A Worldwide Perspective of Atypical Pathogens in
Community-acquired Pneumonia
Forest W. Arnold1, James T. Summersgill1, Andrew S. LaJoie1,2, Paula Peyrani1, Thomas J. Marrie3, Paolo Rossi4,
Francesco Blasi5, Patricia Fernandez6, Thomas M. File, Jr.7, Jordi Rello8, Rosario Menendez9, Lucia Marzoratti10,
Carlos M. Luna11, Julio A. Ramirez1, and the Community-Acquired Pneumonia Organization (CAPO) Investigators*
1
Division of Infectious Diseases, Department of Medicine, and 2Department of Health Promotion and Behavioral Sciences, University of Louisville,
Louisville, Kentucky; 3University of Alberta Hospital, Sturgeon Community Hospital, Grey Nuns Hospital, and Royal Alexandra Hospital,
Edmonton, Alberta, Canada; 4Department of Medicine, S. Maria della Misericordia Hospital, Udine, Italy; 5Istituto Malattie Respiratorio, University
of Milan, Istituto di Ricerca e Cura a Carattere Scientifico, Policlinico, Milan, Italy; 6Instituto Nacional del Torax, Santiago, Chile; 7Summa
Health System, Akron, Ohio; 8Joan XXIII University Hospital, Tarragona, Spain; 9Hospital Universitario La Fe, Valencia, Spain; 10Sanatorio 9 de
Julio, Tucuman, Argentina; and 11Hospital de Clinicas, Buenos Aires, Argentina
Rationale: Controversy still exists in the international literature regarding the need to use antimicrobials covering atypical pathogens
when initially treating hospitalized patients with communityacquired pneumonia (CAP). In different regions of the world, monotherapy with a ␤-lactam antimicrobial is common.
Objectives: We sought to correlate the incidence of CAP due to
atypical pathogens in different regions of the world with the proportion of patients treated with an atypical regimen in those same
regions. In addition, we sought to compare clinical outcomes of
patients with CAP treated with and without atypical coverage.
Methods: A secondary analysis was performed using two comprehensive international databases. World regions were defined as
North America (I), Europe (II), Latin America (III), and Asia and
Africa (IV). Time to reach clinical stability, length of hospital stay,
and mortality were compared between patients treated with and
without atypical coverage.
Measurements and Main Results: The incidence of CAP due to atypical
pathogens from 4,337 patients was 22, 28, 21, and 20% in regions
I–IV, respectively. The proportion of patients treated with atypical
coverage from 2,208 patients was 91, 74, 53, and 10% in regions
I–IV, respectively. Patients treated with atypical coverage had decreased time to clinical stability (3.7 vs. 3.2 d, p ⬍ 0.001), decreased
length of stay (7.1 vs. 6.1 d, p ⬍ 0.01), decreased total mortality
(11.1 vs. 7%, p ⬍ 0.01), and decreased CAP-related mortality (6.4
vs. 3.8%, p ⫽ 0.05).
Conclusions: The significant global presence of atypical pathogens
and the better outcomes associated with antimicrobial regimens
with atypical coverage support empiric therapy for all hospitalized
patients with CAP with a regimen that covers atypical pathogens.
Keywords: pneumonia, mycoplasma; pneumonia, pneumococcal;
atypical; community-acquired infection; empiric antibiotic
In an attempt to help the practicing physician in the selection
of appropriate empiric therapy for hospitalized patients with
(Received in original form March 9, 2006; accepted in final form February 23, 2007)
Parts of this article were presented at the 2004 Infectious Diseases Society of
America conference, October 1, 2004, and at the 2005 International Conference
of the American Thoracic Society, May 24, 2005.
* A complete list of study investigators may be found before the reference section.
Correspondence and requests for reprints should be addressed to Forest W. Arnold,
D.O., Division of Infectious Diseases, University of Louisville, Carmichael Building,
Room 208E, 512 South Hancock Street, Louisville, KY 40292. E-mail: f.arnold@
louisville.edu
This article has an online supplement, which is accessible from this issue’s table
of contents at www.atsjournals.org
Am J Respir Crit Care Med Vol 175. pp 1086–1093, 2007
Originally Published in Press as DOI: 10.1164/rccm.200603-350OC on March 1, 2007
Internet address: www.atsjournals.org
AT A GLANCE COMMENTARY
Scientific Knowledge on the Subject
American Thoracic Society guidelines for communityacquired pneumonia recommend the use of antimicrobials
covering atypical pathogens in all hospitalized patients, but
some studies do not show a benefit.
What This Study Adds to the Field
Atypical pathogens are common causes of communityacquired pneumonia in all regions of the world. Treatment
regimens including coverage for atypical pathogens are
associated with improved patient outcome.
community-acquired pneumonia (CAP), national organizations
in multiple countries have developed guidelines for treating
CAP. Despite these contributions, however, controversy still
exists in the literature regarding the best drug regimen to use
as empiric therapy for CAP. Guidelines from the United States,
Canada, Germany, Japan, and parts of the Latin American region recommend using a regimen that covers atypical pathogens
in all hospitalized patients with CAP (1–5). On the other hand,
guidelines from other countries or regions of the world still
recommend using a ␤-lactam regimen while leaving the addition
of a macrolide as an option (6–9).
