raise the suspicion of foreign body
ingestion. Infants and toddlers are notorious for placing foreign objects in their
mouth and swallowing them. They will also rarely tell their parents or their
physician that they did so. Therefore, consider obtaining neck and chest X-rays
in the evaluation of these patients. Please see Case 9.3 in the following section.
We sometimes forget how fragile senior citizens can be with respiratory
problems such as COPD exacerbation and pneumonia. Disposition decisions in
seniors should be extremely conservative because of their decreased reserve and
compensatory mechanisms. Lung function is impaired along with cardiac
ability to compensate. In addition, they may have multiple medical illnesses and
may live alone or not have means to return if their breathing worsens. We have
152 Learning from medical errors: clinical problems

seen many elderly patients decompensate quickly in the emergency department,

in the clinic, and at home. Please see Case 9.6 below.
It is generally not wise to make drastic changes in respiratory therapy for a
patient that is discharged. Because there are so many respiratory treatment
therapies (i.e. inhalers, nebulizers, steroids, cromolyn, leukotriene antagonists,
theophylline, anticholinergics, and home oxygen), we sometimes utilize as much
of these as possible in order to keep the patient out of the hospital. It must be
noted, however, that some of these therapies do not work immediately, some
cannot be obtained immediately, and some are very difficult to learn to use
immediately. Therefore, the prescription of these therapies sometimes is met
with failure before there is success. Please see Case 9.7 below.
The clinical presentation of congestive heart failure and pulmonary embolism
can often be extremely similar. Shortness of breath, hypoxia, tachycardia, and
leg swelling can be signs and symptoms of both diseases. In addition, patients
with severe congestive heart failure are frequently nonambulatory. This
predisposes them to deep venous thrombosis and further confuses the
diagnosis. Please see Case 9.8 below.
Finally, do not forget that shortness of breath can be a manifestation of
systemic diseases. Adequate respiratory function requires the coordination of
the neurologic, musculoskeletal, respiratory, and cardiovascular systems. There-
fore, not every patient with shortness of breath will have a disease of the
respiratory system. In Case 9.9, we present a ‘zebra’ case of that illustrates this
point.

Errors and interesting cases

Case 9.1 Never too young

• Edwards presents a case, in The M & M Files, of a 17-year-old female who
presented to the emergency department complaining of one week of
intermittent shortness of breath.9 She began taking birth control pills
recently and was diagnosed a few days earlier by her primary physician
with anxiety. Her initial vital signs during the ED visit revealed a
respiratory rate of 28 breaths/minute, a heart rate of 116 beats/minute,
and an oxygen saturation of 96% on room air. The ED physician noted
that the patient had appeared anxious and did not have wheezes on her
lung examination. He ordered a nebulizer treatment and diazepam. The
patient felt better and was discharged home.
• The patient’s breathing became worse after discharge and she was taken
to another ED the following day where blood gases and a lung scan
revealed a pulmonary embolus. Additional history was discovered that
the patient had several immediate relatives who had problems with
hypercoagulability. Although the patient received treatment without long-
term complications, her mother filed a complaint to the initial hospital
administration for the misdiagnosis.
• Although many physicians would not consider the diagnosis of

continued
 Shortness of breath 153

pulmonary embolism in this patient because of her age, assessment of

risk factors is still prudent. This patient was on birth control pills and had
an extensive family history. The first ED physician neglected to obtain
this history. He also did not take notice of the patient’s abnormal vital
signs. The diagnosis, consequently, was not made on the initial visit.

Case 9.2 Interpret with caution

• In another case from Edwards’ The M & M Files, a 32-year-old lady pre-
sented to the emergency department with shortness of breath, nausea,
and diaphoresis.10 She had a near syncopal episode but denied any chest
pain. Her foot was in a short leg cast from an ankle sprain suffered two
weeks before the current visit. She was also a smoker and had a history
of panic attacks. Her vital signs showed a pulse oximetry of 95% on room
air, a pulse of 100 beats/minute, and a respiratory rate of 26 breaths/
minute.
• Because of her numerous risk factors for thromboembolic disease and
her lack of improvement after an albuterol treatment, the ED physician
ordered a work-up for pulmonary embolism. Her electrocardiogram
showed a mild sinus tachycardia. Her arterial blood gas revealed a pCO2
of 34 mm Hg and a pO2 of 90 mm Hg on room air and a ventilation/
perfusion scan was read as ‘low probability’. The physician then
discussed the case with the patient’s primary care doctor and decided to
discharge the patient with the diagnosis of ‘hyperventilation syndrome’.
• When the patient followed up with her primary physician the following
day, she was given the same diagnosis without further work-up. A day
later, she was brought back to the ED in full cardiac arrest and was unable
to be resuscitated. A subsequent autopsy showed that a massive
pulmonary embolus was the cause of death.
• The treating physician based his disposition of the patient on the results
of a ventilation/perfusion scan. In a patient with a high clinical
suspicion for pulmonary embolus, a ‘low probability’ scan should be
met with some skepticism. This patient should have been either treated
for PE or have more tests ordered. In addition, hyperventilation
syndrome is a diagnosis of exclusion. This patient did not safely have the
diagnosis of PE excluded.

Case 9.3 That child will eat anything

• A three-year-old child was brought to our emergency department with
a persistent dry cough that had been present for over three weeks. The
cough was nonproductive and not associated with fevers, wheezing, or
stridor. It also seemed to be worst at night. The child had seen his
pediatrician two times and just finished a course of antibiotics and was

continued
154 Learning from medical errors: clinical problems

also on cough medicine that did not appear to help. The child’s mother
denied any allergy symptoms.
• Physical examination revealed normal vital signs. The child was in no
respiratory distress. His lung examination was normal. A chest-ray was
ordered because of the persistent symptoms. The majority of the neck
was visualized on X-ray due to the child’s small thorax. A nickel-sized
foreign body was seen from the front side on the AP view of the neck.
This was consistent with a probable swallowed nickel in the esophagus.
(A foreign body in the trachea is usually seen in the sagittal plane on the
AP view.) The patient’s mother was surprised to find the cause of the
persistent cough.
• We talked about maintaining a low threshold for ordering X-rays in the
pediatric patient in Chapter 6, Leg pain. A similar statement can be made
concerning the pediatric patient with shortness of breath. Children are
able to give limited histories of their symptoms. In addition, they may
not tell the physician certain important facts (swallowing a foreign body).
Therefore, X-rays are occasionally helpful with the diagnosis.

Case 9.4 X-rays are sometimes behind

• Chest X-rays are frequently invaluable in patients with shortness of breath.
It may give the physician a diagnosis in a matter of seconds. The
limitations of chest X-rays, however, must also be recognized. The
sensitivity of X-rays for pneumonia decreases under certain conditions.
It may not show up if the patient is dehydrated or in the early stages of
aspiration. It also tends to be normal with some specific pneumonias
such as ‘walking’ pneumonia and Pneumocystis carinii pneumonia.
• We saw a healthy 23-year-old male in our emergency department who
had 2 days of pleuritic right chest pain and shortness of breath. He did
not have a fever and his vitals were within normal limits. Physical exam
was also within normal limits. We ordered a chest X-ray and an EKG and
both were unremarkable. The radiologist also interpreted the patient’s
X-ray as normal. The pain improved with an intravenous dose of Toradol
and he was discharged with probable pleurisy.
• He returned to the ED two days later because his symptoms were not
better. His presentation was essentially unchanged and his vitals and
clinical exam were not significantly different. We repeated the chest X-
ray and there was a 1 cm by 1 cm by 1.5 cm hyperdense area on the right
mid-peripheral lung. His complete blood count was normal. This area
was evaluated with a CT scan and interpreted by the radiologist as a
‘necrotic abscess.’
• The patient was admitted to the pulmonary service for intravenous
antibiotics and further evaluation. We were curious to discover the
etiology of this patient’s pathology and visited the patient every day
while he was in the hospital. He improved steadily on antibiotics. All of

continued
 Shortness of breath 155

his cultures (blood, sputum, acid-fast, fungal) were negative. Since the
patient showed steady improvement, a bronchoscopy with biopsy was
not done and the definitive organism was never found. Nevertheless,
this case emphasizes the importance of not relying on an initial X-ray
that is normal.

Case 9.5 Will travel, need airway

• Our colleague and friend, Warren Blount, MD gave us the next case. An
urban ED physician received a phone call one night from an emergency
physician from a small, rural emergency room about 100 miles from his
hospital. The rural physician had a patient with a history of chronic
obstructive pulmonary disease who presented to his hospital in severe
congestive heart failure and pulmonary edema. The caller was request-
ing permission to transfer the patient because his hospital did not have
an intensive care unit.
• The urban physician accepted the patient but was concerned that the
patient’s pCO2 on arterial blood gas was 90 mm Hg. He, therefore,
requested that the transferring physician intubate the patient prior to
transport. The rural physician agreed to do this and the conversation
ended. The urban physician did not hear anything further concerning
this patient until the paramedics radioed his hospital that they were ‘5
minutes out and the patient had stopped breathing.’ Upon arrival, the
patient was immediately intubated and was able to be resuscitated.
Unfortunately, he had a complicated hospital course and subsequently
died.
• It was discovered later that the rural emergency physician worked in the
emergency department on a part-time basis and was not skilled at
intubating patients. He did not even make an attempt at this procedure
prior to sending the patient off. This resulted in the transfer of a critical
patient without proper stabilization. It also represented a violation of the
Emergency Medical Treatment and Active Labor Act (EMTALA). More
importantly, it unnecessarily jeopardized the medical health of the
patient.
• Protection and maintenance of an airway is the most important concept
in Advanced Cardiac Life Support. Hence, it is also important in the
transfer of unstable or potentially unstable patients. Failure to provide
an adequate airway has been a subject of numerous EMTALA violations
and malpractice lawsuits.
156 Learning from medical errors: clinical problems

Case 9.6 Old, fragile, and comorbidities

• Edwards presents a case, in The M & M Files, of an elderly woman, who
would have benefited from conservative respiratory therapy.11 After not
improving on ciprofloxacin and an oral inhaler for a cough and short-
ness of breath, an 83-year-old woman came to the emergency department.
She had a past medical history of hypertension, cerebrovascular
accident, and chronic obstructive pulmonary disease.
• On presentation, her vitals were: temperature 99.7ºF, blood pressure 160/
90 mm Hg, respiratory rate 28 breaths/minute, and oxygen saturation
not recorded. She had decreased breath sounds throughout and
expiratory wheezing but no signs of congestive heart failure. The
physician noted that her skin was warm and confirmed a fever of 101.2ºF
with a rectal thermometer. Tests performed in the ED showed a right
lower lobe pneumonia on chest X-ray and a 15,000 cells/microliter white
blood count with a 2% band forms on the bloodwork.
• After two breathing treatments and an antipyretic, she was feeling better
and the fever had resolved. She wanted to go home and the ED physician
later stated that he reluctantly honored this request. He gave her a dose
of antibiotics and wrote her prescriptions for an antibiotic and oral
steroid. She was told to follow up with her physician in 1–2 days. The
ED physician did not make note that the patient lived alone and did not
record an oxygen saturation at time of discharge.
• The patient became more short of breath that night. Her daughter came
to her assistance and decided to take her to the hospital the following
morning. Shortly later, however, the patient fell while going to the bath-
room and broke her hip. The oxygen saturation at the hospital was 89%.
The patient’s daughter filed a complaint against the initial ED physician
for not admitting her mother.
• There were many clues in this case to suggest that this patient might not
have done well with outpatient treatment. She had already been on
outpatient antibiotics and inhalers without improvement. She was
elderly with multiple chronic medical conditions, including chronic
obstructive pulmonary disease. She lived alone and did not have an easy
means to return to the emergency department. Finally, there was
insufficient chart documentation that this patient was appropriate for
discharge (no oxygen saturation recorded).
• Although the physician argued later that he spent at least 5 minutes in
trying to convince the patient to stay for admission, he did not mention
any of this conversation in his notes. Furthermore, he did not have the
patient sign a form for ‘refusal of care’ or ‘against medical advice’
documenting that the patient did not follow his recommended treatment.
Lastly, he did not consider that the patient might have family members
who shared different opinions of medical treatment from that of the
patient. We have experienced this phenomenon frequently in the treat-
ment of elderly patients, where their children are much more conservative
with their health than the patients are.
 Shortness of breath 157

Case 9.7 The masked acute coronary syndrome

• Another case from Edwards’ The M & M Files shows why it is important
to consider the important interaction between the cardiac and
pulmonary systems.12 A 65-year-old man with a history of COPD came
to the ED with increasing cough and shortness of breath. Home
medications such as inhalers, steroids, blood pressure medicines, and
nitroglycerin did not help his breathing. The triage nurse also noted that
he had chest pain but this was not addressed to the physician. His lips
were cyanotic and his pulse oximetry was 84% on room air. He had
numerous hospitalizations for COPD exacerbation and was in the
process of getting home oxygen.
• The patient perked up with oxygen, intravenous steroids, and nebulizer
treatments in the emergency department. The ED physician did not make
any mention of the patient’s chest pain in his note and did not order any
cardiac test. After three hours of treatment and observation, it was
arranged for oxygen equipment to be delivered to the patient’s home
and he was discharged from the ED.
• Two hours later, before the oxygen had arrived, the patient developed
more severe chest pain. The paramedics found him bradycardic, pulseless,
and with agonal respirations when they arrived. He could not be
resuscitated in the ED. A subsequent lawsuit was filed against the ED
physician.
• Although the ED physician had good intentions in this case by trying to
arrange for home oxygen and keeping the patient out of the hospital, this
patient simply had too many red flags to be discharged home. He had
a history of numerous hospitalizations for similar exacerbations. He failed
inhaler and steroid therapy at home. We agree with Edwards’ opinion
that home oxygen should be prescribed by a primary physician who is
able to follow the patient more closely. We feel that home oxygen can be
arranged once the patient’s exacerbation has resolved, not when it is at
its peak. It is difficult to set up and is supportive therapy and not
corrective. The documentation of cyanotic lips and pulse oximetry in the
low 80s does not suggest to reviewers of the chart that the patient should
be discharged regardless of the amount of clinical improvement in the
ED. Finally, the physician did not address the cardiac status in a patient
with numerous cardiac risk factors and the complaint of chest pain.

Case 9.8 Be careful of look-alikes

• Although congestive heart failure (CHF) and pulmonary embolism can
present with similar signs and symptoms, there are certain findings on
the history or physical that would make a physician more suspicious for
pulmonary embolism. Many of us were taught in medical school to not
look for zebras when the diagnosis is obvious. While this saying is for
the most part true, it may not apply under certain conditions. These

continued
158 Learning from medical errors: clinical problems

include when the clinical manifestations are similar, when two entities
coexist, or when one entity leads to the other. We should be more alert for
pulmonary embolism when there is a history of immobilization, asym-
metric leg swelling, ambiguous chest X-ray, or other risk factors are present.
• McNutt and Tabas present a case, in Foresight, where the patient’s prior
history of CHF confused the physician.13 A 70-year-old man with a
history of CHF and COPD was taken to the ED after developing short-
ness of breath upon walking. This symptom began 3 days prior to his
presentation, after he ran out of his medications. He had twisted his
ankle and was not able to ambulate to get to a pharmacy.
• The ED physician noticed a patient who was tachycardic, diaphoretic,
and gasping for air. He was using accessory breathing muscles and unable
to give a detailed history. His vital signs were: blood pressure 150/100
mm Hg, pulse 110 beats/minute, respiratory rate 40 breaths/minute,
temperature 99.9ºF, oxygen saturation of 83% on room air. Significant
physical exam findings included: jugular venous distention,
hepatojugular reflux, and bilateral pitting edema that was worse on the
right. Slight crackles were heard at the bases.
• The physician ordered a cardiac panel, an EKG, a chest X-ray, and a B-
Type Natriuretic Peptide (BNP) level. The EKG showed a sinus
tachycardia at 110 beats/minute and a right bundle branch block.
Cardiomegaly with a mild right pleural effusion was seen on the chest
X-ray. The physician’s clinical impression was congestive heart failure.
He ordered an aspirin, intravenous furosemide, and nitropaste for
the patient. The patient did not improve significantly with these
medications.
• The BNP level was 240 pg/ml. The physician regarded this level as
confirmation of his suspicion of congestive heart failure. He ordered more
furosemide and a nitroglycerin drip for the patient. The patient’s clinical
status, however, worsened. He was intubated and admitted to the
intensive care unit. Shortly later, he became hypotensive and an
echocardiogram showed right ventricular dilation and hypokinesis. This
was suggestive of pulmonary embolism. Although he was started on
anticoagulants and thrombolytics, he had a prolonged hospital course.
The patient filed a lawsuit against the ED physician and the hospital. The
defense decided to settle the case because they felt it would be difficult
to show that he physician considered the diagnosis of pulmonary
embolism.
• The patient’s presentation in this case was accompanied by several red
flags for pulmonary embolism. These red flags must be recognized and
addressed even if the patient’s symptoms could be explained by CHF
and/or COPD. The patient was recently nonambulatory and had
asymmetric leg swelling. His clinical findings suggested fluid overload
yet his chest X-ray had minimal findings to support this. This raised the
concern of right-sided heart failure without evidence of severe left-sided
heart failure. This combination is occasionally found in pulmonary

continued
 Shortness of breath 159

embolism. Finally, a BNP level between 100 pg/ml and 400 pg/ml
warranted further investigation for etiologies besides heart failure.
B-Type Natriuretic Peptide levels of 200 to 300 pg/ml have been reported
in cases of large pulmonary embolism.

