Idsa Cap

SUPPLEMENT ARTICLE
Infectious Diseases Society of America/American

Thoracic Society Consensus Guidelines on the
Management of Community-Acquired Pneumonia
Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

in Adults
Lionel A. Mandell,1,a Richard G. Wunderink,2,a Antonio Anzueto,3,4 John G. Bartlett,7 G. Douglas Campbell,8
Nathan C. Dean,9,10 Scott F. Dowell,11 Thomas M. File, Jr.12,13 Daniel M. Musher,5,6 Michael S. Niederman,14,15
Antonio Torres,16 and Cynthia G. Whitney11
1
McMaster University Medical School, Hamilton, Ontario, Canada; 2Northwestern University Feinberg School of Medicine, Chicago, Illinois;
3
University of Texas Health Science Center and 4South Texas Veterans Health Care System, San Antonio, and 5Michael E. DeBakey Veterans
Affairs Medical Center and 6Baylor College of Medicine, Houston, Texas; 7Johns Hopkins University School of Medicine, Baltimore, Maryland;
8
Division of Pulmonary, Critical Care, and Sleep Medicine, University of Mississippi School of Medicine, Jackson; 9Division of Pulmonary and
Critical Care Medicine, LDS Hospital, and 10University of Utah, Salt Lake City, Utah; 11Centers for Disease Control and Prevention, Atlanta,
Georgia; 12Northeastern Ohio Universities College of Medicine, Rootstown, and 13Summa Health System, Akron, Ohio; 14State University of New
York at Stony Brook, Stony Brook, and 15Department of Medicine, Winthrop University Hospital, Mineola, New York; and 16Cap de Servei de
Pneumologia i Allèrgia Respiratòria, Institut Clı́nic del Tòrax, Hospital Clı́nic de Barcelona, Facultat de Medicina, Universitat de Barcelona, Institut
d’Investigacions Biomèdiques August Pi i Sunyer, CIBER CB06/06/0028, Barcelona, Spain.
EXECUTIVE SUMMARY priate starting point for consultation by specialists.

Substantial overlap exists among the patients whom
Improving the care of adult patients with community-
these guidelines address and those discussed in the re-
acquired pneumonia (CAP) has been the focus of many
cently published guidelines for health care–associated
different organizations, and several have developed
pneumonia (HCAP). Pneumonia in nonambulatory
guidelines for management of CAP. Two of the most
residents of nursing homes and other long-term care
widely referenced are those of the Infectious Diseases
facilities epidemiologically mirrors hospital-acquired
Society of America (IDSA) and the American Thoracic
pneumonia and should be treated according to the
Society (ATS). In response to confusion regarding dif-
HCAP guidelines. However, certain other patients
ferences between their respective guidelines, the IDSA
whose conditions are included in the designation of
and the ATS convened a joint committee to develop a HCAP are better served by management in accordance
unified CAP guideline document. with CAP guidelines with concern for specific
The guidelines are intended primarily for use by pathogens.
emergency medicine physicians, hospitalists, and pri-
mary care practitioners; however, the extensive litera-
Implementation of Guideline Recommendations
ture evaluation suggests that they are also an appro-
1. Locally adapted guidelines should be imple-

Reprints or correspondence: Dr. Lionel A. Mandell, Div. of Infectious Diseases, mented to improve process of care variables and
McMaster University/Henderson Hospital, 5th Fl., Wing 40, Rm. 503, 711 relevant clinical outcomes. (Strong recommen-
Concession St., Hamilton, Ontario L8V 1C3, Canada (lmandell@mcmaster.ca).
This official statement of the Infectious Diseases Society of America (IDSA) dation; level I evidence.)
and the American Thoracic Society (ATS) was approved by the IDSA Board of
Directors on 5 November 2006 and the ATS Board of Directors on 29 September
2006. It is important to realize that guidelines cannot always account for individual
a
Committee cochairs. variation among patients. They are not intended to supplant physician judgment
Clinical Infectious Diseases 2007; 44:S27–72 with respect to particular patients or special clinical situations. The IDSA considers
2007 by the Infectious Diseases Society of America. All rights reserved. adherence to these guidelines to be voluntary, with the ultimate determination
1058-4838/2007/4405S2-0001$15.00 regarding their application to be made by the physician in the light of each patient’s
DOI: 10.1086/511159 individual circumstances.
IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S27

Enthusiasm for developing these guidelines derives, in large take oral medication and the availability of outpatient support
part, from evidence that previous CAP guidelines have led to resources.
improvement in clinically relevant outcomes. Consistently ben- ICU admission decision.
eficial effects in clinically relevant parameters (listed in table 3) 7. Direct admission to an ICU is required for patients with
followed the introduction of a comprehensive protocol (in- septic shock requiring vasopressors or with acute respi-
cluding a combination of components from table 2) that in- ratory failure requiring intubation and mechanical ven-
creased compliance with published guidelines. The first rec- tilation. (Strong recommendation; level II evidence.)
ommendation, therefore, is that CAP management guidelines 8. Direct admission to an ICU or high-level monitoring unit
be locally adapted and implemented. is recommended for patients with 3 of the minor criteria
Documented benefits. for severe CAP listed in table 4. (Moderate recommen-

2. CAP guidelines should address a comprehensive set of dation; level II evidence.)
elements in the process of care rather than a single element In some studies, a significant percentage of patients with
in isolation. (Strong recommendation; level III evidence.) CAP are transferred to the ICU in the first 24–48 h after hos-
3. Development of local CAP guidelines should be directed pitalization. Mortality and morbidity among these patients ap-
toward improvement in specific and clinically relevant pears to be greater than those among patients admitted directly
outcomes. (Moderate recommendation; level III to the ICU. Conversely, ICU resources are often overstretched
evidence.) in many institutions, and the admission of patients with CAP
who would not directly benefit from ICU care is also problem-
atic. Unfortunately, none of the published criteria for severe
Site-of-Care Decisions
CAP adequately distinguishes these patients from those for
Almost all of the major decisions regarding management of
whom ICU admission is necessary. In the present set of guide-
CAP, including diagnostic and treatment issues, revolve
lines, a new set of criteria has been developed on the basis of
around the initial assessment of severity. Site-of-care decisions
data on individual risks, although the previous ATS criteria
(e.g., hospital vs. outpatient, intensive care unit [ICU] vs.
format is retained. In addition to the 2 major criteria (need
general ward) are important areas for improvement in CAP for mechanical ventilation and septic shock), an expanded set
management. of minor criteria (respiratory rate, 130 breaths/min; arterial
Hospital admission decision. oxygen pressure/fraction of inspired oxygen (PaO2/FiO2) ratio,
4. Severity-of-illness scores, such as the CURB-65 criteria !250; multilobar infiltrates; confusion; blood urea nitrogen
(confusion, uremia, respiratory rate, low blood pressure, level, 120 mg/dL; leukopenia resulting from infection; throm-
age 65 years or greater), or prognostic models, such as bocytopenia; hypothermia; or hypotension requiring aggressive
the Pneumonia Severity Index (PSI), can be used to iden- fluid resuscitation) is proposed (table 4). The presence of at
tify patients with CAP who may be candidates for out- least 3 of these criteria suggests the need for ICU care but will
patient treatment. (Strong recommendation; level I require prospective validation.
evidence.)
5. Objective criteria or scores should always be supple- Diagnostic Testing
mented with physician determination of subjective fac-
tors, including the ability to safely and reliably take oral 9. In addition to a constellation of suggestive clinical fea-
medication and the availability of outpatient support re- tures, a demonstrable infiltrate by chest radiograph or
sources. (Strong recommendation; level II evidence.) other imaging technique, with or without supporting mi-
6. For patients with CURB-65 scores ⭓2, more-intensive crobiological data, is required for the diagnosis of pneu-
treatment—that is, hospitalization or, where appropriate monia. (Moderate recommendation; level III evidence.)
and available, intensive in-home health care services—is Recommended diagnostic tests for etiology.
usually warranted. (Moderate recommendation; level III 10. Patients with CAP should be investigated for specific
evidence.) pathogens that would significantly alter standard (em-
Physicians often admit patients to the hospital who could pirical) management decisions, when the presence of
be well managed as outpatients and who would generally prefer such pathogens is suspected on the basis of clinical and
to be treated as outpatients. Objective scores, such as the CURB- epidemiologic clues. (Strong recommendation; level II
65 score or the PSI, can assist in identifying patients who may evidence.)
be appropriate for outpatient care, but the use of such scores Recommendations for diagnostic testing remain controver-
must be tempered by the physician’s determination of addi- sial. The overall low yield and infrequent positive impact on
tional critical factors, including the ability to safely and reliably clinical care argue against the routine use of common tests,
S28 • CID 2007:44 (Suppl 2) • Mandell et al.

such as blood and sputum cultures. Conversely, these cultures and Drug Administration before making its final recommen-
may have a major impact on the care of an individual patient dation regarding this drug. Recommendations are generally for
and are important for epidemiologic reasons, including the a class of antibiotics rather than for a specific drug, unless
antibiotic susceptibility patterns used to develop treatment outcome data clearly favor one drug. Because overall efficacy
guidelines. A list of clinical indications for more extensive di- remains good for many classes of agents, the more potent drugs
agnostic testing (table 5) was, therefore, developed, primarily are given preference because of their benefit in decreasing the
on the basis of 2 criteria: (1) when the result is likely to change risk of selection for antibiotic resistance.
individual antibiotic management and (2) when the test is likely
to have the highest yield. Outpatient treatment
11. Routine diagnostic tests to identify an etiologic diagnosis 15. Previously healthy and no risk factors for drug-resistant

are optional for outpatients with CAP. (Moderate rec- S. pneumoniae (DRSP) infection:
ommendation; level III evidence.) A. A macrolide (azithromycin, clarithromycin, or
12. Pretreatment blood samples for culture and an expec- erythromycin) (strong recommendation; level I
torated sputum sample for stain and culture (in patients evidence)
with a productive cough) should be obtained from hos- B. Doxycycline (weak recommendation; level III
pitalized patients with the clinical indications listed in evidence)
table 5 but are optional for patients without these con- 16. Presence of comorbidities, such as chronic heart, lung,
ditions. (Moderate recommendation; level I evidence.) liver, or renal disease; diabetes mellitus; alcoholism; ma-
13. Pretreatment Gram stain and culture of expectorated lignancies; asplenia; immunosuppressing conditions or
sputum should be performed only if a good-quality spec- use of immunosuppressing drugs; use of antimicrobials
imen can be obtained and quality performance measures within the previous 3 months (in which case an alter-
for collection, transport, and processing of samples can native from a different class should be selected); or other
be met. (Moderate recommendation; level II evidence.) risks for DRSP infection:
14. Patients with severe CAP, as defined above, should at A. A respiratory fluoroquinolone (moxifloxacin, gem-
least have blood samples drawn for culture, urinary an- ifloxacin, or levofloxacin [750 mg]) (strong rec-
tigen tests for Legionella pneumophila and Streptococcus ommendation; level I evidence)
pneumoniae performed, and expectorated sputum sam- B. A b-lactam plus a macrolide (strong recommen-
ples collected for culture. For intubated patients, an en- dation; level I evidence) (High-dose amoxicillin [e.g.,
dotracheal aspirate sample should be obtained. (Mod- 1 g 3 times daily] or amoxicillin-clavulanate [2 g 2
erate recommendation; level II evidence.) times daily] is preferred; alternatives include cef-
The most clear-cut indication for extensive diagnostic testing triaxone, cefpodoxime, and cefuroxime [500 mg 2
is in the critically ill CAP patient. Such patients should at least times daily]; doxycycline [level II evidence] is an
have blood drawn for culture and an endotracheal aspirate alternative to the macrolide.)
obtained if they are intubated; consideration should be given 17. In regions with a high rate (125%) of infection with
to more extensive testing, including urinary antigen tests for high-level (MIC, ⭓16 mg/mL) macrolide-resistant S.
L. pneumophila and S. pneumoniae and Gram stain and culture pneumoniae, consider the use of alternative agents listed
of expectorated sputum in nonintubated patients. For inpa- above in recommendation 16 for any patient, including
tients without the clinical indications listed in table 5, diagnostic those without comorbidities. (Moderate recommenda-
testing is optional (but should not be considered wrong). tion; level III evidence.)
Antibiotic Treatment Inpatient, non-ICU treatment