The controversy regarding the need to cover for atypical
pathogens is primarily due to two significant unresolved questions. The first is related to the incidence of atypical pathogens
as an etiology of CAP in different regions of the world. Because
the identification of atypical pathogens requires special laboratory tests that are not readily available, studies to evaluate local
incidence of atypical pathogens are difficult to perform (10).
Even when local studies are performed, because the diagnostic
tests for atypical pathogens are not well standardized among
laboratories, it is difficult to compare data on incidence from
different studies using different laboratories. The second is related to the contradictory results found in literature when outcomes are compared in patients with CAP treated with or without coverage for atypical pathogens (11).
In an attempt to investigate some of the controversies in the
field of atypical pathogens in CAP, we designed a study with
the following objectives: first, to evaluate in different regions of
the world the incidence of atypical pathogens in patients with
CAP; second, to define if there is a correlation of the incidence
Arnold, Summersgill, LaJoie, et al.: Worldwide View of Atypical Pneumonia
of atypical pathogens in a region with the proportion of patients
treated with empiric therapy covering atypical pathogens in the
same region; and third, to compare early and late CAP outcomes
for hospitalized patients treated empirically with antimicrobials
with and without coverage for atypical pathogens. Some of the
results of these studies have been previously reported in the
form of abstracts (12, 13).
METHODS
A secondary analysis was performed using two comprehensive international databases: The University of Louisville infectious diseases atypical
pathogens reference laboratory database and the Community-Acquired
Pneumonia Organization (CAPO) database.
University of Louisville Infectious Diseases Atypical
Pathogens Reference Laboratory Database
This database was assembled as a result of the Infectious Diseases
Laboratory serving as a reference laboratory for phase III CAP international clinical trials of antimicrobials from the pharmaceutical companies Abbott, Aventis, Pfizer, and Bristol-Myers Squibb. All study sites
obtained institutional review board approval, and patients with CAP
were enrolled in a prospective manner after informed consent. The
database contains information regarding the incidence of atypical
pathogens in 4,337 patients with CAP from 21 countries. Specimens
were collected from September 1996 until April 2004. All specimens
submitted to the Infectious Diseases Laboratory at University of
Louisville (UL) included oropharyngeal swabs for polymerase chain
reaction (PCR) and culture, acute and convalescent serum samples
for antibody titer determination, and urine specimens for Legionella
pneumophila type 1 antigen detection. The diagnosis of atypical pneumonia was based on one or more positive results from among the
following tests: (1 ) for Mycoplasma pneumoniae, PCR (14) detection
and/or culture from an oropharyngeal swab specimen, or acute IgG
titer (⭓ 1:64), IgM titer (⭓ 1:16), or a fourfold increase in either IgG
or IgM in the convalescent specimen by immunofluorescent antibody
assay (Crowntitre system; Zeus Scientific, Inc., Raritan, NJ); (2 ) for
Chlamydia pneumoniae, PCR (14) detection and culture from an oropharyngeal swab specimen, or acute IgG titer (⭓ 1:512), IgM titer (⭓ 1:10),
or a fourfold increase in either IgG or IgM between the convalescent
specimen by microimmunofluorescence assay (Focus Diagnostics, Inc.,
Cypress, CA); (3 ) for L. pneumophila, PCR (14) detection from an
oropharyngeal swab specimen, culture of a respiratory specimen, direct
fluorescent antibody assay of a respiratory specimen, antigen detection
in urine (Binax, Inc., Portland, ME), or acute IgG, IgM, or IgA titer
(⭓ 1:256), or a fourfold increase in IgG, IgM, or IgA between the acute
and convalescent specimen by serology by immunofluorescent antibody
assay (Zeus Scientific, Inc.). All positive values were determined by
manufacturers’ package insert instructions.
1087
The CAPO Database
This database contains information regarding the management of 2,878
patients with CAP from 39 hospitals in 11 countries (see Table E1 in
the online supplement). Local institutional review board review was
performed, and consent was waived because this was a retrospective,
observational study. Data were collected from patients admitted from
June 2001 through June 2006. In each participating center, primary
investigators randomly selected one or more patients from a list of
hospitalized patients with a diagnosis of CAP. Data were retrospectively
collected on a case report form during hospitalization or after, then
entered into a computer and transferred electronically to the CAPO
coordinating center at UL (www.caposite.com). A group of investigators at UL validated the quality of data by checking for discrepancies
and inconsistencies, and if present, they communicated with the respective primary investigator to correct or clarify the information before
cases were entered into the database.
Study Definitions
Pneumonia was defined as the presence of a new pulmonary infiltrate
plus either a new or increased cough, an abnormal temperature (⬍ 35.6⬚C
or ⬎ 37.8⬚C), or an abnormal serum leukocyte count (leukocytosis, left
shift, or leukopenia as defined by local laboratory values). Severity of
disease was evaluated using the pneumonia severity index score (15).
Atypical coverage was defined as any antibiotic regimen that contained
a macrolide, a tetracycline, or a fluoroquinolone. Clinical stability was
defined per the American Thoracic Society guidelines for CAP as follows: improved clinical signs (improved cough and shortness of breath),
lack of fever for at least 8 hours, improving leukocytosis (decreased at
least 10% from the previous day), and tolerating oral intake (1). Criteria
for clinical stability were evaluated daily during the first 7 days of
hospitalization. Length of stay (LOS) was defined in days and calculated
as the day of discharge minus the day of hospitalization. LOS was
censored at 30 days in an effort to capture only CAP-related LOS. Inhospital all-cause mortality was defined as the total mortality during
hospitalization. CAP-related mortality was defined clinically by each
principal investigator as death due primarily to the pulmonary infection
during hospitalization. Study regions were defined as North America
(region I), Europe (region II), and Latin America (region III), Asia
and Africa (region IV).