Case 9.9 Easier to see the second time around

• A 42-year-old female presented to the ED with shortness of breath for
two days. She also complained of visual problems that had been present
for one year. The emergency physician noted that her vital signs were
normal. He also could not find any remarkable findings on her physical
exam. He, nevertheless, took her complaints seriously and ordered an
extensive work-up.
• The patient was kept and observed in the ED for eight hours. She had
a cardiac work-up and a head CT. She also had a d-dimer ordered. All
of the tests were normal. He could not find any evidence of neurologic,
respiratory, or cardiovascular disease. He told the patient that he did not
have an objective answer to her subjective complaints. She was then
discharged and instructed to follow up with her primary physician that
week.
• Two days later, the patient returned to the ED with increasing shortness
of breath. A different ED physician on duty noticed that her breaths had
become agonal and respiratory failure was imminent. Before he gave her
medications for rapid sequence intubation, he discovered that she could
not open her eyelids, her speech was slurred, and she had weakness of
the upper extremities. He became suspicious that this patient was
suffering from myasthenia gravis. A Tensilon test was performed and
the patient had a dramatic response and avoided intubation. She made
a rapid hospital recovery and is now under the care of a neurologist.
• This patient’s initial presentation was certainly a tough one for any ED
physician. The diagnosis of myasthenia gravis is rarely one made in the
ED. It is also one that is usually not evident unless the patient has a
severe crisis with critical muscle weakness. Nevertheless, this case serves
as a reminder that we must look beyond the heart and lungs for some
causes of shortness of breath.

References
1 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 623.
2 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 443.
3 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 416.
4 Rosen P, Barkin R (1998) Emergency Medicine: concepts and clinical practice (4e), Vol II.
Mosby, St Louis, Missouri, 1874–6.
160 Learning from medical errors: clinical problems

5 Fauci AS, Braunwald E, Isselbacher KJ et al. (1998) Harrison’s Principles of Internal

Medicine (14e). McGraw-Hill, New York, New York, 1403.
6 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 444.
7 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 46.
8 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 44.
9 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 35–6.
10 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 39.
11 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 193–5.
12 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 198–200.
13 McNutt E, Tabas JA (2003) Acute CHF Exacerbations. Foresight: Risk Management for
Emergency Physicians. 58(October): 2–3.
Chapter 10

Syncope

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Syncope is a not uncommon complaint and represents approximately 3% of ED

evaluations annually.1 By definition, it ‘is a sudden, transient loss of conscious-
ness associated with inability to maintain postural tone.’1 An important feature
of syncope that distinguishes it from many other causes of loss of consciousness
is the spontaneous recovery. Near syncope (sometimes called presyncope) can
be caused by the same disease processes and is usually worked up in a
similar fashion. The difference is that there is no loss of consciousness in near
syncope.

Triage
The majority of patients who present with syncope do not have to be seen
emergently. Special populations, however, such as the elderly and those with a
history of or risk factors for cardiac or neurologic disease should be given
immediate attention. Obviously, any patient with abnormal vital signs should
be seen emergently as with any other triage patient. Symptoms accompanying
the syncopal episode such as chest pain (myocardial infarction, aortic dissection,
pulmonary embolism, aortic stenosis), headache (subarachnoid hemorrhage),
and abdominal or back pain (leaking abdominal aortic aneurysm, ruptured
ectopic pregnancy) should be viewed as red flags and should be addressed
immediately.2 In addition, syncope during pregnancy or syncope with shortness
of breath should always raise the suspicion of pulmonary embolism and should
be brought to the physician’s attention. Finally, the patient’s appearance can
serve as a clue to a serious emergency. A pale and lethargic patient may have
passed out due to an active gastrointestinal hemorrhage and needs immediate
stabilization. The patient who has focal weakness or slurred speech, likewise,
needs immediate evaluation for an acute cerebrovascular event. The triage nurse
or physician should examine the patient for any traumatic injuries from the
syncope that may necessitate emergent attention. Finally, the patient with
162 Learning from medical errors: clinical problems

persistent vomiting or diarrhea along with syncope requires expedient fluid

resuscitation.
Most patients with syncope, however, can be triaged on an urgent and not
emergent basis. Inclusion criteria for this group are young (less than 50 years of
age) patients without a history of or risk factors for cardiac or neurologic
disease, patients with classic reflex-mediated (see below) or medicine-related
(see below) syncopes, and patients who are currently asymptomatic. Normal
vital signs and an appearance that is without distress are also important
stipulations for the latter group of patients.

History
Past medical history is extremely important in risk stratifying the patient with
a syncopal episode. Has the patient ever had a syncopal episode in the past? If
so, what kinds of tests were performed and what was their physician’s opinion
of the syncope? Recurrent unexplained or uninvestigated syncope should always
raise concern and lower the threshold for engaging in an extensive evaluation.
A history of cardiac, neurologic, or psychiatric illness may provide clues for
recurrent or more serious causes of syncope. Recent illnesses such as
gastroenteritis or hepatitis are common and usually benign causes of syncope
(from the fluid depletion). Any patient with a cardiac pacemaker and syncope
should be asked about the last time that the pacemaker was checked and should
be considered to have a possibly malfunctioning pacemaker. Most patients are
instructed to carry the phone number of the pacemaker ’s company or
technician with them. Ask them for this information. This enables the physician
to request a technician to come to evaluate the pacemaker. Finally, does the patient
have risk factors for pulmonary embolism or can the patient be pregnant?
Blok writes in Emergency Medicine: a comprehensive study guide, ‘up to 13 percent
of patients diagnosed with pulmonary embolism had an initial syncopal
episode.’3
The syncope history should include details of the events preceding, the events
during, and the events after the episode. Did the event occur after a painful or
fearful incident as in the classic vasovagal syncope? Was there coughing,
urination, defecation, or swallowing preceding the event to suggest situational
syncope? Carotid sinus syncope may occur after shaving, turning the head,
wearing a tight collar shirt, or in those with a history of neck cancers. Has the
patient been having vomiting, diarrhea, or heavy bleeding (e.g. menstrual
bleeding, gastrointestinal bleeding) to suggest orthostatic hypotension? An
infrequent case of orthostatic hypotension might be the patient with acute
adrenal crisis from recent withdrawal of chronic steroids. Has the patient started
any new medications (see below) or had the dosages been recently adjusted on
chronic medications? Did the patient pass out after complaining of a headache
or auras to suggest a migraine?
Ask the patient or witnesses to describe the details of the syncopal event and
the details after the syncopal event. This is important because a lot of people use
the term ‘passing out’ as a ‘catch-all’ phrase. We have heard patients use it for
near syncope episodes, for light-headedness, for vertigo, for seizures, and even
for severe fatigue. Syncope is usually caused by an inciting event that results in
a transient decrease in blood flow to the brain. In the case of cardiac syncope,
 Syncope 163

however, the event may come on suddenly and without warning. After the
syncopal episode, however, ‘the reclined posture of syncope and the response
of autonomic autoregulatory centers reestablish cerebral perfusion, leading to
a spontaneous return of consciousness.’1 Syncope is not usually associated with
post-event confusion, urinary incontinence, or prolonged loss of consciousness.
Mention should be made to the specific scenarios when patients are involved in
single person traumas such as motor vehicle accidents and falls. These incidents
should always be assumed to occur from a syncopal event until further history
suggests otherwise.

Physical exam
All five vital signs should be recorded and the patient placed on a cardiac monitor
if there is suspicion of cardiovascular or neurologic processes. It is helpful to
take blood pressures in different positions and in both arms to screen for
orthostasis, aortic dissection, or subclavian steal syndrome. Repeat vital signs
are often necessary after therapeutic interventions or before discharge to
demonstrate improvement or stability.
The physical examination is focused on the cardiovascular and neurologic
systems to screen for serious etiologies of syncope. An examination for
traumatic injuries should be carried out in the usual fashion if injury during
syncope is suggested. ‘Rectal examination must be performed on all patients to
evaluate for gastrointestinal bleeding.’2

Differential diagnosis
A recent clinical update in Emergency Medicine journal divided the classifications
of syncope into etiologic classes that we found especially useful.4 These classes
included: bradyarrhythmias, ventricular tachyarrhythmias, supraventricular
tachyarrhythmias, hypertrophic cardiomyopathy or long QT syndrome,
neurocardiogenic or situational syncope, carotid sinus hypersensitivity,
cerebrovascular disease, orthostatic hypotension or autonomic dysfunction,
drugs, dehydration, hyperventilation, and other miscellaneous causes.
Blok, in Emergency Medicine: a comprehensive study guide, narrows the above list
into six basic classes.5 We feel that this list condenses the list in the preceding
paragraph without sacrificing any loss of content. These include: reflex-
mediated, orthostatic hypotension, psychiatric, neurologic, medications, and
cardiac. Reflex-mediated syncope is further divided into vasovagal, situational,
and carotid sinus syndrome. Neurologic syncopal episodes may arise from
transient ischemic attacks, subclavian steal syndrome, or migraines. The
majority of cardiac syncopes are either structurally related (e.g. aortic stenosis,
cardiomyopathy) or rhythm-related (e.g. supraventricular tachycardia, heart
blocks) events. Cardiac ischemia, pulmonary embolism, and aortic dissection,
however, are other less frequent causes of cardiac syncope.
Blok also provides a useful list of medicines that are frequently associated
with syncope in Emergency Medicine: a comprehensive study guide.6 This list is
comprised of: antihypertensives, beta-blockers, cardiac glycosides, diuretics,
antidysrhythmics, antipsychotics, antiparkinsonian drugs, antidepressants,
phenothiazines, nitrates, alcohol, and cocaine.
164 Learning from medical errors: clinical problems

Diagnostic tests
In the ED management (or office evaluation) of syncope, there is only one test
that is routinely recommended: the electrocardiogram.7 Other tests are ordered
on a case-based approach. Examples would be a head CT for patients with stroke
or transient ischemic attack symptoms, pregnancy test for women of child-
bearing age, complete blood count for those with bleeding, serum electrolytes
for those with vomiting or diarrhea, and echocardiogram for those with new
heart murmurs. Maintain a low threshold for ordering a CT pulmonary study
or ventilation/perfusion scan for patients with risk factors for pulmonary
embolism and unexplained syncope.
Routine head CT scan, EEG, or lumbar puncture has not been shown
beneficial for patients with syncope unless there are findings to suggest a
neurologic cause.7 Similarly, use of echocardiograms should be limited to those
with possible structural heart disease.7 Patients who are discharged from the
hospital can be referred for outpatient tests such as tilt-table testing, Holter
monitor, and some of the ones discussed above.

Hospital/office course
Patients with suspected cardiac or acute neurologic causes of syncope should
be placed immediately on a cardiac monitor. Some arrhythmias may need
immediate treatment if associated with syncope. For example, we had an elderly
lady present to our ED after having three episodes of syncope. She was noted
to have a third degree heart block on her electrocardiogram. A transcutaneous
pacer was applied while we waited for the cardiologist to place a transvenous
pacer. The more familiar scenario is the supraventricular tachycardia that also
requires immediate treatment.
Studies involving risk stratification in patients with syncope have mainly been
concerned with cardiac syncope. One study identified ‘four significant
predictors of sudden cardiac death or significant dysrhythmia within one year
of a syncopal event.’8 The predictors included an abnormal EKG, age greater
than 45 years, history of ventricular dysrhythmia, and history of congestive heart
failure. Rosen and Barkin, in Emergency Medicine: concepts and clinical practice,
add: ‘any patient who has a history of palpitations or irregular pulse in
proximity to the syncopal event should be considered as having a dysrhythmia
as the cause of the syncope until proved otherwise.’9
Over half of the patients who present to the ED for syncope will have an
uncertain cause for their syncope after the evaluation.8 The physician’s objective
is to group patients into four decision groups after the evaluation. The first group
has a demonstrated serious etiology mandating hospital admission for treat-
ment. The second group may have life-threatening causes (such as cardiac or
neurologic) and also require hospital admission for monitoring and further
testing. The third group is unlikely to have a serious etiology and may be
discharged with further testing as an outpatient. Finally, the last group has a
benign cause and may be discharged.
The elderly patient with syncope should be evaluated with extreme caution.
They tend to have more medical problems and take more medications.
‘Compared with a group of patients between 15 and 59, patients between 60 and
 Syncope 165

90 years of age have a cardiovascular cause of syncope twice the incidence found
in the younger group.’10 Hunt also adds the important point that the elderly also
have more associated injuries with their syncope.11 In Emergency Medicine:
concepts and clinical practice, it is recommended that any patient over 60 years of
age whose syncope cannot be explained as noncardiac be admitted to the
hospital for cardiac monitoring.12
Patients who are felt safe for discharge after the medical evaluation require
certain instructions. Adams and Martin recommend, in Principles and Practice of
Emergency Medicine, that discharged patients should be instructed to not drive
themselves home.13 Other useful instructions for patients who are waiting
further outpatient testing include: no driving or operating machinery, no
swimming, and no operating at heights. We also learned from a neurologist who
was giving discharge instructions to a patient with seizures an additional
helpful advice that also applies to patients with unexplained syncope. He told
the patient not to take baths when they were alone in the house.

Pitfalls
Remember to consider that syncope may have more than one cause. The
traditional evaluation of syncope focused on one etiologic cause. A recent study
published in the Mayo Clinic Proceedings, however, showed that 18% of patients
with syncope had more than one cause. The most frequently found combination
was carotid sinus hypersensitivity along with neurocardiogenic and situational
syncope.14
Inquire about injuries that may have occurred with the syncopal episode.
Physicians will sometimes forget to ask about injuries because they concentrate
on evaluating the cause of the syncopal episode. In particular, ask about head
injuries. These may occur when the patient falls. Head injuries may also precede
and lead to the syncopal event. For example, a recent head injury may cause an
intracranial bleed or a seizure. The seizure may then cause a syncopal episode
or appear like a syncopal episode. Furthermore, if the patient is confused about
the details of the incident, has a change in mental status, has injuries to the face
or scalp, or is demented, assume that a head injury is present unless there are
reliable witnesses that deny any such possibility. Please see the case of the 80-
year-old, demented man in the following section (Case 10.1).
Beware of a ‘crescendo’ pattern of syncope or near syncope. Since the majority
of serious causes of syncope are of cardiac origin, many of the important
historical points explained in Chapter 2 (Chest pain) also apply here. Patients
with risk factors for cardiovascular disease who present with multiple recent
episodes of syncope or near syncope can be considered to have an unstable
angina equivalent version of syncope and may need hospitalization for
monitoring and further evaluation. Please see Case 10.2 below.
A related discussion to the preceding paragraph is new-onset syncope in the
elderly patient or the patient with cardiac risk factors. These patients are at the
greatest risk of harboring a serious cause for syncope and should be considered
to have one until proven otherwise. A low threshold for admission, monitoring,
and further testing should be maintained to prevent the unfortunate death of a
patient from sudden cardiac death (the usual cause of fatality). Please see Case
10.3 below.
166 Learning from medical errors: clinical problems

A large number of medical visits are for patients who have fallen. Our
tendency, sometimes, is to address only the injuries from the fall and never
consider the mechanism of the fall itself. All patients who have fallen should be
asked why they fell. If they are not sure or do not remember why, than they
should be considered to have a syncopal episode until proven otherwise. This
is particularly true with elderly patients who do not engage in strenuous
physical activities and are less inclined to have activity-induced falls than their
younger counterparts. Please see Case 10.3 below.