Empirical antimicrobial therapy. Empirical antibiotic rec- 18. A respiratory fluoroquinolone (strong recommendation;
ommendations (table 7) have not changed significantly from level I evidence)
those in previous guidelines. Increasing evidence has strength- 19. A b-lactam plus a macrolide (strong recommendation;
ened the recommendation for combination empirical therapy level I evidence) (Preferred b-lactam agents include ce-
for severe CAP. Only 1 recently released antibiotic has been fotaxime, ceftriaxone, and ampicillin; ertapenem for se-
added to the recommendations: ertapenem, as an acceptable lected patients; with doxycycline [level III evidence] as an
b-lactam alternative for hospitalized patients with risk factors alternative to the macrolide. A respiratory fluoroquino-
for infection with gram-negative pathogens other than Pseu- lone should be used for penicillin-allergic patients.)
domonas aeruginosa. At present, the committee is awaiting fur- Increasing resistance rates have suggested that empirical
ther evaluation of the safety of telithromycin by the US Food therapy with a macrolide alone can be used only for the treat-

ment of carefully selected hospitalized patients with nonsevere with oseltamivir or zanamivir is recommended for in-
disease and without risk factors for infection with drug-re- fluenza A. (Strong recommendation; level I evidence.)
sistant pathogens. However, such monotherapy cannot be 25. Use of oseltamivir and zanamivir is not recommended
routinely recommended. for patients with uncomplicated influenza with symp-
toms for 148 h (level I evidence), but these drugs may
Inpatient, ICU treatment be used to reduce viral shedding in hospitalized patients
20. A b-lactam (cefotaxime, ceftriaxone, or ampicillin-sul- or for influenza pneumonia. (Moderate recommenda-
bactam) plus either azithromycin (level II evidence) or tion; level III evidence.)
a fluoroquinolone (level I evidence) (strong recommen-
dation) (For penicillin-allergic patients, a respiratory flu- Pandemic influenza

oroquinolone and aztreonam are recommended.) 26. Patients with an illness compatible with influenza and
21. For Pseudomonas infection, use an antipneumococcal, with known exposure to poultry in areas with previous
antipseudomonal b-lactam (piperacillin-tazobactam, ce- H5N1 infection should be tested for H5N1 infection.
fepime, imipenem, or meropenem) plus either cipro- (Moderate recommendation; level III evidence.)
floxacin or levofloxacin (750-mg dose) 27. In patients with suspected H5N1 infection, droplet pre-
or cautions and careful routine infection control measures
the above b-lactam plus an aminoglycoside and should be used until an H5N1 infection is ruled out.
azithromycin (Moderate recommendation; level III evidence.)
or 28. Patients with suspected H5N1 infection should be treated
the above b-lactam plus an aminoglycoside and an an- with oseltamivir (level II evidence) and antibacterial agents
tipneumococcal fluoroquinolone (for penicillin-allergic targeting S. pneumoniae and S. aureus, the most common
patients, substitute aztreonam for the above b-lactam). causes of secondary bacterial pneumonia in patients with
(Moderate recommendation; level III evidence.) influenza (level III evidence). (Moderate recommendation.)
22. For community-acquired methicillin-resistant Staphy-
lococcus aureus infection, add vancomycin or linezolid. Time to first antibiotic dose.
(Moderate recommendation; level III evidence.) 29. For patients admitted through the emergency depart-
Infections with the overwhelming majority of CAP pathogens ment (ED), the first antibiotic dose should be admin-
will be adequately treated by use of the recommended empirical istered while still in the ED. (Moderate recommendation;
regimens. The emergence of methicillin-resistant S. aureus as level III evidence.)
a CAP pathogen and the small but significant incidence of CAP Rather than designating a specific window in which to initiate
due to P. aeruginosa are the exceptions. These pathogens occur treatment, the committee felt that hospitalized patients with
in specific epidemiologic patterns and/or with certain clinical CAP should receive the first antibiotic dose in the ED.
presentations, for which empirical antibiotic coverage may be
warranted. However, diagnostic tests are likely to be of high Switch from intravenous to oral therapy.
yield for these pathogens, allowing early discontinuation of 30. Patients should be switched from intravenous to oral
empirical treatment if results are negative. The risk factors are therapy when they are hemodynamically stable and im-
included in the table 5 recommendations for indications for proving clinically, are able to ingest medications, and
increased diagnostic testing. have a normally functioning gastrointestinal tract.
(Strong recommendation; level II evidence.)
Pathogens suspected on the basis of epidemiologic 31. Patients should be discharged as soon as they are clin-
considerations. ically stable, have no other active medical problems, and
Risk factors for other uncommon etiologies of CAP are listed have a safe environment for continued care. Inpatient
in table 8, and recommendations for treatment are included in observation while receiving oral therapy is not necessary.
table 9. (Moderate recommendation; level II evidence.)
Pathogen-directed therapy. Duration of antibiotic therapy.

23. Once the etiology of CAP has been identified on the 32. Patients with CAP should be treated for a minimum of
basis of reliable microbiological methods, antimicrobial 5 days (level I evidence), should be afebrile for 48–72 h,
therapy should be directed at that pathogen. (Moderate and should have no more than 1 CAP-associated sign of
recommendation; level III evidence.) clinical instability (table 10) before discontinuation of
24. Early treatment (within 48 h of the onset of symptoms) therapy (level II evidence). (Moderate recommendation.)

33. A longer duration of therapy may be needed if initial 40. The intranasally administered live attenuated vaccine is
therapy was not active against the identified pathogen an alternative vaccine formulation for some persons 5–
or if it was complicated by extrapulmonary infection, 49 years of age without chronic underlying diseases, in-
such as meningitis or endocarditis. (Weak recommen- cluding immunodeficiency, asthma, or chronic medical
dation; level III evidence.) conditions. (Strong recommendation; level I evidence.)
41. Health care workers in inpatient and outpatient settings
Other Treatment Considerations and long-term care facilities should receive annual in-
fluenza immunization. (Strong recommendation; level I
34. Patients with CAP who have persistent septic shock de- evidence.)
spite adequate fluid resuscitation should be considered 42. Pneumococcal polysaccharide vaccine is recommended

for treatment with drotrecogin alfa activated within 24 for persons ⭓65 years of age and for those with selected
h of admission. (Weak recommendation; level II high-risk concurrent diseases, according to current Ad-
evidence.) visory Committee on Immunization Practices guidelines.
35. Hypotensive, fluid-resuscitated patients with severe CAP (Strong recommendation; level II evidence.)
should be screened for occult adrenal insufficiency. 43. Vaccination status should be assessed at the time of hos-
(Moderate recommendation; level II evidence.) pital admission for all patients, especially those with
36. Patients with hypoxemia or respiratory distress should medical illnesses. (Moderate recommendation; level III
receive a cautious trial of noninvasive ventilation unless evidence.)
they require immediate intubation because of severe hy- 44. Vaccination may be performed either at hospital dis-
poxemia (PaO2/FiO2 ratio, !150) and bilateral alveolar charge or during outpatient treatment. (Moderate rec-
infiltrates. (Moderate recommendation; level I evidence.) ommendation; level III evidence.)
37. Low-tidal-volume ventilation (6 cm3/kg of ideal body 45. Influenza vaccine should be offered to persons at hospital
weight) should be used for patients undergoing venti- discharge or during outpatient treatment during the fall
lation who have diffuse bilateral pneumonia or acute and winter. (Strong recommendation; level III evidence.)
respiratory distress syndrome. (Strong recommenda- 46. Smoking cessation should be a goal for persons hospi-
tion; level I evidence.) talized with CAP who smoke. (Moderate recommen-
dation; level III evidence.)
47. Smokers who will not quit should also be vaccinated for
Management of Nonresponding Pneumonia
both pneumococcus and influenza. (Weak recommen-
Definitions and classification.
38. The use of a systematic classification of possible causes
48. Cases of pneumonia that are of public health concern
of failure to respond, based on time of onset and type
should be reported immediately to the state or local
of failure (table 11), is recommended. (Moderate rec-
health department. (Strong recommendation; level III
ommendation; level II evidence.)
evidence.)
As many as 15% of patients with CAP may not respond
49. Respiratory hygiene measures, including the use of hand
appropriately to initial antibiotic therapy. A systematic ap-
hygiene and masks or tissues for patients with cough,
proach to these patients (table 11) will help to determine the
should be used in outpatient settings and EDs as a means
cause. Because determination of the cause of failure is more
to reduce the spread of respiratory infections. (Strong
accurate if the original microbiological etiology is known, risk
recommendation; level III evidence.)
factors for nonresponse or deterioration (table 12) figure prom-
inently in the list of situations in which more aggressive and/
INTRODUCTION
or extensive initial diagnostic testing is warranted (table 5).
Improving the care of patients with community-acquired pneu-
Prevention (see table 13) monia (CAP) has been the focus of many different organiza-
tions. Such efforts at improvement in care are warranted, be-
39. All persons ⭓50 years of age, others at risk for influenza cause CAP, together with influenza, remains the seventh leading
complications, household contacts of high-risk persons, cause of death in the United States [1]. According to one es-
and health care workers should receive inactivated in- timate, 915,900 episodes of CAP occur in adults ⭓65 years of
fluenza vaccine as recommended by the Advisory Com- age each year in the United States [2]. Despite advances in
mittee on Immunization Practices, Centers for Disease antimicrobial therapy, rates of mortality due to pneumonia
Control and Prevention. (Strong recommendation; have not decreased significantly since penicillin became rou-
level I evidence.) tinely available [3].