Study Design
All patients in the UL infectious diseases atypical pathogens reference
laboratory were divided into two groups based on the presence or
absence of a positive diagnostic test for atypical organisms. The incidence was calculated for each group. Clinical outcomes were derived
from a separate database, the CAPO database. All patients in the
CAPO database who met criteria for CAP were divided into two groups
based on the presence or absence of atypical coverage in the initial
TABLE 1. THE INCIDENCE OF ATYPICAL PATHOGENS, AND THE PROPORTION OF PATIENTS
TREATED WITH ANTIMICROBIAL THERAPY FOR ATYPICAL PATHOGENS
Globally
Region I
United States, Canada
Region II
Europe
Region III
Latin America
University of Louisville Infectious Diseases Atypical Pathogens Reference Laboratory Database
Total patients with CAP
4,337
3,302
501
331
No. patients with atypical pathogens
975
724
140
71
Incidence of atypical pathogens
22%
22%
28%
21%
Mycoplasma pneumoniae
12%
11%
15%
13%
Chlamydia pneumoniae
7%
8%
7%
6%
Legionella pneumophila
5%
4%
9%
3%
Total patients with CAP
No. patients with therapy for atypical
pathogens
Proportion of patients treated for
atypical pathogens
2,878
The CAPO Database
1,408
Region IV
Asia/Africa
203
40
20%
12%
5%
6%
782
655
33
2,220
1,292
582
343
3
77%
91%
74%
53%
10%
Definition of abbreviation: CAP ⫽ community-acquired pneumonia; CAPO ⫽ Community-Acquired Pneumonia Organization.
1088
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 175 2007
Figure 1. A comparison of patients in
each risk class with (solid bars) and without (shaded bars) coverage for atypical
pathogens (p ⫽ 0.02).
antimicrobial regimen given within the first 24 hours of admission. Time
to clinical stability, LOS, and mortality were compared for each group.
The incidence of atypical pathogens in a certain region was then compared with the frequency of antimicrobial coverage for atypical pathogens in the same region.
whether the groups differed. This type of analysis was also done to
compare the groups on hospital stays of fewer than 30 days and allcause mortality.
Statistical Analysis
The incidence of atypical pathogens and the incidence of initial
antimicrobial therapy with coverage for atypical pathogens for
each region of the world are depicted in Table 1. Although the
range of atypical pathogens is 20 to 28%, the range between
groups receiving therapy for atypical pathogens was 10 to 91%
(p ⬍ 0.01). There were 2,220 patients who received atypical
coverage compared with 658 patients who did not. Totals of
1,100 patients (62%) and 256 patients (58%) in each group were
male, respectively (p ⫽ 0.15). The mean age was 65.2 and 66.4
years, respectively (p ⫽ 0.25). A total of 201 patients (11%) were
admitted to an ICU from the group who received atypical coverage,
whereas 33 patients (8%) were admitted to an ICU from the group
who did not receive atypical coverage (p ⫽ 0.02). The proportion
of patients in each risk class with and without atypical coverage is
outlined in Figure 1. The most common antimicrobial regimens
prescribed in each group are listed in Table 2.
The time to clinical stability for patients with and without
atypical coverage is depicted in Figure 2. The mean time to
clinical stability was 3.2 days (SD, 1.7) for those who received
atypical coverage (n ⫽ 1,186), and 3.7 days (SD, 1.6) for those
who did not receive atypical coverage (n ⫽ 272). The mean
difference, 0.5 days (95% CI, 0.29–0.73), was significant (p ⬍
0.01). The multivariate analysis for time to clinical stability is
shown in Figure 3. Patients with an atypical regimen reached
clinical stability sooner, whereas patients with altered mental
status or admission to an ICU reached clinical stability later.
LOS for up to 30 days for hospitalized patients with and
without atypical coverage is depicted in Figure 4. The mean LOS
for patients who were hospitalized for 30 days or fewer was 8.8
Statistical analyses were performed by SPSS (version 14.0; SPSS, Inc.,
Chicago, IL). Differences in the proportion of patients who were treated
with an empiric regimen that covered atypical pathogens were calculated using a ␹2 test. A ␹2 test was also used to compare patients who
received antimicrobials to cover atypical pathogens and those who did
not for all risk classes. Univariate comparisons of pneumonia severity
index and LOS for each group were made using the nonparametric
Mann-Whitney U test. Kaplan-Meier survival curves were generated
for time to clinical stability and were compared using the log-rank
test (16). Differences in mortality rates were assessed with the ␹2 test.
Multivariate analyses of time to clinical stability and LOS were performed with the Cox proportional hazards method. Logistic regression
was used to examine factors related to mortality. The results of the
statistical analysis of each outcome indicated the unique contribution
of each variable, after controlling for all other variables. The main
factor of interest for all multivariate models was the type of antimicrobial coverage (atypical or nonatypical). Other dichotomous predictor
variables were entered into the model to adjust for the influence of
potential confounding effects. Variables to evaluate severity of disease
included the following: belonging to a high-risk class (III, IV, or V)
(15); admission to an intensive care unit (ICU); the presence of a pleural
effusion or multilobar infiltrate on chest radiograph; and having altered
mental status, hypotension (systolic ⬍ 90 mm Hg or diastolic ⬍ 60 mm
Hg), or tachypnea (respiratory rate ⬎ 30 breaths/min). Variables to
evaluate processes of care included the following: antimicrobial administration within 8 hours of admission, having blood cultures taken within
24 hours of admission, having oxygen status assessed, and being evaluated for pneumococcal vaccine.