Errors and interesting cases

Case 10.1 Search for syncopal injuries
• Edwards in The M & M Files presents the case of an 80-year-old man with
dementia who was brought to the ED by his daughter after a syncopal
episode.15 The daughter found the man sitting on his front porch with
a large abrasion on his forehead and a skin tear on his left arm after he
had gone outside to get the mail and did not return. She had called the
ambulance because he was unsteady with his gait.
• On presentation to the ED, he had normal vital signs. He was able to
ambulate without problems. He denied a headache and complained only
of left shoulder and left knee pain. His daughter stated that he seemed
to be ‘at his baseline mental state’. He told the ED physician that he was
unsure why he fell. The physician ordered X-rays of the chest, pelvis, left
hip, and left shoulder. All of the X-rays were negative. An electrocardio-
gram and routine blood work were unremarkable.
• The ED physician then called the patient’s primary physician to
recommend admission. The primary physician, however, recommended
that the patient be kept in the ED overnight and that he would see the
patient the following morning. The ED physician agreed to this plan and
subsequently signed the patient out to the next ED physician. Early the
next morning, the patient vomited and aspirated his vomitus because he
was having respiratory difficulty. He also responded to questions
sluggishly after he vomited. The new ED physician called the patient’s
primary physician and informed him of the events, after which the
primary physician gave admission orders over the telephone. The
patient was moved to a step-down unit outside of the ED.
• When the primary physician saw the patient the following morning, he
found the patient comatose. A massive subdural hematoma was seen on
the patient’s head CT scan. The patient died the following day. His
daughter questioned the hospital’s administration why the bleed in her
father’s brain had not been found earlier.
• This patient’s syncopal episode was associated with multiple events that
should prompt an emergency physician to consider a head CT. The patient
had a history of dementia making his mental status exam less reliable.
He also had an unobserved fall onto a hard surface and subsequent

continued
 Syncope 167

physical findings of a head injury. Finally, he had other physical findings

that might have suggested a head injury – unsteady gait, vomiting, and
sluggishness.
• In this case, the first ED physician ordered X-rays to search for injuries
from the syncopal episode. He neglected, however, to investigate the
most important injury. The patient had evidence of head trauma and an
unsteady gait. In addition, his history of dementia makes assessment of
his mental status difficult. Therefore, a CT scan is prudent to search for
an intracranial process.

Case 10.2 Caution with multiple or crescendo patterns

• Edwards presents another case, in The M & M Files, where an unstable
pattern of dizziness was not detected.16 While at work, a 46-year-old
nurse manager suddenly developed light-headedness and imbalance as
she was walking. This was followed an hour later by a longer episode
lasting a few minutes. Since she had never experienced anything like this
before, she decided to come to the ED once her shift was completed.
• The patient was well known to the ED physician. She was obese but had
no other medical problems. After performing a neurologic exam and an
EKG, both of which were normal, the ED physician discharged the patient
with a possible diagnosis of labyrinthitis and instructed her to have further
evaluation during the week. This diagnosis was partly from the sugges-
tion of the patient but was not suggested from the physical exam because
the symptoms were not reproducible with changes in head position.
• The patient became unresponsive and went into cardiac arrest that night.
Resuscitative efforts were to no avail and she was pronounced dead in
the emergency department. Her autopsy was normal and the cause of
death was believed to be from a ‘probable cardiac arrhythmia.’
• Although this was certainly a tough case and most physicians would
probably have treated the patient in the same manner, it does serve to
remind us of the difficulty in treating patients with syncope or near
syncope. The rarity of fatal cardiac arrhythmias, particularly in a patient
with very few cardiac risk factors, makes it hard to fault the ED physician
for discharging this patient. We can tell you from our ED experience that
this patient would be a hard sell for admission to a cardiologist or an
internist. The key point of her history, however, was that she had mul-
tiple episodes and this should always lower the clinician’s threshold for
considering admission. If the consulting physicians do not feel that
admission is warranted, then discuss with them about arranging for
outpatient testing as soon as possible. This would have better covered
the ED physician from a legal standpoint. Unfortunately, in this case, the
patient died later that night. The other legal issue is that the physician
made a diagnosis that was not supported from the history or physical
exam but was based on the suggestion of the patient. In the event of a
misdiagnosis, a physician should be prepared to explain how he came
up with the incorrect diagnosis. The physician in this case did not have
any evidence to support the diagnosis of labyrinthitis.
168 Learning from medical errors: clinical problems

Case 10.3 Worth watching

• Edwards demonstrates in The M & M Files that the elderly patient with
syncope should be approached with caution.17 After finishing lunch, a
79-year-old female became pale and leaned over on the table. Her past
medical history was significant for hypertension and hyper-
cholesterolemia but no prior history of syncope or coronary artery disease.
She regained consciousness after one minute and was taken to the ED
by her niece. Her vital signs and physical exam were unremarkable. The
ED physician ordered blood work, an EKG, and a head CT, all of which
were normal. He then discharged the patient with ‘probable vasovagal
syncope, post-prandial’ and told her to follow up with her own
physician in three to five days for further testing.18
• The patient collapsed the following day and was pronounced dead in
the emergency department. Her family contacted an attorney to file a
lawsuit. This case was alarming to us because we have been through a
few almost identical scenarios in our career with the exception of the bad
outcome. It is difficult to convince a physician to admit a patient when
all of the labs are normal and the patient feels great. Since it is not
possible to admit all of these patients, the physician must offer
alternative treatment options to the patient. These may include follow-
ing up with their physician the next day, scheduling tests or monitors as
an outpatient, or asking their physician to see them in the ED. It is
probably not acceptable in the legal community, however, to not discuss
the treatment plan with the primary physician. Particularly in this case
where the patient had no prior history of syncope and had risk factors
for cardiac disease.

References
1 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 352.
2 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 355.
3 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 353.
4 (2003) Clinical Update, Complex Causes of Syncope. Emergency Medicine: Acute
Medicine for the Primary Care Physician. 35(9): 44.
5 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 352.
6 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 354.
7 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 355.
8 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive
study guide (5e). McGraw-Hill, New York, New York, 356.
9 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St Louis, Missouri, 1580.
 Syncope 169

10 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St Louis, Missouri, 1578.
11 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St Louis, Missouri, 1579.
12 Rosen P, Barkin R et al. (1998) Emergency Medicine: concepts and clinical practice (4e).
Mosby, St Louis, Missouri, 1580.
13 Schwartz GR (1999) Principles and Practice of Emergency Medicine (4e). Lippincott
Williams & Wilkins, Philadelphia, Pennsylvania, 1074.
14 (2003) Clinical Update, Complex Causes of Syncope. Emergency Medicine: Acute
Medicine for the Primary Care Physician. 35(9): 44.
15 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 76–9.
16 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 187–9.
17 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 196–7.
18 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 196.
This page intentionally left blank
Chapter 11

Vaginal bleeding

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

This is ‘one of the ten most frequent complaints seen in the emergency depart-
ment. In premenarchal and postmenopausal patients, vaginal bleeding is almost
always the result of a non-hormonal cause.’1 Consequently, evaluation in these
patients should focus on organic causes of the bleeding such as infections,
malignancies, trauma, and bleeding disorders. In fact, Hamilton makes two
dictums that we feel are important for the emergency physician to remember:
‘Vaginal bleeding in the premenarchal patient is sexual abuse until proved
otherwise,’ and ‘postmenopausal bleeding is caused by malignancy until proved
otherwise.’2 Less common causes like infections should also be considered because
gonococcal infection was determined to be the etiology of vaginal bleeding in
30% of sexually active patients with vaginal bleeding in one study.3
Determination of pregnancy status is the main priority in the evaluation of
vaginal bleeding in women of child-bearing years. This is because of the
potential of life-threatening hemorrhage from pregnancy-related complications
such as abortion, ectopic pregnancy, and placental abruption. When the patient
is not pregnant, the physician must entertain other causes of vaginal bleeding
such as the ones discussed in the preceding paragraph. However, hormonal
causes of bleeding (i.e. ovulatory, anovulatory) represent a large percentage of
these patients.

Triage
Initial assessment should focus on hemodynamic stability. Patients with
abnormal vital signs (i.e. hypotension, tachycardia, fever), orthostasis, ill
appearance, abdominal pain, or objective evidence of heavy bleeding should be
brought to the physician’s attention immediately. Those with stable vital
signs who appear comfortable and have no objective evidence of heavy
bleeding can be seen on a less urgent basis. Caution should be taken with the
triage of pregnant patients due to the possibility of severe hemorrhage from
172 Learning from medical errors: clinical problems

pregnancy complications. If in doubt with these patients, bring them back

immediately.
Remember the psychological aspect of this complaint. Regardless of the degree
of bleeding, bleeding itself is a major concern to most individuals. No-one wants
to be left waiting while they are bleeding. Indeed, we have heard many
complaints from patients concerning this issue. Therefore, no bleeding patient
should be triaged as non-emergent and treated without urgency.

History
The most important historical point to ascertain in the patient with vaginal
bleeding is the potential of pregnancy. We consider any female between the ages
of 12 and 50 that has not had a hysterectomy as being capable of pregnancy. It
is important to ask if females within this subgroup have had tubal ligations,
are using birth control pills, are using intrauterine devices, or are using any
other form of contraception. Although all of these factors make it unlikely that
a person is pregnant, we have seen exceptions to all of them. We have also
discovered that pregnancy status based on the patient’s last menstrual period
is often unreliable.
Ask about the onset, extent, and nature of the bleeding. When did the patient
start to bleed? Was it associated with any trauma (e.g. sexual activity)? What
relation is the bleeding to her last menstrual period? Does she have a history of
normal or irregular periods? Try to quantify the amount of bleeding by the number
or pads that the patient is using or have her compare it with her normal menses.
Assessing for symptoms from excessive bleeding is equally important. Does the
patient have shortness of breath, light-headedness, syncope, or fatigue? Lastly,
what kind of blood is present? Is the blood dark or is it bright red? Does it contain
clots or tissue?
For the patients who state that they have a history of abnormal periods, a
detailed past medical history is required. Does the patient have a history of any
thyroid or eating disorders? Has there been a significant change in weight
recently? Is there a personal or family history of fibroids or bleeding disorders?
How about any disease that may be associated with anovulation such as poly-
cystic ovarian syndrome?
Medication history is extremely pertinent in the patient with vaginal bleed-
ing. Is the patient taking birth control pills or getting Depo-Provera shots? This
knowledge not only provides the physician with potential sources of the bleed-
ing but also gives clues on the pregnancy status of the patient. Could the bleeding
be a manifestation of oral anticoagulants?
The final components of a typical history should also be obtained. The
gynecologic history is obviously paramount in the assessment of vaginal bleed-
ing. A detailed menstrual and pregnancy history should be obtained. The blood
type of the patient is necessary in the event of a miscarriage. Is there a history
of ovarian cyst, fibroids, or endometriosis? Has the patient ever had pelvic
inflammatory disease (predisposition to ectopic pregnancy)? Social history is
based on the sexual history of the patient. Surgical history should focus on prior
surgeries for excessive bleeding (thus, providing a clue on the severity of the
bleeding) and prior surgeries for ectopic pregnancies (providing a clue on
possible recurrence).
 Vaginal bleeding 173

Physical exam
Bleeding has a direct effect on the hemodynamic status of a patient. Therefore,
importance of the five vital signs cannot be overly emphasized. In addition, all
bleeding patients should be assessed for orthostasis. Patients with abnormal
vital signs or those who are having heavy bleeding should have their vital signs
rechecked frequently.
Although the evaluation of vaginal bleeding is usually focused on two areas
– the abdomen and the pelvic areas – a few other areas should be checked for
hemodynamic stability. This includes observing the general appearance of the
patient for pallor, fatigue, or volume depletion. The lungs and heart should also
be assessed for compensatory responses to fluid depletion.
The abdomen is palpated for tenderness, distention, or masses. Abdominal
tenderness along with vaginal bleeding in a patient that may be pregnant is an
ectopic pregnancy until proven otherwise. Distention or masses could be from
pregnancy, fibroids, endometriosis, or rarely from ovarian cyst and/or torsion.
The pelvic examination for vaginal bleeding consists of four parts. The first
is the inspection of the external genitalia for any sources of bleeding. This part
is usually more important in the premenopausal female where external trauma
and foreign bodies are common sources of bleeding and in the postmenopausal
female where atropic vaginitis is a common source of bleeding. The second part
is the speculum examination. This part is important for several reasons. It
enables the physician to visualize the vaginal walls and the cervix for possible
sites of bleeding. The entire cervix should be visualized because this enables the
physician to look for the bluish discoloration seen with pregnancy or for cervical
opening. These findings will sometimes help the physician make a quick
diagnosis without diagnostic tests. Please see Case 11.4.
The third part of the pelvic examination is the bimanual exam. This exam
allows for the assessment of the uterine size and the palpation of the adnexas.
We agree with the many authors who state that the examination of the female
lower abdomen is incomplete without a bimanual pelvic examination. This exam,
however, should be used with caution in patients with second or third trimester
pregnancies due to the risk of inducing placental bleeding.
The final part of the pelvic examination for vaginal bleeding is the rectal exam.
This is the most omitted part that we have seen clinically. Although it rarely
contributes to useful information, it takes less than three seconds to perform. In
the event that an ectopic pregnancy or pelvic mass is missed, the plaintiff
attorney may argue that it could have been detected on physical examination of
the rectum. He will also accuse the physician of performing an incomplete exam.

Differential diagnosis
Hamilton categorizes vaginal bleeding into one of six disease processes:1 ‘These
are pregnancy-related, neoplastic, infectious, traumatic, hormonal, and
hematologic.’ Vaginal bleeding during pregnancy can result from spontaneous
abortion, gestational trophoblastic disease, placental abruption, placental
previa, and preterm labor. Tumors that cause vaginal bleeding include: germ
cell tumors, ovarian and uterine tumors, and fibroids. Vaginitis and cervicitis are
two common infectious causes of vaginal bleeding. Bleeding from genital trauma
174 Learning from medical errors: clinical problems

may come from sexual assault, aggressive intercourse, child abuse, and straddle
injuries to the perineum. An occult source of traumatic vaginal bleeding is the
placement of foreign bodies. Hormonal etiologies are thyroid dysfunction,
ovarian cysts, and exogenous hormonal use. Finally, bleeding may arise from a
coagulopathy or a blood dyscrasia.

Diagnostic tests
Pregnancy should always be the first concern in all patients with vaginal bleed-
ing who are capable of pregnancy. Hence, a pregnancy test is mandatory for
these patients. Although a serum pregnancy has greater sensitivity over a urine
pregnancy, both are acceptable in the medical literature and the urine point-of-
care pregnancy test will often give a quick answer. We do recommend that a
serum pregnancy be performed if the patient has had a negative urine
pregnancy test at home and suspicion is still present. Furthermore, a
quantitative human chorionic gonadotropin should be ordered for pregnancies
that are in the first trimester.
A complete blood count may be ordered if the history or the physical exam
is consistent with excessive blood loss. If the patient is on Coumadin or has other
sources of bleeding or bruising, a coagulation panel may be checked. Thyroid
function tests and tests for follicle-stimulating hormone or luteinizing hormone
levels can be ordered at the physician’s discretion. Type and screening for blood
should be reserved for patients with heavy vaginal bleeding or pregnant
patients with unknown blood type and a possible spontaneous abortion.
Vaginal and cervical swabs and cultures should be performed in all sexually
active females.
Do all pregnant females with vaginal bleeding require an emergent pelvic
ultrasound? This important question has raised great controversy and debate
among emergency physicians. The shortage of ultrasound technicians has made
justification of its emergent use paramount. That is to say, will our immediate
management be dictated by the findings on the ultrasound? If the emergent
treatment is not altered, it is hard to justify (to hospital administration and to
the radiology department) calling the ultrasound technician into the hospital.
This statement is particularly true in our current environment where there is a
shortage of ultrasound technicians and their reluctance to remain at hospitals
where they are abused.
Every pregnant female that comes to the emergency department with vaginal
bleeding is concerned about the status of her pregnancy and is under high
emotional stress. This emotional state must be dealt with with great sensitivity
if an emergent ultrasound is not ordered. To the woman or her significant other
involved, the viability of the fetus may be of utmost ‘emergency.’ Also,
remember that pregnancy represents a high medical liability situation because
two lives are involved and the cost of pain, suffering, and productivity of a
newborn is usually enormous.
In general, early second trimester to 24-week pregnancies with vaginal bleed-
ing do not require emergent ultrasound because the management is unlikely to
be altered. Theoretically, ectopic pregnancies do not continue to exist beyond the
first trimester and fetuses do not survive outside of the womb before 24 weeks
so treatments for ectopic pregnancy or emergent delivery are not expected.
 Vaginal bleeding 175

However, there are exceptions to this rule that the emergency physician should
keep in mind. The patient with heavy vaginal bleeding, low hemoglobin, or
abnormal vital signs may require an emergent dilatation and curettage. This
patient should probably receive an emergent ultrasound to document fetal
demise. The patient with no prenatal care should be regarded to be unreliable
regarding the dating of the pregnancy and an emergent ultrasound should be
strongly considered.
The other group of patients with vaginal bleeding during pregnancy who
usually do not require an emergent ultrasound is the one with first trimester
bleeding and a previous ultrasound showing intrauterine pregnancy. Again,
ectopic pregnancy and fetal viability are not theoretical concerns in
these situations. However, as Case 11.1 below illustrates, the accuracy of
prior office ultrasounds must be assessed. Furthermore, as Case 11.2 demon-
strates, an ultrasound can prevent a physician from making an incorrect
assumption.
How about using ultrasound for evaluating patients who have vaginal bleed-
ing who are not pregnant? Ultrasound is excellent for imaging the pelvis and has
good sensitivity for detecting uterine fibroids and ovarian cysts or torsion. It
may also give clues to the diagnosis of endometriosis.