Groups interested in approaches to the management of CAP cordance with CAP guidelines with concern for specific path-
include professional societies, such as the American Thoracic ogens. For example, long-term dialysis alone is a risk for MRSA
Society (ATS) and the Infectious Diseases Society of America infection but does not necessarily predispose patients to infec-
(IDSA); government agencies or their contract agents, such as tion with other HCAP pathogens, such as Pseudomonas aeru-
the Center for Medicare and Medicaid Services and the De- ginosa or Acinetobacter species. On the other hand, certain pa-
partment of Veterans Affairs; and voluntary accrediting agen- tients with chronic obstructive pulmonary disease (COPD) are
cies, such as the Joint Commission on Accreditation of Health- at greater risk for infection with Pseudomonas species but not
care Organizations. In addition, external review groups and MRSA. These issues will be discussed in specific sections below.
consumer groups have chosen CAP outcomes as major quality The committee started with the premise that mortality due
indicators. Such interest has resulted in numerous guidelines to CAP can be decreased. We, therefore, have placed the greatest

for the management of CAP [4]. Some of these guidelines emphasis on aspects of the guidelines that have been associated
represent truly different perspectives, including differences in with decreases in mortality. For this reason, the document fo-
health care systems, in the availability of diagnostic tools or cuses mainly on management and minimizes discussions of
therapeutic agents, or in either the etiology or the antibiotic such factors as pathophysiology, pathogenesis, mechanisms of
susceptibility of common causative microorganisms. The most antibiotic resistance, and virulence factors.
widely referenced guidelines in the United States have been The committee recognizes that the majority of patients with
those published by the ATS [5, 6] and the IDSA [7–9]. CAP are cared for by primary care, hospitalist, and emergency
Differences, both real and imagined, between the ATS and medicine physicians [11], and these guidelines are, therefore,
IDSA guidelines have led to confusion for individual physicians, directed primarily at them. The committee consisted of infec-
as well as for other groups who use these published guidelines tious diseases, pulmonary, and critical care physicians with in-
rather than promulgating their own. In response to this con- terest and expertise in pulmonary infections. The expertise of
cern, the IDSA and the ATS convened a joint committee to the committee and the extensive literature evaluation suggest
develop a unified CAP guideline document. This document that these guidelines are also an appropriate starting point for
represents a consensus of members of both societies, and both consultation by these types of physicians.
governing councils have approved the statement. Although much of the literature cited originates in Europe,
Purpose and scope. The purpose of this document is to these guidelines are oriented toward the United States and Can-
update clinicians with regard to important advances and con- ada. Although the guidelines are generally applicable to other
troversies in the management of patients with CAP. The com- parts of the world, local antibiotic resistance patterns, drug
mittee chose not to address CAP occurring in immunocom- availability, and variations in health care systems suggest that
promised patients, including solid organ, bone marrow, or stem modification of these guidelines is prudent for local use.
cell transplant recipients; patients receiving cancer chemother- Methodology. The process of guideline development
apy or long-term (130 days) high-dose corticosteroid treat- started with the selection of committee cochairs by the presi-
ment; and patients with congenital or acquired immunodefi- dents of the IDSA [12] and ATS [13], in consultation with
ciency or those infected with HIV who have CD4 cell counts other leaders in the respective societies. The committee cochairs
!350 cells/mm3, although many of these patients may be in- were charged with selection of the rest of the committee. The
fected with the same microorganisms. Pneumonia in children IDSA members were those involved in the development of
(⭐18 years of age) is also not addressed. previous IDSA CAP guidelines [9], whereas ATS members were
Substantial overlap exists among the patients these guidelines chosen in consultation with the leadership of the Mycobacteria
address and those discussed in the recently published guidelines Tuberculosis and Pulmonary Infection Assembly, with input
for health care–associated pneumonia (HCAP) [10]. Two issues from the chairs of the Clinical Pulmonary and Critical Care
are pertinent: (1) an increased risk of infection with drug- assemblies. Committee members were chosen to represent dif-
resistant isolates of usual CAP pathogens, such as Streptococcus fering expertise and viewpoints on the various topics. One ac-
pneumoniae, and (2) an increased risk of infection with less knowledged weakness of this document is the lack of repre-
common, usually hospital-associated pathogens, such as Pseu- sentation by primary care, hospitalist, and emergency medicine
domonas and Acinetobacter species and methicillin-resistant physicians.
Staphylococcus aureus (MRSA). Pneumonia in nonambulatory The cochairs generated a general outline of the topics to be
residents of nursing homes and other long-term care facilities covered that was then circulated to committee members for
epidemiologically mirrors hospital-acquired pneumonia and input. A conference phone call was used to review topics and
should be treated according to the HCAP guidelines. However, to discuss evidence grading and the general aims and expec-
certain other patients whose conditions are included under the tations of the document. The topics were divided, and com-
designation of HCAP are better served by management in ac- mittee members were assigned by the cochairs and charged

with presentation of their topic at an initial face-to-face meet- Table 1. Levels of evidence.
ing, as well as with development of a preliminary document
dealing with their topic. Controversial topics were assigned to Evidence level Definition
2 committee members, 1 from each society. Level I (high) Evidence from well-conducted, randomized
controlled trials.
An initial face-to-face meeting of a majority of committee
Level II (moderate) Evidence from well-designed, controlled
members involved presentations of the most controversial top- trials without randomization (including
ics, including admission decisions, diagnostic strategies, and cohort, patient series, and case-control
antibiotic therapy. Prolonged discussions followed each pre- studies). Level II studies also include any
large case series in which systematic
sentation, with consensus regarding the major issues achieved analysis of disease patterns and/or mi-

before moving to the next topic. With input from the rest of crobial etiology was conducted, as well
as reports of data on new therapies that
the committee, each presenter and committee member assigned
were not collected in a randomized
to the less controversial topics prepared an initial draft of their fashion.
section, including grading of the evidence. Iterative drafts of Level III (low) Evidence from case studies and expert
the statement were developed and distributed by e-mail for opinion. In some instances, therapy
recommendations come from antibiotic
critique, followed by multiple revisions by the primary authors. susceptibility data without clinical
A second face-to-face meeting was also held for discussion of observations.
the less controversial areas and further critique of the initial
drafts. Once general agreement on the separate topics was ob-
tained, the cochairs incorporated the separate documents into 12) of the members to consider it to be strong and the majority
a single statement, with substantial editing for style and con- of the others to grade it as moderate.
sistency. The document was then redistributed to committee The implication of a strong recommendation is that most
members to review and update with new information from the patients should receive that intervention. Significant variability
literature up to June 2006. Recommended changes were re- in the management of patients with CAP is well documented.
viewed by all committee members by e-mail and/or conference Some who use guidelines suggest that this variability itself is
phone call and were incorporated into the final document by undesirable. Industrial models suggesting that variability per se
the cochairs. is undesirable may not always be relevant to medicine [15].
This document was then submitted to the societies for ap- Such models do not account for substantial variability among
proval. Each society independently selected reviewers, and patients, nor do they account for variable end points, such as
changes recommended by the reviewers were discussed by the limitation of care in patients with end-stage underlying diseases
committee and incorporated into the final document. The who present with CAP. For this reason, the committee members
guideline was then submitted to the IDSA Governing Council feel strongly that 100% compliance with guidelines is not the
and the ATS Board of Directors for final approval. desired goal. However, the rationale for variation from a
Grading of guideline recommendations. Initially, the com- strongly recommended guideline should be apparent from the
medical record.
mittee decided to grade only the strength of the evidence, using
Conversely, moderate or weak recommendations suggest
a 3-tier scale (table 1) used in a recent guideline from both
that, even if a majority would follow the recommended man-
societies [10]. In response to reviewers’ comments and the
agement, many practitioners may not. Deviation from guide-
maturation of the field of guideline development [14], a sep-
lines may occur for a variety of reasons [16, 17]. One document
arate grading of the strength of the recommendations was
cannot cover all of the variable settings, unique hosts, or ep-
added to the final draft. More extensive and validated criteria,
idemiologic patterns that may dictate alternative management
such as GRADE [14], were impractical for use at this stage.
strategies, and physician judgment should always supersede
The 3-tier scale similar to that used in other IDSA guideline guidelines. This is borne out by the finding that deviation from
documents [12] and familiar to many of the committee mem- guidelines is greatest in the treatment of patients with CAP
bers was therefore chosen. admitted to the ICU [18]. In addition, few of the recommen-
The strength of each recommendation was graded as dations have level I evidence to support them, and most are,
“strong,” “moderate,” or “weak.” Each committee member in- therefore, legitimate topics for future research. Subsequent pub-
dependently graded each recommendation on the basis of not lication of studies documenting that care that deviates from
only the evidence but also expert interpretation and clinical guidelines results in better outcomes will stimulate revision of
applicability. The final grading of each recommendation was a the guidelines. The committee anticipates that this will occur,
composite of the individual committee members’ grades. For and, for this reason, both the ATS and IDSA leaderships have
the final document, a strong recommendation required ⭓6 (of committed to the revision of these guidelines on a regular basis.

We recognize that these guidelines may be used as a measure 19 hospitals randomized between pathway and “conventional”
of quality of care for hospitals and individual practitioners. management found that admission rates among low-risk pa-
Although these guidelines are evidence based, the committee tients at pathway hospitals decreased (from 49% to 31% of
strongly urges that deviations from them not necessarily be patients in Pneumonia Severity Index [PSI] classes I–III; P !
considered substandard care, unless they are accompanied by .01) without differences in patient satisfaction scores or rate of
evidence for worse outcomes in a studied population. readmission [20]. Calculating the PSI score and assigning the
risk class, providing oral clarithromycin, and home nursing
IMPLEMENTATION OF GUIDELINE follow-up significantly (P p .01 ) decreased the number of low-
RECOMMENDATIONS mortality-risk admissions [25]. However, patient satisfaction
among outpatients was lower after implementation of this

guideline, despite survey data that suggested most patients
1. Locally adapted guidelines should be implemented to im-
would prefer outpatient treatment [26]. Of patients discharged
prove process of care variables and relevant clinical out-
from the ED, 9% required hospitalization within 30 days, al-
comes. (Strong recommendation; level I evidence.)
though another study showed lower readmission rates with the
use of a protocol [23]. Admission decision support derived
Enthusiasm for developing this set of CAP guidelines derives,
from the 1993 ATS guideline [5] recommendations, combined
in large part, from evidence that previous CAP guidelines have
with outpatient antibiotic recommendations, reduced the CAP
led to improvement in clinically relevant outcomes [17, 19–
hospitalization rate from 13.6% to 6.4% [23], and admission
21]. Protocol design varies among studies, and the preferable
rates for other diagnoses were unchanged. Not surprisingly, the
randomized, parallel group design has been used in only a small
resultant overall cost of care decreased by half (P p .01).
minority. Confirmatory studies that use randomized, parallel
Protocols using guidelines to decrease the duration of hos-
groups with precisely defined treatments are still needed, but
pitalization have also been successful. Guideline implementa-
a consistent pattern of benefit is found in the other types of
tion in 31 Connecticut hospitals decreased the mean length of
level I studies.
hospital stay (LOS) from 7 to 5 days (P ! .001) [27]. An ED-
Documented benefits. Published protocols have varied in
based protocol decreased the mean LOS from 9.7 to 6.4 days
primary focus and comprehensiveness, and the corresponding
(P ! .0001), with the benefits of guideline implementation
benefits vary from one study to another. However, the most
maintained 3 years after the initial study [22]. A 7-site trial,
impressive aspect of this literature is the consistently beneficial
randomized by physician group, of guideline alone versus the
effect seen in some clinically relevant parameter after the in-
same guideline with a multifaceted implementation strategy
troduction of a protocol that increases compliance with pub-
found that addition of an implementation strategy was asso-
lished guidelines.
ciated with decreased duration of intravenous antibiotic therapy
A decrease in mortality with the introduction of guideline-
and LOS, although neither decrease was statistically significant
based protocols was found in several studies [19, 21]. A 5-year
[28]. Several other studies used guidelines to significantly
study of 28,700 patients with pneumonia who were admitted
shorten the LOS, by an average of 11.5 days [20, 21].
during implementation of a pneumonia guideline demon-
Markers of process of care can also change with the use of
strated that the crude 30-day mortality rate was 3.2% lower
a protocol. The time to first antibiotic dose has been effectively
with the guideline (adjusted OR, 0.69; 95% CI, 0.49–0.97) [19],
decreased with CAP protocols [22, 27, 29]. A randomized, par-
compared with that among patients treated concurrently by
allel group study introduced a pneumonia guideline in 20 of
nonaffiliated physicians. After implemention of a practice
36 small Oklahoma hospitals [29], with the identical protocol
guideline at one Spanish hospital [21], the survival rate at 30
implemented in the remaining hospitals in a second phase.
days was higher (OR, 2.14; 95% CI, 1.23–3.72) than at baseline
Serial measurement of key process measures showed significant
and in comparison with 4 other hospitals without overt pro-
improvement in time to first antibiotic dose and other variables,
tocols. Lower mortality was seen in other studies, although the
first in the initial 20 hospitals and later in the remaining 16
differences were not statistically significant [22, 23]. Studies
hospitals. Implementing a guideline in the ED halved the time
that documented lower mortality emphasized increasing the
to initial antibiotic dose [22].
number of patients receiving guideline-recommended antibi-
otics, confirming results of the multivariate analysis of a ret-
rospective review [24]. 2. CAP guidelines should address a comprehensive set of
When the focus of a guideline was hospitalization, the num- elements in the process of care rather than a single element
ber of less ill patients admitted to the hospital was consistently in isolation. (Strong recommendation; level III evidence.)
found to be lower. Using admission decision support, a pro-
spective study of 11700 emergency department (ED) visits in Common to all of the studies documented above, a com-

