THE AMERICAN
JOURNAL of
MEDICINE
®
The Green Journal
EDITORIAL
287 Peer Review: The Best of the Blemished?
Joseph S. Alpert
CONTENTS
April 2007 Volume 120/Number 4
UPDATE IN OFFICE MANAGEMENT
306 Office Management of Chronic Pain
in the Elderly
Debra K. Weiner
REVIEWS
289 Judicious Antibiotic Use and Intranasal
Corticosteroids in Acute Rhinosinusitis
Catherine Butkus Small, Claus Bachert,
Valerie J. Lund, Augustine Moscatello,
Anjuli S. Nayak, and William E. Berger
Even though many cases are secondary to viral
infections, acute rhinosinusitis is often treated with
antibiotics. Recent studies have documented that inhaled corticosteroids in combination with antibiotics
or as monotherapy provide significant symptom relief and resolution of acute rhinosinusitis.
295 Guideline for Low-Cost Antimicrobial Use in
the Outpatient Setting
Lizbeth A. Hansen, Lee C. Vermeulen,
Sarah Bland, and Tosha B. Wetterneck
The authors have created a guideline for the use
of less expensive but efficacious antibiotics in the
management of common infections.
303 Contemporary Empyema Necessitatis
Scott A. Kono and Trenton D. Nauser
Empyema infection may extend beyond the pleural
space and into the chest wall. Early diagnosis,
combined with effective antimicrobial therapy
as well as early surgical drainage can result
in marked improvement in prognosis for these
patients.
A6
DIAGNOSTIC DILEMMAS
316 Too Much of a Good Thing
Lisa Call Pastel and Jonathan M. Ross
319 A Case for Vaccination
Aima Ahonkai and Noah Lechtzin
IMAGES IN DERMATOLOGY
322 Mystery of the Silk Road
John Patrick Welsh, Christopher B. Skvarka,
Christine Ko, and Carrie Ann Cusack
ECG IMAGE OF THE MONTH
325 A Racing Heart
Julia H. Indik
IMAGES IN RADIOLOGY
328 A Complication of Forceful Nose-Blowing
Amy Fix and Valerie J. Lang
CLINICAL RESEARCH STUDIES
330 Vitamin and Micronutrient Intake and the Risk
of Community-Acquired Pneumonia in US
Women
Mark I. Neuman, Walter C. Willett, and
Gary C. Curhan
364 Outbreak of Pertussis on a College Campus
Allen S. Craig, Seth W. Wright,
Kathryn M. Edwards, John W. Greene,
MaryLou Haynes, Anthony D. Dake, and
William Schaffner
Pertussis is not merely an infection of young children. The authors report a substantial outbreak
of pertussis infections on a university campus.
Guidelines for diagnosis and therapy are suggested.
In a large population of healthy women, higher
intake of vitamins from diet or supplements did
not decrease the risk of pneumonia. Among
smoking women, higher intake of foods rich in
vitamin E did reduce risk of community-acquired
pneumonia.
337 Grip Strength Predicts Cause-Specific Mortality
in Middle-Aged and Elderly Persons
Hideo Sasaki, Fumiyoshi Kasagi,
Michiko Yamada, and Shoichiro Fujita
Maximum grip strength is an accurate and consistent predictor of all causes of mortality in
middle-aged and elderly individuals. This presumably reflects their level of frailty.
343 Thyroid Hormone Use, Hyperthyroidism and
Mortality in Older Women
Douglas C. Bauer, Nicolas Rodondi,
Katie L. Stone, and Teresa A. Hillier
AJM ONLINE
CLINICAL RESEARCH STUDIES
369 Prognostic Stratification of Patients with LeftSided Endocarditis Determined at Admission
José Alberto San Román, Javier López,
Isidre Vilacosta, Marı́a Luaces, Cristina Sarriá,
Ana Revilla, Ricardo Ronderos,
Walter Stoermann, Itziar Gómez, and
Francisco Fernández-Avilés
The prognosis for patients with left-sided endocarditis can be predicted from easily obtained
clinical variables including the presence of heart
failure and myocardial abscess at the time of
presentation.
The use of thyroid hormone replacement in older
women carries no prognostic significance. However, previous history of hyperthyroidism may be
associated with a small increase in all-cause and
cardiovascular mortality.
350 TB in a Low-Incidence Country: Differences
Between New Immigrants, Foreign-Born
Residents and Native Residents
Gerd Laifer, Andreas F. Widmer,
Mathew Simcock, Stefano Bassetti,
Andrej Trampuz, Reno Frei, Michael Tamm,
Manuel Battegay, and Ursula Fluckiger
369 Does Statin Therapy Decrease the Risk for
Bleeding in Patients Who Are Receiving
Warfarin?
James D. Douketis, Magda Melo, Chaim M. Bell,
and Muhammad M. Mamdani
Long-term statin therapy was associated with a
reduced risk of bleeding in patients who were
receiving warfarin for atrial fibrillation.
New immigrants may harbor active but clinically
quiescent tuberculosis infections. Rapid diagnostic tests may be negative, and drug resistance is
more common in these patients than in nativeborn individuals.
The following articles are available only in
the online version of the Journal.
357 Coronary Ischemic Events after First Atrial
Fibrillation: Risk and Survival
Yoko Miyasaka, Marion E. Barnes,
Bernard J. Gersh, Stephen S. Cha, Kent R. Bailey,
James B. Seward, Toshiji Iwasaka, and
Teresa S. M. Tsang
New onset of atrial fibrillation is a marker for the
presence of coronary artery disease. In women,
new atrial fibrillation in the presence of ischemic
heart disease augurs a worsened prognosis.
CLINICAL COMMUNICATION TO
THE EDITOR
e1
An Unusual Case of Dysphagia and Chest Pain
in a Non–HIV Patient: Esophageal Tuberculosis
Bobbak Vahid, Nazmul Huda, and Ali Esmaili
A9
ERRATA
e3
Errata
e17 Porcelain Gallbladder and Cancer: Ethnicity
Explains a Discrepant Literature?
Steven C. Cunningham and H. Richard Alexander
LETTERS
e5
Dosing Frequency of Aspirin and Prevention of
Heart Attacks and Strokes
John K. Amory and David W. Amory
e7
The Reply
James E. Dalen
e9
Enteric Coated Aspirin
Charles M. Grossman
BRIEF OBSERVATION
370 Late Diagnosis of HIV Infection: The Role of Age
and Sex
Michael J. Mugavero, Chelsea Castellano,
David Edelman, and Charles Hicks
Delayed diagnosis of HIV infection is more common in the elderly and in women. It is important to
recognize this population at risk in order to make
an earlier diagnosis of HIV infection.
e11 Prehypertension and Elevated Risk of
Cardiovascular Disease: Physiopathologic
Mechanisms
Flora Affuso, Antonio Ruvolo, and Serafino Fazio
e13 The Safety of Levofloxacin in Patients on
Warfarin
Jeffrey J. Glasheen and Allan V. Prochazka
e15 The Reply
Chaim M. Bell, Lynfa F. Stroud, Alex Kopp, and
Muhammad M. Mamdami
A10
APM PERSPECTIVES
374 The Lost Art of Clinical Skills
Christopher A. Feddock
CLASSIFIED ADS
B1
Positions available
The American Journal of Medicine (2007) 120, 287-288
EDITORIAL
Peer Review: The Best of the Blemished?
One of Winston Churchill’s most memorable quotations
comes from a speech to the House of Commons in which he
remarked that, “Democracy is the worst form of government
except for all the other forms that have been tried.” In my
opinion, the same sentiment applies to the peer review
process employed by biomedical academic journals. As is
the case for Churchill’s democracy, there can be little doubt
that peer review, as currently practiced, has flaws. Its dual
purpose is to assess the quality of manuscripts and to help
authors improve those manuscripts. Most editors of medical
journals believe that employing a wide array of specialty
reviewers is of considerable value in appraising submissions. In addition, many authors— but not all—are grateful
for constructive criticisms that result in improvements to
their published work.
In recent years, both editors and scientists have begun to
examine the peer review process to elucidate bias and error
in the system.1-7 As might be expected, 2 reviewers examining the same manuscript might express diametrically opposed views. Indeed, it is not uncommon for me to see one
reviewer suggest acceptance while another counsels against
it. Studies of the reproducibility of peer review based on
critiques of the same manuscript by different reviewers have
demonstrated considerable disparities between reviewers’
responses.5
Another failing, aside from inconsistency, is delayed
decision-making. Although computerization of the review
process has shortened the time between author submission
and an editorial decision, significant slow-downs continue
to exist. Other commonly cited defects in the peer review
system include reviewer bias, conflict of interest, expense,
and even occasional instances of reviewer dishonesty or
abuse of the author.
Bias exists in all human thinking, and a number of
studies have documented bias in the peer review process.
Such bias may be based on the gender or national origin of
the author, or it may relate to previously held prejudices
concerning the type of work described or the methodology
employed in the study. Today, potential conflicts of interest
are generally an issue in biomedical research as well. Conflicts may be financial or may relate to other factors, such as
concerns about career advancement or jealousy.8
Scientific misconduct also has been documented in the
peer review process. There have been a number of well0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2007.02.013
publicized examples of plagiarism of ideas and of components of a peer-reviewed manuscript. Finally, the peer
review system often fails to uncover examples of scientific fraud on the part of a paper’s authors until 1 or more
manuscripts already have been published and subsequently shown to be false. The recent Korean stem-cell
scandal received front-page coverage in almost every
major newspaper in the world. Some have argued that the
falsified publications of Dr. Hwang Woo-Suk in Science
during 2004 and 2005 demonstrated the weakness of the
present peer review mechanism. However, the opposite
argument also has been advanced. In the end, peer review
triumphed when Dr. Hwang Woo-Suk’s work was shown
to contain falsified data.
Can the process be improved? Given the many potential
problems and abuses attributed to peer review, it is surprising that the system, though deemed to be faulty, has generally functioned well.1,2 It remains the “standard of care”
for most biomedical journals published today.
A number of editors and academicians have suggested
reforms aimed at improving the identified shortcomings of
peer review. Double-blinding of the review process; open
reviews published on the Internet alongside the author’s
response; and focused reviewer training have been suggested and evaluated in a handful of small trials.3,9 Few
existing studies have assessed the efficacy and accuracy of
the peer review process.5 Moreover, the results of these
small trials often have produced conflicting conclusions. No
suggested improvement has emerged as the answer to previously enumerated deficiencies.
So what is an editor to do? In the final analysis, I have
come to believe that like the democratic system, peer review
is an imperfect arrangement, but it is better than anything
else that has been previously tried.
Reviewers are rarely remunerated for their work, and to
my knowledge, there is no current movement to change the
situation. If they were paid, the annual expense would be
enormous, and the readership would likely be required to
contribute. At The American Journal of Medicine, we
remain indebted to colleagues who serve as reviewers. We
also owe a debt of gratitude to readers who share their
comments on manuscripts that we have published—this is
another form of peer review. As I noted in an earlier editorial, the editors of the Green Journal strive to bring you a
288
The American Journal of Medicine, Vol 120, No 4, April 2007
quality product. Their decisions are based on multiple factors, including, but not limited to, reviewers’ comments. We
attempt to anticipate your interest in specific topics, and we
also try to publish material that will be of broad interest to
the internal medicine community.
Overall, the peer review process warrants our confidence; we should trust the reviewers’ and editors’ efforts at
objectivity.1,2,10,11 Those of us who participate in the peer
review process pledge our honesty. If we fail, it stains our
reputations. Similarly, editors and their editorial boards are
committed to the highest ideals of academic honesty and
scholarship. Undoubtedly, The American Journal of Medicine will continue to rely on the peer-review system when
making decisions until a different or modified technique is
conclusively proven to be better.
Joseph S. Alpert, MD
Robert S. and Irene P. Flinn Professor of Medicine
Special Assistant to the Dean
University of Arizona College of Medicine
Tucson
[email protected]
References
1. Smith R. Peer review: a flawed process at the heart of science and
journals. J R Soc Med. 2006;99:178-182.
2. Hojat M, Gonnella JS, Caelleigh AS. Impartial judgment by the “gatekeepers” of science: fallibility and accountability in the peer review
process. Adv Health Sci Educ Theory Pract. 2003;8:75-96.
3. Jefferson T, Alderson P, Wager E, Davidoff F. Effects of editorial peer
review: a systematic review. JAMA. 2002;287:2784-2786.
4. Kearney MH, Freda MC. Nurse editors’ views on the peer review
process. Res Nurs Health. 2005;28:444-452.
5. Kumar PD. How do peer reviewers of journal articles perform? Evaluating the reviewers with a sham paper. J Assoc Physicians India.
1999;47:198-200.
6. Burnham JC. The evolution of editorial peer review. JAMA. 1990;263:
1323-1329.
7. Olson CM. Peer review of the biomedical literature. Am J Emerg Med.
1990;8:356-358.
8. Alpert JS, Furman S, Smaha L. Conflicts of interest: science, money,
and health. Arch Intern Med. 2002;162:635-637.
9. Godlee F. Making reviewers visible: openness, accountability, and
credit. JAMA. 2002;287:2762-2765.
10. Alpert, JS. Dear editor. Am J Med. 2006;119;193.
11. Freshwater D. Editors and publishing: integrity, trust, and faith. J Psychiatr Ment Health Nurs. 2006;13:1-2.
The American Journal of Medicine (2007) 120, 289-294
REVIEW
Judicious Antibiotic Use and Intranasal Corticosteroids in
Acute Rhinosinusitis
Catherine Butkus Small, MD,a Claus Bachert, MD, PhD,b Valerie J. Lund, MS, FRCS, FRCSEd,c
Augustine Moscatello, MD,d Anjuli S. Nayak, MD,e William E. Berger, MD, MBAf
a
Division of Infectious Diseases, Department of Medicine, New York Medical College, Westchester Medical Center, Valhalla;
University of Ghent, Ghent, Belgium; cRoyal National Throat, Nose & Ear Hospital, London, England; dDepartment of
Otolaryngology/Head and Neck Surgery, New York Medical College, Westchester Medical Center, Valhalla; eDepartment of Pediatrics,
University of Illinois College of Medicine, Peoria; fDepartment of Pediatrics, Division of Allergy and Immunology, University of
California, Irvine College of Medicine, Irvine.
b
ABSTRACT
Most patients with symptoms of acute rhinosinusitis are treated with antibiotics. However, many cases of
rhinosinusitis are secondary to viral infections and unlikely to benefit from antibiotic therapy. Inappropriate
use of antibiotics in patients with acute nonbacterial rhinosinusitis contributes to the increase in bacterial
antibiotic resistance. Consequently, safe and effective alternatives to antibiotics are needed in the treatment
of acute rhinosinusitis caused by viral infections. Recent results from controlled trials have shown that
intranasal corticosteroids, used in combination with antibiotics or as monotherapy in selected cases,
provide significant symptom relief and resolution of acute rhinosinusitis. The use of intranasal corticosteroids in acute rhinosinusitis therefore might reduce the inappropriate use of antimicrobial therapy in
acute rhinosinusitis. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Bacterial; Infection; Intranasal corticosteroids; Resistance; Rhinosinusitis; Viral
Acute rhinosinusitis is a common upper respiratory tract
infection seen in ambulatory practice in the United States.1
Annual crude adult rhinosinusitis prevalence rates range
from 14% to 16%.2-4 Total costs related to sinusitis as a
primary or secondary diagnosis approached $5.8 billion in
one study.5,6 Rhinosinusitis is currently the preferred term
because sinusitis is invariably accompanied by nasal airway
inflammation and often preceded by rhinitis.2,5,7-9 Rhinosinusitis is classified as acute (symptoms ⬍4 weeks), subacute
(symptoms 4-12 weeks), and chronic (symptoms ⬎12
weeks),7,10,11 with some patients experiencing recurrent attacks (ⱖ4 episodes yearly ⱖ 7 days).7 Definitive prevalence
rates are lacking because of inconsistencies in definitions
and because not all patients seek professional care.5 Rhinosinusitis may be viral, bacterial, or allergic.8 Frequently,
Requests for reprints should be addressed to Catherine Butkus Small,
MD, Division of Infectious Diseases, Munger Pavilion 245, Westchester
Medical Center, Valhalla, NY 10595.
E-mail address: [email protected].
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.08.035
acute rhinosinusitis is caused by a viral infection,8,12 which
may be complicated by a secondary bacterial infection5,12-15;
representing approximately 20 million cases of acute bacterial rhinosinusitis annually.13,16,17 However, only approximately 38% of adults presenting with symptoms of rhinosinusitis actually may have acute bacterial rhinosinusitis.18
Guidelines for the diagnosis and treatment of acute rhinosinusitis provide algorithms to improve treatment outcomes
and decrease costs.16,17,19 Despite this, approaches used by
physicians to diagnose and treat acute rhinosinusitis vary.
One survey indicated substantial differences in physicians’
use of diagnostic tools, prescription of antibiotics, and referrals to specialists.20 Antibiotics are frequently prescribed
for acute rhinosinusitis, even though the cause may be viral.
As a result, empiric antibiotic therapy has increased, contributing to the current increase in drug resistance among
respiratory pathogens.12,21 Increasing bacterial resistance,
coupled with the decrease in development of new antibiotics, signals a need for interventions to control the increase in
290
morbidity resulting from ineffective antimicrobial therapy
for acute rhinosinusitis.22
This article reviews the pathophysiology, cause, clinical
characteristics, differential diagnosis, and treatment of acute
rhinosinusitis. It also discusses alternative approaches to its
management, focusing on intranasal corticosteroids as both adjunctive treatment and monotherapy.
The American Journal of Medicine, Vol 120, No 4, April 2007
Guidelines for Diagnosis of
Acute Rhinosinusitis
Parameters for the diagnosis and management of acute rhinosinusitis were updated recently by the American Academy of Allergy, Asthma, and Immunology; the American
College of Allergy, Asthma, and
Immunology; and the Joint Council
of Allergy, Asthma, and ImmunolCLINICAL SIGNIFICANCE
ogy.16 The European Academy of
● Most cases of acute rhinosinusitis are
Allergology and Clinical ImmunolPATHOPHYSIOLOGY OF
viral
and
therefore
do
not
respond
to
ogy developed similar guidelines.9
ACUTE RHINOSINUSITIS
The main clinical characterisantibiotics.
Acute rhinosinusitis is charactertics
of acute rhinosinusitis include
ized by inflammation of the para● Inappropriate treatment of acute rhinonasal
congestion, facial pain
nasal sinuses, most commonly the
sinusitis with antibiotics contributes to
and/or
pressure,
rhinorrhea, postmaxillary and ethmoid sinuses
antibiotic resistance.
nasal
drainage,
headache, and
7,10,23
(Figure 1).
Factors contribcough.
Symptoms
in children are
● Intranasal corticosteroids are an effecuting to the development of rhinosimilar
but
include
prolonged
sinusitis include viral, bacterial,
tive and safe treatment for acute rhinocoughing,
irritability,
and
gagging
and, infrequently, fungal infecsinusitis in combination with antibiotics
on
mucus,
leading
to
vomittions; allergic and immunologic
or, in selected cases, as monotherapy.
16,26
ing.
A
position
paper
enreactions; and noninfectious facdorsed
by
the
Centers
for
Disease
● The use of intranasal corticosteroids in
tors, including gastroesophageal
Control and Prevention, the Amer8
the treatment of acute rhinosinusitis
reflux disease. Perennial allergic
ican Academy of Family Physimight help reduce the incidence of anrhinitis also may be a predisposing
cians-American Society of Interfactor because allergic rhinitis
tibiotic resistance.
nal Medicine, and the Infectious
may contribute to rhinosinusitis in
Diseases Society of America reup to 30% of patients with acute
ported 4 signs and symptoms most
maxillary rhinosinusitis and up to
8,24,25
predictive of acute bacterial rhinosinusitis, including puru80% of patients with chronic rhinosinusitis.
lent nasal discharge, maxillary tooth or facial pain (espeDifferentiating viral rhinosinusitis from bacterial rhinocially unilateral), unilateral maxillary sinus tenderness, and
sinusitis is a diagnostic challenge because of their similar
worsening symptoms after initial improvement.27 The
presentations and the insufficient specificity and sensitivity
17
symptoms of chronic rhinosinusitis are more subtle.16 Raof current diagnostic imaging. The overlap among sympdiographic imaging confirms acute rhinosinusitis in patients
toms of acute viral upper respiratory tract infections, seawhose symptoms are vague or persist after therapy, or who
sonal and perennial allergic rhinitis, vasomotor rhinitis, and
have ambiguous physical findings. Computed tomography
rhinitis medicamentosa with symptoms of acute bacterial
is currently the preferred radiographic modality.16
rhinosinusitis also makes an accurate diagnosis problematic.
Figure 1 Anatomy of the paranasal sinuses. (Reproduced with permission from: Nose and paranasal sinus. In: The Merck Manual of
Diagnosis and Therapy. Available at: http://www.merck.com/mrkshared/mmanual/figures/86fig1.jsp. Accessed March 6, 2006.)
Small et al
Management of Acute Rhinosinusitis
291
Figure 2 Algorithm for differentiating bacterial and viral rhinosinusitis and for the primary care management of rhinosinusitis.
(Reproduced with permission from: Sande MA, Gwaltney JM. Acute community-acquired bacterial sinusitis: continuing challenges and
current management. Clin Infect Dis. 2004;39(suppl 3):S151-S158.) CT ⫽ computed tomography; ENT ⫽ ear, nose, and throat.
Differentiating viral from bacterial rhinosinusitis is complicated because viral infection often precedes bacterial
infection.8 Sinus puncture and culture is the gold standard
for microbial identification, but its routine use is not recommended.12,17 Microbiologic assessment of direct cultures
by nasopharyngeal swabs and endoscopic sampling of the
middle meatus is less invasive but may not provide a definitive diagnosis. Thus, the most common diagnostic
course in the office setting involves medical history and
clinical signs and symptoms.12 Duration of symptoms is the
most common way physicians differentiate between acute
viral and bacterial rhinosinusitis. Symptoms associated with
acute bacterial rhinosinusitis worsen after 5 to 7 days or
persist for more than 10 days but less than 12 weeks.2 Other
studies suggest that a 7- to 10-day duration of acute rhinosi-
nusitis may provide a “natural breakpoint,” with acute bacterial
rhinosinusitis more likely when symptoms do not improve or
worsen after this time interval (Figure 2).5,10,12,16
MANAGEMENT OF PATIENTS WITH
ACUTE RHINOSINUSITIS
Symptoms resolve without antibiotics in most patients with
acute rhinosinusitis.10,27 However, acute rhinosinusitis also
is associated with significant physical symptoms, deteriorating quality of life, and impairment of daily functioning,
inducing patients to seek treatment.8 It is unclear which
cases will resolve spontaneously.5 Of treated cases of
acute rhinosinusitis, 85% to 98% of those managed in the
primary care setting are given antibiotics.17 Given that
292
most infections are viral, the high percentage of reported
antibiotic prescriptions suggests that antibiotics are
overused.12,28,29
Antibiotic Therapy
The goals of treating acute bacterial rhinosinusitis include
diminishing tissue edema, reversing sinus ostial obstruction,
and controlling infection.30 Antimicrobial therapy is often
empiric. The European Academy of Allergology and Clinical Immunology recommends that adults with mild symptoms receive symptomatic relief and that patients with
moderate-to-severe symptoms receive intranasal corticosteroids.9 If symptoms worsen after 5 to 7 days or persist after
more than 10 days and are moderate or severe, antibiotic
therapy should be initiated.9 Although current American
treatment guidelines are similar,12,16 exceptions include
patients with specific physical findings or serious complications. In these cases, antibiotics should be started immediately. Infection with resistant pathogens should be considered if symptoms do not improve after 3 to 5 days of
antibiotics.12
Although unnecessary antibiotics contribute to antimicrobial resistance,27 antibiotics are appropriate in certain
patients, including those with acute bacterial infection, meningeal or orbital complications, and immunocompromised
or hospitalized patients.12,31 Antibiotics should be active
against Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis.5,12 However, there is increasing bacterial resistance, especially in S. pneumoniae, to
amoxicillin and other ␤-lactams, trimethoprim/sulfamethoxazole, and macrolides.32 Therefore, the narrowest spectrum antibiotic effective against these pathogens should be
used. Amoxicillin, amoxicillin/clavulanate, cefpodoxime
proxetil, cefdinir, and cefuroxime axetil are currently recommended as first-line antibiotics in acute bacterial
rhinosinusitis.8,17
Other treatments are also recommended for acute rhinosinusitis. Guidelines recommend that patients with uncomplicated acute rhinosinusitis be managed symptomatically,
including antihistamines or decongestants, although antihistamines are recommended only for suspected allergy.16 Mucolytics, saline lavage, expectorants, and homeopathic or
herbal medications are not efficacious.9,33 Decongestants
reduce turbinate swelling and improve nasal patency.9,33
However, prolonged use (⬎3 days) may result in rebound
congestion.34
In acute bacterial rhinosinusitis, a shorter duration of
antibiotic treatment is possible, thereby decreasing resistance and costs.12,35 One study compared trimethoprim/
sulfamethoxazole given for 3 days versus 10 days in
acute bacterial rhinosinusitis with similar success rates.36
Long-term antibiotics often lead to decreased patient
compliance, antibiotic resistance, and increased costs.12
These data support a conservative approach to antibiotics
in acute rhinosinusitis.
The American Journal of Medicine, Vol 120, No 4, April 2007
The Role of Intranasal Corticosteroids
The rationale for intranasal corticosteroids in acute rhinosinusitis resides in their anti-inflammatory properties. Intranasal corticosteroids are a first-line treatment for allergic
rhinitis, which may be a predisposing factor for acute
rhinosinusitis.24
Inflammation and edema of the mucous membranes of
the nasal turbinates and sinus ostia block drainage routes
and impair mucociliary clearance mechanisms. By reducing
inflammation, intranasal corticosteroids foster drainage and
increased aeration.37,38 Many recent studies support the
benefits of intranasal corticosteroids as adjuncts to antimicrobial therapy in acute bacterial rhinosinusitis.37-39 In a
study of children with acute rhinosinusitis, treatment with
amoxicillin/clavulanate for 3 weeks and twice-daily budesonide nasal spray led to significant improvements in cough
and nasal discharge compared with amoxicillin/clavulanate
alone (P ⬍.05).37 In another study, patients aged 12 years or
more with acute bacterial rhinosinusitis confirmed by sinus
computed tomography scan were treated with amoxicillin/
clavulanate plus placebo or mometasone furoate nasal spray
(MFNS) twice daily. Symptom scores revealed that MFNS
treatment decreased the total symptom score and individual
scores for headache, congestion, and facial pain versus
placebo (P ⬍.05).39 Similar results were obtained in outpatients with moderate-to-severe rhinosinusitis treated with
MFNS plus amoxicillin/clavulanate for 3 weeks.40 Thus,
adding intranasal corticosteroids to antibiotics produces a
significant benefit.
Monotherapy with Intranasal Corticosteroids
Because intranasal corticosteroids are an effective adjunctive therapy for acute bacterial rhinosinusitis, they may have
potential as monotherapy for acute rhinosinusitis. One recent trial concluded that MFNS was an effective monotherapy in patients with acute uncomplicated rhinosinusitis.41 The efficacy of MFNS as monotherapy versus
amoxicillin and placebo was evaluated in 981 subjects aged
12 years or more with acute rhinosinusitis for 7 or more but
28 or less days. Patients with acute bacterial rhinosinusitis
were excluded. MFNS 200 ␮g twice daily was superior to
placebo alone and amoxicillin alone in improving symptom
scores (P ⬍.001 and P ⫽ .002, respectively) (Figure 3). The
greatest symptom improvement was for night waking, a
significant finding given the prevalence of sleep impairment
in these patients.
Safety
Studies show that intranasal corticosteroids are safe and
well tolerated in patients with acute rhinosinusitis. The
incidence of adverse events was similar to placebo.41 Common adverse events were headache and epistaxis. A recent
study of intranasal corticosteroids and nasal polyps also
supports their safety.42
Intranasal corticosteroids in pediatric patients do not
change baseline cortisol levels or blunt the response to
Small et al
Management of Acute Rhinosinusitis
293
Figure 3 Efficacy of MFNS as monotherapy in patients with acute rhinosinusitis. (Reproduced with permission from: Meltzer EO,
Bachert C, Staudinger H. Treating acute rhinosinusitis: comparing efficacy and safety of MFNS, amoxicillin, and placebo. J Allergy Clin
Immunol. 2005;116:1289-1295.) MFNS ⫽ mometasone furoate nasal spray.
cosyntropin stimulation, and have little systemic absorption
or suppression of the hypothalamic-pituitary-adrenal axis.43,44
Examination of nasal biopsy specimens after 1 year of therapy
with MFNS in 1 study revealed a decrease in focal metaplasia,
no change in epithelial thickness, and no atrophy.45
CONCLUSIONS
Although many patients with acute rhinosinusitis do not
seek medical care, millions consult physicians each year
requesting treatment and are inappropriately given antibiotics.46 Patients with a cold or flu-like illness for less than 7
days are unlikely to have a bacterial infection and should
not receive antibiotics.12 The inappropriate use of antibiotics can be reduced by better differentiation of acute bacterial
and viral rhinosinusitis.47 In patients who require antibiotics, selection of an appropriate agent and a shorter duration
of therapy is important in preventing resistance.48
Studies reveal that the administration of intranasal corticosteroids with antibiotics provides significant symptom
relief in patients with acute bacterial rhinosinusitis. Because
most cases of acute rhinosinusitis are viral and do not
respond to antibiotics, much of the benefit associated with
this combination may be related to intranasal corticosteroids
alone. This suggestion is supported by a recent study showing that monotherapy with intranasal corticosteroids was
superior to antibiotics alone as empiric treatment. The efficacy of intranasal corticosteroids as monotherapy for acute
rhinosinusitis and their well-documented safety in adult and
pediatric patients suggest that intranasal corticosteroids may
be a useful alternative to antimicrobial therapy in the treatment of acute rhinosinusitis, and may contribute to a decrease in the inappropriate prescribing of antibiotics and to
the development of resistance in this common disease.
References
1. Schappert SM. Ambulatory care visits to physician offices, hospital
outpatient departments, and emergency departments: United States,
1996. Vital Health Stat 13. 1998;134:1-37.
2. Kaliner MA, Osguthorpe JD, Fireman P, et al. Sinusitis: bench to
bedside. Current findings, future directions. J Allergy Clin Immunol.
1997;99:S829-S848.
3. National Center for Health Statistics. Health, United States, 2005.
With chartbook on trends in the health of Americans. Available at:
http://www.cdc.gov/nchs/data/hus/hus05.pdf. Accessed March 6,
2006.
4. Lethbridge-Çejku M, Rose D. Summary health statistics for U.S.
adults: National Health Interview Survey, 2004 [provisional data].
Available at: http://www.cdc.gov/nchs/data/series/sr_10/sr10_228.pdf.
Accessed March 6, 2006.
5. Poole MD. Acute bacterial rhinosinusitis: clinical impact of resistance
and susceptibility. Am J Med. 2004;117(suppl):29S-38S.
6. Ray NF, Baraniuk JN, Thamer M, et al. Healthcare expenditures for
sinusitis in 1996: contributions of asthma, rhinitis, and other airway
disorders. J Allergy Clin Immunol. 1999;103:408-414.
7. Lanza DC, Kennedy DW. Adult rhinosinusitis defined. Otolaryngol
Head Neck Surg. 1997;117:S1-S7.
8. Meltzer EO, Hamilos DL, Hadley JA, et al. Rhinosinusitis: establishing definitions for clinical research and patient care. J Allergy Clin
Immunol. 2004;114(suppl):S155-S212.
9. Fokkens W, Lund V, Bachert C, et al. EAACI position paper on
rhinosinusitis and nasal polyps executive summary. Allergy. 2005;60:
583-601.
10. Hickner JM, Bartlett JG, Besser RE, et al. Principles of appropriate
antibiotic use for acute rhinosinusitis in adults: background. Ann Intern
Med. 2001;134:498-505.
11. International Rhinosinusitis Advisory Board. Infectious rhinosinusitis
in adults: classification, etiology and management. Ear Nose Throat J.
1997;76(suppl 12):1-22.
12. Sande MA, Gwaltney JM. Acute community-acquired bacterial sinusitis: continuing challenges and current management. Clin Infect Dis.
2004;39(suppl 3):S151-S158.
13. Sinus and Health Partnership. Antimicrobial treatment guidelines for
acute bacterial rhinosinusitis. Otolaryngol Head Neck Surg. 2000;123:
1-32.
14. Gwaltney JM Jr. Acute community-acquired sinusitis. Clin Infect Dis.
1996;23:1209-1223.
15. Berg O, Carenfelt C, Rystedt G, et al. Occurrence of asymptomatic
sinusitis in common cold and other acute ENT-infections. Rhinology.
1986;24:223-225.
16. Slavin RG, Spector SL, Bernstein IL, et al. The diagnosis and management of sinusitis: a practice parameter update. J Allergy Clin
Immunol. 2005;116(suppl 6):S13-S47.
294
17. Sinus and Allergy Health Partnership. Antimicrobial treatment guidelines for acute bacterial rhinosinusitis. Otolaryngol Head Neck Surg.
2004;130(suppl 1):1-45.
18. Lau J, Zucker D, Engels EA, et al. Diagnosis and treatment of acute
bacterial rhinosinusitis. Evid Rep Technol Assess (Summ). 1999:1-3.
19. McCracken GH Jr. Clinical practice guidelines for the diagnosis and
treatment of respiratory tract infections. Am J Manag Care. 2001;
7(suppl 6):S183-S191.
20. Trinh N, Ngo HH. Practice variations in the management of sinusitis.
J Otolaryngol. 2000;29:211-217.
21. Karchmer AW. Increased antibiotic resistance in respiratory tract
pathogens: PROTEKT US—an update. Clin Infect Dis. 2004;39(suppl
3):S142-S150.
22. Weber JT. Appropriate use of antimicrobial drugs: a better prescription
is needed. JAMA. 2005;294:2354-2356.
23. Nose and paranasal sinus. In: The Merck Manual of Diagnosis and
Therapy. Available at: http://www.merck.com/mrkshared/mmanual/
figures/86fig1.jsp. Accessed March 6, 2006.
24. Meltzer EO, Szwarcberg J, Pill MW. Allergic rhinitis, asthma, and
rhinosinusitis: diseases of the integrated airway. J Manag Care Pharm.
2004;10:310-317.
25. Spector SL. Overview of comorbid associations of allergic rhinitis.
J Allergy Clin Immunol. 1997;99:S773-S780.
26. American Academy of Pediatrics. Subcommittee on Management of
Sinusitis and Committee on Quality Improvement. Clinical practice
guideline: management of sinusitis. Pediatrics. 2001;108:798-808.
27. Gonzales R, Bartlett JG, Besser RE, et al. Principles of appropriate
antibiotic use for treatment of acute respiratory tract infections in
adults: background, specific aims and methods. Ann Intern Med. 2001;
134:479-486.
28. Stalman W, van Essen GA, van der Graaf Y, de Melker RA. The end
of antibiotic treatment in adults with acute sinusitis-like complaints in
general practice? A placebo-controlled double-blind randomized
doxycycline trial. Br J Gen Pract. 1997;47:794-799.
29. van Buchem FL, Knottnerus JA, Schrijnemaekers VJJ, Peeters MF.
Primary-care-based randomised placebo-controlled trial of antibiotic
treatment in acute maxillary sinusitis. Lancet. 1997;349:683-687.
30. Osguthorpe JD. Adult rhinosinusitis: diagnosis and management. Am
Fam Physician. 2001;63:69-76.
31. Bachert C, Hörmann K, Mösges R, et al. An update on the diagnosis
and treatment of sinusitis and nasal polyposis. Allergy. 2003;58:176191.
32. Felmingham D, Reinert RR, Hirakata Y, et al. Increasing prevalence of
antimicrobial resistance among isolates of Streptococcus pneumoniae
from the PROTEKT surveillance study, and comparative in vitro
activity of the ketolide, telithromycin. J Antimicrob Chemother. 2002;
50(suppl S1):25-37.
33. University of Michigan Health System Guidelines for Clinical Care.
Acute rhinosinusitis in adults. Available at: http://www.mcare.org/
media/pdf_autogen/cpg_rhinosinusitis.pdf. Accessed May 1, 2006.
The American Journal of Medicine, Vol 120, No 4, April 2007
34. Spector SL, Bernstein IL, Li JT, et al. Parameters for the diagnosis and
management of sinusitis. J Allergy Clin Immunol. 1998;102:S107S144.
35. Pichicero M. Short courses of antibiotic in acute otitis media and
sinusitis infections. J Int Med Res. 2000;28(suppl 1):25A-36A.
36. Williams JW Jr, Holleman DR Jr, Samsa GP, et al. Randomized
controlled trial of 3 vs 10 days of trimethoprim/sulfamethoxazole for
acute maxillary sinusitis. JAMA. 1995;273:1015-1021.
37. Barlan IB, Erkan E, Bakir M, et al. Intranasal budesonide spray as an
adjunct to oral antibiotic therapy for acute sinusitis in children. Ann
Allergy Asthma Immunol. 1997;78:598-601.
38. Meltzer EO, Orgel HA, Backhaus JW, et al. Intranasal flunisolide
spray as an adjunct to oral antibiotic therapy for sinusitis. J Allergy
Clin Immunol. 1993;92:812-823.
39. Meltzer EO, Charous BL, Busse WW, et al, and the Nasonex Sinusitis
Group. Added relief in the treatment of acute recurrent sinusitis with
adjunctive mometasone furoate nasal spray. J Allergy Clin Immunol.
2000;106:630-637.
40. Nayak AS, Settipane GA, Pedinoff A, et al, for the Nasonex Sinusitis
Group. Effective dose range of mometasone furoate nasal spray in the
treatment of acute rhinosinusitis. Ann Allergy Asthma Immunol. 2002;
89:271-278.
41. Meltzer EO, Bachert C, Staudinger H. Treating acute rhinosinusitis:
comparing efficacy and safety of mometasone furoate nasal spray,
amoxicillin, and placebo. J Allergy Clin Immunol. 2005;116:12891295.
42. Small CB, Hernandez J, Reyes A, et al. Efficacy and safety of mometasone furoate nasal spray in nasal polyposis. J Allergy Clin
Immunol. 2005;116:1275-1281.
43. Brannan MD, Herron JM, Affrime MB. Safety and tolerability of
once-daily mometasone furoate aqueous nasal spray in children. Clin
Ther. 1997;19:1330-1339.
44. Daley-Yates PT, Kunka RL, Yin Y, et al. Bioavailability of fluticasone
propionate and mometasone furoate aqueous nasal sprays. Eur J Clin
Pharmacol. 2004;60:265-268.
45. Minshall E, Ghaffar O, Cameron L, et al. Assessment by nasal biopsy
of long-term use of mometasone furoate aqueous nasal spray (Nasonex) in the treatment of perennial rhinitis. Otolaryngol Head Neck
Surg. 1998;118:658-664.
46. Butler CC, Kinnersley P, Prout H, et al. Antibiotics and shared decision-making in primary care. J Antimicrob Chemother. 2001;48:435440.
47. Turnidge J. Responsible prescribing for upper respiratory tract infections. Drugs. 2001;61:2065-2077.
48. Brook I, Gooch WM 3rd, Jenkins SG, et al. Medical management of
acute bacterial sinusitis. Recommendations of a clinical advisory committee on pediatric and adult sinusitis. Ann Otol Rhinol Laryngol.
2000;109:2-20.
The American Journal of Medicine (2007) 120, 295-302
REVIEW
Guideline for Low-Cost Antimicrobial Use in the
Outpatient Setting
Lizbeth A. Hansen, PharmD,a,b Lee C. Vermeulen, MS, RPh,a,b Sarah Bland, RPh,a Tosha B. Wetterneck, MDc
a
c
Department of Pharmacy, University of Wisconsin Hospital and Clinics; bSchool of Pharmacy, University of Wisconsin-Madison;
Department of Medicine, University of Wisconsin School of Medicine and Public Health, Madison.
ABSTRACT
In an effort to increase appropriate prescribing of low-cost antimicrobials in the outpatient setting, an
evidence-based guideline was created to identify situations when low-cost medications can be used. A
literature search identified relevant clinical trials describing the efficacy of antimicrobials used in the
outpatient setting. These were analyzed to identify low-cost medications defined as $15 or less. The
information was put into guideline format that includes the level of evidence for recommending the drug
and information about cost. Sixteen common infections and their treatments were included in the guideline.
The efficacy data were similar for the low-cost and higher-cost antimicrobials for all infections included.
We created a low-cost antimicrobial guideline for common infections treated in the outpatient setting. The
treatment options have similar efficacy to higher cost medications. This guideline will serve as an
information source for providers to help them rapidly determine the low-cost treatments for common
infections. In addition, it can serve as a template for the development of similar guidelines in other
therapeutic classes. These guidelines should be customized before implementation at other health care
organizations, with consideration of local resistance patterns, drug availability and patient factors. The
effect of guideline implementation on future prescribing habits and providers’ opinions about availability
of cost information and subsequent conversations with patients and prescribers of medications deserves
further study. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Low-cost antimicrobials; Clinical practice guideline; Outpatient infections
The financial impact on patients of purchasing medications
is a problem of increasing significance. The prevalence of
uninsured and underinsured patients in our healthcare system has been well documented. One of every 3 Americans
under the age of 65 years was uninsured at some time in
2002-2003.1 Even those fortunate enough to receive health
insurance as a benefit of employment have experienced
increasing out-of-pocket expense for the care they receive
as employers reduce coverage to manage ongoing premium
increases.2 These issues not only cause patients economic
distress, but also have the ability to negatively impact clinLee Vermeulen is a member of speaker’s bureaus and/or receives
research funding from Pfizer Labs and Amgen.
Requests for reprints should be addressed to Tosha Wetterneck, MD,
University of Wisconsin Hospital and Clinics, 600 Highland Avenue,
J5/222b CSC, MC 2454, Madison, WI 53792.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.06.012
ical outcomes.3 A recent study showed that 1 in 5 adults
(22%) did not fill at least 1 prescription in the previous year
because of cost.4 Moreover, conversations about the cost of
medications and patients’ out-of-pocket costs occur infrequently despite patients’ interest in discussing this information and physicians’ feelings that this information is important in clinical decision-making.5 Patients most often
reported the reason for not initiating these conversations as
a feeling of discomfort, believing that the discussion may
not lead to a viable alternative or that an alternative might
affect the quality of care provided.6 To pay for expensive
medications, patients reported borrowing money from
friends and family, increasing credit card debt, choosing
medications over food or other essentials, as well as decreased overall adherence to all prescribed medications.7-9
Physicians are also hesitant to initiate these important
conversations about medication cost. In a survey of outpa-
296
The American Journal of Medicine, Vol 120, No 4, April 2007
tient physicians, the providers stated that they had very little
lowest-cost antimicrobials for the treatment of common
requisite knowledge or cost-related resources needed to
infections while maintaining the highest clinical treatment
facilitate cost-effective prescribing.10 However, most phyefficacy. We describe the development of a health system
sicians felt that cost information was important and were
guideline for primary care providers, house staff, and pharwilling to sacrifice some degree of clinical efficacy to make
macist use in the outpatient setting and emergency
therapy more affordable for padepartment.
tients. It also was reported that
physicians learned about cost
CLINICAL SIGNIFICANCE
METHODS
from patients as often as they did
The creation of a low-cost antimifrom pharmacists (45% vs 41%,
● Prescribers are often unaware of lowcrobial prescribing guideline was
respectively). Other sources of
cost treatments.
charged by the hospital’s multidiscost information used by physiciplinary Pharmacy and Therapeu● Many common outpatient infections have
cians were The Medical Letter on
tics Committee. The hospital orgaequally efficacious low-cost therapies.
Drugs and Therapeutics, other
nization was the sole sponsor of
physicians, drug company repre● Low-cost prescribing guidelines can be
the guideline. “Low-cost” was desentatives, advertisements, and
fined as a total patient payment of
created
to
inform
providers
and
provide
10
journals. Medication cost inforunder $15 per prescription, chosen
options
for
patients.
mation is available online or on
to be consistent with the general
software for personal digital assiscost of generic medications or
11
tants ; however, studies evaluatother co-payments within a multiing clinical use of these medication databases (eg, ePocrates
ple-tier formulary, as well as a cost that many patients could
Rx Pro) have not found cost information to be a common
reasonably afford on an acute basis. An initial search was
reason for use.12 Other reasons for physicians not discussing
performed for existing literature on similar guidelines. Findcost are: insufficient time during the visit and lack of inforing none, the 2004 Sanford Guide to Antimicrobial Therapy
mation regarding alternative low-cost medication to offer.6
was used as an initial resource in order to determine comIn fact, providing medication cost information to internal
mon infections that could be managed on an outpatient
medicine physicians along with education about insurance
basis.19 The guide also details the drug therapy considered
coverage and the impact on cost to the patient has been
first and second line and includes references of supporting
shown to improve physician knowledge of medication costs
literature. A literature search was then performed using the
and willingness to consider cost when prescribing.13 TherePubMed database to assess the evidence available to support
fore, educating providers about the cost of medications and
these drug therapy recommendations and to find additional
making information about low-cost alternatives easily acevidence on low-cost alternatives that would also be concessible may be an important means to encourage approprisidered effective. Using keywords specific to the individual
ate prescribing of inexpensive medications.
infection, the search was first limited to practice guidelines.
Antimicrobials are an important medication class for the
If well-established practice guidelines were not found, the
treatment of acute illness. Therefore, the health effects and
limits were adjusted for review articles. If review articles
health care utilization effects of patients not filling prescripwere not found, a primary literature search was performed.
tion medications can be significant, resulting in repeat emerThe search utilized available data through July 2005. Studgency department visits, primary care visits, and readmisies were examined for evidence of antimicrobial efficacy for
sion to the hospital.14-17 Often, a less expensive medication
medications currently available in the US that fell below the
is available that has similar efficacy to the higher cost
chosen $15 ceiling. Cost information was based on the
medication prescribed.18 Generic medications routinely cost
maximum allowable cost for the Wisconsin Medicaid Sysless than brand equivalents with similar potency. Furthertem, a system commonly used to establish reimbursement
more, many medication formularies now have tiered paylevels for generic medications filled for Medicaid patients in
ment systems where patients benefit from lower co-paythe State of Wisconsin.20 Pharmacies pay widely ranging
ments for generic medications or organization-preferred
wholesale prices for generic drugs, and maximum allowable
medications.
cost pricing tables are established by payers to set the
Choosing a low-cost medication directly benefits the
maximum rate of reimbursement for those agents. We chose
patient and serves to lower health care costs in general. We
the Wisconsin maximum allowable cost price list as optherefore undertook a systematic process to identify lowposed to average wholesale price for a consistent and accost antimicrobials that have demonstrated clinical effeccessible standard for evaluation of drug prices.
tiveness similar to other, higher cost alternatives for treatStrength of recommendation was assigned based on a
ment of common infections. The main objective of this
scale that rates evidence based on the letters A-D for the
guideline was to serve as a reliable resource for low-cost
grade, and a number and letter (1a-1c, 2a-2c, 3a-3b, 4 and 5)
antimicrobial prescribing information for physicians pracfor the level of evidence.21 An A grade was paired with
Level 1 evidence and the letters indicate: systematic review
ticing in the outpatient setting. The goal was to identify the
Hansen et al
Low-Cost Antimicrobial Guideline
of randomized controlled trials; individual randomized controlled trials; and all or none studies. The B grade was
paired with level 2 and level 3 evidence, with 2a and 3a
indicating systematic reviews of cohort studies and casecontrol studies, respectively; 2b and 3b indicating individual cohort studies (including low-quality randomized controlled trials) and case-control studies, respectively; and 2c
indicating outcomes research. The C grade was paired with
level 4 evidence indicating case series and poor-quality
case-control and cohort studies, and the D grade was paired
with level 5 evidence indicating expert opinion or based on
bench research (Table). Available evidence was reviewed
by 1 physician and 2 pharmacists, who then came to a
consensus on the assigned strength of recommendation.
The Antimicrobial Use Subcommittee of the Pharmacy
and Therapeutics Committee, a committee with physician
and pharmacist members specializing in infectious diseases,
reviewed the guideline and feedback was incorporated into
the document. The guideline was then presented to the
hospital Pharmacy and Therapeutics Committee for approval. As part of the Committee’s routine guideline implementation approach, print copies of the guideline were distributed to all prescribers (eg, faculty physicians, house
staff, and nurse practitioners) and it was posted on the health
system intranet website, which is accessible broadly to all
staff. Outpatient and emergency department pharmacists
were informed as well to allow for suggestions to the prescriber if the patient could not afford a medication at the
time of dispensing. The guideline is also being used as an
interdisciplinary educational tool for efforts focusing on
maximizing the cost effectiveness of anti-infective
prescribing.
RESULTS
The guideline is displayed in table format with suggested
low-cost antimicrobial selections organized by body system, followed by common infections within the system that
are treated in the outpatient setting (Table). The majority of
the recommended medications were accessible to cash-paying patients for under $15 per course (including an appropriate dispensing fee). In certain situations, an inexpensive
or generic drug was not an appropriate therapeutic choice,
and the medication was noted with a dollar sign ($) to
indicate that the cost exceeds the $15.
For all noted infections, the strength of recommendations
assigned to the low-cost treatments was similar to the
strength of recommendations for the more expensive treatment, demonstrating equivalent evidence for both types of
recommendations. Therefore, the low-cost antimicrobial
guideline goals for creation were achieved. A comments
field was also included for information relevant to medication prescription such as length of therapy comments and
how resistance patterns may change recommendations. The
resultant guideline adhered to the Conference on Guideline
Standardization recommendations for guideline reporting in
the literature.63
297
DISCUSSION
A low-cost antimicrobial guideline was created as a reference for primary care providers and emergency department
physicians to provide information about low-cost treatment
strategies for common infections that maintain high clinical
efficacy. Although generic drug use is increasing, it is apparent that the health care system underutilizes generic
substitution as a means of cost saving.18 Analysis of stateby-state Medicaid prescription drug spending in 2000 identified potential savings of $229 million that could have been
realized from greater use of generic prescription medication.64 Broad generic substitution of outpatient prescription
drugs has been found to save approximately $8.8 billion, or
approximately 11% of drug expenditures for adults.18 The
translation of generic and low-cost prescribing into clinical
practice has been slow, possibly due to the lack of cost
information available to physicians and other providers. The
driving force behind low-cost prescribing has been the development of drug formularies with tiered co-payments and
preferred medications.65 Brand-name medications are
heavily promoted via physician detailing and direct-to-consumer advertising while generic medications are no longer
advertised. Controversy has begun to surface regarding the
benefits of branded medications and whether the newer
drugs are really worth their cost.64,66 While it is unlikely
that these questions will be definitively answered in the near
future, guidelines such as these help to promote the use of
generic medications through an evidence-based, multidisciplinary approach to clinical practice.
The legal implication of a guideline offering 2 standards
of care according to socioeconomic status is an important
consideration when creating a “low-cost” guideline. The
Center for Medicare and Medicaid does not allow separate
standards for treatment of diseases that may discriminate
based on patient socioeconomic status. However, low-cost
alternatives may be offered if the guideline identifies when
high-cost options are preferred due to a difference in clinical
efficacy between the low-cost and higher-cost medication,
and as long as applicability to broad patient populations is
maintained in the guideline. It should be emphasized that, in
general, guidelines are designed as educational tools to be
used as a reference, but are not legally binding. In addition,
it is advantageous to control costs for all patients, not only
patients with less ability to pay for health care costs.
There are limitations to the generalization of this guideline to other health care organizations. The applicability of
medical recommendations in this guideline may depend on
local antimicrobial resistance patterns and drug availability.
Also, the recommendations in this guideline apply only to
immunocompetent, nonpregnant, and nonallergic patients,
although we acknowledge that the same treatments may be
used in immunocompromised patients. Lastly, although
maximum allowable cost was used in this analysis to determine retail drug pricing, US drug prices are proprietary and
vary widely by region, so medications may not be as inexpensive as we have outlined in this guideline.
298
Table
Low-Cost Antimicrobial Recommendations for Outpatient Infections*
Infection
Gastroenteritis/GI
1a. Traveler’s Diarrhea – Mild (1-2 stools/ 24 h with
minor symptoms)22-24
1b. Traveler’s Diarrhea – Moderate (more than 2 stools
within 24 h)22,23
1c. Traveler’s Diarrhea – Severe (fever, bloody stools,
more than six stools within 24 h)22,23
Urogenital System
2. Acute (uncomplicated) lower urinary tract infection
in women25,26
4. Chlamydial Infection28-30
5. Gonococcal Infection28,31,32
6. Genital Herpes – first episode or episodic antiviral
treatment28,33,34
7. Bacterial Vaginosis28,35,36
8. Vulvovaginal Candidiasis28,37,38
9. Trichomonas vaginalis28,39-41
10. Acute or Chronic Bacterial Prostatitis42,43
Comments
Fluids only or loperamide hydrochloride 4 mg PO initially,
then 2 mg PO after each loose stool to a maximum of 8
mg per day or bismuth subsalicylate 4 tablets PO every half
hour, maximum 8 doses (A2a)
If no distressing symptoms: Fluids and loperamide or
bismuth subsalicylate
-if worsening consider single dose ciprofloxacin 500 mg PO
If distressing symptoms or critical trip: Oral fluid,
ciprofloxacin 500 mg PO BID for 72 h with or without
loperamide (A1a)
Increase oral fluids, ciprofloxacin 500 mg PO BID for 1-3
days; avoid loperamide (A1a for treatment; D5 for
classification of “severe”)
● Do not use loperamide if diarrhea is
associated with fever or blood in stool
Trimethoprim-sulfamethoxazole (TMP-SMX) 160/800 mg PO
BID for 3 days (A1a)
● If TMP-SMX resistance is greater than 20% for
community acquired UTI Ciprofloxacin 250 mg
PO twice daily for 3 days25
Ciprofloxacin 500 mg PO BID for 10 days (preferred) or
trimethoprim-sulfamethoxazole 160/80 mg PO for 14 days
(A2a)
Doxycycline 100 mg PO BID for 7 days (A1b)
Ciprofloxacin 500 mg single dose
-if treated or acquired in Hawaii, Asia or Pacific use cefixime
400 mg PO single dose (A1a)
Acyclovir 200 mg PO 5 times daily for 5 days (A1b)
Metronidazole 500 mg PO BID for 5-7 days or metronidazole
2g PO once (A1a)
Clotrimazole 1% cream 5 g intravaginally for 7-14 days ($) or
clotrimazole 100 mg 1 vaginal tablet for 7 days or 2
vaginal tablets for 3 days ($) or fluconazole 150 mg oral
tablet single dose (A1b)
Metronidazole 2 g PO single dose (A1a) or metronidazole 500
mg PO BID for 5-7 days (A1b)
Ciprofloxacin 500 mg PO BID for 28 days for acute or 6-12
weeks for chronic (B2b)
● Resistance to fluoroquinolones in
Campylobacter is increasingly common
● Fluoroquinolones may not be used for longer
than 3 weeks
● If compliance is an issue single-dose
azithromycin 1 g ($) may be more costeffective choice (A1b)
● Chlamydial infection accompanies 10-30% of
gonococcal infections so routine dual therapy
with doxycycline may be cost-effective
● For acute bacterial prostatitis, TMP/SMX 160/
800 mg PO twice daily may be used as well
based on local resistance patterns (B2b)
The American Journal of Medicine, Vol 120, No 4, April 2007
3. Uncomplicated acute pyelonephritis in women25-27
Treatment
Continued
Infection
Respiratory Infections
11. Bronchitis44,45
13. Community-Acquired Pneumonia48-57
Routine antibiotic treatment not justified unless pertussis
infection is suspected or in the situation of acute
exacerbation of chronic bronchitis. (A1a)
Penicillin V 250-500 mg PO 3-4 times daily for 10 days (B2b)
● Limit antibiotic prescribing to patients who
are most likely to have pharyngitis due to
infections with Group A beta-hemolytic
streptococcus as evidenced by diagnostic
testing and/or 2 or more of the following
symptoms: history of fever, tonsillar exudates,
no cough, and tender anterior cervical
lymphoadenopathy47
● Group A streptococcal pharyngitis is usually
self-limited with symptoms disappearing
spontaneously within 3-4 days on onset, even
without antibiotics45
● Antibiotic therapy can safely be postponed for
up to 9 days after onset of symptoms and still
prevent occurrence of the major
nonsuppurative sequela, acute rheumatic
fever45
● Respiratory quinolones are
levofloxacin 750 mg PO daily for 5 days or
moxifloxacin 400 mg PO daily for 10 days
● Advanced macrolides are
azithromycin 500 mg PO on day one then
250 mg PO daily on days 2-5 or
clarithromycin 500 mg PO BID for 10-14
days
● Beta-lactams are
amoxicillin 1 g PO TID for 10 days
amoxicillin/clavulanate extended-release
(Augmentin XR) 2 tablets PO BID for 10
days (equals 2 g amoxicillin plus 125 mg
clavulanate per dose)
299
● Previously healthy patient given no recent antibiotic
therapy: doxycycline 100 mg PO BID for 10-14 days (A1b)
or erythromycin 500 mg PO BID for 10-14 days (A1b)
● Previously healthy patient given recent antibiotic therapy:
respiratory quinolone (A1b) ($) or advanced macrolide
(A1b) and beta-lactam (A1b) ($) or telithromycin 800 mg
PO daily for 7-10 days (A1a) ($)
● Patient with comorbidities (COPD, diabetes, renal or
congestive heart failure, or malignancy) given no recent
antibiotic therapy: respiratory quinolone (A1b) ($) or
advanced macrolide (A1b) ($) or telithromycin 800 mg PO
daily for 7-10 days (A1a) ($)
● Patient with comorbidities given recent antibiotic therapy:
respiratory quinolone (A1b) ($) or advanced macrolide (A1b)
and beta-lactam (A1b) ($) or telithromycin 800 mg PO daily
for 7-10 days (A1a) ($)
● Suspected aspiration with infection: clindamycin 300 mg
PO four times daily for 10 days (B2a) ($) or amoxicillin/
clavulanate 875/125 mg PO BID for 10 days (A1b) ($)
● Influenza with bacterial superinfection: amoxillicin 1 g PO
TID for 10 days or amoxicillin/clavulanate extendedrelease 2 tablets PO BID for 10 days ($)
Comments
Low-Cost Antimicrobial Guideline
12. Pharyngitis (due to group A streptococci)46,47
Treatment
Hansen et al
Table
300
Table
Continued
Infection
58,59
14. Acute Bacterial Sinusitis
Skin Infections
15a. Mild to Moderate Inflammatory Acne60
15c. Severe Papulonodular Acne60
16. Early Lyme Disease61,62
Comments
Amoxicillin 500 mg PO TID for 10 days (A1a)
● Antibiotic therapy should be reserved for
patients with moderately severe symptoms
lasting at least 7-10 days who have maxillary
pain or tenderness in face or teeth
(symptomatic treatment preferred for patients
with mild symptoms)
Topical benzoyl peroxide 5% plus erythromycin 2% solution
applied BID for 8-12 weeks (A1a)
● Treatment continued until no new lesions
develop and then should be slowly
discontinued
● Use over-the-counter benzoyl peroxide and
generic topical erythromycin preparation
Oral tetracycline 500 mg BID for 6-8 weeks (preferred) or
doxycycline 100 mg PO BID for 6-8 weeks (A1a)
No anti-infective strategies suggested (A1a)
Doxycycline 100 mg PO BID for 10 days (A1b)
Mg ⫽ milligrams; kg ⫽ kilograms; PO ⫽ by mouth; BID ⫽ twice daily; TID ⫽ 3 times daily; $ ⫽ over $15 cost for course of therapy.
*All treatments are for adults unless otherwise noted.
● Therapeutic choice is appropriate for the
treatment of early localized or early
disseminated Lyme disease associated with
erythema migrans, in the absence of
neurological involvement or third-degree
atrioventricular heart block.
The American Journal of Medicine, Vol 120, No 4, April 2007
15b. Moderate to Severe Inflammatory Acne60
Treatment
Hansen et al
Low-Cost Antimicrobial Guideline
As with all guidelines, this guideline should be reviewed
on an ongoing basis, ideally at least annually, to incorporate
new treatment evidence, update medication cost information, and reflect the current health care payment status for
medications. While it is unlikely that a medication listed as
low cost will increase in price, new low-cost antimicrobials
become available as brand-name medications lose their
market rights and as new drugs are developed. In fact,
during the development process of this guideline, a number
of new generic antimicrobials came onto the market and
will most likely continue to decrease in price. In addition, it
is unknown what effect Medicare Part D, which started
January 1, 2006, will have on the affordability of prescription medications for Medicare recipients. However, the
overall issue of uninsured and underinsured patients will not
be resolved easily and will likely worsen as prescription
costs continue to soar and new expensive medications continue to be marketed. Lastly, future guideline revisions must
also consider changing microbial resistance patterns that
may necessitate removal of medications that become less
effective over time.