Hospital/office course
Fluid resuscitation and maintaining hemodynamic stability is one of the
primary goals in patients with vaginal bleeding. Once these are obtained, it must
be determined whether the patient requires emergency surgical therapy (e.g.
salpingectomy, hysterectomy, dilatation and curettage, etc.). Medical therapy
for vaginal bleeding consists of hormone treatment (for dysfunctional uterine
bleeding), methotrexate (for stable ectopic pregnancies), blood transfusion (for
excessive blood loss), and Rhogam (for spontaneous abortion of O negative blood
type patients).
Stable patients with nonemergent causes of vaginal bleeding can be treated
as outpatients. Make sure that the patient is not orthostatic or symptomatic upon
discharge and has close follow-up with a gynecologist. Many pregnant patients
with bleeding will have a very early pregnancy and an indeterminate
ultrasound. Explain clearly to these patients that they may still have an ectopic
pregnancy and must follow up with the obstetrician in two days for
further testing. They also must understand to return immediately if the
bleeding increases, the pain worsens, fevers develop, or orthostatic symptoms
occur.

Pitfalls
Consider the presence of heterotopic pregnancy (intrauterine and extrauterine
pregnancy) in patients without prior ultrasound confirmation. Tintinalli reports
in Emergency Medicine: a comprehensive study guide an incidence of 1 per 4,000
pregnancies for heterotopic pregnancies.4 However, ‘patients with a history of
in vitro fertilization may have an incidence as high as 1 per 100 to 200 pregnan-
cies.’4 Case 11.3 provides a good example of how the failure to consider the
possibility of a heterotopic pregnancy can lead to disastrous results.
176 Learning from medical errors: clinical problems

Consideration for heterotopic pregnancy is especially important in the case of

elective abortions. We have experienced that the majority of abortion clinics will
perform vacuum curettage of a first trimester pregnancy without a prior ultra-
sound evaluation for ectopic pregnancy. Their clinic protocol usually involves
examining the evacuated tissue to confirm products of conception but this
evaluation may sometimes be inaccurate or omitted. Furthermore, a complete
and accurate examination does not eliminate the possibility of a heterotopic
pregnancy. Please see Case 11.3 below. Also remember that patients with vaginal
bleeding after an elective abortion can have other more common etiologies such
as uterine perforation, vascular injuries, retain products of conception, cervical
lacerations, and uterine infections.
Remember that the reliability of ultrasounds is very operator-dependent and
technique-dependent. Our experience has taught us that there can be great
variability in ultrasound results based on the experience and skill of the
technician and the technique that they use. If the patient’s history or physical
exam suggests otherwise, consider the utilization of repeat ultrasounds as the
case in the following section illustrates (Case 11.1).

Errors and interesting cases

Case 11.1 I thought it was in . . .

• A 30-year-old female presented with vaginal bleeding and stated that
she was 15 weeks pregnant (dating from two previous ultrasounds that
her obstetrician had performed in her office). Ectopic pregnancy was not
a priority consideration based on the estimation of gestational age and
on the previous ultrasound results. In addition, heterotopic pregnancy
was also unlikely based on the ultrasound results. We initially did not
feel that an ultrasound was needed emergently since she was not bleed-
ing heavily and it would not change our management. However, she
presented during the afternoon hours when the ultrasound technician
was present and we wanted to give the patient a definitive answer. We
received a surprise when the ultrasound showed a 15-week ectopic
pregnancy. This case crossed the borders of medicine as taught in the
textbooks because multiple ultrasounds had missed the ectopic
pregnancy. In addition, the ability of an ectopic pregnancy to exist
beyond the first trimester without rupturing is extremely rare. The
hypothesis from the ultrasound technician is that the obstetrician
performed the ultrasounds without the patient’s bladder being distended
with fluid and this led to the incorrect interpretations of an intrauterine
pregnancy on both occasions. This case taught us that office ultrasounds
are not as accurate as hospital ultrasounds. In fact, we have frequently
ordered hospital ultrasounds on patients who have had office ultrasounds
at their obstetrician’s request.
 Vaginal bleeding 177

Case 11.2 Make sure it is over

• Edwards presents in The M & M Files a case of a woman with vaginal
bleeding who was mistakenly believed to have a fetal demise.5 A 30-
year-old woman with a seven weeks intrauterine pregnancy (confirmed
by ultrasound at six weeks) presented to the ED with several hours of
vaginal bleeding and abdominal cramping. The hospital, like most other
hospitals, did not have ultrasound availability at the late hour at which
the patient presented.
• A small amount of blood in the vaginal vault and a closed cervix was
noted during the speculum exam. The ED physician also removed what
he believed was products of conception and sent it to pathology. Since
the bleeding and pain had stopped and the cervical os was closed, the
physician diagnosed the patient with a completed abortion and went on
to fill out a fetal death certificate.
• After discussing the case with the patient’s on-call obstetrician, it was
decided to start the patient on Methergine and have her follow up in the
obstetrician’s office. The patient faithfully adhered to these instructions
and took the medicine and saw the obstetrician two days later. There, she
was informed that the pathology report stated that no products of
conception were seen. Her obstetrician performed an ultrasound in the
office that showed a viable fetus with good cardiac activity.
• In his discussion, Edwards makes note of two areas where the ED
physician may have made a mistake. The first is assuming that the material
that he removed from the vaginal vault were the products of conception
and not decidual uterine material. Making this clinical distinction is
sometimes difficult for an obstetrician, much less an emergency
physician. ‘The diagnosis of completed abortion, therefore, should be
withheld unless an undeniably obvious fetal shape is seen or an ultra-
sound is performed.’6 The second fault is the ED physician’s decision to
use Methergine for a woman that was not having significant hemorrhage.
The benefit-to-risk ratio of using this drug in this scenario is extremely
low and justification for the drug’s use is difficult. Edwards, however,
defends the physician’s decision not to call in the ultrasound technician
in this event based on her prior ultrasound results.

Case 11.3 Two for one

• A 22-year-old woman came to our ED two days after having an elective
abortion at approximately eight weeks by her last menstrual period. She
was complaining of lower abdominal cramping and moderate vaginal
bleeding. Her arrival was late in the evening when the clinics were closed
and no records were obtainable. Although her vitals signs were stable,
she appeared in moderate distress secondary to the cramping and the
bleeding. Pelvic examination revealed moderate steady bleeding from

continued
178 Learning from medical errors: clinical problems

her cervical os with some small clots. Her uterus also felt extremely tender.
An emergent ultrasound was ordered which showed the presence of
both an ectopic pregnancy in the right adnexa and the presence of retained
products of conception in the uterus. The patient was taken immediately
to surgery and had uncomplicated removal of an unruptured ectopic
pregnancy and complete evacuation of her uterus.

Case 11.4 Do not wait for tests

• This case is from our own files and is not really a case of medical error.
Nevertheless, we included it because it is a good teaching case. Ectopic
pregnancies are like walking time bombs. There are many pregnant
women out there who are walking around with ectopic pregnancies.
Indeed, there are instances in the emergency department where we send
women home after telling them that they may have an ectopic pregnancy
and be in perfect compliance with the standard of care. This occurs when
the pregnancy is extremely early (less than four to five weeks gestational
age) for the ultrasound to be diagnostic. Fortunately, at this early stage,
ectopic pregnancies that do rupture rarely, if ever, cause fatal
exsanguination. More advanced ectopic pregnancies, however, represents
an entire different story. We have seen patients undergo rapid
hemodynamic decompensation from a ruptured ectopic pregnancy.
Therefore, it is necessary to make the diagnosis of these pregnancies as
quickly as possible.
• A case in point involves a 21-year-old female who came to our
emergency department complaining of vaginal bleeding and lower
abdominal pain. She stated that she had her period two weeks ago and
did not believe that she was pregnant. She was mildly hypotensive on
presentation at 96/70 mm Hg. She was also mildly tachycardic at 106
beats per minute. Her abdominal exam revealed tenderness and guard-
ing over the suprapubic area. On pelvic examination, we noticed a bluish
discoloration of the cervix. There was also moderate bleeding from the
patient’s cervix.
• With the combination of these physical findings, we immediately called
the obstetrician for a possible ruptured ectopic pregnancy versus
threatened abortion. The obstetrician took the patient to the operating
room without the assistance of any diagnostic tests from the emergency
department. An intraoperative ultrasound confirmed the presence of a
ruptured ectopic pregnancy. Confirmation of ectopic pregnancy with
ultrasound in the emergency department should be reserved for patients
who are hemodynamically stable.
 Vaginal bleeding 179

Case 11.5 What are we having?

• In the files of interesting cases, there are some pregnant women who will
present with fictitious vaginal bleeding. On our obstetrics rotation in
residency, this occurred several times a week. A pregnant female will
present complaining of vaginal bleeding or abdominal pain. Her physical
exam will reveal no abdominal tenderness or vaginal bleeding. The patient
will then request an ultrasound. When it is performed, her main concern
for the technician is whether the baby is a boy or a girl. The labor and
delivery nurses that we worked with used to write the presenting
complaint of these patients on the board as ‘rule out boy or girl.’

References
1 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 677.
2 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 688.
3 Hamilton GC, Sanders AB, Strange GR et al. (2003) Emergency Medicine: an approach to
clinical problem-solving (2e). Saunders, Philadelphia, Pennsylvania, 678.
4 Tintinalli JE, Kelen GD and Stapczynski JS (2000) Emergency Medicine: a comprehensive:
study guide (5e). McGraw-Hill, New York, New York, 742.
5 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 97–9.
6 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 98.
This page intentionally left blank
Chapter 12

Wounds and lacerations

• Triage
• History
• Physical exam
• Differential diagnosis
• Diagnostic tests
• Hospital/office course
• Pitfalls
• Errors and interesting cases

Wounds and lacerations are extremely common complaints in emergency

medicine and urgent care practices. While most are minor and simple to treat,
some may be threatening to life or limb. They also represent a significant
percentage of malpractice cases. Liability does not necessarily have to result from
death or loss of limb. Other complications that have resulted in legal action
include loss of function, visible deformity, chronic pain, and foreign body
retention.
Many mechanisms can produce a wound or a laceration to the skin. Insect
bites and foreign bodies may cause external breakage of the skin while infected
follicles and sebaceous cysts are internal factors that may violate the skin. In
addition, wounds may start deep within internal organs and may spread
systemically to the skin (i.e. herpes zoster) or deep within the muscles and fascia
(i.e. necrotizing fasciitis).

Triage
Most wounds and lacerations can be triaged as urgent or non-emergent. There
are certain groups, however, who require emergent care. These include:
amputations, wounds with extensive or arterial bleeding, heavily contaminated
wounds, and wounds with signs or symptoms of neurovascular compromise.
In addition, patients with abnormal vital signs such as fever, hypotension, or
tachycardia require attention as soon as possible.
The triage nurse is responsible for covering the wound appropriately to
prevent further contamination of the wound and to slow any possible bleeding.
Some wounds may require splinting to decrease the associated pain or to
prevent the development of further injury. The nurse should also be instructed
to perform periodic checks on neurovascular and bleeding status of wounds if
the patient is waiting for a prolonged time. The triage nurse can also administer
the tetanus immunization if necessary when the physician is comfortable with
this delegation of duty.
182 Learning from medical errors: clinical problems

History
Ask the patient (or those who witnessed the incident) about the details of the
injury. The mechanism of injury is important to assess for expected and potential
injuries. The surroundings of the injury are important to assess for possible foreign
bodies and wound contamination (i.e. broken glass, dirty environment). The
time of injury is important for deciding the most appropriate wound care. The
immunization history and the possibility of intentional injury should be
addressed. Finally, check with your local laws because some wounds (e.g. gun-
shot wounds, knife wounds) require mandatory reporting to the local law
enforcement agencies.
Ask the patient about neurologic or vascular symptoms that may have arisen
from the injury. Any new numbness, weakness, limitations of motion, or
discoloration could be clues to a neurovascular compromise. The physician
should also inquire about associated injuries that may have been overlooked by
the main injury. Patients who state that they may have a foreign body in the
wound are usually accurate. How the patient has taken care of the wound is also
important to document. For example, if the patient has left a wound exposed to
a dirty environment for 12 hours without cleaning it, there is a significantly
increased chance of wound infection regardless of impeccable wound care by
the physician.
Assess for potential problems with wound care. Is the patient allergic to any
local anesthetic? Is the patient on a blood thinner? Does the patient have diabetes
or any other immunosuppressive state? Could the patient’s current medications
interact with the antibiotic that the physician is going to prescribe? Does the
patient smoke cigarettes or have a history of peripheral vascular disease or poor
wound healing? Has the patient had problems with wounds becoming keloids
in the past?

Physical exam
Vital signs are often incomplete or not taken in patients with wounds or
lacerations. We do not recommend this practice because any of them may be a
marker of physical stress on the body and a more serious underlying injury.
Furthermore, patients with wounds and lacerations are more likely to return for
follow-up evaluations and having baseline vitals for comparison are extremely
helpful. An extended vital sign that may be helpful for certain wounds is the
ankle/brachial index (ABI). This is the ankle systolic pressure divided by the
brachial systolic pressure on the same side of the body. The ratio then serves as
an indicator for arterial disease in the specified lower extremity.
All wounds should be examined for location, size, shape, depth, foreign bodies,
and associated injuries. The characteristics of the wound should be consistent
with the mechanism and surroundings of the injury. If not, the physician should
become suspicious of the history. Use descriptive and quantifiable terms in your
notes to avoid any confusion when someone else reviews your chart. Avoid
general terms like ‘deep’ or ‘large’ when describing wounds because they leave
much to subjectivity.
Bullous lesions on the skin should alert the physician as to the possibility of
an underlying systemic inflammatory response syndrome (see ‘Pitfalls’ below).1
 Wounds and lacerations 183

They may also represent local infections such as in the case of bullous impetigo.
Other skin manifestations that may characterize a systemic disease process
include desquamation and blister formation. Desquamation may be found in
toxic shock syndrome or toxic epidermal necrolysis (TEN), while blister
formation may be found in burns and herpetic zoster infection.
As discussed in more detail in the following section, examine all wounds and
lacerations closely for associated injuries to the deeper structures. Is there injury
to nerves, tendons, vessels, etc? It is always important to determine if there is
exposed bone through a wound or a laceration. If exposure is present, then
treatment for an open fracture must be undertaken. Does the wound have necrotic
edges that require debridement? Some causes of tissue necrosis include clostridial
infections, brown recluse spider bites, and tissue that have a very poor vascular
supply.
Finally, all examinations of wounds and lacerations must be accompanied by
a neurovascular examination. Peripheral pulses and circulation must be
evaluated for compromise. Range of motion and muscle strength must be
determined for limitations. Hemostasis must be achieved or bleeding stabilized.
Serial exams of the areas may be appropriate to monitor for the development of
compartment syndrome.

Differential diagnosis
The differential diagnosis of wounds and lacerations must always extend
beyond what is seen on the surface. This means that each wound and each
laceration must be explored for foreign bodies and for injuries to underlying
structures. These structures include tendons, nerves, ligaments, vessels, bone,
and joint spaces. For some specific wounds, the depth of the wound is a huge
determinant on further therapy. For example, neck wounds that violate the
platysma muscle must undergo surgical exploration.
Wounds and lacerations should also be evaluated for associated
complications such as cellulitis, abscesses, or gas gangrene. Always remember
that some wounds may be nontraumatic and result from skin breakdown from
a systemic disease process. Examples of these include venous stasis ulcers,
decubitus ulcers, and TEN.
Some wounds and lacerations are accompanied by psychiatric and social
diagnoses and these must be determined. The classic gesture for suicidal
ideation is the laceration over the wrist. Multiple unexplained knife wounds on
the extremities may be signs of domestic abuse. Gunshot wounds may have
resulted from criminal activities. All of these other problems must be suspected
and investigated because they carry obvious legal implications.
Chronic wounds that are resistant to treatment should raise the clinician’s
suspicion of particular problems. Failure of oral antibiotics for wounds may
suggest the presence of a resilient infection such as osteomyelitis. This is because
osteomyelitis usually requires intravenous antibiotics. Lacerations that do not
close or persistently break down may be harboring a foreign body. Finally,
systemic disease processes such as peripheral vascular disease, uncontrolled
diabetes, and malnutrition may affect wound healing.
Finally, be aware that although some wounds and lacerations may appear
benign on the surface, they may be harboring serious underlying etiologies. The
184 Learning from medical errors: clinical problems

textbook example would be that of necrotizing fasciitis. The superficial appear-

ance of this disease may be normal or relatively benign as underlying tissue is
undergoing rapid destruction. Puncture wounds are another example. Here, a
small break of the skin is the only clue that a deep layer abscess is present. In
the specific case where you have a high injection puncture wound such as that
from a paint gun, there may be severe pressure necrosis of underlying tissue.
Finally, lightning injuries may also have a benign superficial appearance,
masking severe underlying muscle destruction.