Table 2. Elements important for local community-acquired care, because the recommended median LOS was longer than
pneumonia guidelines. the existing LOS for CAP at the study hospitals. The difficulty
in implementing guidelines and changing physician behavior
All patients
has also been documented [28, 33].
Initiation of antibiotic therapy at site of diagnosis for hospitalized
patients Clinically relevant outcome parameters should be evaluated
Antibiotic selection to measure the effect of the local guideline. Outcome param-
Empirical eters that can be used to measure the effect of implementation
Specific of a CAP guideline within an organization are listed in table
Admission decision support 3. Just as it is important not to focus on one aspect of care,
Assessment of oxygenation studying more than one outcome is also important. Improve-

Intensive care unit admission support ments in one area may be offset by worsening in a related area;
Smoking cessation
for example, decreasing admission of low-acuity patients might
Influenza and pneumococcal vaccine administration
increase the number of return visits to the ED or hospital
Follow-up evaluation
readmissions [25].
Inpatients only
Diagnostic studies
Timing SITE-OF-CARE DECISIONS
Types of studies
Almost all of the major decisions regarding management of
Prophylaxis against thromboembolic disease
CAP, including diagnostic and treatment issues, revolve around
Early mobilization
Thoracentesis for patients with significant parapneumonic the initial assessment of severity. We have, therefore, organized
effusions the guidelines to address this issue first.
Discharge decision support Hospital admission decision. The initial management de-
Patient education cision after diagnosis is to determine the site of care—outpa-
tient, hospitalization in a medical ward, or admission to an
ICU. The decision to admit the patient is the most costly issue
prehensive protocol was developed and implemented, rather
in the management of CAP, because the cost of inpatient care
than one addressing a single aspect of CAP care. No study has
for pneumonia is up to 25 times greater than that of outpatient
documented that simply changing 1 metric, such as time to
care [34] and consumes the majority of the estimated $8.4–
first antibiotic dose, is associated with a decrease in mortality.
$10 billion spent yearly on treatment.
Elements important in CAP guidelines are listed in table 2. Of
Other reasons for avoiding unnecessary admissions are that
these, rapid and appropriate empirical antibiotic therapy is con-
sistently associated with improved outcome. We have also inpatients at low risk for death who are treated in the outpatient
cluded elements of good care for general medical inpatients, setting are able to resume normal activity sooner than those
such as early mobilization [30] and prophylaxis against throm- who are hospitalized, and 80% are reported to prefer outpatient
boembolic disease [31]. Although local guidelines need not therapy [26, 35]. Hospitalization also increases the risk of
include all elements, a logical constellation of elements should
be addressed.
Table 3. Clinically relevant outcome parameters in community-
acquired pneumonia.
3. Development of local CAP guidelines should be directed
Mortality
toward improvement in specific and clinically relevant out-
Rate of hospital admission
comes. (Moderate recommendation; level III evidence.) Rate of intensive care unit admission
Delayed transfer to the intensive care unit
In instituting CAP protocol guidelines, the outcomes most
Treatment failure
relevant to the individual center or medical system should be
Drug toxicity and adverse effects
addressed first. Unless a desire to change clinically relevant Antibiotic resistance in common pathogens
outcomes exists, adherence to guidelines will be low, and in- Length of stay
stitutional resources committed to implement the guideline are Thirty-day readmission rate
likely to be insufficient. Guidelines for the treatment of pneu- Unscheduled return to emergency department or primary
monia must use approaches that differ from current practice physician office
and must be successfully implemented before process of care Return to work/school/normal activities
and outcomes can change. For example, Rhew et al. [32] de- Patient satisfaction
Cost of care
signed a guideline to decrease LOS that was unlikely to change

thromboembolic events and superinfection by more-virulent least 2 of the following 3 conditions: tachypnea, diastolic hy-
or resistant hospital bacteria [36]. potension, and an elevated blood urea nitrogen (BUN) level.
These criteria appear to function well except among patients
with underlying renal insufficiency and among elderly patients
4. Severity-of-illness scores, such as the CURB-65 criteria [52, 53].
(confusion, uremia, respiratory rate, low blood pressure, The most recent modification of the BTS criteria includes 5
age 65 years or greater), or prognostic models, such as easily measurable factors [45]. Multivariate analysis of 1068
the PSI, can be used to identify patients with CAP who patients identified the following factors as indicators of in-
may be candidates for outpatient treatment. (Strong rec- creased mortality: confusion (based on a specific mental test
ommendation; level I evidence.) or disorientation to person, place, or time), BUN level 17

mmol/L (20 mg/dL), respiratory rate ⭓30 breaths/min, low
Significant variation in admission rates among hospitals and blood pressure (systolic, !90 mm Hg; or diastolic, ⭐60 mm
among individual physicians is well documented. Physicians Hg), and age ⭓65 years; this gave rise to the acronym CURB-
often overestimate severity and hospitalize a significant number 65. In the derivation and validation cohorts, the 30-day mor-
of patients at low risk for death [20, 37, 38]. Because of these tality among patients with 0, 1, or 2 factors was 0.7%, 2.1%,
issues, interest in objective site-of-care criteria has led to at- and 9.2%, respectively. Mortality was higher when 3, 4, or 5
tempts by a number of groups to develop such criteria [39– factors were present and was reported as 14.5%, 40%, and 57%,
48]. The relative merits and limitations of various proposed respectively. The authors suggested that patients with a CURB-
criteria have been carefully evaluated [49]. The 2 most inter- 65 score of 0–1 be treated as outpatients, that those with a
esting are the PSI [42] and the British Thoracic Society (BTS) score of 2 be admitted to the wards, and that patients with a
criteria [39, 45]. score of ⭓3 often required ICU care. A simplified version
The PSI is based on derivation and validation cohorts of (CRB-65), which does not require testing for BUN level, may
14,199 and 38,039 hospitalized patients with CAP, respectively, be appropriate for decision making in a primary care practi-
plus an additional 2287 combined inpatients and outpatients tioner’s office [54].
[42]. The PSI stratifies patients into 5 mortality risk classes, The use of objective admission criteria clearly can decrease
and its ability to predict mortality has been confirmed in mul- the number of patients hospitalized with CAP [20, 23, 25, 55].
tiple subsequent studies. On the basis of associated mortality Whether the PSI or the CURB-65 score is superior is unclear,
rates, it has been suggested that risk class I and II patients because no randomized trials of alternative admission criteria
should be treated as outpatients, risk class III patients should exist. When compared in the same population, the PSI classified
be treated in an observation unit or with a short hospitalization, a slightly larger percentage of patients with CAP in the low-
and risk class IV and V patients should be treated as inpatients risk categories, compared with the CURB or CURB-65 criteria,
[42]. while remaining associated with a similar low mortality rate
Yealy et al. [50] conducted a cluster-randomized trial of among patients categorized as low risk [56]. Several factors are
low-, moderate-, and high-intensity processes of guideline im- important in this comparison. The PSI includes 20 different
plementation in 32 EDs in the United States. Their guideline variables and, therefore, relies on the availability of scoring
used the PSI for admission decision support and included rec- sheets, limiting its practicality in a busy ED [55]. In contrast,
ommendations for antibiotic therapy, timing of first antibiotic the CURB-65 criteria are easily remembered. However, CURB-
dose, measurement of oxygen saturation, and blood cultures 65 has not been as extensively studied as the PSI, especially
for admitted patients. EDs with moderate- to high-intensity with prospective validation in other patient populations (e.g.,
guideline implementation demonstrated more outpatient treat- the indigent inner-city population), and has not been specifi-
ment of low-risk patients and higher compliance with antibiotic cally studied as a means of reducing hospital admission rates.
recommendations. No differences were found in mortality rate, In EDs with sufficient decision support resources (either human
rate of hospitalization, median time to return to work or usual or computerized), the benefit of greater experience with the
activities, or patient satisfaction. This study differs from those PSI score may favor its use for screening patients who may be
reporting a mortality rate difference [19, 21] in that many candidates for outpatient management [50, 57–59].
hospitalized patients with pneumonia were not included. In
addition, EDs with low-intensity guideline implementation
formed the comparison group, rather than EDs practicing non- 5. Objective criteria or scores should always be supple-
guideline, usual pneumonia care. mented with physician determination of subjective fac-
The BTS original criteria of 1987 have subsequently been tors, including the ability to safely and reliably take oral
modified [39, 51]. In the initial study, risk of death was in- medication and the availability of outpatient support re-
creased 21-fold if a patient, at the time of admission, had at sources. (Strong recommendation; level II evidence.)

Studies show that certain patients with low PSI or CURB- such as neuromuscular or sickle cell disease, may require hos-
65 scores [20, 60, 61] require hospital admission, even to the pitalization but not affect the PSI score.
ICU [49, 62, 63]. Both scores depend on certain assumptions. The necessary reliance on dichotomous predictor variables
One is that the main rationale for admission of a patient with (abnormal vs. normal) in most criteria and the heavy reliance
CAP is risk of death. This assumption is clearly not valid in on age as a surrogate in the PSI score may oversimplify their
all cases. Another is that the laboratory and vital signs used for use for admission decisions. For example, a previously healthy
scoring are stable over time rather than indicative of transient 25-year-old patient with severe hypotension and tachycardia
abnormalities. This is also not true in all cases. Therefore, dy- and no additional pertinent prognostic factors would be placed
namic assessment over several hours of observation may be in risk class II, whereas a 70-year-old man with a history of
more accurate than a score derived at a single point in time. localized prostate cancer diagnosed 10 months earlier and no