We describe an example of a multidisciplinary evidencebased guideline process that developed useful low-cost prescribing information for providers in our healthcare system.
It is hoped that this process can be used to guide future
low-cost guideline endeavors in other disease areas to more
broadly benefit patients and providers. Future research
should be focused on direct provider and patient interventions to improve the use of low-cost care guidelines and
whether the availability of low-cost guidelines improves the
occurrence and content of discussions about out-of-pocket
medications costs and, ultimately, the patient’s filling of the
prescription and adherence to the prescribed regimen. Further, future research measuring the impact of guideline
implementation on clinical and economic outcomes should
be conducted.
References
1. One in Three: Non-Elderly Americans without Health Insurance,
2002-2003. Washington, DC: Families USA Foundation; 2004.
2. Regopoulos LE, Trude S. Employers shift rising health care costs to
workers: no long-term solution in sight. Issue Brief Cent Stud Health
Syst Change. 2004;83:1-4.
3. Heisler M, Langa KN, Eby EL, Fendrick AM, Kabeto MU, Piette JD.
The health effects of restricting prescription medication use because of
cost. Med Care. 2004;42:626-634.
4. Harris Interactive. The impact of tiered co-pays: a survey of patients
and pharmacists. Research report. September 2003. Available at:
http://www.harrisinteractive.com/news/newsletters/healthnews/HI_
HealthCareNews2001Vol1_iss32.pdf. Accessed March 18, 2005.
5. Alexander GC, Casalino LP, Meltzer DO. Patient-physician communication about out-of-pocket costs. JAMA. 2003;290:953-958.
6. Alexander GC, Casalino LP, Tseng C-W, McFadden D, Meltzer DO.
Barriers to patient-physician communication about out-of-pocket
costs. J Gen Intern Med. 2004;19:856-860.
7. Piette JD, Heisler M, Wagner TH. Problems paying out-of-pocket
medication costs among older adults with diabetes. Diabetes Care.
2004;27:384-391.
301
8. Piette JD, Heisler M. Problems due to medication costs among VA and
non-VA patients with chronic illnesses. Am J Manag Care. 2004;10:
861-868.
9. Heisler M, Wagner TH, Piette JD. Patient strategies to cope with high
prescription medication costs: who is cutting back on necessities,
increasing debt, or underusing medications. J Behav Med. 2005;28:
43-51.
10. Reichert S, Simon T, Halm EA. Physicians’ attitudes about prescribing
and knowledge of the costs of common medications. Arch Intern Med.
2000;160:2799-2803.
11. Clauson KA, Seamon MJ, Clauson AS, Van TB. Evaluation of drug
information databases for personal digital assistants. Am J Health Syst
Pharm. 2004;61:1015-1024.
12. Rothschild JM, Lee TH, Bae T, Bates DW. Clinician use of a palmtop
drug reference guide. J Am Med Inform Assoc. 2002;9:223-229.
13. Korn LM, Reichert S, Simon T, Halm EA. Improving physicians’
knowledge of the costs of common medication and willingness to
consider costs when prescribing. J Gen Intern Med. 2003;18:31-37.
14. Depiero AD, Ochsenschlager DW, Chamberlain JM. Analysis of pediatric hospitalizations after emergency department release as a quality
improvement tool. Ann Emerg Med. 2002;39:159-163.
15. Taira DA, Iwane KA, Chung RS. Prescription drugs: elderly enrollee
reports of financial access, receipt of free samples, and discussion of
generic equivalents to type of coverage. Am J Manag Care. 2003;9:
305-312.
16. Matsuli D, Joubert GI, Dykxhoorn S, Rieder MJ. Compliance with
prescription filling in the pediatric emergency department. Arch Pediatr Adolesc Med. 2000;154:195-198.
17. Thomas EJ, Burstin HR, O’Neil AC, Orav EJ, Brennan TA. Patient
noncompliance with medical advice after the emergency department
visit. Ann Emerg Med. 1996;27:49-55.
18. Hass JA, Phillips KA, Gerstenberger EP, Seger AC. Potential savings
from substituting generic drugs for brand-name drugs: medical expenditure panel survey, 1997-2000. Ann Intern Med. 2005;142:891-897.
19. Gilbert DN, Moellering RC, Eiopoulos GM, Sande MA. The Sanford
Guide to Antimicrobial Therapy, 34th ed. Hyde Park, VT: Antimicrobial Therapy, Inc; 2004.
20. Wisconsin Medicaid. Wisconsin Department of Health and Family
Services Website. Available at: http://dhfs.wisconsin.gov/medicaid4/
pharmacy/data_tables/legend_mac.asp. Accessed July 5, 2005.
21. Sackett DL, Strauss SE, Richardson WS, Rosenberg W, Haynes RB.
Evidence-Based Medicine: How to Practice and Teach EBM, 2nd ed.
Oxford, UK: Churchill Livingstone; 2000.
22. Cheng AC, Thielman NM. Update on traveler’s diarrhea. Curr Infect
Dis Rep. 2002;4:70-77.
23. Adachi JA, Ostrosky-Zeichner L, DuPont HL, Ericsson CD. Empirical
antimicrobial therapy for traveler’s diarrhea. Clin Infect Dis. 2000;31:
1079-1083.
24. Caeiro JP, DuPont HL, Albrecht H, Ericsson CD. Oral rehydration
therapy plus loperamide versus loperamide alone in the treatment of
traveler’s diarrhea. Clin Infect Dis. 1999;28:1286-1289.
25. Fihn SD. Acute uncomplicated urinary tract infections in women.
N Engl J Med. 2003;349:259-266.
26. Warren JW, Abrutyn E, Hebel R, Johnson JR, Schaeffer AJ, Stamm
WE. Guidelines for antimicrobial treatment of uncomplicated acute
bacterial cystitis and acute pyelonephritis in women. Clin Infect Dis.
1999;29:745-758.
27. Gupta K, Hooton TM, Stamm WE. Increasing antimicrobial resistance
and the management of uncomplicated community-acquired urinary
tract infections. Ann Intern Med. 2001;135:41-50.
28. Centers for Disease Control and Prevention. Sexually transmitted
disease guidelines 2002. MMWR Recomm Rep. 2002;51:1-78.
29. Horner PJ, Caul EO. National guidelines for the management of
Chlamydia trachomatis genital tract infection. Sex Transm Infect.
1999;75:S4-S8.
30. Hillis SD, Coles FBD, Litchfield B, et al. Doxycycline and azithromycin for prevention of chlamydial persistence or recurrence one
302
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
month after treatment in women: a use-effectiveness study in public
health settings. Sex Transm Dis. 1998;25:5-11.
Bignell C. National guidelines for the management of gonorrhoea in
adults. Sex Transm Infect. 1999;75:S13-S15.
Echols RM, Heyd A, O’Keeffe BJ, Schacht P. Single-dose ciprofloxacin for the treatment of uncomplicated gonorrhea: a worldwide summary. Sex Transm Dis. 1996;23:165-167.
Barton S, Brown D, Cowan F, et al. National guidelines for the
management of genital herpes. Sex Transm Infect. 1999;75:S24-S28.
Fife KH, Barbarash RA, Rudolph T, Degregorio B, Roth R. The
valacyclovir international herpes simplex virus study group. Valacyclovir versus acyclovir in the treatment of first-episode genital herpes
infection: results of an international, multicenter, double-blind, randomized clinical trial. Sex Transm Dis. 1997;24:481-486.
Hay P. National guidelines for the management of bacterial vaginosis.
Sex Transm Infect. 1999;75:S16-S18.
Lugo-Miro VI, Green M, Mazur L. Comparison of different metronidazole therapeutic regimens for bacterial vaginosis. A meta-analysis.
JAMA. 1992;268:92-95.
Daniels D, Forster G. National guidelines for the management of
vulvovaginal candidiasis. Sex Transm Infect. 1999;75:S19-S20.
De Punzio C, Garutti P, Mollica G, Nappi C, Piccoli R, Genazzani AR.
Fluconazole 150 mg single dose versus itraconazole 200 mg per day for
3 days in the treatment of acute vaginal candidiasis: a double-blind
randomized study. Eur J Obstet Gynecol Reprod Biol. 2003;106:193-197.
Sherrard J. National guidelines for the management of Trichomonas
vaginalis. Sex Transm Infect. 1999;75:S21-S23.
Gülmezoglu AM. Interventions for trichomoniasis in pregnancy. Cochrane Database Syst Rev. 2002;3.
Hager WD, Brown ST, Kraus SJ, Kleris GS, Perkins GJ, Henderson
M. Metronidazole for vaginal trichomoniasis. Seven-day vs singledose regimens. JAMA. 1980;12:1219-1220.
Walker P, Wilson J. National guidelines for the management of prostatitis. Sex Transm Infect. 1999;75:S46-S50.
Fowler JE. Antimicrobial therapy for bacterial and nonbacterial prostatitis. Urology. 2002;20(Suppl 6A):S24-S26.
Smucny J, Fahey T, Becker L, Glazier R. Antibiotics for acute bronchitis. The Cochrane Database of Systematic Review [database online].
Cochrane Acute Respiratory Infections Group; 2000. Updated December 15, 2003.
Gonzales R, Bartlett JG, Besser RE, et al. Principles of appropriate
antibiotic use for the treatment of uncomplicated acute bronchitis:
background. Ann Emerg Med. 2001;37:720-727.
Bisno AL, Gerber MA, Gwaltney JM Jr, Kaplan EL, Schwartz RH.
Practice guidelines for the diagnosis and management of group a
streptococcal pharyngitis. Clin Infect Dis. 2002;35:113-125.
Cooper RJ, Hoffman JR, Bartlett JG, et al. Principles of appropriate
antibiotic use for acute pharyngitis in adults: background. Ann Emerg
Med. 2001;37:711-719.
Mandell LA, Bartlett JG, Dowell SF, File TM, Musher DM, Whitney
C. Update of practice guideline for the management of communityacquired pneumonia in immunocompetent adults. Clin Infect Dis.
2003;37:1405-1433.
Ailania RK, Agastya G, Ailani RK, Mukunda BN, Shekar R. Doxycycline is a cost-effective therapy for hospitalized patients with community-acquired pneumonia. Arch Intern Med. 1999;159:266-270.
The American Journal of Medicine, Vol 120, No 4, April 2007
50. Gleason PP, Kapoor WN, Stone RA, et al. Medical outcomes and
antimicrobial costs with the use of the American thoracic society
guidelines for outpatients with community-acquired pneumonia.
JAMA. 1997;278:32-39.
51. Dunbar LM, Wunderink RG, Habib MP, et al. High-dose, short-course
levofloxacin for community-acquired pneumonia: a new treatment
paradigm. Clin Infect Dis. 2003;37:752-760.
52. Dunbar LM, Khashab MM, Kahn JB, Zadeikis N, Ziang JX, Tennenberg AM. Efficacy of 750-mg, 5-day levofloxacin in the treatment of
community-acquired pneumonia caused by atypical pathogens. Curr
Med Res Opin. 2004:20:555-563.
53. Sánchez F, Mesna J, Martínez A, et al. Is azithromycin the first-choice
macrolide for treatment of community-acquired pneumonia? Clin Infect Dis. 2003;36:1239-1245.
54. Kasbekar N, Acharya PS. Telithromycin: the first ketolide for the
treatment of respiratory infections. Am J Health Syst Pharm. 2005;62:
905-916.
55. Levison ME, Manura CT, Lorber B, et al. Clindamycin compared with
penicillin for the treatment of anaerobic lung abscess. Ann Intern Med.
1983;98:466-471.
56. Perlino CA. Metronidazole vs clindamycin treatment of anaerobic
pulmonary infection. Failure of metronidazole therapy. Arch Intern
Med. 1981;141:1424-1427.
57. Kapila R, Sen P, Salaki J, Louria DB. Evaluation of clindamycin and
other antibiotics in the treatment of anaerobic bacterial infections of
the lung. J Infect Dis. 1977;135(suppl):S58-S64.
58. Hickner JM, Barlett JG, Besser RE, Gonzales R, Hoffman JR, Sande
M. Principles of appropriate antibiotic use for acute rhinosinusitis in
adults: background. Ann Emerg Med. 2001;37:703-710.
59. Ip S, Fu L, Balk E, Chew P, DeVine D, Lau J. Update on Acute
Bacterial Rhinosinusitis. Prepared for: Agency for Healthcare Research and Quality. U.S. Department of Health and Human Services.
Rockville, MD; 2005.
60. Haider A, Shaw JC. Treatment of acne vulgaris. JAMA. 2004;292;726735.
61. Wormser GP, Nadelman RB, Dattwyler RJ, et al. Practice guidelines
for the treatment of Lyme disease. Clin Infect Dis. 2000;31(suppl
1):S1-S14.
62. Wormser GP, Ramanathan R, Nowakoaski JM, et al. Duration of
antibiotic therapy for early Lyme disease. Ann Intern Med. 2003;138:
697-704.
63. Shiffman RN, Shekelle P, Overhage M, Slutsky J, Grimshaw J, Deshpande AM. Standardized reporting of clinical practice guidelines: a
proposal from the conference on guideline standardization. Ann Intern
Med. 2003;139:493-498.
64. Fischer MA, Avorn J. Economic consequences of underuse of generic
drugs: evidence from Medicaid and implications for prescription drug
benefit plans. Health Serv Res. 2003;38:1051-1063.
65. Nutescu EA, Park HY, Walton SM, et al. Factors that influence
prescribing within a therapeutic drug class. J Eval Clin Pract. 2005;
11:357-365.
66. Lichtenberg FR. Are the benefits of newer drugs worth their cost?
Evidence from the 1996 MEPS. Health Aff. (Millwood) 2001;20:241251.
The American Journal of Medicine (2007) 120, 303-305
REVIEW
Contemporary Empyema Necessitatis
Scott A. Kono, DO,a Trenton D. Nauser, MDb
a
Department of Internal Medicine, University of Kansas Medical Center, Kansas City, Kans; bDivision of Pulmonary and Critical Care
Medicine, Department of Veterans Affairs Medical Center, Kansas City, Mo.
ABSTRACT
Empyema necessitatis is defined by the extension of an empyema through the parietal pleura, into
surrounding tissue. Clinical manifestations are generally subacute, representing the indolent nature of the
most commonly implicated pathogens (Mycobacterium tuberculosis and Actinomyces israelii). Treatment
with antimicrobials and surgical debridement has drastically reduced mortality; however, with proper
initial assessment and treatment of pulmonary infection, this rare complication can often be avoided. We
describe a patient with empyema necessitatis to illustrate the importance of timely diagnosis and treatment
of parapneumonic pleural effusion, and the need to consider this diagnosis in patients presenting with
constitutional symptoms and a chest wall mass. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Empyema; Pulmonary infection; Chest wall mass; Pleural effusion
A parapneumonic pleural effusion is a complication of bacterial pneumonia, lung abscess, or bronchiectasis. At least
40% of patients with pneumonia have an associated effusion.1 The evolution of a parapneumonic effusion is a continuum of stages. Dependent upon the stage, the fluid varies
from a clear, straw-colored fluid to frank pus (empyema).
In the first or exudative stage, a lower respiratory infection may increase pulmonary interstitial fluid, which crosses
the visceral pleura. When the resorptive capacity of the
pleural space is exceeded, pleural fluid accumulates. Pleural
fluid pH and glucose are normal and the lactate dehydrogenase (LDH) usually less than 500 U/L. These effusions are
uncomplicated, do not require drainage, and often resolve
after appropriate antimicrobial therapy.
The second stage is fibropurulent, in which bacteria from
the contiguous pulmonary infection invade the pleura. The
pleural fluid becomes increasingly turbid and viscous due
to leukocytes, cellular debris, and fibrin. Diagnostically,
the pleural effusion has a pH ⬍7.2, glucose ⬍40 mg/dL,
and LDH ⬎1000 U/L. These complicated parapneumonic
effusions and empyemas require antimicrobials and
drainage for control of inflammation and pleural sepsis.
Requests for reprints should be addressed to Scott A. Kono, DO,
Department of Internal Medicine, University of Kansas Medical Center,
3901 Rainbow Blvd., Kansas City, KS 66160.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.09.019
A tendency exists toward forming limiting membranes,
or loculations, which may need evacuation via thoracotomy rather than simple tube thoracostomy.
In managing pleural effusions, thoracentesis should be
done immediately to help direct therapy.2,3 If untreated, an
empyema may rupture through the parietal pleura and dissect into the surrounding tissue with resultant abscess formation (empyema necessitatis).
CASE REPORT
A 52-year-old man with no medical history was admitted
with new onset of painful left chest mass, and 2 weeks of
cough, dyspnea, and fever. One week before admission, he
completed a 5-day course of azithromycin for presumptive
pneumonia. A chest radiograph was not performed. He
regularly used cigarettes, ethanol and marijuana, and denied
ill contacts.
On admission, he appeared moderately distressed. Vital
signs included temperature 38.7°C, respiratory rate 24
breaths per minute, pulse 85 beats per minute, blood pressure 154/83 mm Hg, and SpO2 95% on room air. His
dentition was poor. A 7 ⫻ 8 cm nonerythematous, fluctuant,
tender left anterior chest mass was noted. Left lung examination revealed absent tactile fremitus, percussion dullness,
and diminished breath sounds. Extremity examination revealed no clubbing.
304
The American Journal of Medicine, Vol 120, No 4, April 2007
contribute to this disease.6-8 Extension occurs along the path
Laboratory investigation revealed WBC 15,400/mm3
with 70% neutrophils and 8% bands. Alkaline phosphatase
of least resistance. A painful anterior chest wall mass, typwas 165 IU/L. A large left pleural effusion was evident on
ically between the second and sixth intercostal space, is the
chest radiograph (Figure 1). Computed tomography conmost common presenting symptom.4,5 Alternative areas of
firmed a large loculated effusion with direct anterior extenextension include the bronchi, esophagus, breast, diaphragm,
sion into the chest wall and adjaretroperitoneum, and groin.4-6,9-19
Tuberculosis is the most comcent rib destruction (Figures 2, 3).
monly reported etiology, followed
Thoracentesis revealed a grossly
CLINICAL SIGNIFICANCE
by actinomycosis.5 The chronic
purulent fluid with glucose 1 mg/
nature of these infections corredL, LDH 7019 IU/L, and WBC
● Pleural effusions require immediate and
lates well with the typical slow
48,500/mm3 with 76% neutroaccurate diagnostic evaluation to miniphils. The pH was spuriously eleprogression of disease. Constitumize morbidity and mortality.
vated due to air accumulation
tional symptoms typically begin
● Empyema necessitatis, as described in
within the syringe. Pleural fluid
1-2 months before development of
Gram’s stain was negative.
a tender chest wall mass, although
this case, is a rare complication of an
Piperacillin-tazobactam was inisymptoms up to 6 years prior have
untreated or under-treated parapneumotiated and the patient taken for
been reported. More acute presennic effusion.
video-assisted thoracoscopic surtations are associated with pyo● Suspect empyema necessitatis in any pagery where extensive exudates and
genic bacteria.4,5
tient presenting with a chest wall mass
adhesions were removed from the
Although anaerobic organisms
and underlying pulmonary infection.
chest wall and lung surface. Turbid
are the predominant pathogens
pleural fluid was drained and aderecovered from empyema in gen● Successful therapy for empyema necesquate lung expansion achieved. All
eral,20 anaerobes other than Actisitatis
is
dependent
upon
prompt
anticultures remained negative and the
nomyces israelii are only rarely
microbial and surgical intervention.
patient recovered uneventfully after
associated with empyema necessicompleting 6 weeks of broad spectatis. Often, no causative organism
trum antibiotics.
is identified, which may reflect
difficulty in culturing anaerobic bacteria.4,12 Other causal
organisms include Streptococcus pneumoniae, Staphylococcus
DISCUSSION
aureus, Streptococcus milleri, Pseudomonas cepacia, and MyEmpyema necessitatis is rare, with the majority of cases
cobacterium avium-intracellulare.5,15,16
described in the preantibiotic era. In 1640 Guillan de BailTreatment of empyema necessitatis involves antimicrolon reported the first case, a syphilitic aneurysm with disbials
combined with surgical drainage.4,5 Antibiotics should
charge of pus through the chest wall.4 Sindel reported 115
be initiated to cover potential polymicrobial and anaerobic
subjects in 1940,4 and a recent review cited only 26 cases
pleuropulmonary infection, and tailored when sensitivities
since 1966.5
become available. Empiric antitubercular medications should
Increased pressure within the loculation, chronic inflambe considered.
mation, and necrosis with erosion and fluid extension all
Before antibiotics, mortality was 87% among those with
tuberculosis as the causative agent and 28% in those with a
Figure 1 Chest radiograph demonstrating a large left pleural
effusion with contralateral shift of mediastinal structures.
Figure 2 CT scan demonstrating anterolateral communication
of empyema into subcutaneous tissue.
Kono and Nauser
Contemporary Empyema Necessitatis
Figure 3 CT scan revealing anterior focal rib destruction secondary to empyema involvement.
pyogenic organism.4 Since the advent of modern antimicrobial and surgical therapy, death has not been reported.
Our patient represents a rare complication of parapneumonic pleural effusion, a common illness. Outpatient antibiotics may have treated his pneumonia, but left a parapneumonic pleural effusion inadequately treated. Negative
pleural fluid culture may have resulted from sterilization by
antibiotics before sampling or from a lack of laboratory
sensitivity in detecting anaerobic organisms. Alternatively,
complicated parapneumonic effusions can be sterile because
bacteria are cleared rapidly from the pleural space.21
The ‘empyema of necessity’ should be considered in any
patient presenting with chest mass and constitutional symptoms. However, timely pleural fluid analysis can prevent
this complication. In 1893, Duckworth conveyed fittingly,
“ . . . it is not common for empyema to open spontaneously,
and it is a less common occurrence now than it used to be,
for the reason that nowadays the hypodermic syringe makes
the diagnosis certain and the pus is soon withdrawn.”22
References
1. Light RW, Girard WM, Jenkinson SG, et al. Parapneumonic effusions.
Am J Med. 1980;69:507-512.
2. Sahn SA, Light RW. The sun should never set on a parapneumonic
effusion. Chest. 1989;95:945-947.
305
3. Colice GL, Curtis A, Deslauriers J, et al. Medical and surgical treatment of parapneumonic effusions. An evidence-based guideline.
Chest. 2000;18:1158-1171.
4. Sindel EA. Empyema necessitatis. Q Bull Seaview Hosp. 1940;6:1-49.
5. Freeman AF, Ben-Ami T, Shulman ST. Streptococcus pneumoniae
empyema necessitatis. Pediatr Infect Dis J. 2004;23:177-179.
6. Marks MI, Eickhoff TC. Empyema necessitatis. Am Rev Respir Dis.
1970;101:759-761.
7. Bernstein A, White FZ. Unusual physical findings in pleural effusion:
intrathoracic manometric studies. Ann Intern Med. 1952;37:733-738.
8. Haddad CJ, Sim WK. Empyema necessitatis. Am Fam Physician.
1989;40:149-152.
9. Bassiri AG, Girgis RE, Theodore J. Actinomyces odontolyticus thoracopulmonary infections. Two cases in lung and heart-lung transplant
recipients and a review of the literature. Chest. 1996;109:1109-1111.
10. Davies AL, Messerschmidt W, Piland JH, Lally J. Thoracic actinomycosis presenting as empyema necessitatis. Del Med J. 1987;59:649657.
11. Dershaw DD. Actinomycosis of the chest wall: ultrasound findings in
empyema necessitatis. Chest. 1984;86:779-780.
12. Hockensmith ML, Mellman DL, Aronsen EL. Fusobacterium nucleatum empyema necessitans. Clin Infect Dis. 1999;29:1596-1598.
13. Hoover EL, Hsu HK, Minnard E. Contained thoracic empyema presenting as an anterior chest wall abscess: empyema necessitatis revisited. J Natl Med Assoc. 1989;81:912-917.
14. Jover F, Andreu L, Cuadrado J, Montaagud J, Merino J. Tuberculous
empyema necessitatis in a man infected with the human immunodeficiency virus. South Med J. 2002;95:751-752.
15. Marinella MA, Harrington GD, Standiford TJ. Empyema necessitans
due to Streptococcus milleri. Clin Infect Dis. 1996;23:203-204.
16. Noyes BE, Michaels MG, Kurland G, Armitage JM, Orenstein DM.
Pseudomonas cepacia empyema necessitatis after lung transplantation
in two patients with cystic fibrosis. Chest. 1994;105:1888-1891.
17. Reyes CV, Thompson KS, Jensen J. Fine needle aspiration biopsy of
mastitis secondary to empyema necessitatis. Acta Cytol. 1999;43:873876.
18. Sahn SA, Iseman MD. Tuberculous empyema. Semin Respir Infect.
1999;14:82-87.
19. Sakomoto T, Miyamoto Y, Nishio W, Matsuoka H, Tsubota N. Empyema necessitatis into the retroperitoneal space. Ann Thorac Surg.
2002;73:1954-1956.
20. Bartlett JG, Finegold SM. Anaerobic infections of the lung and pleural
space. Am Rev Respir Dis. 1974;110:56-77.
21. Mavroudis C, Tomlinson JR, Wilson C, et al. Experimental aerobic
anaerobic thoracic empyema in the guinea pig. Ann Thorac Surg.
1987;43:298.
22. Duckworth SP. Case of empyema which opened externally in the
seventh right intercostal space. Int Clin Philadelphia. 1893;1:19-25.
The American Journal of Medicine (2007) 120, 306-315
UPDATE IN OFFICE MANAGEMENT
Office Management of Chronic Pain in the Elderly
Debra K. Weiner, MD
Division of Geriatric Medicine, Department of Medicine, Department of Psychiatry, and Department of Anesthesiology,
University of Pittsburgh, Pittsburgh, Penn.
ABSTRACT
Chronic pain plagues older adults more than any other age group; thus, practitioners must be able to
approach this problem with confidence and skill. This article reviews the assessment and treatment of the
most common chronic nonmalignant pain conditions that affect older adults—myofascial pain, generalized
osteoarthritis, chronic low back pain (CLBP), fibromyalgia syndrome, and peripheral neuropathy. Specific
topics include essential components of the physical examination; how and when to use basic and advanced
imaging in older adults with CLBP; a stepped care approach to treating older adults with generalized
osteoarthritis and CLBP, including noninvasive and invasive management techniques; how to diagnose
and treat myofascial pain; strategies to identify the older adult with fibromyalgia syndrome and avoid
unnecessary “diagnostic” testing; pharmacological treatment for the older adult with peripheral neuropathy; identification and treatment of other factors such as dementia and depression that may significantly
influence response to pain treatment; and when to refer the patient to a pain specialist. While common,
chronic pain is not a normal part of aging, and it should be treated with an emphasis on improved physical
function and quality of life. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Chronic pain; Elderly; Evaluation; Treatment; Low back pain; Fibromyalgia; Syndrome; Myofascial
pain; Peripheral neuropathy; Generalized osteoarthritis
Chronic pain is a treatable problem affecting an estimated
50% of community-dwelling older adults and over 75% of
long-term care facility residents. Because of its numerous
consequences, including impaired physical function, depression, anxiety, disrupted sleep and appetite, and excessive use of health care resources, chronic pain should be
treated aggressively. Chronic pain may accompany both
malignant and nonmalignant disorders. This article focuses
on the evaluation and management of chronic nonmalignant
pain.
DEFINITION AND PATHOGENESIS OF CHRONIC
NONMALIGNANT PAIN
speaking there are 2 types of conditions that underlie its
pathogenesis—nociceptive and neuropathic. The nervous
system is responsible for the perception of pain. Nociceptive
pain is associated with tissue damage and a normal nervous
system (eg, pain associated with osteoarthritis), while neuropathic pain is associated with physiological nervous system dysfunction (eg, diabetic neuropathy, postherpetic neuralgia). Not infrequently, these 2 types of pain coexist. The
discussion that follows focuses on the most common
chronic nonmalignant pain disorders—myofascial pain, osteoarthritis, chronic low back pain, fibromyalgia syndrome,
and peripheral neuropathy.
Chronic pain is defined as pain that persists beyond the
expected time of healing, or more than 3 to 6 months. While
a number of disorders may cause chronic pain, generally
GENERAL APPROACH TO ASSESSMENT
AND TREATMENT
Requests for reprints should be addressed to Debra K. Weiner, MD,
UPMC Pain Medicine at Centre Commons, 5750 Centre Ave., Suite 400,
Pittsburgh, PA 15206.
E-mail address: [email protected]
To afford an optimal response to therapy, the practitioner
must keep 3 general principles in mind, and communicate
them to their older patients in order to establish reasonable
treatment expectations:
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.05.048
Weiner
●
●
●
Chronic Pain in the Elderly
Chronic pain is a syndrome with many potential contributors, all of which require treatment to afford an optimal
clinical outcome.
Chronic pain is treatable but not curable; improvement is
the rule, not the exception.
It is often possible to improve functional ability to a
greater extent than the severity of pain is reduced.
EVALUATION AND TREATMENT OF SPECIFIC
CHRONIC
PAIN DISORDERS
Myofascial Pain
Definition and Pathogenesis. Because myofascial pain
(MP) occurs in the majority of patients with chronic pain
disorders, learning how to recognize, diagnose, and treat
this condition is essential. Broadly defined, myofascial pain
is pain in physiologically abnormal muscles. The pathognomonic features of this disorder are taut bands and trigger
points elicited by firm palpation.1 Patients may present with
a range of pain symptoms such as aching, soreness, stabbing, or burning. The descriptors used by patients suffering
from myofascial pain may be similar to those of neuropathic
pain, and sometimes these disorders occur together; thus,
hands-on assessment is critical to accurate diagnosis. Myofascial pain is commonly associated with degenerative
changes of the axial skeleton and other factors (eg, spinal
deformity, such as scoliosis or kyphosis) that may irritate
nerves. As a result, the musculature supplied by these
nerves does not receive adequate trophic factors and it, in
turn, becomes irritable.2
Evaluation. Careful physical examination is required to
diagnose myofascial pain. Historical clues include: relief of
pain by exerting mild pressure over the area (eg, massaging
the affected area), engaging in low level activity, or application of heat or vapocoolant spray; and worsening of pain
with excessive activity. Physical examination performed by
firmly palpating involved muscles may reveal both latent
and active trigger points.1 Palpation of latent trigger points
may cause tenderness, but may not reproduce the radiation
of pain that occurs when active trigger points are palpated.
Taut bands can be identified by firmly palpating the muscle
along the same plane but in the axis that runs perpendicular
to the direction of the muscle fibers. The examiner may need
to apply pressure in this area for 10-15 seconds before the
characteristic pain pattern is reproduced.3 Local twitch responses, due to sharp contractions of the taut band (not the
entire muscle) initiated by an intense physical stimulus such
as needle insertion or plucking of the band may also be
noted.1 Sensory abnormalities include hypersensitivity and
allodynia (ie, pain caused by a stimulus that does not normally provoke pain) of the trigger points. Autonomic phenomena, such as temperature change, piloerection, and
sweating, may be present in the affected area but are not
required for diagnosis.
307
Treatment. The first step in the treatment of MP is identification and modification of perpetuating factors, such as
posture, sleeping position, and body mechanics during activity. Early involvement of a physical therapist skilled in
the treatment of myofascial pain disorders is crucial. Physical therapy techniques commonly employed include gentle
sustained stretching of involved muscles to reduce focal
contractions and inactivate trigger points, sometimes in
combination with a vapocoolant spray; exercise programs
that include graded stretching and strengthening; and teaching of flare self-management techniques.4 In order to facilitate physical therapy-directed rehabilitation, trigger point
deactivation may be useful. This can be accomplished with
either “dry needling” (ie, deactivating the trigger point with
a needle alone) or injection of the trigger point with an
anesthetic or steroid. Either technique may be effective as
long as a local twitch response (ie, transient contraction of
a group of tense muscle fibers [taut band] that traverse a
trigger point) is achieved.5 This author also has had success
with a variety of acupuncture techniques. In refractory
cases, referral to a multidisciplinary pain center may be
helpful.
Osteoarthritis and Chronic Low Back Pain
Definition and Pathogenesis. Osteoarthritis (OA) is commonly referred to as degenerative arthritis, but this is somewhat of a misnomer. Accumulating evidence indicates that
inflammation plays an important role in the pathogenesis of
OA. While virtually all older adults have radiographic evidence of OA, most of these individuals are asymptomatic,
and the severity of OA on radiograph correlates poorly with
symptom severity. In the case of chronic low back pain
(CLBP), degenerative disease of the discs and facets is only
one factor that contributes to pain; thus, its evaluation and
treatment should be approached comprehensively.
Evaluation. Proper treatment of appendicular OA (eg,
knees, hips) and CLBP requires a detailed evaluation. Several rheumatologic disorders can cause generalized pain and
may be confused with OA. Table 1 summarizes clinical
features of common rheumatologic conditions.
The first step in evaluating low back pain is the identification of “red flags” (eg, fever, unintentional weight loss,
sudden change in pain quality) indicative of a serious underlying disorder such as malignancy or spinal infection
(Figure 1). These should be screened with a targeted history
and physical examination. If a serious condition is suspected, diagnostic imaging should be pursued promptly.
Conversely, in the absence of red flags, imaging is rarely
indicated.6 Practitioners receive little training in musculoskeletal assessment, thus, basic and advanced imaging (ie,
radiographs and magnetic resonance imaging [MRI]/computerized tomography [CT], respectively) is often relied
upon as an initial diagnostic test. This approach, however, is
likely to lead to incorrect management because degenerative
abnormalities are a nonspecific finding in older adults with
Differentiation of Osteoarthritis from other Common Rheumatologic Disorders: History, Physical Examination and Other Diagnostic Features
History
308
Table 1
Physical Examination
AM Stiffness
Location of Pain
Synovitis
Extrasynovial Disease
Osteoarthritis
Generally short-lived,
eg, ⬍30 minutes
Absent or mild
None related to arthritis itself
Pseudogout
Pseudo-rheumatoid
pattern may be
associated with
prolonged AM
stiffness
Weight-bearing appendicular joints,
cervical and lumbar spine, DIPs,
PIPs and 1st CMC. MCP and wrist
involvement go against OA.
Knee and wrist are most common
locations; disease is often
symmetrical
Acute flares are intensely
inflammatory
Gout
Pseudo-rheumatoid
pattern may be
associated with
prolonged AM
stiffness
Joints of the lower extremities are
most often involved, especially 1st
MTP; disease is typically
asymmetrical
Acute flares are intensely
inflammatory
Chondrocalcinosis on
radiographs; eye deposits,
bursitis, tendonitis, carpal
and cubital tunnel syndromes
may occur. Tophaceous soft
tissue deposits uncommon.
Tophi may deposit in soft
tissues.
Rheumatoid
Arthritis
Prolonged, eg, ⬎30
minutes. Duration
of stiffness is used
as one parameter
of disease activity
Any synovial joint. The lumbar
spine is typically spared.
Present.
Systemic Lupus
Erythematosus
Not a prominent
feature.
Generally absent; arthralgias
are more common than
arthritis.
Fibromyalgia
Syndrome
Generally short-lived,
eg, ⬍30 minutes
Depends upon tissues involved—
may or may not be limited to
joints. Comorbid fibromyalgia is not
uncommon.
Typically diffuse. Worst symptoms
often involve the axial skeleton.
Polymyalgia
Rheumatica
May be prolonged,
lasting several
hours
Typically proximal—eg, shoulder
girdle, hip girdle, neck. If
headaches, jaw claudication, or
prominent systemic symptoms (eg,
fever), consider temporal arteritis.
Absent. Joints themselves
are not involved, although
patients experience pain in
joints and soft tissues.
May occur, especially in
small joints of hands.
Not uncommon; rheumatoid
nodules can develop in soft
tissues. Many other possible
manifestations including
anemia, vasculitis (skin
lesions, peripheral
neuropathy, pericarditis,
visceral arteritis, palpable
purpura), pulmonary disease,
etc.
Common—eg, anemia, skin
rash, pleuritis, peritonitis,
pericarditis, nephritis,
meningitis, etc.
Many other disorders may
coexist (Table 4)
Occurs if comorbid temporal
arteritis and relates to
involvement of arteries (eg,
Raynaud’s phenomenon,
bruits, claudication).
Other Diagnostic
Features/Comments
Because OA is ubiquitous in older
adults, radiographs should be used to
rule out other disorders, not to
diagnose OA.
Chondrocalcinosis may be
asymptomatic. Identification of
intracellular CPPD crystals
offers a definitive diagnosis in acute
flares. Acute and chronic forms occur.
Hyperuricemia may be asymptomatic.
Serum uric acid cannot diagnose gout.
Identification of intracellular
monosodium urate monohydrate
crystals offers a definitive diagnosis in
acute flares.
Patients may be seronegative. If
disease is suspected, patient should
promptly be referred to a
rheumatologist to retard disease
progression.
Anyone with suspected SLE should
promptly be referred to a
rheumatologist.
Fibromyalgia syndrome is not a
diagnosis of exclusion, but one based
upon careful history and physical
examination
(see text).
Because the erythrocyte sedimentation
is very nonspecific, this test should be
used to assist with confirmation of a
suspected diagnosis. Note that cases
of PMR and TA with a normal ESR have
been reported.
CMC ⫽ carpo-metacarpal joint; CPPD ⫽ calcium pyrophosphate dihydrate; DIP ⫽ distal interphalangeal joint; ESR ⫽ erythrocyte sedimentation rate; MCP ⫽ metacarpophalangeal joint;
MTP ⫽ metatarsophalangeal joint; OA ⫽ osteoarthritis; PIP ⫽ proximal interphalangeal joint; PMR ⫽ polymyalgia rheumatica; SLE ⫽ systemic lupus erythematosus; TA ⫽ temporal arteritis.
The American Journal of Medicine, Vol 120, No 4, April 2007
Disorder
Weiner
Chronic Pain in the Elderly
309
Comprehensive Patient History
Evidence of
Lumbar Spinal Stenosis?
NO
YES
Red Flags?
Symptoms Severe
and/or associated with
recent neurological deficits?
NO
Basic or Advanced
Imaging;
Treat per findings
Comprehensive Physical Examination
of the Low Back and Legs
NO
YES
YES
Fibromyalgia
Physical Therapy
+
Analgesics
Advanced Imaging
+
Surgical Referral
Sacroiliac pain,
Myofascial pain
Rheumatology or
Pain Clinic
Prominent
Hip Disease
Orthopedist
Physical Therapy
Adjunctive Modalities and Medications
Figure 1
Evaluation and treatment of the older adult with persistent low back pain.
back pain. Indeed, virtually 100% of older adults have
degenerative pathology, whether or not they have pain.
When used as a screening tool, imaging should be thought
of as a way to demonstrate the absence of disease (eg,
Table 2
compression fractures, metastatic bone disease, disk space
infection) rather than as a way to diagnose the cause of pain.
Moderate to severe central spinal canal stenosis identified by MRI is also as common in older adults with pain as
Essential Clinical History Questions for Older Adults with Chronic Low Back Pain
Question
Potential Diagnostic Clue(s) Obtained
1. Can you show me where your back hurts?
If patient places hand to right or left of midline, over sacrum rather than lumbar
spine, this suggests sacroiliac joint syndrome (look for associated scoliosis, hip
or knee disease, leg length discrepancy), inflammatory disorder, or sacral
insufficiency fracture.
Improvement in fetal position suggests spinal stenosis. Worsening in fetal
position suggests sacroiliac disease because of joint compression in this
position.
Buttocks involvement can be associated with hip disease, piriformis myofascial
pain (often sharp or burning), or spinal stenosis and requires contextual
evaluation.
Groin pain can be associated with intrinsic hip disease, local myofascial
pathology, sacroiliac joint syndrome, or an insufficiency fracture.
Posterior radiation is consistent with sciatica (sharp) or spinal stenosis (dull).
Lateral thigh radiation suggests tensor fascia lata/iliotibial band pain (not past
the knee) or gluteus minimus (past the knee “pseudo-sciatica”) myofascial pain.
Lateral leg pain with paresthesias or numbness suggests L5 radiculopathy.
Anterior thigh pain suggests hip disease, meralgia paresthetica, quadriceps
strain with knee osteoarthritis, or L2/3/4 radiculopathy.
Worsening with walking suggests spinal stenosis or vasogenic claudication.
Improvement with walking suggests myofascial pathology or neuropathic pain.
Prolonged walking may worsen myofascial pain. Degenerative disease may be
associated with initial pain/stiffness, then improvement and worsening with
excessive use.
Patients with fibromyalgia syndrome often have prominent axial pain, and may
present with a chief complaint of severe low back pain, but in fact low back pain
is just one of many sites of pain.
2. Does the pain get better or worse when
you curl up in bed?
3. Does the pain go into your buttocks?
If “yes”: Is the pain sharp or dull?
4. Do you have pain in your groin?
5. Does the pain shoot down your leg(s)?
If “yes”: In what part of your leg do you
feel the pain? Is the pain sharp or dull?
6. Is the pain made better or worse with
walking?
7. Do you sometimes feel that you have pain
all over?
Reprinted from: Weiner DK, Cayea D. Low back pain and its contributors in older adults: a practical approach to evaluation and treatment. In: Pain
in Older Persons, Progress in Pain Research and Management, Vol. 35. Seattle, Washington: IASP Press; 2005:332, with permission from Debra Weiner and
IASP Press.
310
Table 3
The American Journal of Medicine, Vol 120, No 4, April 2007
Physical Examination Abnormalities in Older Adults with Low Back Pain
Finding
Operational Definition
Examination Technique
Fibromyalgia tender
points
Presence of pain when
approximately 4 kg
of force is applied
to defined tender
points.
Functional leg
length
discrepancy
Pelvic asymmetry
Scoliosis (lateral/
rotational)
Lateral/rotational
curvature of
thoracolumbar spine
Sacroiliac joint
pain
Pain with direct
palpation of
sacroiliac joint or
with Patrick’s test
Myofascial pain,
piriformis
Presence of pain on
deep palpation of
piriformis.
Myofascial pain,
tensor fascia lata
(TFL) ⫾ iliotibial
(IT) band pain
Kyphosis
Presence of pain on
deep palpation of
TFL or IT band.
Have patient sit comfortably on examination table, arms resting in lap.
Tell patient that you are going to apply pressure at several points on the body, and that
you want to know if pressure on any point causes pain.
Examine the following points bilaterally, using enough pressure to blanch thumb nail:
(1) Occiput at suboccipital muscle insertions
(2) Low cervical at the anterior aspects of the intertransverse spaces at C5-C7
(3) Trapezius, midpoint of upper border
(4) Supraspinatus at origins, above the scapular spine near the medial border
(5) 2nd rib at the 2nd costochondral junction, just lateral to the junction on the upper
surfaces
(6) Lateral epicondyle 2 cm distal to the epicondyle
(7) Medial fat pad of the knee, proximal to joint line
(8) Greater trochanter, just posterior to the trochanteric prominence
(9) Gluteal at upper outer quadrant of buttocks in anterior fold of muscle
Have patient stand with both feet on floor, shoes removed. Ask him to stand with feet
together, and as erect as possible.
Kneel behind patient. With palms parallel to floor, and fingers extended, place lateral
surface of index finger of both hands atop pelvic brim bilaterally. Eyes should be level
with hands.
Determine if right and left thumbs are at different heights.
Have patient stand on floor with shoes removed. Stand behind patient. Run index finger
along spinous processes (do not lift hand between vertebrae) a series of 3 times. If you
do not detect scoliosis, then:
Ask patient to bend forward. Determine if there is asymmetry in height of paraspinal
musculature.
Direct Palpation: Have patient stand on floor with shoes removed. Stand behind patient.
Exert firm pressure over sacroiliac joint, first on one side, then the other. Palpate right
joint with right thumb, standing to left side of patient; palpate left joint with left
thumb, standing to right of patient.
Patrick’s (FABER) Test—Have the patient lie supine on the examining table and place the
foot of involved side on opposite knee. Then slowly lower the test leg in abduction
toward the examining table. If patient reports pain in back (not groin, buttocks or leg),
then test is positive.
Have patient lay supine on examination table.
Have patient flex right hip and knee, keeping sole of foot on table.
Cross bent leg over opposite leg; again place sole on table and exert mild medially
directed pressure on lateral aspect of knee to put piriformis in stretch.
Exert firm pressure (4 kg) over middle extent of piriformis.
Repeat examination on opposite side.
Have patient lying supine on examination table. Using thumbs of both hands, exert firm
pressure (4 kg) over full extent of TFL and IT band.
Repeat examination on opposite side.
Myofascial pain of
paralumbar
musculature
Vertebral body pain
Hip disease
Deformity of thoracic
spine creating
forward flexed
posture
Presence of pain on
deep palpation of
paralumbar
musculature.
Presence of pain on
firm palpation of
lumbar spinous
processes
Pain and restricted
motion of hip
Have patient stand on floor with shoes removed. Ask him to stand fully erect.
Inspect posture from the side.
Have patient stand on floor with shoes removed. Stand behind and to left of patient and
brace patient in front with left arm. Palpate full extent of right paravertebral musculature
with right thumb. Exert approximately 4 kg force. Repeat, palpating the left paravertebral
musculature.
Position yourself behind patient, as for examination of paravertebral musculature above.
Using dominant thumb, firmly palpate spinous processes L1-L5.
Hip Internal Rotation—Have patient lie supine on examining table with hip and knee bent
to 90°. Put the hip into maximum internal rotation (should be ⬎15°) and ask patient if he
experiences pain.
Patrick’s Test—As above, but pain may be in groin or leg.
Modified and reprinted from: Weiner DK, Sakamoto S, Perera S, Breuer P. Chronic low back pain in older adults: prevalence, reliability, and validity
of physical examination findings. J Am Geriatr Soc. 2006;54:11-20; with permission from Debra Weiner and Blackwell Publishing.
Weiner
Chronic Pain in the Elderly
311
Surgery
Strong Opioids
Other NSAIDs, Weak Opioids
Non-acetylated salicylates
Topical Preparations,
Cognitive Behavioral Therapy,
Interdisciplinary Pain Therapy,
CAM Modalities
Acetaminophen
Injections
Education, Exercise, Weight Loss, Assistive Devices
Figure 2 Stepped care approach to the treatment of axial and
appendicular osteoarthritis.
in those without pain.7 In part for this reason, failed back
surgery syndrome occurs in up to 40% of patients when
treatment decisions are based on the results of advanced
imaging alone. Factors that increase the risk of failed back
surgery syndrome include scoliosis, hip disease, and osteoporotic compression fractures, all of which are common in
older adults.8,9 Because these factors can alter gait or spinal
biomechanics, their close relationship with low back pain is
not surprising.
Essential elements of the history and physical examination for the older adult with CLBP are shown in Tables 2
and 3. Because low back pain is a syndrome, typically more
than one abnormality will be uncovered.
Treatment. Treatment of appendicular OA and CLBP
should be directed toward each of its individual contributors. In addition to treating the physical causes, careful
attention should be paid to the many psychosocial factors
that can impact pain and function, such as anxiety, depression, and fear. A stepped care approach to the treatment of
appendicular OA and CLBP is shown in Figure 2. Note on
the right side of Figure 2, “topical preparations, cognitivebehavioral therapy, interdisciplinary pain treatment, and
complementary and alternative modalities (CAM).” These
treatments should be considered alone or in combination
with any of the individual step components described below. Discussion of CAM is beyond the scope of this article.
Interested readers are referred to a recent review of CAM
for chronic pain in older adults.10
Step 1 includes patient education, exercise,11 weight loss,
and prescription of assistive devices (eg, canes, walkers).
Step 2 focuses on injection therapies. Even though injections are invasive, the risk of these procedures when performed by experienced clinicians is less than that of systemic analgesics. Injections should be considered for the
older adult with pain in 1 or 2 joints, eg, bilateral knee
osteoarthritis. The value of corticosteroid injections for the
patient with CLBP is less clear. The strongest evidence for
effectiveness is in the setting of a herniated disk associated
with radiculopathy. A critical review of minimally invasive
procedures for the patient with low back pain is provided
elsewhere.12 In general, injection therapies should be
viewed as a tool to enhance compliance with rehabilitation
efforts, which represent the mainstay of CLBP and generalized OA treatment.
Steps 3 through 6 include a range of pharmacologic
options. If regularly scheduled acetaminophen fails, nonacetylated salicylates (eg, salsalate, choline magnesium trisalicylate) should be considered because of their superior
side-effect profile as compared with traditional nonsteroidal
anti-inflammatory drugs (NSAIDs). Medications in Step 5
(other NSAIDs and weak opioids) include propoxyphene,
codeine, hydrocodone, and tramadol. Propoxyphene is generally contraindicated in older adults because it is no more
effective than placebo and retains many of the toxicities of
more potent and effective opioids. Codeine is an effective
antitussive but a weak analgesic. Before prescribing either
weak or strong opioids, patients must be counseled about
their potential side effects, and those that are preventable
should be addressed. For example, opioids increase the risk
of falls; therefore, physical therapy to enhance the patient’s
mobility and stability is recommended before opioids are
prescribed. In addition, a stimulant laxative (eg, senna)
should be taken at the first sign of constipation. Detailed
recommendations regarding dosing of nonopioid and opioid
analgesics, precautions, and potential drug-drug interactions
have been published by the American Geriatrics Society.13
Step 7 includes surgical procedures. In general, total joint
replacement is considered after noninvasive or minimally
invasive strategies have failed to control pain. Guidelines
regarding when to pursue surgical treatment for refractory
low back pain are less clear. Back surgery is elective in the
vast majority of cases and rates of failed back surgery
syndrome are substantial, with estimates ranging from 5%
Table 4 Symptoms and Disorders that Commonly Coexist in
Patients with Fibromyalgia Syndrome
Postexertional pain
Restless legs
Sleep apnea
Anxiety and depression
Dysesthesias and paresthesias
Impaired memory and concentration
Vague auditory, vestibular, and ocular symptoms
Intolerance or allergies to multiple medications
Palpitations
Dyspnea
Regional pain syndromes
● Tension and migraine headaches
● Atypical chest pain
● Temporomandibular symptoms
● Myofascial pain
● Pelvic pain
● Dyspareunia
Irritable bowel syndrome
Cold intolerance
Urinary frequency with interstitial cystitis
Subjective joint swelling
Fluid retention
Unexplained bruising
312
Table 5
First-line Medications for Neuropathic Pain
Key Drug-Disease
Interactions
Medication
Pharmacokinetics
Key Drug-Drug Interactions
Tricyclic
Antidepressants
(TCAs)
Nortriptyline
Desipramine
(Amitriptyline)
Hepatic metabolism;
higher levels of
active metabolites
in the elderly
Antipsychotics, anticholinergics,
selective serotonin reuptake
inhibitors (SSRI), sedativehypnotics, antiarrhythmics,
MAO inhibitors, clonidine,
antiretrovirals
Anticonvulsants
Gabapentin
Opioids
Opioids and
opioid-like
drugs
Oxycodone
Morphine
Tramadol
Anticholinergics, sedativehypnotics, anxiolytics, CYP2D6
inhibitors, SSRIs, TCAs,
muscle relaxants
Ileus, chronic
obstructive pulmonary
disease, dementia,
prostatic hypertrophy
Topical
anesthetics
Topical lidocaine
patch 5%
Renal elimination;
prolonged halflife with renal
impairment
Hepatic metabolism
and renal
elimination;
plasma levels may
be higher in the
elderly
Very little systemic
absorption
Myocardial infarction, QT
prolongation, AV
block, bundle branch
block, ileus, prostatic
hypertrophy,
glaucoma, seizure
disorder, dementia
Dementia, ataxia
Class I antiarrhythmics
Important Adverse Effects
Arrhythmia, cardiac conduction block,
orthostatic hypotension, urinary
retention, constipation, cognitive
impairment; adverse withdrawal
events after abrupt
discontinuation; death if torsades
de pointes
Somnolence, dizziness, peripheral
edema, increased appetite and
weight gain, adverse withdrawal
events after abrupt discontinuation
Constipation, sedation,
nausea/vomiting, respiratory
depression, nervous system
symptoms, pruritis, adverse
withdrawal events after abrupt
discontinuation
Skin rash
Reprinted from: Schmader KE, Dworkin RH. Clinical features and treatment of postherpetic neuralgia and peripheral neuropathy in older adults. In: Gibson SJ, Weiner DK, eds. Pain in Older Persons, Progress
in Pain Research and Management, Vol. 35. Seattle, Washington: IASP Press; 2005:363, with permission from IASP Press.
The American Journal of Medicine, Vol 120, No 4, April 2007
Medication Class
Weiner
Chronic Pain in the Elderly
Table 6
313
Dosage of First-Line Medications for Neuropathic Pain*
Duration of Adequate
Trial
Medication
Beginning Dosage
Titration
Maximum Dosage
Gabapentin
100-300 mg qhs
Lidocaine patch
5%
Maximum of 3
patches daily for a
maximum of 12
hours
2.5-15 mg every 4
hours as needed
Increase by 100-300 mg
a day every 1-7 days as
tolerated in divided
doses up to tid
None needed
3600 mg daily (1200 mg
tid); reduce if creatinine
clearance less than 60
mL/min
Maximum of 3 patches
daily for a maximum of
12 hours
3-8 weeks for titration
plus 1-2 weeks at
maximum tolerated
dosage
2 weeks
After 1-2 weeks, convert
total daily dosage to
long-acting opioid
analgesic and continue
short-acting medication
as needed
Increase by 25-50 mg
daily in divided doses
every 3-7 days as
tolerated
Increase by 10-25 mg
daily every 3-7 days as
tolerated
No maximum with careful
titration; consider
evaluation by pain
specialist at dosages
exceeding 120-180 mg
daily
400 mg daily (100 mg
qid); in patients over 75
years of age, 300 mg
daily in divided doses
75-150 mg daily; if blood
level of active drug and
its metabolite is below
100 ng/ml, continue
titration with caution
4-6 weeks
Opioid
analgesics
(dosages
given are for
morphine)
Tramadol
25 mg once daily
Tricyclic
antidepressants,
especially
nortriptyline
or
desipramine
10-25 mg qhs
4 weeks
6-8 weeks with at least
1-2 weeks at maximum
tolerated dosage
Reprinted from: Schmader KE, Dworkin RH. Clinical features and treatment of postherpetic neuralgia and peripheral neuropathy in older adults.
In: Gibson SJ, Weiner DK, eds. Pain in Older Persons, Progress in Pain Research and Management, Vol. 35. Seattle, Washington: IASP Press; 2005:365, with
permission from IASP Press. Originally adapted with permission from: Dworkin RH, Allen RR, Argoff CR, Backonja M, et al. Arch Neurol. 2003;60:1524-1534.
to 40%. Thus, back surgery should be considered for clearcut cases, such as deterioration of neurological function or
a herniated disc with anatomical location correlating precisely with symptom location and failure to respond to
nonsurgical management. Otherwise, back surgery should
be considered cautiously. The value of minimally invasive
surgical procedures for the treatment of low back pain is
currently under investigation.
Fibromyalgia Syndrome
Definition and Pathogenesis. Fibromyalgia syndrome
(FMS) affects 7% of women between the ages of 60 and 79
years.14 FMS is characterized by generalized pain and characteristic tender points on physical examination (Table 3).
Most patients with FMS also suffer from morning stiffness,
fatigue, and nonrestorative sleep. Once thought of as an
illness based in psychopathology, it is now recognized that
FMS is associated with dysregulation of the central nervous
system. That is, the nervous system of the FMS patient does
not process pain normally.
Evaluation. A careful history and physical examination is
the key to FMS diagnosis. It is not a diagnosis of exclusion.
A number of other disorders may coexist in patients with
FMS, such as migraine headaches, depression, and sleep
apnea. Table 4 lists these disorders as well as other symptoms from which patients with FMS suffer. Often, patients
report pain for decades, even during childhood (eg, “growing
pains”). There is increasing evidence that FMS is a heritable
disorder, thus practitioners should obtain a thorough family
history in the older adult with suspected FMS.
A combination of characteristic symptoms along with
palpation-elicited tenderness supports a diagnosis of FMS.
The presence of ⱖ11 of 18 characteristic tender points has
been used for classification purposes (ie, distinguishing patients with FMS from those with other rheumatologic disorders), but this exact number is not required to make a
diagnosis. Typically, patients with FMS have widespread
tenderness (ie, pain elicited by palpation when the examiner
uses enough pressure to blanch the thumbnail bed).
Treatment. Aerobic exercise is an important component of
treatment. The symptoms that most interfere with patients’
quality of life include pain, fatigue, limited activity tolerance, and sleep disturbance. A variety of medications, such
as tricyclic antidepressants, selective serotonin reuptake inhibitors, and cyclobenzaprine, may be effective in targeting
these symptoms. For detailed recommendations on the treatment of fibromyalgia, the reader is referred to guidelines
recently published by the American Pain Society.15 Patients
314
with refractory symptoms should be referred for interdisciplinary treatment that may include cognitive-behavioral
therapy, instruction in flare self-management techniques
and how to pace engagement in aerobic exercise, and methods to enhance sleep hygiene. Recently pregabalin also has
been approved.
Peripheral Neuropathy
Definition and Pathogenesis. Neuropathic pain is defined
as pain “initiated or caused by a primary lesion or dysfunction in the nervous system.”16 The most common causes in
older adults are diabetes mellitus and reactivation of herpes
zoster (ie, postherpetic neuralgia). Axial arthritis (eg, cervical and lumbar spondylosis) associated with radiculopathy
is also considered by many experts to be a form of neuropathic pain. Other causes of peripheral neuropathy in older
adults include alcoholic polyneuropathy, chemotherapy-induced polyneuropathy, entrapment neuropathies, postmastectomy pain, post-thoracotomy pain, nerve compression or
infiltration by tumor, phantom limb pain, postradiation plexopathy, and trigeminal neuralgia.17 Central poststroke pain
may also present with symptoms that mimic peripheral
neuropathy.
Evaluation. Although several specialized tests are available
for the assessment of neuropathic pain, such as quantitative
sensory testing and electromyography/nerve conduction, the
history and physical examination remain the cornerstone of
the evaluation. Patients with peripheral neuropathy often
complain of constant or intermittent pain (typically worse at
night and with inactivity), stimulus evoked-pain such as
Table 7 Guiding Principles for Caregivers of Older Adults
with Chronic Pain and Comorbid Dementia
● Chronic nonmalignant pain is not in and of itself harmful,
and should not preclude performing routine activities.
● Try to use distraction as much as possible to lessen
patient’s focus on pain.
● If anxiety is a prominent component of pain behavior, use
gentle touch (eg, rubbing the person’s back) in an effort
to calm the patient.
● Ask the patient’s primary provider or physical therapist
about available nonpharmacologic approaches to pain
management (eg, gentle massage, balms and ointments,
relaxation techniques). Systemic pharmacologic agents
fraught with potential toxicities should be avoided as
much as possible, although judicious and carefully
monitored use of these agents should be considered on a
case by case basis.
● The caregiver and patient should try performing relaxation
exercises together to diminish stress of both parties.
● The caregiver should try to be supportive and reassuring
when the patient voices pain complaints. These complaints
should not be brushed off, but attended to in a positive
way.
The American Journal of Medicine, Vol 120, No 4, April 2007
Table 8 Treatment Outcome Parameters for the Older Adult
with Chronic Pain
Pain interference with performance of:
● Basic activities of daily living
● Instrumental activities of daily living
● Discretionary activities
Mobility/activity level
Energy level
Appetite
Sleep
Mood – eg, irritability, depression
Interpersonal interactions
Attention and concentration
Frequency of prn analgesic ingestion
Pain severity
allodynia or hyperalgesia, pruritis or tingling, and sensory
loss. A comprehensive neurological examination should be
performed on all older adults with neuropathic pain. When
the cause is not obvious (eg, postmastectomy pain, postthoracotomy pain), a thorough general physical examination
should be performed and any signs of malignancy should be
pursued diagnostically. If no cause can be found, the following laboratory studies should be obtained: vitamin B12
and folate levels, fasting blood glucose, and serum/urine
electrophoresis. If a cause still cannot be identified, the
patient should be referred to a neurologist.
Treatment. Neuropathic pain is typically more refractory to
treatment efforts than nociceptive pain. A number of pharmacologic options are available. Tables 5 and 6 summarize
medications and dosing guidelines for the treatment of neuropathic pain.
THE ROLE OF DEPRESSION AND DEMENTIA IN
PAIN MANAGEMENT
In 1965, the Gate Control Theory of Pain was born, and its
principles continue to guide pain management practice today. This theory emphasizes the central role of the brain in
pain processing and accounts for the effectiveness of cognitive behavioral therapy in management of chronic pain.
When the brain functions abnormally, as in the case of
depression or dementia, pain control becomes more challenging. Thus, practitioners that care for older adults with
chronic pain must be as aggressive about recognizing and
treating these disorders as they are about the pain itself.
While depression is eminently treatable, the presence of
dementia in the older adult with chronic pain poses a considerable challenge. Often, those with dementia have exaggerated fear responses to pain which, in turn, can intensify
the pain experience. The pharmacologic and nonpharmacologic modalities that have already been discussed should
also be offered to the older adult with dementia. Education
should be emphasized, especially education of caregivers.