Diagnostic tests
X-rays may be performed to look for fractures and foreign bodies. Remember
that metal and glass are usually seen, while plastic, wood, and other organic
materials are not.2 X-rays can also be obtained to look for gas in the soft tissue
of wounds. The presence of gas may signify a more serious infection such as
clostridial myonecrosis or gas gangrene. Alternative tests include Computed
Tomography, magnetic resonance imaging, or ultrasound. A complete blood cell
count along with wound and blood cultures should be considered for all
patients who are candidates for hospital admission.
As discussed in the previous section, there are additional tests that may be
appropriate for specific situations. A triple phase bone scan is much more
sensitive for ruling out osteomyelitis than X-rays. Arterial/brachial indices may
be taken (see above) but arterial evaluation with color flow Doppler or
angiography is much more sensitive and may be needed in those with severe
peripheral vascular disease. Serum creatine phosphokinase and urine myoglobin
may be helpful in those with suspected muscle injury. Lastly, tissue biopsies may
be utilized to confirm systemic disease processes.

Hospital/office course
Certain types of wounds require consultation with a wound specialist. James
Palombaro, MD lists several in Just the Facts in Emergency Medicine.2 These include
any wound that: involves the tarsal plate of the eyelid or lacrimal duct, involves
an open fracture or joint space, is associated with multiple trauma that need
surgical admission, involves the face and require extensive plastic
reconstruction, is associated with amputations, is associated with loss of
function, involves tendons, nerves, or vessels, and involves a significant loss of
epidermis.2
Documentation of wound care is extremely important because of the certainty
of a small percentage of wounds that will have complications. There is a familiar
saying in medicine these days that states that ‘the only way to never have a
malpractice suit is to never see a patient.’ We believe that an analogy can be
made for wound care. The only way to never have a wound complication is to
never take care of one. Therefore, Guidner and Leinen write, in the ‘Legal
Corner’ of EMpulse, that there are four essential elements to document in wound
care.3 The first is the examination for foreign bodies. The second is that copious
irrigation was used for the wound. The third is the status of the neurovascular
examination. Tendon integrity is the final component to address.
We have a minimal standard documentation regimen for all wound repairs.
 Wounds and lacerations 185

This regimen consists of five steps. Step 1 is to state that the wound was prepped
and draped in normal sterile fashion. Step 2 is to state the type of anesthetic used.
Step 3 is to state that the wound was thoroughly irrigated with normal saline.
Step 4 is to describe the exploration of the wound for foreign bodies and
associated injuries. This also includes whether any debridement or manipula-
tion of the wound was performed. Finally, Step 5 is how the wound was closed
(specifically, the type and number of sutures used). Adhering to these five steps
reminds us to perform each step with every wound repair. In addition to the five
steps, tetanus immunization status must be documented.
In repairing wounds that are over cosmetically sensitive areas, patients should
have explained the risk for scarring. This explanation should be documented in
the chart. There are patients who return surprised when their repair becomes a
scar and are upset that they were never told that it could happen. Furthermore,
cosmetically sensitive areas may vary among patients, so try to get a feel if
appearance is going to be a concern. We repaired a laceration on the back of
the hand of a 62-year-old lady once who asked ten times if there was going
to be a scar on her hand. We respect this concern although it is not typical or
expected.
Discharge instructions for wounds and lacerations are extremely important to
prevent complications from being undetected. Most of us are used to telling the
patient to return or call if having increasing pain, fever, swelling, redness, or
discharge from the wound. However, as Case 12.11 below shows, asking the
patient to call back may sometimes get you into more trouble.

Pitfalls
On occasions, patients will present with lacerations that they request a plastic
surgeon to repair. The types of lacerations for these requests will range from a
complicated laceration on the face to a simple laceration on the arm. Regardless
of the patient’s insurance status and your personal opinion of the laceration, you
should honor the request and make the call to the plastic surgeon. The surgeon
may refuse or state that they are busy and may take a while to respond (they are
unlikely to leave their office or surgery to come suture a patient). You can relay
this conversation to the patient and many will change their mind and have you
suture them versus waiting. The liability in not making the phone call arises
when the appearance is not to their liking and they sue you for the cosmetic
result from not having a plastic surgeon perform the closure.
Be particularly careful with the management of puncture wounds. They often
appear benign on the surface but harbor acute emergencies underneath. It is
difficult to estimate the depth involved in a puncture wound. Likewise,
assessments of injuries to tendons, nerves, and vessels are also challenging. Special
mention should be made of paint gun injuries due to the chemical irritation and
the possibility of developing compartment syndrome. Please see Case 12.5 below.
Also, as Case 12.9 shows, underestimation of the depth of puncture wounds
may be severe and lead to disastrous consequences.
Always keep the potential for intentional injury in the back of your mind for
patients with wounds and lacerations. Classic injury patterns like lacerations on
the wrist should raise suspicion despite the patient’s denial of intentional injury.
If repetitive questioning is not utilized in these situations, intervention for
186 Learning from medical errors: clinical problems

suicide prevention will be missed. The subsequent legal payout could be huge
as Case 12.1 in the following section illustrates.
With the exceptions of bullous impetigo and erysipelas, be aware of bullous
lesions as manifestations of serious disease processes. These include both
infectious processes such as necrotizing fasciitis or gas gangrene and non-
infectious processes like toxic epidermal necrolysis. Stevens advises that
‘violaceous bullae should prompt the physician to look hard for necrotizing
fasciitis or, alternatively, gas gangrene.’4
We believe that necrotizing fasciitis is one of the most difficult diagnoses to
make in medicine. It may present with fever and deep-seated pain as the only
initial complaints and have no cutaneous signs on physical exam.5 ‘An alarming
50% of patients with an ultimate diagnosis of group A streptococcal necrotizing
fasciitis do not have a portal of entry.’6 The diagnosis is often missed during the
initial presentation because the infection is located within the deep tissues and
not obvious by visual inspection. ‘Frequently, these patients are in shock and
have organ failure by the time a diagnosis is made.’6
Retained foreign bodies are major constituents of lawsuits concerning wound
care. ‘The most common foreign body in a wound is soil. Clay-contaminated
soils and soils with large amounts of organic material have a high potential for
infection.’7 They are problematic for several reasons. They are associated with
an increasing frequency of wound infections and abscesses. They may move and
subsequently cause neurovascular compromise. Finally, they may cause chronic
pain or, in the case of road grit, produce disfiguring wounds with tattooing.8 We
have discussed the importance of performing and documenting a thorough search
for foreign bodies in every wound. Please see Case 12.2 below.
The other types of foreign bodies that may be problematic are the sharp ones
such as glass. If undetected or not removed, they may sever nerves, tendons,
ligaments, and vessels. In an article published in The American Journal of Diseases
in Children, Avner and Baker recommend that X-rays should be ordered with any
wound with glass involvement with the exception of very superficial wounds.9
Please see Case 12.6 below. As a final reminder, always consider the possibility
of a foreign body if the instrument causing the laceration is uncertain. Please see
Case 12.12 below.
In addition, lacerations caused by sharp objects such as glass should be
assumed to be associated with nerve, tendon, and vessel injuries until proven
otherwise. This usually means a detailed neurovascular examination but may
include color Doppler ultrasound studies or angiograms in some cases. Please
see Case 12.7 below.
Wounds and lacerations in young children can be problematic in many aspects.
Young children have fear of the unknown and will often not stay still for a good
examination or for wound or laceration repair. They may have injuries in areas
that are considered sacred for them such as the genital or rectal area. Please see
Case 12.3 below. They may have been instructed by their parents to not allow
strangers to touch them in these areas. Sedation (general anesthesia in
some instances) is often required for these examinations and many
physicians are hesitant about sedating children. Furthermore, some parents
are fearful of sedation for their children and will refuse it. In these instances,
we will often defer the examination of genital wounds and lacerations in
young females to the gynecologist because we know they will allow only one
 Wounds and lacerations 187

physician examination and it is best to have the most experienced one perform
it.
If possible, always follow up yourself or have someone from your group follow
up your wound repairs. Through the years, we have encountered a number of
post-surgical complications in the office or in the emergency department.
Whenever these complications occur, it is always difficult to get another surgeon
to see the patient if the operating surgeon is not available. In general, no-one
wants to be ‘dragged into someone else’s mess.’ The dictum also often applies
vice versa as many physicians do not wish other physicians to interfere with
their work. The second physician may not be familiar or approve of the tech-
nique that was used. She may make the wound worse or be critical of the wound.
In addition, many physicians want to see the end result of their work (to obtain
satisfaction from a job well done or to learn from mistakes). Therefore it is best
to follow up your own wound and laceration repair in your own office. Please
see Case 12.4 below.
The invention of Dermabond made the treatment of simple lacerations quicker
and less painful. It is easy to use and is a very strong adhesive once applied. Its
applications, however, are limited to simple, straight, clean, wounds that are
under minimal or no tension. Wounds that are contaminated, complex, or deep
are usually not good candidates for Dermabond. Please see Case 12.8 below.
Wounds and lacerations over joints should be treated with caution because of
the possible development of joint infections. Septic joints may be as occult as
necrotizing fasciitis and are sometimes misdiagnosed as cellulitis, gout, or bursitis.
A misdiagnosis, however, is not acceptable because of the extreme morbidity
associated with septic joints. Please see Case 12.10 below.

Errors and interesting cases

Case 12.1 Cannot read minds

• We read about a case where a man presented with bilateral wrist
lacerations. He stated that they occurred after he was grabbing for some
pipes with both hands. He was questioned by the triage nurse and by the
emergency physician separately whether the injuries were intentional.
He gave adamant denials to both parties and these were documented in
the chart. The patient was treated and then found dead three days later
after he had hanged himself. The patient did not have any family but a
probate lawyer who came to his estate contacted a plaintiff attorney to
file a suit on the estate’s behalf. The physician and the hospital were both
subsequently found not at fault because of their thorough documenta-
tion. The courts deemed the physician as unable to be a ‘mind reader’
for his patients.
• We mentioned some types of wounds that are occasionally accompanied
by legal issues. In managing these wounds, the physician must
anticipate the legal issues and document carefully. The verdict in the
above case would probably have been against the physician if she had
not been thorough with her history and documentation.
188 Learning from medical errors: clinical problems

Case 12.2 Pollution is not the solution

• Edwards presents a case, in The M & M Files, where a wound repair had
an adverse outcome. The case is also an example of poor chart documen-
tation.10 A 20-year-old college student was pushed off his bicycle as he
was riding along a canal trail. He fell into some bushes and suffered
multiple facial lacerations. The patient was taken to the emergency
department where the lacerations were sutured and he was given routine
follow-up instructions.
• The patient returned to the ED two days later with a low-grade fever and
swelling and redness over his wounds. The second ED physician removed
the sutures and noticed a moderate amount of purulent discharge from
the wounds. There were also bits of wood and bark in the wounds. When
he probed the wound further, he was able to extract four or five twigs
(the largest being 1.8 cm in length). He irrigated the wounds and left
them open. He then admitted the patient for intravenous antibiotics. The
wounds were secondarily closed by a plastic surgeon five days later but
the result was unsatisfactory and the patient pursued a lawsuit against
the first ED physician.
• The initial ED physician did not document in his chart any exploration
of the wound for foreign bodies. He also did not consider other treat-
ment options such as X-rays or a consultation with a plastic surgeon for
multiple wounds on the face (see above discussion). Finally, Edwards
brings up the discussion of prophylactic antibiotics in this case. While
there is controversy concerning the routine use of prophylactic
antibiotics, there are certain specifics of this case that would probably
make most physicians more prone to use it. The patient’s accident occurred
on soil and in an environment filled with organic material. Both
scenarios are more likely to lead to a wound infection. Furthermore,
multiple wounds over a cosmetically sensitive area warrant, at a
minimum, consideration of the use of prophylactic antibiotics.

Case 12.3 They may not let you look

• Although wounds and lacerations in children are sometimes difficult to
examine, the physician must find alternatives and must not abandon
care. Medical and legal risks will arise for those physicians who choose
to do so. Edwards presents a case, in The M & M Files, of a seven-year-
old girl who fell off monkey bars and landed on steel pipes between her
legs.11 She was brought to the ED with blood soaked towels between her
upper thighs and did not want to lie on the stretcher.
• The patient resisted the examination by the ED physician. Even with the
assistance of the child’s mother and two nurses, a suboptimal examination
was performed before the mother’s anger ended the efforts. The physician
noted a laceration to the labia minora. There was no active bleeding. He
then waited to see if the child could urinate. Shortly later, she did and

continued
 Wounds and lacerations 189

was also walking without any problems. There was a small amount of
blood in her urine. The physician then discharged the patient and
instructed the mother to have the child follow up with her pediatrician
as needed.
• The mother brought the child to her pediatrician after she noted that the
child had continued to ooze blood from the perineum throughout the
night. After the pediatrician had asked a gynecologist to see the patient,
the patient was taken to the operating room under conscious sedation
to have a laceration on her introitus repaired with eight sutures. Both the
mother and the gynecologist addressed their concern to the hospital that
the ED physician did not perform an adequate examination.
• The sedation of children is not a common occurrence unless you work
in a pediatric emergency department or see a high volume of pediatric
patients. This process is almost never met with total ease and comfort.
In this case, however, discharging a patient from your emergency
department with an unknown source of bleeding is more ‘gut wrench-
ing’ than the use of sedation. What if this child had bled to death at home
from a pelvic hematoma that suddenly became unstable? How about a
laceration that is not repaired in a timely fashion and leads to permanent
deformity or infection? Sexual or urinary problems may also be
complications. Imagine the financial payout for such a lawsuit. The take
home message is that the physician should do what needs to be done,
regardless of her discomfort in doing it.

Case 12.4 Check on your work

• Edwards presents another case, in The M & M Files, of a wound compli-
cation that could have been avoided if the physician had followed up on
his repair.12 A three-year-old boy suffered a small 1.5 cm scalp laceration
after tripping and hitting his head on a television set. There were no
neurologic deficits and the boy was behaving normally, so the ED
physician decided to clean the wound with peroxide and close it with
two staples. He felt that this was quicker and avoided the necessity of
giving the boy a painful injection of lidocaine. Everyone appeared to be
satisfied with the wound repair.
• When the mother took the boy to his pediatrician’s office five days later
to have the staples taken out, they found out that the pediatrician did not
have the equipment or the experience to remove the staples. After
obtaining a staple remover from an emergency department, the
pediatrician seemed to have problems using it and the wound was soon
bleeding and the child was screaming. The child was sent to see a
surgeon that afternoon and he also had problems taking out the staples,
which had now seem to become intertwined subcutaneously. Under
conscious sedation, the surgeon was able to open the wound and remove
the staples. The wound was then closed secondarily. The mother called
the hospital to complain and refused to pay the hospital bill.

continued
190 Learning from medical errors: clinical problems

• In our opinion, two errors were made in this case. Neither of the errors,
however, involved the medical care of the emergency physicians. Small
scalp lacerations are usually closed with staples. The good intentions of
avoiding a painful lidocaine injection or the risk associated with
conscious sedation for a minor laceration is totally appropriate. This is
witnessed by the satisfaction at discharge and the absence of a wound
repair complication. The first error, however, was in the ED physician’s
disposition instructions. He failed to advise the mother to bring the boy
back to the ED for staple removal. Although most physicians know how
to remove staples, this case was the rare exception that the ED physician
did not expect. The second error involves the pediatrician’s failed
attempt at removing the staples. Even a relatively easy procedure like
this should probably not be attempted if you have never done it before
or do not have the proper supervision.