The difference in the proportions of patients who reached clinical
stability within 7 days was compared for the two groups and the 95%
confidence interval (CI) around the difference was used to conclude
RESULTS
TABLE 2. THE FOUR MOST COMMON ANTIMICROBIAL REGIMENS USED IN PATIENTS WITH
AND WITHOUT COVERAGE FOR ATYPICAL PATHOGENS
Antimicrobial Regimen
Regimen
Regimen
Regimen
Regimen
1
2
3
4
Atypical Coverage (n ⫽ 2,878)
No Atypical Coverage (n ⫽ 658)
ß-Lactam ⫹ macrolide, 1,130 (51%)
Quinolone, 681 (31%)
ß-Lactam ⫹ quinolone, 240 (11%)
Macrolide, 54 (2%)
ß-Lactam, 553 (84%)
ß-Lactam ⫹ clindamycin, 34 (5%)
ß-Lactam ⫹ TMP/SMX, 18 (3%)
ß-Lactam ⫹ gentamicin, 16 (2%)
Definition of abbreviation: TMP/SMX ⫽ trimethoprim/sulfamethoxazole.
Arnold, Summersgill, LaJoie, et al.: Worldwide View of Atypical Pneumonia
Figure 2. Time to clinical stability for patients with (solid line) and without (dashed line) coverage for atypical pathogens (p ⬍ 0.01). The p
value is an expression of the log-rank test comparing the two KaplanMeier curves.
days (SD, 7.2) for those who received atypical coverage (n ⫽
2,220) and 9.6 days (SD, 7.0) for those who did not receive
atypical coverage (n ⫽ 658). The mean difference, 0.72 days
(95% CI, 0.96–1.33), was significant (p ⬍ 0.02). If the LOS was
censored at 14 days, then a mean difference of 0.95 days was
more significant (6.1 vs. 7.1 d; 95% CI, 0.56–1.34; p ⬍ 0.01). The
multivariate analysis for LOS is shown in Figure 5. Patients with
an atypical regimen were discharged sooner, whereas patients
admitted to an ICU were discharged later.
1089
Total in-hospital mortality for patients with and without atypical coverage is depicted in Figure 6. Patients with an atypical
regimen died less often. The multivariate analysis for total inhospital mortality is shown in Figure 7. Patients died less often
when administered an atypical regimen or given antimicrobials
within 8 hours of admission. Patients died more often if they
had a higher risk class, altered mental status, or hypotension.
The multivariate analysis for CAP-related in-hospital mortality
is shown in Figure E1 and Tables E5, E9, and E13. A comparison
of the percentage difference of four outcomes for patients who
received treatment for atypical pathogens compared with those
who did not is represented in Figure 8. Patients prescribed an
atypical regimen reached clinical stability sooner and died less
often, but did not necessarily have a shorter LOS.
The outcomes for regions I, II, and III are depicted in Figures
E2–E13. No region-specific data were analyzed for region IV
(Asia and Africa) because this region had minimal representation in our study. An atypical regimen was not significant in any
region when a subanalysis was performed, although the p values
did approach 0.05 for mortality in regions I and III (Tables E4
and E12). An elevated risk class and change in mental status
were associated with worse outcomes in region I (Tables E2–E5)
and region II (Tables E6–E9). In region III, an elevated risk
class and admission to an ICU were associated with a longer
time to clinical stability and LOS, whereas a change in mental
status was associated with a longer time to clinical stability and
higher mortality (Tables E10–E13).
DISCUSSION
This study indicates that, although the incidence of atypical
pathogens is relatively similar in all regions of the world, there
are significant differences in the proportion of patients who
are treated with an empiric regimen that covers for atypical
Figure 3. The hazard ratio (HR)
for time to clinical stability adjusted for baseline imbalances
using Cox proportional hazards regression. Hypotension
includes systolic pressure ⬍ 90
mm Hg or diastolic pressure
⬍ 60 mm Hg. Tachypnea includes respiratory rate ⬎ 30
breaths/minute. Blood cultures
were obtained within 24 hours
of admission. CI ⫽ confidence
interval; ICU ⫽ intensive care
unit.
1090
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 175 2007
Figure 4. Length of stay for hospitalized patients with (solid line) and
without (dashed line) coverage for atypical pathogens (p ⬍ 0.01). The
p value is an expression of the log-rank test comparing the two KaplanMeier curves.
pathogens. This study also indicates that, in hospitalized patients
with CAP, initial empiric therapy with coverage for atypical
pathogens is associated with decreased time to clinical stability,
decreased duration of hospitalization, and decreased patient
mortality.
Using a comprehensive workup for atypical pathogens, we
were able to define a very similar incidence in all regions, with
approximately one of four patients with CAP with evidence of
atypical pathogens. Even though M. pneumoniae was the most
common atypical pathogen identified in all regions of the world,
the second most common was C. pneumoniae in North and South
America, and L. pneumophila in Europe. Overall, these results
are likely to be generalizable because the data on the incidence
of atypical pathogens are from multiple sites, from a large number of patients with CAP, and were generated using the same
laboratory over a 7-year time period.