Components of care-giver education are provided in Table 7.
Weiner
Chronic Pain in the Elderly
HOW TO DETERMINE WHETHER PAIN
TREATMENT IS EFFECTIVE
Pain intensity is only one of many parameters that may
improve when chronic pain is managed successfully. Patients with chronic nonmalignant pain should anticipate, on
average, no more than 30%-50% reduction in pain intensity.
Physical function, mood, endurance, sleep, appetite, and
interpersonal interactions may respond even more substantially to treatment, as may overall quality of life. As part of
the initial assessment, therefore, the practitioner should
clearly identify treatment goals in order to determine the
effectiveness of therapy. Although treatment goals must be
individualized, a list of potential outcome parameters is
provided in Table 8.
HOW AND WHEN TO REFER TO A
PAIN SPECIALIST
The decision regarding when and to whom to refer the older
adult with refractory pain depends upon the patient’s goals.
In general, referral to a pain specialist should be considered
when the patient continues to experience disabling pain
despite efforts to control symptoms with medications and
other therapeutic modalities (eg, injections, physical therapy). Pain specialists also vary widely in their training and
approach to treatment. Some specialists focus exclusively
on injection procedures, while others work with an interdisciplinary team that utilizes a rehabilitative approach. Because of the multifactorial nature of chronic pain syndromes
in older adults, the latter strategy is favored and appears to
be most effective.
References
1. Simons DG, Travell JG, Simons LS, eds. Travell & Simons’ Myofascial Pain and Dysfunction—The Trigger Point Manual, Volume I.
Upper Half of Body, Volume II. Lower Half of Body. Second Ed.
Baltimore: Williams & Wilkins; 1999.
2. Gunn CC. Neuropathic myofascial pain syndromes. In: Loeser JD,
Butler SH, Chapman CR, et al. Bonica’s Management of Pain. Third
Ed. Philadelphia: Lippincott Williams & Wilkins; 2001.
315
3. Gerwin RD. Classification, epidemiology, and natural history of myofascial pain syndrome. Curr Pain Headache Rep. 2001;5:412-20.
4. Borg-Stein J. Treatment of fibromyalgia, myofascial pain, and related
disorders. Phys Med Rehab Clin North Am. 2006;17:491-510.
5. Hong CZ. Lidocaine injection versus dry needling to myofascial trigger point: the importance of the local twitch response. Am J Phys Med
Rehab. 1994;73:256-63.
6. Jarvik JG, Deyo RA. Diagnostic evaluation of low back pain with
emphasis on imaging. Ann Intern Med. 2002;137:586-97.
7. Jarvik JJ, Hollingworth W, Heagerty P, Haynor DR, Deyo RA. The
longitudinal assessment of imaging and disability of the back
(LAIDBack) study: baseline data. Spine. 2001;26:1158-66.
8. Airaksinen O, Herno A, Turunen V, Saari T, Suomlainen O. Surgical
outcome of 438 patient treated surgically for lumbar spinal stenosis.
Spine. 1997;22:2278-82.
9. Frazier DD, Lipson SJ, Fossel AH, Katz JN. Associations between
spinal deformity and outcomes after decompression for spinal stenosis.
Spine. 1997;22:2025-9.
10. Prestwood K. Complementary and alternative medicine approaches to
pain in older persons. In: Gibson SJ, Weiner DK, eds. Pain in Older
Persons, Progress in Pain Research and Management. Seattle: IASP
Press; 2005.
11. American Geriatrics Society Panel on Exercise and Osteoarthritis.
Exercise prescription for older adults with osteoarthritis pain: Consensus practice recommendations. J Am Geriatr Soc. 2001;49:808-23.
12. Bernstein CD, Lateef B, Fine PG. Interventional pain management
procedures in older patients. In: Gibson SJ, Weiner DK, eds. Pain in
Older Persons, Progress in Pain Research and Management. Seattle:
IASP Press; 2005:263-83.
13. AGS Panel on Persistent Pain in Older Persons. The Management of
Persistent Pain in Older Persons. American Geriatrics Society. J Am
Geriatr Soc. 2002;50:S205-S224.
14. Wolfe F, Ross K, Anderson J, Russell IJ, Hebert L. The prevalence and
characteristics of fibromyalgia in the general population. Arthritis
Rheum. 1995;38:19-28.
15. Burckhardt C, Goldenberg DL, Crofford L, et al. Guideline for the
Management of Fibromyalgia Syndrome Pain in Adults and Children—APS Clinical Practice Guidelines Series. Fourth Ed. Glenview,
Illinois: American Pain Society; 2005.
16. Turk DC, Okifuji A. Pain terms and taxonomies of pain. In: Loeser JD,
ed. Bonica’s Management of Pain. Third Ed. Philadelphia: Lippincott
Williams & Wilkins; 2001:17-25.
17. Schmader KE, Dworkin RH. Clinical features and treatment of postherpetic neuralgia and peripheral neuropathy in older adults. In: Gibson SJ, Weiner DK, eds. Pain in Older Persons, Progress in Pain
Research and Management. Seattle: IASP Press; 2005:355-75.
The American Journal of Medicine (2007) 120, 316-318
DIAGNOSTIC DILEMMA
Charles M. Wiener, MD, Section Editor
Too Much of a Good Thing
Lisa Call Pastel, MD, Jonathan M. Ross, MD, FACP
Dartmouth-Hitchcock Medical Center, Lebanon, NH.
PRESENTATION
The increasing use of supplements in the US challenges
physicians to understand the effects of and quantify the
usage of these readily available medications. Unfortunately,
all too often physicians don’t inquire about supplement
usage, and patients don’t volunteer the information.
Our patient was a 46-year-old man who presented with
neutropenia and anemia following multiple surgical complications that began after a gastric bypass 4 years earlier
(Figure 1). He developed an incisional hernia, a small bowel
obstruction, and an abdominal abscess and had difficulties
with wound healing. His most recent surgery— 6 months
earlier—left him with an open abdominal wound that was
allowed to heal by secondary intention.
After his gastric bypass he had lost over 100 pounds with
near resolution of diabetes, hypertension, and obstructive
sleep apnea. He was wheelchair bound and on pain medications due to chronic back and knee pain. He was previously employed as a journalist until he became disabled
from chronic pain. Despite his medical problems, he homeschooled his daughters full time.
At presentation, he denied past or present infections,
fevers, fatigue, or dyspnea. The patient was taking multiple,
potentially conflicting medications and supplements. His
medications included: methadone, hydrocodone, diazepam,
methocarbamol, acetaminophen, cephalexin, aspirin, furosemide, hydrochlorothiazide, metoprolol, a daily multivitamin, vitamin B-12 supplementation, calcium citrate, vitamin A, zinc, and vitamin C. He had no known allergies. He
was a past smoker, having quit 11 years ago, but he denied
past or present alcohol use. The family history was notable
for coronary artery disease.
On examination the patient appeared well. He was afebrile with normal vital signs. He was anicteric and had no
lymphadenopathy; cardiopulmonary exam was normal. His
abdomen had an open midline abdominal wound without
Requests for reprints should be addressed to Lisa Call Pastel, MD,
Department of Internal Medicine, Dartmout-Hitchcock Medical Center,
One Medical Center Drive, Lebanon, NH 03756.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.11.024
apparent infection. There was no palpable hepatosplenomegaly or peripheral edema.
His laboratory tests were significant for a low white
blood count of 1.9 ⫻ 103/␮L with an absolute neutrophil
count of 500 and hemoglobin of 9.0 g/dL. His smear
showed anisocytosis, poikilocytosis, occasional target cells,
ovalocytes, and teardrop cells. Iron studies were consistent
with anemia of chronic disease. Thyroid-stimulating hormone was slightly elevated; however, the remainder of the
thyroid panel was normal. ANA, HIV, c-ANCA, p-ANCA,
and hepatitis serologies were all normal (Table 1).
ASSESSMENT
The assessment was medication-related neutropenia. Hydrochlorothiazide, cephalexin, metoprolol, aspirin, and acetaminophen were discontinued without improvement in his
neutropenia; iron therapy was initiated. Bone marrow biopsy demonstrated normocellular marrow with myeloid hypoplasia and adequate iron stores. In anticipation of further
abdominal surgery, the patient was given granulocyte colony stimulating factor, which had a modest response. The
workup outlined above continued over the next year.
DIAGNOSIS
Eight months later, the patient was subsequently admitted to
the hospital for neutropenic fever. At this time, his medications were thoroughly reviewed again. The medical team
discovered that for the past year the patient had been taking
50 mg of zinc 4 times a day to aid in wound healing. At his
previous presentation, he did not tell his outpatient physician the dosage of zinc that he was taking. Serum copper
and ceruloplasmin—the protein which carries copper—
were found to be extremely low at ⬍0.1 ug/mL (0.75-1.45
ug/mL normal range) and 1.3 mg/dL (22.9-43.1 mg/dL
normal range), respectively. These findings were consistent
with zinc-induced copper deficiency.
MANAGEMENT
Within 1 month, discontinuation of zinc and supplementation with copper (3 mg copper sulfate 3 times a day) in-
Pastel and Ross
Zinc Toxicity
Gastric
bypass
317
Hernia/obstruction
Abscess drained
Wound left open
Iron replacement
Presentation
Admit for
neutropenic fever
2001
7/2003
1/2004
5/2004
1/2005
WBC
Hct (%)
7.1 x 103
31.8
5.9 x 103
27.0
1.9 x 10 3
30.0
1.3 X 103
35.7
Figure 1 Timeline of presentation. This timeline demonstrates the patient’s development of neutropenia and anemia in the context of his
complicated surgical issues. Over a period of 6 months he became significantly neutropenic and anemic. WBC ⫽ white blood count. Hct ⫽
Hematocrit.
creased the patient’s white count to 4.1 ⫻ 103/␮L and
hemoglobin to 12.4 g/dL. Supplementation with copper was
planned for a total of 6 months. Copper and ceruloplasmin
levels were not repeated after supplementation.
Zinc toxicity causing copper deficiency, and resultant
neutropenia and anemia is rare, reported in less than 30
cases since 1977.1 Zinc excess occurs when doses exceed
50-100 mg/day for prolonged periods, estimated at 6 weeks
to several months.2 This has been observed in patients
taking zinc for treatment of acne,3 the common cold,4 and
wound healing.5 It also has occurred in psychiatric patients
who have consumed pennies (after 1982).6
The US Department of Agriculture recommends 9-11
mg of zinc per day for adults, with an upper limit of 40
mg/day.4 Intake of zinc lozenges should be limited to less
than 50 mg/day.1 At our institution, doses for wound healing are usually 50 mg 3 times a day for 2 weeks only.7 This
patient had been taking slightly more than the recommended
dose (200 mg/day) for wound healing and for far too long.
There is limited evidence concerning zinc and wound healing; however, the research suggests that patients who are
zinc deficient have the best response to supplementation.4
The mechanism by which zinc toxicity causes copper
deficiency involves the metal-binding ligand metallothionein, which is found in enterocytes. It binds to both zinc and
copper but prefers copper. Excess zinc induces this ligand,
and bound copper is lost when enterocytes are shed.1 Figure
2 shows the structure of metallothionein, as determined by
x-ray crystallography. Zinc and copper bind at available
cysteines.8
It is unknown exactly how copper deficiency causes
neutropenia; however, some research suggests that it interferes with maturation of neutrophils from the promyelocyte/
myelocyte stage to more mature forms.10 It also has been
suggested that copper deficiency somehow marks the neutrophils as foreign, and they are subsequently destroyed.10
Anemia results when reduced levels of ceruloplasmin
oxidize ferrous (2⫹) to ferric (3⫹) iron for subsequent binding to transferrin. Additionally, the copper-dependent enzyme cytochrome-c oxidase further reduces ferric iron to be
used for heme synthesis. A disruption in these steps of heme
synthesis can result in sideroblastic anemia, which has been
reported in several cases of copper deficiency. In the setting
of copper deficiency, it also has been observed that there are
reduced levels of the enzyme superoxide dismutase that
Table 1
Results of Laboratory Tests
Normal
Range
At Presentation
White cell count (⫻103/␮L)
Hemoglobin (g/dL)
Hematocrit (%)
Mean corpuscular volume (FL)
Red Cell Distribution
Width (%)
Peripheral Smear
4.0-10.0
14.0-17.0
40-54
78.2-95.8
16.0
1.9
9.0
30.0
85.0
11.5-15.0
Sodium (mEq/L)
Potassium (mEq/L)
Chloride (mEq/L)
Carbon dioxide (mEq/L)
Urea nitrogen (mg/dL)
Creatinine (mg/dL)
Glucose (mg/dL)
Calcium (mg/dL)
Magnesium (mEq/dL)
Phosphorus (mg/dL)
Total bilirubin (mg/dL)
Direct bibirubin (mg/dL)
Aspartate Aminotransferase
(units/L)
Alanine Aminotransferase
(units/L)
Alkaline Phosphatase (units/L)
Albumin (g/dL)
Erythrocyte Sedimentation rate
(mm/h)
Thyroid-stimulating hormone
(mIU/L)
Iron (ug/dL)
Total Iron Binding Capacity
(ug/dL)
Transferrin Saturation (%)
Ferritin (ng/mL)
Erythropoietin (mU/mL)
c-ANCA
p-ANCA
HIV
Hepatitis A, B, and C
Anti-nuclear Antibody
135-145
3.5-5.0
98-107
22-31
10-20
0.8-1.5
70-110
8.5-10.5
1.3-2.0
2.5-4.5
0.2-1.3
0.0-0.3
15-50
Anisocytosis,
poikilocytosis,
occasional
target cells,
occasional
oval and
teardrop cells
143
3.5
103
30
16
0.5
93
9.0
1.44
4.5
1.4
0.1
22
Variable
10-56
17
40-120
3.2-5.2
0-15
75
4.0
50
0.27-4.20
45-160
250-450
20-50
30-400
4-24
4.26
9
171
5
754
30
Negative
Negative
Negative
Negative
Negative
318
The American Journal of Medicine, Vol 120, No 4, April 2007
Figure 2 Structure of Metallothionein. Copper and zinc bind at the available cysteine molecules. There is a binding preference for copper.
Szilayi and Z Fenselau C. Molecular Dynamics Simulation of Metallothionein-Drug Complexes. Drug Metabolism and
Disposition.2000;28(2):174-179.
maintains red blood cell membrane integrity. As a result
increased red blood cell turnover is observed with copper
deficiency.6
Treatment for copper deficiency consists of discontinuation of zinc therapy and supplementation with copper. This
results in normalized hematologic indices within months.1
This patient’s hematologic problems emphasize the importance of a complete medication review including specific
doses of supplementary medications. This is particularly
important with increasing use of over-the-counter medications and supplements, typically not reported by patients
and not quantified by physicians.
References
1. Irving JA, Mattman A, Lockitch G, et al. Element of caution: a case of
reversible cytopenias associated with excessive zinc supplementation.
CMAJ. 2003;169(2):129-31.
2. Fosmire GJ. Zinc toxicity. Am J Clin Nutr. 1990;51(2):225-7.
3. Igic PG, Lee E, Harper W, et al. Toxic effects associated with Consumption of Zinc. Mayo Clin Proc. 2002;77(7):713-6.
4. National Institutes of Health [homepage on the Internet]. Health Info –
Dietary Supplement Fact Sheets. Last updated Dec. 2002. Available
from: http://ods.od.nih.gov/factsheets/cc/zinc.html#rda.
5. Todd LM, Godber IM, Gunn IR. Iatrogenic copper deficiency causing
anaemia and neutropenia. Ann Clin Biochem. 2004;41(Pt 5):414-6.
6. Willis MS, Monaghan SA, Miller ML, et al. Zinc-induced copper
deficiency: a report of three cases initially recognized on bone marrow
examination. Am J Clin Pathol. 2005;123(1):125-31.
7. Nolette P. Dartmouth-Hitchcock Medical Center Wound Care Team.
Oct 2005.
8. Szilayi Z and Fenselau C. Molecular Dynamics Simulation of Metallothionein-Drug Complexes. Drug Metabolism and Disposition. 2000;
28(2):174-179.
9. Brewer GJ, Hill GM, Prasad AS, et al. Oral zinc therapy for Wilson’s
disease. Ann Intern Med. 1983;99(3):314-9.
10. Percival SS. Neutropenia caused by copper deficiency: possible mechanisms of action. Nutr Rev. 1995;053(3):59-66.
The American Journal of Medicine (2007) 120, 319-321
DIAGNOSTIC DILEMMA: INFECTIOUS DISEASE
Charles M. Wiener, MD, Section Editor
A Case for Vaccination
Aima Ahonkai, MD and Noah Lechtzin, MD, MHS
Osler Medical Service, Johns Hopkins Hospital, Baltimore, MD
PRESENTATION
Patients occasionally question the wisdom of immunization,
but the case of a 56-year-old woman offers sober testimony
to the potential value of pneumococcal vaccination. She
presented to the emergency department after 1 to 2 days of
nausea, vomiting, confusion, and incontinence. Family
members reported symptoms of a flu-like illness marked by
fever, malaise, and myalgias in the preceding week. On
admission, the patient had a fever of 103.1°F (39.5°) compounded by hypotension (95/54 mm Hg), tachycardia (160
beats per minute), hypoxia (oxygen saturation on room air
was 93%), and altered mental status. She had a history of
hypertension, and 25 years earlier, she had Hodgkin’s lymphoma, for which she had undergone splenectomy, chemotherapy, and radiation. The cancer was in remission, according to a post-staging laparotomy. Despite her history of
Hodgkin’s lymphoma and splenectomy, the patient had
never received pneumococcal vaccine.
ASSESSMENT
The patient was lethargic and confused but had no focal
neurologic deficits or meningeal signs. She had pharyngeal
erythema but no exudates. Her lungs were clear, and her
abdomen was soft. Weak pulses and poor skin turgor with
mottling of the skin was noted on both of her upper and
lower extremities. Her initial Acute Physiology and Chronic
Health Evaluation (APACHE) II score was 14, predicting a
49.7% risk of death. Laboratory testing was notable for a
white blood cell count of 7.6 ⫻ 103/cells/mm3 with 28%
bands and 58% polymorphonuclear leukocytes; acute renal
failure (creatinine, 1.6 mg/dL); coagulopathy (international
normalized ratio ⬎1.5); and mild lactic acidosis (serum
lactate, 3.6 mmol/L). Her initial chest radiograph showed no
infiltrates.
In the medical intensive care unit, the patient was aggressively resuscitated with fluids. Nonetheless, her hypotension and acidosis worsened (peak serum lactate was 10.6
Requests for reprints should be addressed to Charles M. Wiener, MD.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2007.02.011
mmol/L), and acute respiratory failure required treatment
with norepinephrine, endotracheal intubation, and mechanical ventilation. Stress-dose corticosteroids were administered, as was a 96-hour infusion of activated protein C.
Vancomycin, ceftriaxone, and doxycycline also were administered as empirical treatment for sepsis and meningitis.
Lumbar puncture was deferred, given the risk of bleeding
while on activated protein C.
Initial blood cultures grew penicillin-resistant Streptococcus pneumoniae in 2 of 2 cultures within 6 hours of
presentation; doxycycline was discontinued. Sputum and
urine cultures remained negative.
Within 24 hours of admission, the patient developed
purpuric skin lesions on her upper extremities, lower extremities, and nose, with patchy areas of ecchymoses and
bullae. She suffered multi-organ-system dysfunction with
acute renal failure (peak creatinine, 3.6 mg/dL), hepatitis
(peak aspartate aminotransferase, 1015 IU/L; peak alanine
aminotransferase, 594 IU/L), respiratory failure, and disseminated intravascular coagulopathy (D-dimer, 55 mg/L;
lactate dehydrogenase, 1186 U/L; international normalized
ratio, 2.1; activated partial thromboplastin time, 54 seconds;
and fibrinogen, 163 mg/dL). Her skin lesions became increasingly necrotic, and she eventually developed dry gangrene of her digits and feet (Figures 1-3).
DIAGNOSIS
Purpura fulminans secondary to streptococcal sepsis was
diagnosed. This syndrome, distinguished by disseminated
intravascular coagulation and purpuric skin lesions, is a rare
complication of overwhelming septic shock from severe
bacterial infection. Or, it can arise in conjunction with
inherited homozygous deficiencies of protein C or protein
S— unusual conditions— or, on occasion, autoimmune disease, including antiphospholipid antibody syndrome. Warfarin-induced skin necrosis and cholestasis have infrequently been associated with purpura fulminans, as well.
The striking skin lesions result from perivascular hemorrhage and necrosis triggered by dermal microvascular
thromboses. Deep-tissue skin necrosis is often extensive,
320
The American Journal of Medicine, Vol 120, No 4, April 2007
Figure 1 The patient had extensive gangrene of the distal phalanges and plantar surface of the foot.
causing gangrene of extremities. Amputation is a likely
outcome.1 Even with advances in critical care medicine,
mortality from infection-related purpura fulminans still approaches 50%.2
A 10-year retrospective chart review found that Neisseria meningitides was the most common etiologic agent in
purpura fulminans in infants and adolescents, while Streptococcus pneumoniae was the most common etiologic agent
among adults.3 Still, case reports and case series describing
meningococcal purpura fulminans in pediatric populations
appear much more commonly in the literature than do cases
generated by S. pneumonia infection. Infants and children
are likely more vulnerable because of congenital and acquired immunodeficiencies and coagulation defects.
MANAGEMENT
Several approaches have been proposed and studied for the
treatment of sepsis-related purpura fulminans, including
plasmapheresis, antiendotoxin therapy, anticytokine therapies, heparin, antithrombin III, thrombolysis, and fasciotomy, although none have been examined in a randomized fashion. Undoubtedly, the ideal management of the
condition includes early antibiotic use, aggressive fluid resuscitation, ventilatory support, inotropic support, tight glucose control, and steroid replacement as necessary.
Figure 2
digits.
Hemorrhagic bullae progressed to gangrene in the
Figure 3 Geographic cutaneous infarctions and bullae from microvascular thromboses are classic lesions in this condition.
Activated protein C may also have a role in limiting
morbidity. Protein C, which has anticoagulant, fibrinolytic,
and anti-inflammatory properties, seems to play a central
role in modulating the sepsis disease process, and it has
been shown to reduce mortality in severely septic patients.4
Though it is not clear which patients with severe sepsis are
at risk for developing purpura fulminans, the data support
greater peril for those infected with meningococci or Streptococcus species. In patients with meningococcal sepsis,
increased levels of plasminogen activator inhibitor-1 and
decreased levels of protein C are linked with the development of purpuric skin lesions and poor prognosis. In fact,
levels of protein C approximate those seen in homozygous
protein C deficiency.5-7
Most of the literature describing protein C replacement
in purpura fulminans applies to patients with meningococcal
321
sepsis. The data suggest that protein C replacement and
anticoagulation improve mortality and morbidity with regard to the extent of skin necrosis and the need for limb
amputation. One prospective study included 12 patients,
aged 3 months to 27 years, who were admitted to an intensive care unit with severe meningococcemia, septic shock,
purpura fulminans, disseminated intravascular coagulation,
and protein C levels below 3%. In addition to conventional
treatment, they each received continuous infusions of protein C concentrate and unfractionated heparin. Survival was
100%, and only 2 patients required amputations. One began
protein C 48 hours after admission, and the other began
treatment at 72 hours; the rest of the patients were given
protein C by the 15th hour of admission.8
In an open-label prospective study, protein C concentrate
was administered to 36 patients, aged 3 months to 72 years,
with severe meningococcemia, purpura fulminans, and multiorgan failure.9 Survival was 98%, and the overall amputation rate was 12%, compared with a predicted amputation
risk of 33%. The amputation rate among patients who were
treated with protein C concentrate sooner—within 24 hours
of admission—was 6.5%.
Since 1970, there have been 10 case reports of purpura
fulminans associated with S. pneumoniae. Most occurred in
asplenic patients, and there is little data on the efficacy of
activated protein C and anticoagulation on morbidity and
mortality in this group.10-13
Our patient is one of very few people reported to receive
prompt treatment with activated protein C for pneumococcal-related purpura fulminans. Despite aggressive management, including early treatment with protein C, our patient
developed extensive gangrene requiring amputation of all
digits on both hands and ultimately, bilateral foot
amputations.
It is possible that purpura fulminans in patients with
streptococcal infection may be modulated by additional or
unique factors that play more significant roles than activated
protein C. Pneumococci do not contain endotoxin, which is
central in the initiation of the coagulation cascade in purpura fulminans spurred by gram negative organisms. The
cell wall does, nevertheless, stimulate production of proinflammatory cytokines and activate complement through the
alternative pathway. It has been postulated that circulating
complement-activating immune complexes incite immunemediated destruction in pneumococcal disease, though direct vascular damage cannot be ruled out.14
This case underscores the importance of preventing severe bacterial infections with a vaccination program in asplenic patients.15 It also suggests that activated protein C
may not have the same role in meningococcal and pneumococcal purpura fulminans. Alternate treatments for purpura
fulminans in pneumococcal sepsis deserve further
evaluation.
References
1. Smith OP, White B. Infectious purpura fulminans: diagnosis and
treatment. Br J Haematol. 1999;104:202-207.
2. Giraud T, Dhainaut JF, Schremmer B, et al. Adult overwhelming
meningococcal purpura. A study of 35 cases, 1977-1989. Arch Intern
Med. 1991;151:310-316.
3. Warner PM, Kagan RJ, Yakuboff KP, et al. Current management of
purpura fulminans: a multicenter study. J Burn Care Rehabil. 2003;
24:119-126.
4. Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of
recombinant human activated protein C for severe sepsis. N Engl
J Med. 2001;344:699-709.
5. Hermans PW, Hibberd ML, Booy R, et al. 4G/5G promoter polymorphism in the plasminogen-activator-inhibitor-1 gene and outcome of
meningococcal disease. Meningococcal Research Group. Lancet.
1999;354:556-560.
6. Powars D, Larsen R, Johnson J, et al. Epidemic meningococcemia and
purpura fulminans with induced protein C deficiency. Clin Infect Dis.
1993;17:254-261.
7. Fijnvandraat K, Derkx B, Bijlmer R, et al. Coagulation activation and
tissue necrosis in meningococcal septic shock: severely reduced protein C levels predict a high mortality. Thromb Haemost. 1995;73:1520.
8. Smith OP, White B, Vaughan D, et al. Use of protein C concentrate,
heparin, and hemodiafiltration in meningococcus-induced purpura fulminans. Lancet. 1997;350:1590-1593.
9. White B, Livingstone W, Murphy C, et al. An open label study of the
role of adjuvant hemostatic support with protein C replacement therapy in purpura fulminans-associated meningococcemia. Blood. 2000;
96:3719-3724.
10. Noguera A, Fortuny C, Pons M, Cambra F, Palomeque A. Pneumococcal-associated purpura fulminans in a healthy infant. Pediatr
Emerg Care. 2004;20:528-530.
11. Shperber Y, Geller E, Rudick V, Orda R. Purpura fulminans associated
with overwhelming sepsis and disseminated intravascular coagulation
following splenectomy. Isr J Med Sci.1989;25:657-659.
12. Murphy CB, Noeller K. Purpura fulminans secondary to pneumococcal sepsis in an asplenic patient. J Am Podiatr Med Assoc. 1993;83:
43-46.
13. Hautekeete ML, Berneman ZN, Bieger R, et al. Purpura fulminans in
pneumococcal sepsis. Arch Intern Med. 1986;146:497-499.
14. Rintala E, Kauppila M, Seppala OP, et al. Protein C substitution in
sepsis-associated purpura fulminans. Crit Care Med. 2000;28:23732378.
15. Prevention of pneumococcal disease: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm
Rep. 1997 Apr 4;46(RR-8):1-24.
The American Journal of Medicine (2007) 120, 322-324
IMAGES IN DERMATOLOGY
Parwathi “Uma” Paniker, MD, Section Editor
Mystery of the Silk Road
John Patrick Welsh, MD,a Christopher B. Skvarka, MD,a Christine Ko, MD,b Carrie Ann Cusack, MDa
a
Drexel University College of Medicine, Department of Dermatology, Philadelphia, Pa; bYale University School of Medicine,
Department of Dermatology, New Haven, Conn.
PRESENTATION
Patients often present to the dermatologist’s office with skin
conditions that may be indicative of multiple diseases.
When there is no specific test for a particular disease,
diagnosis may be difficult. Our patient was a 34-year-old
white man with a medical history significant for recurrent
aphthous stomatitis. He presented with fever (101.5°F); oral
and genital ulcerations; arthralgias; and painful lower extremity rash for the past 2 weeks. He denied eye pain or
vision changes, known drug allergies, or contact with others
who had been ill. He also reported that his father demonstrated similar symptomatology for years without diagnosis.
His recent medications included only acyclovir and levofloxacin dispensed by the emergency department for his
worsening symptoms.
ASSESSMENT
On examination, the patient exhibited scattered 0.1- to
0.3-cm herpetiform ulcerations, each with a surrounding red
border on the soft palate, labial/gingival mucosae, and penis
(shaft, glans, and coronal sulcus) (Figures 1 and 2). He also
manifested multiple tender, blanchable, red-pink, 2.0- to
4.0-cm noduloplaques on the extensor surface of his bilateral lower extremities (dorsal feet to dorsal thigh) and a few
small follicular pustules on the back and chest (Figure 3).
His bilateral ankle joints and right wrist joint exhibited
noticeable edema and decreased active/passive range of
motion.
A punch biopsy from 1 of the tender leg nodules showed
a primarily lobular panniculitis with associated deep mid-tolarge artery vasculitis. Gastroenterology survey showed
multiple small ulcers at the antrum, ileocecal valve, and
ascending/descending colon. Further test results were negative for the following serologic markers in the patient:
human immunodeficiency virus, herpes simplex virus, hepatitis B and C, rapid plasmin reagin, antinuclear antibody,
rheumatoid factor, double-stranded deoxyribonucleic acid,
perinuclear antineutrophil cytoplasmic antibody, central antineutrophil cytoplasmic antibody, and human leukocyte
antigen-B*51. After treatment with methylprednisolone and
indomethacin, our patient’s symptoms dramatically improved. He was subsequently maintained on azathioprine.
DIAGNOSIS
Behçet’s disease was first diagnosed 2500 years ago by the
Greek physician Hippocrates of Kos,1 but it did not receive
its official clinical designation until 1937. Turkish dermatologist Hulusi Behçet identified the triad of orogenital
aphthosis, iritis, and hypopyon as emblematic of a uniform
condition.2 Although the disease is found globally, the majority of cases concentrate along the ancient Silk Road
(eastern Asia to the Mediterranean Sea between latitudes 30
and 45 degrees North). Prevalence ranges from 80 to 370
cases per 100,000 in Turkey to 0.12 to 0.33 per 100,000 in
the United States.3 Behçet’s disease often manifests itself in
the third or fourth decade of life.4 There is a variable sexual
difference with female predominance in Japan/Korea and
male predominance in the Middle East.3 Familial clustering
also has been documented.5
An incurable condition, Behçet’s disease can be difficult
to diagnose because there is no specific test; often physicians may have to rule out several other conditions before
settling on a diagnosis.1 Behçet’s disease is marked by both
mucocutaneous and systemic involvement. In 1990, the
International Study Group reached consensus regarding currently accepted diagnostic criteria, which include the following:
●
Requests for reprints should be addressed to John Patrick Welsh, MD,
Drexel University, College of Medicine, Department of Dermatology, 219 N.
Broad Street, 4th Floor, Philadelphia, PA 19107.
E-mail address: [email protected].
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.11.007
●
Recurrent oral aphthae (minor, major, or herpetiform)
with at least 3 flares in any 12-month period.
Two of the following: recurrent genital aphthae/scarring,
anterior/posterior uveitis, cutaneous lesions, or a positive
pathergy test result.6
Welsh et al
Behcet’s Disease
Figure 1
323
Oral ulcers.
Although oral/genital aphthae remain the seminal cutaneous facet of Behçet’s, other common manifestations include erythema nodosum-, Sweet’s-, and pyoderma gangrenosum-like lesions; pseudofolliculitis/acne; and palpable
purpura. Often the condition may be complicated by a
cutaneous vasculitis, predominantly venulitis or thrombophlebitis, with relative sparing of the arterial component.7
On occasion, a necrotizing panarteritis involving small- and
medium-sized arteries at the dermal-subcutaneous junction
simulating cutaneous polyarteritis nodosa may also be
seen.8-10
Although the mucocutaneous component is paramount,
Behçet’s also may present with ocular, articular, neurologic,
and gastrointestinal involvement.3 Large-vessel vascular
complications develop in approximately 35% of patients
with Behçet’s disease. Involvement of arteries in vital organs portends a poorer prognosis frequently necessitating
immunosuppressive therapy.9-11 The most serious ocular
complication is posterior uveitis, which may result in blindness. In fact, Behçet’s disease is the leading cause of blindness in Japan.12 Other ocular complications include retinal
vaso-occlusion, anterior uveitis, hypopyon (purulence in the
dependent anterior chamber), and secondary changes such
Figure 3
as cataracts/glaucoma.13 Articular manifestations occur in
approximately half of all Behçet’s cases and include a
nonerosive, asymmetric oligoarthritis involving the ankles,
knees, and/or wrists. Neurologic complications, only
present in one-quarter of cases, are not part of the International Study Group criteria but herald severe morbidity
(meningitis/meningoencephalitis, cranial palsies, brain stem
lesions, psychiatric disturbances).3 Gastrointestinal involvement is often marked by abdominal pain, nausea, diarrhea,
and prominent ulcerations at the ileocecal region and ascending and descending colon.14 Perforation is the most
worrisome complication of gastrointestinal extension of
Behçet’s disease. Our case mirrors these findings with the
patient developing an associated systemic vasculitis (gastrointestinal extension) requiring azathioprine treatment.
When considering a diagnosis of Behçet’s disease, one
also must entertain the following differential:
●
●
Figure 2
Genital ulcers.
Erythema nodosum-like lesions.
Orogenital ulcers—aphthous stomatitis, herpes simplex virus, herpangina, hand-foot-and-mouth disease, varicellazoster virus, syphilitic chancre, deep fungal infection,
lupus erythematosus, vasculitis, and carcinoma.
Erythema nodosum-like lesions—acute sarcoidosis
(Lofgren’s syndrome), drug eruption (especially sulfonamides or oral contraception), bacterial/fungal infection,
inflammatory bowel disease, and lymphoma.4
Although the pathogenesis of Behçet’s has not been
completely elucidated, recent research has revealed a number of hereditary and environmental risk factors. Early theorists proposed an infectious cause, and although serum
herpes simplex virus antibodies are increased over healthy
controls, any direct link is purely speculative.6 Patients with
Behçet’s disease show increased hypersensitivity in cutaneous and plasma monocytes to Streptococcus sanguis antigens (which share homologies with heat shock proteins).15
324
The theory is that stressed oral mucosal cells, after bacterial
challenge, express antigenic heat shock proteins leading to
attack by antimucosal T lymphocytes.15 This molecular
mimicry model is loosely proven by the fact that prophylactic penicillin has been shown to decrease the frequency
and duration of mucocutaneous lesions.15
Despite 2 decades of genetic research, no single causative locus has been discovered. The most frequently recurring associated genetic region is human leukocyte antigenB*51 (yet it is present in only 13% of white patients and
60% of all Behçet’s cases).16 Other implicated immunologic
modulators include tumor necrosis factor, MICA alleles in
the major histocompatibility complex, induced endothelial
cell adhesion molecule-1, soluble CD8, interleukins 1 and 2,
and impaired T and B lymphocytes.16 The pathophysiology
of Behçet’s disease is further marked by vasculitis and
neutrophil hyperactivity (ie, increased oxidative burst and
chemotaxis). The hallmark of this enhanced neutrophil activity is the pathergy test, whereby a sterile needle prick
induces a cutaneous pustular vasculitis within 24 hours of
the intradermal trauma.
MANAGEMENT
Treatment options are dependent on both the site and the
severity of lesions. Mild disease generally responds to palliative and supportive case. Recurrent mucocutaneous lesions may be treated with a combination of nonsteroidal
anti-inflammatory agents, topical steroids, and colchicine
(ie, “first line”).17 Severe mucocutaneous and systemic
Behçet’s disease has been controlled with systemic steroids,
immunosuppressives (azathioprine/cyclophosphamide/cyclosporine), thalidomide, interferon-alpha, and anti-tumor
necrosis factor monoclonal antibodies (ie, “second line”).17
In addition to pharmacotherapy, exercise also helps patients with Behçet’s disease by the maintenance of strong
and flexible joints.1 The goal of all Behçet’s treatment is the
decrease in frequency and intensity of flares, thus limiting
morbidity and serious complications.
The American Journal of Medicine, Vol 120, No 4, April 2007
References
1. National Institutes of Arthritis and Muscoloskeletal and Skin Diseases.
Questions and answers about Behcet’s disease. Available at: http://
www.niams.nih.gov/hi/topics/behcets/behcets.htm. Accessed October
5, 2006.
2. Behçet H. Uber rezidiverende, apthose durch ein Virus verursachte
Geschwure am Mund, am Auge und an der Genitalen. Dermatol
Wochenschr. 1937;105:1152-1157.
3. Sakane T, Takeno M, Suzuki N, et al. Behçet’s disease. N Engl J Med.
1999;341:1284-1291.
4. Wolff K, Johnson RA, Suurmond D. Behcet’s syndrome. In: Ko C, ed.
Fitzpatrick’s Color Atlas & Synopsis of Clinical Dermatology. 5th ed.
New York: McGraw-Hill Medical Publishing; 2005:368-370.
5. Gul A, Inanc M, Ocal L, et al. Familial aggregation of Behcet’s disease
in Turkey. Ann Rheum Dis. 2000;59:622-625.
6. International Study Group for Behçet’s Disease. Criteria for diagnosis
of Behçet’s disease. Lancet. 1990;335:1078-1080.
7. Chen KR, Kawahara Y, Miyakawa S, et al. Cutaneous vasculitis in
Behçet’s disease: a clinical and histopathological study of 20 patients.
J Am Acad Dermatol. 1997;36:689-696.
8. Diaz-Perez JL, Winkelmann RK. Cutaneous periarteritis nodosa. Arch
Dermatol. 1974;110:407-414.
9. Liao YH, Hsiao GH, Hsiao CH. Behçet’s disease with cutaneous
changes resembling polyarteritis nodosa. Br J Dermatol. 1999;140:
368-369.
10. Agarwal V, Sachdev A, Singh R, et al. Behçet’s disease with relapsing
cutaneous polyarteritis nodosa-like lesions, responsive to oral cyclosporine therapy. Dermatology Online Journal. Available at: http://
dermatology.cdlib.org. Accessed May 20, 2006.
11. Huong DLT, Wechsler B, Papo T, et al. Arterial lesions in Behçet’s
disease: a study in 25 patients. Int J Dermatol. 1995;34:190-192.
12. Johns Hopkins Vasculitis Center. Types of vasculitis, Behçet’s disease. Available at: http://vasculitis.med.jhu.edu/typesof/behcets.html.
Accessed October 5, 2006.
13. Verity D, Wallace G, Vaughn R, et al. Behçet’s disease: from Hippocrates to the third millennium. Br J Ophthamol. 2003;87:1175-1183.
14. Bayraktar Y, Ozaslan E, Van Thiel D. Gastrointestinal manifestations
of Behçet’s disease. J Clin Gastroenterol. 2000;30:144-154.
15. Calguneri M, Ertenli I, Kiraz S, et al. Effect of prophylactic benzathine
penicillin on mucocutaneous symptoms of Behçet’s disease. Dermatology. 1996;192:125-128.
16. Verity D, Vaughn R, Marr J, et al. Behçet’s disease, The Silk Road,
and HLA-B51: historical and geographical perspectives. Tissue Antigens. 1999;54:213-220.
17. Goker B, Goker H. Current therapy for Behçet’s disease. Am J Ther.
2002;9:465-470.
The American Journal of Medicine (2007) 120, 325-327
ECG IMAGE OF THE MONTH
A Racing Heart
Julia H. Indik, MD, PhD, Section Editor
Sarver Heart Center, University of Arizona, Tucson
PRESENTATION
ASSESSMENT
A straightforward maneuver served 2 important purposes
for an 80-year-old man with a history of palpitations. He
had been having symptoms of a racing heartbeat for the past
year. Daily, he would feel his heart suddenly race, making
him short of breath. The episode was usually provoked by a
change in posture, such as bending over. One year earlier,
he had been started on the beta-adrenergic blocker metoprolol for these symptoms, but it had not been effective. His
past medical history included diabetes mellitus type 2 with
peripheral neuropathy, ischemic heart disease, hypertension, and hyperlipidemia. Treatment for these conditions
consisted of a diuretic, an ACE inhibitor, a statin medication, metformin, and insulin. Two years earlier, the patient
underwent 4-vessel coronary artery bypass graft surgery.
His blood pressure was 110/60 mm Hg with a heart rate of
120 beats per minute.
The patient’s ECG showed a wide complex tachycardia at a
rate of 120 beats per minute (Figure 1). However, the QRS
complexes are wide due to right bundle branch block. A
small deflection is seen in the early part of the T wave.
To help determine diagnosis, carotid sinus massage was
performed with the 12-lead ECG running (Figure 2). Tachycardia abruptly broke; the last 2 P waves became slightly
slower and evidence of block was evident in the AV node
(no QRS complex was present). Several seconds later, sinus
rhythm was established at a rate of 90 beats per minute with
a prolonged PR interval of 300 msec and right bundle
branch block (Figure 3).
Julia H. Indik, MD, PhD, Sarver Heart Center, University of Arizona,
1501 North Campbell Avenue, Tucson, AZ 85724.
E-mail address: [email protected]
DIAGNOSIS
This patient had atrial tachycardia, a form of supraventricular tachycardia. The finding of 2 nonconducted P waves
just as tachycardia terminated clinched the diagnosis, proving that the patient did not have AV reentrant tachycardia
from a bypass tract. In that situation, tachycardia cannot
continue without participation of both the atria and the
ventricles. Also, the P waves, wide-based and upright in
Figure 1 A wide complex tachycardia was seen in this ECG from a patient with daily symptoms of palpitations. Note the small deflection
at the beginning of the T wave (arrows shown in lead V3).
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2007.02.017
326
The American Journal of Medicine, Vol 120, No 4, April 2007
Figure 2 Carotid sinus massage terminated the tachycardia, with the last 2 P waves not conducted through the AV node (arrows shown
in leads II, AVF, and V2).
Figure 3
Sinus rhythm at a rate of 90 beats per minute was restored after the tachycardia terminated.
Indik
A Racing Heart
leads II and AVF, were very similar to the sinus P waves of
Figure 3. If the patient did have AV nodal reentrant tachycardia, the P waves would be the result of conduction
backwards from the AV node, and in that situation, they are
very narrow. Therefore, we were left with the diagnosis of
atrial tachycardia, which may have originated from an area
near the sinus node or from the crista terminalis. The PR
interval is relatively long, and this is probably due to a delay
in conduction through the AV node, which could be caused
by medications, such as beta-adrenergic blockers, or by
aging.
Most atrial tachycardias are due to automaticity, and they
cannot be stopped by carotid sinus massage. However, some
types of atrial tachycardia can be terminated with this maneuver. In particular, with triggered activity, there is an
excess of intracellular calcium that can prompt extra beats,
known as delayed afterdepolarizations.1 Usually atrial
tachycardia due to triggered activity is related to catecholamine excess, which can be caused by agents used to treat
asthma or by digoxin intoxication, but this patient was not
taking these medications.1 Instead, the finding of P waves
similar to the sinus P wave suggests the possibility that this
patient’s atrial tachycardia is due to a reentry mechanism—a circuit—possibly near the sinus node or the crista
terminalis; this type of atrial tachycardia can also be
abruptly terminated by a vagal maneuver such as carotid
sinus massage.1
MANAGEMENT
Atrial tachycardia may respond to medical therapy, which
can include beta-adrenergic blockers and calcium channel
blockers. Antiarrhythmic therapy can also be effective.
327
However, it is often the case that these tachycardias become
incessant or refractory to medical therapy, in which case,
catheter ablation is indicated.
It is important to remember useful techniques such as
carotid sinus massage to identify the mechanism of a tachycardia, as well as to terminate it. Pressure over the carotid
sinus activates baroreceptors, leading to a vagal reflex response from the autonomic nervous system. Gentle pressure
for 5 seconds is applied over the area where the carotid
pulse is strongest, which is just under the angle of the jaw.2
The right side should be tried first, as patients are more
likely to be responsive to this side, but if there is no response, then the left carotid should be tried. Carotid sinus
massage should be avoided in a patient with a carotid bruit,
known carotid stenosis, or a prior history of stroke or transient ischemic attack. Nonetheless, the risk of inducing a
stroke or transient ischemic attack, even in the elderly, is
very low, at less than 0.2%.3 Rare instances of ventricular
tachycardia and even ventricular fibrillation have also been
reported in the literature.2,4 Overall, though, carotid sinus
massage remains a very simple and useful technique for
diagnosis and termination of tachycardia.
References
1. Josephson ME. Clinical cardiac electrophysiology, Techniques and
Interpretations. 3rd ed. Philadelphia, PA: Lippincott Williams &
Wilkins; 250 –262.
2. Schweitzer P, Teichholz LE. Carotid sinus massage. Its diagnostic and
therapeutic value in arrhythmias. Am J Med. 1985;78:645-654.
3. Walsh T, Clinch D, Costelloe A, et al. Carotid sinus massage— how
safe is it? Age Ageing. 2006;35:518-520.
4. Deepak SM, Jenkins NP, Davidson NC, Bennett DH, Mushahwar SS.
Ventricular fibrillation induced by carotid sinus massage without preceding bradycardia. Europace. 2005;7:638-640.
The American Journal of Medicine (2007) 120, 328-329
IMAGES IN RADIOLOGY
Michael A. Bettmann, MD
A Complication of Forceful Nose-Blowing
Amy Fix, MD, Valerie J. Lang, MD
PRESENTATION
Four days prior to presentation, our patient was feeling well
when he forcefully blew his nose and felt a “pop.” Immediately following this incident, he began to experience pain
above his left eye. Over the next 4 days, the intensity of his
headache increased and worsened each time he blew his
nose. Two days prior to presentation he developed a clear
discharge from his right nostril which recurred when he
leaned forward. On the day of admission to an outside
hospital, the patient’s headache became severe, and he had
an episode of vomiting. At the outside hospital, over the
course of 3 hours, he became febrile to 38.2° C, and his
mental status declined to the point of unresponsiveness. A
head CT revealed diffuse su achnoid air. He was given a
dose of ceftriaxone and transferred to the university hospital
for further evaluation. The 73-year-old male patient had a
history of type II diabetes and remote cutaneous melanoma.
According to the family, he had no recent history of head or
facial trauma and no recent viral illness. They reported tha
he regularly blew his nose forcefully.
with 4020 white blood cells and 240 red blood cells. White
blood cell differential showed 86% neutrophils and 14%
lymphocytes. Cerebrospinal fluid glucose was 89 (with serum glucose 245), and protein was 860. Dipstick of the clear
nasal discharge was 3⫹ positive for glucose. Gram stain of
a cytospin sample of the cerebrospinal fluid showed a few
gram negative coccobacilli that could not be further
characterized.
DIAGNOSIS
Pneumocephalus was originally described by Chiari in 1884
by autopsy. With the advent of radiography, Luckett (1913)
ASSESSMENTS
On presentation to the university hospital, his temperature
was 37.8° C, heart rate 117, blood pressure 196/100, respiratory rate 16, and oxygen saturation 94% on 2 liters nasal
cannula. He opened his eyes to voice and moved all of his
extremities purposefully, but he did not follow commands
or interact in any other way. His pupils were equal and
reactive. His neck was stiff with forward flexion. He had no
rash, raccoon eyes, or Battle sign. Further head, neck, cardiac, pulmonary, and abdominal exams were normal. White
blood cell count was 13,700. Repeat head CT confirmed
pneumocephalus (Figure 1). However, there was no mass or
skull base defect noted on CT or on MRI done the following
day. A lumbar puncture revealed cloudy cerebrospinal fluid
Requests for reprints should be addressed to Valerie J. Lang, MD,
University of Rochester School of Medicine & Dentistry, Department of
Medicine, Hospital Medicine Division, 601 Elmwood Avenue, Box MEDHMD, Rochester, NY 14642.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.11.016
Figure 1 CT scan on hospital day 1 demonstrating diffuse noncompressive subarachnoid pneumocephalus.
Fix and Lang
Images in Radiology
published a case of air in the ventricles of a living man who
had been struck by a trolley car. Since that time, multiple
case reports have described pneumocephalus, mainly related to trauma. Some of these cases also had coexistant
cerebrospinal fluid rhinorrhea or otorrhea, and some developed meningitis as a result of the internal-external communication. In the absence of trauma, other identified causes of
pneumocephalus have included neoplasm, surgery, nasogastric tube insertion, bag-valve mask ventilation, nasotracheal
intubation, lumbar puncture, otitis media, post-radiation necrosis, and cerebrospinal fluid infection with gas-producing
or unidentified organisms.1-5 Several reports also have described pneumocephalus related to barotrauma from air
travel or diving. One report by Chan et al in 2000 identified
an elderly patient presenting with confusion after blowing
his nose during an airplane flight; pneumocephalus was
identified on CT scan.6
Skull base dehiscences are common, and patients with
such congenital defects could easily incur dural rupture at
these sites with Valsalva-like actions such as forceful noseblowing.7 In 1991, Tolley and Schwartz described a case of
chronic cerebral spinal fluid rhinorrhea in a patient with a
habit of forceful nose-blowing,7 and Orrell et al. reported a
case of septic cavernous sinus thrombosis and meningitis in
a patient with history of forceful nose-blowing.8 In 2004,
Richards et al. described pneumocephalus related to forceful nose-blowing in 2 patients who were found to have
hyperpneumatized mastoid bones on CT scan.9
Pneumocephalus together with meningitis leads to a limited differential: infection with gas-producing organisms or
a dural rupture, allowing both bacteria and air into the skull.
Our patient’s forceful nose-blowing followed by a “pop”
and frontal headache were consistent with dural rupture
leading to his pneumocephalus.
His clear nasal discharge was not tested for beta-2transferrin, which is a more accurate, but less widely available, test than glucose oxidase for identifying cerebrospinal
fluid. However, the timing, character, positive glucose, and
positional nature of the discharge were consistent with cerebrospinal fluid rhinorrhea. His persistence in forcefully
blowing his nose after the initial “pop” and discharge likely
forced intranasal colonizing bacteria through the skull base
329
defect, causing meningitis. It was felt that, based on the
cytospin gram stain, the likely etiology of his meningitis
was Haemophilus influenzae, a common gram negative colonizer of the nares and pharynx.
MANAGEMENT
Our patient was treated empirically with ceftriaxone, ampicillin, and vancomycin. Over the next 12 hours, he demonstrated dramatic clinical improvement, and his mental status
returned to normal. His neck stiffness resolved by hospital
day 2. With instructions to avoid nose blowing, the patient’s
clear rhinorrhea resolved spontaneously on hospital day 4.
On hospital day 7, he underwent a nuclear cisternogram,
which did not reveal a cerebrospinal fluid leak, consistent
with spontaneous resolution of a small defect. A follow-up
CT on day 10 revealed complete resolution of the subarachnoid air. Final cerebrospinal fluid culture was negative,
consistent with the administration of antibiotics several
hours before the specimen was obtained. The patient completed a 14-day course of ceftriaxone. The other antibiotics
were discontinued, and he had no recurrence of his
symptoms.
References
1. North JB. On the importance of intracranial air. Br J Surg. 1971;58(11):
826-9.
2. Goyal M, Sharma R, Berry M. Diffuse pneumocephalus due to meningitis: CT findings. Pediatr Radiol. 1996;26(4):278-9.
3. Maliwan N. Spontaneous pnemocephalus associated with mixed aerobic-anaerobic bacterial meningitis. J Infect Dis. 1985;152(4):847-8.
4. Turgut S, Ercan I, Alkan Z, Cakir B. A case of pneumocephalus and
meningitis as a complication of silent otitis media. Ear Nose Throat J.
2004;83(1):50-2.
5. Chan YP, Yau CY, Lewis RP, Kinirons MT. Acute confusion secondary
to pneumocephalus in an elderly patient. Age Ageing. 2000;29:365-7.
6. Jayaram S, Jadhav S, Rathod D, Tarvade S, Soman A. Meningitis: An
unusual cause of pneumocephalus. J Assoc Physicians India. 2004;52:
67-8.
7. Tolley NS, Schwartz P. Nose-blowing and CSF rhinorrhea. Lancet.
1991;337(8736):302.
8. Orrell RW, Guthrie JA, Lamb JT. Nose-blowing and CSF rhinorrhea.
Lancet. 1991;337(8744):804.
9. Richards SD, Saeed SR, Laitt R, Ramsden RT. Hypercellularity of the
mastoid as a cause of spontaneous pneumocephalus. J Laryngol Otol.
2004;118:474-6.
The American Journal of Medicine (2007) 120, 330-336
CLINICAL RESEARCH STUDY
Vitamin and Micronutrient Intake and the Risk of
Community-Acquired Pneumonia in US Women
Mark I. Neuman, MD, MPH,a,b Walter C. Willett, MD, DrPH,b,c,d,e Gary C. Curhan, MD, ScDb,c,d
a
Division of Emergency Medicine, Children’s Hospital, Boston, Mass; bChanning Laboratory, cDepartment of Medicine, Brigham and
Women’s Hospital/Harvard Medical School, Boston, Mass; dDepartments of Epidemiology and eNutrition, Harvard School of Public
Health, Boston, Mass.
ABSTRACT
BACKGROUND: There are limited data regarding the role of dietary and supplemental vitamin intake and
the risk of community-acquired pneumonia.
METHODS: We prospectively examined, during a 10-year period, the association between dietary and
supplemmin intake and the risk of community-acquired pneumonia among 83,165 women in
Nurses’ Health Study II who were between the ages of 27 and 44 years in 1991. We excluded women who
had pneumonia before 1991, those who did not provide complete dietary information, or those with a
history of cancer, cardiovascular disease, or asthma. Self-administered food frequency questionnaires were
used to assess dietary and supplemental vitamin intake. Cases of pneumonia required a diagnosis by a
physician and confirmation with a chest radiograph. The independent associations between specific
vitamins and pneumonia risk were evaluated.
RESULTS: There were 925 new cases of community-acquired pneumonia during 650,377 person-years
of follow up. After adjusting for age, cigarette smoking, body mass index, physical activity, total energy
intake, and alcohol consumption, there were no associations between dietary or total intake of any
individual vitamin and risk of community-acquired pneumonia. Specifically, women in the highest quintile
of vitamin A intake did not have a significantly lower risk of pneumonia than women in the lowest quintile
(multivariate relative risk [RR] ⫽ 0.88; 95% confidence interval [CI], 0.70-1.09, P for trend ⫽ .16).
Similarly, vitamin C (RR ⫽ 0.94; 95% CI, 0.76-1.16, P for trend ⫽ .81) and E (RR ⫽ 0.95; 95% CI,
0.76-1.17, P for trend ⫽ .74) intake did not alter risk of pneumonia.
CONCLUSIONS: Higher vitamin intake from diet and supplements is unlikely to reduce pneumonia risk
in well nourished women. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Vitamin; Micronutrient; Diet; Pneumonia
Community-acquired pneumonia is a common illness associated with significant morbidity and mortality.1-4 Many
factors, particularly waning immune responses and the onset of age-associated organ dysfunction, account for an
increase in susceptibility to respiratory tract infection in
older adults.5-7 Nutritional factors play a major role in the
This work was supported by Grant CA050385 from the National
Institutes of Health.
Requests for reprints should be addressed to Mark I. Neuman, MD,
Division of Emergency Medicine, Children’s Hospital, 300 Longwood
Ave, Boston, MA 02115.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.06.045
immune responses, and may affect the risk of development
of respiratory infections, particularly community-acquired
pneumonia.5,8,9 There is a lack of information regarding the
risk factors for the development of pneumonia, particularly
in well nourished individuals.
Individuals with vitamin deficiencies have been shown to
be at increased risk for respiratory infections,10 and vitamin
supplementation, in particular vitamins such as C and E,
have been shown to decrease the rate of respiratory infections in certain populations.11-16 Other studies have found
no benefit with vitamin supplementation in reducing the rate
of respiratory infections.6,17-20 Merchant and colleagues
demonstrated that vitamin E intake from food sources, but
Neuman et al
Vitamin Intake and Pneumonia
not total intake, was inversely associated with pneumonia
risk in US men.15,17 We prospectively studied over 80,000
well nourished women to determine whether vitamin intake
alters risk of community-acquired pneumonia.
331
in 1991. For each food, the participant indicated how often
they consumed a specified portion of food (eg, 6 to 8 ounces
of fish). The 9 potential response categories ranged from
“never, or less than once per month” to “6 or more times per
day.” Specific nutrient intakes were computed from the
reported frequency of consumpMETHODS
tion of each specified unit of food
or beverage using published data
Study Population
CLINICAL SIGNIFICANCE
on the nutrient content of the specDetails of the study design and
ified portions.23 Because total en● In well nourished women, vitamin intake
data collection used in the Nurses’
ergy intake for a given person
Health Study II (NHS II) have
did not reduce the risk of community21
tends to be fixed within a narrow
been previously published. The
acquired pneumonia.
range, variations in nutrient intake
NHS II study began in 1989 when
●
Among
smoking
women,
higher
intake
of
are largely attributable to changes
116,671 U.S. female registered
in diet composition, not total food
foods
rich
in
vitamin
E
reduced
the
risk
nurses aged 25-42 years returned a
consumption. Therefore, all comof community-acquired pneumonia.
mailed questionnaire. At the time
puted nutrient intakes were adof enrollment, participants projusted for total energy intake.24
vided a detailed medical history
Energy adjustment also reduces
including diagnosed diseases,
variation introduced by questionnaire responses that undermedications, and information on lifestyle factors including
report or over-report intake, thus improving the accuracy of
smoking, physical activity, and alcohol use. Information on
nutrient measurements.
dietary and supplemental vitamin intake was first ascertained in 1991 and updated every 4 years using a semiThe validity of the self-reported data from the food
quantitative food frequency questionnaire. Women were
frequency questionnaire has been demonstrated in the
excluded from the analysis if they had incomplete questionNurses’ Health Study I and Health Professionals Folnaires (12,360), had pneumonia before the baseline in 1991
low-Up Study cohorts.25-27 Compared with diet records,
(14,156), died before 1991 (37), or if they had a history of
the food frequency questionnaire was a good measure of
conditions known to affect pneumonia risk (6953), includspecific vitamin intake (correlation coefficients range
ing cancer, cardiovascular disease (myocardial infarction,
from 0.52 to 0.85).23,28,29 We assessed the following
stroke, or arterial surgery), or asthma diagnosed either benutrients: thiamine, riboflavin, niacin, pyridoxine, folate,
fore or during the study period.
vitamins B12, A, C, D, E, total carotenoids, alpha-carotene, beta-carotene, beta-cryptoxanthine, lycopene, lutein
Identification of Cases of Pneumonia
and zeaxanthin.
We considered a case to be physician-diagnosed pneuTo account for wide ranges in supplemental vitamin
monia if confirmed by chest radiography, and included
intake, specific vitamins were further categorized; pyridoxonly the first documented event of community-acquired
ine (⬍1.5 mg, 1.5-4 mg, 5-9 mg, 10-39 mg, 40-99 mg,
pneumonia occurring between June 1, 1991 and May 31,
100⫹ mg/day), vitamin B12 (⬍2.6 mg, 2.6-7.9 mg, 8-10
2001. Women who reported pneumonia were sent a supmg, 11-13 mg, 14⫹ mg/day), vitamin A (⬍5000 IU, 5000plementary questionnaire asking whether the pneumonia
7499 IU, 7500-9999 IU, 10,000-14,999 IU, and 15,000⫹
diagnosis had been confirmed by radiograph. To examine
IU/day), vitamin C (⬍90 mg, 90-249 mg, 250-499 mg,
the validity of self-reported pneumonia during the first 2
500-999 mg, 1000-1499 mg, and 1500⫹ mg/day).
years of follow-up, a study physician blinded to exposure
status examined the medical records of 76 women who
Assessment of Other Covariates
had reported pneumonia. A radiology report of a pulmoCovariates considered in the multivariate model included
nary infiltrate confirmed the presence of pneumonia in
age, body mass index (BMI), cigarette smoking, physical
82% of the cases.22 After the first 2 years of follow-up,
activity, and alcohol intake. BMI was calculated as weight
medical records were obtained from all women who reported radiologically diagnosed pneumonia. We rein kilograms divided by height in meters squared using the
viewed records from a sample of 99 confirmed cases and
reported height of the women at the start of the study and
found only one that was potentially hospital-acquired.
updated weight. Cigarette smoking was assessed on the
Therefore, we considered all the cases to have commubasis of whether the woman smoked and the amount
nity-acquired pneumonia.
smoked. Physical activity was assessed as the number of
MET-hours per week, the time invested in an activity every
Ascertainment of Vitamin Intake
week multiplied by the energy expenditure required by the
activity. Alcohol intake was classified by the amount the
A semi-quantitative food frequency questionnaire was used
to estimate vitamin and nutrient intake every 4 years starting
participant consumed per day.