Case 12.5 Guns can cause serious injuries

• Selbst and Korin present a case, in Preventing Malpractice Lawsuits in
Pediatric Emergency Medicine, where a painter injected paint thinner into
his fingertip with a high-pressure paint gun. 13 The ED physician
discharged this relatively benign appearing puncture wound. He returned
a few hours later and required surgical exploration, irrigation, and
eventual amputation of the finger. The ED physician was found liable
for $77,500 for the patient’s injuries. It is important to remember that
surgical consultation is mandatory for all foreign body injection injuries
because compartment syndrome and tissue necrosis are significant risks.
• A related situation that we have seen a handful of times in our careers
is the infiltration of contrast dye from an intravenous line during a CT
scan procedure. We have discussed the management of these cases with
numerous surgeons and have not received a consensus treatment plan.
Nevertheless, these cases should be discussed with the on-call surgeon
or plastic surgeon.

Case 12.6 Be sharp with glass injuries

• Another case from Selbst and Korin’s Preventing Malpractice Lawsuits in
Pediatric Emergency Medicine concerns a female who suffered permanent
injury from a retained foreign body.14 After stepping on a piece of glass
on the beach, a world-class athlete came to the ED where two physicians
were unsuccessful at removing the foreign body. This resulted in
permanent nerve damage in her foot with numbness and ‘clawed toes.’
She, subsequently, filed a lawsuit against the hospital and the ED
physicians and was awarded $5.2 million. A retrial later reduced the
award to $350,000.
• A thorough and extensive attempt should be made to remove any sharp

continued
 Wounds and lacerations 191

retained foreign body. This is due to the high potential of the object injuring
adjacent neurovascular structures. If repeated attempts are unsuccess-
ful, consider consulting a surgeon, using a fluoroscopy-guided technique,
or splinting the area to prevent movement of the object.

Case 12.7 Cutting edge injuries

• Also from Preventing Malpractice Lawsuits in Pediatric Emergency Medicine
by Selbst and Korin is a case of a complicated wound cause by a glass
injury.15 A 10-year-old boy was brought to the ED by ambulance after he
injured his foot by kicking a plate glass window. Although the paramedics
noted substantial blood loss, the ED physician’s note did not reflect this.
The physician sutured a blood vessel, closed the wound, and discharged
the patient.
• He was unable to move his foot later. During surgery, he was found to
have severed four tendons and a sustained a partially lacerated artery.
The physician was accused of performing an inadequate examination
and not detecting the patient’s neurovascular deficits. A jury award of
$238,000 was given to the plaintiff.
• Physicians sometime become distracted with the repair of wounds and
lose focus of the secondary injuries. The ED physician in this case
probably did not perform a neurovascular examination or he would have
detected the child’s tendon injuries. He was focused on repairing the
injured vessel that he saw and assumed that this was the lone source of
the bleeding. This assumption is dangerous to make, particularly in the
presence of an injury that is caused by a sharp object. His defense would
have been much stronger and even, perhaps, victorious had he
documented the absence of any of these complications on his examination.

Case 12.8 Do not seal the poison

• A five-year-old boy was brought to the ED after getting bitten by the
family dog. The wound appeared straight on the lower cheek but it was
difficult to determine the depth of the wound. The ED physician closed
the wound with Dermabond, placed the patient on Keflex, and gave
specific instructions to return if signs of infection developed.
• Two days later, the patient returned with increased facial swelling. He
had formed a large facial abscess with fluctulance and required drainage
from a plastic surgeon. The purulent material could not drain because
the Dermabond had formed such a locked-in seal. Culture from the wound
drainage grew the organism Alcaligenes, which was resistant to Keflex.
• The complications from this case could have occurred regardless of the
degree of wound care. However, certain precautions should be taken in
managing animal bites, especially of the face. Dermabond is probably
not a good choice for closure because it is difficult to predict the

continued
192 Learning from medical errors: clinical problems

cleanliness or the depth of the wound. Specific antibiotics (as recom-

mended by the Sanford Antimicrobial Guide) should be used. In this case,
although Keflex is an excellent antibiotic for most skin infections, it is not
the best choice for animal bites. Finally, animal bites to the face have the
potential for disastrous cosmetic deformities. Therefore, complications
must be detected as early as possible and re-evaluation in 24 hours is
prudent.

Case 12.9 How deep is it?

• In another case from Selbst and Korin’s Preventing Malpractice Lawsuits
in Pediatric Emergency Medicine, the importance of misjudging the depth
of a puncture wound is emphasized.16 After an 18-year-old man was
stabbed by an ice pick, the ED physician discharged him with the
diagnosis of ‘superficial wound’ and told him to return if the symptoms
worsened. He returned three hours later with vomiting and abdominal
pain. He was found to have a duodenal perforation during surgery and
had it repaired. The surgeon, however, missed a second perforation to
the duodenum. He also missed lacerations to the right kidney and right
renal vein. These lesions led to continued blood loss and the patient’s
cardiac arrest four hours later.
• The ED physician and the hospital were sued for misdiagnosis and
delayed treatment. The surgeon was also sued for missing the other
injuries. Although the hospital settled before the trial, the jury verdict
was in favor of the physicians. This decision was surprising to us based
on the above details. We feel that puncture wounds to the chest, back,
or abdomen should be treated with extreme caution because of the
potential of organ injury and excessive bleeding. If the depth cannot be
clearly visualized, consider diagnostic tests such as CT scan or ultra-
sound.

Case 12.10 Neighboring injuries

• Another example from Selbst and Korin’s Preventing Malpractice Law-
suits in Pediatric Emergency Medicine involves a wound to an area that is
susceptible to complications.17 A patient was treated in the ED for a
contusion with an anti-inflammatory agent after he fell and scraped his
knee on a nail. On a return visit, the knee was swollen and warm and
his white blood cell count was mildly elevated. A tetanus shot and an X-
ray were ordered and the patient was given antibiotics for an early
infection. Instructions were to return in 10 days for a follow-up visit.
• The patient saw an orthopedist five days later and was found to have an
infected joint. There was redness and swelling of the joint and foul-
smelling purulent discharge was aspirated. Gas gangrene had developed
within the knee joint and the patient required an amputation at the knee.

continued
 Wounds and lacerations 193

The patient filed a lawsuit against the ED physician for failure to instruct
him to return earlier to monitor the infection. Despite arguments from
the physician and the nurse that the patient was told to come back if his
symptoms worsened, there was no documentation of this. Furthermore,
the plaintiff’s expert witness argued that the patient would probably
have developed sepsis and died if he had waited 10 days. The physician
was subsequently found to be negligent and his appeal was unsuccess-
ful.
• The plaintiff’s arguments were certainly very strong and valid in this
case. This is true despite the absence of the claim that the ED physician
made a misdiagnosis. Although septic arthritis would have been an
extremely, almost impossible, diagnosis to make on the first visit, it should
have been considered on the second visit. Had it been considered, the
patient’s treatment would have been much more conservative. This would
have consisted of hospitalization for intravenous antibiotics, orthopedic
consultation, joint needle aspiration, or instructions to return in 24 hours
for recheck. Conservative treatment would have probably decreased the
morbidity of this injury.

Case 12.11 Words of faulty wisdom

• Patients are frequently told to call if their symptoms worsen. This advice
is becoming a large medical liability because of the inexperienced phone
medical advice that is given by medical staff employees. Some offices
and hospitals have frequent turnover in their staff members and have
multiple members performing phone triage. This combination often leads
to inconsistent and inexperienced phone advice that may contribute to
medical errors and medical malpractice. It has become tempting to tell
patients to come back and not call if their symptoms worsen.
• Selbst and Korin present a case in Preventing Malpractice Lawsuits in
Pediatric Emergency Medicine where an incorrect phone triage led to a
complication that was not detected.18 A boy suffered an open compound
finger fracture while playing basketball. The ED physician cleaned and
sutured the wound and gave the patient written discharge instructions.
When the boy called later to report that his finger had become swollen
and painful, a nurse told him that these symptoms are not unusual. He
received the same response the following day and was told to see his
private physician.
• When he went to see another physician, he was diagnosed with a
staphylococcal infection in the joint sheath of the finger. He was
admitted to the hospital for treatment. He later successfully sued the
emergency department for not recognizing the symptoms of an infection
and not providing proper follow-up care.
• This case was interesting in that the treating physician probably did not
treat an open fracture adequately (i.e. wound irrigation and cleansing in

continued
194 Learning from medical errors: clinical problems

the operating room and parental antibiotics). The infection most likely
resulted from this clinical failure. The physician was not sued, however,
for maltreatment. Instead, the entire department was found negligible
for wound after-care. Does that mean that a lawsuit would not have been
filed if the infection occurred but the patient was told to come back
immediately? This case should remind us that medical-legal medicine is
more than a physician’s evaluation and treatment of a patient.

Case 12.12 What was the culprit?

• Gardner and Mendelson present a case in Foresight where a laceration
from an unknown instrument results in a wound complication.19 A 24-
year-old man was walking barefoot on a beach when he fell into a hole.
A sharp, burning sensation was immediately felt in his foot. He wrapped
the foot with his shirt to control the moderate bleeding and went to the
ED.
• The ED physician found a 10 cm laceration on the plantar surface that
extended from the metatarsals to the heel. He documented that there
were normal vascular, motor, and sensory functions in the foot. The
wound was cleaned, sutured, and splinted. The patient was instructed
to return in seven to 10 days for suture removal.
• The patient returned four days later with increasing pain and oozing
from the wound. A different ED physician noticed that there was pus
draining between the sutures. An X-ray revealed a 4 cm linear foreign
body and several smaller gravel-type foreign bodies. The patient was
taken to surgery where a long sea urchin spine and several small rocks
were removed. The patient had a long and complicated hospital course.
He underwent multiple surgeries, skin grafting, physical therapy, and
was left with a limp. A lawsuit was filed against the initial ED physician.
Because of the absent documentation of the physician’s consideration of
a retained foreign body, the case was settled for $450,000.
• It was not certain what caused the laceration in this patient. The
environment surrounding the injury must be taken into consideration.
The patient was walking barefoot on a beach. There are many sharp objects
that are found indigenously or from pollution on the beach. These include
broken glass (e.g. from beer bottles) and sharp appendages from ocean
animals. Therefore, a careful search for a foreign body must be initiated.

References
1 Stevens DL (2003) Skin and soft tissue infections. Infections in Medicine. 20(10): 484.
2 Cline DM, Ma OJ, Tintinalli JE et al. (2001) Just the Facts in Emergency Medicine. McGraw-
Hill, New York, New York, 64.
3 Guidner G and Leinen A (2004) Malpractice or misfortune: we, the jury, find the
emergency physician… EMpulse. 9(1): 8.
 Wounds and lacerations 195

4 Stevens DL (2003) Skin and soft tissue infections. Infections in Medicine. 20(10): 492.
5 Stevens DL (2003) Skin and soft tissue infections. Infections in Medicine. 20(10): 492–
3.
6 Stevens DL (2003) Skin and soft tissue infections. Infections in Medicine. 20(10): 493.
7 Cline DM, Ma OJ, Tintinalli JE et al. (2001) Just the Facts in Emergency Medicine. McGraw-
Hill, New York, New York, 63.
8 Cline DM, Ma OJ, Tintinalli JE et al. (2001) Just the Facts in Emergency Medicine. McGraw-
Hill, New York, New York, 77.
9 Avner JR, Baker MD (1992) Lacerations involving glass: the role of routine radiographs.
American Journal of Diseases in Children. 146: 600–602.
10 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 135.
11 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 138–9.
12 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency
medicine. Hanley & Belfus, Inc., Philadelphia, Pennsylvania, 142–4.
13 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 76.
14 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 74–5.
15 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 75.
16 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 126–7.
17 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 147–8.
18 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 150–51.
19 Gardner AF, Mendelson DJ (2002) Avoidable. Foresight: Risk Management for Emergency
Physicians. 5(October): 3–5.
This page intentionally left blank
Chapter 13

Legal issues involving radiology

• Responsibilities of X-ray interpretation

• Making sure that you have the right X-ray study
• Should nurses order X-rays?
• Responsibilities of CT interpretation
• Loss of radiology reports
• Check radiology reports on all of your studies
• Talk to your radiologist and ask for his help
• Do not let one X-ray finding distract you
• Pediatric X-rays
• Radiology studies are not innocuous

The use of diagnostic radiology is involved in almost every field of medicine.

Its use has become increasingly popular as the improvement in technology has
made radiology equipment more compact and more affordable. Most outpatient
offices now have X-ray capabilities and some have even incorporated ultrasound
and CT scanners into their practices. For most urban emergency departments,
ultrasound, CT scanners, and even sometimes MRI (magnetic resonance imaging)
scanners have become standard equipment. Over the past few years, imaging
equipment have also become portable, as exemplified by ‘scanning fairs’ where
people can obtain whole body CT scanning from a scanner located on a large
truck.
The increased availability and use of radiologic imaging means that there are
more studies for physicians to interpret. Since some of these studies are
performed in the outpatient setting or are performed in the hospital after
‘typical’ working hours when radiologists are not present, it also means that a
significant amount of studies will be interpreted by non-radiologists. This,
undoubtedly, creates a potential for medical errors because most physicians do
not receive extensive radiology training during their residency. Those who train
in the primary care setting typically do not order enough studies to gain
valuable experience, while those who train in the hospital setting usually have
the comfort of radiologist or radiology residents to interpret their films.

Responsibilities of X-ray interpretation

In most emergency departments and outpatient clinics in the United States, the
treating physician is responsible for making the initial X-ray interpretation. Very
few centers have radiologists at their site to give simultaneous interpretations
or have the ability to send films to the radiologist’s home for interpretation.
198 Learning from medical errors: clinical problems

Therefore, the accountability of non-radiologist physicians in making X-ray

interpretations has been an issue in legal medicine for many years.
While we are certainly not held to the same standards of interpreting X-rays
as are the radiologists, we are expected to show competence in X-ray
interpretation. This competence is varied among the different medical specialties.
Although family practitioners or internists are expected to be able to recognize
the findings of a pneumonia or those of congestive heart failure on a chest X-
ray, they are less expected to detect a small pneumothorax or a
pneumomediastinum. These latter two findings typically occur in the trauma
patient and do not represent the ‘normal’ patient for the primary care physician.
The emergency physician, however, should be competent in spotting these two
findings on a chest X-ray. Similarly, the ED physician may not detect a subtle
Salter-Harris fracture that an orthopedist would be expected to see.
There are a few methods to prevent a misreading on an X-ray from turning
into a medical liability. The first and most important is to tell the patient that
your interpretation is only a preliminary reading and that you will contact her
if the radiologist discovers a different finding. The statement requires that you
make sure that you follow up on the X-ray report. Emergency physicians may
sometimes be off duty for several days in a row. Waiting until you return to the
hospital to check on a suspicious X-ray is not good practice. We realize that
sometimes it cannot be helped (e.g. vacation, report has not been read, etc.), but
you should make an attempt to obtain the report as soon as possible. We have
a dictation service at our hospital. Therefore, whenever we will not be returning
to the hospital for several days, we will write the necessary demographics to get
the report over the phone and the patient’s contact number in case we need to
reach her. In the same manner, most private offices have two methods to have
X-rays performed in the office read by a radiologist. Most are sent out by courier
to the radiologist and are read within several days. Some, however, can be picked
up by a special courier on request and taken to the radiologist for a wet read.
Obviously, this latter method involves more expenditure and cannot be requested
with every film. It should be reserved for questionable films where the treatment
would be dictated by the immediate reading. The final point is to treat the patient
conservatively if you are not sure of the X-ray. If you clinically suspect a fracture,
splint the involved area even though the X-ray is negative. If you are not sure
whether a pneumonia is present, either treat the patient with antibiotics or hold
antibiotics and bring him back tomorrow when you will have the results of the
wet read.
Another scenario that occurs commonly is when the radiologist discovers a
significant finding that was missed by the ordering physician. Since most
radiologists interpret X-rays after the disposition of the patient has been already
made, this delayed finding may develop into legal liability. Who is responsible
for informing the patient about this finding?
The following is how the readings of emergency X-rays are read in our facility.
We believe that the process in not unlike that of most other facilities. However,
as we will point out, there are many potential flaws with this process. After an
X-ray is performed, the ED physician interprets it and makes her treatment and
disposition based on the reading. The patient is often told that the reading is
preliminary and that they will be contacted if the radiologist disagrees with it.
Hours to days later, after the film has been read by a radiologist, the dictated
 Legal issues involving radiology 199

report is placed in the ordering physician’s mailbox. The radiologist typically

does not call the physician with every positive finding but only those that may
be life-threatening.
The most obvious pitfall is the delay in the radiologist’s interpretation of the
film. This means that fractures may go without splinting, pneumonias may not
be treated, foreign bodies not removed, etc. The delay may be compounded if
the ED physician does shift work and may be away from the hospital for several
days. Furthermore, for emergency departments that have temporary physicians,
these physicians may not have mailboxes and may never see an official report.
Also, in EDs that have more than single physician coverage, reports occasionally
go to the wrong physician’s mailbox.
Communication among physicians who review the X-ray is also not optimal
for the prevention of medical error. As we mentioned, the radiologist will usually
call the physician with life-threatening findings such as a cervical spine fracture
or a pneumothorax. Fractures, foreign bodies, and pneumonias are less frequently
discussed with the ordering physician. These entities represent a significant
amount of malpractice lawsuits when not addressed in a timely and appropriate
fashion. Even when the radiologists do call the emergency department to
discuss the X-ray findings, they will often give the report to a nurse or to a
different ED physician. This creates a dilemma, in our opinion, because neither
of these individuals is aware of the patient. Without knowing the patient, they
are more likely to forget about the report if they are in the midst of doing some-
thing important. Furthermore, if they are distracted, they may feel no obligation
to take a corrective initiative (e.g. pull the chart, review it, make a callback)
because there is no patient–physician bond.