Although all North American guidelines strongly recommend
covering typical and atypical pathogens in all hospitalized patients with CAP, 9% of hospitalized patients in North America
(region I) did not receive empiric therapy that covered atypical
pathogens. We documented a lack of coverage for atypical pathogens in 26% of patients in region II, 47% of patients in region
III, and 90% of patients in region IV. The significant difference
in the proportion of hospitalized patients that receive atypical
coverage may be a reflection of the emphasis on coverage for
atypical pathogens given by a particular country’s national CAP
guidelines.
Published studies on the effect of covering atypical pathogens
in total patient mortality offer contradictory results. A beneficial
effect of an atypical regimen on total mortality has been reported
in six studies with a combined total study population of 17,192
patients hospitalized with CAP (17–22). On the other hand, a
lack of beneficial effect on total mortality has been reported in
three studies with a combined total study population of 12,735
patients hospitalized with CAP (23–25). Our research of patients
hospitalized with CAP adds another study in which a favorable
effect in outcomes can be documented with a regimen that covers
atypical pathogens. This current controversy may be explained
in part due to the fact that total mortality in some patients with
CAP is due to factors unrelated to initial antimicrobial therapy.
Initial antimicrobial therapy with or without atypical coverage
will have no influence in a patient that died of factors not related
to the pulmonary infection. It has been estimated that mortality
may not be directly related to the pulmonary infection in up to
half of hospitalized patients with CAP (26, 27).
Figure 5. The hazard ratio (HR) for
length of stay adjusted for baseline
imbalances using Cox proportional
hazards regression. Hypotension includes systolic pressure ⬍ 90 mm Hg
or diastolic pressure ⬍ 60 mm Hg.
Tachypnea includes respiratory rate
⬎ 30 breaths/minute. Blood cultures
were obtained within 24 hours of
admission.
Arnold, Summersgill, LaJoie, et al.: Worldwide View of Atypical Pneumonia
Figure 6. Total and community-acquired pneumonia (CAP)–related inhospital mortality for patients with (solid bars, n ⫽ 2,220) and without
(shaded bars, n ⫽ 658) coverage for atypical pathogens.
One of the strengths of our study is that a beneficial effect
on mortality for patients treated with an atypical regimen was
found for all-cause mortality as well as for only CAP-related
mortality. In an attempt to control for other factors that may
confound the outcome of mortality, we evaluated all patients
for certain recognized factors that are associated with mortality
in hospitalized patients with CAP. The results of our multivariate
analysis were consistent with the published literature indicating
an increased risk for mortality in patients with elevated risk
class, admission to ICU, multilobar infiltrates, pleural effusion,
altered mental status, increased respiratory rate, and hypotension. All of these variables characterized severity of pneumonia
at the time of presentation and cannot be modified. We identified
three variables that were amenable to modification which were
associated with a decreased risk for mortality: antibiotic adminis-
1091
tration within 8 hours, performing an oxygenation assessment,
and empiric therapy with a regimen that covered atypical pathogens. Another strength of our study is the generalizability of the
CAPO database study population with an overall mortality rate
of greater than 10%. Pharmaceutical trials often exclude patients
who are likely to die shortly after admission; hence, our study
is more consistent with “real life” populations.
Because the outcomes of mortality and length of stay may
be influenced by factors not directly related to the pulmonary
infection, some investigators consider that the best outcome to
evaluate the response to antimicrobial therapy is time to clinical
stability (28). Our study is the first one to evaluate the effect of
covering for atypical pathogens in time to clinical stability. We
documented a significant decrease in time to clinical stability in
patients who were treated with atypical coverage. Using the
CAPO international database, we recently demonstrated a
strong correlation of patient risk class not only with mortality
and LOS but also with time to clinical stability (29). In the
current study, the beneficial effect of atypical coverage on time
to clinical stability was maintained after adjusting for risk class.
In our population of hospitalized patients with CAP, empiric
therapy with a regimen that covers atypical pathogens produced
a consistent benefit for the early study outcomes (time to clinical
stability and LOS), as well as for the late study outcomes (total
mortality and CAP-related mortality). The beneficial effect of
antibiotics with atypical coverage is more difficult to be recognized in ambulatory patients with CAP because, in this population, time to clinical stability is not measured and mortality is a
very rare outcome. This may explain why a recent meta-analysis
failed to find a beneficial effect of a regimen with coverage for
atypical pathogens in patients with mild CAP (30).
The primary limitation of this study is the retrospective nature, and subsequently the lack of control for unmeasured confounders. A prospective study would also allow for patients in
each risk class to be distributed equally between the two groups:
those who would receive coverage for atypical pathogens and
Figure 7. The odds ratio (OR) for total
mortality adjusted for baseline imbalances using logistic regression. Hypotension includes systolic pressure ⬍ 90
mm Hg or diastolic pressure ⬍ 60 mm
Hg. Tachypnea includes respiratory
rate ⬎ 30 breaths/minute. Blood cultures were obtained within 24 hours
of admission.
1092
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 175 2007
Figure 8. A comparison of the percent differences of four outcomes for patients who received treatment with and without coverage
for atypical pathogens.
those who would not. A major limitation of the etiologic findings
is that the samples came from patients enrolled in drug trials.