Although advantageous to making decisions regarding hospital other problems would be placed in risk class IV [42]. Finally,
admission, sole reliance on a score for the hospital admission patient satisfaction was lower among patients treated outside
decision is unsafe. the hospital in one study with a PSI-based intervention group
Reasons for the admission of low-mortality-risk patients fall [25], suggesting that the savings resulting from use of the PSI
into 4 categories: (1) complications of the pneumonia itself, may be overestimated and that physicians should consider ad-
(2) exacerbation of underlying diseases(s), (3) inability to re- ditional factors not measured by the PSI.
liably take oral medications or receive outpatient care, and/or
(4) multiple risk factors falling just above or below thresholds
6. For patients with CURB-65 scores ⭓2, more-intensive
for the score [62]. Use of the PSI score in clinical trials has treatment—that is, hospitalization or, where appropriate
demonstrated some of its limitations, which may be equally and available, intensive in-home health care services—is
applicable to other scoring techniques. A modification of the usually warranted. (Moderate recommendation; level III
original PSI score was needed when it was applied to the ad- evidence.)
mission decision. An arterial saturation of !90% or an arterial
oxygen pressure (PaO2) of !60 mm Hg as a complication of Although the PSI and CURB-65 criteria are valuable aids in
the pneumonia, was added as a sole indicator for admission avoiding inappropriate admissions of low-mortality-risk pa-
for patients in risk classes I–III as an added “margin of safety” tients, another important role of these criteria may be to help
in one trial [42]. In addition to patients who required hospital identify patients at high risk who would benefit from hospi-
admission because of hypoxemia, a subsequent study identified talization. The committee preferred the CURB-65 criteria be-
patients in low PSI risk classes (I–III) who needed hospital cause of ease of use and because they were designed to measure
admission because of shock, decompensated coexisting ill- illness severity more than the likelihood of mortality. Patients
nesses, pleural effusion, inability to maintain oral intake, social with a CURB-65 score ⭓2 are not only at increased risk of
problems (the patient was dependent or no caregiver was avail- death but also are likely to have clinically important physiologic
able), and lack of response to previous adequate empirical an- derangements requiring active intervention. These patients
tibiotic therapy [64]. Of 178 patients in low PSI risk classes should usually be considered for hospitalization or for aggres-
who were treated as inpatients, 106 (60%) presented with at sive in-home care, where available. In a cohort of ∼3000 pa-
least 1 of these factors. Other medical or psychosocial needs tients, the mortality with a CURB-65 score of 0 was only 1.2%,
whereas 3–4 points were associated with 31% mortality [45].
requiring hospital care include intractable vomiting, injection
Because the PSI score is not based as directly on severity of
drug abuse, severe psychiatric illness, homelessness, poor over-
illness as are the CURB-65 criteria, a threshold for patients who
all functional status [65], and cognitive dysfunction [61, 66].
would require hospital admission or intensive outpatient treat-
The PSI score is based on a history of diseases that increase
ment is harder to define. The higher the score, the greater the
risk of death, whereas the CURB-65 score does not directly
need for hospitalization. However, even a patient who meets
address underlying disease. However, pneumonia may exac-
criteria for risk class V on the basis of very old age and multiple
erbate an underlying disease, such as obstructive lung disease,
stable chronic illnesses may be successfully managed as an out-
congestive heart failure, or diabetes mellitus, which, by them-
patient [23].
selves, may require hospital admission [60, 67]. Atlas et al. [25]
were able to reduce hospital admissions among patients in PSI
ICU admission decision.
risk classes I–III from 58% in a retrospective control group to
43% in a PSI-based intervention group. Ten of 94 patients in 7. Direct admission to an ICU is required for patients with
the latter group (compared with 0 patients in the control pop- septic shock requiring vasopressors or with acute respi-
ulation) were subsequently admitted, several for reasons un- ratory failure requiring intubation and mechanical ven-
related to their pneumonia. Also, the presence of rare illnesses, tilation. (Strong recommendation; level II evidence.)

8. Direct admission to an ICU or high-level monitoring unit Table 4. Criteria for severe community-acquired pneumonia.
is recommended for patients with 3 of the minor criteria
for severe CAP listed in table 4. (Moderate recommen- Minor criteriaa
dation; level II evidence.) Respiratory rateb ⭓30 breaths/min
PaO2/FiO2 ratiob ⭐250
The second-level admission decision is whether to place the Multilobar infiltrates
patient in the ICU or a high-level monitoring unit rather than Confusion/disorientation
on a general medical floor. Approximately 10% of hospitalized Uremia (BUN level, ⭓20 mg/dL)
Leukopeniac (WBC count, !4000 cells/mm3)
patients with CAP require ICU admission [68–70], but the
Thrombocytopenia (platelet count, !100,000 cells/mm3)
indications vary strikingly among patients, physicians, hospi-
Hypothermia (core temperature, !36C)

tals, and different health care systems. Some of the variability
Hypotension requiring aggressive fluid resuscitation
among institutions results from the availability of high-level Major criteria
monitoring or intermediate care units appropriate for patients Invasive mechanical ventilation
at increased risk of complications. Because respiratory failure Septic shock with the need for vasopressors
is the major reason for delayed transfer to the ICU, simple
NOTE. BUN, blood urea nitrogen; PaO2/FiO2, arterial oxygen pressure/frac-
cardiac monitoring units would not meet the criteria for a high- tion of inspired oxygen; WBC, white blood cell.
a
level monitoring unit for patients with severe CAP. One of the Other criteria to consider include hypoglycemia (in nondiabetic patients),
acute alcoholism/alcoholic withdrawal, hyponatremia, unexplained metabolic
most important determinants of the need for ICU care is the acidosis or elevated lactate level, cirrhosis, and asplenia.
presence of chronic comorbid conditions [68–72]. However, b
A need for noninvasive ventilation can substitute for a respiratory rate 130
approximately one-third of patients with severe CAP were pre- breaths/min or a PaO2/FiO2 ratio !250.
c
As a result of infection alone.
viously healthy [73].
The rationale for specifically defining severe CAP is 4-fold:
• Appropriate placement of patients optimizes use of limited admission include the initial ATS definition of severe CAP [5]
ICU resources. and its subsequent modification [6, 82], the CURB criteria [39,
• Transfer to the ICU for delayed respiratory failure or delayed 45], and PSI severity class V (or IV and V) [42]. However,
onset of septic shock is associated with increased mortality none of these criteria has been prospectively validated for the
[74]. Although low-acuity ICU admissions do occur, the ICU admission decision. Recently, these criteria were retro-
major concern is initial admission to the general medical spectively evaluated in a cohort of patients with CAP admitted
unit, with subsequent transfer to the ICU. As many as 45% to the ICU [63]. All were found to be both overly sensitive and
of patients with CAP who ultimately require ICU admission nonspecific in comparison with the original clinical decision
were initially admitted to a non-ICU setting [75]. Many to admit to the ICU. Revisions of the criteria or alternative
delayed transfers to the ICU represent rapidly progressive criteria were, therefore, recommended.
pneumonia that is not obvious on admission. However, For the revised criteria, the structure of the modified ATS
some have subtle findings, including those included in the criteria for severe CAP was retained [6]. The 2 major criteria—
minor criteria in table 4, which might warrant direct ad- mechanical ventilation with endotracheal intubation and septic
mission to the ICU. shock requiring vasopressors—are absolute indications for ad-
• The distribution of microbial etiologies differs from that of mission to an ICU.
CAP in general [76–79], with significant implications for In contrast, the need for ICU admission is less straightfor-
diagnostic testing and empirical antibiotic choices. Avoid- ward for patients who do not meet the major criteria. On the
ance of inappropriate antibiotic therapy has also been as- basis of the published operating characteristics of the criteria,
sociated with lower mortality [80, 81]. no single set of minor criteria is adequate to define severe CAP.
• Patients with CAP appropriate for immunomodulatory Both the ATS minor criteria [75] and the CURB criteria [45]
treatment must be identified. The systemic inflammatory have validity when predicting which patients will be at increased
response/severe sepsis criteria typically used for generic sep- risk of death. Therefore, the ATS minor criteria and the CURB
sis trials may not be adequate when applied specifically to variables were included in the new proposed minor criteria
severe CAP [82]. For example, patients with unilateral lobar (table 4). Age, by itself, was not felt to be an appropriate factor
pneumonia may have hypoxemia severe enough to meet for the ICU admission decision, but the remainder of the
criteria for acute lung injury but not have a systemic CURB-65 criteria [45] were retained as minor criteria (with
response. the exception of hypotension requiring vasopressors as a major
Several criteria have been proposed to define severe CAP. criterion). Rather than the complex criteria for confusion in
Most case series have defined it simply as CAP that necessitates the original CURB studies, the definition of confusion should
ICU admission. Objective criteria to identify patients for ICU be new-onset disorientation to person, place, or time.

Three additional minor criteria were added. Leukopenia the presence of pneumonia in patients without obvious signs
(white blood cell count, !4000 cells/mm3) resulting from CAP of pneumonia and unsuspected hypoxemia in patients with
has consistently been associated with excess mortality, as well diagnosed pneumonia [42, 97, 98].
as with an increased risk of complications such as acute re- A chest radiograph is required for the routine evaluation of
spiratory distress syndrome (ARDS) [77, 79, 83–87]. In addi- patients who are likely to have pneumonia, to establish the
tion, leukopenia is seen not only in bacteremic pneumococcal diagnosis and to aid in differentiating CAP from other common
disease but also in gram-negative CAP [88, 89]. When leu- causes of cough and fever, such as acute bronchitis. Chest ra-
kopenia occurs in patients with a history of alcohol abuse, the diographs are sometimes useful for suggesting the etiologic
adverse manifestations of septic shock and ARDS may be de- agent, prognosis, alternative diagnoses, and associated condi-
layed or masked. Therefore, these patients were thought to tions. Rarely, the admission chest radiograph is clear, but the

benefit from ICU monitoring. The coagulation system is often patient’s toxic appearance suggests more than bronchitis. CT
activated in CAP, and development of thrombocytopenia scans may be more sensitive, but the clinical significance of
(platelet count, !100,000 cells/mm3) is also associated with a these findings when findings of radiography are negative is
worse prognosis [86, 90–92]. Nonexposure hypothermia (core unclear [99]. For patients who are hospitalized for suspected
temperature, !36C) also carries an ominous prognosis in CAP pneumonia but who have negative chest radiography findings,
[83, 93]. The committee felt that there was sufficient justifi- it may be reasonable to treat their condition presumptively with
cation for including these additional factors as minor criteria. antibiotics and repeat the imaging in 24–48 h.
Other factors associated with increased mortality due to CAP Microbiological studies may support the diagnosis of pneu-
were also considered, including acute alcohol ingestion and monia due to an infectious agent, but routine tests are fre-
delirium tremens [79, 85, 94], hypoglycemia and hyperglyce- quently falsely negative and are often nonspecific. A history of
mia, occult metabolic acidosis or elevated lactate levels [91], recent travel or endemic exposure, if routinely sought, may
and hyponatremia [95]. However, many of these criteria overlap identify specific potential etiologies that would otherwise be
with those selected. Future studies validating the proposed cri- unexpected as a cause of CAP (see table 8) [100].
teria should record these factors as well, to determine whether
addition or substitution improves the predictive value of our Recommended Diagnostic Tests for Etiology
proposed criteria.
With the addition of more minor criteria, the threshold for
10. Patients with CAP should be investigated for specific
ICU admission was felt to be the presence of at least 3 minor
pathogens that would significantly alter standard (em-
criteria, based on the mortality association with the CURB
pirical) management decisions, when the presence of
criteria. Selecting 2 criteria appears to be too nonspecific, as is
such pathogens is suspected on the basis of clinical and
demonstrated by the initial ATS criteria [5]. Whether each of
epidemiologic clues. (Strong recommendation; level II
the criteria is of equal weight is also not clear. Therefore, pro-
evidence.)
spective validation of this set of criteria is clearly needed.
The need for diagnostic testing to determine the etiology of

DIAGNOSTIC TESTING
CAP can be justified from several perspectives. The primary
reason for such testing is if results will change the antibiotic
9. In addition to a constellation of suggestive clinical fea- management for an individual patient. The spectrum of anti-
tures, a demonstrable infiltrate by chest radiograph or biotic therapy can be broadened, narrowed, or completely al-
other imaging technique, with or without supporting mi- tered on the basis of diagnostic testing. The alteration in therapy
crobiological data, is required for the diagnosis of pneu- that is potentially most beneficial to the individual is an es-
monia. (Moderate recommendation; level III evidence.) calation or switch of the usual empirical regimen because of
unusual pathogens (e.g., endemic fungi or Mycobacterium tu-
The diagnosis of CAP is based on the presence of select berculosis) or antibiotic resistance issues. Broad empirical cov-
clinical features (e.g., cough, fever, sputum production, and erage, such as that recommended in these guidelines, would
pleuritic chest pain) and is supported by imaging of the lung, not provide the optimal treatment for certain infections, such
usually by chest radiography. Physical examination to detect as psittacosis or tularemia. Increased mortality [80] and in-
rales or bronchial breath sounds is an important component creased risk of clinical failure [81, 101] are more common with
of the evaluation but is less sensitive and specific than chest inappropriate antibiotic therapy. Management of initial anti-
radiographs [96]. Both clinical features and physical exam find- biotic failure is greatly facilitated by an etiologic diagnosis at
ings may be lacking or altered in elderly patients. All patients admission. De-escalation or narrowing of antibiotic therapy on
should be screened by pulse oximetry, which may suggest both the basis of diagnostic testing is less likely to decrease an in-