332
Data Analysis
Person-time of follow-up was calculated as the time between the return of the 1991 questionnaire until the first
report of community-acquired pneumonia, death, or the end
of the study period (May 31, 2001). We first examined
age-adjusted models for the association between vitamin
intake and the risk of pneumonia. Cox proportional hazards
multivariate models with updating of exposure variables
were used to estimate multivariate relative risks (RR). The
age-adjusted relative risk of pneumonia was calculated per
quintile of specific vitamin intake, with the referent group
being the first quintile. Additionally, due to wide variation
in specific vitamin intake, we evaluated the association
between pyridoxine, vitamins B12, A, and C and pneumonia risk by further categorizing these variables into wider
contrasting groups.
The multivariate model adjusted for age, BMI (⬍21
kg/m2, 21-22.9, 23-24.9, 25-29.9, 30⫹), alcohol intake (0
gm/day, 0.1-5 gm, 5-9.9 gm, 10-14.9 gm, 15-29.9 gm, 30⫹
gm), cigarette smoking (never, past, current smoker of 1-14
cigarettes per day, 15-24 per day, or 25⫹ per day), and
physical activity (in quintiles of metabolic equivalents
[METS] per day). Additionally, to assess for possible confounding, we further adjusted for multivitamin use (0, 1-2,
3-5, 6-9, ⱖ10 multivitamin tablets per week).
We further analyzed the association between dietary vitamin intake (excluding supplements) and the risk of community-acquired pneumonia. We also studied the association between vitamin E intake (total intake, as well as diet
alone) and risk of community-acquired pneumonia among
smokers, as vitamin E intake has been shown to decrease the
risk of pneumonia in laboratory animals exposed to
smoke.30 This analysis was further adjusted for intake of
vegetable fat (in quintiles), as vegetable fat may also be rich
in vitamin E. Lastly, we studied the association between
multivitamin use (0, 1-2, 3-5, 6-9, ⱖ10 multivitamin tablets
per week) and the risk of community-acquired pneumonia.
We used the Mantel extension test to calculate tests for
trends across quintiles of intake using the respective median
values.31 SAS statistical software (SAS Institute Inc., Cary,
NC) was used for all analyses. Two-sided P-values ⬍.05
were considered significant.
This study was approved by the Human Subjects Committee of the Harvard School of Public Health.
RESULTS
During 10 years of follow-up (650,377 person-years), there
were 925 new cases of nonfatal community-acquired pneumonia. Women with higher vitamin intake were less likely
to be current smokers and exercised more at baseline in
1991 than women with lower vitamin intake (Table 1).
Forty-two percent of participants reported the use of supplemental multivitamins: 1-2 per week (7%), 3-5 (12%), 6-9
(21%), and 10⫹ per week (2%).
After adjusting for age, women in the highest quintile of
total vitamin A intake were less likely to develop pneumo-
The American Journal of Medicine, Vol 120, No 4, April 2007
nia than women in the lowest quintile (RR ⫽ 0.79; 95% CI,
0.64-0.98; P for trend ⫽ .02) (Table 2). This association was
no longer significant after further adjusting for smoking,
BMI, alcohol use, and physical activity (RR 0.88; 95% CI,
0.70-1.09; P for trend ⫽ .16). After multivariate adjustment,
there were no significant associations between vitamin C
intake (P, trend ⫽ .81) or vitamin E intake (P, trend ⫽ .74)
and risk of community-acquired pneumonia. There were
also no significant associations between intake of other
vitamins and micronutrients and community-acquired pneumonia risk. Inclusion of women with cancer, cardiovascular
disease, and asthma did not materially change our results.
There were no associations between vitamin intake from
diet alone (excluding intake of supplements) and risk of
community-acquired pneumonia (data not shown). Additionally, even after classification of specific vitamins into
categories (eg, vitamin C ⬍90 mg vs 1500⫹ mg) there were
no associations with risk of community-acquired pneumonia for pyridoxine (P value comparing extreme categories ⫽ .29), vitamin B12 (P ⫽ .34), vitamin A (P ⫽ .15), and
vitamin C (P ⫽ .67).
Among smokers, total vitamin E intake did not affect risk
of community-acquired pneumonia (multivariate P for trend
.40). Among smokers, after we excluded those who took
vitamin E supplements, women in the top quintile of vitamin E intake from food sources had a lower risk of community-acquired pneumonia (multivariate RR 0.46; 95% CI,
0.23-0.90; P for trend ⫽ .04) (Table 3). There was no material change after further adjusting for vegetable fat (RR
0.45; 95% CI, 0.21-0.93; P for trend ⫽ .055).
Multivitamin use did not lower the risk of communityacquired pneumonia. Compared with nonusers, the multivariate RR for women taking 10 or more multivitamin
tablets per week was 0.75 (95% CI, 0.45-1.26; P for
trend ⫽ .84) (Table 4).
DISCUSSION
We found that vitamin intake was not associated with community-acquired pneumonia in well-nourished, young and
middle-aged adult US women. Additionally, dietary vitamin
intake as well as multi-vitamin use did not appear to alter
risk. Although many studies have found that intakes of
specific vitamins reduce risk of pneumonia and respiratory
infections, these studies have been primarily conducted in
the elderly or poorly nourished individuals.
Nutritional status is an important determinant of immune
function.32,33 The beneficial effects of adequate nutrition
appear to be mediated through mechanisms such as increasing the number of T-cell subsets and natural killer cells,
improved lymphocyte proliferation response to mitogens,
interleukin-2 production, and antibody response to natural
killer cell activity.17,34 Individuals with vitamin deficiencies
are at increased risk for respiratory infections,10 and vitamin
supplementation has been shown to decrease the rate of
respiratory infections in certain populations.11-16,18-20
Neuman et al
Table 1
Vitamin Intake and Pneumonia
333
Age-Adjusted Characteristics of Women in the Top, Middle and Bottom Quintiles of Specific Vitamin Intake at Baseline (1991)
Smoking:
Current Smokers %
Alcohol Intake† g/d
Physical Activity†
MET/wk
BMI† kg/m2
1.3
1.8
5.1
18.8
8.7
10.6
4.0
2.7
2.8
18.4
20.8
23.8
24.7
24.3
24.1
1.4
2.3
6.0
17.6
10.1
10.6
3.6
2.8
2.8
17.7
20.8
23.8
24.5
24.5
24.2
19.2
27.8
55.9
14.3
11.4
11.4
2.9
3.3
3.1
18.4
20.3
24.7
24.3
24.6
24.2
1.7
2.5
8.8
17.0
9.9
10.6
3.0
3.2
3.0
16.5
22.0
24.0
24.5
24.4
24.3
4.0
7.0
17.0
13.1
11.8
12.0
3.4
3.1
2.7
20.1
19.9
22.9
24.1
24.7
24.4
5288
11095
21512
16.4
11.0
9.7
3.3
3.2
2.8
16.0
20.6
27.7
24.9
24.4
24.1
75
158
525
16.4
10.3
11.3
3.2
3.1
3.3
15.7
21.1
26.0
24.9
24.4
23.9
15.8
11.3
10.4
3.2
3.3
3.0
16.9
21.6
24.3
24.5
24.7
24.1
3130
7540
16456
15.5
11.3
9.8
3.1
3.2
3.0
16.0
20.1
27.9
24.9
24.4
24.1
160
577
1544
15.3
12.0
9.3
3.4
3.2
2.8
17.9
19.8
26.1
24.7
24.4
24.2
1554
3578
7432
15.2
11.0
10.6
2.9
3.2
3.1
15.6
20.1
28.4
24.8
24.3
24.1
16.5
11.0
9.6
3.5
3.2
2.6
18.0
20.8
24.9
24.6
24.4
24.2
3305
7227
15917
12.9
12.0
11.8
2.8
3.1
3.4
19.1
20.6
23.8
24.8
24.3
24.4
942
2566
5968
12.9
11.4
12.3
2.4
3.2
3.7
15.6
20.4
28.0
24.9
24.4
24.0
224
371
866
17.8
10.2
9.0
3.1
3.3
2.6
15.9
22.3
24.0
25.0
24.2
24.2
128
316
742
16.7
10.6
9.2
3.5
3.1
2.5
17.9
21.4
23.9
24.7
24.4
24.2
Median Intake*
Thiamine (mg/day)
Quintile 1
Quintile 3
Quintile 5
Riboflavin (mg/day)
Quintile 1
Quintile 3
Quintile 5
Niacin (mg/day)
Quintile 1
Quintile 3
Quintile 5
Pyridoxine (mg/day)
Quintile 1
Quintile 3
Quintile 5
Vitamin B12 (␮g/day)
Quintile 1
Quintile 3
Quintile 5
Vitamin A (IU/day)
Quintile 1
Quintile 3
Quintile 5
Vitamin C (mg/day)
Quintile 1
Quintile 3
Quintile 5
Vitamin E (mg/day)
Quintile 1
Quintile 3
Quintile 5
Total Carotenoids (IU/day)
Quintile 1
Quintile 3
Quintile 5
Alpha-carotene (␮g/day)
Quintile 1
Quintile 3
Quintile 5
Beta-carotene (␮g/day)
Quintile 1
Quintile 3
Quintile 5
Beta-cryptoxanthine (␮g/day)
Quintile 1
Quintile 3
Quintile 5
Lycopene (␮g/day)
Quintile 1
Quintile 3
Quintile 5
Lutein and Zeaxanthin (␮g/day)
Quintile 1
Quintile 3
Quintile 5
Folate (␮g/day)
Quintile 1
Quintile 3
Quintile 5
Vitamin D (IU/day)
Quintile 1
Quintile 3
Quintile 5
4.9
8.1
45.9
6.4
35.5
103.9
*Based in intake from food and supplements in 1991.
†Values are means.
334
Table 2
The American Journal of Medicine, Vol 120, No 4, April 2007
Relative Risks (RR) for Total Vitamin Intake and Community-Acquired Pneumonia
Quintile of Intake from All Sources
Thiamine
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Riboflavin
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Pyridoxine
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Niacin
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Vitamin B12
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Vitamin A
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Vitamin C
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Vitamin E
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Total Carotenoids
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Alpha-carotene
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Beta-carotene
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Beta-cryptoxanthine
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Lycopene
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Lutein ⫹ Zeaxanthine
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Folate
No. of cases (n)
Age adjusted RR
Multivariate adjusted
Vitamin D
No. of cases (n)
Age adjusted RR
Multivariate adjusted
P-value,
test for trend
1
2
3
4
5
RR*
181
Ref
Ref
174
0.94 (.76-1.16)
0.96 (.78-1.20)
212
1.14 (.94-1.40)
1.22 (.99-1.50)
192
1.06 (.86-1.30)
1.14 (.92-1.41)
166
0.92 (.75-1.14)
0.98 (.79-1.22)
.33
.60
RR*
176
Ref
Ref
188
1.05 (.85-1.29)
1.07 (.86-1.32)
213
1.17 (.96-1.43)
1.21 (.98-1.49)
170
0.96 (.77-1.18)
1.03 (.83-1.29)
178
1.02 (.83-1.25)
1.07 (.86-1.33)
.61
.99
RR*
176
Ref
Ref
191
1.08 (.88-1.32)
1.10 (.89-1.36)
200
1.13 (.93-1.39)
1.17 (.95-1.44)
178
1.02 (.83-1.25)
1.08 (.87-1.34)
180
1.05 (.85-1.29)
1.11 (.89-1.37)
.93
.72
RR*
182
Ref
Ref
184
1.01 (.82-1.24)
0.98 (.79-1.21)
200
1.09 (.89-1.33)
1.05 (.86-1.30)
195
1.07 (.87-1.31)
1.08 (.88-1.33)
164
0.93 (.76-1.15)
0.94 (.75-1.17)
.50
.69
RR*
198
Ref
Ref
179
0.90 (.73-1.10)
0.89 (.72-1.10)
198
1.00 (.82-1.21)
0.98 (.80-1.20)
191
0.98 (.80-1.19)
0.96 (.78-1.18)
159
0.83 (.67-1.02)
0.85 (.69-1.06)
.17
.30
RR*
201
Ref
Ref
191
0.96 (.79-1.17)
1.01 (.83-1.24)
197
0.99 (.81-1.12)
1.08 (.88-1.32)
184
0.94 (.77-1.15)
1.02 (.82-1.25)
152
0.79 (.64-.98)
0.88 (.70-1.09)
.02
.16
RR*
204
Ref
Ref
178
0.87 (.71-1.06)
0.90 (.73-1.11)
189
0.92 (.76-1.12)
1.00 (.82-1.23)
184
0.93 (.76-1.13)
1.02 (.83-1.26)
170
0.88 (.72-1.08)
0.94 (.76-1.16)
.56
.81
RR*
204
Ref
Ref
171
0.85 (.69-1.04)
0.87 (.71-1.07)
182
0.89 (.73-1.09)
0.95 (.77-1.16)
194
0.96 (.79-1.17)
1.01 (.82-1.24)
170
0.88 (.72-1.08)
0.95 (.76-1.17)
.51
.74
RR*
193
Ref
Ref
203
1.04 (.86-1.27)
1.08 (.88-1.33)
191
1.00 (.81-1.22)
1.06 (.86-1.30)
174
0.92 (.75-1.13)
1.00 (.81-1.24)
164
0.88 (.71-1.08)
0.97 (.78-1.20)
.06
.34
RR*
190
Ref
Ref
203
1.17 (.96-1.42)
1.23 (1.00-1.51)
163
0.97 (.78-1.19)
1.01 (.81-1.25)
193
1.07 (.87-1.31)
1.16 (.94-1.43)
176
0.99 (.80-1.22)
1.05 (.85-1.31)
.51
.94
RR*
196
Ref
Ref
194
1.02 (.84-1.25)
1.08 (.88-1.32)
208
1.01 (.83-1.23)
1.07 (.87-1.32)
160
0.92 (.75-1.13)
1.01 (.82-1.22)
167
0.90 (.73-1.11)
0.98 (.79-1.22)
.13
.45
RR*
198
Ref
Ref
180
0.99 (.81-1.21)
1.04 (.84-1.28)
202
1.00 (.82-1.22)
1.03 (.84-1.27)
167
0.85 (.69-1.05)
0.91 (.73-1.13)
178
0.94 (.76-1.15)
1.03 (.83-1.27)
.36
.99
RR*
178
Ref
Ref
198
1.04 (.84-1.27)
1.08 (.88-1.34)
174
1.00 (.82-1.24)
1.04 (.84-1.29)
180
1.01 (.82-1.24)
1.07 (.86-1.32)
195
1.08 (.88-1.33)
1.12 (.91-1.39)
.50
.35
RR*
183
Ref
Ref
192
1.04 (.85-1.27)
1.08 (.88-1.33)
211
1.14 (.94-1.40)
1.22 (.99-1.50)
164
0.95 (.77-1.17)
1.04 (.83-1.29)
175
1.04 (.85-1.29)
1.17 (.94-1.45)
.98
.31
RR*
182
Ref
Ref
193
1.07 (.88-1.32)
1.13 (.92-1.40)
176
0.95 (.77-1.17)
1.03 (.83-1.28)
189
1.05 (.85-1.28)
1.15 (.93-1.42)
181
1.01 (.82-1.24)
1.13 (.91-1.40)
.98
.36
RR*
197
Ref
Ref
177
0.88 (.72-1.08)
0.91 (.74-1.12)
187
0.93 (.76-1.13)
0.97 (.79-1.20)
182
0.91 (.74-1.10)
0.96 (.78-1.18)
182
0.92 (.76-1.13)
1.02 (.83-1.26)
1.0
1.0
RR ⫽ relative risk (95% confidence interval [CI]).
Dietary information for vitamin E and folate were missing for 4 participants from 1999 questionnaire preceding the development of pneumonia.
*Multivariate model adjusted for age in months, smoking (never smokers, past smokers and current smokers who smoked 1-14, 15-24, or ⱖ25
cigarettes/d), BMI (⬍21, 21-22.9, 23-24.9, 25-29.9, and ⱖ30 kg/m2), alcohol use (never, 0.1-4.9, 5.0-14.9, 15.0-29.9, and ⱖ30 gm/d), and physical
activity (quintiles of MET).
Neuman et al
Vitamin Intake and Pneumonia
335
Table 3 Relative Risks (RR) for Dietary Vitamin E Intake and Community-Acquired Pneumonia among Current Smokers
(n ⫽ 12,003)
Quintile of Vitamin E Intake from Diet Only
Age-adjusted RR
Multivariate adjusted RR*
1
2
3
4
5
P-value,
test for trend
Ref
Ref
0.68 (.40-1.18)
0.62 (.35-1.11)
0.58 (.32-1.03)
0.57 (.31-1.05)
0.64 (.37-1.13)
0.72 (.41-1.28)
0.44 (.23-.84)
0.46 (.23-.90)
.01
.04
*Multivariate model adjusted for age in months, BMI (⬍21, 21-22.9, 23-24.9, 25-29.9, and ⱖ30 kg/m2), alcohol use (never, 0.1-4.9, 5.0-14.9,
15.0-29.9, and ⱖ30 gm/d), and physical activity (quintiles of MET).
Micronutrients can improve cell-mediated immunity and
reduce oxidative stress.35 Vitamin E supplementation in
healthy, well-nourished adults increased delayed hypersensitivity responses, response to hepatitis B vaccine, lymphocyte proliferation, and decreased formation of immunosuppressive prostaglandins.17,36 Vitamin C regenerates the
antioxidant form of vitamin E and is critical for the killing
of pathogens by neutrophils.17 Vitamin B6 (pyridoxine)
supplementation enhanced lymphocyte proliferation and interleukin-2 levels in young women.37
Merchant and colleagues reported that vitamin E intake
from food sources was inversely associated with pneumonia
risk among U.S. men (multivariate relative risk comparing
extreme quintiles ⫽ 0.58, 95% CI, 0.39-0.86, P-value for
trend ⫽ .01), but no association was observed when vitamin
E from supplements was included.17 In contrast, we found
that dietary vitamin E intake was not associated with community-acquired pneumonia in healthy, well-nourished
women. Both of these large prospective cohort studies
found that intake of other vitamins (including dietary and
supplemental intake) was not associated with risk of pneumonia. Additionally, both studies found that multivitamin
use was not associated with pneumonia risk. Although certain factors, such as excessive weight gain and cigarette
smoking, have been found to increase the risk of community-acquired pneumonia in both men and women, other
factors such as sedentary lifestyle and obesity were associated with pneumonia risk only among women.22 There
might be differences between males and females that may
affect susceptibility to pneumonia, however, there is insufficient information available. Our results in women, along
with Merchant’s study of pneumonia in older men, suggest
Table 4
that higher vitamin intake is unlikely to reduce the risk of
community-acquired pneumonia in well-nourished adults.17
Pacht et al found a deficiency of vitamin E in the alveolar
fluid of cigarette smokers,30 and a recent laboratory-based
study showed that vitamin E supplementation was protective against the development of pneumonia among rats
exposed to cigarette smoke.38 We found that among smokers, dietary, but not total, intake of vitamin E was inversely
associated with risk of community-acquired pneumonia. A
possible reason for this observation is that foods containing
vitamin E may have some other beneficial nutrient. Vitamin
E-rich foods (ie, nuts) also contain fatty acids, which have
been shown to modulate inflammation and immunity, and
may alter risk of respiratory infections and pneumonia.39-42
Dietary information is unlikely to be influenced by recall
bias because it was gathered prospectively. Misclassification of the diagnosis of community-acquired pneumonia is
certainly possible, however, we included only participants
with physician-diagnosed and radiographically confirmed
pneumonia. We were unable to distinguish bacterial and
viral pneumonia, but even in the best of circumstances, the
microbiological etiology of pneumonia is difficult to establish.43-45 Additionally, because some of the nurses were
working in a hospital setting, it is possible that more than
1% of the cases may have been “hospital-acquired pneumonia”. Lastly, our results are generalizable to healthy, young
and middle-aged women.
In conclusion, vitamin intake does not alter communityacquired pneumonia risk in healthy young and middle-aged
women. Among smokers, higher intake of foods rich in
vitamin E may reduce the risk of community-acquired
pneumonia.
Relative Risks (RR) for Multivitamin Intake and Community-Acquired Pneumonia
Multivitamin Intake (# of Tablets) per Week
Age adjusted RR
Multivariate adjusted RR*
0
1-2
3-5
6-9
10⫹
P-value,
test for trend
Ref
Ref
0.97 (.76-1.25)
0.97 (.75-1.26)
1.01 (.83-1.23)
0.99 (.80-1.21)
1.03 (.88-1.21)
1.05 (.90-1.24)
0.75 (.45-1.23)
0.75 (.45-1.26)
0.95
0.84
*Multivariate model adjusted for age in months, smoking (never smokers, past smokers and current smokers who smoked 1-14, 15-24, or ⱖ25
cigarettes/d), BMI (⬍21, 21-22.9, 23-24.9, 25-29.9, and ⱖ30 kg/m2), alcohol use (never, 0.1-4.9, 5.0-14.9, 15.0-29.9, and ⱖ30 gm/d), and physical
activity (quintiles of MET).
336
References
1. National Center for Injury Prevention and Control. Activity Report
2001: CDC’s Unintentional Injury and Prevention Program. Atlanta,
GA: Centers for Disease Control and Prevention; 2002.
2. American Lung Association. Pneumonia Fact Sheet. 2002.
3. File TM Jr, Tan JS. Pneumonia in older adults: reversing the trend.
JAMA. 2005;294:2760-2763.
4. Fry AM, Shay DK, Holman RC, Curns AT, Anderson LJ. Trends in
hospitalizations for pneumonia among persons aged 65 years or older
in the United States, 1988-2002. JAMA. 2005;294:2712-2719.
5. Miller RA. The aging immune system: primer and prospectus. Science.
1996;273:70-74.
6. Graat JM, Schouten EG, Kok FJ. Effect of daily vitamin E and
multivitamin-mineral supplementation on acute respiratory tract infections in elderly persons: a randomized controlled trial. JAMA. 2002;
288:715-721.
7. Meyer KC. Lung infections and aging. Ageing Res Rev. 2004;3:55-67.
8. Lesourd B, Mazari L. Nutrition and immunity in the elderly. Proc Nutr
Soc. 1999;58:685-695.
9. Mazari L, Lesourd BM. Nutritional influences on immune response in
healthy aged persons. Mech Ageing Dev. 1998;104:25-40.
10. Reyes H, Villalpando S, Perez-Cuevas R, et al. Frequency and determinants of vitamin A deficiency in children under 5 years of age with
pneumonia. Arch Med Res. 2002;33:180-185.
11. Hemila H. Vitamin C supplementation and respiratory infections: a
systematic review. Mil Med. 2004;169:920-925.
12. Hemila H, Douglas RM. Vitamin C and acute respiratory infections.
Int J Tuberc Lung Dis. 1999;3:756-761.
13. Hemila H, Kaprio J. Vitamin E and respiratory tract infections in
elderly persons. JAMA. 2004;292:2834; author reply 2834.
14. Hemila H, Virtamo J, Albanes D, Kaprio J. Vitamin E and betacarotene supplementation and hospital-treated pneumonia incidence in
male smokers. Chest. 2004;125:557-565.
15. Meydani SN, Han SN, Hamer DH. Vitamin e and respiratory infection
in the elderly. Ann N Y Acad Sci. 2004;1031:214-222.
16. Sempertegui F, Estrella B, Camaniero V, et al. The beneficial effects
of weekly low-dose vitamin A supplementation on acute lower respiratory infections and diarrhea in Ecuadorian children. Pediatrics.
1999;104(1):e1.
17. Merchant AT, Curhan G, Bendich A, Singh VN, Willett WC, Fawzi
WW. Vitamin intake is not associated with community-acquired pneumonia in U.S. men. J Nutr. 2004;134:439-444.
18. Fawzi WW, Mbise R, Spiegelman D, Fataki M, Hertzmark E, Ndossi
G. Vitamin A supplements and diarrheal and respiratory tract infections among children in Dar es Salaam, Tanzania. J Pediatr. 2000;
137:660-667.
19. Fawzi WW, Mbise RL, Fataki MR, et al. Vitamin A supplementation
and severity of pneumonia in children admitted to the hospital in Dar
es Salaam, Tanzania. Am J Clin Nutr. 1998;68:187-192.
20. Kjolhede CL, Chew FJ, Gadomski AM, Marroquin DP. Clinical trial
of vitamin A as adjuvant treatment for lower respiratory tract infections. J Pediatr. 1995;126:807-812.
21. Curhan GC, Chertow GM, Willett WC, et al. Birth weight and adult
hypertension and obesity in women. Circulation. 1996;94:1310-1315.
22. Baik I, Curhan GC, Rimm EB, Bendich A, Willett WC, Fawzi WW.
A prospective study of age and lifestyle factors in relation to community-acquired pneumonia in US men and women. Arch Intern Med.
2000;160:3082-3088.
23. Willett WC, Sampson L, Stampfer MJ, et al. Reproducibility and
validity of a semiquantitative food frequency questionnaire. Am J
Epidemiol. 1985;122:51-65.
24. Willett W, Stampfer MJ. Total energy intake: implications for epidemiologic analyses. Am J Epidemiol. 1986;124:17-27.
25. Salvini S, Hunter DJ, Sampson L, et al. Food-based validation of a
dietary questionnaire: the effects of week-to-week variation in food
consumption. Int J Epidemiol. 1989;18:858-867.
The American Journal of Medicine, Vol 120, No 4, April 2007
26. Rimm EB, Giovannucci EL, Stampfer MJ, Colditz GA, Litin LB,
Willett WC. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male
health professionals. Am J Epidemiol. 1992;135:1114-1126; discussion 1127-1136.
27. Feskanich D, Rimm EB, Giovannucci EL, et al. Reproducibility and
validity of food intake measurements from a semiquantitative food
frequency questionnaire. J Am Diet Assoc. 1993;93:790-796.
28. Stryker WS, Kaplan LA, Stein EA, Stampfer MJ, Sober A, Willett
WC. The relation of diet, cigarette smoking, and alcohol consumption
to plasma beta-carotene and alpha-tocopherol levels. Am J Epidemiol.
1988;127:283-296.
29. Jacques PF, Sulsky SI, Sadowski JA, Phillips JC, Rush D, Willett WC.
Comparison of micronutrient intake measured by a dietary questionnaire and biochemical indicators of micronutrient status. Am J Clin
Nutr. 1993;57:182-189.
30. Pacht ER, Kaseki H, Mohammed JR, Cornwell DG, Davis WB. Deficiency of vitamin E in the alveolar fluid of cigarette smokers. Influence on alveolar macrophage cytotoxicity. J Clin Invest. 1986;77:789796.
31. Mantel N. Chi-square tests with one degree of freedom: extensions of
the Mantel-Haenszel procedure. J Am Stat Assoc. 1963;58:690-700.
32. Hamer DH, Meydani SN. Immune function in the elderly. Arch Intern
Med. 2001;161:482-483.
33. Meydani SN, Leka LS, Fine BC, et al. Vitamin E and respiratory tract
infections in elderly nursing home residents: a randomized controlled
trial. JAMA. 2004;292:828-836.
34. Pike J, Chandra RK. Effect of vitamin and trace element supplementation on immune indices in healthy elderly. Int J Vitam Nutr Res.
1995;65:117-121.
35. Evans P, Halliwell B. Micronutrients: oxidant/antioxidant status. Br J
Nutr. 2001;85(Suppl 2):S67-S74.
36. Meydani SN, Meydani M, Blumberg JB, et al. Vitamin E supplementation and in vivo immune response in healthy elderly subjects. A
randomized controlled trial. JAMA. 1997;277:1380-1386.
37. Bendich A, Zilberboim R: Drug/nutrient interactions and immune
function. In: Armente V, ed. Handbook of Drug-Nutrient Interactions,
5th ed. Totowa, NJ: Humana Press; 2004:468.
38. Ozlu T, Cay M, Akbulut A, Yekeler H, Naziroglu M, Aksakal M. The
facilitating effect of cigarette smoke on the colonization of instilled
bacteria into the tracheal lumen in rats and the improving influence of
supplementary vitamin E on this process. Respirology. 1999;4:245248.
39. Baughman RP, Stein E, MacGee J, Rashkin M, Sahebjami H. Changes
in fatty acids in phospholipids of the bronchoalveolar fluid in bacterial
pneumonia and in adult respiratory distress syndrome. Clin Chem.
1984;30:521-523.
40. Calder PC, Davis J, Yaqoob P, Pala H, Thies F, Newsholme EA.
Dietary fish oil suppresses human colon tumour growth in athymic
mice. Clin Sci (Lond). 1998;94:303-311.
41. James MJ, Gibson RA, Cleland LG. Dietary polyunsaturated fatty
acids and inflammatory mediator production. Am J Clin Nutr. 2000;
71(1 Suppl):343S-348S.
42. Venuta A, Spano C, Laudizi L, Bettelli F, Beverelli A, Turchetto E.
Essential fatty acids: the effects of dietary supplementation among
children with recurrent respiratory infections. J Int Med Res. 1996;24:
325-330.
43. Korppi M. Non-specific host response markers in the differentiation
between pneumococcal and viral pneumonia: what is the most accurate
combination? Pediatr Int. 2004;46:545-550.
44. Swingler GH. Radiologic differentiation between bacterial and viral
lower respiratory infection in children: a systematic literature review.
Clin Pediatr (Phila). 2000;39:627-633.
45. Bettenay FA, de Campo JF, McCrossin DB. Differentiating bacterial
from viral pneumonias in children. Pediatr Radiol. 1988;18:453-454.
The American Journal of Medicine (2007) 120, 337-342
CLINICAL RESEARCH STUDY
Grip Strength Predicts Cause-Specific Mortality in
Middle-Aged and Elderly Persons
Hideo Sasaki, MD, PhD,a,b Fumiyoshi Kasagi, PhD,c Michiko Yamada, MD, PhD,a Shoichiro Fujita, PhDd
a
Department of Clinical Studies, Radiation Effects Research Foundation, Hiroshima, Japan; bHealth Management and Promotion
Center, Hiroshima Atomic Bomb Casualty Council, Hiroshima, Japan; cDepartment of Epidemiology and dDepartment of Statistics,
Radiation Effects Research Foundation, Hiroshima, Japan.
ABSTRACT
PURPOSE: Handgrip strength is a simple measurement used to estimate overall muscle strength but
might also serve as a predictor of health-related prognosis. We investigated grip strength-mortality
association in a longitudinal study.
METHODS: A total of 4912 persons (1695 men and 3217 women), 35 to 74 years old at baseline, were
the subjects of this study. Members of the Adult Health Study (AHS) cohort in Hiroshima, Japan, these
individuals underwent a battery of physiological tests, including handgrip-strength testing, between July
1970 and June 1972. Mortality was followed until the end of 1999. Estimates of relative risk (RR) of
mortality associated with grip strength were adjusted for potentially confounding factors by Cox proportional hazard analysis.
RESULTS: Multivariate-adjusted RR of all causes of death, except for external causes, for the highest
quintile of grip strength in men was 0.52 (95% confidence interval [CI], 0.33-0.80) for the age group 35-54
years, 0.72 (95% CI, 0.53-0.98) for the ages 55-64 years, and 0.67 (95% CI, 0.49-0.91) for the ages 65-74
years. These figures were significantly lower than the RR for the reference group (the third quintile).
Similar trends were observed in women. Multivariate-adjusted RR of all causes of death except external
causes for each 5-kg increment of grip strength was significantly low (RR: 0.89, 95% CI, 0.86-0.92 for
men, RR: 0.87, 95% CI, 0.83-0.92 for women). Multivariate-adjusted RR for heart disease, stroke, and
pneumonia in men was 0.85 (95% CI, 0.79-0.93), 0.90 (95% CI, 0.83-0.99), and 0.85 (95% CI, 0.75-0.98),
respectively. RR for each 5-kg increment of grip strength remained 0.92 (95% CI, 0.87-0.96), even after
more than 20 years of follow-up.
CONCLUSION: Grip strength is an accurate and consistent predictor of all causes of mortality in
middle-aged and elderly persons. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Grip strength; Mortality; Prognosis; Cohort study
Grip-strength measurement is useful for assessing approximate overall muscle strength of middle-aged and elderly
people, and longitudinal studies revealed that it also predicts
health-related prognosis.1-13 The prognostic outcomes in
these studies included functional limitation, functional decline,4,5 activities of daily living dependence,6,13 and mortality.1-3,8,9,11,12 The majority of the mortality studies utiRequests for reprints should be addressed to Hideo Sasaki, MD, PhD,
Health Management and Promotion Center, Hiroshima A-Bomb Casualty
Council, 3-8-6, Senda-machi, Naka-ku, Hiroshima 730-0052, Japan.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.04.018
lized total mortality, with only one analyzing association
between grip strength and cause-specific mortality.1
The subjects in such studies, however, were restricted to
hospitalized patients5,12 or elderly people.1,3,6 The number
of studies on association between grip strength and prognosis conducted on middle-aged, relatively healthy subjects of
both sexes is limited.2,11
Further, most of the studies’ follow-up periods were less
than 10 years after baseline measurement of grip strength.
Only 2 studies, one in Canada and the other in Hawaii,
followed-up subjects for a longer duration, 13 and 30 years,
respectively.2,9
338
The American Journal of Medicine, Vol 120, No 4, April 2007
The Adult Health Study (AHS) is a longitudinal cohort
Other Measurements
study conducted at the Radiation Effects Research FoundaClinical examinations included taking of medical history,
tion (RERF) in Hiroshima and Nagasaki. Grip strength was
general physical examinations, anthropometrical examinameasured with other physiological function tests during
tions including height and body weight measurements, and
1970-1972 to assess the overall physiological condition of
laboratory tests. Body mass index (BMI) was calculated as
4912 persons in Hiroshima.14 This
body weight (kg) divided by the
population-based study is characsquare of height (m). Physical exterized as one of the largest cohort
aminations included blood presCLINICAL SIGNIFICANCE
studies of middle-aged and elderly
sure measurement taken by sphygpersons, with a long follow-up of
momanometer at the right arm
● Grip strength is a strong and consistent
more than 25 years.
with the subject in a sitting posipredictor of all causes of mortality in
The aforementioned studies of
tion. General laboratory tests conmiddle-aged and elderly persons.
grip strength-mortality association
sisted of peripheral blood test and
● In accordance with the present findings,
did not clarify whether such assodetermination of serum cholesciation is universal. The present
regular exercise for improvement of
terol and other biochemical items.
study was conducted using the
physical and musculoskeletal fitness is
Information about smoking habits
large cohort to clarify grip
and alcohol intake was obtained
eagerly recommended to improve
strength-mortality association in
from self-administrated questionprognosis.
more detail: first, by investigating
naires conducted during the period
association in an age and sex-spe1965-1968. Categories for smokcific manner, second, by studying
ing habits were “never,” “former,”
cause-specific association, and third, by determining time
and “current.” For current smokers, a question was included
trends of this association after dividing the follow-up into
about quantity of cigarettes smoked. Categories for alcohol
intake were “never” and “former/current.” Individual radidifferent time periods.
ation dose estimates were based on RERF’s 1986 Dosimetry
System (DS86).16
METHODS
Subjects were members of the AHS cohort of RERF. The
AHS was begun in 1958 by RERF’s predecessor, the
Atomic Bomb Casualty Commission (ABCC), as a clinical
cohort study to investigate the long-term medical and biological effects of exposure to atomic-bomb radiation among
the survivors and unexposed controls in Hiroshima and
Nagasaki. Approximately 20,000 persons were invited to
participate in biennial health examinations conducted by
ABCC (and later, RERF) clinical physicians. A detailed
description of the examinations, which included clinical
evaluations and routine laboratory determinations, is available elsewhere.15
Mortality Follow-up
Mortality was followed-up for the entire study sample from
the time of grip-strength examination in 1970-1972 until the
end of 1999. Deaths were routinely identified through Japan’s koseki (obligatory household registry) system, and
ascertainment is essentially complete. Causes of death were
obtained from death certificates. Underlying causes of death
were classified into the following categories: cancer (140208 by International Classification of Diseases [ICD] 8th or
9th), heart disease (390-429 by ICD 8th or 9th), coronary
heart disease (CHD, 410 by ICD 8th or 9th), stroke (430-438
by ICD 8th or 9th), cerebral infarction (CI, 433 and 434 by
Handgrip Measurement
Current study subjects were derived from 6129 persons in
the AHS cohort who underwent a battery of noninvasive
age-related physiological tests between July 1970 and June
1972 in Hiroshima.14 Handgrip strength was measured 2
times for both left and right hands with subjects in a standing position using a dynamometer in units of kilograms.
Grip devices were calibrated with known weights. Subjects
held the dynamometer at thigh level and were encouraged to
exert the strongest possible force. The maximum grip
strength among all measurements was used for the present
analysis. Analysis was restricted to 4912 subjects (1695
men and 3217 women) in Hiroshima who were 35 to 74
years old at the time of examination and who had completed
the grip-strength measurements, as well as provided other
clinical information (Table 1).
Table 1
Clinical Characteristics of the Study Subjects
Number of subjects
Age at examination (years)
Systolic blood pressure
(mm Hg)
Diastolic blood pressure
(mm Hg)
Total cholesterol (mg/dL)
Body mass index (kg/m2)
Mean grip strength (kg)
Current smoker (%)
Cigarettes smoked (per day)
Current alcohol drinker (%)
Men
Women
1689
55.5 ⫾ 11.1
130.9 ⫾ 23.6
3209
53.9 ⫾ 10.7
125.5 ⫾ 23.4
81.9 ⫾ 26.1
78.7 ⫾ 26.2
183.7 ⫾ 34.3
21.7 ⫾ 3.1
46.4 ⫾ 9.0
68.4
18.6
70.0
200.7 ⫾ 37.6
22.6 ⫾ 3.6
29.2 ⫾ 6.3
13.9
10.1
15.7
Each continuous variable is shown as mean value ⫾ SD.
Sasaki et al
Grip Strength and Mortality Prognosis
339
Analysis was carried out separately by sex. Deaths within
the first 2 years after baseline were excluded from analysis.
Secular trends of multivariate-adjusted RR of all causes
of death, except external causes, were calculated by dividing the follow-up period from baseline examination (within
5 years, after 5 years, after 10 years, after 15 years, and after
20 years).
RESULTS
Figure 1 Average grip strength (kg) by age- and sex-specific
categories.
ICD 8th or 9th), cerebral hemorrhage (CH, 431 by ICD 8th or
9th), and pneumonia (480-487 by ICD 8th or 9th). Pneumonia was chosen because it is one of the major causes of
death among elderly persons in Japan.
Statistical Analysis
Relative risk (RR) of mortality associated with grip strength
was adjusted multivariately for potentially confounding factors by Cox proportional hazard models, both in men and
women. The factors considered in the models were age at
baseline, systolic blood pressure, BMI, smoking and drinking habits, serum cholesterol level, and radiation dose. Allowance for the effect of radiation dose was necessary
because the study population included atomic bomb survivors exposed to radiation. When RR of all causes of mortality except external deaths was estimated according to the
sex- and age-specific quintile of grip strength, the third
quintile was established as a reference. For trend test, the
mean value of grip strength in each quintile was assigned to
the category. In the case of cause-specific mortalities, RR
was calculated for each 5-kg increment of grip strength.
Table 2
Average age at examination was 55.5 years in men and 53.9
years in women. Average grip strength was 46.4 kg in men
and 29.2 kg in women. According to average grip strength
by age- and sex-specific categories, a gradual decrease was
apparent in both sexes. Decrease of grip strength from 35-44
years to 65-74 years was 11.1 kg in men and 8.8 kg in
women (Figure 1). Other clinical characteristics are shown
in Table 1. Men tend to have higher systolic and diastolic
blood pressure, and are more likely than women to smoke
and to drink alcohol.
Over the 27 years of follow-up, 2483 deaths occurred
besides those from external causes, which included trauma
and suicide. Numbers of deaths from cancer, heart disease,
stroke, and pneumonia were 784, 518, 435, and 191, respectively (Table 2).
Multivariate-adjusted RR was compared among different
grip-strength categories divided into quintiles for each age
and sex group, using the third quintile group as reference. In
all age and sex groups, a declining trend of mortality by
increment of grip strength was observed. RR of mortality
for the highest quintile in men was 0.52 (95% confidence
interval [CI], 0.33-0.80) for the age group 35-54 years, 0.72
(95% CI, 0.53-0.98) for ages 55-64 years, and 0.67 (95%
CI, 0.49-0.91) for ages 65-74 years, which was significantly
lower than that of the reference group. RR of mortality for
the lowest quintile in men aged 55-64 years (RR 1.38, 95%
CI, 1.01-1.89) and aged 65-74 years (RR 1.38, 95% CI,
1.01-1.88) was significantly higher than that of the reference
group (Figure 2).
RR of mortality for the lowest quintile in women aged
35-54 years (RR 1.39, 95% CI, 1.02-1.90) and aged 65-74
Number of Deaths by Cause During Follow-up Period of 30 Years
Number of subjects at baseline
Mean (⫾ SD) age at baseline (years)
All death except external causes
Cancer
Heart
CHD
Stroke
CI
CH
Pneumonia
Men
Women
Total
1689
55.5 ⫾ 11.1
1081
391
185
87
172
74
41
82
3209
53.9 ⫾ 10.7
1402
393
333
122
263
105
55
109
4898
54.4 ⫾ 10.9
2483
784
518
209
435
179
96
191
CHD ⫽ coronary heart disease; CI ⫽ cerebral infarction; CH ⫽ cerebral hemorrhage.
340
The American Journal of Medicine, Vol 120, No 4, April 2007
Table 3 Relative Risk (RR) of Death for Each 5-kg
Increment of Grip in Men
All death
except
external
causes
Cancer
Heart
CHD
Stroke
CI
CH
Pneumonia
Age-adjusted RR
Multivariateadjusted RR
0.89
(0.86-0.93)
0.89
(0.86-0.92)
0.95
0.86
0.85
0.92
0.90
0.86
0.83
(0.90-1.01)
(0.79-0.94)
(0.75-0.96)
(0.84-1.00)
(0.79-1.03)
(0.73-1.03)
(0.73-0.94)
0.94
0.85
0.83
0.90
0.90
0.85
0.85
(0.88-1.00)
(0.79-0.93)
(0.74-0.94)
(0.83-0.99)
(0.79-0.96)
(0.72-1.01)
(0.75-0.98)
RR was adjusted for age, systolic blood pressure, body mass index,
total cholesterol, smoking habits, alcohol consumption, and radiation
dose. Figures in parentheses are 95% confidence intervals (CI).
CHD ⫽ coronary heart disease; CI ⫽ cerebral infarction; CH ⫽ cerebral
hemorrhage.
Figure 2 Multivariate-adjusted relative risk (RR) among different grip strength categories divided into quintiles in each age and
sex group, using the third quintile group as reference.
years (RR 1.54, 95% CI, 1.20-1.98) was significantly higher
than that of the reference group (Figure 2).
Age-adjusted RR of all causes of death, except external
causes, for each 5-kg increment of grip strength in men was
significantly low (RR 0.89, 95% CI, 0.86-0.92). Age-adjusted RR for heart disease, CHD, and pneumonia was 0.86
(95% CI, 0.79-0.94), 0.85 (95% CI, 0.75-0.96), and 0.83
(95% CI, 0.73-0.94), respectively. Multivariate-adjusted RR
of all causes of death, except external causes, was not
changed in age-adjusted values, although RR of stroke (RR
0.90, 95% CI, 0.83-0.99) and cerebral infarction (RR 0.90,
95% CI, 0.79-0.96) became statistically significant (Table 3).
Age-adjusted RR of all causes of death, except external
causes, for each 5-kg increment of grip strength in women
also was significantly low (RR 0.89, 95% CI, 0.84-93).
Age-adjusted RR for heart disease, stroke, and cerebral
infarction was 0.82 (95% CI, 0.75-0.91), 0.87 (95% CI,
0.78-0.97), and 0.81 (95% CI, 0.68-0.97), respectively.
Age-adjusted RR for CHD was not significant (RR 0.90,
95% CI, 0.76-1.06). Multivariate-adjusted RR was not significantly different in age-adjusted values (Table 4). Multivariate-adjusted RR of all causes of death, except external
causes, for each 5-kg increment of grip strength in women
(RR 0.87, 95% CI, 0.83-0.92) was not very different from
that observed in men.
Secular trends of multivariate-adjusted RR of all causes
of death, except external causes, were analyzed. RR tended
to move to unity through the follow-up period, from 0.80
(95% CI, 0.74-0.87), during the initial 5 years after baseline
examination, to 0.92 (95% CI, 0.87-0.96), at more than 20
years of follow-up. However, even after 20 years of followup, RR was significantly lower for every 5-kg increment of
grip strength (Table 5).
DISCUSSION
Grip strength was a strong and consistent predictor of all
causes of mortality in middle-aged and elderly persons in
Table 4 Relative Risk (RR) of Death for Each 5-kg
Increment of Grip in Women
All death
except
external
causes
Cancer
Heart
CHD
Stroke
CI
CH
Pneumonia
Age-adjusted RR
Multivariateadjusted RR
0.89
(0.84-0.93)
0.87
(0.83-0.92)
1.00
0.82
0.90
0.87
0.81
0.83
0.86
(0.91-1.09)
(0.75-0.91)
(0.76-1.06)
(0.78-0.97)
(0.68-0.97)
(0.66-1.05)
(0.73-1.03)
0.99
0.80
0.88
0.85
0.80
0.80
0.87
(0.90-1.08)
(0.72-0.88)
(0.75-1.04)
(0.76-0.95)
(0.68-0.96)
(0.63-1.02)
(0.73-1.04)
RR was adjusted for age, systolic blood pressure, body mass index,
total cholesterol, smoking habits, alcohol consumption, and radiation
dose. Figures in parentheses are 95% confidence intervals (CI).
CHD ⫽ coronary heart disease; CI ⫽ cerebral infarction; CH ⫽ cerebral
hemorrhage.
Sasaki et al
Grip Strength and Mortality Prognosis
Table 5 Secular Trends of Multivariate-Adjusted Relative
Risk (RR) of all Death Except External Causes
Follow-up Period
RR
95% CI
ⱕ5 years since baseline
⬎5 years
⬎10 years
⬎15 years
⬎20 years
0.80
0.90
0.91
0.92
0.92
(0.74-0.87)
(0.87-0.93)
(0.88-0.95)
(0.88-0.96)
(0.87-0.96)
RR was estimated for each 5-kg increment of grip strength.
CI ⫽ confidence interval.
Japan at more than 20 years of follow-up from baseline.
This association was observed in all age categories from 35
to 74 years and in both sexes. It predicted not only all causes
of mortality, but also heart disease and stroke mortality.
Grip strength may be representative of overall muscular
strength because it is highly correlated with other muscular
strength measures, including elbow flexion, knee extension,
trunk flexion, and trunk extension.17 Maintenance of muscle
strength throughout life reduces the prevalence of functional
limitations that might closely relate to early death, especially in the elderly.18-21
Muscular strength is related to skeletal muscle mass,
which also is a significant site of glucose disposal and
insulin action.22 Elevated insulin levels in persons with
lower grip strength are thought to be the central feature of
insulin-resistance syndrome.23 Muscle is also the greatest
reserve of protein in the body and plays an important role in
immunity and other functions. People with poor muscular
strength, therefore, may be more prone to injurious accidents, and their recovery from acute disease or injury may
be compromised.1 Thus, measures of muscular strength can
be considered markers of the risk of metabolic disorders,
which are ultimately related to risk of premature mortality,
especially from cardiovascular diseases. Because muscle
strength is closely related to regular physical activity, regular daily exercise to maintain muscle strength is
recommended.
Most early studies analyzed data from hospitalized patients or the elderly population.1,3,5,6,12 The present analysis
concludes that the protective effects of higher grip strength
were apparent both in men and women, and in a wide-range
of age distribution. The results also show that such protective effects were observed even in nonhospitalized and
middle-aged individuals, which demonstrates the importance of maintenance of muscle strength in the middle-age
period for better health in later life.
Differences in muscle strength and body composition are
known to exist between men and women, and from early
adulthood on, women have, on average, 30% to 40% less
muscle strength than men.23 Several studies investigated the
grip-mortality association in women, 2 of which studied
elderly women aged 65 years or older, and found an inverse
relationship.1-3 Katzmarzyk and Craig failed to show that
grip strength was predictive of mortality in middle-aged
341
women, although there was a 49% increased risk of death in
the lowest quartile of grip strength in men.2 To the best of
our knowledge, our study is the first to show a significant
grip-mortality association in middle-aged women. However, it is noteworthy that there was a difference in relationship between grip strength and mortality between the
sexes. In men, both ends of grip strength distribution are
associated with mortality; in women, however, low grip
strength is a risk for premature death and high grip strength
is not protective. These differences may be due to the lower
and narrower distribution of grip strength in women compared with the distribution in men. Age-adjusted and multivariate-adjusted RR for each 5-kg increment of grip
strength for CHD was significantly low in men, but not in
women. This may be due to the relatively small number of
female CHD cases, because hypertensive heart disease is
more prevalent than CHD in Japan, especially in women.
The impact of grip strength on mortality prognosis persisted for more than 20 years. Grip strength is shown to be
a tracking phenomenon in which the measure remains consistent even if physical activity varies during the observation period.24 Therefore, the lower RR of premature death
for people with higher grip strength is stable for long observational periods.
The major limitation of the study is that many factors
that might affect the observed relationship between grip
strength and mortality, such as presence of diabetes, income, education, physical activity, and so on were not
included in the analysis. Therefore, the present findings
could be subject to residual confounding.
Half of the subjects in this study were exposed to A-bomb
radiation. However, average grip strength among radiationdose groups did not differ (data not shown), and association
between grip strength and mortality was independent from
radiation dose (P ⬎.05). Thus, the grip-mortality association
can be generalized to other human populations.
CONCLUSION
Grip strength is a strong and consistent predictor of all
causes of mortality in middle-aged and elderly persons,
even at more than 20 years of follow-up from baseline. In
accordance with the present findings, regular exercise for
improvement of physical and musculoskeletal fitness is eagerly recommended to improve prognosis.
ACKNOWLEDGMENT
This publication is based on research performed at the
Radiation Effects Research Foundation (RERF), Hiroshima
and Nagasaki, Japan. RERF is a private nonprofit foundation funded equally by the Japanese Ministry of Health,
Labour and Welfare (MHLW) and the US Department of
Energy (DOE), the latter through the National Academy of
Sciences. This publication was supported by RERF Research Protocol RP2-75.
342
References
1. Rantanen T, Volpato S, Ferrucci L, et al. Handgrip strength and
cause-specific and total mortality in older disabled women: exploring
the mechanism. J Am Geriatr Soc. 2003;51:636-641.
2. Katzmarzyk PT, Craig CL. Musculoskeletal fitness and risk of mortality. Med Sci Sports Exerc. 2002;34:740-744.
3. Snib SA, Markides KS, Ray L, et al. Handgrip strength and mortality
in older Mexican Americans. J Am Geriatr Soc. 2002;50:1250-1256.
4. Rantanen T, Guralnik JM, Foley D, et al. Midlife handgrip strength as
a predictor of old age disability. JAMA. 1999;281:558-560.
5. Humphreys J, de la Maza P, Hirsch S, et al. Muscle strength as a
predictor of loss of functional status in hospitalized patients. Nutrition.
2002;18:616-620.
6. Rantanen T, Avlund K, Suominen H, et al. Muscle strength as a
predictor of onset of ADL dependence in people aged 75 years. Aging
Clin Exp Res. 2002;14(3 suppl):10-15.
7. Davis JW, Ross PD, Preston SD, et al. Strength, physical activity, and
body mass index: relationship to performance-based measures and
activities of daily living among older Japanese women in Hawaii. J Am
Geriatr Soc. 1998;46:274-279.
8. Rantanen T. Muscle strength, disability and mortality. Scand J Med Sci
Sports. 2003;13:3-8.
9. Rantanen T, Harris T, Leveille SG, et al. Muscle strength and body
mass index as long-term predictors of mortality in initially healthy
men. J Gerontol A Biol Sci Med Sci. 2000;55:M168-M173.
10. Fujita Y, Nakamura Y, Hiraoka J, et al. Physical-strength tests and
mortality among visitors to health promotion centers in Japan. J Clin
Epidemiol. 1995;48:1349-1359.
11. Phillips P. Grip strength, mental performance and nutritional status as
indicators of mortality risk among female geriatric patients. Age Aging.
1986;15:53-56.
12. Bohannon RW. Dynamometer measurements of handgrip strength
predict multiple outcomes. Percept Mot Skills. 2001;93:323-328.
The American Journal of Medicine, Vol 120, No 4, April 2007
13. Rantanen T, Masaki K, Izmirlian G, et al. Grip strength changes
over 27 yr in Japanese-American men. J Appl Physiol. 1998;85:
2047-2053.
14. Belsky JL, Moriyama IM, Fujita S, et al. Aging studies in atomic bomb
survivors. RERF Technical Report. 11-78, 1978.
15. Sawada H, Kodama K, Shimizu Y, Kato H. Adult Health Study Report
6: Results of six examination cycles, 1968-80, Hiroshima and Nagasaki. RERF Technical Report. 3-86, 1986.
16. Roesch WC, Ed. Reassessment of Atomic Bomb Radiation Dosimetry
in Hiroshima and Nagasaki. Final Report. Hiroshima, Japan: Radiation Effects Research Foundation, 1987.
17. Rantanen T, Era P, Kauppinen M, et al. Maximal isometric muscle
strength and socio-economic status, health and physical activity in
75-year-old persons. J Aging Phys Act. 1994;2:206-220.
18. Metter EJ, Talbot LA, Schrager M, et al. Skeletal muscle strength as
a predictor of all-cause mortality in healthy men. J Gerontol A Biol Sci
Med Sci. 2002;57:B359-B365.
19. Morey MC, Pieper CF, Cornoni-Huntley J. Physical fitness and functional limitations in community-dwelling older adults. Med Sci Sports
Exerc. 1998;30:715-723.
20. Huang Y, Macera CA, Blair SN, et al. Physical fitness, physical
activity, and functional limitation in adults aged 40 and older. Med Sci
Sports Exerc. 1998;30:1430-1435.
21. Brill PA, Macera CA, Davis DR, et al. Muscular strength and physical
function. Med Sci Sports Exerc. 2000;32:412-416.
22. Lazarus R, Sparrow D, Weiss ST. Handgrip strength and insulin
levels: cross-sectional and prospective associations in the Normative
Aging Study. Metabolism. 1997;46:1266-1269.
23. Lindle RS, Metter EJ, Lynch NA, et al. Age and gender comparisons
of muscle strength in 654 women and men aged 20-93 yr. J Appl
Physiol. 1997;83:1581-1587.
24. Fortier MD, Katzmarzyk PT, Malina RM, et al. Seven-year stability of
physical activity and musculoskeletal fitness in the Canadian population. Med Sci Sports Exerc. 2001;33:1905-1911.
The American Journal of Medicine (2007) 120, 343-349
CLINICAL RESEARCH STUDY
Thyroid Hormone Use, Hyperthyroidism and Mortality in
Older Women
Douglas C. Bauer, MD,a,b Nicolas Rodondi, MD, MAS,b,c Katie L. Stone, PhD,a Teresa A. Hillier, MD, MS,d for the Study
of Osteoporotic Fractures Research Group: Universities of California (San Francisco), Pittsburgh, Minnesota
(Minneapolis), and Kaiser Permanente Center for Health Research, Portland
a
Division of General Internal Medicine, University of California, San Francisco; bDepartment of Epidemiology and Biostatistics,
University of California, San Francisco; cUniversity Outpatient Clinic, Department of Community Medicine and Public Health,
University of Lausanne, Switzerland; dDepartment of Endocrinology, Center for Health Research Northwest/Hawaii, Kaiser
Permanente Northwest, Portland, Ore.
ABSTRACT
PURPOSE: Thyroid dysfunction is common, particularly among older women. The safety of thyroid
hormone use and long-term prognosis of hyperthyroidism remain controversial. We performed a prospective cohort study to examine the relationship among thyroid hormone use, previous hyperthyroidism,
abnormal thyroid function, and mortality.
METHODS: We studied 9449 community-dwelling white women aged ⱖ65 years followed for 12 years. For
analyses of thyroid function, we performed a nested case-cohort in 487 women using a third-generation thyroidstimulating hormone assay. Causes of death were adjudicated based on death certificates and hospital records.
RESULTS: Twelve percent of the 9449 women took thyroid hormone at baseline, and the mean duration
of thyroid hormone use was 15.8 years; 9.4% of participants reported a history of hyperthyroidism. During
12 years of follow-up, 3159 women died (33%). In multivariate analysis, mortality among users of thyroid
hormone was similar to that observed for nonusers (relative hazard [RH] 1.11, 95% confidence interval
[CI], 0.98-1.24, P ⫽ .09). Previous hyperthyroidism was associated with a higher risk of all-cause mortality
(RH 1.20, 95% CI, 1.06-1.36), particularly cardiovascular mortality (RH 1.46, 95% CI, 1.20-1.77). Low
(ⱕ0.5 mU/L) or high (⬎5 mU/L) thyroid-stimulating hormone levels were not associated with excess total
or cause-specific mortality, but the power to detect these relationships was limited.
CONCLUSIONS: Among older women, thyroid hormone use is not associated significantly with excess
mortality, but previous hyperthyroidism may be associated with a small increase in all-cause and cardiovascular mortality. Additional long-term studies of hyperthyroidism and its treatment should further
explore these findings. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Thyroid diseases; Thyroid hormones; Thyrotropin; Mortality; Cardiovascular disease; Cohort study
Thyroid dysfunction is common, particularly among older
women and other high-risk groups.1 Prevalence estimates of
thyroid dysfunction (both hyper- and hypothyroidism) in
This work was supported by Public Service Grants K08 AG00629
(NIA), and AG05407, AR35582, AG05394, AR35584, AR35583. Dr.
Rodondi was supported by a grant from the Swiss National Foundation
(PBLAB-102353).
Requests for reprints should be addressed to Douglas C. Bauer, MD,
University of California, San Francisco 185 Berry, #5700, San Francisco,
CA 94107.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.04.034
older women range from 5%-10%,2,3 and 10% of women
over age 65 years in the US take thyroid hormone,4 a large
proportion with biochemical evidence of over- or underreplacement.5-7
The clinical consequences of hypothyroidism or hyperthyroidism have been primarily studied in those with overt
disease (symptomatic). Individuals with untreated overt thyroid dysfunction have a reduced life expectancy.8,9 Mortality rates among women with subclinical thyroid dysfunction
have not been well studied, but even mild forms of thyroid
disease may be associated with greater mortality for several
344
The American Journal of Medicine, Vol 120, No 4, April 2007
reasons. First, high thyroid-stimulating hormone is associFor the analyses of thyroid-stimulating hormone and
ated with hyperlipidemia,6,10,11 which could promote coromortality we used a case-cohort design25 using a random
nary disease. Second, abnormalities of cardiac contractility
sample of 487 women nested in the entire cohort of 9449
and arrhythmias, particularly atrial fibrillation,12 have been
women. Our analysis set for thyroid-stimulating hormone
described among individuals with low thyroid-stimulating
measurements consisted of 147 death (cases) and 340 surhormone, but the clinical signifivivors (controls). From the 1123
cance of these findings is uncerdeaths that had occurred as of
tain.13 Finally, low thyroid-stimuMarch 1995 (mean follow-up ⫽ 6.7
CLINICAL SIGNIFICANCE
lating hormone is associated with
years), we randomly selected 92
an increased risk of fracture in
cases. In addition, we randomly
● Long-term use of thyroid hormone did
older women, and hip fractures in
identified 400 women from the ennot affect all-cause, cardiovascular, or
particular are associated with extire cohort, of whom 58 had died as
cancer mortality.
cess mortality.4 The treatment of
of March 1995 and were analyzed
● Previous hyperthyroidism may be associhyperthyroidism also might be harmas cases, as recommended for this
ful, as increased mortality among indesign.25 We excluded 2 survivors
ated with an increase in mortality.
dividuals with hyperthyroidism
and 3 cases because of missing
treated with radioiodine or anti-thybaseline data.
roid drugs has been reported.14,15
Measurements
The relationship between thyroid hormone use and morParticipants were interviewed and examined at one of the
tality is also unclear, but a number of older studies have
clinical centers during the baseline visit in 1986-1988. At
linked thyroid hormone use to a variety of adverse outthat visit, detailed data about physician-diagnosed medical
comes, including cancer16 and coronary disease.17 It should
conditions and past medication use (before baseline) were
be noted that the use of nonsynthetic thyroid hormone
collected, and baseline medication use at the time of the
preparations was common in these older studies, and their
visit, including thyroid hormone use, was confirmed by
findings have not been confirmed.18,19 Studies of the crossexamination of pill bottles by trained interviewers. Particisectional20,21 and longitudinal22 relationship between abnorpants were asked specifically about previous diagnoses of
mal thyroid-stimulating hormone levels and cardiovascular
hyperthyroidism or Graves’ disease ever in their life. Infordisease are similarly conflicting. A recent population-based
mation on prior treatment for hyperthyroidism was not
study found that an elevated thyroid-stimulating hormone
collected.
with normal thyroid hormone levels (subclinical hypothyroidism) was associated with a greater prevalence of aortic
Ascertainment of Deaths. Women were contacted by postatherosclerosis.23
card every 4 months. If a participant died, copies of the
To further examine the relationship among use of thyroid
official death certificate and hospital records, if available,
hormone, history of hyperthyroidism, abnormal thyroid
were collected. The cause of death was then adjudicated by
function, and mortality in older women, we performed a
a physician-investigator at the Coordinating Center (Uniprospective analysis in the Study of Osteoporotic Fractures
versity of California, San Francisco) who was unaware of
examining the risk of mortality among those with and withthe participant’s thyroid status or other predictor variout thyroid conditions.
ables. Causes of death were assigned by ICD-9 codes.