These trials used enrollment criteria that limit the generalizability of the etiologic data presented in this study. Because this was
not a randomized trial, an important limitation of the outcomes
analysis is the possibility of indication bias. Despite efforts to
control for confounders, treating physicians may have selected
atypical coverage for subgroups of patients who were destined
to have better outcomes. Our analysis was also limited by the
low number of patients with CAP enrolled from Asia and Africa.
The implication of finding a 22% global incidence of atypical
pneumonia may affect future research in CAP. For example,
additional studies addressing the use of procalcitonin levels to
guide duration of antimicrobial therapy will need to consider a
comprehensive evaluation for atypical pathogens because more
than 50% of patients with low procalcitonin levels were found
to have atypical pneumonia (31, 32). As compliance with CAP
guidelines has been associated with improved outcomes (33),
future studies in that area might consider defining compliance
based on guidelines that recommend covering atypical pathogens
for all hospitalized patients with CAP (1, 2).
In summary, this study has three primary conclusions. First,
using comprehensive diagnostic tests, atypical pathogens were
found to be present in a considerable and similar proportion of
patients with CAP from all regions of the world. Second, there
were significant variations in the number of hospitalized patients
with CAP who were treated with empiric therapy that cover
for atypical pathogens. Third, patients who were treated with
antimicrobial regimens with coverage for atypical pathogens had
a shorter time to clinical stability, a decreased length of hospitalization, and a decreased mortality. Our data strongly support the
concept that all hospitalized patients with CAP should receive
empiric therapy with a regimen that covers typical and atypical
pathogens.
Conflict of Interest Statement : F.W.A. has received $500 funding from Pfizer and
$750 from Ortho-McNeil for speaking within the last 3 years. J.T.S. does not have
a financial relationship with a commercial entity that has an interest in the subject
of this manuscript. A.S.L. does not have a financial relationship with a commercial
entity that has an interest in the subject of this manuscript. P.P. does not have a
financial relationship with a commercial entity that has an interest in the subject
of this manuscript. T.J.M. does not have a financial relationship with a commercial
entity that has an interest in the subject of this manuscript. P.R. does not have a
financial relationship with a commercial entity that has an interest in the subject
of this manuscript. F.B. has received consulting fees from Pfizer, Sanofi-Aventis
(S-A), Altana, Wyeth, Schering-Plough (S-P), and GlaxoSmithKline (GSK); lecture
fees from Bayer, S-A, Pfizer, Abbott, Altana, GSK; and grant support from Pfizer,
Altana, and Abbott. P.F. does not have a financial relationship with a commercial
entity that has an interest in the subject of this manuscript. T.M.F. in the last
3 years has received funding from the following companies for the following
relationships: for recent research funding from Ortho-MacNeil, Oscient, Pfizer,
S-A; for consultancies from Bayer, GSK, Merck, Ortho-MacNeil, Oscient, Pfizer,
S-A, S-P, Wyeth; for Speaker’s Bureaus from Abbott, GSK, Merck, Ortho McNeil,
Oscient, Pfizer, S-A, S-P, and Wyeth. The aggregate amount received for the
preceding 3 years exceeds $10,000 for each of the pharmaceutical manufacturers
listed. J.R. does not have a financial relationship with a commercial entity that
has an interest in the subject of this manuscript. R.M. has received $2,000 for
speaking in a conference sponsored by Pfizer in 2005. L.M. does not have a
financial relationship with a commercial entity that has an interest in the subject
of this manuscript. C.M.L. does not have a financial relationship with a commercial
entity that has an interest in the subject of this manuscript. J.A.R. in the preceding
3 years has received a grant from Bayer/S-P in the amounts of $355,000 for a
clinical trial from October 2004 through April 2006. He also received a grant from
Pfizer Pharmaceuticals in the amount of $250,000 for a clinical trial that is currently
in progress.
Acknowledgment : The authors acknowledge the assistance of Elizabeth Smigielski, Associate Professor with the Kornhauser Health Sciences Library at the University of Louisville, and Ginny Sciortino with the Division of Infectious Diseases at
the University of Louisville.
CAPO investigators and affiliations: Dr. Raul Nakamatsu, Veterans Affairs Medical
Center, Louisville, KY; Dr. John Myers and Dr. Guy Brock, University of Louisville,
KY; Dr. Jose Bordon, Providence Hospital, Washington, DC; Dr. Peter Gross,
Hackensack University Medical Center, Hackensack, NJ; Dr. Karl Weiss,
Maisonneuve-Rosemont Hospital, University of Montreal, Montreal, Canada; Dr.