Case 13.1 Passing the buck

• A radiologist called an ED physician during one of his shifts to tell him
that a pneumonia was seen on a chest X-ray from a patient who was
treated the previous night. The ED physician meant to pull the chart and
review the treatment given but was sidetracked by a couple of critical
patients. He eventually forgot about the report. The radiologist dictated
that he gave the report to the ED physician.
• Two days later, the patient came back to the ED in severe respiratory
distress. He had not been treated for the pneumonia. Shortly afterwards,
he went into respiratory arrest and died. His family filed a lawsuit against
the initial treating ED physician, the ED physician who took the report,
and the radiologist. The ED physician who took the report was
eventually dropped from the suit because he had no relationship with
the patient.

Cases like this are controversial and may be viewed differently among courts.
Ideally, it would be best if a radiologist could read an X-ray with minimal delay
from the ordering physician’s interpretation. In the era of digital radiology and
teleradiology capabilities, we believe that this will soon be accomplished. Until
then, we must work to achieve a more consistent and cooperative protocol for
following X-rays. The ordering physician and the radiologist should bear the
200 Learning from medical errors: clinical problems

primary responsibility for contacting the patient with any significant finding
(life-threatening or not). These two physicians are the ones that are on the
patient’s chart and the ones that are billing the patient. Therefore, they have
unquestionably formed a physician–patient relationship and should assume
responsibility. Getting another physician involved who did not treat the patient
does not relieve them from this responsibility.

Case 13.2 Needed communication

• This is an example from Selbst and Korin’s Preventing Malpractice Law-
suits in Pediatric Emergency Medicine of a radiology interpretation that
was not communicated because it was not life-threatening.1 A four-year-
old child was brought to the ED with an arm injury. The physician read
the X-rays as negative and the child was discharged with an elastic
bandage. The radiologist looked at the film the following day and
dictated that there was a fracture of the distal humerus. He did not
contact the ED physician or the patient’s family. Subsequently, the child
required an open reduction and removal of calcium deposits months
later. The family sued the radiologist, who denied that he was
responsible for contacting the patient. The court, however, felt that
medical professionals should ‘coordinate their efforts in a manner that
best serves their patient’s well being.’1

We feel that there should have been a default mechanism to prevent the fracture
from not being treated for months. The ordering physician must make it a habit
to check the reports on all of his X-rays. The radiologist must also make an effort
to communicate any positive findings to the ordering physician or the patient.
In addition, another trick that we have learned in the ED is to tell the patient to
follow up with their physician in a few days and obtain the dictated report.

Case 13.3 Did you hear the report?

• In contrast, Selbst and Korin also present a case, in Preventing Malpractice
Lawsuits in Pediatric Emergency Medicine, where the radiologist was not
found liable for failure to communicate his findings.1 After an 18-year-
old came to the ED with cold symptoms and chest congestion, she was
discharged after the ED physician read her chest X-ray as normal. The
dictation from the radiologist, however, noted that there was a possible
mediastinal mass. When the patient returned to the ED to see the same
physician two days later, it was unlikely that the physician was aware
of this X-ray finding because he did not discuss it with the patient.
• The patient was diagnosed with Hodgkin’s lymphoma and had
chemotherapy that impaired her fertility. She filed a lawsuit against the
ED physician for failure to inform her of the mass and the radiologist for
failure to communicate his findings to her physicians. The plaintiff

continued
 Legal issues involving radiology 201

affirmed that an earlier diagnosis would have made her eligible for more
conservative treatment without fertility complications. The radiologist
claimed that he had no duty to make verbal contact. The ED physician,
on the other hand, claimed that the mass seen on the chest X-ray was not
the lymphoma. A settlement of $500,000 was paid by the hospital and the
ED physician while the radiologist was cleared.

It is obvious that there is no consistency with court rulings in these cases. The
two physicians in the prior case certainly each had a degree of arrogance in their
defense arguments. The legal consequences of missing the diagnosis of cancer
are so huge in the medical community that most radiologists would have
contacted the treating physician. We certainly feel that this radiologist fell below
the ‘standard of care’ in this regard. As for the ED physician, his argument that
the mass on the X-ray was not the lymphoma did not go well with the jury.
Because he missed the diagnosis, it is his burden of proof to show that this mass
was not the lymphoma (which he was unlikely to do). We can see that these
types of cases can result in variable outcomes. Regardless of the physician’s role
in treating the patient, he will improve his chance of winning a lawsuit if he
shows initiative in contacting the patient. Those that appear indifferent and those
without initiative are likely to be found liable.

Making sure that you have the correct X-ray study

We talked about the many checks that nurses must perform before administer-
ing a medication to a patient earlier in this book. Although not as complicated,
a checking system is also required for interpreting X-rays. The physician must
make sure that the film is of the right patient; the film is of the correct body part;
the film is of adequate quality (e.g. there is no jewelry to hide potential
pathology); and the film contains the adequate amount of views (e.g. a cervical
spine is incomplete without an odontoid view).
Having an X-ray with the wrong body part can lead to disastrous consequences.

Case 13.4 Choosing the right side

• An orthopedist operated on the wrong shoulder after an X-ray
technician had mislabeled a humeral fracture. The technician shot a
picture of the right shoulder but labeled it as the left. The patient did not
read the consent completely before signing it after being examined by
the physician assistant. The physician did not see the patient until the
patient was under general anesthesia and, subsequently, did not realize
the error until the left shoulder was opened. The case was settled out of
court.
202 Learning from medical errors: clinical problems

Case 13.5 Taking the right pictures

• Edwards presents another case, in The M & M Files, where the X-ray was
of the incorrect body part. A 25-year-old woman came to the ED after a
volleyball injury to her left ring finger.2 After being triaged, she was sent
directly to the radiology department to get an X-ray. The film was obtained
before the patient saw the physician. When the physician examined the
patient, he noted that there was ‘marked swelling with ecchymosis’ of
the finger and believed that an injury more serious than a sprain might
be present.
• After he read the X-ray as negative, however, he told the patient that she
had a sprain. The finger was splinted and the patient was discharged
with the instructions to wear the splint for three days. When there was
no improvement four days later, the patient went to her family
physician. Her physician obtained the hospital X-ray and discovered that
it was performed on the middle finger. After obtaining a correct X-ray,
an avulsion fracture was visualized. The patient filed a complaint against
the hospital and the ED physician. The discussion of whether the film
should have been ordered is in the following section. Although the nurse
ordered an X-ray of the incorrect body part, it is the physician’s
responsibility to make sure that the correct area was radiographed.

Should nurses order X-rays?

The freedom of triage nurses to order X-rays before the patient sees the
physician has it pros and cons. The pros are mainly to save time for the patient
and the physician. Most patients do not like to wait in the waiting room and then
wait after they are seen for the X-ray to be done. Having the X-ray ready for the
physician to interpret immediately after he examines the patient makes the
patient’s visit time substantially shorter. In addition, if an obvious finding such
as a fracture or a pneumothorax is seen, the patient can be temporarily treated
in the waiting room or can be upgraded in the triage order, respectively. We have
also discovered that affording nurses this additional responsibility usually gives
them more satisfaction in their jobs.
There are also many good reasons for restricting the ability of nurses to order
X-rays. X-rays may be ordered that are not needed (e.g. a knee X-ray when the
Ottawa criteria are not met). Resources are not used efficiently. The patient with
a more emergent need for an X-ray may have to wait unnecessarily. Cost-
effective healthcare is not being implemented. Also, the patient and the X-ray
technician are subjected to unnecessary radiation exposure.
As in the case above, the wrong body part was ordered. The physician can
avoid these errors by examining the patient first and then reviewing the X-ray
with a focus on the clinical area involved. If the area of suspect is not adequately
seen on the film, then additional films are required. The physician in the above
case did not focus on the clinical area involved when he reviewed the X-rays.
In addition, suboptimal X-rays may be ordered because the nurse does not have
 Legal issues involving radiology 203

the experience with the pathophysiology of the injury. An example is when a

patient severely twists his ankle and injures his tibia/fibia syndesmosis. The
patient may complain of pain in the mid lower leg and have the nurse order an
X-ray of the tibia and fibia. This X-ray may not detect a subtle fracture of the
ankle and will miss a Jones fracture. These are two injuries that may occur with
an injury to the tibia/fibia syndesmosis.

Responsibilities of CT interpretation
Most hospitals bestow the responsibility of CT imaging interpretations to
radiologists. Studies ordered on a stat basis are either given to the radiologist
to read or are sent through the Internet to the radiologist’s home if he is not in
the hospital. Consequently, delays in CT interpretation occur frequently. For
some patients with diseases such as acute hemorrhagic stroke and perforated
abdominal viscus, interpretations must be made as soon as possible. This leads
to the underlying question of this section: who is responsible for making these
timely CT interpretations?

Case 13.6 Read it while it is still wet

• Edwards presents a case, in The M & M Files, where a delayed CT inter-
pretation leads to an adverse outcome and a settlement against the
hospital.3 A 52-year-old man was working on a hayloft when he fell six
feet and struck his forehead on a beam. He did not lose consciousness
and was able to walk to his house after the fall but had a three cm
laceration on his scalp. He was driven to the ED by a friend and was met
there by his wife. His initial neurologic exam was normal and the ED
physician started to repair the scalp laceration. The patient then
complained of a headache. The physician ordered a head CT despite a
repeat neurologic exam that was without change.
• Because the incident occurred on a weekend, there was no radiologist in
the hospital. The study had to be sent through the internet to a larger
hospital that was 30 miles away for a radiologist to read. The larger
hospital was busier and had its own studies for the on-duty radiologist
to read so it was not unusual for the formal reading to take hours. (In
addition, radiology has become an increasingly interventional field so
the radiologist may be occupied with performing procedures.) The ED
physician did not feel comfortable in reading CT scans and decided to
wait on the formal reading before committing to any further medical
action.
• After one-and-a-half hours had passed without a reading on the patient’s
head CT, the patient started to vomit and acted strangely. His wife then
informed the ED physician that he was on a blood thinner. The physician
immediately went to re-evaluate the patient. The patient was now
sluggish in answering questions and had one larger pupil. This prompted
the ED physician to call the radiologist for an immediate reading. After
several minutes, the radiologist informed the ED physician that the patient
had a subdural hematoma.
continued
204 Learning from medical errors: clinical problems

• The patient was intubated and transported by helicopter to a larger facility

for neurosurgical care. The patient did make a recovery but was left with
permanent neurologic deficits. The hospital was contacted by the
family’s attorney concerning settlement for the delay in care and the
medical injuries resulting from it.
• What were the potentially correctable pitfalls in this case? The easiest
one to correct is to routinely ask any patient who has sustained a head
injury if they are taking anti-coagulants. Patients taking blood thinners
are at an increased risk of intracranial hemorrhage and warrant a lower
threshold for receiving a head CT (we obtain a head CT in every patient
under these conditions). In addition, the interval from physician
evaluation to obtaining the results of the scan should be kept as short as
possible. Had the ED physician in this case known that the patient was
taking a blood thinner, he might have contacted the radiologist
immediately to obtain an interpretation sooner or he may have made an
attempt to read the CT scan.
• This brings us to the second pitfall in this case. Although the interpret-
ation of head CTs can sometimes be complicated even for a radiologist
looking for subtle infarcts or bleeds, major hemorrhages or significant
brain edema are usually not difficult to detect for most emergency
physicians. This is especially true for those that are large enough to have
midline shifts and cause physical signs suggesting an impending brain
herniation. Edwards concurs with our comments by stating that ‘life-
threatening intracranial collections of blood and significant midline shifts
are easy to interpret, and are within the bailiwick of all emergency
physicians.’3 Therefore, physicians should recognize what is regarded as
‘expected knowledge’ concerning radiology interpretation because they
will likely be held responsible in the court of law. For physicians who are
not comfortable with reading CT scans, we recommend that they become
well acquainted with their radiologists and CT technicians. These
individuals can be extremely helpful in showing you the pathologic find-
ings on each study. We also try to make it a habit to review scans that
have pathology in order to increase our knowledge base. There are also
many continuing medical education courses on emergency radiology.
• The last area of pitfall in this case is the hospital’s arrangement with the
radiology group in interpreting their CT scans. It is medically dangerous
to have stat CT scans from an emergency department not read within an
hour and, preferably, these should be read within 30 minutes after the
study is performed. Although there will be some exceptions as we
discussed above, these must be kept to an extreme minimum. If a
hospital routinely takes over an hour to have a radiologist interpret CT
scans performed under emergent conditions, then it is probably placing
itself at a greater medical liability than if the hospital did not have a CT
scan. If the hospital did not have a CT scanner, then it could divert cases
like this one to a facility that is able to provide a higher level of care. Jauch
writes in Cardiovascular and Neurovascular Emergencies: implications for

continued
 Legal issues involving radiology 205

clinical practice that ‘it is the responsibility of hospital systems to provide

timely (within 25 minutes of arrival) and skilled interpretation (within
45 minutes of arrival) of CT images in acute stroke patients, and
emergency physicians should not be responsible for the interpretation of
these images.’4

Loss of radiology reports

Stat or wet reads on radiology studies are often given by radiologist to
emergency physicians. Since these interpretations are usually given verbally,
access to them is limited to the receiving physician and, at times, the dictation
line. If the physician receiving the report does not write it down or passes the
information to the next physician at sign out, reports may be lost or very difficult
to obtain. This is especially true whenever patients are transferred to another
facility and the films are not sent and the facility uses a different dictation
system.
Our friend, Jeff Abraham, MD, tells the story of a complication that occurred
during a transfer where insufficient information was passed.

Case 13.7 That information would have been helpful

• A 38-year-old nurse was treated at a community hospital after falling six
feet from a platform while painting. He hit his head and buttocks during
the fall. He had loss of consciousness, vomiting, and complained mainly
of a severe headache. He had a CT scan of the head and pelvis at the
community hospital. The radiologist read these scans and gave the
reports to the treating emergency physician. The initial emergency
physician reported to the oncoming ED physician that the head scan was
normal but the patient was still vomiting. He neglected to inform the
oncoming physician that a CT of the pelvis was performed. There was
also no record of this scan on the chart since the order and report were
taken verbally.
• Since the community hospital did not have the services of a
neurosurgeon, the decision was made to transfer the patient to a regional
medical center. The ED physician contacted the accepting physician and
relayed to him all of the details of the patient’s care with the exception
of the pelvic CT. All of the patient’s paperwork was copied and sent with
the transfer. After the patient arrived at the larger facility, he asked the
emergency staff the results of his pelvic CT. The emergency staff tried
unsuccessfully to obtain this result. The initial ED physician and the
interpreting radiologist had both gone off call and were not answering
their pagers. Furthermore, neither the second hospital nor the current
ED physician at the first hospital knew how to access the radiology
dictation system. After a two-hour delay, they were finally able to access

continued
206 Learning from medical errors: clinical problems

the dictation line to discover that the patient had suffered a nondisplaced
pelvic fracture. A pelvic vein thrombosis was also seen on the CT scan.
• The patient was started on anticoagulants and admitted for his
traumatic injuries. He had no subsequent complications from his
traumatic injuries but suffered chronic pelvic pain, which he attributed
to the delay in treating the pelvic vein thrombosis. A subsequent lawsuit
against both hospitals and emergency physicians was settled for a
substantial amount.

Check radiology reports on all of your studies

In the ideal world, physicians should look at every radiology study that they
order. We confess that this is not practical because most of us would be lost
looking at CT scans, MRI scans, or nuclear studies. However, we feel that
physicians should make the effort to look at and interpret every X-ray that they
order because it is a skill in which the public expects us to be competent. It is
also a good way to increase one’s knowledge base. The clinician would be
surprised how often he will find something that the radiologist has missed. As
the clinician, you have the advantage of a clinical history and physical
examination.
With the above comments being said, it is not good enough to just look at the
film. Because most of us are not radiologists, we will frequently miss X-ray details
that are not common or obvious. These details may be inconsequential (e.g.
accessory rib) or may have great significance (e.g. pneumothorax). There also
may be findings that are inconsequential now (e.g. pulmonary nodule) but may
have great significance in the future (e.g. lung cancer). Regardless of the nature
of these findings, it is the ordering physician’s responsibility to trace down this
report.