Cardiovascular causes of death included codes 390-459,
METHODS
while cancer causes of death included ICD-9 codes 140Subjects and Design
240. All other codes were classified as “other.”
Subjects in this prospective study were participants in the
Thyroid-Stimulating Hormone Assays. Sera were colStudy of Osteoporotic Fractures (SOF), a population-based
lected from each participant at the baseline visit and stored
cohort study begun in 1986-1988 of 9704 communityat ⫺190°C. In 1995, thyroid-stimulating hormone assays
dwelling white women aged 65 years and over at entry.
were performed on thawed baseline sera from 2 groups:
These women were recruited for SOF from populationbased listings at 4 clinical centers irrespective of osteopowomen who died during follow-up (n ⫽ 147) and women
rosis status (Portland, Ore; Minneapolis, Minn; Baltimore,
randomly selected from the cohort (n ⫽ 340). Thyroid-stimMd; and the Monongahela Valley near Pittsburgh, Penn).
ulating hormone is highly stable in frozen sera over proOnly women with prior bilateral hip replacement were exlonged periods.5,26
cluded. The SOF study has been described in detail elseThyroid-stimulating hormone levels were measured uswhere.24 Our overall design was a prospective cohort study
ing a highly sensitive, third-generation chemiluminescent
to assess the relationship among history of hyperthyroidism,
assay (Endocrine Science, Calabasas, CA). The normal
thyroid hormone use, and mortality. These analyses inrange for this assay is 0.5-5.5 mU/L; the minimum detection
cluded all women with complete baseline data who were not
level is 0.01 mU/L. At thyroid-stimulating hormone conlost to follow-up (99% of the cohort).
centrations of 0.5 mU/L, the intra- and interassay coeffi-
Bauer et al
Thyroid Disease and Mortality
cients of variation (CV) are 4.7% and 6.3%, respectively.
Thirty randomly selected specimens were blindly submitted
for duplicate analysis. The correlation between these 2 thyroid-stimulating hormone results was high (r ⫽ 0.95).
All hormone analyses were performed without knowledge of the participant’s vital status or thyroid history.
Analysis
Differences in baseline characteristics were tested with t test
or chi-squared test. Thyroid-stimulating hormone levels
were then categorized as low (ⱕ0.5 mU/L), normal (0.5-5.5
mU/L), or high (⬎5.5 mU/L) based on the normal range
provided by the manufacturer. Additional analyses were
performed using a low thyroid-stimulating hormone cutpoint of ⱕ0.1 mU/L, which is highly correlated with a
suppressed response to thyroid-stimulating hormone stimulation and is the best indicator of thyroid excess among
community-dwelling adults.27
The relationships among hyperthyroidism, thyroid hormone use, and mortality were analyzed with Cox proportional hazards models (SAS Institute, Cary, NC). The
subcohort analyses examining the relationship between thyroid-stimulating hormone levels and mortality utilized proportional hazards models that take account of the casecohort sampling design (Epicure, Hirosoft International,
Seattle WA).25
Multivariate models were constructed to account for potential confounders that were selected on the basis of either
biologic plausibility (eg, hypertension, diabetes) or strong
association (P ⬍.05) with hyperthyroidism or thyroid-stimulating hormone level (eg, age). Results are reported as
relative hazards (RH), with 95% confidence intervals (CI).
RESULTS
During a mean follow-up of 11.9 years, with ascertainment
of vital status 99% complete, 3159 (33%) women died. The
most common causes of death were cardiovascular
(n ⫽ 1210) and cancer (n ⫽ 818).
Baseline Characteristics
At the baseline visit, 12% of the 9449 women took thyroid
hormone, and 9.4% reported previous hyperthyroidism (Table 1). Compared with women without previous hyperthyroidism, women who reported hyperthyroidism were older,
less likely to report good or excellent health, less likely to
drink alcohol, more likely to be hypertensive, and more
likely to use estrogen and thyroid hormone.
The baseline characteristics of the subcohort with thyroid-stimulating hormone measurements, including the 147
women who died and 340 who were alive as of March 1995,
are shown in Table 2. Compared with the surviving women,
those who died were older, lighter, less active, reported
poorer health status, and more likely to be hypertensive.
345
Table 1 Previous Hyperthyroidism and Mortality: Baseline
Characteristics of Entire Cohort
Variable
Age (years ⫾ SD)
Weight (kg ⫾ SD)
Self-reported health
status (good or
excellent, %)
Baseline cigarette
use (%)
Baseline alcohol
use (%)
Physical activity
(kcal/years)
Diabetes (%)
Hypertension (%)
Baseline oral
estrogen
use (%)
Thyroid hormone
use (%)
At baseline
Before baseline
History of
Hyperthyroidism
(n ⫽ 891)
No History of
Hyperthyroidism
(n ⫽ 8558)
72.3 ⫾ 5.6†
66.7 ⫾ 11.8
76.3†
71.6 ⫾ 5.3
67.1 ⫾ 12.6
84.3
10.4
10.1
67.0*
70.1
1570 ⫾ 1716*
1612 ⫾ 1671
8.0
42.6†
15.5*
6.7
37.9
13.7
48.7†
20.6†
8.1
5.9
*P ⬍.05 compared with women without hyperthyroidism.
†P ⬍.01 compared with women without hyperthyroidism.
Previous Hyperthyroidism and Mortality
In age-adjusted analyses in the entire cohort, previous hyperthyroidism was associated with an increased risk of total
mortality (RH 1.29, 95% CI, 1.16-1.44). After further adjustment for thyroid hormone use, weight, health status,
alcohol and tobacco use, physical activity, diabetes, hypertension, and baseline use of estrogen and thiazides, mortality was 20% higher (95% CI, 6%-36%) among women
reporting previous hyperthyroidism (Table 3). In multivariate analyses, previous hyperthyroidism was associated with
increased cardiovascular mortality (RH 1.46, 95% CI, 1.201.77). The associations between hyperthyroidism and cancer or other causes of death were not statistically significant
(Table 3).
The relationship between previous hyperthyroidism and
mortality were similar among the smaller subset of women
with thyroid-stimulating hormone measurements. After adjustment for thyroid-stimulating hormone levels and other
factors, total mortality was still elevated among women
with previous hyperthyroidism (RH 1.36), but without
reaching statistical significance (95% CI, 0.64-2.93), perhaps as a result of the small sample size. Likewise, the risk
of cardiovascular mortality remained elevated among
women reporting previous hyperthyroidism, without being
statistically significant (RH 2.06, 95% CI, 0.72-5.94).
Thyroid Hormone Use and Mortality
At the baseline visit, 12% of women were thyroid hormone
users with a mean duration use of 15.8 years. In age-
346
The American Journal of Medicine, Vol 120, No 4, April 2007
Table 2 Thyroid-Stimulating Hormone and Mortality: Baseline Characteristics in Subcohort with Thyroid-Stimulating
Hormone Measurements
Variable
Women Who Were Alive
at the End of
Follow-up (n ⫽ 340)
Women Who Died During
Follow-up (n ⫽ 147)
Age (years ⫾ SD)
Weight (kg ⫾ SD)
Self-reported health status (good or excellent, %)
Baseline cigarette use (%)
Baseline alcohol use (%)
Physical activity (kcal/years ⫾ SD)
Diabetes (%)
Hypertension (%)
Baseline oral estrogen use (%)
History of hyperthyroidism (%)
Baseline thyroid hormone use (%)
Duration of thyroid hormone use among users at baseline (years ⫾ SD)
Thyroid-stimulating hormone (mU/L ⫾ SD)
Low thyroid-stimulating hormone (ⱕ0.5 mU/L, %)
High thyroid-stimulating hormone (⬎ 5.5 mU/L, %)
71.5 ⫾ 5.0
67.6 ⫾ 12.2
85.0
8.6
68.5
1570 ⫾ 1631
8.3
36.5
9.8
8.8
11.0
14.8 ⫾ 15.3
2.4 ⫾ 3.4
10.6
7.1
75.0 ⫾ 5.5†
64.8 ⫾ 14.0*
72.1†
13.6
65.3
1096 ⫾ 1466†
12.2
49.0†
9.7
10.5
10.3
19.2 ⫾ 19.9
2.3 ⫾ 3.7
11.6
8.2
*P ⬍.05 compared with women who were alive at the end of follow-up.
†P ⬍.01 compared with women who were alive at the end of follow-up.
adjusted analyses, thyroid hormone use at baseline was
associated with total mortality (RH 1.20, 95% CI, 1.091.34). However, in multivariate analysis, mortality among
baseline users of thyroid hormone was similar to that observed for nonusers (RH 1.11, 95% CI, 0.98-1.24, P ⫽ .09)
(Table 3). Similar analyses of cause-specific death also
revealed no statistically significant associations between
thyroid hormone use and cardiovascular or cancer mortality.
Thyroid hormone use before baseline was not associated
with total or cause-specific mortality.
Thyroid-Stimulating Hormone Level and Death
In the subcohort, we found no significant associations among
low, normal, or high thyroid-stimulating hormone values and
Table 3
total or cause-specific mortality (Table 4). For example, in
multivariate analysis, compared with women with normal thyroid-stimulating hormone (0.5-5.5 mU/L), the relative hazard
of any death was 0.86 (95% CI, 0.39-1.87) among women with
low thyroid-stimulating hormone (ⱕ0.5 mU/L) and 1.23 (95%
CI, 0.55-2.74) among women with high thyroid-stimulating
hormone. Results were qualitatively similar among subjects
with thyroid-stimulating hormone levels ⱕ0.1 mU/L (total
mortality RH 0.63, 95% CI, 0.23-1.73). Excluding subjects
with thyroid-stimulating hormone ⱕ0.5 mU/L who used thyroid hormone (86%) yielded similar results. When examined
as a continuous variable, neither thyroid-stimulating hormone
nor log thyroid-stimulating hormone was related to subsequent
mortality.
History of Hyperthyroidism, Thyroid Hormone Use and Mortality: Multivariate Results among Entire Cohort*
Variable (unit)
History of hyperthyroidism (yes)
Thyroid hormone use (yes)
At baseline
Before baseline
Age (SD)
Weight (SD)
Self-reported health status (fair or poor)
Baseline cigarette use (yes)
Baseline alcohol use (yes)
Physical activity (SD)
Diabetes (yes)
Hypertension (yes)
Baseline oral estrogen use (yes)
Baseline thiazide use (yes)
All-cause Mortality
RH (95% CI)
CV Deaths
RH (95% CI)
Cancer Deaths
RH (95% CI)
Other Deaths
RH (95% CI)
1.20 (1.06-1.36)
1.46 (1.20-1.77)
1.08 (0.83-1.41)
1.13 (0.91-1.40)
1.11
0.99
1.71
0.93
1.57
2.03
0.83
0.92
1.78
1.45
0.88
1.45
1.11
0.81
2.03
0.93
1.77
1.89
0.80
0.87
2.29
1.68
0.77
1.00
1.19
1.12
1.53
1.02
1.51
2.68
0.86
0.96
1.34
1.19
0.82
0.99
1.13
1.05
2.05
0.83
1.67
2.24
0.72
0.91
2.10
1.68
1.02
0.78
(0.98-1.24)
(0.86-1.15)
(1.65-1.76)
(0.89-0.97)
(1.44-1.72)
(1.82-2.26)
(0.77-0.89)
(0.88-0.96)
(1.58-2.00)
(1.31-1.61)
(0.78-0.99)
(1.31-1.61)
*The multivariate models include all covariates listed.
†RH ⫽ Relative hazards for death (95% confidence interval [CI]).
(0.92-1.34)
(0.63-1.04)
(1.93-2.14)
(0.87-0.99)
(1.54-2.04)
(1.55-2.29)
(0.70-0.90)
(0.82-0.95)
(1.92-2.74)
(1.42-2.00)
(0.63-0.94)
(0.84-1.20)
(0.94-1.50)
(0.85-1.46)
(1.42-1.64)
(0.95-1.10)
(1.26-1.82)
(2.22-3.24)
(0.74-1.01)
(0.89-1.04)
(1.02-1.76)
(0.95-1.49)
(0.66-1.03)
(0.77-1.26)
(0.93-1.37)
(0.83-1.34)
(1.94-2.16)
(0.76-0.89)
(1.44-1.94)
(1.86-2.69)
(0.64-0.82)
(0.84-0.97)
(1.73-2.56)
(1.42-1.98)
(0.85-1.23)
(0.65-0.94)
Bauer et al
Thyroid Disease and Mortality
347
Table 4 Thyroid-Stimulating Hormone Level and Risk of Death: Multivariate Results among Subcohort with Thyroid-Stimulating
Hormone Measurements
Relative Hazards for Death (95% CI)*
Thyroid-stimulating hormone
level (mU/L)
ⱕ0.5
⬎0.5-5.5
⬎ 5.5
History of hyperthyroidism
Baseline thyroid hormone use
Total Mortality (n ⫽ 147)
CV Deaths (n ⫽ 54)
Cancer Deaths (n ⫽ 55)
Other Deaths (n ⫽ 39)
0.86
1.0
1.23
1.36
0.69
0.91
1.0
0.92
2.06
0.64
1.16
1.0
1.17
1.23
0.60
0.52
1.0
1.91
1.05
1.04
(0.39-1.87)
(0.55-2.74)
(0.64-2.93)
(0.29-1.66)
(0.28-2.95)
(0.25-3.34)
(0.72-5.94)
(0.16-2.59)
(0.40-3.40)
(0.38-3.60)
(0.42-3.63)
(0.17-2.10)
(0.10-2.63)
(0.55-6.60)
(0.25-4.43)
(0.21-5.25)
*Adjusted for variables shown plus age, weight, and baseline use of estrogen.
Although limited by the small number of deaths, analyses of cause-specific death revealed no significant associations between thyroid-stimulating hormone level and mortality. There were too few breast cancer deaths among the
women with thyroid-stimulating hormone measurements
(n ⫽ 8) to examine the relationship between thyroid-stimulating hormone level and breast cancer. There was a trend
towards increased noncardiovascular noncancer mortality
among women with high thyroid-stimulating hormone (RH
1.91), but this relationship did not reach statistical significance (95% CI, 0.55-6.60).
DISCUSSION
In this large prospective study of community-dwelling older
women, we found that prior hyperthyroidism (or its treatment) was associated with a small increased risk of total
mortality, particularly cardiovascular mortality. Conversely,
long-term users and nonusers of thyroid hormone had similar rates of all-cause, cardiovascular, and cancer mortality.
Among women with baseline thyroid-stimulating hormone
measurements, levels above or below the normal range were
not associated with total or cause-specific mortality.
These data provide reassuring evidence that the large
number of postmenopausal women taking thyroid hormone
do not face excess risk of mortality. Previous clinical studies
of the association between thyroid hormone use and morbidity and mortality have been conflicting. For example, one
prospective cohort study of 1180 individuals on thyroxine
(90% women) found an increased incidence of coronary
disease hospital admissions over 5 years compared with the
general population.17 However, this risk was limited to
those aged younger than 65 years. Another study of 1462
women found no evidence of an increased risk of coronary
disease among thyroid hormone users, but this study had
little power to detect mortality relationships, as all of the 22
long-term thyroid hormone users were alive 12 years later.18
Our data with a large number of thyroid hormone users and
long-term prospective follow-up provide further evidence
that thyroid hormone use is not associated with an increased
risk of cardiovascular or cancer mortality.
Although there have been numerous reports that subtle
abnormalities of thyroid function are not physiologic and
may result in cardiac,12 skeletal,4 and neuropsychiatric abnormalities,28-30 but with limited and conflicting data,1,13
we found no evidence that abnormal thyroid-stimulating
hormone was associated with excess all-cause mortality.
Older pathological studies have suggested that thyroid disease, particularly autoimmune thyroiditis (Hashimoto’s),
may be associated with an increased prevalence of cardiovascular disease.31,32 Possible mechanisms postulated to
account for the association between thyroid autoimmunity
and coronary disease include both direct effects on coronary
vasculature33 and indirect effects on serum lipids. However,
previous clinical studies of the association between thyroid
disease and cardiovascular disease have been conflicting.13
In a study of 945 consecutive hospital admissions, Tieche
et al found that the prevalence of coronary disease was
significantly higher among adults with elevated thyroidstimulating hormone (ⱖ4 mU/L) compared with agematched controls with normal thyroid-stimulating hormone.21 Another large study from Whickham, England
found a weak association between elevated thyroid-stimulating hormone and minor (but not major) EKG abnormalities in women,20 but there was no association with coronary
disease during a reanalysis of the same cohort after 20 years
of follow-up.22 Consistent with our results, the Whickham
study also found no evidence that women with elevated
thyroid-stimulating hormone or antithyroid antibodies experienced greater total mortality (RR 0.90, 95% CI,
0.65-1.22).
Other studies have found similar associations between
hyperthyroidism and increased cardiovascular risk. One
population-based study of 1191 individuals found an association between low thyroid-stimulating hormone and mortality, largely accounted for by an increase in cardiovascular
disease.34 Another study of 1049 patients admitted from an
emergency department found that compared with those with
normal T3 values, patients with high serum T3 concentrations at the time of admission were more likely to have
angina pectoris or myocardial infarction (OR 2.6) and were
much more likely to develop subsequent cardiovascular
348
events during a 3-year follow-up.35 Moreover, hyperthyroidism has been associated with an abnormal left ventricular function36 and atrial fibrillation. For example, in the
Framingham cohort, individuals with low thyroid-stimulating hormone were 3-fold more likely to develop atrial fibrillation during follow-up,7 which may lead to death by
heart failure or stroke. Lastly, it has been postulated that
treatment for hyperthyroidism itself may be associated with
adverse cardiovascular outcomes. One population-based
study found an unexpectedly high total and cardiovascular
mortality rate among 7209 subjects with hyperthyroidism
treated with radioactive iodine.14 However, this study did
not clarify whether this increased mortality was attributable
to radioiodine or hyperthyroidism per se, as it only compared these rates of death to the general population. It has
not been reported that antithyroid medications might be
associated with increased cardiovascular disease or cardiovascular mortality.37,38 Similar to these previous studies, we
were not able to test these hypotheses, as we did not collect
information on specific treatments for hyperthyroidism, but
our results could indicate that hyperthyroidism itself and not
its treatment is responsible. Further large studies are needed
to clarify whether hyperthyroidism itself or its treatment is
associated with cardiovascular mortality.
Some, but not all, previous studies have suggested that
thyroid disease is associated with malignancy, particularly
breast cancer. In the study by Petersen et al noted above, 3%
of thyroid hormone users developed cancer compared with
5.3% of nonusers, but there was no adjustment for age or
other potential confounders.18 More recently, several studies have demonstrated an association between thyroid size
or autoimmunity and breast cancer.39,40
Despite its relatively large size and high follow-up rates,
our study did have several limitations. Our assessment of
hyperthyroidism was by self-report, and previous studies
have found that hyperthyroidism was over-reported by older
women.41 Misclassification of thyroid conditions, such as
inability to report a specific thyroid condition, may have
biased our results, but such misclassification would be expected to underestimate the effect of any observed associations. We did confirm all self-reported thyroid hormone
use, but were unable to analyze specific thyroid hormone
preparations. We did not measure T4 levels in this subcohort and so were unable to differentiate abnormal thyroidstimulating hormone related to overt or subclinical thyroid
disease. Furthermore, we did not measure thyroid autoantibodies and therefore may have overlooked more specific
associations with pure thyroid autoimmunity. In the subcohort of 487 women, we had limited power to examine the
relationship between thyroid-stimulating hormone levels
and mortality, and were also not able to examine in detail
the relationship between thyroid-stimulating hormone levels
and cause-specific mortality because of the small number of
each specific cause of death.
In summary, in this large population-based prospective
study, we found that previous hyperthyroidism was associ-
The American Journal of Medicine, Vol 120, No 4, April 2007
ated with a small increase in mortality, particularly due to
cardiovascular disease. Low or high thyroid-stimulating
hormone was not associated with mortality in this cohort.
Long-term users and nonusers of thyroid hormone had similar rates of all-cause, cardiovascular, and cancer mortality.
Additional long-term studies of hyperthyroidism and its
treatment are needed to confirm and further explore these
findings.
References
1. Helfand M. Screening for subclinical thyroid dysfunction in nonpregnant adults: a summary of the evidence for the US Preventive Services
Task Force. Ann Intern Med. 2004;140:128-141.
2. Sawin CT, Castelli WP, Hershman JM, McNamara P, Bacharach P.
The aging thyroid. Thyroid deficiency in the Framingham Study. Arch
Intern Med. 1985;145:1386-1388.
3. Hollowell JG, Staehling NW, Flanders WD, et al. Serum TSH, T(4),
and thyroid antibodies in the United States population (1988 to 1994):
National Health and Nutrition Examination Survey (NHANES III).
J Clin Endocrinol Metab. 2002;87:489-499.
4. Bauer DC, Ettinger B, Nevitt MC, Stone KL. Risk for fracture in
women with low serum levels of thyroid-stimulating hormone. Ann
Intern Med. 2001;134:561-568.
5. Stall G, Harris S, Sokoll LJ, Dawson-Hughes B. Accelerated bone loss
in hypothyroid patients overtreated with l-thyroxine. Ann Intern Med.
1990;113:265-269.
6. Canaris GJ, Manowitz NR, Mayor G, Ridgway EC. The Colorado
thyroid disease prevalence study. Arch Intern Med. 2000;160:526-534.
7. Sawin CT, Geller A, Wolf PA, et al. Low serum thyrotropin concentrations as a risk factor for atrial fibrillation in older persons. N Engl
J Med. 1994;331:1249-1252.
8. Barker DJ, Phillips DI. Current incidence of thyrotoxicosis and past
prevalence of goitre in 12 British towns. Lancet. 1984;2:567-570.
9. Goldman MB, Monson RR, Maloof F. Cancer mortality in women
with thyroid disease. Cancer Res. 1990;50:2283-2289.
10. Bauer DC, Ettinger B, Browner WS. Thyroid functions and serum
lipids in older women: a population-based study. Am J Med. 1998;
104:546-551.
11. Althaus BU, Staub JJ, Ryff-De Leche A, Oberhansli A, Stahelin HB.
LDL/HDL-changes in subclinical hypothyroidism: possible risk factors for coronary heart disease. Clin Endocrinol (Oxf). 1988;28:
157-163.
12. Sawin C. Subclinical hyperthyroidism and atrial fibrillation. Thyroid.
2002;12:501-503.
13. Surks MI, Ortiz E, Daniels GH, et al. Subclinical thyroid disease:
scientific review and guidelines for diagnosis and management. JAMA.
2004;291:228-238.
14. Franklyn JA, Maisonneuve P, Sheppard MC, Betteridge J, Boyle P.
Mortality after the treatment of hyperthyroidism with radioactive iodine. N Engl J Med. 1998;338:712-718.
15. Ron E, Doody MM, Becker DV, et al. Cancer mortality following
treatment for adult hyperthyroidism. Cooperative Thyrotoxicosis Therapy Follow-up Study Group. JAMA. 1998;280:347-355.
16. Kapdi CC, Wolfe JN, Shapiro S, et al. Breast cancer. Relationship to
thyroid supplements for hypothyroidism. JAMA. 1976;236:1124-1127.
17. Leese GP, Jung RT, Guthrie C, Waugh N, Browning MC. Morbidity
in patients on L-thyroxine: a comparison of those with a normal TSH
to those with a suppressed TSH. Clin Endocrinol (Oxf). 1992;37:
500-503.
18. Petersen K, Bengtsson C, Lapidus L, Lindstedt G, Nystrom E. Morbidity, mortality, and quality of life for patients treated with levothyroxine. Arch Intern Med. 1990;150:2077-2081.
19. Shapiro S, Slone D, Kaufman DW, et al. Use of thyroid supplements
in relation to the risk of breast cancer. JAMA. 1980;244:1685-1687.
Bauer et al
Thyroid Disease and Mortality
20. Tunbridge WMG, Evered DC, Hall R, et al. The spectrum of thyroid
disease in a community: the Whickham survey. Clin Endocrinol.
1977;7:481-493.
21. Tieche M, Lupi GA, Gutzwiller F, Grob PJ, Studer H, Burgi H.
Borderline low thyroid function and thyroid autoimmunity. Risk factors for coronary heart disease? Br Heart J. 1981;46:202-206.
22. Vanderpump MP, Tunbridge WM, French JM, et al. The development
of ischemic heart disease in relation to autoimmune thyroid disease in
a 20-year follow-up study of an English community. Thyroid. 1996;
6:155-160.
23. Hak A, Pols H, Visser T, Drexage H, Hoffman A, Witteman J.
Subclinical hypothyroidism is an independent risk factor for atherosclerosis and myocardial infarction in elderly women: the Rotterdam
Study. Ann Intern Med. 2000;132:270-278.
24. Cummings S, Nevitt M, Browner W, et al. Risk factors for hip fracture
in white women. N Engl J Med. 1995;332:767-773.
25. Prentice RL. A case-cohort design for epidemiologic cohort studies
and disease prevention trials. Biometrika. 1986;73:1-11.
26. Kashiwai T, Ichihara K, Tamaki H, et al. The stability of immunological and biological activity of human thyrotropin in buffer: its temperature-dependent dissociation into subunits during freezing. Scand
J Clin Lab Invest. 1991;51:417-423.
27. Spencer C, Schwarzbein D, Guttler R, LoPresti J, Nicoloff J. Thyrotropin (TSH)-releasing hormone stimulation test responses employing
third and fourth generation TSH assays. J Clin Endocrinol Metab.
1992;76:494-498.
28. Cooper D, Halpern R, Wood L, Levin A, Ridgway E. L-thyroxine therapy
in subclinical hypothyroidism. Ann Intern Med. 1984;101:18-24.
29. Kong WM, Sheikh MH, Lumb PJ, et al. A 6-month randomized trial
of thyroxine treatment in women with mild subclinical hypothyroidism. Am J Med. 2002;112:348-354.
30. Jaeschke R, Guyatt G, Gerstein H, et al. Does treatment with Lthyroxine influence health status in middle-aged and older adults with
subclinical hypothyroidism? J Gen Intern Med. 1996;11:744-749.
349
31. Bastenie PA, Vanhaelst L, Neve P. Coronary-artery disease in hypothyroidism. Lancet. 1967;2:1221-1222.
32. Salabe GB, Salabe-Lotz H, Puletti M, Milani C. Hypothyroidism and
thyroid autoimmunity in acute myocardial infarction (AMI). Thyroidology. 1989;1:67-72.
33. Sellitti DF, Dennison D, Akamizu T, Doi SQ, Kohn LD, Koshiyama
H. Thyrotropin regulation of cyclic adenosine monophosphate production in human coronary artery smooth muscle cells. Thyroid. 2000;10:
219-225.
34. Parle JV, Maisonneuve P, Sheppard MC, Boyle P, Franklyn JA.
Prediction of all-cause and cardiovascular mortality in elderly people
from one low serum thyrotropin result: a 10-year cohort study. Lancet.
2001;358:861-865.
35. Peters A, Ehlers M, Blank B, et al. Excess triiodothyronine as a risk
factor of coronary events. Arch Intern Med. 2000;160:1993-1999.
36. Forfar JC, Muir AL, Sawers SA, Toft AD. Abnormal left ventricular
function in hyperthyroidism: evidence for a possible reversible cardiomyopathy. N Engl J Med. 1982;307:1165-1170.
37. Cooper DS. Antithyroid drugs. N Engl J Med. 2005;352:905-917.
38. Bartalena L, Bogazzi F, Martino E. Adverse effects of thyroid
hormone preparations and antithyroid drugs. Drug Saf. 1996;
15:53-63.
39. Smyth PP, Shering SG, Kilbane MT, et al. Serum thyroid peroxidase
autoantibodies, thyroid volume, and outcome in breast carcinoma.
J Clin Endocrinol Metab. 1998;83:2711-2716.
40. Giani C, Fierabracci P, Bonacci R, et al. Relationship between
breast cancer and thyroid disease: relevance of autoimmune thyroid
disorders in breast malignancy. J Clin Endocrinol Metab. 1996;81:
990-994.
41. Brix TH, Kyvik KO, Hegedus L. Validity of self-reported hyperthyroidism and hypothyroidism: comparison of self-reported questionnaire data with medical record review. Thyroid. 2001;11:769-773.
The American Journal of Medicine (2007) 120, 350-356
CLINICAL RESEARCH STUDY
TB in a Low-Incidence Country: Differences Between New
Immigrants, Foreign-Born Residents and Native Residents
Gerd Laifer, MD,a Andreas F. Widmer, MD, MS,a Mathew Simcock, MSc,b Stefano Bassetti, MD,a Andrej Trampuz, MD,a
Reno Frei, MD,c Michael Tamm, MD,d Manuel Battegay, MD,a Ursula Fluckiger, MDa
a
Division of Infectious Diseases and Hospital Epidemiology, bBasel Institute of Epidemiology, cDivision of Microbiology, and dDivision
of Pneumology, University Hospital Basel, Switzerland.
ABSTRACT
BACKGROUND: New immigrants and foreign-born residents add to the burden of pulmonary tuberculosis (TB) in low-incidence countries. The highest TB rates have been found among recent immigrants.
Active screening programs are likely to change the clinical presentation of TB, but the extent of the
difference between immigrant and resident populations has not been studied prospectively.
METHODS: Adult new immigrants were screened upon entry to 1 of 5 immigration centers in Switzerland. Immigrants with abnormal chest radiographs were enrolled and compared in a cohort study to
consecutive admitted foreign-born residents from moderate-to-high incidence countries and native residents presenting with suspected TB.
RESULTS: Of 42,601 new immigrants screened, 112 had chest radiographs suspicious for TB. They were
compared with foreign-born residents (n ⫽ 118) and native residents (n ⫽ 155) with suspected TB (n ⫽ 385
patients included). Active TB was confirmed in 40.5% of all patients (immigrants 38.4%, foreign-born
residents 50%, native residents 34.8%). Clinical signs and symptoms of TB and laboratory markers of
inflammation were significantly less common in immigrants than in the other groups with normal results
in ⬎70%. The proportion of positive results on rapid testing to detect M. tuberculosis (MTB) in 3
respiratory specimens was significantly lower in immigrants (34.9% for acid-fast staining; 55.8% for
polymerase chain reaction) compared with foreign-born residents (76.2% and 89.1%, respectively) and
native residents (83.3% and 90.9%, respectively). Isoniazid resistance and multi-drug resistance were more
prevalent in immigrants.
CONCLUSION: New immigrants with TB detected in a screening program are often asymptomatic and
have a low yield of rapid diagnostic tests but are at higher risk for resistant MTB strains. Postmigration
follow-up of pulmonary infiltrates is essential in order to control TB among immigrants, even in the
absence of clinical and laboratory signs of infection. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Foreign-born; Clinical presentation; Immigrants; Residents; Screening; Tuberculosis
More than one-third of the world’s population is estimated
to be infected with Mycobacterium tuberculosis complex
(MTB). Eight million new cases of pulmonary tuberculosis
(TB) and approximately 2 million deaths attributable to TB
are reported each year.1 In many developed countries, new
immigrants and foreign-born residents are increasingly conRequests for reprints should be addressed to Gerd Laifer, MD, Division
of Infectious Diseases and Hospital Epidemiology, University Hospital
Basel, Switzerland; Petersgraben 4, 4031 Basel, Switzerland.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.10.025
tributing to the burden of TB in the host country. The
proportion of TB cases in these 2 groups exceeds 50% in the
US and parts of Europe.2-4 In particular, new immigrants
who are screened for TB as they enter a country show high
TB prevalences (256-504/100,000 population per year).3,5-8
Consistently high incidences (33-85/100,000 population per
year) also have been reported for foreign-born residents in
low-prevalence countries.3,6,9,10 Among the native residents
in these countries, the TB incidence is below 10/100,000
population per year.3,10 The risk of disease for foreign-born
Laifer et al
New Immigrants with Tuberculosis
residents is highest in the first 5 years after arrival, with 50%
of cases occurring in this period.11
351
1 of the 5 State immigration centers (Basel) in Switzerland
(active screening). Foreign-born residents (ie, persons with
work permit in Switzerland and their families, students and
Prompt identification, early effective treatment, and isotourists) are not actively screened for tuberculosis in Switlation of TB patients are important strategies in preventing
zerland. Like the native residents, they were referred at the
an increase in the occurrence of
discretion of the general practitioTB in the general population of a
ners (passive case detection). Delow-incidence country. The most
mographic data (age, sex, country
CLINICAL SIGNIFICANCE
important criteria for establishing
of origin), duration of isolation,
an early diagnosis are positive relaboratory results, and microbio● Immigrants add substantially to the
sults in rapid diagnostic tests on
logical data were collected proburden of tuberculosis (TB) in lowrespiratory specimens (acid-fast
spectively and recorded in a dataincidence countries.
smear and polymerase chain reacbase. Additional data (fever,
tion) and a case definition that is
● Screening programs to detect TB in imweight loss, night sweats, cough,
based on clinical and radiographic
sputum production, dyspnea, and
migrants are recommended as a measure
signs and risk factors. Although
auscultatory findings) were exto support TB elimination efforts.
scoring systems and decision trees
tracted by chart review. Countries
have been developed to more ad● Postmigration follow-up of pulmonary
of origin were categorized into
equately predict patients at high
infiltrates
is
essential
in
order
to
control
those with a low notification rate
risk for active TB,12-15 they are
TB among immigrants, even in the abof TB (⬍24/100,000 population
difficult to apply in clinical
sence of clinical signs of infection.
per year) and those with a moderpractice.
ate to high notification rate of TB
The clinical presentation of TB
(ⱖ24/100,000 population per
depends largely on how far the
year), as reported by the 1998 redisease has progressed in an individual. The diagnosis is
port of the World Health Organization (WHO). The study
particularly challenging in immigrants who are actively
was reviewed and approved by the Ethics Committee of the
screened upon entry to their host country. We hypothesized
University of Basel.
that active screening changes the presentation of TB compared with passive case detection and that many actively
Microbiology
screened immigrants presenting with TB are asymptomatic.
Respiratory specimens were digested, decontaminated, conThe extent of this difference in comparison with the resident
centrated, and stained with auramine-rhodamine and then
population has not been investigated in prospective studies
examined with fluorescent microscopy. Ziehl-Neelsen acidbefore.
fast staining (AFS) was used to confirm the presence of
Active screening programs are currently under debate beacid-fast bacilli. Mycobacterial cultures were performed by
cause of their costs and the limited evidence of a reduction in
sample-inoculation on to Loewenstein-Jensen media and
transmitting TB in the host country. If most immigrants with
into liquid media (radiometric Bactec 460 TB system,
TB are asymptomatic, and thus cannot be diagnosed by passive
Becton Dickinson, Franklin Lakes, NJ, used from 1997 to
case detection, active screening programs should be reinforced.
2000; BacT/ALERT 3D system, BioMérieux Inc., Durham,
This would aid in identification of large numbers of individuals
NC, used from 2001 through 2004) according to the
who will benefit from therapeutic interventions and limit the
manufacturers’ instructions and incubated at 37°C for 8
transmission at least within the migrant community living
weeks. Polymerase chain reaction (PCR) testing for MTB
under crowded conditions.
was performed by using the Roche Amplicor Mycobacterium test with a Cobas Amplicor analyzer (Roche DiMETHODS
agnostics Systems, Basel, Switzerland) according to the
manufacturer’s instructions on concentrated decontamiPatient Population
nated specimens.
From January 1997 through July 2004, all patients admitted
to our hospital who were placed in respiratory isolation for
Study Definitions
suspected TB were included in this cohort study. The University Hospital Basel is a 780-bed primary and tertiary care
Active TB was defined as one or more positive cultures of
center with approximately 23,000 admissions annually. The
respiratory specimens for MTB (proven TB), or as culturepatients were categorized into new immigrants who were
negative with a compatible abnormal chest radiograph and
actively screened for TB upon entry into Switzerland, forclinical and radiological improvement after therapy for 2-3
eign-born residents who emigrated from countries with a
months with a standard combination therapy and without
moderate to high prevalence of TB, and native Swiss resiany alternative diagnosis (probable TB needing treatment).
dents. Immigrants were referred to our hospital if the
Inactive TB was defined as negative cultures and an abnorscreening chest radiograph was suspect for TB upon entry to
mal chest radiograph, which was stable after a standard
352
therapy for 2-3 months in suspected TB cases without alternative diagnosis. An alternative diagnosis was defined as
a diagnosis of an infection other than TB, cancer, or noninfectious diseases diagnosed by treating physicians and
explaining the abnormal chest radiograph and the clinical
presentations. Multi-drug-resistance (MDR) was defined as
the resistance of MTB to both isoniazid and rifampin. Excessive alcohol use was defined as consumption of ⬎14
standard drinks per week for men and ⬎7 standard drinks
per week for women, according to the criteria for “at risk”
drinking established by the National Institute on Alcohol
Abuse and Alcoholism.
Statistical Methods
The analyses were carried out by using SAS software (SAS
Institute Inc., Cary, NC). All predefined associations that
were considered to be important predictors were investigated. They consisted of using t tests between continuous
data variables and chi-squared tests between categorical
data variables. In situations where the cell counts fell below
The American Journal of Medicine, Vol 120, No 4, April 2007
5 in 20% or more of the cells, Fisher’s exact test was
performed. To overcome the problem inherent in multiple
testing, P-values were adjusted by using the Bonferroni
correction method.
RESULTS
Patient Characteristics
Between January 1997 and July 2004, 397 patients with
suspected pulmonary TB were included in the study.
Twelve patients (3.0%) were excluded because of incomplete data. The distribution of the patients and the diagnoses
are shown in the Figure.
There were 42,601 new immigrants ⬎16 years of age
who were actively screened by chest radiograph upon entry
to Switzerland at the Basel immigration center (median age
26 years; range 16-98 years). Of these, 112 (0.26%) were
referred to the hospital with pulmonary infiltrates suspicious
for TB. Table 1 shows the demographic characteristics and
risk factors for TB.
Figure Diagnoses of patients in respiratory isolation with pulmonary infiltrates suspicious for pulmonary tuberculosis. TB ⫽ tuberculosis;
NTM ⫽ nontuberculous mycobacteria.
Laifer et al
New Immigrants with Tuberculosis
353
Table 1 Demographic Characteristics of New Immigrants,
Foreign-born Residents and Native Residents with Active
Pulmonary Tuberculosis
Characteristic
Age (years)
Mean range
Male sex (%)
Area of origin (%)
Africa
Asia
Latin America
Eastern Europe
Southern Europe
Central Europe
TB risk factors (%)
HIV positive
Excess alcohol use
Diabetes
Cancer
History of TB (%)
Immigrant Foreign-born
Group
Native Group
Group
(n ⫽ 54)
(n ⫽ 43) (n ⫽ 59)
30.6
16-49
90.7
35.1
16-86
61.0
50.2**
23-88
72.2
18.6
9.3
—
62.8
9.3
—
0*
0*
0
0
0
14.0*
15.3
16.9
10.2
32.2
25.4
—
25.4
12.5
10.2
5.1
1.7
1.7
—
—
—
—
—
100
59.3**
18.4**
27.7**
7.4
1.9
3.7
34.8% in native residents). A prior history of tuberculosis
was reported more often among new immigrants than
among either foreign-born or native residents. Significantly
more native residents with active TB had risk factors for
TB, compared with new immigrants or foreign-born residents with active TB (Table 1). Excess alcohol use was the
predominant risk factor.
Clinical, Radiological, and Laboratory
Presentation
TB ⫽ pulmonary tuberculosis; HIV ⫽ human immunodeficiency
virus.
*P ⬍.05 immigrants vs native and vs foreign-born.
**P ⬍.05 native vs foreign-born and vs immigrants.
Active TB was confirmed in 40.5% of all patients isolated for suspected TB (38.4% in the screening group of
new immigrants, 50.0% in foreign-born residents, and
All clinical signs of infection (ie, fever, night sweats, weight
loss, cough, sputum production, and dyspnea) were significantly less common in the screening group of new immigrants with active TB than they were in foreign-born and
native residents with active TB (Table 2). The clinical
presentation of TB in foreign-born and native residents did
not differ with the exception of dyspnea, which was reported more often in foreign-born residents. Pulmonary
auscultation was less likely to be abnormal in the screening
group (new immigrants) compared with the foreign-born
and native residents (Table 2). Radiological presentation
was similar in the 3 groups with the exception of more
multilobular involvement in native residents (Table 2),
whereas foreign-born residents presented more often with
cavitations (Table 2). Systemic signs of inflammation, as
measured by C-reactive protein (CRP) and neutrophil
counts, were significantly more elevated in native residents
with active TB than they were in foreign-born residents with
Table 2 Disease Characteristics of New Immigrants, Foreign-born Residents, and Native
Residents with Active Pulmonary Tuberculosis
Disease Characteristic
Fever ⬎38.3 n (%)
Night sweats n (%)
Weight loss n (%)
Cough n (%)
Sputum n (%)
Dyspnea n (%)
Crackles n (%)
Infiltrate
Upper lobe n (%)
⬎1 Lobe n (%)
Bilateral n (%)
Cavitation n (%)
Neutrophils
Mean (109/L)
Hemoglobin
Mean (g/L)
CRP Mean (mg/L)
Albumin Mean (g/L)
Case fatality rate (%)
New Immigrant
(n ⫽ 43)
Foreign-born
(n ⫽ 59)
Native
(n ⫽ 54)
7/42
7/40
11/40
19/42
10/42
3/39
6/42
40/56
30/53
32/52
51/56
43/55
10/55
20/56
(71.4)
(56.6)
(61.5)
(91.1)
(78.2)
(18.2)†
(35.7)
40/52
20/51
30/52
47/52
41/52
24/52
21/52
(76.9)
(39.2)
(57.7)
(90.4)
(78.8)
(46.2)
(40.4)
37/41 (90.2)
3/41 (7.3)
9/41 (22.0)
10/41 (24.4)
5.83*
46/57 (80.7)
7/57 (12.3)
12/57 (21.1)
26/57 (45.6)§
6.36†
34/50
15/50
22/50
17/50
9.18
(68.0)
(30.0)‡
(44.0)
(34.0)
14.2*
12.2
12.0
17.0*
39.3*
0*
67.1†
32.0†
1/59 (1.7%)†
90.7
27.1
6/54 (11.1%)
(16.7)*
(17.5)*
(27.5)*
(45.2)*
(23.8)*
(7.7)*
(14.3)*
CRP ⫽ C-reactive protein.
*P ⬍.05 immigrants vs foreign-born and vs native.
†P ⬍.05 foreign-born vs native.
‡P ⬍.05 native vs immigrants.
§P ⬍.05 foreign-born vs immigrants.
354
The American Journal of Medicine, Vol 120, No 4, April 2007
Table 3
Microbiological Data for Patients with Active Pulmonary Tuberculosis
Results According to Patient Category
Immigrant
Type of Data
Proportion of positive acid-fast
smear
Proportion of positive PCR
Proportion of positive culture
Foreign-born
Per specimen
n ⫽ 134
ⱖ1 Positive per
patients
n ⫽ 43
Per specimen
n ⫽ 170
23.1%*
34.9%*
57.3%
35.4%*
52.2%*
55.8%*
76.7%*
81.4%
92.4%
Native
ⱖ1 Positive per
patients
n ⫽ 59
Per specimen
n ⫽ 128
ⱖ1 Positive per
patients
n ⫽ 54
76.2%
75.6%
83.3%
89.1%
100%
80.8%
85.8%
90.9%
98.1%
PCR ⫽ polymerase chain reaction.
*P ⬍.05; immigrants vs foreign-born residents and vs native residents.
active TB (Table 2). Systemic signs of inflammation were
low in the screening group of new immigrants (Table 2) and
were absent in a substantial proportion of patients with
active TB (normal CRP in 73.8%; normal neutrophil count
in 71.4%) in this population. Furthermore, hemoglobin levels and albumin levels were lower in foreign-born and
native residents with active TB than they were in new
immigrants (Table 2). In-hospital mortality was increased
among native residents with active TB (11.1% vs 0% in new
immigrants and 1.7% in foreign-born residents; Table 2).
Microbiology
A total of 1158 respiratory specimens were collected for
mycobacterial diagnostics (mean 3.0 specimens/patient).
The proportion of positive rapid tests for MTB (AFS and
PCR) was significantly lower in new immigrants in all
specimens of patients with active TB (23.1% for AFS;
35.4% for PCR) compared with foreign-born residents
(57.3% and 81.4%, respectively; Table 3) and native residents (75.6% and 80.8%, respectively; Table 3). Similar
results were obtained from cultures for MTB (proportion of
positive results in 52.2% of specimens of new immigrants
vs. 92.4% in foreign-born residents and 85.8% in native
residents; Table 3). Accordingly, the proportion of patients
with active disease and at least one positive rapid test for
MTB was lower in new immigrants (34.9% for AFS, 55.8%
for PCR) compared with foreign-born residents (76.2% and
89.1%, respectively; Table 3) and native residents (83.3%
and 90.9% respectively; Table 3). Specificity was 98% or
higher throughout all 3 groups for rapid diagnostic tests,
as well as for cultures. A significant proportion of new
immigrants had culture-negative TB (23.3%), whereas
only 1/113 patients in the other 2 groups had culturenegative TB.
Drug Resistance
Isoniazid resistance and multi-drug resistance were more
prevalent in the new immigrant population (21.9% and
6.3%, respectively) than they were in foreign-born residents
(10.2% and 1.7%; P ⬍.05) and native residents (5.8% and
1.9%; P ⬍.05).
DISCUSSION
In our study, we demonstrated that immigrants in screening
programs present with significantly fewer clinical signs
compared with foreign-born residents and native residents
living in a low-incidence country. This is not surprising
because the latter 2 groups present with a disease, whereas
the new immigrants were screened irrespective of clinical
symptoms. The important result of this study is that the vast
majority (⬎70%) of new immigrants screened upon entry
show no clinical signs or symptoms of TB at all and present
with normal laboratory markers of inflammation despite
having active TB. The extent to which new immigrants are
asymptomatic is a new finding. Previous retrospective studies report up to 50% of actively screened immigrants to be
“symptom-free.”16 Moreover, the disappointing low yield
of rapid tests for MTB underscores the diagnostic dilemma
for this population and the need of a postmigration
follow-up.
Prompt identification and effective isolation of patients
with active TB is considered to have a high priority in TB
control policies, especially in immigrants. Patients are less
contagious while still smear-negative17 and have a lower
morbidity and mortality. In the US regulations require immigrants to undergo TB screening in their countries of
origin, and travel to the US is permitted only if sputum
smears are negative in individuals with pulmonary infiltrates. Almost 7% of immigrants who arrived in the US and
were referred for further medical evaluation had active
TB.18 Furthermore, more than 80% of new immigrants in
the US who are subsequently diagnosed with active TB
were screened within 6 months before arrival in the US with
negative results, and 50% were given a diagnosis within 30
days after arrival.8 The ability of current overseas screening
programs to detect TB based on chest radiographs and AFS
is low19 and will miss persons with AFS-negative active
TB, who constitute the majority (65.1%) of new immigrants
Laifer et al
New Immigrants with Tuberculosis
in our study. Despite its superior accuracy compared with
AFS20 and its high negative predictive value in refugees
with culture-positive TB,5 PCR testing for MTB is not
recommended in this setting. A strategy including chest
radiographs upon arrival in the host country and a clinical
work-up including repeated PCR testing in case of TBsuspicion may be superior, because the yield of AFS in the
screening population in our study was disappointingly low
and was even lower than previously reported in another
screening population of immigrants in the Netherlands.21
Despite posing a substantial financial burden to host countries with a low incidence of TB, similar strategies have
been shown to be cost effective.22,23
Both the underdiagnosis and overdiagnosis of patients at
risk for TB make actual isolation guidelines difficult to
implement and not fully effective. Importantly, scoring systems and decision trees to identify patients at high risk for
active TB12-14 are not validated in screening populations
and may not apply to this setting. In institutions that must
deal with great numbers of TB patients, TB is confirmed in
one of 7-10 patients isolated.4,24,25 In our single-center
study, active TB was confirmed in 40.5% of all isolated
patients and initially missed in only 5%, which is in accordance with published data from the literature.15
The sensitivities of single cultures for MTB range from
50% to 81%26 and are in line with the results obtained in our
study. Low bacillary populations and temporal variations in
the number of bacilli being expelled are possible explanations for a failure to isolate MTB.
About 17% of all TB cases among foreign-born persons
in the US are culture-negative.10 In our study, especially
new immigrants in the screening group presented with a
culture-negative TB, whereas the overall rate was only
7.1%, confirming the results of a retrospective study comparing active and passive screening for TB.16
The prevalence of drug-resistant MTB can reach up to
37% and is highest in Eastern Europe and parts of Asia.27 In
the US, the primary resistance rate to isoniazid in foreignborn residents was reported to be 11.6%, with multi-drug
resistance being 1.7%.10 Similar results were obtained in
studies from the UK7 and from Texas,28 and these are in
accordance with the rates obtained in our study. Importantly, especially in new immigrants the proportion of drugresistant MTB-strains is high, as shown in our study.
Although the study was not designed to detect differences in mortality rates, native residents who died were
older and had more risk factors for TB, including excess
alcohol intake and HIV co-infection. Treating physicians
possibly suspect TB less often in native residents in a
low-incidence country, which may contribute to a delayed
hospital admission, a more advanced disease at the time of
presentation, and an increased case fatality rate. The higher
proportion of HIV-infection in native residents with TB
may explain the more frequent multilobular involvement on
chest radiograph and is in concordance with previous reports from the literature.10
355
Our study shows several strengths. All 3 groups in the
study consisted of a similar number of patients (more than
100 per group) with a high overall rate of confirmed TB
(40.5%) and a similar TB rate in each group. The standard
of care and the diagnostic approach of all patients in respiratory isolation were the same over a period of almost 8
years. There are 2 potential limitations to this study. First,
this study is a single center study, which may not represent
other settings in the Western world. Second, the communication with persons from foreign countries is often difficult,
which may introduce a possible bias in taking an accurate
patient’s history.
In conclusion, foreign-born residents and native residents
with active TB do not show significant differences in the
clinical, radiological, and mycobacterial presentation in a
low-incidence country. In contrast, new immigrants
screened upon entry present without clinical signs and
symptoms of TB in the majority of cases. Because the
diagnostic yield of AFS in this population is low, a clinical
work-up including better diagnostic tools like PCR testing
and a postmigration follow-up is important in order to
control TB among immigrants with pulmonary infiltrates,
even in the absence of clinical and laboratory signs of
infection.
References
1. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus
statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO Global Surveillance and Monitoring Project. JAMA. 1999;282:677-686.
2. Centers for Disease Control and Prevention. Trends in tuberculosis
morbidity—United States, 1992-2002. MMWR Morb Mortal Wkly
Rep. 2003;52:217-222.
3. Maurer AM, Seiler AJ. Diagnosis and treatment of pulmonary tuberculosis in asylum seekers and refugees in the Bern canton 1993-1997.
Schweiz Rundsch Med Prax. 2000;89:271-274.
4. Bloom BR. Tuberculosis—the global view. N Engl J Med. 2002;346:
1434-1435.
5. Laifer G, Widmer AF, Frei R, Zimmerli W, Fluckiger U. Polymerase
chain reaction for Mycobacterium tuberculosis: impact on clinical
management of refugees with pulmonary infiltrates. Chest. 2004;125:
981-986.
6. Van den Brande P, Uydebrouck M, Vermeire P, Demedts M. Tuberculosis in asylum seekers in Belgium. VRGT (Flemish Lung and
Tuberculosis Association). Eur Respir J. 1997;10:610-614.
7. Control and prevention of tuberculosis in the United Kingdom: code of
practice 2000. Joint Tuberculosis Committee of the British Thoracic
Society. Thorax. 2000;55:887-901.
8. Thorpe LE, Laserson K, Cookson S, et al. Infectious tuberculosis
among newly arrived refugees in the United States. N Engl J Med.
2004;350:2105-2106.
9. Wilcke JT, Poulsen S, Askgaard DS, Enevoldsen HK, Ronne T,
Kok-Jensen A. Tuberculosis in a cohort of Vietnamese refugees after
arrival in Denmark 1979-1982. Int J Tuberc Lung Dis. 1998;2:219224.
10. Talbot EA, Moore M, McCray E, Binkin NJ. Tuberculosis among
foreign-born persons in the United States, 1993-1998. JAMA. 2000;
284:2894-2900.
11. Binkin NJ, Zuber PL, Wells CD, Tipple MA, Castro KG. Overseas
screening for tuberculosis in immigrants and refugees to the United
States: current status. Clin Infect Dis. 1996;23:1226-1232.
12. Tattevin P, Casalino E, Fleury L, Egmann G, Ruel M, Bouvet E. The
validity of medical history, classic symptoms, and chest radiographs in
356
13.
14.
15.
16.
17.
18.
19.
20.
predicting pulmonary tuberculosis: derivation of a pulmonary
tuberculosis prediction model. Chest. 1999;115:1248-1253.
El-Solh A, Mylotte J, Sherif S, Serghani J, Grant BJ. Validity of a
decision tree for predicting active pulmonary tuberculosis. Am J Respir
Crit Care Med. 1997;155:1711-1716.
Wisnivesky JP, Kaplan J, Henschke C, McGinn TG, Crystal RG.
Evaluation of clinical parameters to predict Mycobacterium tuberculosis in inpatients. Arch Intern Med. 2000;160:2471-2476.
Wisnivesky JP, Henschke C, Balentine J, Willner C, Deloire AM,
McGinn TG. Prospective validation of a prediction model for isolating
inpatients with suspected pulmonary tuberculosis. Arch Intern Med.
2005;165:453-457.
Monney M, Zellweger JP. Active and passive screening for tuberculosis in Vaud Canton, Switzerland. Swiss Med Wkly. 2005;135:469474.
Behr MA, Warren SA, Salamon H, et al. Transmission of Mycobacterium tuberculosis from patients smear-negative for acid-fast bacilli.
Lancet. 1999;353:444-449.
DeRiemer K, Chin DP, Schecter GF, Reingold AL. Tuberculosis
among immigrants and refugees. Arch Intern Med. 1998;158:753-760.
Maloney SA, Fielding KL, Laserson KF, et al. Assessing the performance of overseas tuberculosis screening programs. Arch Intern Med.
2006;166:234-240.
Jonas V, Longiaru M. Detection of Mycobacterium tuberculosis by
molecular methods. Clin Lab Med. 1997;17:119-128.
The American Journal of Medicine, Vol 120, No 4, April 2007
21. Verver S, Bwire R, Borgdorff MW. Screening for pulmonary tuberculosis among immigrants: estimated effect on severity of disease and
duration of infectiousness. Int J Tuberc Lung Dis. 2001;6:419-425.
22. Schwartzman K, Menzies D. Tuberculosis screening of immigrants to
low-prevalence countries. A cost-effectiveness analysis. Am J Respir
Crit Care Med. 2000;161:780-789.
23. Verver S, van Soolingen D, Borgdorff MW. Effect of screening of
immigrants on tuberculosis transmission. Int J Tuberc Lung Dis. 2002;
6:121-129.
24. Bock NN, McGowan JE Jr, Ahn J, Tapia J, Blumberg HM. Clinical
predictors of tuberculosis as a guide for a respiratory isolation policy.
Am J Respir Crit Care Med. 1996;154:1468-1472.
25. McGowan JE Jr. Nosocomial tuberculosis: new progress in control and
prevention. Clin Infect Dis. 1995;21:489-505.
26. Levy H, Feldman C, Sacho H, van der Meulen H, Kallenbach J,
Koornhof H. A reevaluation of sputum microscopy and culture in the
diagnosis of pulmonary tuberculosis. Chest. 1989;95:1193-1197.
27. Espinal MA, Laszlo A, Simonsen L, et al. Global trends in resistance
to antituberculosis drugs. World Health Organization-International
Union against Tuberculosis and Lung Disease Working Group on
Anti-Tuberculosis Drug Resistance Surveillance. N Engl J Med. 2001;
344:1294-1303.
28. El Sahly HM, Adams GJ, Soini H, Teeter L, Musser JM, Graviss EA.
Epidemiologic differences between United States- and foreign-born
tuberculosis patients in Houston, Texas. J Infect Dis. 2001;183:461468.
The American Journal of Medicine (2007) 120, 357-363
CLINICAL RESEARCH STUDY
Coronary Ischemic Events after First Atrial Fibrillation:
Risk and Survival
Yoko Miyasaka, MD, PhD,a,b Marion E. Barnes, MSc,a Bernard J. Gersh, MB, ChB, Dphil,a Stephen S. Cha, MS,c
Kent R. Bailey, PhD,c James B. Seward, MD,a Toshiji Iwasaka, MD, PhD,b Teresa S. M. Tsang, MDa
a
c
Division of Cardiovascular Diseases and Internal Medicine, and bSection of Biostatistics, Mayo Clinic, Rochester, Minn.
Cardiovascular Division, Department of Medicine II, Kansai Medical University, Osaka, Japan.
ABSTRACT
PURPOSE: We aimed to determine the long-term, gender-specific incidence and mortality risk of
coronary ischemic events after first atrial fibrillation (AF).
METHODS: In this longitudinal cohort study, adult residents of Olmsted County, Minnesota, with an
electrocardiogram-confirmed AF first documented in 1980 to 2000 and without prior coronary heart
disease, were followed to 2004. The primary outcome was first coronary events (angina with
angiographic confirmation, unstable angina, nonfatal myocardial infarction, or coronary death).
Sex-specific incidence of coronary ischemic events and survival after development of such events
were assessed using Cox proportional hazards modeling. Kaplan-Meier estimates of risks for coronary
ischemic events were compared with those predicted by the Framingham equation.
RESULTS: Of the 2768 subjects (mean age 71 years, 48% were men), 463 (17%) had a first coronary
event during a follow-up of 6.0 ⫾ 5.2 years. The unadjusted incidence was 31 per 1000 person-years, and
there was no difference between men and women. The incidence was higher in men (hazard ratio 1.32,
P ⫽ .004) after adjusting for age. The 10-year event estimates were 22% and 19% in men and women,
respectively, by our Kaplan-Meier analyses, and 21% and 11%, respectively, by Framingham risk
equation. The mortality risk after coronary events was higher in women (hazard ratio 2.99 vs 2.33;
P ⫽ .044), even after multiple adjustment.
CONCLUSIONS: First AF marks a high risk for new coronary ischemic events in both men and women.
AF conferred additional risk for coronary events beyond conventional risk prediction in women only. The
excess mortality risk associated with the development of coronary events was significantly greater in
women. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Atrial fibrillation; Coronary heart disease; Incidence; Prognosis
Atrial fibrillation (AF) is a growing public health problem1 that has reached epidemic proportions.2,3 It is a
potent risk factor for stroke,4,5 congestive heart failure,6,7
and possibly cognitive dysfunction.8,9 Although studies
have shown that the risk of acute coronary ischemic
events at the time of first AF is low,10,11 there are no
This study was supported by the American Heart Association National
Scientist Development Grant. There is no financial conflict on the part of
any author.
Requests for reprints should be addressed to Teresa S. M. Tsang, MD,
Mayo Clinic, Division of Cardiovascular Diseases, 200 First Street SW,
Rochester, MN 55905.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.06.042
published longer-term data. There is a paucity of data with
regard to the incidence and timing of coronary events after first
AF, as well as the predictors for such events. In addition, it is
unknown whether gender differences exist in the risk of
post-AF coronary ischemic events and survival after
these events. We conducted a community-based longitudinal cohort study that spanned 2 decades and assessed
the gender-specific incidence, timing, and predictors of
coronary ischemic events after first diagnosis of AF
in patients without prior evidence of coronary heart disease. We determined the gender-specific mortality risk
conferred by the development of coronary ischemic
events.
358
METHODS
Study Setting
The American Journal of Medicine, Vol 120, No 4, April 2007
no history or evidence of coronary heart disease before or at
the time of AF diagnosis.