Table 5. Clinical indications for more extensive diagnostic testing.
Blood Sputum Legionella Pneumococcal

Indication culture culture UAT UAT Other
Intensive care unit admission X X X X Xa
Failure of outpatient antibiotic therapy X X X
b
Cavitary infiltrates X X X
Leukopenia X X
Active alcohol abuse X X X X
Chronic severe liver disease X X

Severe obstructive/structural lung disease X
Asplenia (anatomic or functional) X X
Recent travel (within past 2 weeks) X Xc
d
Positive Legionella UAT result X NA
Positive pneumococcal UAT result X X NA
Pleural effusion X X X X Xe
NOTE. NA, not applicable; UAT, urinary antigen test.

a
Endotracheal aspirate if intubated, possibly bronchoscopy or nonbronchoscopic bronchoalveolar lavage.
b
Fungal and tuberculosis cultures.
c
See table 8 for details.
d
Special media for Legionella.
e
Thoracentesis and pleural fluid cultures.
dividual’s risk of death but may decrease cost, drug adverse is not streamlined when possible [104, 105] or when inappro-
effects, and antibiotic resistance pressure. priate escalation occurs [95]. In clinical practice, narrowing of
Some etiologic diagnoses have important epidemiologic im- antibiotic therapy is, unfortunately, unusual, but the committee
plications, such as documentation of severe acute respiratory strongly recommends this as best medical practice. The pos-
syndrome (SARS), influenza, legionnaires disease, or agents of sibility of polymicrobial CAP and the potential benefit of com-
bioterrorism. Diagnostic testing for these infections may affect bination therapy for bacteremic pneumococcal pneumonia
not only the individual but also many other people. Although have complicated the decision to narrow antibiotic therapy.
pneumonia etiologies that should be reported to public health Delays in starting antibiotic therapy that result from the need
officials vary by state, in general, most states’ health regulations to obtain specimens, complications of invasive diagnostic pro-
require reporting of legionnaires disease, SARS, psittacosis, cedures, and unneeded antibiotic changes and additional testing
avian influenza (H5N1), and possible agents of bioterrorism for false-positive tests are also important considerations.
(plague, tularemia, and anthrax). In addition, specific diag- The general recommendation of the committee is to strongly
nostic testing and reporting are important for pneumonia cases encourage diagnostic testing whenever the result is likely to
of any etiology thought to be part of a cluster or caused by change individual antibiotic management. For other patients
pathogens not endemic to the area. with CAP, the recommendations for diagnostic testing focus
There are also societal reasons for encouraging diagnostic on patients in whom the diagnostic yield is thought to be
testing. The antibiotic recommendations in the present guide- greatest. These 2 priorities often overlap. Recommendations for
lines are based on culture results and sensitivity patterns from patients in whom routine diagnostic testing is indicated for the
patients with positive etiologic diagnoses [102]. Without the above reasons are listed in table 5. Because of the emphasis on
accumulated information available from these culture results, clinical relevance, a variety of diagnostic tests that may be ac-
trends in antibiotic resistance are more difficult to track, and curate but the results of which are not available in a time
empirical antibiotic recommendations are less likely to be window to allow clinical decisions are neither recommended
accurate. nor discussed.
The main downside of extensive diagnostic testing of all
patients with CAP is cost, which is driven by the poor quality
11. Routine diagnostic tests to identify an etiologic diagnosis
of most sputum microbiological samples and the low yield of
are optional for outpatients with CAP. (Moderate rec-
positive culture results in many groups of patients with CAP.
ommendation; level III evidence.)
A clear need for improved diagnostic testing in CAP, most likely
using molecular methodology rather than culture, has been Retrospective studies of outpatient CAP management usually
recognized by the National Institutes of Health [103]. show that diagnostic tests to define an etiologic pathogen are
The cost-benefit ratio is even worse when antibiotic therapy infrequently performed, yet most patients do well with empir-

ical antibiotic treatment [42, 106]. Exceptions to this general may have been unique to the specific antibiotic choice
rule may apply to some pathogens important for epidemiologic (erythromycin).
reasons or management decisions. The availability of rapid The lack of benefit overall in this trial should not be inter-
point-of-care diagnostic tests, specific treatment and chemo- preted as a lack of benefit for an individual patient. Therefore,
prevention, and epidemiologic importance make influenza test- performing diagnostic tests is never incorrect or a breach of
ing the most logical. Influenza is often suspected on the basis the standard of care. However, information from cohort and
of typical symptoms during the proper season in the presence observational studies may be used to define patient groups in
of an epidemic. However, respiratory syncytial virus (RSV) can which the diagnostic yield is increased. Patient groups in which
cause a similar syndrome and often occurs in the same clinical routine diagnostic testing is indicated and the recommended
scenario [107]. Rapid diagnostic tests may be indicated when tests are listed in table 5.

the diagnosis is uncertain and when distinguishing influenza Blood cultures. Pretreatment blood cultures yielded posi-
A from influenza B is important for therapeutic decisions. tive results for a probable pathogen in 5%–14% in large series
Other infections that are important to verify with diagnostic of nonselected patients hospitalized with CAP [104, 105, 109–
studies because of epidemiologic implications or because they 111]. The yield of blood cultures is, therefore, relatively low
require unique therapeutic intervention are SARS and avian (although it is similar to yields in other serious infections), and,
(H5N1) influenza, disease caused by agents of bioterrorism, when management decisions are analyzed, the impact of pos-
Legionella infection, community-acquired MRSA (CA-MRSA) itive blood cultures is minor [104, 105]. The most common
infection, M. tuberculosis infection, or endemic fungal infection. blood culture isolate in all CAP studies is S. pneumoniae. Be-
Attempts to establish an etiologic diagnosis are also appropriate cause this bacterial organism is always considered to be the
in selected cases associated with outbreaks, specific risk factors, most likely pathogen, positive blood culture results have not
or atypical presentations. clearly led to better outcomes or improvements in antibiotic
selection [105, 112]. False-positive blood culture results are
12. Pretreatment blood samples for culture and an expec- associated with prolonged hospital stay, possibly related to
torated sputum sample for stain and culture (in patients changes in management based on preliminary results showing
with a productive cough) should be obtained from hos- gram-positive cocci, which eventually prove to be coagulase-
pitalized patients with the clinical indications listed in negative staphylococci [95, 109]. In addition, false-positive
table 5 but are optional for patients without these con- blood culture results have led to significantly more vancomycin
ditions. (Moderate recommendation; level I evidence.) use [95].
13. Pretreatment Gram stain and culture of expectorated For these reasons, blood cultures are optional for all hos-
sputum should be performed only if a good-quality spec- pitalized patients with CAP but should be performed selectively
imen can be obtained and quality performance measures (table 5). The yield for positive blood culture results is halved
for collection, transport, and processing of samples can by prior antibiotic therapy [95]. Therefore, when performed,
be met. (Moderate recommendation; level II evidence.) samples for blood culture should be obtained before antibiotic
14. Patients with severe CAP, as defined above, should at administration. However, when multiple risk factors for bac-
least have blood samples drawn for culture, urinary an- teremia are present, blood culture results after initiation of
tigen tests for Legionella pneumophila and S. pneumoniae antibiotic therapy are still positive in up to 15% of cases [95]
performed, and expectorated sputum samples collected and are, therefore, still warranted in these cases, despite the
for culture. For intubated patients, an endotracheal as- lower yield.
pirate sample should be obtained. (Moderate recom- The strongest indication for blood cultures is severe CAP.
mendation; level II evidence.) Patients with severe CAP are more likely to be infected with
pathogens other than S. pneumoniae, including S. aureus, P.
The only randomized controlled trial of diagnostic strategy aeruginosa, and other gram-negative bacilli [77–80, 95, 113,
in CAP has demonstrated no statistically significant differences 114]. Many of the factors predictive of positive blood culture
in mortality rate or LOS between patients receiving pathogen- results [95] overlap with risk factors for severe CAP (table 4).
directed therapy and patients receiving empirical therapy [108]. Therefore, blood cultures are recommended for all patients with
However, pathogen-directed therapy was associated with lower severe CAP because of the higher yield, the greater possibility
mortality among the small number of patients admitted to the of the presence of pathogens not covered by the usual empirical
ICU. The study was performed in a country with a low inci- antibiotic therapy, and the increased potential to affect anti-
dence of antibiotic resistance, which may limit its applicability biotic management.
to areas with higher levels of resistance. Adverse effects were Blood cultures are also indicated when patients have a host
significantly more common in the empirical therapy group but defect in the ability to clear bacteremia—for example, as a result

of asplenia or complement deficiencies. Patients with chronic Because the best specimens are collected and processed be-
liver disease also are more likely to have bacteremia with CAP fore antibiotics are given, the time to consider obtaining ex-
[95]. Leukopenia is also associated with a high incidence of pectorated sputum specimens from patients with factors listed
bacteremia [79, 95]. in table 5 is before initiation of antibiotic therapy. Once again,
Respiratory tract specimen Gram stain and culture. the best indication for more extensive respiratory tract cultures
The yield of sputum bacterial cultures is variable and strongly is severe CAP. Gram stain and culture of endotracheal aspirates
influenced by the quality of the entire process, including spec- from intubated patients with CAP produce different results
imen collection, transport, rapid processing, satisfactory use of than expectorated sputum from non-ICU patients [76, 120].
cytologic criteria, absence of prior antibiotic therapy, and skill Many of the pathogens in the broader microbiological spectrum
in interpretation. The yield of S. pneumoniae, for example, is of severe CAP are unaffected by a single dose of antibiotics,