Delfino Legnani, Ospedale L. Sacco, Milan, Italy; Dr. Roberto Cosentini, Policlinico,
Milan, Italy; Dr. Maria Bodi, Hospital Universitario Joan XXIII, Tarragona, Spain;
Dr. Antoni Torres, Instituto de Neumonologia y Cirugia Toracica, Barcelona, Spain;
Dr. Jose Porras, Hospital Sant Pau i Santa Tecla, Tarragona, Spain; Dr. Harmut
Lode, City Hosp. E.v.Behring/Lungenklinik Heckeshorn, Berlin, Germany; Dr. Jorge
Roig, Hospital Nostra Senyora de Meritxell, Escaldes, Andorra; Dr. Guillermo
Benchetrit, IDIM A. Lanari, Buenos Aires, Argentina; Dr. Gustavo Lopardo, Hospital
Profesor Bernardo Houssay, Buenos Aires, Argentina; Dr. Lautaro de Vedia, Hospital
Francisco J. Muniz, Buenos Aires, Argentina; Dr. Jorge Corral, Hospital Dr Oscar
Alende, Mar del Plata, Argentina; Dr. Jorge Martinez, Instituto Medico Platense,
La Plata, Argentina; Dr. Jose Gonzalez, Hospital Enrique Tornu, Buenos Aires,
Argentina; Dr. Alejandro Videla, Hospital Universitario Austral, Buenos Aires,
Argentina; Dr. Carlos Victorio, Clinica Uruguay, Entre Rios, Argentina; Dr. Eduardo
Rodriguez, Hospital Español de La Plata, La Plata, Argentina; Dr. Maria Rodriguez,
Hospital Rodolfo Rossi, La Plata, Argentina; Dr. Gur Levy, Hospital Universitario
de Caracas, Caracas, Venezuela; Dr. Federico Arteta, Hospital Luis Gomez LopezAscardio, Barquisimeto, Venezuela; Dr. Alejandro Diaz Fuenzalida, Pontifica Universidad de Chile, Santiago, Chile; Dr. Maria Parada, Clinica las Condes, Santiago,
Chile; Dr. Juan Luna, Hospital Nacional Roosevelt, Guatemala.
References
1. Niederman MS, Mandell LA, Anzueto A, Bass JB, Broughton WA,
Campbell GD, Dean N, File T, Fine MJ, Gross PA, et al. Guidelines
for the management of adults with community-acquired pneumonia.
Am J Respir Crit Care Med 2001;163:1730–1754.
2. Mandell LA, Marrie TJ, Grossman RE, Chow AW, Hyland RH. Canadian guidelines for the management of community-acquired pneumonia: an evidence-based update by the Canadian Infectious Diseases
Society and the Canadian Thoracic Society. Clin Infect Dis 2000;31:
383–421.
3. Yanagihara K, Kohno S, Matsusima T. Japanese guidelines for the management of community-acquired pneumonia. Int J Antimicrob Agents
2001;18:S45–S48.
4. Vogel F, Worth H, Adam D, Elies W, Ewig S, Höffken G, Lode H,
Lorenz J, Scholz H, Stille W, et al. for the Paul Ehrlich Society for
Chemotherapy and the German Respiratory Association. Rational
treatment of bacterial respiratory tract infections. [in German] Chemotherapie Journal. 2000;1:3–23.
Arnold, Summersgill, LaJoie, et al.: Worldwide View of Atypical Pneumonia
5. Luna CM, Ramirez J, Lopez H, Mazzei JA, Abreu de Oliveira JC, Pereira
J, Jardim JR, Gonzales P, Lisboa C, Maldonado D, et al.; Grupo
de trabajo de la Asociacion Latinoamericana del Torax (ALAT).
Recomendaciones ALAT sobre la neumonia adquirida en la comunidad. [in Spanish] Arch Bronconeumol 2001;37:340–348.
6. Feldman C, Bateman ED, Klugman KP, Richards G, Maartens G, and
Ras G, for the Working Groups of the South African Pulmonology
Society and the Antibiotic Study Group of South Africa. Management
of community-acquired pneumonia in adults. S Afr Med J 1996;86:
1152–1163.
7. Memish Z, Shibl A, Ahmed Q. Guidelines for the management of
community-acquired pneumonia in Saudi Arabia: a model for the
Middle East region. Int J Antimicrob Agents 2002;20:S1–12.
8. French Infectious Diseases Society (SPILF). Guidelines of the French
Infectious Diseases Society. What should the initial antibiotherapy for
acute community-acquired pneumonia be? How should it be reassessed
in case of failure, given the evolution of responsible pathogens and the
resistance of pneumococci? Should combined treatment be used? Med
Mal Infect 2001;31:357–363.
9. IMPACT working group. Reducing bacterial resistance with IMPACT—
Interhospital Multi-disciplinary Programme on Antimicrobial Chemotherapy, 2nd ed. Hong Kong: Hong Kong University and Hong Kong
Hospital Authority; 2001. pp. 34–35.
10. Ruiz M, Ewig S, Marcos MA, Martinez JA, Arancibia F, Mensa J, Torres
A. Etiology of community-acquired pneumonia: impact of age, comorbidity, and severity. Am J Respir Crit Care Med 1999;160:397–405.
11. Oosterheert JJ, Bonten MJM, Hak E, Schneider MME, Hoepelman IM.
How good is the evidence for the recommended empirical antimicrobial treatment of patients hospitalized because of community-acquired
pneumonia? A systematic review. J Antimicrob Chemother 2003;52:
555–563.
12. Arnold F, Summersgill J, Blasi F, File T, Fernandez P, Porras J, Feldman
C, Ramirez J. Community-acquired pneumonia due to atypical pathogens: a worldwide comparison of the incidence and initial empiric
therapy. Results from the CAPO international cohort study [abstract].
Proc Am Thorac Soc 2005;2:A46.
13. Arnold F, Summergill J, Cohen MG, Ramirez J. Incidence and empiric
therapy for atypical pathogens of community-acquired pneumonia:
an international perspective [abstract]. Final Program and Abstracts
of the Annual Meeting of the Infectious Diseases Society of America
(IDSA) 2004; Boston, MA. A423.