The community-based cohort study was approved by the
Definition and Ascertainment of Outcome
Mayo Foundation Institutional Review Board. Olmsted
Coronary Ischemic Events. The primary outcome of inCounty, Minnesota, is well suited
terest was a first documented corfor the conduct of studies with
onary ischemic event, defined as
long-term follow-up because of a
angina pectoris with angiographic
CLINICAL SIGNIFICANCE
number of unique features.12 Geoconfirmation, unstable angina,
graphically, the community is relnonfatal acute myocardial infarc● Patients diagnosed with first atrial fibrilatively isolated from other urban
tion, or coronary death occurring
lation constitute a high-risk group for
centers, and medical care is delivafter the date of first AF. We
subsequent new coronary ischemic
ered by only a few health care
chose well-defined end points
events.
providers, principally the Mayo
with high specificity and opted to
Clinic and its associated hospitals.
● Atrial fibrillation conferred additional
exclude the patients who had preThe majority of Olmsted County
risk for the development of coronary
sumed angina but without angioresidents return to the Mayo
graphic confirmation of obstrucischemic events, beyond conventional
Clinic on a regular basis, allowing
tive coronary artery disease.
risk prediction, in women only.
capture of health-related events. A
Coronary artery disease was deprevious study showed that 96%
● The excess mortality risk associated with
fined as an angiographic finding of
of Olmsted County women resithe development of new coronary events
atherosclerotic narrowing of least
dents aged 65 to 74 years returned
post-atrial fibrillation was greater in
50% in any of the 3 main arterial
to the Mayo Clinic within a 3-year
women.
distributions. Unstable angina was
period.12 For each patient at the
defined as the presence of new or
Mayo Clinic, a unified medical
accelerated ischemic symptoms
record containing details of all inwith or without electrocardiopatient and outpatient encounters is maintained. Coded digraphic changes, but without elevation of cardiac enzymes.
agnoses are electronically entered into a medical diagnostic
Acute myocardial infarction was considered present if at
index, allowing easy identification of all patients with a
least 2 of the 3 diagnostic criteria were fulfilled: compatible
diagnosis of interest. An electronic electrocardiogram dataclinical presentation, diagnostic cardiac enzyme/biomarker
base has been in use at Mayo since 1976, allowing rapid
levels, and electrocardiographic changes consistent with
identification of the coded interpretation and diagnoses for
myocardial infarction. Coronary death was defined as death
each electrocardiogram. All electrocardiograms can also be
attributable to fatal acute myocardial infarction and death
retrieved for direct review.
with the primary cause being coronary heart disease (based
on documentation per death certificates or physician adjudication by chart review). The ascertainment of outcomes
Incident Atrial Fibrillation Cohort and
was accomplished through comprehensive review of the
Study Population
medical records in addition to cross-referencing the multiple
The medical records of adult residents of Olmsted County,
administrative databases to identify any inconsistencies.
Minnesota, who had a first AF documented between January
1, 1980, and December 31, 2000, in any of the Mayo
administrative databases (medical index, surgical index,
electrocardiographic, and echocardiographic databases)
were reviewed and followed forward in the medical records
to March 2004 by 1 cardiologist (Y. M.). For the purpose of
this study, we excluded subjects who had any history or
evidence of coronary heart disease and subjects who sustained a first coronary ischemic event on the same day as the
first AF. Because the unit record system at Mayo dated back
to the early 1900s, any electrocardiogram performed in any
patient was contained within the unit record. Any AF that
occurred before the establishment of administrative databases would have been identified from the chart review
process. Final inclusion in the study population required the
following: The AF was sustained, allowing confirmation by
12-lead electrocardiogram; the AF episode was verified as
the first recognized AF event for the person; and there was
Death. The secondary outcome of interest was death. Ascertainment of death was accomplished through comprehensive review of the medical records, death certificates, Vital
Status Information from Mayo Registration, Minnesota
State Death Tapes, and Social Security Death Index.
Statistical Analysis. Baseline characteristics were summarized by means and standard deviations or frequency percentages and compared between subjects with or without
subsequent development of coronary ischemic events by
using age-adjusted Cox proportional hazards analyses for
both sexes. Average annual incidence rates were reported by
sex and 10-year age interval, and summaries were tabulated
as cases per 1000 person-years. The overall and sex-specific
cumulative incidence of first coronary ischemic events after
first AF was estimated using the Kaplan-Meier method.
Incidence was assessed as a function of age and sex, using
Miyasaka et al
Risk of Coronary Ischemic Events in AF
a Cox proportional hazards model for time to first coronary
ischemic event. Cox models based on clinical variables at
the time of first AF were also developed using backward
stepwise selection. The clinical covariates considered included type of AF (paroxysmal vs chronic) at initial diagnosis, body mass index, systolic or diastolic blood pressure,
history of systemic hypertension, dyslipidemia, diabetes
mellitus, smoking, peripheral vascular disease, stroke, congestive heart failure, valvular heart disease, chronic obstructive pulmonary disease, obstructive sleep apnea, regular
alcohol use, hyperthyroidism, chronic renal disease, and
malignancy.
We also analyzed whether AF provides additional predictive information, beyond conventional risk factor consideration, for the development of subsequent coronary ischemic events. We compared our sex-specific coronary
incidence rates, estimated by Kaplan-Meier analysis, with
those predicted by the Framingham risk equation13 at 10
years.
The association of first coronary ischemic event with
subsequent survival was estimated using time-dependent
proportional hazards models, controlling for age and sex,
and with and without clinical variables. The same clinical
Table 1
359
covariates were considered, and the same selection method
was used. The effect of interim coronary revascularization
(percutaneous coronary intervention or coronary artery bypass grafting) on mortality was accounted for by including
it as a time-dependent covariate in the model. Mortality
ratios relative to the general Minnesota white population
were calculated by dividing the observed number of mortality events by an “expected number” of deaths. This expected number was calculated as the sum of each individual’s cumulative hazard, based on the standard age, sex, and
calendar-year specific life-table survival probability up to
the individual’s follow-up time. Confidence intervals (CIs)
were based on the Poisson distribution, and gender comparisons were based on a likelihood ratio chi-square test. All
tests of significance were 2-tailed, and a P value less than
.05 was considered statistically significant.
RESULTS
A total of 4618 subjects (51% were men; mean age 73
years) were confirmed to have developed a first AF during
1980 to 2000. We excluded 1776 subjects (38%) because
they had a history of coronary heart disease and 74 subjects
Baseline Characteristics of the Study Population Stratified by Subsequent Development of Coronary Ischemic Events
Men
Women
Variable
No Events (n ⫽ 1091)
Had Events (n ⫽ 227)
No Events (n ⫽ 1214)
Had Events (n ⫽ 236)
Age (y)
Paroxysmal AF, N (%)
BMI (kg/m2)
Systemic hypertension, N (%)
Hypertensive therapy, N (%)
Systolic BP (mm Hg)
Diastolic BP (mm Hg)
Dyslipidemia, N (%)
Lipid-lowering therapy, N (%)
Total cholesterol (mg/dL)§
HDL cholesterol (mg/dL)§
Diabetes mellitus, N (%)
Smoking, N (%)
Peripheral artery disease, N (%)
Stroke, N (%)
Prior CHF, N (%)
VHD, N (%)
COPD, N (%)
Obstructive sleep apnea, N (%)
Regular alcohol use, N (%)
Hyperthyroidism, N (%)
Chronic renal disease, N (%)
History of malignancy, N (%)
66 ⫾ 18
797 (73)
27.3 ⫾ 5.4
691 (63)
373 (34)
133 ⫾ 18
78 ⫾ 11
344 (32)
19 (1.7)
191 ⫾ 47
45 ⫾ 14
125 (11)
762 (70)
72 (6.6)
79 (7.2)
54 (4.9)
142 (13)
245 (22)
29 (2.7)
225 (21)
9 (0.8)
137 (13)
314 (29)
69 ⫾ 13‡
151 (67)
28.4 ⫾ 5.2†
172 (76)‡
91 (40)†
138 ⫾ 20†
79 ⫾ 11
87 (38)†
4 (1.8)
198 ⫾ 45
44 ⫾ 12
37 (16)†
174 (77)*
19 (8.4)*
18 (7.9)
9 (4.0)
39 (17)*
60 (26)
9 (4.0)*
47 (21)
2 (0.9)
36 (16)†
46 (20)
76 ⫾ 14
825 (68)
26.5 ⫾ 6.9
961 (79)
613 (50)
139 ⫾ 21
78 ⫾ 11
329 (27)
20 (1.6)
204 ⫾ 45
55 ⫾ 17
143 (12)
424 (35)
80 (6.6)
91 (7.5)
89 (7.3)
264 (22)
187 (15)
6 (0.5)
72 (5.9)
21 (1.7)
99 (8.2)
364 (30)
77 ⫾ 11‡
147 (62)
27.4 ⫾ 7.3*
215 (91)‡
136 (58)‡
147 ⫾ 19‡
80 ⫾ 10
64 (27)
3 (1.3)
211 ⫾ 46
52 ⫾ 14
49 (21)‡
79 (33)
30 (13)‡
17 (7.2)
18 (7.6)*
58 (25)
36 (15)
2 (0.8)
9 (3.8)
5 (2.1)
26 (11)‡
48 (20)
AF ⫽ atrial fibrillation; BMI ⫽ body mass index; BP ⫽ blood pressure; CHF ⫽ congestive heart failure; COPD ⫽ chronic obstructive pulmonary
disease; HDL ⫽ high-density lipoprotein; VHD ⫽ valvular heart disease.
Values are given as mean ⫾ standard deviation or number (percentage). P value for events status differences in each characteristics by Cox
proportional hazards analysis after adjusting for age.
*.01 ⬍ P ⬍.05 vs no events.
†.001 ⬍ P ⬍.01 vs no events.
‡P ⬍.001 vs no events.
§The available number of data were 1999 for total cholesterol and 1143 for HDL cholesterol.
360
The American Journal of Medicine, Vol 120, No 4, April 2007
confirmation (72 men [32%], 45 women [19%]), unstable
angina (45 men [20%], 59 women [25%]), nonfatal acute
myocardial infarction (66 men [29%], 79 women [34%]), and
coronary death (44 men [19%], 53 women [22%]).
Incidence, Timing, and Predictors of Coronary
Ischemic Events
Figure 1 The unadjusted incidence of first coronary ischemic
events by time from first AF diagnosis among men (left) and
women (right). AF ⫽ atrial fibrillation.
(2%) because they had a coronary ischemic event that occurred on the same day as the first AF. The remaining 2768
subjects (48% were men, mean age 71 years, range 18-107
years) constituted the study population. Of these, 463 patients (17%) had a total of 841 coronary ischemic events
during a mean follow-up period of 6.0 ⫾ 5.2 years, and 1700
died. The proportion of our cohort followed to death or to
within 2 years of the last follow-up of this study (March
2004) was 92.4%. The baseline characteristics of the study
population, stratified by sex and subsequent coronary ischemic events, are shown in Table 1.
Distribution of Coronary Ischemic Events
The type of first coronary ischemic events (227 men, 236
women) were as follows: angina pectoris with angiographic
The unadjusted incidence of coronary ischemic events after
AF was 31 per 1000 person-years and was not significantly
different between men and women (30.2 vs 31.4 per 1000
person-years; P ⫽ .68). The risk of new coronary ischemic
events was greatest during the first year of AF (cumulative
incidence 4.7%, 95% CI, 3.9-5.6) and plateaued thereafter
(2.5% per year after first year, cumulative event rate 14.4%
at 5 years, 95% CI, 12.8-15.9) (Figure 1) (Table 2). When
adjusted for age, male sex was associated with increased
risk of coronary ischemic events (hazard ratio [HR] 1.32,
95% CI, 1.09-1.60, P ⫽ .004). When age stratified, the incidence of coronary ischemic events increased throughout
the age range in women but did not increase in men after
age 85 years or more (Table 2). When a sex-specific correction for “age 85 years or more” was added to the model
with a linear term for age and sex, there was a significant
(P ⫽ .028) downward correction for men but no apparent
departure from linearity for women (P ⫽ .06 for the difference between the corrections). The overall incidence of
coronary ischemic events did not differ between men and
women, likely because the mean age of women was 10
years older (76 vs 66 years).
In a multivariable Cox model for the prediction of first
coronary events (Table 3), advancing age, male sex, higher
systolic blood pressure, and history of systemic hyperten-
Table 2 Age-and Sex-Specific Incidence Rates and Kaplan-Meier Estimated Cumulative Incidence of Coronary Ischemic Events for
Patients Diagnosed with First Atrial Fibrillation
Kaplan-Meier Estimated Cumulative
Incidence
Overall
Men
⬍55 y
55-64 y
65-74 y
75-84 y
ⱖ85 y
Overall
Women
⬍55 y
55-64 y
65-74 y
75-84 y
ⱖ85 y
Overall
N
Person-y
Incidence Rate/1000 Person-y*
1 y, % (95% CI)
5 y, % (95% CI)
2768
15046.53
30.77 (463)
4.7 (3.9-5.6)
14.4 (12.8-15.9)
299
207
333
323
156
1318
2573.86
1428.25
1829.45
1321.64
371.87
7525.07
12.04
27.31
38.81
52.96
43.03
30.17
(31)
(39)
(71)
(70)
(16)
(227)
0.7
2.5
8.4
6.3
5.0
4.8
(0.0-1.7)
(0.3-4.7)
(5.3-11.5)
(3.3-9.2)
(0.9-9.0)
(3.5-6.0)
5.4
11.8
19.7
19.5
17.2
14.5
(2.5-8.2)
(6.6-16.7)
(14.7-24.4)
(14.0-24.7)
(7.7-25.7)
(12.2-16.7)
101
139
288
533
389
1450
965.59
1058.37
1826.05
2589.81
1081.63
7521.45
10.36
21.73
29.02
35.52
53.62
31.38
(10)
(23)
(53)
(92)
(58)
(236)
1.0
5.4
4.2
4.6
6.1
4.7
(0.0-2.9)
(1.4-9.2)
(1.7-6.6)
(2.7-6.4)
(3.5-8.7)
(3.5-5.8)
3.2
10.2
10.9
16.1
22.7
14.3
(0.0-6.6)
(4.5-15.6)
(6.9-14.9)
(12.3-19.7)
(16.3-28.6)
(12.1-16.4)
CI ⫽ confidence interval.
*Data are presented as incidence rate followed in parentheses by the actual number of cases observed.
Miyasaka et al
Risk of Coronary Ischemic Events in AF
361
Table 3 Overall and Sex-Specific Multivariable-Adjusted Hazard Ratios for the Coronary Ischemic Events in Patients Diagnosed
with First Atrial Fibrillation
Overall
Age/10 y
Men
Systolic BP/10 mm Hg
Systemic hypertension
Diabetes mellitus
Peripheral artery disease
Valvular heart disease
Obstructive sleep apnea
Chronic renal disease
Men
Women
HR
95% CI
HR
95% CI
HR
95% CI
1.30‡
1.37†
1.06*
1.65‡
1.76‡
1.56†
1.39†
1.97*
1.47†
1.20-1.42
1.13-1.67
1.01-1.12
1.23-2.21
1.38-2.24
1.14-2.13
1.11-1.75
1.07-3.64
1.11-1.95
1.32‡
1.04
1.47*
1.50*
1.39
1.39
1.97
1.34
1.18-1.46
0.96-1.13
1.02-2.12
1.04-2.17
0.85-2.28
0.98-1.97
0.99-3.89
0.91-1.96
1.29‡
1.07
2.23†
2.04‡
1.67*
1.44*
2.43
1.79†
1.13-1.47
0.99-1.14
1.33-3.73
1.47-2.82
1.12-2.50
1.06-1.95
0.59-9.97
1.17-2.72
BP ⫽ blood pressure, HR ⫽ hazard ratio, CI ⫽ confidence interval.
*.01 ⬍ P ⬍.05.
†.001 ⬍ P ⬍.01.
‡P ⬍.001.
sion, diabetes mellitus, peripheral artery disease, valvular
heart disease, obstructive sleep apnea, and chronic renal
disease were significant independent predictors. The type of
AF, paroxysmal or chronic AF, was not significant. Separate
models for men and women including these same variables
appeared similar (Table 3), and none of the interactions
between the risk factors and the gender were significant.
A prediction formula for the probability of coronary
events within 1, 5, or 10 years based on this model is
given in the Appendix, and Table 4 shows an array of
sample calculations.
Gender Differences in the Risk of Coronary
Events after First Atrial Fibrillation
Although men have higher age-adjusted incidence of coronary ischemic events after first AF, how our sex-specific
coronary incidence rates relate to those predicted by the
Framingham risk equation is of interest. Restricting the
analysis to the subgroup aged 30 to 74 years (N ⫽ 1403),
the same age range for which the Framingham coronary risk
equation was developed,13 the estimations of coronary event
rates at 10 years were 22% (95% CI, 18-26) for men and
19% (95% CI, 14-23) for women by Kaplan-Meier analyses, versus 21% for men and 11% for women by Framingham risk equation.13
Table 4
Impact of Coronary Ischemic
Events on Survival
In time-dependent Cox analysis adjusted for age, the hazard
for death was highly significantly associated with the occurrence of post-AF coronary ischemic events in each sex
(P ⬍.0001). When an overall model was fit, with age, sex,
coronary events, and an interaction between coronary
events and sex, the HR associated with events was significantly higher in women (HR 2.99, 95% CI, 2.53-3.53 vs HR
2.33, 95% CI, 1.94-2.81, P ⫽ .044). Before the occurrence
of a coronary ischemic event, the age-adjusted HR for men
versus women was 1.23 (95% CI, 1.10-1.37), whereas after
the occurrence of a coronary ischemic event, there was no
significant difference in age-adjusted mortality risk. After
adjustment for age, clinical risk factors, and coronary revascularization as additional time-dependent covariates, the
occurrence of a coronary ischemic event remained a highly
significant independent predictor of death in both sexes
(men, HR 2.71, 95% CI, 2.22-3.32; women, HR 3.55, 95%
CI, 2.99-4.22; P ⫽ .033).
Compared with the general Minnesota white population,
the relative mortality HRs before coronary ischemic events
were 1.56 (95% CI, 1.47-1.66) in men and 1.71 (95% CI,
1.62-1.82) in women (P ⫽ .10). After the development of a
first coronary ischemic event, the corresponding relative
Probability of Coronary Ischemic Event within 5 Years Using a Prediction Formula* (Appendix)
Men
Women
Risk Factors
None
HTN
HTN ⫹ DM
HTN ⫹ DM ⫹ VHD
None
HTN
HTN ⫹ DM
HTN ⫹ DM ⫹ VHD
Age 60 y
Age 70 y
Age 80 y
7%
9%
11%
11%
14%
17%
18%
23%
29%
24%
30%
37%
5%
6%
8%
8%
10%
13%
13%
17%
22%
18%
23%
29%
HTN ⫽ systemic hypertension; DM ⫽ diabetes mellitus; VHD ⫽ valvular heart disease.
*Imputing the mean value of systolic blood pressure (137 mm Hg) for these estimations.
362
Table 5
The American Journal of Medicine, Vol 120, No 4, April 2007
Mortality Ratios Relative to the General Minnesota White Population after Atrial Fibrillation Diagnosis
Men
Overall
Pre-events
Post-events
Women
Observed*
Expected†
Ratio‡ (95% CI)
Observed*
Expected†
Ratio‡ (95% CI)
710
563
147
411.67
361.11
50.56
1.72 (1.62-1.84)
1.56 (1.47-1.66)
2.91 (2.35-3.66)
990
806
184
516.40
470.99
45.41
1.92 (1.81-2.04)§
1.71 (1.62-1.82)
4.05 (3.23-5.36)储
CI ⫽ confidence interval.
*Observed is number of deaths in the absence of coronary ischemic events (pre-events) or occurring after coronary ischemic events
(post-events).
†Expected is the cumulative hazard (based on age and sex using Minnesota white population) for follow-up before or in the absence of coronary
ischemic events (pre-events) or after the occurrence of coronary ischemic events (post-events).
‡Ratio of observed to expected represents hazard ratio relative to the general population.
§.01 ⬍ P ⬍.05 vs men.
储P ⬍.001 vs men.
HRs were 2.91 (95% CI, 2.35-3.66) in men and 4.05 (95%
CI, 3.23-5.36) in women (P ⬍.001) (Table 5).
DISCUSSION
In our community-based cohort, the first AF marks a high
risk for new coronary ischemic events in both men and
women. In contrast with men, AF conferred an additional
risk of ischemic events in women beyond conventional risk
prediction. The excess mortality risk associated with the
development of new coronary ischemic events was greater
in women.
Incidence of Coronary Ischemic Events After
First Atrial Fibrillation
Coronary heart disease has been regarded as an independent
risk factor of AF,14,15 although the data supporting this
connection came from only a small handful of studies, and
the event rate was modest.16,17 In the Framingham Heart
Study, the 2-year age-adjusted incidence for chronic AF
among patients with coronary heart disease, per 1000, was
8.9 for men and 2.3 for women.16 However, little is known
regarding AF as a precursor of new coronary ischemic
events in patients without prior evidence of coronary heart
disease.
Our study showed that the ischemic event rate was high
after manifestation of first AF, and that the unadjusted
incidence did not differ between men and women. Incidence, when adjusted for age, was higher in men. In fact,
the incidence of post-AF coronary ischemic events was
higher than that of post-AF stroke in the same cohort
(10% at 5 years).18 The precise reasons for the high
ischemic event rate after first AF among patients with no
prior coronary heart disease could not be well delineated,
but a likely explanation is that persons who develop AF
frequently harbor the same cardiovascular risk factors for
coronary ischemic events. In our cohort, more than 70%
had hypertension, 30% had dyslipidemia, and 13% had
diabetes. It is unclear why some develop ischemic events
first and others develop AF first.
In a multivariate Cox model, higher systolic blood
pressure, systemic hypertension, diabetes mellitus, and
obstructive sleep apnea were significant independent predictors of new coronary ischemic events after first AF.
Early diagnosis and control of these treatable conditions
may have implications on cardiovascular outcomes in
patients who have AF.
Although the age-adjusted incidence of new coronary
events post-AF was higher in men; AF actually conferred
an additional risk (⬃70%) of new coronary events in
women only, based on the comparison of our KaplanMeier 10-year event rates and those predicted by the
Framingham risk equation. Our data set does not lend
itself readily to analyses for potential explanatory factors.
The underlying reasons for this gender difference in the
impact of AF on subsequent coronary risk warrant further
investigation.
Impact of Coronary Ischemic
Events on Survival
AF is known to be associated with excess mortality
risk.14,19-22 Our study showed that the development of a first
coronary ischemic event further increased the mortality risk,
even after multivariable adjustment. In addition, the mortality risk relative to the general population was higher in
women than in men. This was similar to our previous
finding of greater relative mortality risk in women associated with post-AF stroke.18 The reasons for these differences could not be determined in this study but clearly
require further studies.
LIMITATIONS
There were inherent biases associated with the retrospective design. It is possible that the incidence of AF may
have been underestimated, given that some patients may
not have received care at the Mayo Clinic, although the
underestimation would likely be small because the Mayo
Clinic is the principal health care provider and referral
center for Olmsted County.12 Also, we may have under-
Miyasaka et al
Risk of Coronary Ischemic Events in AF
estimated the coronary event rates because we did not
include those who had presumed angina but did not have
angiographic confirmation. The population of Olmsted
County, Minnesota, is predominantly white and not racially ethnically diverse.
CONCLUSIONS
Patients diagnosed with a first AF constitute a high-risk
group for subsequent new coronary ischemic events. Although the age-adjusted incidence of such post-AF coronary ischemic events was higher in men, AF actually
conferred an additional risk for the development of ischemic events, beyond conventional risk prediction, in
women only. Further, the excess mortality risk associated
with the development of post-AF new coronary events
was greater in women.
See appendix online at http://www.amjmed.com
References
1. Chugh SS, Blackshear JL, Shen WK, et al. Epidemiology and natural
history of atrial fibrillation: clinical implications. J Am Coll Cardiol.
2001;37:371-378.
2. Braunwald E. Shattuck lecture— cardiovascular medicine at the turn of
the millennium: triumphs, concerns, and opportunities. N Engl J Med.
1997;337:1360-1369.
3. Alpert JS. Atrial fibrillation: a growth industry in the 21st century. Eur
Heart J. 2000;21:1207-1208.
4. Whisnant JP, Wiebers DO, O’Fallon WM, et al. A population-based
model of risk factors for ischemic stroke: Rochester, Minnesota. Neurology. 1996;47:1420-1428.
5. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent
risk factor for stroke: the Framingham Study. Stroke. 1991;22:983988.
6. Scheinman MM. Atrial fibrillation and congestive heart failure: the
intersection of two common diseases. Circulation. 1998;98:941-942.
7. Wang TJ, Larson MG, Levy D, et al. Temporal relations of atrial
fibrillation and congestive heart failure and their joint influence on
mortality: the Framingham Heart Study. Circulation. 2003;107:29202925.
8. Kilander L, Andren B, Nyman H, et al. Atrial fibrillation is an independent determinant of low cognitive function: a cross-sectional study
in elderly men. Stroke. 1998;29:1816-1820.
363
9. Ott A, Breteler MM, de Bruyne MC, et al. Atrial fibrillation and
dementia in a population-based study. The Rotterdam Study. Stroke.
1997;28:316-321.
10. Zimetbaum PJ, Josephson ME, McDonald MJ, et al. Incidence and
predictors of myocardial infarction among patients with atrial fibrillation. J Am Coll Cardiol. 2000;36:1223-1227.
11. Friedman HZ, Weber-Bornstein N, Deboe SF, Mancini GB. Cardiac
care unit admission criteria for suspected acute myocardial infarction
in new-onset atrial fibrillation. Am J Cardiol. 1987;59:866-869.
12. Melton LJ 3rd. History of the Rochester Epidemiology Project. Mayo
Clin Proc. 1996;71:266-274.
13. Wilson PW, D’Agostino RB, Levy D, et al. Prediction of coronary
heart disease using risk factor categories. Circulation. 1998;97:18371847.
14. Krahn AD, Manfreda J, Tate RB, et al. The natural history of atrial
fibrillation: incidence, risk factors, and prognosis in the Manitoba
Follow-Up Study. Am J Med. 1995;98:476-484.
15. Tsang TS, Barnes ME, Bailey KR, et al. Left atrial volume: important
risk marker of incident atrial fibrillation in 1655 older men and
women. Mayo Clin Proc. 2001;76:467-475.
16. Kannel WB, Abbott RD, Savage DD, McNamara PM. Coronary heart
disease and atrial fibrillation: the Framingham Study. Am Heart J.
1983;106:389-396.
17. Orndahl G, Thulesius O, Hood B. Incidence of persistent atrial fibrillation and conduction defects in coronary heart disease. Am Heart J.
1972;84:120-131.
18. Miyasaka Y, Barnes ME, Gersh BJ, et al. Time trends of ischemic
stroke incidence and mortality in patients diagnosed with first atrial
fibrillation in 1980 to 2000. Report of a Community-Based Study.
Stroke. 2005;36:2362-2366.
19. Benjamin EJ, Wolf PA, D’Agostino RB, et al. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation.
1998;98:946-952.
20. Stewart S, Hart CL, Hole DJ, McMurray JJ. A population-based study
of the long-term risks associated with atrial fibrillation: 20-year follow-up of the Renfrew/Paisley study. Am J Med. 2002;113:359-364.
21. Vidaillet H, Granada JF, Chyou PH, et al. A population-based study of
mortality among patients with atrial fibrillation or flutter. Am J Med.
2002;113:365-370.
22. Gajewski J, Singer RB. Mortality in an insured population with atrial
fibrillation. JAMA. 1981;245:1540-1544.
23. McKee PA, Castelli WP, McNamara PM, Kannel WB. The natural
history of congestive heart failure: the Framingham study. N Engl
J Med. 1971;285:1441-1446.
363.e1
APPENDIX
Definition of Covariates
Chronic versus paroxysmal AF was defined by whether
there were recognizable intervening episodes of sinus
rhythm. Systemic hypertension was defined by a physician’s diagnosis, need for antihypertensive therapy, or systolic blood pressure greater than 140 mm Hg or diastolic
blood pressure greater than 90 mm Hg on at least 2 occasions that were not associated with acute illness or injury.
Dyslipidemia was defined by a total cholesterol level of 200
mg/dL or greater, triglyceride level of 150 mg/dL or greater,
low-density lipoprotein cholesterol level of 130 mg/dL or
greater or high-density lipoprotein cholesterol less than 40
mg/dL on 2 or more occasions, or treatment with lipidlowering agents. Diabetes mellitus was defined by physician’s diagnosis and treatment with insulin or oral hypoglycemic agents. Smoking history was classified as past (⬎6
months prior) or current smoker. Stroke included development of any type of stroke, as defined by clinical documentation of the diagnosis with or without confirmatory findings
on imaging studies. Congestive heart failure was defined by
the presence of 2 major or 1 major and 2 minor Framingham
The American Journal of Medicine, Vol 120, No 4, April 2007
criteria.23 Valvular heart disease was defined by greater than
mild stenosis or regurgitation by echocardiography or prior
valve repair/replacement. Regular alcohol use was defined
by self-reported consumption of more than 1 drink per day
regularly. Chronic renal disease, chronic obstructive pulmonary disease, obstructive sleep apnea, and hyperthyroidism
were defined by these clinical diagnoses in the medical
records.
Prediction Formula for the Probability of
Coronary Ischemic Event
The proportional-hazards model-based probability of having had a coronary ischemic event within year 1, 5, or 10 of
first AF diagnosis, as defined by the multivariate model in
Table 3, is estimated as follows:
PROB ⫽ 1 ⫺ (Pi)eXB,
where XB ⫽ 0.265*(age/10) ⫹ 0.316*{1 if male} ⫹
0.059*(systolic blood pressure/10) ⫹ 0.500*{1 if systemic
hypertension} ⫹ 0.563*{1 if diabetes mellitus} ⫹ 0.446*{1
if peripheral artery disease} ⫹ 0.332*{1 if valvular heart
disease} ⫹ 0.679*{1 if obstructive sleep apnea} ⫹
0.385*{1 if chronic renal disease} ⫺ 3.450, and (P1, P5,
P10) ⫽ (.9608, .8682, .7467).
The American Journal of Medicine (2007) 120, 364-368
CLINICAL RESEARCH STUDY
Outbreak of Pertussis on a College Campus
Allen S. Craig, MD,a,b Seth W. Wright, MD, MPH,c Kathryn M. Edwards, MD,d John W. Greene, MD,d
MaryLou Haynes, RN,c Anthony D. Dake, MD,c William Schaffner, MDe
a
Office of Workforce and Career Development, Centers for Disease Control and Prevention, Atlanta, Ga; bTennessee Department of
Health, Nashville; Departments of cEmergency Medicine, dPediatrics, and ePreventive Medicine, Vanderbilt University School of
Medicine, Nashville, Tenn.
ABSTRACT
BACKGROUND: Pertussis is increasing among adolescents and adults despite universal childhood vaccination. This investigation describes an outbreak of pertussis among undergraduate students and assesses
the burden of cough illness on a college campus.
METHODS: Students presenting with prolonged cough were evaluated with culture, polymerase chain
reaction (PCR), and serology. An e-mail survey was performed to determine the burden of cough illness
on campus.
RESULTS: Thirty-seven undergraduates were evaluated. Their mean duration of cough was 28 days. No
student had cultures positive for B. pertussis; one was PCR positive. Ten (27%) had serologic values
consistent with acute pertussis infection. The e-mail survey was returned by 225/500 (45%) students. Of
these, 66 (29%; 95% confidence interval [CI], 23%-36%) reported a cough of 2 weeks or longer duration
during the fall semester. A conservative estimate showed that the campus-wide incidence of a cough illness
meeting the Centers for Disease Control and Prevention case definition for pertussis was 13% (95% CI,
10%-16%) during the fall semester.
CONCLUSIONS: Adolescents and young adults are susceptible to pertussis infection. This study demonstrates that there was a substantial rate of pertussis infection during an outbreak on a college campus.
Our findings support the routine use of the acellular pertussis vaccine in adolescents and adults. © 2007
Elsevier Inc. All rights reserved.
KEYWORDS: Bordetella pertussis; College health; Respiratory tract infections; Whooping cough
INTRODUCTION
Adolescents and adults are at risk for pertussis infection
because of waning immunity following childhood vaccination.1-4 Pertussis in adolescents and adults is often atypical,
consisting primarily of a prolonged cough. However, the
diagnosis is usually not considered by the patient or the
treating clinician. Studies in a variety of populations have
documented that pertussis is quite common, ranging from
12% to 26% of adolescents and adults with prolonged
cough.1,4-6 Thus, it is not surprising that infections would
occur on college campuses.7 Over a study period of 2½
years, Mink and colleagues8 examined 130 students presentRequests for reprints should be addressed to Allen S. Craig, MD,
Tennessee Department of Health, 425 5th Avenue North, Nashville, TN
37243.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.06.035
ing to the University of California Los Angeles Student
Health Service with cough of 1 week or longer and found
that 26% had laboratory evidence of an acute pertussis
infection.
Pertussis was suspected when clinicians at the Vanderbilt
University Student Health Service noted an increase in the
number of students presenting with a prolonged cough. An
investigation was begun to confirm the presumptive diagnosis of pertussis and to determine the burden of prolonged
cough illness on the campus.
METHODS
Setting
Located in Nashville, Tenn, Vanderbilt University is a private university with over 5800 undergraduate students en-
Craig et al
Pertussis Outbreak on Campus
rolled. A large majority of undergraduates live on campus in
student housing. The Student Health Service, staffed by
physicians and certified nurse practitioners, provides primary care services to the students at no charge. Approval for
this outbreak investigation was given by the Vanderbilt
University Institutional Review
Board.
365
hours, where they were plated on Bordet-Gengou, BordetGengou with penicillin, and buffered charcoal yeast extract
with cephalexin agar at 35°C. They were read over a 7-10
day period. Suspicious colonies were transferred to a second
Bordet-Gengou plate, and DFA was performed. DFA was
also performed on slides streaked
from the original nasopharyngeal
swab and examined using stanCLINICAL SIGNIFICANCE
dard techniques.9 The PCR was
Surveillance for Pertussis
performed at the Centers for DisCase Finding. In mid-October an
● Pertussis is increasing as a cause of
ease Control and Prevention
increase in the number of students
cough illness in adults and adolescents.
(CDC) using simultaneous detecpresenting to the Student Health
●
The
clinical
presentation
of
pertussis
is
tion and discrimination of
Service with cough was noted. A
often
atypical,
consisting
primarily
of
a
B. pertussis and B. parapertussis as
prospective assessment of cough
described by van der Zee et al.10
prolonged
cough.
illness was conducted in collaboSerum samples were tested for
ration with the Tennessee Depart● Clinicians must maintain a high index of
antibodies
to pertussis toxin (PT)
ment of Health during the 4-week
suspicion in patients with a cough lastby
enzyme-linked
immunoassay
period between October 24 and
ing
⬎2
weeks.
(ELISA).
Acute
and
convalescent
November 22 in students with a
samples
were
stored
at
⫺20°C and
cough of 2 weeks or longer. Deanalyzed
concurrently.
A standard
mographic data collected included
reference
human
antiserum
with
the age, sex, smoking status, and immunization history. The
an
assigned
value
of
200
ELISA
units/mL
was
run
with
the
duration of cough at the time of presentation and associated
test
serum
on
each
plate.
The
lower
limit
of
detection
was
2
symptoms, including sore throat, rhinorrhea, fever, sputum
ELISA
units
for
antibody
to
PT;
specimens
with
activity
production, inspiratory whoops, paroxysms of cough, nightbelow the lower limit of detection were assigned a value of
time awakening by cough, and post-tussive emesis, were
1 ELISA unit. Results were expressed in ELISA units per
obtained. Students were asked if they previously had sought
milliliter.
medical care for their cough, if they were already taking
antibiotics for the cough illness, and if they had any known
Case Definition
exposure to a person with a cough of 1 week or longer
Cases of pertussis were defined using the CDC outbreak
duration.
case definition consisting of an unexplained cough illness
Nasopharyngeal swab specimens were obtained for
lasting 2 or more weeks with or without paroxysms of
Bordetella pertussis culture, direct fluorescent antibody ascough, whoops, or post-tussive emesis.11 This case definisay (DFA), and polymerase chain reaction (PCR) from the
tion is commonly used when 5 or more cases are clustered
first 20 students because of limited resources and laboratory
in time or space and, when used during an outbreak, has a
availability. An acute blood specimen was obtained for
sensitivity ranging from 84% to 100% and specificity of
pertussis antibody titers from all students with a prolonged
35% to 90%.12,13
cough. A convalescent sample was obtained approximately
A laboratory-confirmed case consisted of any student
4 to 12 weeks later. The students were followed until the
with
a positive culture, DFA, or PCR assay for B. pertussis,
resolution of the cough.
or serologic studies consistent with acute B. pertussis infection. Serologic assays were considered positive if the PT
Survey of Students. An e-mail survey of a random sample
antibody level showed a 4-fold or greater increase with at
of 500 undergraduate students was performed to assess the
least one value above a cutoff value, or a single PT antibody
extent of cough illness on campus during the outbreak. The
titer was above a cutoff value. The initial sample obtained
students were randomized by computer, based upon the last
was frequently a mid-illness sample.1 The cutoff value for
digit of their computer account number. The e-mail survey
PT antibody was determined from an analysis of 5409
was sent to the students the first week following the Christhealthy volunteers aged 10-49 years who had pertussis semas break. In the survey, students were asked if they had a
rology obtained as part of the Third National Health and
cough of 2 weeks or longer duration during the fall semesNutrition Survey (NHANES III).14 These PT antibody valter, whether they had sought medical care for their cough,
ues were transformed using the log10 function and a mixand whether they had missed school because of the cough.
ture model was developed that assumed a normal distribuLaboratory Methods
tion for the population. A cutoff point was then chosen to
Nasopharyngeal swab specimens were obtained using calattain high model specificity. When the cutoff point was
cium alginate swabs. Swabs were placed in both charcoal
applied to data collected in prospective pertussis surveilagar transport media and saline, and transported to the
lance studies with culture-confirmed disease, the cutoff had
Tennessee Department of Health Laboratory within 24
an 80% sensitivity and 93% specificity.14 A PT antibody
366
The American Journal of Medicine, Vol 120, No 4, April 2007
Table 1 Comparison of Patients with Cough of at Least 2
Weeks Duration Evaluated in the Student Health Service with
and without Pertussis
Laboratory Evidence of Pertussis
Characteristic
Age, mean ⫾ SD
Female, n (%)
Current smoker, n (%)
Duration of cough at
presentation, mean ⫾ SD
number of days
Total duration of cough,
mean ⫾ SD number of
days
Paroxysmal cough, n (%)
Presence of whoops, n (%)
Post-tussive emesis, n (%)
Awakened by cough, n (%)
History of fever, n (%)
Sputum production, n (%)
Sore throat, n (%)
Rhinorrhea, n (%)
Exposed to persistent
cough, n (%)
Yes
(n ⫽ 10)
No
(n ⫽ 27)
P-value
19.8 ⫾ 1.1 19.6 ⫾ 1.1
4 (40)
17 (63)
1 (10)
3 (11)
37.6 ⫾ 19 24.2 ⫾ 13
.62
.27
.99
.02
73.8 ⫾ 31
46.7 ⫾ 21
.02
10
6
4
9
2
8
6
6
8
25
17
3
19
9
19
21
24
13
.99
.99
.07
.39
.69
.69
.41
.07
.14
(100)
(60)
(40)
(90)
(20)
(80)
(60)
(60)
(80)
(93)
(63)
(11)
(70)
(33)
(70)
(78)
(89)
(48)
value of ⱖ94 ELISA units/mL was considered positive for
this investigation. Students lost to follow-up with a low
initial titer and negative PCR or culture were considered in
the analysis to be free of pertussis.
Statistical Analysis
Continuous data were compared with the unpaired Student’s
t test. Categorical data were compared with the Fisher’s
exact test. Statistical analysis was performed with the Statview 4.5 (Abacus Concepts, Inc. Berkeley, CA) and Instat
2.01 (Graphpad Software, San Diego, CA) statistical software packages. All tests were 2-tailed; a P-value of ⱕ.05
was considered significant.
RESULTS
During the 4-week surveillance period, all patients with
cough for 2 weeks or longer who presented to the Student
Health Services were recruited for diagnostic studies; 37
undergraduate students consented. Of these, 32 (86%) returned for follow-up serologic evaluation. The age range of
the 37 students was 18-22 years (median 20 years); 21
(57%) were female. All students reported full childhood
immunization. Eighteen (49%) had been seen previously by
a health care practitioner for this coughing illness and 15
(41%) were taking antibiotics at the time of enrollment. The
mean duration of cough at the time of enrollment was 28
days (range 14-83 days). There was no notable clustering of
cases within a single dormitory, classroom, or extra-curricular activity.
Table 1 compares the characteristics of students with and
without laboratory evidence of pertussis. Students with laboratory evidence of pertussis had a significantly longer
duration of cough at presentation and significantly longer
time to cough resolution. There were no other significant
differences between the groups, including the presence of
paroxysmal cough, whoops, or post-tussive emesis.
None of the 20 students evaluated by nasopharyngeal
swab had a positive culture or DFA smear; one had a
positive PCR test. Ten of the 37 students (27%) had serologic values indicative of an acute pertussis infection; 2 of
these had 4-fold titer increase in PT antibody levels with at
least one value ⱖ94 ELISA units/mL, and the remaining 8
had single antibody levels that equaled or exceeded the
cutoff value of 94 ELISA units/mL (Table 2). The student
with a positive PCR also had serologic evidence diagnostic
of pertussis 18 days after cough onset.
Table 2 Serologic Results of Students Presenting with
Prolonged Cough in ELISA units/mL, n ⫽ 37
Student Number
Acute PT*
Antibody
Convalescent
PT Antibody
Positive
Serology
1
2
3
4†
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
13
621
1
7
119
10
11
38
11
207
6
2
3
2
3
1
58
90
3
67
76
2
1
2
22
1
21
53
7
5
721
11
173
363
2
383
231
13
115
1
174
45
12
7
21
10
55
No
Yes
No
Yes
Yes
No
No
No
No
Yes
No
No
No
No
No
No
No
No
No
Yes
No
No
No
No
No
No
No
No
No
No
Yes
No
Yes
Yes
No
Yes
Yes
*Pertussis toxin.
†Positive polymerase chain reaction.
3
3
2
13
92
3
295
57
1
2
73
1
54
9
6
281
16
104
246
1
218
6
Craig et al
Pertussis Outbreak on Campus
Replies to the student e-mail survey were received from
225 (45%) of the 500 students. Of these, 66 (29%) stated
that they had had a cough of 2 weeks or longer duration
during the fall semester. The severity of the cough was
sufficient that 26 (39%) had sought medical attention for the
cough and 31 (47%) of 66 had missed class because of the
cough. Based on these results, it was estimated that as many
as 1704 (29%; 95% confidence interval [CI], 23%-36%) of
the 5877 total undergraduate students could have met the
CDC case definition for pertussis during the fall semester.
More conservatively, if it is assumed that none of the e-mail
nonresponders had a persistent cough, 776 (13.2%; 95% CI,
10%-16%) students met the case definition.
DISCUSSION
Immunity following childhood pertussis vaccination rapidly
wanes and outbreaks of pertussis occur even in populations
with high rates of childhood immunization.15,16 Adolescents
appear to be at particular risk for infection and outbreaks
with attack rates ⬎20% have occurred in high school settings.16 Sporadic cases in adults and adolescents are also
quite common but are rarely diagnosed. For instance, Cromer and colleagues followed a cohort of adolescents for 5
years and found that the annual incidence of serologically
defined pertussis infection was 6.1%.3 None were diagnosed
with clinical pertussis during that interval.
The high frequency of laboratory-diagnosed pertussis in
students presenting to the Student Health Center for treatment of prolonged cough illness, along with the results of
the e-mail survey, support the conclusion that the observed
cases were part of an outbreak of pertussis rather than
recognition of endemic cases. Over one quarter of the patients with cough presenting to the Student Health Service
and evaluated for pertussis during the 4-week surveillance
period had laboratory evidence of a pertussis infection. Our
27% occurrence rate may also be somewhat conservative,
because some students had received antibiotics before specimens were obtained and students who did not return for
follow-up serology in our study may have been misdiagnosed as being disease free.
The students with laboratory evidence of pertussis were
clinically quite similar to the patients who did not meet the
definition, reinforcing the difficulty facing clinicians attempting to distinguish pertussis from other causes of persistent cough, as well as the limitations of the current diagnostic methods. The largest difference was in the overall
time until resolution of the cough. Patients with laboratory
evidence of confirmed pertussis coughed, on average, about
27 days longer than those without laboratory-confirmed
pertussis.
The results of the e-mail survey indicated that prolonged
cough illness was widespread on the campus during the
period of the outbreak. During an outbreak, the CDC defines
a clinical case of pertussis as a cough of 2 weeks duration or
longer, with or without paroxysms of cough, whoops, or
post-tussive emesis.11 Almost 30% of the survey responders
367
reported having had a cough illness that met this definition
during the fall semester when it was documented that pertussis was on campus. The severity of the cough also appeared to be noteworthy, as over one-third of the affected
students had sought medical care and almost half had
missed class because of the cough. A conservative estimate
indicated that between 10% and 16% of the student body
may have had pertussis during the fall semester.
The diagnosis of pertussis in adolescents and adults is
difficult. Adults generally do not present for care until they
have been coughing for a prolonged period and their cultures are often negative at that point in their illness.1,4,17-19
None of the 20 students evaluated had a positive culture,
which is not surprising given its low sensitivity of 1%-15%,
particularly after the first week of cough.16,17 Serologic
diagnosis may be useful, although it is not widely available,
is cumbersome, and interpretation of results is problematic
as there are no universally accepted diagnostic criteria.20
The “acute” titer is usually obtained mid-illness and subsequent 4-fold increases in antibody are relatively uncommon;
sensitivity of serologic testing has been reported to be 80%
and is less dependent on timing of the specimen.1,14 Indeed,
it is common for the follow-up titer to fall rather than
increase.1 In this investigation we rapidly analyzed serum
specimens for pertussis toxin antibodies on the first individuals we evaluated and were able to confirm within several
days that pertussis was present. The availability of a research laboratory within the Vanderbilt University Medical
Center with an established pertussis serologic assay greatly
facilitated the investigation. PCR can also be a useful diagnostic tool, but it is not currently commercially available,
has decreased sensitivity after 3 weeks of cough illness, and
its sensitivity (42%-61%) and specificity (88%-97%) vary
greatly among laboratories and are subject to differences in
test interpretation.21,22 Only 1 of the 20 tested subjects had
a positive PCR, although 7 of the 20 had serologic evidence
of pertussis. In the past, DFA had been widely performed in
outbreak settings, however, the usefulness of the test is
limited by difficulty in interpretation and cross-reactions
with other antigens.17 None of the DFA studies in this
investigation were positive.
Although antimicrobial prophylaxis with erythromycin
or azithromycin during an outbreak appears to prevent
transmission of the organism,15,16 it was not deemed practical to implement widespread chemoprophylaxis in this
outbreak because the disease involved virtually the entire
undergraduate body and there was an upcoming winter
break. However, prophylaxis of roommates and other close
contacts of infected individuals was undertaken and should
be considered in similar situations.
Adolescents and adults traditionally have not been revaccinated since the last pertussis booster dose delivered at 4 to
6 years of age. Acellular vaccines have been shown to be
safe and immunogenic in adolescents and adults and have
recently been licensed for use in older populations.23 The
CDC’s Advisory Committee on Immunization Practices
368
(ACIP) has recently made new recommendations for a routine preadolescent booster dose of acellular pertussis vaccine (combined with diphtheria and tetanus toxoid [Tdap])
at age 11-12 years and at the time of the routine decennial
booster in adults. As shown by our outbreak, routine vaccination of adolescents and adults with Tdap is clearly
warranted. Pertussis in adults and adolescents is usually not
life-threatening, but the cough is typically prolonged, and
infected people seek medical care and may lose time from
work or school.24 An additional reason for vaccination of
adolescents and adults is that they are thought to be a
common source of infection in infants and children.25,26
ACKNOWLEDGMENTS
We thank Carol Worthington, Tennessee Department of
Health, Laboratory Services, for performing pertussis cultures and direct fluorescent antibody testing; Gary N.
Sanden, Bacterial and Mycotic Diseases, National Center
for Infectious Diseases, Centers for Disease Control and
Prevention, for performing PCR testing; and Peter M.
Strebel, National Immunization Program, Centers for Disease Control and Prevention, for advice.
References
1. Wright SW, Edwards KM, Decker MD, Zeldin MH. Pertussis infection
in adults with persistent cough. JAMA. 1995;273:1044-1046.
2. Guris D, Strebel PM, Bardenheier B, et al. Changing epidemiology of
pertussis in the United States: increasing reported incidence among
adolescents and adults, 1990-1996. Clin Infect Dis. 1999;28:12301237.
3. Cromer BA, Goydos J, Hackell J, et al. Unrecognized pertussis infection in adolescents. Am J Dis Child. 1993;147:575-577.
4. Nennig ME, Shinefield HR, Edwards KM, et al. Prevalence and incidence of adult pertussis in an urban population. JAMA. 1996;275:
1672-1674.
5. Robertson PW, Goldberg H, Jarvie BH, et al. Bordetella pertussis
infection: a cause of persistent cough in adults. Med J Aust. 1987;146:
522-525.
6. Schmitt-Grohe S, Cherry JD, Heininger U, et al. Pertussis in German
adults. Clin Infect Dis. 1995;21:860-866.
7. Jackson LA, Cherry JD, Wang SP, Grayston JT. Frequency of serological evidence of Bordetella infections and mixed infections with
other respiratory pathogens in university students with cough illnesses.
Clin Infect Dis. 2000;31:3-6.
8. Mink CM, Cherry JD, Christenson P, et al. A search for Bordetella
pertussis infection in university students. Clin Infect Dis. 1992;14:
464-471.
9. Murray PR, Baron EJ, Pfaller ME, et al. Manual of Clinical Microbiology. Washington, DC: American Society for Microbiology; 1995.
The American Journal of Medicine, Vol 120, No 4, April 2007
10. van der Zee A, Agterberg C, Peeters M, et al. Polymerase chain
reaction assay for pertussis: simultaneous detection and discrimination
of Bordetella pertussis and Bordetella parapertussis. J Clin Microbiol.
1993;31:2134-2140.
11. Case definitions for infectious conditions under public health surveillance. Centers for Disease Control and Prevention. MMWR Recomm
Rep. 1997;46:1-55.
12. Strebel PM, Cochi SL, Farizo KM, et al. Pertussis in Missouri: evaluation of nasopharyngeal culture, direct fluorescent antibody testing,
and clinical case definitions in the diagnosis of pertussis. Clin Infect
Dis. 1993;16:276-285.
13. Patriarca PA, Biellik RJ, Sanden G, et al. Sensitivity and specificity of
clinical case definitions for pertussis. Am J Public Health. 1988;78:
833-836.
14. Baughman AL, Bisgard KM, Edwards KM, et al. Establishment of
diagnostic cutoff points for levels of serum antibodies to pertussis
toxin, filamentous hemagglutinin, and fimbriae in adolescents and
adults in the United States. Clin Diagn Lab Immunol. 2004;11:
1045-1053.
15. Christie CD, Glover AM, Willke MJ, Marx ML, Reising SF, Hutchinson NM. Containment of pertussis in the regional pediatric hospital
during the Greater Cincinnati epidemic of 1993. Infect Control Hosp
Epidemiol. 1995;16:556-563.
16. Pertussis outbreaks—Massachusetts and Maryland, 1992. MMWR
Morb Mortal Wkly Rep. 1993;42:197-200.
17. Mattoo S, Cherry JD. Molecular pathogenesis, epidemiology, and
clinical manifestations of respiratory infections due to Bordetella pertussis and other Bordetella subspecies. Clin Microbiol Rev. 2005;18:
326-382.
18. Dworkin MS. Adults are whooping, but are internists listening? Ann
Intern Med. 2005;142:832-835.
19. Senzilet LD, Halperin SA, Spika JS, et al. Pertussis is a frequent cause
of prolonged cough illness in adults and adolescents. Clin Infect Dis.
2001;32:1691-1697.
20. Hewlett EL, Edwards KM. Clinical practice. Pertussis—not just for
kids. N Engl J Med. 2005;352:1215-1222.
21. Heininger U, Schmidt-Schlapfer G, Cherry JD, Stehr K. Clinical validation of a polymerase chain reaction assay for the diagnosis of
pertussis by comparison with serology, culture, and symptoms during
a large pertussis vaccine efficacy trial. Pediatrics. 2000;105:E31.
22. Lievano FA, Reynolds MA, Waring AL, et al. Issues associated with
and recommendations for using PCR to detect outbreaks of pertussis.
J Clin Microbiol. 2002;40:2801-2805.
23. Ward JI, Cherry JD, Chang SJ, et al. Efficacy of an acellular pertussis
vaccine among adolescents and adults. N Engl J Med. 2005;353:
1555-1563.
24. Lee GM, Lett S, Schauer S, et al. Societal costs and morbidity of
pertussis in adolescents and adults. Clin Infect Dis. 2004;39:
1572-1580.
25. Edwards KM. Overview of pertussis: focus on epidemiology, sources
of infection, and long term protection after infant vaccination. Pediatr
Infect Dis J. 2005;24:S104-S108.
26. Halperin SA. Pertussis—a disease and vaccine for all ages. N Engl
J Med. 2005;353:1615-1617.
The American Journal of Medicine (2007) 120, 369.e1-369.e7
CLINICAL RESEARCH STUDY
Prognostic Stratification of Patients with Left-Sided
Endocarditis Determined at Admission
José Alberto San Román, MD, PhD,a Javier López, MD,a Isidre Vilacosta, MD, PhD,b María Luaces, MD,b
Cristina Sarriá, MD, PhD,c Ana Revilla, MD,a Ricardo Ronderos, MD,d Walter Stoermann, MD,e Itziar Gómez, MD,a
Francisco Fernández-Avilés, MD, PhDa
a
c
Instituto de Ciencias del Corazón, Hospital Clínico Universitario, Valladolid, Spain; bHospital Clínico San Carlos, Madrid, Spain;
Hospital la Princesa, Madrid, Spain; dHospital San Juan de Dios, La Plata, Argentina; eCIMAC, San Juan, Argentina.
ABSTRACT
BACKGROUND: The prognosis of patients with left-sided endocarditis remains poor despite the progress
of surgical techniques. Identification of high-risk patients within the first days after admission to the
hospital would permit a more aggressive therapeutic approach.
METHODS: We designed a prospective multicenter study to find out the clinical, microbiologic, and
echocardiographic characteristics available within 72 hours of admission that might define the profile of
high-risk patients. Of 444 episodes, 317 left-sided endocarditis cases were included and 76 variables were
assessed. Events were surgery in the active phase of the disease and in-hospital death. A stepwise logistic
regression analysis was undertaken to determine variables predictive of events.
RESULTS: Multivariate analysis of the clinical variables found to have statistical significance in the
univariate analysis identified the following as predictive: patient referred from another hospital (odds ratio
[OR]: 1.8; confidence interval [CI], 1.1-2.9), atrioventricular block (OR: 2.5; CI, 1.1-5.9), acute onset (OR:
1.7; CI, 1.1-2.9), and heart failure at admission (OR: 2.3; CI, 1.4-3.8). When the echocardiographic and
microbiological variables statistically significant in the univariate analysis were introduced, the presence
of heart failure at admission (OR: 2.9; CI, 1.8-4.8), periannular complications (OR: 1.8; CI, 1.1-3.1), and
Staphylococcus aureus infection (OR: 2.0; CI, 1.1-3.8) retained prognostic power. Risk could be accurately
stratified when combining the 3 variables with predictive power: 0 variables present: 25% of risk; 1
variable present: 38% to 49% of risk; 2 variables present: 56% to 66% of risk; and 3 variables present: 79%
of risk.
CONCLUSIONS: The risk of patients with left-sided endocarditis can be accurately stratified with the
assessment of variables easily available within 72 hours of admission to the hospital. © 2007 Elsevier Inc.
All rights reserved.
KEYWORDS: Left-sided endocarditis; Prognosis; Risk stratification
Every physician accustomed to treating patients with leftsided endocarditis is aware not only of its devastating nature
but also of its uncertain prognosis. Despite profound knowledge of the disease and a strictly correct management ac-
The present study was financed in part by the Red de Centros Cardiovasculares, which is supported by the Instituto de Salud Carlos III.
Requests for reprints should be addressed to José Alberto San Román
Calvar, Instituto de Ciencias del Corazón (ICICOR), Hospital Clínico
Universitario, C/ Ramón y Cajal 3, 47005 Valladolid.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.05.071
cording to accepted guidelines,1,2 a feeling of frustration
predominates if rates of mortality are rigorously examined.
Before the inclusion of antibiotics in the therapeutic armamentarium, left-sided endocarditis was almost always a
fatal disease. With antibiotics, uncontrolled infection decreased significantly and hemodynamic deterioration
emerged as the main cause of death. Two decades ago,
surgery in the active phase of the disease in patients who
met strict criteria was introduced in the therapeutic approach, and death again decreased.3,4 Since then, mortality
rates remain unchanged.3-8 At this point, different therapeu-
369.e2
The American Journal of Medicine, Vol 120, No 4, April 2007
tic options must be tested with the aim of improving progECHOCARDIOGRAPHY
nosis. We hypothesize that an early aggressive surgical
All patients underwent transthoracic and transesophageal
therapeutic approach in patients with poor prognosis would
echocardiography. After introduction of the probe into the
be beneficial. To test this hypothesis, we first have to idenesophagus, a systematic approach to cardiac structures was
tify the profile of patients with poor prognosis within the
used. We focused on special areas of interest, namely,
initial 72 hours of admission to the
native cardiac valves and their anhospital. Bearing the preceding
nuli, prosthesis and their annuli,
considerations in mind, we have unand mitroaortic continuity. DoppCLINICAL SIGNIFICANCE
dertaken a prospective multicenter
ler was used to interrogate the hestudy to identify early epidemiomodynamic status of the valves,
● Our model can help clinicians care for
logic, clinical, electrocardiographic,
presence or absence of flow within
patients with endocarditis and predict
radiographic, analytic, microbiologthe cavities, and communication
their risk.
ical, and echocardiographic predicbetween chambers. From an echo● The 3 cornerstones in the diagnosis of
tors of poor prognosis.
cardiographic perspective, a vegeendocarditis are pivotal regarding progtation was defined as a thrombusnosis:
clinical
examination,
blood
cullike mass with shaggy echoes and
PATIENT POPULATION
erratic motion independent of that
tures,
and
echocardiography.
In 1996, 5 tertiary care centers
of the valve. The vegetation was
with surgical facilities initiated a
● Patients with heart failure, periannular
measured in various planes. The
prospective study. For every pacomplications, and/or Staphylococcus
maximal diameter and area were
tient, a standardized case report
aureus infection should be closely
used for subsequent analysis. In
form with 18 epidemiologic, 8
followed.
case of multiple vegetations, the
clinical, 10 analytic, 4 radiolargest was measured. Periannular
graphic, 6 electrocardiographic,
● Hospitals without surgical facilities
complications were defined as fol14 microbiological, and 16 echoshould refer these patients to tertiary
lows:11-13 abscess, well-delineated
cardiographic variables was comcare centers without delay.
perivalvular area of reduced
pleted. Between 1996 and 2003,
echodensity with no flow; pseudo●
A
multidisciplinary
team
including
car441 patients who met Duke criteaneurysm, echo-free perivalvular
9
diac surgeons should take care of these
ria (406 with definite and 35 with
pouch with flow in its interior; and
patients.
possible criteria) were included.
fistula, narrow communication beOf these patients, 333 had lefttween 2 adjacent chambers. Once
sided endocarditis. Sixteen paa periannular lesion was seen, it
tients with septic shock at admiswas further analyzed by subtle movements of the transducer
sion were excluded from further analysis to avoid an
to better determine its morphologic characteristics.
obvious bias because it was an absolute indication for urgent surgery. The remaining 317 patients made up our study
population. The characteristics of these patients are shown
DEFINITION OF EVENTS
in Table 1.
Death and surgery in the active phase of the disease were
To ensure consecutive enrollment, all patients who unregarded as events. Any death occurring during hospital stay
derwent echocardiography to rule out endocarditis were
was considered an event irrespective of the final cause.
followed until a diagnosis was established. In patients with
Surgery in the active phase, that is, before adequate antibia final diagnosis of endocarditis, clinical data at admission
otic treatment was completed according to established
were collected. Patients were interviewed regarding their
guidelines,1,2 was performed when any of the following
past and current clinical history. All patients underwent a
occurred: left-sided heart failure unresponsive to maximal
detailed history, standard physical examination, electrocarmedical treatment; septic shock; or uncontrolled infection
diogram, blood analysis, urine analysis, set of 3 blood culdefined as persistent bacteriemia or fever lasting more than
tures, and transthoracic and transesophageal echocardiogra10
7 days after starting adequate antibiotic treatment when
phy. All were done within 24 hours of admission except
other causes had been excluded. The isolated echocardioechocardiography, which depends on expert availability; in
graphic finding of a periannular complication (abscess,
any case, it was performed within 48 hours. Blood culture
pseudoaneurysm, fistula) was not considered an absolute
results were reported within 72 hours. Therefore, all variindication for surgery.
ables entered in the statistical model were available within
72 hours of admission.