only 40%–50% from sputum cultures from patients with bac- unlike S. pneumoniae. In addition, an endotracheal aspirate
teremic pneumococcal pneumonia in studies performed a few does not require patient cooperation, is clearly a lower respi-
decades ago [115, 116]. A more recent study of 100 cases of ratory tract sample, and is less likely to be contaminated by
bacteremic pneumococcal pneumonia found that sputum spec- oropharyngeal colonizers. Nosocomial tracheal colonization is
imens were not submitted in 31% of cases and were judged as not an issue if the sample is obtained soon after intubation.
inadequate in another 16% of cases [117]. When patients re- Therefore, culture and Gram stain of endotracheal aspirates are
ceiving antibiotics for 124 h were excluded, Gram stain showed recommended for patients intubated for severe CAP. In addi-
pneumococci in 63% of sputum specimens, and culture results tion to routine cultures, a specific request for culture of re-
were positive in 86%. For patients who had received no anti- spiratory secretions on buffered charcoal yeast extract agar to
biotics, the Gram stain was read as being consistent with pneu- isolate Legionella species may be useful in this subset of patients
mococci in 80% of cases, and sputum culture results were with severe CAP in areas where Legionella is endemic, as well
positive in 93%. as in patients with a recent travel history [130].
Although there are favorable reports of the utility of Gram The fact that a respiratory tract culture result is negative does
stain [118], a meta-analysis showed a low yield, considering not mean that it has no value. Failure to detect S. aureus or
the number of patients with adequate specimens and definitive gram-negative bacilli in good-quality specimens is strong evi-
results [119]. Recent data show that an adequate specimen with dence against the presence of these pathogens. Growth inhi-
a predominant morphotype on Gram stain was found in only bition by antibiotics is lower with these pathogens than with
14% of 1669 hospitalized patients with CAP [120]. Higher PSI S. pneumoniae, but specimens obtained after initiation of an-
scores did not predict higher yield. However, a positive Gram tibiotic therapy are harder to interpret, with the possibility of
stain was highly predictive of a subsequent positive culture colonization. Necrotizing or cavitary pneumonia is a risk for
result. CA-MRSA infection, and sputum samples should be obtained
The benefit of a sputum Gram stain is, therefore, 2-fold. in all cases. Negative Gram stain and culture results should be
First, it broadens initial empirical coverage for less common adequate to withhold or stop treatment for MRSA infection.
etiologies, such as infection with S. aureus or gram-negative Severe COPD and alcoholism are major risk factors for in-
organisms. This indication is probably the most important, fection with P. aeruginosa and other gram-negative pathogens
because it will lead to less inappropriate antibiotic therapy. [131]. Once again, Gram stain and culture of an adequate spu-
Second, it can validate the subsequent sputum culture results. tum specimen are usually adequate to exclude the need for
Forty percent or more of patients are unable to produce any empirical coverage of these pathogens.
sputum or to produce sputum in a timely manner [108, 120]. A sputum culture in patients with suspected legionnaires
The yield of cultures is substantially higher with endotracheal disease is important, because the identification of Legionella
aspirates, bronchoscopic sampling, or transthoracic needle as- species implies the possibility of an environmental source to
pirates [120–126], although specimens obtained after initiation which other susceptible individuals may be exposed. Localized
of antibiotic therapy are unreliable and must be interpreted community outbreaks of legionnaires disease might be recog-
carefully [120, 127, 128]. Interpretation is improved with quan- nized by clinicians or local health departments because ⭓2
titative cultures of respiratory secretions from any source (spu- patients might be admitted to the same hospital. However,
tum, tracheal aspirations, and bronchoscopic aspirations) or outbreaks of legionnaires disease associated with hotels or cruise
by interpretation based on semiquantitative culture results [122, ships [132–134] are rarely detected by individual clinicians,
123, 129]. Because of the significant influence on diagnostic because travelers typically disperse from the source of infection
yield and cost effectiveness, careful attention to the details of before developing symptoms. Therefore, a travel history should
specimen handling and processing are critical if sputum cul- be actively sought from patients with CAP, and Legionella test-
tures are obtained. ing should be performed for those who have traveled in the 2

weeks before the onset of symptoms. Urinary antigen tests may pneumonia when samples for culture cannot be obtained in a
be adequate to diagnose and treat an individual, but efforts to timely fashion or when antibiotic therapy has already been
obtain a sputum specimen for culture are still indicated to initiated. Serial specimens from patients with known bacter-
facilitate epidemiologic tracking. The availability of a culture emia were still positive for pneumococcal urinary antigen in
isolate of Legionella dramatically improves the likelihood that 83% of cases after 3 days of therapy [147]. Comparisons with
an environmental source of Legionella can be identified and Gram stain show that these 2 rapidly available tests often do
remediated [135–137]. The yield of sputum culture is increased not overlap, with only 28% concordance (25 of 88) among
to 43%–57% when associated with a positive urinary antigen patients when results of either test were positive [140]. Only
test result [138, 139]. ∼50% of Binax pneumococcal urinary antigen–positive patients
Attempts to obtain a sample for sputum culture from a can be diagnosed by conventional methods [140, 150]. Dis-

patient with a positive pneumococcal urinary antigen test result advantages include cost (approximately $30 per specimen), al-
may be indicated for similar reasons. Patients with a productive though this is offset by increased diagnosis-related group–based
cough and positive urinary antigen test results have positive reimbursement for coding for pneumococcal pneumonia, and
sputum culture results in as many as 40%–80% of cases [140– the lack of an organism for in vitro susceptibility tests. False-
143]. In these cases, not only can sensitivity testing confirm positive results have been seen in children with chronic respi-
the appropriate choice for the individual patient, but important ratory diseases who are colonized with S. pneumoniae [151]
data regarding local community antibiotic resistance rates can and in patients with an episode of CAP within the previous 3
also be acquired. months [152], but they do not appear to be a significant prob-
Other cultures. Patients with pleural effusions 15 cm in lem in colonized patients with COPD [140, 152].
height on a lateral upright chest radiograph [111] should un- For Legionella, several urinary antigen assays are available,
dergo thoracentesis to yield material for Gram stain and culture but all detect only L. pneumophila serogroup 1. Although this
for aerobic and anaerobic bacteria. The yield with pleural fluid particular serogroup accounts for 80%–95% of community-
cultures is low, but the impact on management decisions is acquired cases of legionnaires disease [138, 153] in many areas
substantial, in terms of both antibiotic choice and the need for of North America, other species and serogroups predominate
drainage. in specific locales [154, 155]. Prior studies of culture-proven
Nonbronchoscopic bronchoalveolar lavage (BAL) in the ED legionnaires disease indicate a sensitivity of 70%–90% and a
has been studied in a small, randomized trial of intubated specificity of nearly 99% for detection of L. pneumophila se-
patients with CAP [144]. A high percentage (87%) of non- rogroup 1. The urine is positive for antigen on day 1 of illness
bronchoscopic BAL culture results were positive, even in some and continues to be positive for weeks [138, 150].
patients who had already received their first dose of antibiotics. The major issue with urinary bacterial antigen detection is
Unfortunately, tracheal aspirates were obtained from only a whether the tests allow narrowing of empirical antibiotic ther-
third of patients in the control group, but they all were culture apy to a single specific agent. The recommended empirical
positive. Therefore, it is unclear that endotracheal aspirates are antibiotic regimens will cover both of these microorganisms.
inferior to nonbronchoscopic BAL. The use of bronchoscopic Results of a small observational study suggest that therapy with
BAL, protected specimen brushing, or transthoracic lung as- a macrolide alone is adequate for hospitalized patients with
piration has not been prospectively studied for initial manage- CAP who test positive for L. pneumophila urinary antigen [156].
ment of patients with CAP [123]. The best indications are for Further research is needed in this area.
immunocompromised patients with CAP or for patients with In contrast, rapid antigen detection tests for influenza, which
CAP in whom therapy failed [101, 145]. can also provide an etiologic diagnosis within 15–30 min, can
Antigen tests. Urinary antigen tests are commercially avail- lead to consideration of antiviral therapy. Test performance
able and have been cleared by the US Food and Drug Admin- varies according to the test used, sample type, duration of ill-
istration (FDA) for detection of S. pneumoniae and L. pneumo- ness, and patient age. Most show a sensitivity of 50%–70% in
phila serogroup 1 [138, 140, 146–149]. Urinary antigen testing adults and a specificity approaching 100% [157–159]. Advan-
appears to have a higher diagnostic yield in patients with more tages include the high specificity, the ability of some assays to
severe illness [139, 140]. distinguish between influenza A and B, the rapidity with which
For pneumococcal pneumonia, the principal advantages of the results can be obtained, the possibly reduced use of anti-
antigen tests are rapidity (∼15 min), simplicity, reasonable spec- bacterial agents, and the utility of establishing this diagnosis
ificity in adults, and the ability to detect pneumococcal pneu- for epidemiologic purposes, especially in hospitalized patients
monia after antibiotic therapy has been started. Studies in adults who may require infection control precautions. Disadvantages
show a sensitivity of 50%–80% and a specificity of 190% [146, include cost (approximately $30 per specimen), high rates of
149, 150]. This is an attractive test for detecting pneumococcal false-negative test results, false-positive assays with adenovirus

infections, and the fact that the sensitivity is not superior to Table 6. Most common etiologies of community-acquired
physician judgment among patients with typical symptoms dur- pneumonia.
ing an influenza epidemic [157, 158, 160].
Patient type Etiology
Direct fluorescent antibody tests are available for influenza
and RSV and require ∼2 h. For influenza virus, the sensitivity Outpatient Streptococcus pneumoniae
is better than with the point-of-care tests (85%–95%). They Mycoplasma pneumoniae
Haemophilus influenzae
will detect animal subtypes such as H5N1 and, thus, may be
Chlamydophila pneumoniae
preferred for hospitalized patients [161, 162]. For RSV, direct a
Respiratory viruses
fluorescent antibody tests are so insensitive (sensitivity, 20%–
Inpatient (non-ICU) S. pneumoniae
30%) in adults that they are rarely of value [163].

M. pneumoniae
Acute-phase serologic testing. The standard for diagnosis C. pneumoniae
of infection with most atypical pathogens, including Chlamy- H. influenzae
dophila pneumoniae, Mycoplasma pneumoniae, and Legionella Legionella species
species other than L. pneumophila, relies on acute- and con- Aspiration
valescent-phase serologic testing. Most studies use a microim- Respiratory virusesa
munofluorescence serologic test, but this test shows poor re- Inpatient (ICU) S. pneumoniae
producibility [164]. Management of patients on the basis of a Staphylococcus aureus
single acute-phase titer is unreliable [165], and initial antibiotic Legionella species
Gram-negative bacilli
therapy will be completed before the earliest time point to check
H. influenzae
a convalescent-phase specimen.
PCR. A new PCR test (BD ProbeTec ET Legionella pneumo- NOTE. Based on collective data from recent studies [171]. ICU, intensive
care unit.
phila; Becton Dickinson) that will detect all serotypes of L. a
Influenza A and B, adenovirus, respiratory syncytial virus, and
pneumophila in sputum is now cleared by the FDA, but exten- parainfluenza.
sive published clinical experience is lacking. Most PCR reagents
for other respiratory pathogens (except Mycobacterium species)
are “home grown,” with requirements for use based on com- correlates with the presenting clinical manifestations) is not
pliance with NCCLS criteria for analytical validity [166]. De- specific enough to reliably predict the etiology of CAP [172–
spite the increasing use of these tests for atypical pathogens 174]. Even if a microbial etiology is identified, debate continues
[167, 168], a 2001 review by the Centers for Disease Control with regard to pathogen-specific treatment, because recent
and Prevention (CDC) of diagnostic assays for detection of C. studies suggest coinfection by atypical pathogens (such as C.
pneumoniae indicated that, of the 18 PCR reagents, only 4 pneumoniae, Legionella species, and viruses) and more tradi-
satisfied the criteria for a validated test [166]. The diagnostic tional bacteria [120, 175]. However, the importance of treating
criteria defined in this review are particularly important for use multiple infecting organisms has not been firmly established.
in prospective studies of CAP, because most prior reports used The majority of antibiotics released in the past several de-
liberal criteria, which resulted in exaggerated rates. For SARS, cades have an FDA indication for CAP, making the choice of
several PCR assays have been developed, but these tests are antibiotics potentially overwhelming. Selection of antimicrobial
inadequate because of high rates of false-negative assays in early regimens for empirical therapy is based on prediction of the
stages of infection [169, 170]. most likely pathogen(s) and knowledge of local susceptibility
patterns. Recommendations are generally for a class of anti-
ANTIBIOTIC TREATMENT biotics rather than a specific drug, unless outcome data clearly
favor one drug. Because overall efficacy remains good for many
A major goal of therapy is eradication of the infecting organism,
classes of agents, the more potent drugs are given preference
with resultant resolution of clinical disease. As such, antimi-
crobials are a mainstay of treatment. Appropriate drug selection because of their benefit in decreasing the risk of selection for
is dependent on the causative pathogen and its antibiotic sus- antibiotic resistance. Other factors for consideration of specific
ceptibility. Acute pneumonia may be caused by a wide variety antimicrobials include pharmacokinetics/pharmacodynamics,
of pathogens (table 6). However, until more accurate and rapid compliance, safety, and cost.
diagnostic methods are available, the initial treatment for most
patients will remain empirical. Recommendations for therapy Likely Pathogens in CAP
(table 7) apply to most cases; however, physicians should con- Although CAP may be caused by a myriad of pathogens, a
sider specific risk factors for each patient (table 8). A syndromic limited number of agents are responsible for most cases. The
approach to therapy (under the assumption that an etiology emergence of newly recognized pathogens, such as the novel