Case 13.8 Hunt down reports

• This case involves a lost radiology report from Physician’s Medical Law
Letter.5 A man presented to the ED with abdominal bloating, weakness,
and vomiting. Abdominal and chest X-rays were done in the ED. He was
then admitted by his family physician. The patient was treated for anemia
and gastrointestinal bleeding while in the hospital and for two months
after hospital discharge. His family physician did not receive the
radiology report, which disclosed the presence of an abdominal aortic
aneurysm.
• The patient returned to the ED months later with continued
gastrointestinal bleeding. His aortic aneurysm was still not picked up
and he was sent home. The aneurysm ruptured a few hours later and the
patient underwent an emergency surgery. The surgery, however, was too
late and the patient suffered end-organ damage from the massive blood
loss. His inevitable death occurred shortly afterwards.
continued
 Legal issues involving radiology 207

• The plaintiff experts argued that the family physician was at fault for not
picking up the aneurysm earlier. The family physician, in turn, blamed
the hospital for not getting the report to him. The jury determined that
both parties were held liable and each was given a percentage of the
fault. This case involves several aspects of a suit that is difficult to win.
The plaintiff was successful at creating dissension among the defendants
and having them work against each other. The jury questioned the
attempts that the family physician made in obtaining the radiology
report. Finally, the plaintiff could have probably successfully sued the
ED physician for missing the aneurysm on the film that he had ordered.

Case 13.9 Treating something that is not there

• Another case involves a 72-year-old female who was brought into our
ED with shortness of breath. She had no history of heart or lung
problems and had been hospitalized for six days the previous week for
‘pneumonia.’ There was no history of tobacco use. Her initial oxygen
saturation was 93% on room air but the rest of her vital signs were
unremarkable. Scattered expiratory wheezes were present but there was
no evidence of consolidation. Her chest X-ray showed a small hyperdense
area next to the lower right heart border. The rest of her labs and her EKG
were unremarkable.
• The chest X-ray finding did not strike us as a typical pneumonia. Her
blood work also was not suggestive of this diagnosis. We reviewed her
hospital chart and found out that she had received daily chest X-rays
while in the hospital and was indeed treated for pneumonia. All five
chest X-rays were interpreted by the radiologist as normal with the
exception of a pericardial fat pad. The admitting physician had apparently
never reviewed these reports. This made us skeptical about the
diagnosis of pneumonia. We ordered a CT pulmonary study. The patient
was subsequently found to have multiple pulmonary emboli in both
lungs including one in the right pulmonary artery. She was also found
to have a deep venous thrombosis in her leg. Although these findings
may all have developed after or during the hospitalization, the
misdiagnosis of pneumonia is suspected based on the radiology reports.

Case 13.10 Going on a search mission

• The final example is one from our own files. We had a pregnant patient
in our practice that was about six weeks pregnant based on her last
menstrual period. We had ordered an outpatient ultrasound (at the
obstetrician’s request) one week before her first scheduled prenatal
appointment. The patient went to the local hospital’s radiology
appointment to obtain her ultrasound. Through the years, we have
continued
208 Learning from medical errors: clinical problems

discovered that most ultrasound technicians are adept at interpreting

their scans and will relay any important findings to the physician
immediately. This technician performed the scan, which showed an
ectopic pregnancy, but chose to tell the patient to go home after the
procedure rather than detaining her. This was despite the fact that there
was a request for a ‘wet read’ on the ultrasound order.
• After we had not heard a report three hours after her scheduled
procedure, we telephoned the radiology department to check on the
results. The scan was then given to the radiologist who confirmed that
an ectopic pregnancy was seen. We immediately tried to call the patient’s
home phone number but it had been disconnected. We then called her
next contact number and reached a neighbor. The neighbor said that they
lived in the country and that their houses were 10 miles apart. She did
not have transportation to go to the patient’s house. As a last resort, we
drove to the patient’s home, which was one hour from the hospital, and
arrived at her house at 9:30 p.m. After walking through three huge bark-
ing dogs, we finally reached her doorstep and informed her to return to
the hospital to be treated by an obstetrician. The importance of getting
a radiology report in preventing a fatal complication for this woman was
the lesson of this case.

Talk to your radiologist and ask for his help

Our radiology colleagues are almost like computers for two reasons. They are
extremely informative and can always give you a huge differential concerning
a particular radiographic finding. This differential will sometimes include
diagnoses that are extremely obscure and unusual; however, it will also
contain the correct diagnosis the majority of the time. Therefore, it behooves you
to ask them what a finding could mean or represent. There are also times when
they will expand their list of differential diagnoses after receiving some
more clinical data from you. Constructive conversing is beneficial to both
clinicians.
Radiologists are also similar to computers because they are sometimes only
as good as the information that is given to them. It is frustrating to evaluate a
vague patient because it is like working in the dark. In the same manner, asking
a radiologist to evaluate a study without giving adequate clinical information
also imposes a broad and difficult task for him. For example, some physicians
do not realize that some CT studies consist of hundreds of axial slices. That is
a lot of pictures to evaluate and scan in detail. You are more likely to
improve your yield if you can give your radiologist a direction on which to
concentrate.
 Legal issues involving radiology 209

Case 13.11 Getting by with a little help from your friends

• The following is a good case from our own clinical experiences. A 56-
year-old woman presented to the ED with the complaint of sudden onset
of epigastric pain 20 hours prior to her presentation. She had nausea but
no associated fevers, vomiting, or diarrhea. There was some radiation of
the pain to her back. She had a history of thyroid cancer many years ago
but denied any chronic medical illnesses or prior abdominal surgeries.
• Although her vital signs were unremarkable with the exception of mild
hypertension, she appeared in moderate discomfort. The ED physician
ordered a cardiac evaluation along with liver function tests and lipase.
The tests were all negative and the patient received transient relief with
a GI cocktail. However, the pain recurred and appeared to be more
severe. The ED physician then administered intravenous Demerol and
Phenergan and ordered a CT scan of the abdomen to look for an
abdominal aortic aneurysm. The radiologist reported the CT scan as
normal.
• At shift change, the patient still had abdominal pain that was severe and
of uncertain etiology. Because of the first physician’s continuing
concern, we ordered an abdominal ultrasound to evaluate for gallstones.
We also called the oncoming radiologist to inform him that the patient
was having pain in the epigastric area. Although the ultrasound was
positive for gallstones, the more significant finding was a 2 cm hypodense
lesion at the head of the pancreas. Both of these findings were present
on second review of the CT scan. The patient was admitted and a sub-
sequent biopsy of the lesion confirmed a carcinoma of the pancreas.
Clinically, it was difficult to distinguish whether the gallstones or the
tumor was responsible for the patient’s pain. An answer to the possible
cause of the patient’s pain, however, could have been provided to the
patient based on the CT scan if more clinical information was provided
to the first radiologist.

Discussing the clinical case with the radiologist can also be helpful because it
gives him a chance to recommend additional tests.

Case 13.12 Can we see what we need to?

• Edwards presents a case, in The M & M Files, where a discussion with
the radiologist might have prevented a misinterpretation of an X-ray.6 A
32-year-old man was brought to the ED in full spinal immobilization and
complaining of bilateral ankle pain. He did not have any head or neck
injury and denied any other pain.
• The ED physician ordered X-rays of the neck and both ankles. The
radiologist was still present in the hospital and sent a wet reading of the
X-rays, which were normal. The ED physician reassured the patient with

continued
210 Learning from medical errors: clinical problems

these reports and ordered elastic wraps for both of the patient’s ankles.
The patient was discharged in a wheelchair because he could not bear
any weight on the left ankle or heel. He was instructed to see his physician
during the next office day if there was no improvement.
• After the weekend had passed, the patient could still not bear weight.
His family physician noted that his right heel was now grossly swollen
and ecchymotic. The physician became suspicious and reviewed the X-
rays from the hospital. He noticed that the calcaneus did not appear
normal and ordered calcaneal views of the foot, which confirmed the
presence of a fracture.
• The patient complained to the hospital administration and threatened to
sue the hospital if he was billed. He questioned why a neck X-ray was
ordered when he was not hurting there and why calcaneal views were
not obtained. He was extremely upset that he was out of work for more
than three months. Although the patient’s disability would probably not
be altered had the fracture been detected earlier, the patient’s frustration
might have been avoided if the ED physician and radiologist had
examined the specific area of injury with finer detail. This might have
included the addition of calcaneal X-ray films.

Do not let one X-ray finding distract you

This is one of the fundamental rules that you learn on your radiology rotation.
We are all taught to perform a history and physical exam in a systematic fashion.
Likewise, radiologists are taught to look at a film in a similar fashion. As a
consequence, you will often obtain a radiograph report that lists incidental find-
ings that you were not looking for. The most common example in our practice
is the finding of a pulmonary nodule on a chest X-ray that was ordered to look
for an acute pulmonary process. Acute findings on X-rays may sometimes be
missed if the clinician becomes distracted by other findings.

Case 13.13 Only seeing what you want to see

• We had a recent case that highlights the above discussion. A 67-year-old
male with a history of chronic obstructive pulmonary disease and lung
cancer went to his primary physician’s office with shortness of breath
and cough. He had developed increasing shortness of breath for the past
three days and had an increased cough with purulent sputum
production. He was afebrile in the office and the physician ordered a
chest X-ray. The physician was concerned that the patient was
developing pneumonia and commented on his note that there were no
infiltrates seen on the X-ray. He was also impressed with the significant
amount of scarring seen on the patient’s lungs. The patient was discharged
with antibiotics for the treatment of acute bronchitis.

continued
 Legal issues involving radiology 211

• The radiologist called the clinic the following day and reported that there
was a new development of a 75% pneumothorax on the patient’s right
chest. The patient was contacted by the clinic and instructed to go to the
emergency department. We saw him there and placed a chest
thoracostomy tube in his right chest, which re-expanded his lung. The
patient’s primary physician was so focused on looking for hyperdensities
(e.g. infiltrates) and distracted by the pulmonary scarring that he over-
looked the hyperlucencies (e.g. pneumothorax).

Pediatric X-rays
Pediatrics films can be especially challenging. Children still have growth plates
and injuries through these areas can be damaging to long-term growth and may
be difficult to detect. Furthermore, some are almost impossible to see on X-ray
(Salter-Harris 1). Physicians who must interpret pediatric films must understand
the classifications and implications of the Salter-Harris fractures. Indeed, missed
pediatric radiographic interpretations represent a significant percentage of
malpractice claims and many medical centers now employ a pediatric
radiologist to interpret these films.

Case 13.14 More at stake

• Starr presents a case, in Cortlandt Forum, where a missed interpretation
led to a large settlement.7 One morning, just after returning from a
vacation a few hours earlier, an experienced radiologist was given an X-
ray of a pelvis and femur from a 12-year-old boy. The boy had complained
of left leg pain with activity for months and was taken by his mother to
see his family physician. His physician, in turn, had ordered these X-rays
to ‘rule out developmental and congenital disorders.’8
• The radiologist had interpreted these films as normal so the family
physician reassured the mother and told her that the boy could return
to full activity. Over the next 22 months, the boy’s pain became
progressively worse, causing his physician to order hip X-rays. There
was now clear evidence of damage to the articular cartilage and
chondrolysis. These changes caused limitation of hip motion and a
permanent deformity. The child was most likely going to need a hip
replacement in the future.
• The parents obtained a personal injury attorney and filed a lawsuit against
the radiologist. Another radiologist reviewed the initial films and
testified that there was bilateral grade 1 slipped capital femoral
epiphysis. The defense, however, obtained their own expert radiologist
who countered by saying that ‘missing the subtle changes on the film
was still within the standard of care and did not signify negligence.’9 The
case dragged on for approximately five years. The patient was now a

continued
212 Learning from medical errors: clinical problems

teenager and had failed reconstructive surgery and would permanently

walk with a cane and require hip replacement every 10 years. These
injuries led to an eventual settlement of $2.7 million that was to be spread
throughout the patient’s lifetime to help cover the cost of his medical
expenses.
• This case is certainly a difficult one, not only for those involved, but also
for the readers of the case. Without a doubt, an error was made and this
left a patient permanently injured and worthy of financial
compensation. It is controversial, however, whether the error by the
radiologist was preventable. Nevertheless, it reminds us that pediatric
films usually entail higher medical risks than adult films because of the
increased difficulty in interpretation and the greater potential of
permanent injury if the pathology is not detected.

Radiology studies are not innocuous

Every physician and almost every patient are aware of the radiation exposure
of radiology studies. With the exception of MRI and ultrasound, the other
common radiology imaging modalities all carry some degree of radiation
exposure. This will frequently become an issue with pediatric patients (effects
on bone growth) and pregnant patients (effects on fetal organ development). In
fact, some centers have protocols requiring pregnancy tests before any imaging.
While radiation effects are seldom overlooked, other potential dangers of
radiologic studies are often underestimated. The risk of anaphylaxis to
intravenous contrast dye is significant and must be ascertained in every patient.
For patients with brittle congestive heart failure, an intravenous contrast dye
load may tip them over into critical hemodynamic decompensation.

Case 13.15 Did not take much to tip over

• We saw a patient during our residency for shortness of breath and
pleuritic chest pain. The radiologist read her chest X-ray as normal. After
she had received 125 ml of intravenous contrast for a CT pulmonary
study, she went into respiratory arrest and had copious frothy sputum
during the intubation. We discovered later that she had a history of flash
pulmonary edema. Fortunately, she recovered uneventfully.

The chemical effects of the dye may be physically detrimental. We have seen a
handful of cases of infiltrating intravenous lines that leaked contrast dye into
the subcutaneous tissue. This could lead to the severe complication of tissue
necrosis from a severe inflammatory reaction. Contrast dye may also tip a
patient with borderline renal function into acute renal failure. Lastly, contrast
dye may be excreted in breast milk and is harmful to breast-fed infants if this
activity is not ascertained.
 Legal issues involving radiology 213

Case 13.16 Contrast dye is a prescription medicine

• This case is from our own practice and it serves as a good reminder that
when you order a radiology study requiring contrast dye, you are
essentially prescribing the contrast dye (just like any other medicine)
and you should be aware of its dangers. A 36-year-old woman was sent
by her obstetrician to see us after she started complaining of shortness
of breath. She had just had a cesarean section three days prior and her
physician was concerned about a pulmonary embolus.
• When we saw her, she also had 2–3 plus pitting edema in both lower
extremities and was slightly tachycardic but not hypoxic. We informed
her that we shared the concern of her obstetrician and would like to
order a test to check for pulmonary embolism. We asked her if she was
breast-feeding and she said yes. She was then told that the contrast dye
is usually out of her system in one day but we recommend that she hold
breast-feeding for two days. Immediately, she burst into tears and did
not want to stop breast-feeding.
• Searching for another option, we went to the radiology department and
asked the radiologist about a ventilation-perfusion scan for this
evaluation. The radiologist checked with the technician and told us that
the dye used for nuclear scan would require a minimum of three days
to get out of her system.
• Still searching for alternatives for this patient, we called the obstetrician
for suggestions. He felt that one of the two tests had to be done and the
patient had to make an informed decision. Subsequently, we gave the
important data to the patient and she chose to have the CT scan.
Incidentally, it was normal and the patient felt that it was unnecessary
in retrospect. However, we were able to sleep better at night knowing
that the test was normal.

References
1 Selbst SM and Korin JB (1999) Preventing Malpractice Lawsuits in Pediatric Emergency
Medicine. American College of Emergency Physicians, Dallas, Texas, 127.
2 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency medi-
cine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 149–50.
3 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency medi-
cine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 80–82.
4 Jauch EC (2004) Cardiovascular and Neurovascular Emergencies: implications for clinical
practice. EMCREG-International, Cincinnati, Ohio, 64.
5 Priest W (2003) If a physician does not receive a copy of an abnormal X-ray and ra-
diology report, who is liable – the physician or the hospital? Physician’s Medical Law
Letter. December: 2–3.
6 Edwards FJ (2002) The M & M Files: morbidity and mortality rounds in emergency medi-
cine. Hanley & Belfus Inc., Philadelphia, Pennsylvania, 145–6.
7 Starr DS (2004) Exhausted physician misses a key diagnosis. Cortlandt Forum. 17(4): 86–
7.
8 Starr DS (2004) Exhausted physician misses a key diagnosis. Cortlandt Forum. 17(4): 86.
9 Starr DS (2004) Exhausted physician misses a key diagnosis. Cortlandt Forum. 17(4): 87.
This page intentionally left blank