Patients referred from another institution were not sent
DEFINITION OF TERMS
straight to surgery but were evaluated and treated according
to our protocol. They underwent operation only when they
Acute onset was defined as the span of time between the
met any of the 3 requirements described afterward.
onset of symptoms and the admission to the hospital less
San Román et al
Table 1
369.e3
Clinical Characteristics of the 317 Patients
Age (y)
Male gender
Referred
Nosocomial acquisition
Previous cardiopathy
Degenerative
Prosthesis
Rheumatic
Comorbid conditions
Diabetes mellitus
Cancer
Chronic anemia
Renal insufficiency
Dermatopathy
Intravenous drug users
Immunocompromised state
HIV
Previous endocarditis
Clinical manifestations:
Acute onset
Fever
Dyspnea
Heart failure
Abnormal skin findings
Splenomegaly
Renal insufficiency
Stroke
Hematologic findings:
Serum creatinine (mg/dL)
Blood urea nitrogen (mg/dL)
Hemoglobin (g/dL)
Hematocrit (%)
WCC ⫻ 103/mm3
WCC ⬎ 103/␮L
Sedimentation globular rate (mm)
Platelets ⫻ 103/mm3
Urine findings
Hematuria
Proteinuria
Radiographic findings
Cardiac enlargement
Septic emboli
Pleural effusion
Pulmonary congestion
Electrocardiographic findings
Atrioventricular block
Left bundle-branch block
Right bundle-branch block
Atrial fibrillation
Acute myocardial infarction
Supraventricular tachycardia
Total
(n ⫽ 317)
No Events
(n ⫽ 187)
Events
(n ⫽ 130)
P Value
57 ⫾ 16
209 (66%)
118 (28%)
89 (29%)
202 (65%)
29 (9%)
124 (40%)
32 (10%)
57 ⫾ 16
128 (68%)
57 (31%)
45 (25%)
121 (66%)
16 (9%)
72 (39%)
17 (9%)
58 ⫾ 16
81 (62%)
61 (47%)
34 (43%)
81 (64%)
13 (10%)
52 (41%)
15 (12%)
.82
.26
.003
.08
.72
.65
.76
.47
50
21
36
22
5
16
21
10
28
(16%)
(7%)
(11%)
(7%)
(2%)
(5%)
(7%)
(3%)
(9%)
30
9
16
10
4
11
10
5
16
(16%)
(5%)
(9%)
(5%)
(2%)
(6%)
(5%)
(3%)
(9%)
20
12
20
12
1
5
11
5
12
(15%)
(9%)
(15%)
(9%)
(1%)
(4%)
(9%)
(4%)
(9%)
.86
.19
.06
.27
.65
.58
.39
.75
.80
133
223
150
126
45
35
43
30
(42%)
(70%)
(48%)
(40%)
(14%)
(11%)
(14%)
(9%)
70
135
78
61
28
22
23
13
(39%)
(72%)
(42%)
(33%)
(15%)
(12%)
(12%)
(7%)
63
88
72
69
17
13
20
17
(49%)
(68%)
(56%)
(53%)
(13%)
(10%)
(15%)
(13%)
.08
.39
.01
⬍.001
.64
.62
.43
.07
1.49 ⫾ 1.19
55 ⫾ 38
11.2 ⫾ 2.1
33.8 ⫾ 5.9
12.2 ⫾ 5.3
79 (62%)
67 ⫾ 31
207 ⫾ 10
.94
.78
.50
.46
.59
.50
.23
.15
.02
.17
1.4 ⫾ 1.6
54 ⫾ 42
11.2 ⫾ 2.1
34 ⫾ 6
11.9 ⫾ 7.1
185 (59%)
63.5 ⫾ 32
216 ⫾ 94
1.5 ⫾ 1.8
53 ⫾ 45
11.3 ⫾ 2.1
34.3 ⫾ 6.1
11.7 ⫾ 8.1
106 (58%)
62 ⫾ 32
222 ⫾ 87
135 (51%)
56 (21%)
90 (57%)
29 (18%)
45 (43%)
27 (25%)
177
1
83
120
(56%)
(0.03%)
(26%)
(38%)
95
0
48
51
(51%)
(0%)
(26%)
(27%)
82
2
35
69
(64%)
(1.5%)
(27%)
(54%)
.02
.17
.74
⬍.001
29
13
6
35
1
10
(9%)
(4%)
(2%)
(11%)
(0.03%)
(3%)
12
5
4
20
1
5
(7%)
(3%)
(2%)
(11%)
(0.5%)
(3%)
17
8
2
15
0
5
(13%)
(6%)
(1.5%)
(12%)
(0%)
(4%)
.04
.21
.99
.82
.99
.75
HIV ⫽ human immunodeficiency virus; WCC ⫽ white cell count; SGR ⫽ sedimentation globular rate.
Univariate analysis comparing the event and no event group. Data are presented as number of patients (%) or mean ⫾ standard deviation.
than 15 days. Antibiotic treatment was considered adequate
when high-dose intravenous antibiotic combinations known
to be bactericidal in vitro against the isolated microorganisms were used. Empiric antibiotic regimens were chosen
for culture-negative cases. Renal failure was defined as
serum creatinine equal to or higher than 2 mg/dL. Atrio-
ventricular block included first-, second-, and third-degree block. A diagnosis of heart failure was established
according to accepted criteria.14 Elective surgery was
defined when a patient underwent surgery after the antibiotic regimen was completed. The latter was not considered an event.
369.e4
STATISTICAL ANALYSIS
The statistical analysis included a comparison between 2
groups of patients, divided according to the presence or
absence of events. Continuous variables were presented as
mean value ⫾ standard deviation, and the groups were compared by a 2-tailed Student t test or Wilcoxon rank-sum test.
Categoric variables were expressed as a frequency and a
percentage and were compared by chi-square test and Fisher
exact test when necessary. In patients with more than one
event, only the first one was considered. Multivariate analysis considering events as the dependent variable was performed by a logistic regression model by means of a backward stepwise method. In consecutive steps, variables that
were statistically significant in the univariate analysis, and
others considered clinically relevant, were included in the
logistic regression. First, we included clinical variables,
then microbiologic variables, and finally, the echocardiographic variables were added to the model. No more than 1
variable per 10 outcome events was entered in the logistic
models to avoid overfitting. The goodness-of-fit of the final
model was examined by using the Hosmer-Lemeshow test
and model discrimination with the concordance index (Cindex). The odds ratio (OR) and 95% confidence intervals
(CIs) for each variable were expressed. All reported P
values are 2-sided. Significance was set at a P value of less
than .05. The Statistical Package for the Social Sciences
statistical software package (version 11.0, SPSS Inc, Chicago, Ill) was used.
RESULTS
Events
A total of 130 patients had an event before discharge; 65
died and 65 underwent operation in the active phase. The
more frequent causes of death were uncontrolled infection
(n ⫽ 24) and heart failure unresponsive to maximal medical
treatment (n ⫽ 11). A total of 41 of 65 patients whose event
was death did not go to surgery because the operative risk
was considered prohibitive after a careful assessment was
carried out by a multidisciplinary team composed of surgeons, clinicians, and experts in endocarditis. The remaining 24 died in the following days of elective surgery.
Univariate and Multivariate Analysis of
Epidemiologic, Analytic, Electrocardiographic,
Radiographic, and Clinical Variables
The univariate analysis identified 1 epidemiologic, 2 clinical, 1 electrocardiographic, 1 analytic, and 2 radiographic
variables significantly more frequently in patients with an
event (Table 1). Multivariate analysis of epidemiologic,
clinical, radiographic, electrocardiographic, and analytic
variables (Table 2) showed that only the following retained
statistical significance: referral from another institution,
acute onset, atrioventricular block, and heart failure at
admission.
The American Journal of Medicine, Vol 120, No 4, April 2007
Table 2 Multivariate Analysis of Epidemiologic, Clinical,
Electrocardiograhsa⬎⫽ odds ratio; CI ⫽ confidence interval.
Multivariate Analysis Adding the
Microbiologic and Echocardiographic Variables
The statistical univariate analysis of the microbiological and
echocardiographic variables is shown in Table 3. After
adding the microbiologic and echocardiographic variables
found to be significant in the univariate analysis to the
multivariate statistical model, only 3 variables retained
prognostic power: heart failure at admission, periannular
complications, and Staphylococcus aureus as the causative
microorganism (Table 4). The Hosmer-Lemeshow goodness-of-fit test yielded a P value of .81. Calculation of
model discrimination by using the concordance index was
0.70 (CI 95%, 0.63-0.75).
To rule out that our results are biased by the fact that ours
are tertiary care centers with many referred patients, we
analyzed referred and nonreferred patients separately by
using the same statistical model. We found the same predictors of poor outcome. Of note, heart failure was the most
powerful in the referred population (OR 6.6, CI, 2.7-16.2)
and periannular complications in the nonreferred population
(OR 2.5, CI, 1.3-5).
Stratification of Risk
The factors encountered to bear prognostic importance in
the multivariate analysis were combined to stratify risk.
Patients without heart failure, periannular complications, or
Staphylococcus aureus infection had the best prognosis with
a 25% probability of having an event. If 1 of these factors
was present, the risk increased to 38% to 49%. Risk was
56% to 66% in patients with 2 factors, and it increased to
79% when the 3 factors were present (Figure 1).
DISCUSSION
Our results show that the risk of patients with left-sided
endocarditis can be stratified accurately with the assessment
of variables easily available within 72 hours of admission to
the hospital. Some practical consequences can be drawn
from this study. First, the 3 cornerstones in the diagnosis of
San Román et al
Table 3
369.e5
Baseline Microbiologic and Echocardiographic Characteristics of the 317 Patients
Positive blood cultures at
admission
Streptococcus viridans
Streptococcus bovis
Other Streptococci
Staphylococcus aureus
Methicillin-resistant
Methicillin-sensitive
Coagulase-negative Staphylococci
HACEK
Enterococci
Polymicrobial
Fungi
Anaerobes
Gram-negative bacilli
Negative cultures
Others
Multivalvular
Prosthetic
Valve involved
Aortic native valve
Mitral native valve
Aortic mechanical prothesis
Mitral mechanical prothesis
Mitral bioprosthesis
Aortic bioprosthesis
Vegetations
Major diameter (mm)
Area (cm2)
Periannular complications:
Abscess
Pseudoaneurysm
Fistula
Valve perforation
Ejection fraction
Total
(n ⫽ 317)
No events
(n ⫽ 187)
Events
(n ⫽ 130)
214 (67%)
127 (67%)
87 (67%)
.91
18
5
7
29
27
2
21
1
12
3
4
1
6
20
2
30
51
.20
.89
.61
.03
54
14
21
53
47
6
48
1
25
16
6
3
15
48
13
51
114
(17%)
(4%)
(7%)
(17%)
(15%)
(2%)
(15%)
(0.03%)
(8%)
(5%)
(2%)
(1%)
(5%)
(15%)
(4%)
(16%)
(36%)
110 (35%)
95 (30%)
36 (11%)
55 (17%)
16 (5%)
10 (3%)
238 (79%)
13.9 ⫾ 7.7
90 (54%)
90 (28%)
39 (14%)
37 (13%)
29 (10%)
28 (9%)
62 ⫾ 12
36 (20%)
9 (5%)
14 (8%)
24 (13%)
20 (11%)
4 (2%)
27 (15%)
0
13 (7%)
13 (7%)
2 (1%)
2 (1%)
9 (5%)
28 (15%)
8 (4%)
21 (12%)
63 (34%)
59 (32%)
65 (35%)
17 (9%)
35 (19%)
30 (18%)
7 (4%)
137 (77%)
13.4 ⫾ 7.4
1.1 ⫾ 1.3
35 (23%)
13 (8%)
17 (11%)
16 (10%)
20 (11%)
62 ⫾ 12
(14%)
(4%)
(5%)
(23%)
(21%)
(2%)
(16%)
(1%)
(9%)
(2%)
(3%)
(1%)
(5%)
(16%)
(2%)
(22%)
(39%)
51 (39%)
30 (23%)
19 (15%)
20 (15%)
25 (17%)
9 (7%)
101 (82%)
14.9 ⫾ 8.1
1.7 ⫾ 1.1
46 (37%)
26 (21%)
20 (16%)
13 (11%)
8 (7%)
62 ⫾ 12
P value
.392
.68
.41
.60
.11
.23
.99
.99
.93
.21
.09
.31
.16
.03
.13
.44
.80
.31
.35
.14
.08
.007
.002
.20
.95
.18
.88
HACEK ⫽ Haemophilus, Actinobacillus, Cardiobacterium, Eikenella, Kingella.
Univariate analysis comparing the event and no event group. Data are presented as number of patients (%) or mean ⫾ standard deviation.
endocarditis10 are also pivotal regarding prognosis: clinical
examination, blood cultures, and echocardiography. Second, patients with heart failure, periannular complications,
and/or Staphylococcus aureus infection should be closely
followed. Third, hospitals without surgical facilities should
refer these patients to tertiary care centers without delay.
Finally, a multidisciplinary team including cardiac surgeons
should take care of these patients because it can be anticipated that surgery will be needed as part of the therapeutic
strategy in many of them.
We demonstrate that risk can be accurately, easily, and
rapidly stratified from simple clinical, microbiological,
and echocardiographic variables; the more variables
present, the higher the risk. This could help clinicians in
one of the most challenging decisions in clinical practice,
which is to choose between a watchful and active waiting
or a combined medico-surgical approach in patients with
active endocarditis. Unquestionably, this decision must
be individualized, and many factors that cannot be trans-
Table 4 Stepwise Multivariate Analysis Including All
Variables in the 317 patients with Left-sided Endocarditis
OR ⫽ odds ratio; CI ⫽ confidence interval.
369.e6
The American Journal of Medicine, Vol 120, No 4, April 2007
Figure Risk stratification according to the presence of variables found to be significant in the multivariate analysis. HF ⫽ heart failure;
PAC ⫽ periannular complications; SA ⫽ Staphylococcus aureus.
lated into a statistical model must be taken into account.
In this sense, it must be emphasized that “there is still as
much art as science in the care of patients with endocarditis.”15 We hope that our experience could help to tip the
balance in favor of science.
The inclusion of surgery in the active phase of the
disease as an event must be argued. Our protocol established a tight criteria for a patient to be sent to surgery
before finishing the antibiotic treatment. These criteria
were based on studies that demonstrate a grim prognosis
if surgery is not carried out7,16-18; in other words, had
those patients not underwent operation, they would have
died in all likelihood.
Other investigations dealing with endocarditis have attempted to identify predictors of poor prognosis during
hospital stay and are in agreement with our results.7,19,20
Hasburn and colleagues7 developed a weighted scoring system with 5 independent predictors of 6-month mortality
applicable to patients with complicated left-sided native
valve endocarditis. They identified the following prognostic
factors: Charlson comorbidity scale score of 2 or greater,
altered mental status, congestive heart failure, infection due
to an organism other than Streptococcus viridans, or no
surgery. According to the number of these variables present,
they classified patients in 4 groups with mortality rates of
5%, 15%, 31%, and 59%. This classification system worked
well in predicting mortality in a subsequent validation cohort group.
Our work is unique for the following reasons: Our patients were consecutively and prospectively included in the
study. All patients underwent transesophageal echocardiography. A high number of variables were analyzed. A total of
76 variables have been introduced in the statistical analysis,
including the majority of those previously found to be
predictive of poor prognosis. Only variables available
within 72 hours of admission were included for analysis.
The statistical model was designed with the purpose of
mirroring the clinical approach to a patient. Thus, we designed the statistical analysis in the same manner we are
confronted with a real patient. Most other studies consider
variables altogether, forgetting that in clinical practice all
those variables come available to the physician step by step.
We first take the clinical history and examine the patient;
then, we obtain an electrocardiogram and a blood and urine
analysis, and in the following hours we have the results of
blood cultures and transesophageal echocardiography
available.
Other authors have attempted to establish the profile of
patients with a high probability of having events including
a longer follow-up.8,14,18,21-24 Although useful for the management of patients, these studies do not help in the decision-making process in the early phase of the disease. Moreover, their results are heterogeneous given the differences in
the recruitment of patients (right- and left-sided endocarditis,5,7,18,19,22,25-32 complicated,27,33-35 or noncomplicated
endocarditis), follow-up period (from hospitalization to
20 years36), variables included,32,37-40 nature of the study
(retrospective5,7,17,18,26,32,41 or prospective), and percentage
of patients who underwent transesophageal echocardiography (43%42 to 90%28).
Of course, this is just one more step in the fight against
endocarditis. The prognostic model we developed must be
prospectively validated in an independent population. Although several nonrandomized investigations suggest that
an aggressive medico-surgical strategy could improve prognosis, this will need to be proven in a large randomized
study.
San Román et al
References
1. Bonow R, Carabello B, Leon AC, et al. ACC/AHA guidelines for the
management of patients with valvular heart disease. Executive summary. A report of the American College of Cardiology/American
Heart Association task force on practice guidelines (committee on
management of patients with valvular heart disease). J Am Coll Cardiol. 1998;32:1486-1582.
2. Horstkotte D, Piper C, Vogt J, et al. The Task Force on Infective
Endocarditis of the European Society of Cardiology. Guidelines on
Prevention, Diagnosis and Treatment of Infective endocarditis. Eur
Heart J. 2004;25:267-276.
3. Durack DT. Evaluating and optimizing outcomes of surgery for endocarditis. JAMA. 2003;290:3250-3251.
4. Oakley CM. The clinical spectrum and prognosis of native valve
endocarditis in non-addicts. Eur Heart J. 1995;16:1454-1455.
5. Wallace S, Walton B, Kharbanda R, et al. Mortality from infective
endocarditis: clinical predictors of outcome.Heart. 2002;88:53-60.
6. Erbel R, Liu F, Rohmann J, Kupferwasser I. Identification of high-risk
subgroups in infective endocarditis and the role of echocardiography.
Eur Heart J. 1995;16:588-602.
7. Hasburn R, Vikram H, Barakat L, et al. Complicated left-sided native
valve endocarditis. Risk classification for mortality.JAMA. 2003;289:
1933-1940.
8. Bouza E, Menasalvas A, Muñoz P, et al. Infective endocarditis: a
prospective study at the end of the twentieth century. Medicine. 2001;
80:298-307.
9. Durack DT, Lukes AS, Bright DK. The Duke Endocarditis Service:
new criteria for the diagosis of infective endocarditis: utilization of
specific echocardiographic findings. Am J Med. 1994;96:200-209.
10. Bayer A, Bolger A, Tauber K, et al. Diagnosis and management of
infective endocarditis and its complications. Circulation. 1998;98:
2936-2948.
11. San Román JA, Vilacosta I, Sarriá C, et al. Clinical course, microbiologic profile, and diagnosis of periannular complications in prosthetic
valve endocarditis. Am J Cardiol. 1999;83:1075-1079.
12. Graupner C, Vilacosta I, San Román JA, et al. Periannular extension
of infective endocarditis. J Am Coll Cardiol. 2002;39:1204-1211.
13. San Román JA, Vilacosta I, Zamorano JL, et al. Transesophageal
echocardiography in right-sided endocarditis. J Am Coll Cardiol.
1993;21:1226-1230.
14. Cleland JF, Erdmann E, Ferrari OM, et al. The Task Force on Heart
Failure of the European Society of Cardiology. Guidelines for the
diagnosis of heart failure. Eur Heart J. 1995;16:741-751.
15. DiSesa VJ. Art and science in the management of endocarditis. Ann
Thorac Surg. 1991;51:6-7.
16. Bishara J, Leibovici L, Gartman-Israel D, et al. Long-term outcome of
infective endocarditis: the impact from early surgical intervention.
Clin Infect Dis. 2001;33:1636-1643.
17. Netzer R, Zollinger E, Seiler C, Cerny A. Infective endocarditis:
clinical spectrum, presentation and outcome. An analysis of 212 cases
1980-1995. Heart. 2000;84:25-30.
18. Netzer R, Altwwegg S, Zollinger E, et al. Infective endocarditis:
determinants of long term outcome. Heart. 2002;88:61-66.
19. Granowitz E, Longworth D. Risk stratification and bedside prognostication in infective endocarditis. JAMA. 2003;289:1991-1993.
20. Mansur AJ, Grinberg M, Cardoso HA, et al. Determinants of prognosis
in 300 episodes of infective endocarditis. Thorac Cardiol Surg.1996;
44:2-10.
21. Delahaye F, Ecochard R, de Gevigney G, et al. The long term prognosis of infective endocarditis. Eur Heart J.1995;16:48-53.
369.e7
22. Schulz R, Werner J.B, Andreas S, et al. Clinical outcome and echocardiographic findings of native and prosthetic valve endocarditis in
the 1990s. Eur Heart J. 1996;17:281-288.
23. Tornos MP, Permanyer-Miralda G, Olona M, et al. Long term complications of native valve infective endocarditis in non-addicts. A
15-year follow-up study. Ann Intern Med. 1992;117:567-572.
24. Mansur AJ, Dal Bo C, Fukushima JT, et al. Relapses, recurrences,
valve replacements and mortality during the long term follow-up after
infective endocarditis. Am Heart J. 2001;141:78-86.
25. Castillo JC, Anguita MP, Ramírez A, et al. Clinical characteristics and
evolution of recurrent infectious endocarditis in non drug addicts. Rev
Esp Cardiol. 2000;53:344-352.
26. Cartón JA, Asensi V, Maradona JA, et al. Endocarditis infecciosa
sobre válvula natural: perfil epidemiológico y análisis de la mortalidad
entre los años 1984 y 1993. Med Clin (Barc).1995;104:493-499.
27. Chan KL. Early clinical course and long-term outcome of patients with
infective endocarditis complicated by perivalvular abscess. CMAJ.
2002;16:19-24.
28. Hoen B, Alla F, Selton-Suty C, et al. Changing profile of infective
endocarditis. Results of a 1-year survey in France. JAMA. 2002;288:
75-81.
29. Hecht S, Berger M. Right-sided endocarditis in intravenous drug users.
Ann Intern Med. 1992;117:560-566.
30. Sandre RM, Shafran SD. Infective endocarditis: review of 135 cases
over 9 years. Clin Infect Dis. 1996;22:276-286.
31. Chu VH, Cabel CH, Benjamin DK, et al. Early predictors of inhospital death in infective endocarditis. Circulation. 2004;109:17451749.
32. Goldman ME, Fisher EA, Winters S, et al. Early identification of
patients with native infectious endocarditis at risk for mayor complications by initial clinical presentation and baseline echocardiography.
Int J Cardiol.1995;52:257-264.
33. Karth G, Koreny M, Binder T, et al. Complicated infective endocarditis necessitating ICU admission: clinical course and prognosis. Crit
Care. 2002;6:149-154.
34. Holenarasipur R, Buenconsejo J, Hasbun R, Quagliarello V. Impact of
valve surgery on 6-month mortality in adults with complicated, leftsided native valve endocarditis. JAMA. 2003;290:3207-3214.
35. Baddour LM. Twelve-year review of recurrent native-valve infective
endocarditis: a disease of the modern antibiotic era. Rev Infect
Dis.1988;10:1163-1170.
36. Tornos P, Almirante B, Olona M, et al. Clinical outcome and longterm prognosis of late prosthetic valve endocarditis: a 20-year experience. Clin Infect Dis.1997;24:381-386.
37. Millaire A, Leroy O, Gaday V, et al. Incidence and prognosis of
embolic events and metastatic infections in infective endocarditis. Eur
Heart J. 1997;18:677-684.
38. Mügge A, Daniel W. Echocardiographic assessment of vegetations in
patients with infective endocarditis. Echocardiography. 1995;12:651661.
39. Aranki S, Adams D, Rizzo R, et al. Determinants of early mortality
and late survival in mitral valve endocarditis. Circulation.1994;92:
143-149.
40. Fernández-Guerrero ML, Verdejo C, Azofra J, de Gárgolas M. Hospital-acquired infectious endocarditis not associated with cardiac surgery: an emerging problem. Clin Infect Dis. 1995;20:16-23.
41. Zamorano J, Sanz J, Moreno R, et al. Comparison of outcome in
patients with culture negative versus culture positive active infective
endocarditis. Am J Cardiol. 2001;87:1423-1425.
42. Casabé JH, Deschle H, Cortés C, et al. Predictors of hospital mortality
in 186 cases of active infective endocarditis treated in a tertiary
medical center (1992-2001). Rev Esp Cardiol. 2003;56:578-585.
The American Journal of Medicine (2007) 120, 369.e9-369.e14
CLINICAL RESEARCH STUDY
Does Statin Therapy Decrease the Risk for Bleeding in
Patients Who Are Receiving Warfarin?
James D. Douketis, MD, FRCP(C),a Magda Melo, BSPhrm, MSc,b Chaim M. Bell, MD, PhD, FRCP(C),b,c,d,f
Muhammad M. Mamdani, PharmD, MA, MPHb,c,e
a
Department of Medicine, McMaster University and St. Joseph’s Healthcare, Hamilton, Ontario, Canada; bInstitute for Clinical
Evaluative Sciences, Toronto, Ontario, Canada; cDepartment of Medicine, University of Toronto, Ontario, Canada; dDivision of
General Internal Medicine, St. Michael’s Hospital, Toronto, Ontario, Canada; eDepartment of Pharmacy, University of Toronto,
Ontario, Canada; fDepartment of Health Policy, Management and Evaluation, University of Toronto, Ontario, Canada.
ABSTRACT
PURPOSE: Recent observations in patients with atrial fibrillation who are receiving warfarin suggest that
concomitant treatment with a 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor (statin) decreases the risk for bleeding.
METHODS: We conducted a population-based, nested case-control study using the linked administrative
databases of Ontario, Canada, to assess whether statin use decreases the risk of bleeding in warfarin users.
Eligible patients were Ontario residents, age 66 years or more, with atrial fibrillation who were prescribed
warfarin between April 1, 1994, and December 31, 2001. Patients were followed until hospitalization for
upper gastrointestinal or intracranial bleeding, study end (March 31, 2002), discontinuation of warfarin, or
death. Cases were matched to controls by age and sex. Logistic regression analysis was used to estimate
odds ratios (ORs) and 95% confidence intervals (CIs) for the association between bleeding and statin use.
RESULTS: We identified 79,207 warfarin users with atrial fibrillation. There were 1518 cases with an
upper gastrointestinal or intracranial bleed and 15,100 matched controls without bleeding. Long-term (ⱖ1
year) statin use was associated with a lower risk for any bleeding (OR ⫽ 0.80; 95% CI, 0.66-0.97).
However, there was no association between bleeding and recent (⬍6 months) statin use (OR ⫽ 1.04; 95%
CI, 0.74-1.48) or statin use of any duration (OR: 0.91; 95% CI, 0.77-1.07), suggesting potential confounding of the association between statin use and bleeding by a health-user effect.
CONCLUSION: Long-term statin use may be associated with a decreased risk for bleeding in warfarin
users with atrial fibrillation. Additional research is needed to further explore this putative association.
© 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Oral anticoagulants; Statins; Bleeding; Case-control study
3-Hydroxy-3-methylglutaryl coenzyme A reductase inhibitor drugs (statins) are widely used for the primary and
Dr. Bell holds a Canadian Institutes of Health Research Phase 2 ClinicianScientist Award. Dr. Mamdani is currently an employee of Pfizer Inc, the
manufacturer of one statin. The involvement of Dr. Mamdani in the development of this study and article was preceded by (⬃6 months) and was not
influenced by his subsequent decision to work for Pfizer. Dr. Douketis, Dr.
Bell, and Ms. Melo have no conflict of interest or financial disclosures.
Requests for reprints should be addressed to James D. Douketis, MD,
FRCP(C), St. Joseph’s Hospital, 50 Charlton Ave, Hamilton, ON, Canada,
L8N 4A6.
E-mail address: [email protected].
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.06.008
secondary prevention of acute coronary syndromes, stroke,
and other cardiovascular disease,1,2 and in recent years
statin use has increased sharply.3 In addition to their lipidlowering effects, statins are considered to have several
pleiotropic effects unrelated to lipid lowering, which include plaque stabilization in patients with an acute coronary
syndrome.2
Recent observations have suggested that statin use may
modify the risk for bleeding in patients who are receiving
anticoagulants and in patients with acute intracerebral bleeding. An analysis of the SPORTIF III and V Study databases
assessed risk factors for bleeding in 7329 patients with chronic
369.e10
The American Journal of Medicine, Vol 120, No 4, April 2007
atrial fibrillation who received warfarin, administered to
of study period (March 31, 2002), patient death, or disachieve a target international normalized ratio (INR) of 2.0
continuation of warfarin, whichever came first. Patients
to 3.0, or ximelagatran, 36 mg twice daily.4 Multivariate
were not studied during their first year of eligibility for
logistic regression analysis found, unexpectedly, that statin
prescription drugs, corresponding to their 66th year, to
use at the time of study enrollment conferred a 40% risk
avoid incomplete medication records. Discontinuation of
reduction for major bleeding in
warfarin was deemed to have ocboth warfarin-treated patients (odds
curred if there was more than
ratio [OR] ⫽ 0.60; 95% confi180 days from the previous preCLINICAL SIGNIFICANCE
dence interval [CI], 0.41-0.87)
scription. In these cases of warand ximelagatran-treated patients
farin discontinuation, the period
● Recent observations suggest that statin
(OR ⫽ 0.61; 95% CI, 0.39-0.96).
of observation was counted until
use may be associated with a decreased
The reason for this observation is
90 days after the last warfarin
risk for bleeding in patients who are renot clear and is counterintuitive
prescription date. A priori, we
ceiving oral anticoagulant therapy.
when one considers that statins
separated warfarin users into 3
have antiplatelet and other antigroups based on their duration of
● By using a linked administrative datathrombotic properties that would
warfarin use before the outcome:
base, we found that long-term (ⱖ1 year)
tend to promote bleeding.5,6
any warfarin user; long-term
statin use was associated with a lower
To test whether statin use has a
warfarin users, based on at least
risk for any bleeding (OR ⫽ 0.80; 95% CI,
protective effect on clinically sig6 months of continuous warfarin
0.66-0.97), although a healthy user efnificant bleeding in anticoagulated
use; and recent warfarin users,
fect may also explain these findings.
patients, we performed a populabased on less than 6 months of
tion-based, nested case-control
warfarin use.
study involving patients with
chronic atrial fibrillation who were receiving warfarin
Definition of Case Patients
therapy.
Within the cohort of warfarin users with atrial fibrillation,
case patients were defined as those registered in the CIHI
METHODS
DAD as having been admitted to hospital with a diagnosis
of upper gastrointestinal or intracranial hemorrhage beStudy Design and Data Sources
tween April 1, 1994, and March 31, 2002 (Appendix 1).
We used a large linked population-based administrative database in Ontario, Canada, to perform a nested case-control study
from April 1, 1994, to March 31, 2002. Prescriptions were
Definition of Control Patients
identified from the Ontario Drug Benefit Program, which
From the cohort of continuous warfarin users who were not
records prescription medications dispensed to residents of
admitted to a hospital between April 1, 1994, and March 31,
Ontario who are age 65 years and older. Hospital admissions
2002, we randomly selected 10 controls for each case,
for upper gastrointestinal and intracranial bleeding were idenmatched for age (⫾1 year) and sex. If fewer than 10 controls
tified from the Canadian Institute for Health Information Disexisted, we analyzed only those controls and did not alter
charge Abstract Database (CIHI DAD), which provides dethe matching process. A case patient could serve as a control
tailed documentation of all hospital admissions in Ontario.
for a different case patient before his or her date of admisBaseline patient demographic and other clinical characteristics
sion, and each control could be used for multiple case
were obtained from the Ontario Registered Persons Database.
patients.10,11
These databases are linked through encrypted patient identification numbers.
Ethics approval was obtained from the Sunnybrook and
Definition of Exposure
Women’s College Health Science Centre and used the proThe exposure of interest was use of a statin drug before the
tocols of the Institute for Clinical Evaluative Sciences to
index bleeding event. The statin drugs included in the Onsafeguard patient confidentiality.
tario formulary at that time were atorvastatin, cerivastatin,
fluvastatin, lovastatin, pravastatin, and simvastatin. A priori,
Cohort Definition
a statin user was classified as any statin user, in whom there
We studied patients with a history of atrial fibrillation
was at least 1 prescription claim during the 2 years before
based on hospitalization records for the 3 years before
the index bleed; long-term statin user, in whom there was at
cohort entry. Patient prescription records enabled us to
least 1 statin claim in each of the 6-month periods of the
define a period of warfarin use based on previously pubprevious year; recent statin user, in whom there was at least
lished methods.7-9 The period of observation for the pa1 prescription claim during the 180 days before the index
tient cohort began with any prescription of warfarin after
bleed but no claim in the remaining 6 months of the year
the age of 66 years and ended with a hospital admission
for upper gastrointestinal or intracranial bleeding, the end
before the event.
Douketis et al
Table 1
Statin Therapy and Bleeding Risk
369.e11
Characteristics of Study Subjects
Subjects, n
Age, mean (SD)
Male
Mean days of follow-up
Prior bleeding event
Products used in the past
year, mean (SD)
LTC facility
Prescriptions in the
previous 120 d
ACE inhibitors
Angiotensin receptor
antagonists
Aspirin
Antibiotics (systemic)
␤-blockers
Calcium channel blockers
Clopidogrel
Digoxin
Diuretics
Fibrates
Gastroprotectors
CYP 2C9 inducers
CYP 2C9 inhibitors
NSAIDs
Steroids
Ticlopidine
Any Warfarin Users
Long-term Warfarin Users
Recent Warfarin Users
Cases
Controls
Cases
Controls
Cases
Controls
1518
77.2 (7.2)
45.7
612.9
3.7
16.1 (8.0)
15,100
76.9 (7.0)
45.7
737.6
1.5
13.6 (6.9)
1003
76.9 (6.9)
45.7
895.2
3.3
15.9 (8.0)
9993
76.9 (6.7)
45.7
924.7
1.6
13.3 (6.7)
513
77.7 (7.7)
46.0
63.2
4.3
16.6 (7.9)
4921
77.7 (7.2)
46.0
82.8
1.7
14.8 (7.2)
11.7
7.3
11.7
8.2
11.5
8.5
52.4
3.3
46.9
3.4
52.0
4.4
46.5
3.9
53.4
1.2
47.8
2.9
11.4
39.5
30.8
33.2
0.3
63.3
64.3
2.2
32.7
1.4
9.0
16.7
30.4
1.0
5.4
32.6
30.7
32.8
0.1
64.5
59.2
1.3
23.8
2.0
9.1
9.0
27.8
0.3
6.2
37.9
28.2
32.2
0.1
63.1
64.4
1.9
30.0
1.6
7.5
16.4
31.5
0.3
3.3
31.6
29.4
32.0
0.1
65.3
58.5
1.4
23.2
2.2
8.2
8.7
27.7
0.1
21.6
42.3
36.1
35.3
0.6
63.6
64.1
2.9
38.0
1.0
11.9
17.4
28.5
2.3
15.4
36.9
37.7
32.7
0.4
61.4
58.1
1.5
26.8
1.8
13.1
11.2
28.0
1.0
SD ⫽ standard deviation; LTC ⫽ long-term care; ACE ⫽ angiotensin-converting enzyme; NSAID ⫽ nonsteroidal anti-inflammatory drug;
CYP ⫽ cytochrome P450. Values are percentages, unless stated otherwise. CYP 2C9 inducers: rifampin, phenytoin. CYP 2C9 inhibitors: amiodarone,
losartan, tolbutamide, cimetidine, fluvastatin, fluconazole.
Exposure to Interacting Medications and
Covariates
To adjust for concomitant drug use that might be associated
with an increased risk for bleeding or that might interfere
with warfarin metabolism, drug use during the 120 days
before the index bleed event was included in the regression
analysis.12 To adjust for potential patient-related confounders that might impact on bleeding, several comorbid conditions, as outlined in Appendix 2, were included in the
regression analyses. Residence in a long-term care facility
was also identified as a covariate, because this might be an
indicator of comorbidity.13 Adjustment also was made for
the number of different drugs prescribed in the year before
the index bleed, which is an additional measure of
comorbidity.14
Statistical Analysis
Conditional logistic regression analysis was performed to
estimate the ORs and corresponding 95% CIs for the association between admission to a hospital for upper gastrointestinal or intracranial bleeding and use of a statin drug. We
also performed a subgroup analyses on patients with upper
gastrointestinal bleeding, because this condition has been
extensively studied. In the multivariate analysis, adjustment
was made for the covariates outlined in Appendix 2. To
avoid overfitting the model, drugs that could have confounded the model were grouped by mechanism rather than
inserted individually. All analyses were performed using
SAS version 8.2 (SAS Institute, Cary, NC).
RESULTS
Study Subject Characteristics
We identified 79,207 warfarin users with a history of atrial
fibrillation, of whom 52% were women. The mean age at
cohort entry was 79 years.
Statin Use and Risk for Any Bleeding (Upper
Gastrointestinal or Intracranial)
During the period of observation, from April 1, 1994, to
March 31, 2002, 1518 individuals from the warfarin cohort
were admitted to a hospital with an upper gastrointestinal or
intracranial bleed (cases) and were matched to 15,100 controls without a bleeding event (Table 1). Hospital admissions for upper gastrointestinal bleeding (n ⫽ 1201) were
more common than for intracranial bleeding (n ⫽ 319) in
369.e12
Table 2
The American Journal of Medicine, Vol 120, No 4, April 2007
Association Between Statin Use and Bleeding (Upper Gastrointestinal Bleed or Intracranial Bleed) Among Warfarin Users
Warfarin Users
Statin Use
Crude OR (95% CI)
Adjusted OR (95% CI)§
Any use
Any use*
Long-term use†
Recent use‡
Any use*
Long-term use†
Recent use‡
Any use*
Long-term use†
Recent use‡
1.03 (0.88-1.20)
0.87 (0.72-1.05)
1.22 (0.87-1.72)
0.92 (0.76-1.13)
0.85 (0.68-1.07)
0.89 (0.54-1.47)
1.07 (0.82-1.39)
0.88 (0.63-1.22)
1.05 (0.67-1.65)
0.91 (0.77-1.07)
0.80 (0.66-0.97)
1.04 (0.74, 1.48)
0.82 (0.67-1.00)
0.79 (0.63, 0.99)
0.77 (0.46, 1.30)
1.02 (0.78-1.34)
0.82 (0.58-1.15)
1.06 (0.66-1.68)
Long-term
Recent use
OR ⫽ odds ratio; CI ⫽ confidence interval. Any warfarin use: subject who had at least a claim for warfarin during the study period before the bleeding
event (includes long-term users and recent users). Long-term warfarin user: subject who used warfarin continuously for at least 6 months. Recent warfarin
user: subject who used warfarin for less than 6 months.
*Any use of statins: any use in the previous 2 years of the index bleeding.
†Long-term use of statins: at least one statin claim in each of the 6-month periods of the previous year.
‡Recent use of statin: at least one statin claim in the 6 months before index, but none in the 6 months before that.
§Adjusted for number of drugs used in the past year, history of upper gastrointestinal bleeding or intracranial bleeding, residency in a long-term care
facility at cohort entry, and use of ACE inhibitors, angiotensin receptor blockers, aspirin, NSAIDs, ticlopidine, steroids, diuretics, fibrates, gastroprotectors, digoxin, CYP 2C9 inducers, CYP 2C9 inhibitors, systemic antibiotics, beta-blockers, calcium channel blockers, and clopidogrel in the 120 days
before index.
our patient population. The analyses were run separately for
upper gastrointestinal bleeding and intracranial bleeding.
Two patients had both events but contributed only once to
the “any bleed” category.
For our primary analysis examining the association of
any statin use in the previous 2 years in those also prescribed warfarin for upper gastrointestinal or intracranial
bleeding, the multivariate analysis demonstrated no statistically significant effect (OR ⫽ 0.91, 95% CI, 0.77-1.07)
(Table 2). However, those prescribed warfarin and on longterm statins had a statistically significant reduced risk for
upper gastrointestinal or intracranial bleeding (OR ⫽ 0.80,
95% CI, 0.66-0.97). The statistically significant reduced risk
was also found in long-term warfarin and statin users
(OR ⫽ 0.79, 95% CI, 0.63-0.99). All other combinations of
warfarin and statin use demonstrated no statistically significant difference (Table 2). Furthermore, when we restricted
our analyses to upper gastrointestinal bleeding events, longterm statin use was also associated with a reduced risk of
bleeding among warfarin users (OR ⫽ 0.78, 95% CI, 0.630.97), but not recent statin use (OR ⫽ 1.10, 95% CI, 0.741.66) or any statin use (OR ⫽ 0.92, 95% CI, 0.77-1.11).
DISCUSSION
The principal finding of this study is that in warfarin users,
long-term statin use seems to be associated with a decreased
risk for bleeding complications. However, because new use
of a statin does not seem to confer a protective effect against
warfarin-associated bleeding, the possibility exists that a
protective effect of long-term statin use against bleeding
may be related to a “healthy user effect.” Thus, long-term
statin users or their physicians may be more attentive to
personal health risks and more likely to have other unmeasured determinants that decrease bleeding risk (eg, better
warfarin anticoagulation control) than statin nonusers. Fur-
thermore, the lack of a biologically plausible mechanism
further suggests that an association between statin use and
bleeding should be interpreted with caution and may be
spurious.
Previous randomized, controlled trials of statin therapy
did not assess bleeding as an outcome and, therefore, cannot
be used to examine the effects of statin use on bleeding.2 In
a small retrospective cohort study involving 54 patients with
type 2 diabetes, there was a trend toward delayed vitreous
hemorrhage in statin users compared with nonusers (21.4
years vs 16.2 years, P ⫽ .09).15 In studies that have assessed
the effects of statin therapy on vascular reactivity in patients
with subarachnoid hemorrhage, there is preliminary evidence that statins may reduce bleeding, possibly through an
ameliorating effect on vasospasm.16-18 Taken together,
there is no compelling evidence that statins have an effect
on modifying bleeding risk in non-anticoagulated patients.
This study has potential limitations. First, we acknowledge the potential for inaccurate ascertainment of cases with
bleeding, which is germane to administrative database studies. The accuracy of the diagnostic codes we used to capture
upper gastrointestinal and intracranial bleeds has not been
validated in the Ontario hospital discharge database and was
not feasible in our study. However, validation studies performed elsewhere, including in another Canadian jurisdiction, have shown that the positive predictive value for capturing intracranial bleeds with an International
Classification of Diseases, Ninth Revision (ICD-9) code is
77% to 98% when the ICD-9 code was the most responsible
diagnosis for the hospitalization.19-22 Similarly, the accuracy of ICD-9 codes for capturing upper gastrointestinal
bleeds was 59% to 100% in other studies and greater than
81% when ICD-9 codes specific for bleeding sites were
used.23,24 Furthermore, there is excellent agreement between administrative hospitalization data and chart audits in
Douketis et al
Statin Therapy and Bleeding Risk
other clinical domains with the database used in this
study.25 Second, our analyses assumed a statin drug class
effect because of the exploratory nature of our study and
because we considered our study unlikely to have sufficient
power to assess bleeding risk with individual statins. Preparation-specific protective or adverse effects may occur because some statins (fluvastatin, lovastatin, simvastatin) have
been found to potentiate the anticoagulant effect of warfarin
and possibly increase the risk for bleeding.26 We cannot
exclude the possibility that the effects we observed may
differ according to individual statin preparations. Third, we
excluded patients with atrial fibrillation who were aged less
than 66 years and, therefore, cannot apply our findings to
such patients. Fourth, we did not include anticoagulation
control, as measured by the INR, as a variable in our
regression analysis because this was not feasible. However,
correlating INR level and bleeding risk is problematic even
in randomized trials, unless regular INR testing is mandated.4 Furthermore, bleeding can occur irrespective of the
INR level, and the INR at the time of clinical presentation
may not reflect the INR at the time of a bleed.27,28 Finally,
we could not assess the effect of using nonprescription
(“over-the-counter”) drugs that affect hemostasis, such as
aspirin or over-the-counter nonsteroidal anti-inflammatory
drugs, on the association between statin use and bleeding
risk.
To summarize, long-term statin use seems to be associated with a decreased risk for bleeding in warfarin users
with atrial fibrillation. However, potential confounders and
a lack of biologic plausibility warrants additional studies to
further investigate the putative association between statin
use and bleeding risk in warfarin users.
APPENDIX 1
International Classification of Diseases, Ninth Revision,
Clinical Modification Diagnostic Codes Used to Identify
Cases with Upper Gastrointestinal or Intracranial Bleeding
ICD-9 Codes
431: Intracerebral hemorrhage
531.0: Gastric ulcer—acute with hemorrhage
531.2: Gastric ulcer—acute with hemorrhage and perforation
531.4: Gastric ulcer—chronic or unspecified with hemorrhage
531.6: Gastric ulcer—chronic or unspecified with hemorrhage
and perforation
532.0: Duodenal ulcer—acute with hemorrhage
532.2: Duodenal ulcer—acute with hemorrhage and
perforation
532.4: Duodenal ulcer—chronic or unspecified with
hemorrhage
532.6: Duodenal ulcer—chronic or unspecified with
hemorrhage and perforation
533.0: Peptic ulcer—acute with hemorrhage
533.2: Peptic ulcer—acute with hemorrhage and perforation
533.4: Peptic ulcer—chronic or unspecified with hemorrhage
533.6: Peptic ulcer—chronic or unspecified with hemorrhage
and perforation
534.0: Gastrojejunal ulcer—acute with hemorrhage
369.e13
ICD-9 Codes
534.2: Gastrojejunal ulcer—acute with hemorrhage and
perforation
534.4: Gastrojejunal ulcer—chronic or unspecified with
hemorrhage
534.6: Gastrojejunal ulcer—chronic or unspecified with
hemorrhage and perforation
578.0: Hematemesis
578.1: Blood in stool
578.9: Hemorrhage of gastrointestinal tract, unspecified
ICD-9 ⫽ International Classification of Diseases, Ninth Revision.
APPENDIX 2
Potential Confounders for Association Between Statin Use
and Bleeding in Patients with Atrial Fibrillation Who Are
Receiving Warfarin
Potential Confounders
Prior event in the previous 3 years of cohort entry
- prior gastrointestinal bleeding
- prior intracerebral bleeding
Number of prescription drugs dispensed in the year before
cohort entry
Drug use in the 120 days preceding the event
- Aspirin
- Clopidogrel
- Ticlopidine
- Nonsteroidal anti-inflammatory drugs
- COX-2 inhibitors
- Cytochrome P450 2C9 inhibitors (amiodarone, losartan,
tolbutamide, cimetidine, fluvastatin, fluconazole)
- Cytochrome P450 2C9 inducers (rifampin, phenytoin)
- Fibrates
- Digoxin
- Diuretics
- ␤-blockers
- Calcium channel blockers
- Angiotensin-converting enzyme inhibitors
- Angiotensin receptor blockers
- Systemic antibiotics
- Corticosteroids
- Gastroprotective drugs
COX-2 ⫽ cyclooxygenase-2.
References
1. Briel M, Studer M, Glass TR, Bucher HC. Effects of statins on stroke
prevention in patients with and without coronary heart disease: a
meta-analysis of randomized controlled trials. Am J Med. 2004;117:
596-606.
2. Cheung BMY, Lauder IJ, Lau C-P, Kumana CR. Meta-analysis of
large randomized controlled trials to evaluate the impact of statins on
cardiovascular outcomes. Br J Clin Pharmacol. 2003;57:640-651.
3. Tu JV, Gong Y. Trends in treatment and outcomes for acute stroke
patients in Ontario, 1992-1998. Arch Intern Med. 2003;163;293-297.
4. Douketis JD, Arneklev K, Goldhaber SZ, et al. Ximelagatran versus
warfarin: a comparison of bleeding outcomes in patients with nonvalvular atrial fibrillation. Arch Intern Med. 2006;166:853-859.
5. Szapary L, Horvath B, Marton Z, et al. Short-term effect of low-dose
atorvastatin on haemorrheological parameters, platelet aggregation and
endothelial function in patients with cerebrovascular disease and hyperlipidemia. CNS Drugs. 2004;18:165-172.
6. Undas A, Brummel KE, Musial J, Mann KG, Szceklik A. Simvastatin
depresses blood clotting by inhibiting activation of prothrombin, Fac-
369.e14
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
tor V, and Factor XIII and by enhancing Factor Va inactivation.
Circulation. 2001;103:2248-2253.
Juurlink DN, Mamdani M, Kopp A, Laupacis A, Redelmeier DA.
Drug-drug interactions among elderly patients hospitalized for drug
toxicity. JAMA. 2003;289:1652-1658.
Ray JG, Mamdani MM. Statins and the risk of idiopathic venous
thromboembolism. Br J Clin Pharmacol. 2003;50:460.
Mamdani MM, van Walraven C, Bica A, Williams JI, Naylor CD. Is
there an association between lipid-lowering drugs and cholecystectomy? Am J Med. 2000;108:418-421.
Lubin JH, Gail MH. Biased selection of controls for case-control
analyses of cohort studies. Biometrics. 1984;40:63-75.
Langholz B. Case-control study, nested. In: Armitage P, Colton C, eds.
Encyclopedia of Biostatistics. Chichester, England: John Wiley &
Sons Ltd; 1998;514-519.
Michalets E. Update: clinically significant cytochrome P-450 drug
interactions. Pharmacotherapy. 1998;18:84-112.
Bravo G, Dubois MF, Hebet R, De Wals P, Messier L. A prospective
evaluation of the Charlson Comorbidity Index for use in long-term
care patients. J Am Geriatr Soc. 2002;50:740-745.
Schneeweiss S, Seeger JD, Maclure M, Wang PS, Avorn J, Glynn RJ.
Performance of comorbidity scores to control for confounding in epidemiologic studies using claims data. Am J Epidemiol. 2001;154:854-864.
Banerjee S, Denniston AK, Gibson JM, Dodson PM. Does cardiovascular therapy affect the onset and recurrence of preretinal and vitreous
haemorrhage in diabetic eye disease? Eye. 2004;18:821-825.
Tseng M-Y, Czosnyka M, Richards H, Pickard JD, Kirkpatrick PJ.
Effects of acute treatment with pravastatin on cerebral vasospasm,
autoregulation, and delayed ischemic deficits after aneurysmal subarachnoid hemorrhage. Stroke. 2005;36:1627-1632.
Parra A, Kreiter KT, Williams S, et al. Effect of prior statin use on
functional outcome and delayed vasospasm after acute aneurysmal
subarachnoid hemorrhage: a matched controlled cohort study. Neurosurgery. 2005;56:476-484.
The American Journal of Medicine, Vol 120, No 4, April 2007
18. Singhal AB, Topcuoglu MA, Dorer DJ, Ogilvy CS, Carter BS, Koroshetz WJ. SSRI and statin use increases the risk for vasospasm after
subarachnoid hemorrhage. Neurology. 2005;64:1008-1013.
19. Tirschwell DL, Longstreth WT Jr. Validating administrative data in
stroke research. Stroke. 2002;33:2465-2470.
20. Rosamond WD, Folsom AR, Chambless LE, et al. Stroke incidence
and survival among middle-aged adults: 9-year follow-up of the Atherosclerosis Risk in Communities (ARIC) cohort. Stroke. 1999;30:
736-743.
21. Leone MA, Capponi A, Varrasi C, Tarletti R, Monaco F. Accuracy of
the ICD-9 codes for identifying TIA and stroke in an Italian automated
database. Neurol Sci. 2004;25:281-288.
22. Broderick J, Brott T, Kothari R, et al. The Greater Cincinnati/Northern
Kentucky Stroke Study: preliminary first-ever and total incidence rates
of stroke among blacks. Stroke. 1998;29:415-421.
23. Cattaruzzi C, Troncon MG, Agostinis L, Garcia Rodriguez LA. Positive predictive value of ICD-9th codes for upper gastrointestinal
bleeding and perforation in the Sistema Informativo Sanitario Regionale database. J Clin Epidemiol. 1999;52:499-502.
24. Raiford DS, Perez GS, Garcia Rodriguez LA. Positive predictive value
of ICD-9 codes in the identification of cases of complicated peptic
ulcer disease in the Saskatchewan hospital automated database. Epidemiology. 1996;7:101-104.
25. Cardiovascular Health and Services in Ontario: An ICES Atlas. Toronto, Canada: Institute for Clinical Evaluative Sciences; 1999.
26. Andrus MR. Oral anticoagulant drug interactions with statins: case
report of fluvastatin and review of the literature. Pharmacotherapy.
2004;24:285-290.
27. Kucher N, Connolly S, Beckman JA, et al. International normalized
ratio increase before warfarin-associated hemorrhage: brief and subtle.
Arch Intern Med. 2004;164:2176-2179.
28. Fanikos J, Grasso-Correnti N, Shah R, Kucher N, Goldhaber SZ.
Major bleeding complications in a specialized anticoagulation service.
Am J Cardiol. 2005;96:595-598.
The American Journal of Medicine (2007) 120, e1-e2
CLINICAL COMMUNICATION TO THE EDITOR
An Unusual Case of Dysphagia and Chest Pain
in a Non–HIV Patient: Esophageal
Tuberculosis
CASE REPORT
A 25-year-old woman presented with substernal chest pain
and dysphagia lasting 4 days. The patient described the pain
as a sharp substernal pain with no radiation. Pain intensity
only increased by swallowing food. She also reported a
2.5-kg weight loss in the past 3 months. The patient immigrated to the United States from India approximately 10
years previously. Physical examination revealed a temperature of 97.7°F, blood pressure of 138/86 mm Hg, heart rate
of 94 beats/min, and respiratory rate of 18 breaths/min.
Chest, heart, and abdominal examination results were unremarkable. Laboratory investigation revealed a white
blood cell count of 6.7 103/mL, hemoglobin level of 13
Requests for reprints should be addressed to Bobbak Vahid, MD, 834
Walnut Street, Suite 650, Philadelphia, PA 19107.
E-mail address: [email protected].
Figure 1
g/dL, and normal blood chemistry. Human immunodeficiency virus screening antibody was negative. A computed
tomography scan of the chest was performed showing a
subcarinal mass with a hypodense center that was inseparable from the esophagus (Figure 1). Flexible esophagoscopy showed a single deep linear ulcer in the middle third of
the esophagus. A biopsy of the subcarinal mass showed a
necrotizing granulomatous inflammation. Tissue cultures
grew Mycobacterium tuberculosis complex. The diagnosis
of esophageal tuberculosis was made. A gastrostomy tube
was placed, and treatment with isoniazid, rifampin, ethambutol, and pyrazinamide was started.
DISCUSSION
Esophageal tuberculosis is rare. In autopsy studies, esophageal
involvement was found in only 0.15% of patient who died of
tuberculosis.1 The primary esophageal tuberculosis is defined
as esophageal involvement with no other systemic manifestation of tuberculosis. Primary esophageal tuberculosis is rare,
because squamous epithelium and peristalsis of esophagus
have a strong protective effect against tuberculosis.2 Secondary
esophageal tuberculosis is more common. Secondary esopha-
CT scan of chest showing subcarinal mass with hypodense center involving the esophagus (arrow).
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
e2
geal tuberculosis can occur because of direct extension of
mediastinal, pulmonary, and spinal tuberculosis.3
The symptoms of esophageal tuberculosis include dysphagia, odynophagia, chest pain, low-grade fever, and
weight loss.2,4 Infection with human immunodeficiency virus, immunosuppressive therapy, and immigration from tuberculosis-endemic areas are risk factors for esophageal
tuberculosis.5 The reported complications of esophageal tuberculosis are aspiration pneumonia,6 fatal hematemesis,6
esophagotracheal fistula,5 esophagomediastinal fistula,5
traction diverticula,3 and esophageal strictures.3
Approximately 65% of patients with esophageal tuberculosis have nonspecific findings on chest radiograph suggestive of tuberculosis. Enlarged lymph nodes with a hypodense center on computed tomography scan of the chest that
are characteristic of tuberculous lymphadenitis were observed in the majority of these patients.3
Diagnosis of esophageal tuberculosis can be made by
esophagoscopy or biopsy of associated mediastinal lymphadenopathy.4 A differential diagnosis of esophageal tuberculosis includes esophageal carcinoma, Crohn’s disease,7
moniliasis, actinomycosis, syphilis, and esophageal injury
secondary to ingestion of caustic material.8 Treatment of
esophageal tuberculosis with antituberculous antibiotics and
surgical repair has been reported to be effective.8 Even the
patients with esophageal tuberculosis complicated with
esophagotracheal and esophagomediastinal fistulas were
safely treated with antituberculous antibiotics alone.5
In patients with a risk factor for tuberculosis, esophageal
tuberculosis should be included in the differential diagnosis
of dysphagia and odynophagia even despite its rarity.
The American Journal of Medicine, Vol 120, No 4, April 2007
Bobbak Vahid, MD
Department of Pulmonary and Critical Care Medicine
Thomas Jefferson University Hospital
Philadelphia, Penn
Nazmul Huda, MD
Ali Esmaili, MD
Department of Medicine
Thomas Jefferson University Hospital
Philadelphia, Penn
doi:10.1016/j.amjmed.2005.12.026
References
1. Fujiwara T, Yoshida Y, Yamada S, et al. A case of primary esophageal
tuberculosis diagnosed by identification of Mycobacteria in paraffinembedded esophageal biopsy specimens by polymerase chain reaction.
J Gastroenterol. 2003;38:74-78.
2. Malathi S, So S. Esophageal tuberculosis. Indian J Pediatr. 2004;71:
457-458.
3. Nagi B, Lal A, Kochhar R, et al. Imaging of esophageal tuberculosis: a
review of 23 cases. Acta Radiol. 2003;44:329-333.
4. Mokoena T, Shama DM, Ngakane H, Bryer JV. Oesophageal tuberculosis: a review of eleven cases. Postgrad Med J. 1992;68:110-115.
5. Devarbhavi HC, Alvares JF, Radhikavedi M. Esophageal tuberculosis
associated with esophagotracheal or esophagomediastinal fistula: report
of 10 cases. Gastrointest Endosc. 2003;57:588-592.
6. Abid S, Jafri W, Hamid S, et al. Endoscopic features of esophageal
tuberculosis. Gastrointest Endosc. 2003;57:759-762.
7. Griga T, Duchna HW, Orth M, et al. Tuberculous involvement of the
oesophagus with oesophagobroncheal fistula. Dig Liver Dis. 2002;34:
528-531.
8. Fang H, Lin T, Cheng C, et al. Esophageal tuberculosis: a rare presentation with massive hematemesis. Ann Thorac Surg. 1999;68:23442346.
The American Journal of Medicine (2007) 120, e3-e4
ERRATA
Due to technical difficulties with The American Journal of
Medicine Website, the authors for several articles were
correctly listed in the print journal but were incorrect on the
Journal’s Website.
INCORRECT INTERNET CITATIONS
Wiener CM, Antonarakis ES, Wung PK, Durand DJ,
Leyngold I, Meyerson DA. An Atypical Complication
of Atypical Pneumonia. Am J Med. 2006;119:
824-827.
Wiener CM, Burris G, Fraig M, Judson MA. Trouble Below
the Surface. Am J Med. 2006;119:828-831.
Wiener CM, Huston KK, Nielsen ME, Gelber AC. High
Pressure Surgery. Am J Med. 2006;119:740-742.
Pendleton RC, Rodgers GM, Wiener CM. A Necessary
Detour. Am J Med. 2006;119:651-653.
Scialla JJ, Hemmes AR, Wiener CM. A Pressing Situation.
Am J Med. 2005;118:1347-1349.
Aull FB, Mathiasen H. What can poetry teach physicians?
Am J Med. 2005;118:1310.
Stephens S, Brown CD, Wiener CM. Host to an Intruder.
Am J Med. 2005;118:1102-1104.
Leary P, Dunbar K, Wiener CM. A Very Close Call. Am J
Med. 2005;118:968-971.
Chen TM, Paniker P. A Much-Delayed Complication. Am J
Med. 2005;118:833-835
Faris JE, Brown CD, Wiener CM. Time Tells the Tale. Am J
Med. 2005;118:840-842.
Gianankos D, Mathiasen H. The Sheila Experience. Am J
Med. 2005;118:934.
Antonovich DD, Callen JP, Paniker P. No Walk in the Park.
Am J Med. 2005;118:715-716.
Makaryus AN, Friedman GH, Bettmann MA. Single coronary artery. Am J Med. 2005;118:717.
Helton, TJ, Bavry AA, Indik JH. Acute Myocardial Infarction. Am J Med. 2005;118:718-719.
Scialla TJ, Cavanaugh KL, Wiener, CM. The Rare Needle in a Diagnostic Haystack. Am J Med.