14. Ramirez JA, Ahkee S, Tolentino A, Miller RD, Summersgill JT. Diagnosis of Legionella pneumophila, Mycoplasma pneumoniae, or Chlamydia pneumoniae lower respiratory infection using the polymerase
chain reaction on a single throat swab. Diagn Microbiol Infect Dis
1996;24:7–14.
15. Fine MJ, Auble TE, Yealy DM, Hanusa DM, Weissfeld LA, Singer DE,
Coley CM, Marrie TJ, Wishwa KN. A prediction rule to identify lowrisk patients with community-acquired pneumonia. N Engl J Med
1997;336:243–250.
16. Zar JH. Biostatistical analysis. In: Zar JH, editor. Two-sampled hypotheses, 4th ed. Upper Saddle River, NJ: Prentice Hall; 1999. pp. 122–160.
17. Trowbridge JF, Artymowicz RJ, Lee CE, Brown PB, Farber MS, Kernan
W, Beaulieu JE, Goldman MP, Yanak F, Lewis JS, et al. Antimicrobial
selection and length of hospital stay in patients with communityacquired pneumonia. J Clin Outcomes Manag 2002;9:613–619.
18. Stahl JE, Barza M, DesJardin J, Martin R, Eckman MH. Effect of macrolides as part of initial empiric therapy on length of stay in patients
1093
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
hospitalized with community-acquired pneumonia. Arch Intern Med
1999;159:2576–2580.
Dudas V, Hopefl A, Jacobs R, Guglielmo BJ. Antimicrobial selection
for hospitalized patients with presumed community-acquired pneumonia: a survey of non-teaching US community hospitals. Ann Pharmacother 2000;34:446–452.
Mufson MA, Stanek RJ. Bacteremic pneumococcal pneumonia in one
American city: a 20-year longitudinal study, 1978–1997. Am J Med
1999;107:34S–43S.
Martinez JA, Horcajada JP, Almela M, Marco F, Soriano A, Garcı́a E,
Marco MA, Torres A, Mensa J. Addition of a macrolide to a
␤-lactam–based empirical antibiotic regimen is associated with lower
in-hospital mortality for patients with bacteremic pneumococcal pneumonia. Clin Infect Dis 2003;36:389–395.
Gleason PP, Meehan TP, Fine JM, Galusha DH, Fine MJ. Associations
between initial antimicrobial therapy and medical outcomes for hospitalized elderly patients with pneumonia. Arch Intern Med 1999;159:
2562–2572.
Frei CR, Koeller JM, Burgess DS, Talbert RL, Johnsrud MT. Impact of
atypical coverage for patients with community-acquired pneumonia managed on the medical ward: results from the United States CommunityAcquired Pneumonia Project. Pharmacotherapy 2003;23:1167–1174.
Houck PM, MacLehose RF, Niederman MS, Lowry JK. Empiric antibiotic therapy and mortality among medicare pneumonia inpatients in
10 western states. Chest 2001;119:1420–1426.
Burgess DS, Lewis JS. Effect of macrolides as part of initial empiric
therapy on medical outcomes for hospitalized patients with communityacquired pneumonia. Clin Ther 2000;22:872–878.
Menéndez R, Ferrando D, Vallés JM, Martinez E, Perpiñá M. Initial
risk class and length of hospital stay in community-acquired pneumonia. Eur Respir J 2001;18:151–156.
Arnold F, Jain S, Christensen D, Ramirez J; CAPO Investigators.
Pneumonia-related versus pneumonia-unrelated length of stay in hospitalized patients with pneumonia after switch therapy: results from
the Community-Acquired Pneumonia Organization (CAPO) international cohort study [abstract]. Am J Respir Crit Care Med 2003;167:
A560.
Barlow GD, Lamping DL, Davey PG, Nathwani D. Evaluation of outcomes in community-acquired pneumonia: a guide for patients, physicians, and policy-makers. Lancet Infect Dis 2003;3:476–488.
Arnold FW, LaJoie AS, Marrie T, Rossi P, Blasi F, Luna CM, Fernandez
P, Porras J, Weiss K, Feldman C, et al. The pneumonia severity index
predicts time to clinical stability in patients with community-acquired
pneumonia. Int J Tuberc Lung Dis. 2006;10:739–743
Mills GD, Oehley MR, Arrol B. Effectiveness of B-lactam antibiotics
compared with antibiotics active against atypical pathogens in nonsevere community-acquired pneumonia: meta-analysis. BMJ 2005;330:
456–462.
Christ-Crain M, Stolz D, Bingisser R, Muller C, Miedinger D, Huber
PR, Zimmerli W, Harbarth S, Tamm M, Muller B. Procalcitonin guidance of antibiotic therapy in community-acquired pneumonia: a randomized trial. Am J Respir Crit Care Med 2006;174:84–93.
Wunderink RGA. CAP on antibiotic duration. Am J Respir Crit Care
Med 2006;174:3–5.
Menendez R, Torres A, Zalacain R, Aspa J, Martin-Villasclaras JJ,
Borderias L, Benitez-Moya JM, Ruiz-Manzano J, Rodriguez de Castro F,
Blanquer J, et al. Guidelines for the treatment of community-acquired
pneumonia: predictors of adherence and outcome. Am J Respir Crit
Care Med 2005;172:757–762.