2005;118:720-722.
Roberts K, Mathiasen H. Walt Whitman, Civil War Nurse.
Am J Med. 2005;118:787.
Chen TM, Paniker P. An Unpleasant Momento. Am J Med.
2005;118:604-605.
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
S0002-9343(07)00098-8
Hsiao E, Charles R, Magid C, Nivatpumin P, Wiener CM.
An Unusual Development. Am J Med. 2005;118:
609-611.
Tauscher A, Paniker P. Rapid Resolution of a Common
Problem. Am J Med. 2005;118:475-476.
Ko N, Averett L, Brown C, Wiener CM. An Unexpected
Finding. Am J Med. 2005;118:477-479.
Stephens, RS, Brown C, Wiener CM. Persistent Lower
Abdominal and Groin Pain: What Is the Diagnosis?
Am J Med. 2005;118:364-367.
Scatena L, Paniker P. What Triggered This Sudden Eruption? Am J Med. 2005;118:362-363.
CORRECTED CITATIONS
Antonarakis ES, Wung PK, Durand DJ, Leyngold I, Meyerson DA. An Atypical Complication of Atypical Pneumonia. Am J Med. 2006;119:824-827.
Burris G, Fraig M, Judson MA. Trouble Below the Surface.
Am J Med. 2006;119:828-831.
Huston KK, Nielsen ME, Gelber AC. High Pressure Surgery. Am J Med. 2006;119:740-742.
Pendleton RC, Rodgers GM. A Necessary Detour. Am J
Med. 2006;119:651-653.
Scialla JJ, Hemmes AR. A Pressing Situation. Am J Med.
2005;118:1347-1349.
Aull FB. What can poetry teach physicians? Am J Med.
2005;118:1310.
Stephens S, Brown CD. Host to an Intruder. Am J Med.
2005;118:1102-1104.
Leary P, Dunbar K. A Very Close Call. Am J Med.
2005;118:968-971.
Chen TM. A Much-Delayed Complication. Am J Med.
2005;118:833-835
Faris JE, Brown CD. Time Tells the Tale. Am J Med.
2005;118:840-842.
Gianankos D. The Sheila Experience. Am J Med.
2005;118:934.
Antonovich DD, Callen JP. No Walk in the Park. Am J Med.
2005;118:715-716.
Makaryus AN, Friedman GH. Single coronary artery. Am J
Med. 2005;118:717.
e4
Helton, TJ, Bavry AA. Acute Myocardial Infarction. Am J
Med. 2005;118:718-719.
Scialla TJ, Cavanaugh KL. The Rare Needle in a Diagnostic
Haystack. Am J Med. 2005;118:720-722.
Roberts K. Walt Whitman, Civil War Nurse. Am J Med.
2005;118:787.
Chen TM. An Unpleasant Memento. Am J Med.
2005;118:604-605.
Hsiao E, Charles R, Magid C, Nivatpumin P. An Unusual
Development. Am J Med. 2005;118:609-611.
The American Journal of Medicine, Vol 120, No 4, April 2007
Tauscher A. Rapid Resolution of a Common Problem. Am J
Med. 2005;118:475-476.
Ko N, Averett L, Brown C. An Unexpected Finding. Am J
Med. 2005;118:477-479.
Stephens, RS, Brown C. Persistent Lower Abdominal and
Groin Pain: What Is the Diagnosis? Am J Med.
2005;118:364-367.
Scatena L. What Triggered This Sudden Eruption? Am J
Med. 2005;118:362-363.
The American Journal of Medicine (2007) 120, e5
LETTER
Dosing Frequency of Aspirin and Prevention
of Heart Attacks and Strokes
suggested by others,8,9 are warranted to see whether such an
approach might help prevent additional MI and stroke.
John K. Amory, MD
David W. Amory, MD, PhD
To the Editor:
We enjoyed the review by Dalen1 regarding the use of
aspirin in the prevention of myocardial infarction (MI) and
stroke and would like to raise the issue of dosing frequency
in the discussion of the ideal aspirin regimen. Because
aspirin irreversibly acetylates platelet cyclooxygenase-1, it
prevents platelet activation and aggregation for the lifespan
of the platelets (10 days), and it accomplishes this within 60
minutes of dosing.2 However, because approximately 10%
of the circulating platelet pool is replenished daily and the
serum-half life of aspirin is only a few hours, once-daily
aspirin dosing has the potential to leave a significant fraction of the platelet pool uninhibited, especially 24 hours
after dosing. Moreover, it has been shown that the presence
of as little as 10% to 15% of platelets with normal cyclooxygenase-1 activity may result in normal hemostasis.3-5
Therefore, it seems possible that 81 mg twice-daily dosing
of aspirin may be superior to once-daily dosing, because it
would reduce the percentage of uninhibited platelets in the
circulation. This could be an important issue because there
is an increased incidence of both MI6 and stroke7 in the
early morning hours when platelet aggregability and blood
pressure are highest—and usually close to 24 hours after
most patients have taken their once-daily dose of aspirin.
Future studies of once versus twice-daily aspirin therapy, as
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
University of Washington, Seattle
doi:10.1016/j.amjmed.2006.04.023
References
1. Dalen JE. Aspirin to prevent heart attack and stroke: what’s the right
dose? Am J Med. 2006:119:198-202.
2. Patrono C, Coller B, FitzGerald GA, et al. Platelet-active drugs: the
relationship among dose, effectiveness and side effects. Chest. 2004;
126:234S-264S.
3. DiMinno G, Silver MJ, Murphy S. Monitoring of the entry of new
platelets into the circulation after ingestion of aspirin. Blood. 1983;61:
1081-1085.
4. Patrono C, Ciabattoni G, Patrignani P, et al. Clinical pharmacology of
platelet cyclooxygenase inhibition. Circulation. 1985;72:1177-1184.
5. Catella-Lawson F, Reilly MP, Kapoor SC, et al. Cyclooxygenase inhibitors and the antiplatelet effects of aspirin. N Engl J Med. 2001;345:
1809-1817.
6. Ridker PM, Manson JE, Buring JE, et al. Circadian variation of acute
myocardial infarction and the effect of low-dose aspirin in a randomized
trial of physicians. Circulation. 1990;82:897-902.
7. Elliott WJ. Circadian variation in the timing of stroke onset. A metaanalysis. Stroke. 1998;29:992-996.
8. Krizbacher I, Koppan M, Bodis J. Aspirin for stroke prevention taken in
the evening? Stroke. 2004;35:2760-2761. Letter to the Editor.
9. Krizbacher I, Ajtay Z, Koppan M, Bodis J. Can the time of taking
aspirin influence the frequency of cardiovascular events? Am J Cardiol.
2005;96:608-610.
The American Journal of Medicine (2007) 120, e7
LETTER
The Reply:
Amory and Amory raise some important issues regarding
the dosing of aspirin. I agree that a clinical trial of aspirin
81 mg twice per day versus 162 mg once per day versus
placebo would be of great interest. An appropriate trial
would be for primary prevention of myocardial infarction in
women aged more than 50 years. All the previous trials of
primary prevention in women have used doses of aspirin of
100 mg per day or less, and they have failed to prevent
myocardial infarction.1
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
James E. Dalen, MD, MPH
Associate Editor
American Journal of Medicine
Tucson, Ariz
doi:10.1016/j.amjmed.2006.05.023
Reference
1. Berger JS, Roncaglioni MC, Avanzini F, et al. Aspirin for the primary
prevention of cardiovascular events in women and men. JAMA. 2006;
295:306-313.
The American Journal of Medicine (2007) 120, e9
LETTER
Enteric Coated Aspirin
To the Editor:
The review “Aspirin to Prevent Heart Attack and Stroke:
What’s the Right Dose?” by Dalen1 was welcome. It presented more facts than “fiction,” not always found in the
literature on the use of acetylsalicylic acid (ASA). Missing
was a mention of enteric-coated acetylsalicylic acid
(ECASA). As a long-time student of ASA therapy (⬃70
years) and a still-working, long-time practitioner (60 years)
using ASA in large numbers of patients (first with some
food and in the last 25 years as ECASA), I have yet to see
an acute gastrointestinal bleed that clearly resulted from
ECASA use. Perhaps such events occurred, and the patients
decided to see other doctors. Twenty-five years ago it was
established endoscopically in young volunteers that ECASA
caused far less gastric irritation and ulceration than ASA.
It may be of interest that during a visit to Berlin, Ger-
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
many, in 1995, a search of a large drug formulary yielded no
information about ECASA. When “magen” (German for
stomach) was searched, there was indeed a mention of
ECASA. It has been mentioned in the literature that a study
of ECASA in continental Europe was limited by the fact
that it was not used in Sweden or Hungary. In a letter to
Lancet in 1994,2 it was pointed out that no reports of
significant gastrointestinal bleeding (compared with placebos) were found.
Charles M. Grossman, MD
Legacy Health System
Department of Medicine
Portland, Ore
doi:10.1016/j.amjmed.2006.04.033
References
1. Dalen JE. Aspirin to prevent heart attack and stroke: what’s the right
dose? Am J Med. 2006;119:198-202.
2. Grossman CM. NSAIDS and gut toxicity. (Letters). Lancet. 1994;344:57.
The American Journal of Medicine (2007) 120, e11
LETTER
Prehypertension and Elevated Risk
of Cardiovascular Disease:
Physiopathologic Mechanisms
To the Editor:
Kshirsagar and colleagues1 recently reported a strong
association between prehypertensive blood pressure levels
and cardiovascular disease. Perhaps a recent study of ours
could cast some light on the mechanism underlying the
increased risk of cardiovascular disease in subjects with
prehypertensive blood pressure levels.2 We performed
Doppler echocardiography and submaximal bicycle ergometry in 20 nonsmoking, sedentary, prehypertensive subjects
and in 20 age, sex-matched, nonsmoking, sedentary, normotensive subjects to investigate the association between
the systolic blood pressure response to exercise (SBPRE)
and hypertensive target organ damage. An exaggerated
SBPRE and a normal SBPRE were diagnosed from the
mean ⫹2 standard deviation of systolic blood pressure at
100 Watts in normotensive subjects. This workload may be
considered to reflect normal daily activities. As expected,
resting blood pressure and systemic vascular resistance
were higher in prehypertensive subjects. Approximately
half the latter group had an exaggerated SBPRE. There were
no differences in age, gender, body mass index, or glycemia
and lipid levels between normotensive and prehypertensive
subjects either with an exaggerated SBPRE or normal
SBPRE. Resting blood pressure and systemic vascular re-
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
sistance were similarly increased in prehypertensive subjects irrespective of SBPRE versus normotensive subjects.
Compared with normotensive subjects, prehypertensive
subjects with an exaggerated SBPRE had a significantly
greater left ventricular relative wall thickness, mostly because of a small cavity, a significantly longer left ventricular
isovolumic relaxation time, and a significantly greater
global arterial stiffness, as estimated by pulse pressure/left
ventricular stoke volume ratio. These findings suggest that
an exaggerated SBPRE is frequent among prehypertensive
subjects and is associated with cardiovascular remodeling,
which may herald cardiovascular disease.
This findings provide a link between prehypertension
and elevated risk of cardiovascular disease.
Flora Affuso, MD
Antonio Ruvolo, MD
Serafino Fazio, MD
Department of Internal Medicine
Federico II University of Naples
Naples, Italy
doi:10.1016/j.amjmed.2006.02.032
References
1. Kshirsagar AV, Carpenter M, Bang H, Wyatt SB, Clindres RE. Blood
pressure usually considered normal is associated with an elevated risk of
cardiovascular disease. Am J Med. 2006;119:133-141.
2. Fazio S, Palmieri EA, Izzo R, et al. An exaggerated systolic blood pressure
response to exercise is associated with cardiovascular remodeling in subjects with prehypertension. Ital Heart J. 2005;6(11):886-892.
The American Journal of Medicine (2007) 120, e13
LETTER
The Safety of Levofloxacin in
Patients on Warfarin
To the Editor:
We would like to commend Stroud and colleagues for
their excellent work regarding the safety of levofloxacin use
in elderly patients on warfarin.1 Their large populationbased study attempts to rectify the conflicting evidence
regarding the warfarin-levofloxacin interaction. They correctly note that prior case reports and case series appear to
contradict data from more controlled trials. They also note
a paucity of studies evaluating this interaction in acutely ill
individuals, their so-called “real-world practice.”
We have recently published data on the impact of the
addition of antibiotics to warfarin in a real-world setting by
studying acutely ill patients who were on stable warfarin
doses and had stable international normalized ratios
(INRs).2 In our retrospective cohort study, the incidence of
supratherapeutic INR was 33% in those who were given
levofloxacin (n ⫽ 27) with a mean increase in INR of .85.
Nineteen percent and 11% of patients had an elevation of
their INR to ⬎ 4.0 and ⬎ 5.0, respectively. Consequently,
our study and that of others clearly suggest that there is an
interaction in terms of elevation of INRs.
There are several limitations to Stroud et al’s study that
may have contributed to their negative findings. First, their
study design does not allow for knowledge of laboratory
monitoring either before or after commencement of the
antibiotic and as such the stability of the patients’ anticoagulation. Second, they were unable to document potential
anticipatory dosing changes (ie, lowering of warfarin doses)
or more frequent monitoring after adding an antibiotic. It is
possible that the reason patients did not bleed was not
because of a lack of interaction but rather because of increased monitoring and dose adjustments. Third, significant
elevations of the INR may have occurred, which may not
have led to hospitalization for bleeding, but still put the
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
patient at higher risk for a hemorrhagic complication.
Fourth, the study database did not evaluate emergency department or clinic visits, so it is possible that some less
severe hemorrhages were missed. Finally, it is also possible
that co-existing medication changes around the time of
illness could have reduced the hypoprothrombinemic effects
of warfarin, thus offsetting the potentiating effects of
levofloxacin.
Even though our data, like Stroud and colleagues’, did
not demonstrate an increase in hemorrhagic events, this
likely relates to the small sample size (warfarin and levofloxacin users) in both studies. Still, we did show a very
significant increase in INR, which is a strong risk factor for
warfarin-related bleeding, and we believe that the remaining
body of literature still strongly supports the interaction between warfarin and levofloxacin in acutely ill patients.
Thus, the results of this research should not alter the clinical
dictum that close monitoring of INR levels after the addition
of any medication, including quinolone antibiotics, is
mandatory.
Jeffrey J. Glasheen, MD
Division of General Internal Medicine
Department of Medicine
University of Colorado at Denver Health Sciences Center
Allan V. Prochazka, MD, MSc
Division of General Internal Medicine
Department of Medicine
Denver VA Medical Center
University of Colorado at Denver Health Sciences Center
doi:10.1016/j.amjmed.2006.02.034
References
1. Stroud LF, Mamdami MM, Kopp A, Bell CM. The safety of levofloxacin in elderly patients on warfarin. Am J Med. 2005;118:1417.e71417.e12.
2. Glasheen JJ, Fugit RV, Prochazka AV. The risk of overanticoagulation
with antibiotic use in outpatients on stable warfarin regimens. J Gen
Intern Med. 2005;20:653-656.
The American Journal of Medicine (2007) 120, e15-e16
LETTER
The Reply:
Glasheen and Prochazka highlight some limitations that
they suggest explain our study’s negative findings.1 Specifically, they describe the absence of clinical information and
our inability to document warfarin dose changes. They posit
that co-existing medication changes may have had a protective effect on patients and our failure to include less
severe patient outcomes could have biased the findings.
The administrative data we used was detailed, but we
were unable to control for changes in drug dosage or patient
monitoring. Indeed, significant elevations in international
normalized ratio (INR) could have occurred in many patients. However, the association between INR and hemorrhage is not perfect. First, assessing the effect of INR level
is problematic unless data are available for the time period
just preceding and at the time of the bleed. This is often not
feasible, even in prospective trials.2 Second, the INR level
may increase as a consequence of bleeding. Therefore, data
at the time of the hemorrhage may not reflect the effect of
over-anticoagulation. Third, bleeding can occur irrespective
of the INR level, whether supra-therapeutic (⬎4.5) or subtherapeutic (⬍2.0).3,4 Moreover, our model incorporated
drugs interacting with warfarin. Finally, our analysis did
control for physician consultations to address the intensity
of anticoagulation monitoring.
Including less severe outcomes such as clinic appointments or emergency department visits may have captured
more adverse events and increased statistical power. However, we could then be criticized for heterogeneous composite endpoints that include diverse severity levels ranging
from an added physician visit to death (considerably different outcomes when considering a patient’s perspective).
Had we demonstrated an adverse effect in this circumstance,
others might argue for more severe clinically relevant
endpoints.
Glasheen and Prochazka also compare their study results
with our work.5 Some of the variation in findings may relate
to the different designs, choice of comparators, and statistical power. Our study used a population-based, nested
case-control method. There were 4269 people who were on
chronic warfarin therapy matched to 17,048 control patients
requiring treatment with an antibiotic for similar disease
indications. We accounted for many covariates. In contrast,
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
Glasheen et al used a cohort from a single Veteran’s Affairs
Center. They assessed INR levels in 27 patients on warfarin
also prescribed levofloxacin. They compared their findings
with 3 other antibiotics that also demonstrated significant
INR elevations (possibly implicating the effect of the illness). Further, most patients had no significant INR elevation. No hemorrhagic outcomes were observed.
We are unsure whether the main reason both our studies
did not demonstrate increased hemorrhagic events for patients prescribed levofloxacin while on warfarin is due to
small sample size. Indeed, it is difficult to detect these
events in population-based research. However, we were
able to demonstrate an effect in patients prescribed cefuroxime, a drug with no previous reports of warfarin interaction. Much of the criticism raised by Glasheen and
Prochazka, such as warfarin dosing, INR monitoring, and
interacting drugs, are part of usual clinical practice. By
including them in a real-world analysis, we maximize the
generalizability of our findings. Still, we agree with their
caution that patients on warfarin should be closely monitored after the addition of any medication.
Chaim M. Bell, MD, PhDa,b
Lynfa F. Stroud, MDc
Alex Kopp, BAd
Muhammad M. Mamdami, PharmD, MA, MPHe,f
a
Departments of Medicine and Health Policy Management and
Evaluation
University of Toronto
b
The institute for Clinical Evaluative Sciences
Department of Medicine
St. Michael’s Hospital
Toronto, Canada
c
Department of Medicine
University of Toronto
Toronto, Canada
d
The Institute for Clinical Evaluative Sciences
Toronto, Canada
e
Department of Pharmacy
University of Toronto
Toronto, Canada
f
Pfizer Inc.
New York, New York
doi:10.1016/j.amjmed.2006.06.030
e16
References
1. Stroud LF, Mamdami MM, Kopp A, Bell CM. The safety of levofloxacin in elderly patients on warfarin. Am J Med. 2005;118:1417.e71417.e12.
2. Douketis JD, Arneklev K, Goldhaber SZ, et al. Comparison of bleeding
in patients with nonvalvular atrial fibrillation treated with ximelagatran
or warfarin: assessment of incidence, case-fatality rate, time course and
sites of bleeding, and risk factors for bleeding. Arch Intern Med.
2006;166:853-859.
The American Journal of Medicine, Vol 120, No 4, April 2007
3. Kucher N, Connolly S, Beckman JA, et al. International normalized
ratio increase before warfarin-associated hemorrhage: brief and subtle.
Arch Intern Med 2004;164:2176-2179.
4. Fanikos J, Grasso-Correnti H, Shah R, et al. Major bleeding complications in a specialized anticoagulation service. Am J Cardiol. 2005;96:
595-598.
5. Glasheen JJ, Fugit RV, Prochazka AV. The risk of overanticoagulation
with antibiotic use in outpatients on stable warfarin regimens. J Gen
Intern Med. 2005;20:653-656.
The American Journal of Medicine (2007) 120, e17-e18
LETTER
Porcelain Gallbladder and Cancer: Ethnicity
Explains a Discrepant Literature?
Table 2 Two-by-Two Table Comparing Risk of Cancer
According to Ethnicity
To the Editor:
Lee and colleagues1 are to be congratulated for an eloquent report of a case of porcelain gallbladder. Citing 2
recent studies, they discuss the possibility that the rate of
gallbladder cancer occurring in calcified (porcelain) gallbladders is lower than previously thought (0%2-5%3 vs
12%4-62%5). However, 2 important aspects of this sparse
and inconsistent literature warrant further comment.
First, the conclusions of 2 recent cited studies suggesting
that the incidence of cancer does not depend on the presence
of porcelain gallbladder should be regarded with caution.
Although the number of patients undergoing cholecystectomy in those North American studies was 5- to 10-fold
larger than in the largest older study (South American),5 the
number of patients with porcelain gallbladders was actually
smaller in the North American studies compared with the
South American study. In such a small cohort (17 patients as
in 1 North American study), for example, the reported rate
of cancer per porcelain gallbladder would increase from 0%
to 6% or 12% if just 1 or 2 additional patients were diagnosed with cancer, approaching the rate of the older studies
and rendering their conclusion incorrect.2
The second point regards a possible explanation for the
inconsistent literature. Consider the hypothesis that recent
North American investigators,2,3 conducting well-designed
retrospective studies, found such different rates compared
with those reported previously because of the influence of
ethnic differences between cohorts, a factor that is known to
Table 1
First
Author
5
Etala
Ashur7
Shimizu8*
Kane9
Cornell4
Stephen3
Towfigh2
Cancer
No cancer
Total
Asia/Latin America
US
Total
20
32
52
6
51
57
26
83
109
By a Fisher exact test (2-tail), P ⫽ .0007.
affect the rate of gallbladder cancer.6 Table 1 shows a global
review of the rates of cancer in porcelain gallbladders according to the ethnicity of the patient population reported.
By a Fisher exact test comparing the rate of cancer in
patients with porcelain gallbladder in North America versus
Asia or Latin America, ethnicity is associated with the
presence of cancer occurring in a porcelain gallbladder
(Table 2) (P ⫽ .0007). This statistic is even more significant when nonsystematic case series are excluded, considering only the formal cholecystectomy series and holds
regardless of whether the cases in the Stephen study are
counted as porcelain gallbladders (0% rate of concomitant
cancer in their study) or selective mucosal calcification (5%
rate of concomitant cancer). Certainly, the reason that more
recent series have reported lower rates of cancer in patients
with porcelain gallbladder could be the more careful and
systematic nature of recent studies of large patient cohorts
or the fact that the diseased organs are being recognized
earlier with widely available computed tomography
scanning.
Global Review of Cancer Occurring in Porcelain Gallbladders
Year
County
N Cholecystectomies
N Porcelain
N Cancer
Cancer/Porcelain
1962
1978
1989
1984
1959
2001
2001
Argentina
Israel
Japan
US
US
US
US
1786
NSCS
NSCS
NSCS
45
25,900
10,741
26
5
21
9
16
17
15
16
1
3
4
2
0
0
62%
20%
14%
44%
13%
0%
0%
NSCS ⫽ nonsystematic case series.
*Nine cases from the US literature9 are excluded, leaving 21 Japanese cases.
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
e18
The American Journal of Medicine, Vol 120, No 4, April 2007
The patient reported by Lee and colleagues1 proceeded
through uneventful cholecystectomy. Discussions of decreased cancer risk in porcelain gallbladders, such as by Lee
and colleagues, have led others to conclude that routine
cholecystectomy may not be indicated for asymptomatic
patients with porcelain gallbladder.2,3 We believe that when
counseling patients regarding the possibility that they harbor an occult gallbladder carcinoma within a porcelain gallbladder, physicians should consider ethnicity.
Steven C. Cunningham, MD
H. Richard Alexander, MD
Department of Surgery
University of Maryland Medical Center
Baltimore
doi:10.1016/j.amjmed.2006.05.028
References
1. Lee TC, Liu KL, Lai IR, Wang HP. Diagnosing porcelain gallbladder.
Am J Med. 2005;118:1171-1172.
2. Towfigh S, McFadden DW, Cortina GR, et al. Porcelain gallbladder is
not associated with gallbladder carcinoma. Am Surg. 2001;67:7-10.
3. Stephen AE, Berger DL. Carcinoma in the porcelain gallbladder: a
relationship revisited. Surgery. 2001;129:699-703.
4. Cornell CM, Clarke R. Vicarious calcification involving the gallbladder.
Ann Surg. 1959;149:267-272.
5. Etala E. [Gallbladder cancer]. Prensa Med Argent. 1962;49:2283-2299.
6. Lazcano-Ponce EC, Miquel JF, Munoz N, et al. Epidemiology and molecular pathology of gallbladder cancer. CA Cancer J Clin. 2001;51:349-364.
7. Ashur H, Siegal B, Oland Y, Adam YG. Calcified gallbladder (porcelain
gallbladder). Arch Surg. 1978;113:594-596.
8. Shimizu M, Miura J, Tanaka T, Itoh H, Saitoh Y. Porcelain gallbladder:
relation between its type by ultrasound and incidence of cancer. J Clin
Gastroenterol. 1989;11:471-476.
9. Kane RA, Jacobs R, Katz J, Costello P. Porcelain gallbladder: ultrasound and CT appearance. Radiology. 1984;152:137-141.
The American Journal of Medicine (2007) 120, 370-373
BRIEF OBSERVATION
Late Diagnosis of HIV Infection: The Role of Age and Sex
Michael J. Mugavero, MD, MHS,a Chelsea Castellano, BS,b David Edelman, MD, MHS,c,d Charles Hicks, MDb
a
Division of Infectious Diseases and International Health, Department of Medicine, University of Alabama at Birmingham; bDivision of
Infectious Diseases and International Health, Department of Medicine, Duke University, Durham, NC; cCenter for Health Services
Research in Primary Care, Durham VA Medical Center, Durham, NC; and dDivision of General Internal Medicine, Department of
Medicine, Duke University, Durham, NC.
ABSTRACT
BACKGROUND: Late diagnosis of human immunodeficiency virus (HIV) infection is detrimental to
infected persons and to the public health. The objective of this study was to identify factors associated with
late diagnosis of HIV infection, defined as an initial CD4 count ⬍200 cells/␮L, in a cohort of recently
diagnosed persons. Additionally, we evaluated factors associated with HIV infection being diagnosed
during hospitalization.
METHODS: This was a cross-sectional study of a university-based HIV clinic in the southeastern US.
Patients with newly diagnosed HIV infection evaluated at the Duke University HIV clinic between October
2002 and August 2004 were included in this analysis. Socio-demographic variables, site of HIV diagnosis,
opportunistic infections present at diagnosis, initial CD4 count, and initial HIV RNA level were recorded
for study subjects.
RESULTS: Forty-nine percent of subjects met the immunologic definition of AIDS at the time of HIV
diagnosis (CD4 count ⬍200 cells/␮L). In multivariable logistic regression analyses, older patients were
more likely to be diagnosed with a CD4 count ⬍200 cells/␮L (adjusted odds ratio [AOR] 1.72, 95%
confidence interval [CI], 1.12-2.64, P ⫽ .01), and older patients (AOR 1.79, 95% CI, 1.07-3.12, P ⫽ .03)
and women (AOR 6.74, 95% CI, 2.08-21.81, P ⫽ 0.001) were more likely to be diagnosed during
hospitalization.
CONCLUSIONS: Late diagnosis of HIV infection is a considerable problem, particularly for older
patients. Inpatient diagnosis of HIV infection is significantly more common among women and older
patients. Improved HIV testing strategies may allow for more timely diagnosis of HIV infection, which
may benefit both the infected individual and society. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: AIDS; Diagnosis; HIV; Prevention; Rural; South
Highly active antiretroviral therapy (HAART) has proven
effective in reducing morbidity and mortality associated
with human immunodeficiency virus (HIV) infection.1 Despite increased availability of HAART in the United States,
These data were presented in part at the 3rd International AIDS Society
Conference on HIV Pathogenesis and Treatment in Rio de Janeiro, Brazil
July 25-28, 2005.
Supported by Agency for Healthcare Research and Quality (AHRQ)
training Grant T32 HS000079 and NIH K24 Grant NIAID 5 K24
AI001608.
Requests for reprints should be addressed to Charles Hicks, MD,
Division of Infectious Diseases and International Health; Duke University
Medical Center, Box 3360, Durham, NC 27710.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2006.05.050
many HIV-infected persons do not fully benefit from therapy due to late diagnosis. Approximately 25% of the estimated 1.1 million persons living with HIV in the United
States are unaware they are HIV-infected.2 Further, recent
data from the Centers for Disease Control and Prevention
(CDC) indicate that approximately half the persons meeting
the case definition for acquired immune deficiency syndrome (AIDS) had been diagnosed with HIV infection less
than a year before their AIDS diagnosis.3 This late diagnosis
of HIV infection is detrimental both to the individual and to
the public health.
Studies assessing factors associated with late presentation for HIV care have focused on patients living in large
urban settings in the Northeast and West.4,5 However, the
Mugavero et al
Late Diagnosis of HIV Infection
371
South has the highest rates of new HIV infections and is
RESULTS
home to 39% of persons living with AIDS in the US.6
Data were available for 113 patients who met study incluAdditionally, some socio-demographic patient characterission criteria (Table 1). Initial CD4 counts were ⬍200
tics, such as rural living, are common in the South. Over two
cells/␮L in 49% of subjects, and 35% were diagnosed durthirds of patients with AIDS and residing in non-metropoling hospitalization.
itan areas live in the South.7 A
study done over a decade ago
Initial CD4 Count <200
showed that only 12% of reCells/␮L
CLINICAL SIGNIFICANCE
cently diagnosed HIV-infected
In bivariable and multivariable
patients in South Carolina had
● Forty-nine percent of recently diagnosed
analyses, older patients were more
8
CD4 counts ⬍200 cells/␮L.
HIV-infected patients had clinical or imlikely to be diagnosed with a CD4
This suggests that late diagnosis
munologic AIDS (CD4 T lymphocyte
count ⬍200 cells/␮L (adjusted
was relatively infrequent in that
count ⬍200 cells/␮L).
odds ratio [AOR] 1.72, 95% conera. Much has changed in the
fidence interval [CI], 1.12-2.64,
past 10 years regarding features
● Older patients were more likely to have
P ⫽ .01) (Table 2).
of persons living with HIV/
an AIDS diagnosis at the time HIV was
AIDS, including increased prevdiagnosed.
HIV Diagnosis during
alence rates among AfricanHospitalization
● Older patients and women were more
Americans,
Latinos,
and
In bivariable analysis, this outlikely to be diagnosed with HIV during
women.6 Further, public health
come was more common among
hospitalization.
funding has decreased over this
older patients (P ⫽ .02), females
period. Given the importance of
● These findings support the recent CDC
(P ⬍.001), and minorities (P ⫽ .02).
the timing of HIV diagnosis for
call for routine, nationwide HIV testing
In multivariable analysis, HIV dithe individual and the public
in primary care settings.
agnosis during hospitalization was
health, an assessment of HIV dimore common among older paagnosis among a Southern cohort
tients (AOR 1.79, 95% CI, 1.07in the current era seemed
3.12, P ⫽ .03) and females (AOR
appropriate.
METHODS
Study Design
We evaluated medical records of clinic patients initially
presenting for care at the Duke University HIV clinic between October 2002 and August 2004. Persons whose HIV
infection had been diagnosed within 6 months of presentation to the clinic were included in this analysis. Abstracted
data included socio-demographic and clinical information.
Non-Caucasians were classified as racial/ethnic minorities.
Consistent with the CDC classification, persons not residing
in a metropolitan statistical area (population ⬍50,000) were
considered to reside in nonmetropolitan locales.6,7 Age was
categorized in 10-year increments. The study was approved
by the Duke University Institutional Review Board.
Statistical Analysis
Bivariable analyses were done using Pearson’s chi-square
test and logistic regression. The study had two primary
outcome measures; an initial CD4 count ⬍200 cells/␮L and
an initial diagnosis of HIV infection during hospitalization.
Multivariable logistic regression analyses were performed
for both outcome variables. Predictor variables were identified a priori and included age, sex, race/ethnicity, insurance status, and nonmetropolitan residence. All statistical
analyses were performed using SAS E-Guide Version 2.05
for Windows (SAS Institute, Cary, NC).
Table 1 Patient Characteristics of 113 Recently Diagnosed
HIV-Infected Subjects Establishing Care at the Duke
University HIV Clinic
Characteristic
Age (range 17-61 years)
Female
Race
Caucasian
Minority*
Unknown
Reported mode of HIV acquisition
Men who have sex with men
Heterosexual contact
Intravenous drug use
Other/unknown
Uninsured
Non-metropolitan
Site of diagnosis
Inpatient HIV diagnosis
Outpatient HIV diagnosis
Unknown
Opportunistic infection at diagnosis†
HIV RNA ⬎100,000 copies/mL
CD4 Lymphocyte count ⬍200 cells/␮L
Mean ⫾ SD or
Frequency (%)
36.1 ⫾ 10
31 (27.4%)
35 (31.0%)
71 (62.8%)
7 (6.2%)
33
32
8
40
44
32
(29.2%)
(28.3%)
(7.1%)
(35.4%)
(38.9%)
(28.3%)
39
54
20
20
66
55
(34.5%)
(47.8%)
(17.7%)
(17.9%)
(58.9%)
(48.7%)
*87% African American, 10% Latino.
†Opportunistic infections: Pneumocystis carinii pneumonia (PCP) 9,
Cryptococcal meningitis 5, Lymphoma 2, Cryptosporidial diarrhea 1,
CNS toxoplasmosis 1, Progressive multifocal leukoencephalopathy 1,
Kaposi’s sarcoma 1.
372
Table 2
The American Journal of Medicine, Vol 120, No 4, April 2007
Bivariable and Multivariable Logistic Regression Analyses*
CD4 Count ⬍200 cells/␮L (n ⫽ 113)
Inpatient HIV Diagnosis (n ⫽ 93)
Bivariable Analysis
Bivariable Analysis
Multivariable Analysis
Characteristic
OR (95% CI)
Age† (range
17-61 years)
Female
Minority
Uninsured
Non-metropolitan
1.74 (1.16-2.60) .005
1.72 (1.12-2.64) .01
1.63 (1.07-2.56)
1.68
1.37
0.94
1.28
1.30
1.56
1.06
1.19
6.71
3.27
1.71
1.58
(0.73-3.88)
(0.61-3.08)
(0.44-2.00)
(0.57-2.91)
P-value OR (95% CI)
.22
.45
.87
.55
(0.51-3.35)
(0.63-3.87)
(0.46-2.45)
(0.49-2.87)
P-value OR (95% CI)
.59
.34
.89
.71
Multivariable Analysis
P-value OR (95% CI)
.02
(2.52-17.87) ⬍.001
(1.20-8.87)
.02
(0.74-4.00)
.21
(0.63-3.93)
.33
P-value
1.79 (1.07-3.12)
.03
6.74
2.63
2.99
1.32
.001
.10
.05
.61
(2.08-21.81)
(0.82-8.43)
(0.99-9.00)
(0.44-3.99)
*Variables associated with immunologic AIDS (CD4 count ⬍200 cells/␮L) and inpatient HIV diagnosis among 113 recently diagnosed HIV-infected
subjects establishing care at the Duke University HIV Clinic.
†Odds ratios for 10-year increase in age.
6.74, 95% CI, 2.08-21.81, P ⫽ .001), whereas lack of insurance was of borderline statistical significance (AOR 2.99,
95% CI, 0.99-9.00, P ⫽ 0.05) (Table 2).
DISCUSSION
Despite efforts to increase HIV prevention and awareness,
half the recently diagnosed HIV-infected patients evaluated
at a university-based HIV clinic in North Carolina had
initial CD4 counts ⬍200 cells/␮L, compared with only 12%
of those reported a decade earlier from South Carolina.8
This may relate to changes in socio-demographic characteristics among persons living with HIV/AIDS, such that persons now at greatest risk are less likely to be tested due to
issues of stigma and lower perceived risk. The ready availability of effective antiretroviral medications may also play
a role by providing a sense of security to at-risk populations,
thereby reducing the urgency with which they seek HIV
testing. Further, decreases in public health spending over
this time may also contribute to this finding.
Late diagnosis of HIV infection is detrimental both to the
individual and to society. HIV-infected patients whose CD4
counts are below 200 cells/␮L experience increased morbidity and mortality, diminished responses to antiretroviral
therapy, and higher health care expenditures.9,10 Focusing
prevention of transmission strategies on HIV-infected persons reduces rates of high-risk sexual behaviors that might
transmit HIV to others.11 Patients diagnosed late have
“missed out” on prevention programs for years while unaware of their infection. Moreover, antiretroviral therapy
that reduces HIV RNA levels almost certainly reduces risk
of transmission to sexual partners.12 The delay in initiating
antiretroviral therapy due to late diagnosis thus likely generates more secondary HIV infections.
Testing based on perceived risk of HIV infection, rather
than broader screening approaches, likely contributes to the
problem. This study shows that older patients and women
are more likely to be diagnosed during hospitalization. This
suggests that these populations may have lower perceived
risk for HIV infection by both themselves and by their
providers, and are therefore less likely to be tested for HIV
as outpatients.
As a single, academic medical center study, our findings
may not be generalizable to other medical settings or regions of the country. However, our findings are consistent
with other recently reported estimates, including a cohort
of patients in a large health maintenance organization.4,5
Given the relatively modest sample size, it is possible that
there was insufficient power to identify other important
associations.
These findings support the notion that late diagnosis of
HIV infection is a significant problem, particularly among
older patients. Additionally, these data indicate that diagnosis of HIV infection during hospitalization is common,
especially among older patients and women. Recent initiatives call for HIV testing to become a more routine part of
medical care,13 in part because of the problem of late HIV
diagnosis substantiated by this study. Routine rather than
risk-based HIV testing is advocated, because high-risk behaviors are frequently not identified in primary care encounters.5 Prospective studies assessing the impact of routine
screening in various medical settings are ongoing and will
be important in developing an evidence-based HIV screening policy.
References
1. Palella FJ Jr, Delaney KM, Moorman AC, et al. Declining morbidity
and mortality among patients with advanced human immunodeficiency
virus infection. HIV Outpatient Study Investigators. N Engl J Med.
1998;338:853-860.
2. Glynn M, Rhodes P. Estimated HIV prevalence in the United States at
the end of 2003. In: Program and Abstracts of the National HIV
Prevention Conference (Abstract 595), Atlanta, GA, June 2005.
3. Centers for Disease Control and Prevention. Late vs. early testing of
HIV—16 sites, United States, 2000-2003. MMWR Morb Mortal Wkly
Rep. 2003;52:581-586.
4. Dybul M, Bolan R, Condoluci D, et al. Evaluation of initial CD4⫹ T
cell counts in individuals with newly diagnosed human immunodeficiency virus infection, by sex and race, in urban settings. J Infect Dis.
2002;185:1818-1821.
5. Klein D, Hurley LB, Merrill D, Quesenberry CP Jr. Review of medical
encounters in the 5 years before a diagnosis of HIV-1 infection:
Mugavero et al
6.
7.
8.
9.
Late Diagnosis of HIV Infection
implications for early detection. J Acquir Immune Defic Syndr.
2003;32:143-152.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
Report, 2003 (Vol. 15). Atlanta, GA: US Department of Health and
Human Services, Centers for Disease Control and Prevention; 2004:
1-46.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance
in Urban and Nonurban Areas L206 slide series. Available at: http://
www.cdc.gov/hiv/graphics/rural-urban.htm. Accessed March 1, 2006.
Luby S, Jones J, Horan J. Using CD4 counts to evaluate the stages and
epidemiology of HIV infection in South Carolina public clinic patients. Am J Public Health. Mar 1994;84(3):377-381.
Egger M, May M, Chene G, et al. Prognosis of HIV-1-infected patients
starting highly active antiretroviral therapy: a collaborative analysis of
prospective studies. Lancet. 2002;360:119-129.
373
10. Chen RY, Westfall AO, Accort NA, et al. Distribution of health care
expenditures for HIV-infected patients. Clin Infect Dis. 2006;42:10031010.
11. Marks G, Crepaz N, Senterfitt JW, Janssen RS. Meta-analysis of
high-risk sexual behavior in persons aware and unaware they are
infected with HIV in the United States: implications for HIV
prevention programs. J Acquir Immune Defic Syndr. 2005;39:446453.
12. Quinn TC, Wawer MJ, Sewankambo N, et al. Viral load and heterosexual
transmission of human immunodeficiency virus type 1. Rakai Project
Study Group. N Engl J Med. 2000;342:921-929.
13. Centers for Disease Control and Prevention. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in
health-care settings. MMWR Morb Mortal Wkly Rep. 2006;55(RR14);
1-17.
APM Perspectives
APM Perspectives
The Association of Professors of Medicine (APM) is the national organization of departments of internal
medicine at the US medical schools and numerous affiliated teaching hospitals as represented by chairs and
appointed leaders. As the official sponsor of The American Journal of Medicine, the association invites
authors to publish commentaries on issues concerning academic internal medicine.
For the latest information about departments of internal medicine, please visit APM’s website at
www.im.org/APM.
The Lost Art of Clinical Skills
Christopher A. Feddock, MD, MS
Department of Internal Medicine, University of Kentucky, Lexington.
With the rapid extension of laboratory tests of
greater accuracy, there is a tendency for some clinicians and hence for some students in reaching a
diagnosis to rely more on laboratory reports and less
on the history of the illness, the examination and
behavior of the patient, and clinical judgment. While
in many cases laboratory findings are invaluable for
reaching correct conclusions, the student should
never be allowed to forget that it takes a man, not a
machine, to understand a man.1
Although seemingly contemporary, this critique
about the use of clinical skills in medical education was
made in 1946. By the end of World War II, x-rays,
which formerly entailed several minutes of exposure
time, could be performed in a matter of seconds and, for
the first time, arterial blood gas measurement was possible. As attractive as these technologies seemed in the
1940s, the advancements of medical technology since
then are even more alluring. Yet, technology seems to
be replacing basic medical skills rather than complementing them. In “Introduction to Clinical Medicine: A
Time for Consensus and Integration,” Omori and colleagues discuss concerns over history-taking and physical examination instruction, particularly during the
first 2 years of medical school.2 The authors recommend a national collaboration on the integration of
clinical skills education through medical school curriculum. Although this is a critical initial step, beyond the
poor acquisition of basic clinical skills is the documented decline of some rudimentary skills after the
second year of medical school.3,4 Why should historyRequests for reprints should be addressed to Christopher A.
Feddock, MD, Department of Internal Medicine and Pediatrics, University of Kentucky, K522 Kentucky Clinic, Lexington, KY 405360284.
E-mail address: [email protected]
taking and physical examination skills crest when they
remain esoteric concepts and plateau or decrease when
they should be used in actual practice of medicine?
Regardless of the formal educational curriculum, the
skills should improve if they are used to guide patient
care decisions.
Multiple editorials have decried the loss of clinical
skills and medical educators justify not only teaching
but also using basic history-taking and physical examination skills in routine patient care.5-7 However, some
skeptics suggest that emphasis on clinical skills is from
a bygone era and that the availability of advanced
imaging techniques and laboratory tests have supplanted ambiguous history and physical findings.7-9
Numerous studies and meta-analyses describe the poor
operating characteristics of most history-taking and
physical examination findings. For example, in a metaanalysis identifying the symptoms, signs, and tests most
useful in diagnosing congestive heart failure (CHF), no
single history-taking or physical examination findings
provided adequate discrimination.10 The most discerning features of CHF—such as the presence of paroxysmal nocturnal dyspnea, an S3 gallop, or jugular venous
distention— have such a low incidence (each documented in ⬍50% of patients with CHF) that their presence is not particularly helpful in diagnosing CHF. In
fact, the best single predictor of CHF was found to be
a B-type natriuretic peptide (BNP) value of 250 pg/mL
with a sensitivity of 89% and a specificity of 81%.
However, physicians rarely make decisions based
upon one isolated finding, and the overall gestalt of the
patient’s illness is perhaps most important. In the CHF
study, a high initial clinical suspicion (based upon the
complete history and physical) was the most predictive
element in the diagnosis of CHF.10 In fact, the authors
concluded that ordering a BNP level was useful only in
cases with an equivocal clinical suspicion. The impor-
0002-9343/$ -see front matter © 2007 The Association of Professors of Medicine. All rights reserved.
doi:10.1016/j.amjmed.2007.01.023
Feddock
Clinical Skills
tance of clinical judgment was even more apparent in a
study of patients presenting to an outpatient clinic with
either abdominal pain or chest pain.11 Physicians correctly classified 93% of patients with abdominal pain
and 98% of patients with chest pain as having either an
organic or nonorganic etiology solely on the basis of
their initial clinical judgment and before ordering any
diagnostic tests. These findings clearly indicate that
clinical skills are important, but can technology replace
thorough history-taking and physical examinations?
Contrary to the paramount importance often ascribed to technology, numerous studies have demonstrated that technology has not necessarily improved
the quality of patient care. Successive cohorts of autopsies over the last century have consistently demonstrated similar rates of misdiagnosis despite the use of
advanced diagnostic procedures.12,13 For example, unknown malignancies were diagnosed in 36.5% of autopsies in 1923, compared with 41% in 1972 and 44%
in 1998.13 In another study of 167 patients who died
during a stay in the intensive care unit, autopsies uncovered a major diagnostic error in 32% of patients
despite extensive diagnostic testing.14 In fact, the use of
imaging was noted to contribute to misdiagnosis. Physicians missed the diagnosis of endocarditis in 9 patients despite ordering echocardiograms in the week
before each of their deaths. By contrast, 3 patients
diagnosed with endocarditis by echocardiogram subsequently had their diagnoses refuted during autopsy. In
another autopsy study, the authors determined that ultrasound and computed tomography scanning provided
misleading information for 7% of patients, whereas
history and physical examinations rarely misled physicians (⬍1% for history and 2% for physical
examination).12
More recently, the Medicare program evaluated the
use of imaging services nationwide and found a threefold variation in the number of examinations obtained
across the United States.15 Despite substantial differences in the use of imaging studies, no demonstrable
changes in quality or patient outcomes were noted.
Although recent technological advances offer numerous tools to aid in diagnosis, their use should be guided
by thorough history and physical examinations. Studies
have consistently demonstrated that history-taking and
physical examinations are the most important factors in
arriving at a correct diagnosis, whereas lab tests and
imaging studies play only minor roles.12,16 The aforementioned study of autopsy results concluded that history-taking and physical examinations provided conclusive information for determining the main diagnosis
in 73% and 62% of patients, respectively.12 By comparison, imaging techniques provided conclusive information for diagnoses of 35% of patients and standard
lab tests for 22% of patients. In fact, inaccurate, incomplete, or misinterpreted patient histories are among the
375
leading causes for diagnostic errors.8,17 Physicians acknowledge the poor discrimination of individual history and physical examination findings but often neglect to consider the true sensitivity and specificity of
imaging and other tests.6,7 Using technology becomes a
“gold standard” for diagnosis instead of an adjunct to
clinical judgment.
THE STATE OF CLINICAL SKILLS
Inconsistencies between laboratory findings and
clinical data go undetected simply because too many
physicians are insufficiently disciplined in the
proper use of clinical skills and in the analysis of
clinical data. Too often, palpably illogical laboratory
findings are accepted without question.6
Despite the importance of history-taking and physical examination, clinical skills education has decreased
since the 1960s, with deficiencies beginning in medical
school and continuing through residency and into practice.3-5,9,18-22 With the erosion of thorough historytaking and physical examinations, clinical reasoning
(the ability to develop the gestalt impression) has also
decreased.2,5 Although essential to determining the correct diagnosis, appropriate history-taking and physical
examination are futile without the ability to interpret
gathered information. History-taking is frequently limited by close-ended questions that fail to gather specific
details critical to clinical decision-making.5-8 For example, instead of determining the severity, context,
alleviating factors, exacerbating factors, and chronology of dyspnea, its presence becomes a yes or no
response. Beyond the clear decline in skills, medical
students who demonstrate proper technical ability have
serious deficiencies in clinical reasoning before they
enter residency training.5-8,21 In a study comparing student scores on an objective structured clinical examination (OSCE), which requires a complete organ system examination and a clinical performance
examination (CPX), which requires a focused physical
based on the patient’s history, a large discrepancy was
noted among student scores.21 Students who excelled in
the technical examination (OSCE) performed inconsistently when deciding on the appropriate physical examination elements during the patient scenario (CPX).
In fact, the two scores showed no correlation on individual cases. Apparently, being able to perform a physical examination correctly during an OSCE (the standard used by most medical schools) does not translate
into appropriate use of those skills in patient care.
CLINICAL SKILLS IN TEACHING HOSPITALS
A glimpse of clinical education during medical school
and residency illuminates several reasons for the overall decline in clinical decision-making skills. Before the
376
The American Journal of Medicine, Vol 120, No 4, April 2007
1970s, internal medicine rounds with the attending physician were considered the platform for demonstrating
history-taking, physical examinations, and clinical decision-making.5,22 Attending physicians demonstrated
clinical skills, refined a physician-in-training’s techniques, and corrected errors or misinterpretations. By
1978, the rate of performing bedside examinations during teaching time had decreased to 16%, and many
physicians estimate that number has decreased even
further.5,22 Today, physicians-in-training rarely see attending physicians demonstrate history-taking and
physical examination techniques and apply their findings to clinical decision-making.5,22,23 Rounds at many
hospitals have shifted from bedside interactions with
patients focused on the patient’s history and physical
examination to conference room learning focused on
the patient’s latest laboratory and imaging results.5,22 A
recent national survey of medical students inquired
about the quality of teaching during the inpatient component of their internal medicine clerkships.23 Onethird of students reported that their attendings rarely or
never saw new patients with the team, not including the
number who saw but did not examine the patients with
the team.
Not only are senior physicians failing to demonstrate
clinical skills, but physicians-in-training are rarely observed taking histories or performing physical examinations, or offered feedback on their clinical
skills.6,19,22,24,25 Less than 50% of medical students
reported that the attending physician observed them
performing some aspect of a physical examination, and
less than 20% reported they were observed interviewing a patient.24 Even when faculty members do observe
a student’s performance, many do not identify the errors made by those students during their examinations.25 Observation and feedback are essential because
physicians-in-training commit a high number of errors
in routine history-taking and physical examinations.8,17,19,20 These errors have significant consequences; in one study, an attending physician’s physical examination ultimately changed the diagnosis or
disease management for one-quarter of the patients
admitted to an internal medicine service, providing
proof that any feedback on clinical skills is beneficial.20
In another study of the accuracy of resident presentations, resident physical examination skills directly correlated to the amount of time attending physicians spent
with them at the bedside examining patients.18 However, physical examination skills did not correlate with
classic measures of medical knowledge, such as resident in-training examination scores or prior medical
school class rank, indicating those measurements do
not automatically translate to clinical skills.
Despite its decline, both medical educators and physicians-in-training still consider bedside teaching one of
the most valuable learning tools.22,23 Although many
lament the changes in medical education focus, medical
schools overall have done little to improve the way
clinical skills are taught.6,7,26,27 Since the 1980s, innovations in medical education have attempted to refocus
the activities of medical schools and teaching hospitals
back on education.26-28 New modalities, such as standardized patients, simulators, and computer technology, have been included to improve the teaching of
clinical skills. However, these recent instructional
methods have one common factor—they require less
faculty involvement. In a recent case study, faculty
members at one medical school consistently made curricular decisions that would best preserve faculty research time, resulting in a persistent decrease over the
last half-century in student-faculty contact.27 Identifying faculty who will take the time to teach medical
students is now one of the most substantial difficulties
faced by course directors.28 Although many of the new
teaching technologies offer outstanding, nonintimidating platforms for students to refine their skills, they
cannot replace patient interactions supervised by
faculty.29
CONCLUSION
Current movements to improve clinical skills education
are essential and long overdue. Acknowledging that the
provision of high quality medical care requires strong
clinical skills and clinical reasoning, the Association of
American Medical Colleges established the Task Force
on the Clinical Skill Education of Medical Students.
Several recommendations from this task force emphasize the importance of clinical skills.29 First, the task
force acknowledged that clinical skills education is a
longitudinal process that must be taught throughout
medical school, residency training, and postgraduate
work. Clinical skills are often ignored after the second
year of medical school, even though they should be
continuously reinforced and advanced to include more
sophisticated techniques. Second, the task force emphasized that faculty members must take the primary role
in teaching clinical skills. The development of clinical
skills requires close mentorship with someone who can
not only teach the specified skills, but also assess and
provide feedback to the student. Third, the task force
recommended that the evaluation of clinical skills must
be patient-centered. Simulators and computerized technology are useful adjuncts to teaching clinical skills,
but to deliver high-quality patient care, students must
learn with actual patients.
Emphasizing clinical skills is as significant as acknowledging the impact of the hidden curriculum perpetuated in many clinical arenas.5,30 Students and residents will value what their faculty mentors value and
their observations of faculty-patient interactions demonstrate what skills are truly important to succeed as a
physician. Rushed rounds with attending physicians
Feddock
Clinical Skills
exaggerate the importance of ancillary testing at the
expense of clinical skills. With a greater reliance on
laboratory studies and imaging, the details of history
and physical examinations become inconsequential and
no longer change the diagnostic approach.7,9 Physician
educators must not only dedicate themselves to teaching clinical skills, but also emulating those skills in
everyday work. They must integrate new technology
into clinical education without diverting attention from
patients. Expectations for physicians-in-training should
include a progressive improvement in clinical skills,
guided by faculty feedback to correct and advance
those skills.
Although clinical faculty members play an essential
role in promoting the development of necessary clinical
skills in physicians-in-training, they must have sufficient support from medical school or teaching hospital
leadership, as well as individual departments. Even
motivated teaching faculty find numerous barriers
within the administration of the medical school or
teaching hospital.28,31-33 Medical education, particularly basic clinical skills education, is a low priority,
especially in comparison to the income-generating endeavors of research and clinical work.28,31 In general,
faculty are given little time to dedicate to teaching
duties, so they either risk salary cuts or they teach “on
their own time” after completing their clinical or research activities.32 At one medical school, it was estimated that, after paying for fringe benefits, faculty
members were compensated approximately $16 for
each hour they spent teaching.31 In addition, teaching
activities often do not contribute significantly to promotion and tenure decisions.33 Many medical school
promotion and tenure committees have difficulty accepting the expanded definition of scholarship that rewards certain teaching activities.34 Ultimately, medical
schools and teaching hospitals must instill value in
teaching basic clinical skills by providing physicians
the time and the compensation for these activities.
Teaching clinical skills is time-intensive and requires dedicated faculty who are able to demonstrate,
teach, and provide feedback. Current faculty members
may lack the clinical expertise to be successful, adding
to the difficulty in finding strong physician models
proficient in teaching clinical skills. Most current medical school or teaching hospital faculty completed their
education during a time of diminished emphasis on
clinical skills, so they may lack confidence in their own
skills.4,22 Teaching is not an intuitive endeavor for all
physicians, and faculty are limited in their ability to
assess clinical skills and provide quality feedback to
physicians-in-training.24,25 Faculty development is necessary to improve the current clinical skills of medical
school faculty and provide proper instruction on the
teaching and evaluation of clinical skills.
377
Although history-taking and physical examination
skills are often considered rudimentary, they serve as
the foundation for all clinical decision-making and their
significance should not be disregarded or forgotten.
Modern technology has improved physician understanding of ailments and created new tools to use in
diagnostic paradigms, but the technology is not infallible.7,12-16 Indiscriminate use of new technology will
not improve health care but will only contribute to
spiraling health care costs.15 The enhancement of clinical skills curricula must be accompanied by focusing
clinical training back on patients and away from computerized data. In the words of Sir William Osler, “it is
a safe rule to have no teaching without a patient for a
text, and the best teaching is that taught by the patient
himself.”35
References
1. Allen, RB. Medical Education and the Changing Order. New
York: The Commonwealth Fund; 1946.
2. Omori DM, Wong RY, Antonelli MA, Hemmer PA. Introduction
to clinical medicine: a time for consensus and integration. Am J
Med. 2005;118:189-194.
3. Pfeiffer C, Madray H, Ardolino A, Willms J. The rise and fall of
students’ skill in obtaining a medical history. Med Educ. 1998:
32:283-288.
4. Vukanovic-Criley JM, Criley S, Warde CM, et al. Competency
in cardiac examination skills in medical students, trainees, physicians and faculty. Arch Intern Med. 2006;166:610-616.
5. Ahmed MEK. What is happening to bedside clinical teaching?
Med Educ. 2002;36:1185-1188.
6. Engel GL. Are medical schools neglecting clinical skills? JAMA.
1976;236:861-863.
7. Bordage G. Where are the history and physical? Can Med Assoc
J. 1995;152:1595-1598.
8. Wiener S, Nathanson M. Physical examination: frequently observed errors. JAMA. 1976;236:852-855.
9. Mangione S, Nieman LZ, Gracely E, Kaye D. The teaching and
practice of cardiac auscultation during internal medicine and
cardiology training. Ann Intern Med. 1993;119:47-54.
10. Wang CS, Fitzgerald JM, Schulzer M, et al. Does this dyspneic
patient in the emergency department have congestive heart failure? JAMA. 2005;294:1944-1956.
11. Martina B, Bucheli B, Stotz M, et al. First clinical judgment by
primary care physicians distinguishes well between nonorganic
and organic causes of abdominal or chest pain. J Gen Intern Med.
1997;12:459-465.
12. Kirch W, Schafii C. Misdiagnosis at a university hospital in four
medical eras: report on 400 cases. Medicine (Baltimore). 1996;
75:29-40.
13. Burton EC, Troxclair DA, Newman WP. Autopsy diagnoses of
malignant neoplasms: how often are clinical diagnoses incorrect?
JAMA. 1998;280:1245-1248.
14. Combes A, Mokhtari M, Couvelard A, et al. Clinical and autopsy
diagnoses in the intensive care unit: a prospective study. Arch
Intern Med. 2004;164:389-392.
15. Iglehart JK. The new era of medical imaging—progress and
pitfalls. N Engl J Med. 2006;354:2822-2828.
16. Peterson MC, Holbrook JH, Hales DV, et al. Contributions of the
history, physical examination, and laboratory investigation in
making medical diagnoses. West J Med. 1992;156:163-165.
17. Bordage G. Why did I miss the diagnosis? Some cognitive
explanations and educational implications. Acad Med.
1999;74(10 Suppl):S138-S143.
378
The American Journal of Medicine, Vol 120, No 4, April 2007
18. Johnson J, Carpenter JL. Medical house staff performance in
physical examination. Arch Intern Med. 1986;146:937-941.
19. Wray NP, Friedland JA. Detection and correction of house staff
error in physical diagnosis. JAMA. 1983;249:1035-1037.
20. Reilly BM. Physical examination in the care of medical inpatients: an observational study. Lancet. 2003;362:1100-1105.
21. Wilkerson L, Lee M. Assessing physical examination skills of
senior medical students: knowing how versus knowing when.
Acad Med. 2003;78(10 Suppl):S30-S32.
22. Ramani S, Orlander JD, Strunin L, Barber TW. Whither bedside
teaching? A focus-group study of clinical teachers. Acad Med.
2003;78:384-390.
23. Guarino CM, Ko CY, Baker LC. Impact of instructional practices
on student satisfaction with attendings’ teaching in the inpatient
component of internal medicine clerkships. J Gen Intern Med.
2006;21:7-12.
24. Holmboe ES, Fiebach NH, Galaty LA, Huot S. The effectiveness
of a focused educational intervention on resident evaluations
from faculty: a randomized controlled trial. J Gen Intern Med.
2001;16:427-434.
25. Noel GL, Herbers JE, Caplow MP. How well do internal medicine faculty members evaluate the clinical skills of residents?
Ann Intern Med. 1992;117:757-765.
26. Bloom SW. Structure and ideology in medical education: an
analysis of resistance to change. J Health Soc Behav. 1988;29:
294-306.
27. Cuban L. How Scholars Trumped Teachers: Change Without
Reform in University Curriculum, Teaching and Research, 18901990. New York: Teachers College Press; 1999.
28. Mangan KS. Physician, teach thyself. Chron High Educ. 2003;
50:A18.
29. Association of American Medical Colleges. The AAMC Project
on the Clinical Education of Medical Students: Clinical Skills
Education. Washington, DC: AAMC; 2005.
30. Hafferty FW. Beyond curriculum reform: confronting medicine’s hidden curriculum. Acad Med. 1998;73:403-407.
31. Shea S, Nickerson KG, Tenebaum J, et al. Compensation to a
department of medicine and its faculty members for the teaching
of medical students and house staff. N Engl J Med. 1996;334:
162-167.
32. Sheffield JV, Wipf JE, Buchwald D. Work activities of clinicianeducators. J Gen Intern Med. 1998;13:406-409.
33. Batshaw ML, Plotnick LP, Petty BG, et al. Academic promotion
at a medical school: experience at Johns Hopkins University
School of Medicine. N Engl J Med. 1988;318:741-747.
34. Nora LM, Pomeroy C, Curry TE, et al. Revising appointment,
promotion, and tenure procedures to incorporate an expanded
definition of scholarship: the University of Kentucky College of
Medicine experience. Acad Med. 2000;75:913-924.
35. Osler W. Aequanimatus, and Other Addresses. Philadelphia: P.
Blakiston’s Son & Co; 1904.