Table 7. Recommended empirical antibiotics for community- and Moraxella catarrhalis, generally in patients who have un-
acquired pneumonia. derlying bronchopulmonary disease, and S. aureus, especially
during an influenza outbreak. Risks for infection with Enter-
Outpatient treatment
obacteriaceae species and P. aeruginosa as etiologies for CAP
1. Previously healthy and no use of antimicrobials within the
previous 3 months are chronic oral steroid administration or severe underlying
A macrolide (strong recommendation; level I evidence) bronchopulmonary disease, alcoholism, and frequent antibiotic
Doxycyline (weak recommendation; level III evidence) therapy [79, 131], whereas recent hospitalization would define
2. Presence of comorbidities such as chronic heart, lung, liver cases as HCAP. Less common causes of pneumonia include,
or renal disease; diabetes mellitus; alcoholism; malignan- but are by no means limited to, Streptococcus pyogenes, Neisseria
cies; asplenia; immunosuppressing conditions or use of
meningitidis, Pasteurella multocida, and H. influenzae type b.

immunosuppressing drugs; or use of antimicrobials within
the previous 3 months (in which case an alternative from a The “atypical” organisms, so called because they are not
different class should be selected) detectable on Gram stain or cultivatable on standard bacteri-
A respiratory fluoroquinolone (moxifloxacin, gemifloxacin, or ologic media, include M. pneumoniae, C. pneumoniae, Legion-
levofloxacin [750 mg]) (strong recommendation; level I
evidence) ella species, and respiratory viruses. With the exception of Le-
A b-lactam plus a macrolide (strong recommendation; level I gionella species, these microorganisms are common causes of
evidence) pneumonia, especially among outpatients. However, these path-
3. In regions with a high rate (125%) of infection with high-level ogens are not often identified in clinical practice because, with
(MIC ⭓16 mg/mL) macrolide-resistant Streptococcus pneu-
moniae, consider use of alternative agents listed above in
a few exceptions, such as L. pneumophila and influenza virus,
(2) for patients without comorbidities (moderate recommen- no specific, rapid, or standardized tests for their detection exist.
dation; level III evidence) Although influenza remains the predominant viral cause of
Inpatients, non-ICU treatment CAP in adults, other commonly recognized viruses include RSV
A respiratory fluoroquinolone (strong recommendation; level I
[107], adenovirus, and parainfluenza virus, as well as less com-
evidence)
A b-lactam plus a macrolide (strong recommendation; level I
mon viruses, including human metapneumovirus, herpes sim-
evidence) plex virus, varicella-zoster virus, SARS-associated coronavirus,
Inpatients, ICU treatment and measles virus. In a recent study of immunocompetent adult
A b-lactam (cefotaxime, ceftriaxone, or ampicillin-sulbactam) patients admitted to the hospital with CAP, 18% had evidence
plus either azithromycin (level II evidence) or a respiratory
of a viral etiology, and, in 9%, a respiratory virus was the only
fluoroquinolone (level I evidence) (strong recommendation)
(for penicillin-allergic patients, a respiratory fluoroquinolone pathogen identified [176]. Studies that include outpatients find
and aztreonam are recommended) viral pneumonia rates as high as 36% [167]. The frequency of
Special concerns other etiologic agents—for example, M. tuberculosis, Chlamy-
If Pseudomonas is a consideration dophila psittaci (psittacosis), Coxiella burnetii (Q fever), Fran-
An antipneumococcal, antipseudomonal b-lactam (piperacillin-
cisella tularensis (tularemia), Bordetella pertussis (whooping
tazobactam, cefepime, imipenem, or meropenem) plus
either ciprofloxacin or levofloxacin (750 mg) cough), and endemic fungi (Histoplasma capsulatum, Cocci-
or dioides immitis, Cryptococcus neoformans, and Blastomyces hom-
The above b-lactam plus an aminoglycoside and azithromycin inis)—is largely determined by the epidemiologic setting (table
or 8) but rarely exceeds 2%–3% total [113, 177]. The exception
The above b-lactam plus an aminoglycoside and an antipneu- may be endemic fungi in the appropriate geographic distri-
mococcal fluoroquinolone (for penicillin-allergic patients,
substitute aztreonam for above b-lactam)
bution [100].
(moderate recommendation; level III evidence) The need for specific anaerobic coverage for CAP is generally
If CA-MRSA is a consideration, add vancomycin or linezolid overestimated. Anaerobic bacteria cannot be detected by di-
(moderate recommendation; level III evidence) agnostic techniques in current use. Anaerobic coverage is clearly
NOTE. CA-MRSA, community-acquired methicillin-resistant Staphylococ- indicated only in the classic aspiration pleuropulmonary syn-
cus aureus; ICU, intensive care unit. drome in patients with a history of loss of consciousness as a
result of alcohol/drug overdose or after seizures in patients with
SARS-associated coronavirus [170], continually increases the concomitant gingival disease or esophogeal motility disorders.
challenge for appropriate management. Antibiotic trials have not demonstrated a need to specifically
Table 6 lists the most common causes of CAP, in decreasing treat these organisms in the majority of CAP cases. Small-
order of frequency of occurrence and stratified for severity of volume aspiration at the time of intubation should be ade-
illness as judged by site of care (ambulatory vs. hospitalized). quately handled by standard empirical severe CAP treatment
S. pneumoniae is the most frequently isolated pathogen. Other [178] and by the high oxygen tension provided by mechanical
bacterial causes include nontypeable Haemophilus influenzae ventilation.

Table 8. Epidemiologic conditions and/or risk factors related to specific pathogens in community-acquired
pneumonia.
Condition Commonly encountered pathogen(s)

Alcoholism Streptococcus pneumoniae, oral anaerobes, Klebsiella
pneumoniae, Acinetobacter species, Mycobacterium
tuberculosis
COPD and/or smoking Haemophilus influenzae, Pseudomonas aeruginosa,
Legionella species, S. pneumoniae, Moraxella carar-
rhalis, Chlamydophila pneumoniae
Aspiration Gram-negative enteric pathogens, oral anaerobes

Lung abscess CA-MRSA, oral anaerobes, endemic fungal pneumonia,
M. tuberculosis, atypical mycobacteria
Exposure to bat or bird droppings Histoplasma capsulatum
Exposure to birds Chlamydophila psittaci (if poultry: avian influenza)
Exposure to rabbits Francisella tularensis
Exposure to farm animals or parturient cats Coxiella burnetti (Q fever)
HIV infection (early) S. pneumoniae, H. influenzae, M. tuberculosis
HIV infection (late) The pathogens listed for early infection plus Pneumocys-
tis jirovecii, Cryptococcus, Histoplasma, Aspergillus,
atypical mycobacteria (especially Mycobacterium
kansasii), P. aeruginosa, H. influenzae
Hotel or cruise ship stay in previous 2 weeks Legionella species
Travel to or residence in southwestern United States Coccidioides species, Hantavirus
Travel to or residence in Southeast and East Asia Burkholderia pseudomallei, avian influenza, SARS
Influenza active in community Influenza, S. pneumoniae, Staphylococcus aureus,
H. influenzae
Cough 12 weeks with whoop or posttussive Bordetella pertussis
vomiting
Structural lung disease (e.g., bronchiectasis) Pseudomonas aeruginosa, Burkholderia cepacia, S. aureus
Injection drug use S. aureus, anaerobes, M. tuberculosis, S. pneumoniae
Endobronchial obstruction Anaerobes, S. pneumoniae, H. influenzae, S. aureus
In context of bioterrorism Bacillus anthracis (anthrax), Yersinia pestis (plague),
Francisella tularensis (tularemia)
NOTE. CA-MRSA, community-acquired methicillin-resistant Staphylococcus aureus; COPD, chronic obstructive pulmonary dis-
ease; SARS, severe acute respiratory syndrome.
Antibiotic Resistance Issues for pneumonia is uncertain, and few well-controlled studies
Resistance to commonly used antibiotics for CAP presents an- have examined the impact of in vitro resistance on clinical
other major consideration in choosing empirical therapy. Re- outcomes of CAP. Published studies are limited by small sample
sistance patterns clearly vary by geography. Local antibiotic sizes, biases inherent in observational design, and the relative
prescribing patterns are a likely explanation [179–181]. How- infrequency of isolates exhibiting high-level resistance [183–
ever, clonal spread of resistant strains is well documented. 185]. Current levels of b-lactam resistance do not generally
Therefore, antibiotic recommendations must be modified on result in CAP treatment failures when appropriate agents (i.e.,
the basis of local susceptibility patterns. The most reliable amoxicillin, ceftriaxone, or cefotaxime) and doses are used,
source is state/provincial or municipal health department re- even in the presence of bacteremia [112, 186]. The available
gional data, if available. Local hospital antibiograms are gen- data suggest that the clinically relevant level of penicillin resis-
erally the most accessible source of data but may suffer from tance is a MIC of at least 4 mg/L [3]. One report suggested
small numbers of isolates. that, if cefuroxime is used to treat pneumococcal bacteremia
Drug-resistant S. pneumoniae (DRSP). The emergence of when the organism is resistant in vitro, the outcome is worse
drug-resistant pneumococcal isolates is well documented. The than with other therapies [112]. Other discordant therapies,
incidence of resistance appears to have stabilized somewhat in including penicillin, did not have an impact on mortality. Data
the past few years. Resistance to penicillin and cephalosporins exist suggesting that resistance to macrolides [187–189] and
may even be decreasing, whereas macrolide resistance continues older fluoroquinolones (ciprofloxacin and levofloxacin) [180,
to increase [179, 182]. However, the clinical relevance of DRSP 190, 191] results in clinical failure. To date, no failures have

been reported for the newer fluoroquinolones (moxifloxacin monia (gingivitis and a risk for loss of consciousness, such as
and gemifloxacin). seizures or alcohol abuse, or esophogeal motility disorders).
Risk factors for infection with b-lactam–resistant S. pneu- Diagnosis is usually straightforward, with high yields from spu-
moniae include age !2 years or 165 years, b-lactam therapy tum and blood cultures in this characteristic clinical scenario.
within the previous 3 months, alcoholism, medical comorbid- CA-MRSA CAP remains rare in most communities but is ex-
ities, immunosuppressive illness or therapy, and exposure to a pected to be an emerging problem in CAP treatment.
child in a day care center [112, 192–194]. Although the relative
predictive value of these risk factors is unclear, recent treatment Empirical Antimicrobial Therapy
with antimicrobials is likely the most significant. Recent therapy Outpatient treatment. The following regimens are recom-
or repeated courses of therapy with b-lactams, macrolides, or mended for outpatient treatment on the basis of the listed

fluoroquinolones are risk factors for pneumococcal resistance clinical risks.
to the same class of antibiotic [181, 193, 195, 196]. One study
15. Previously healthy and no risk factors for DRSP infec-
found that use of either a b-lactam or macrolide within the
tion:
previous 6 months predicted an increased likelihood that, if
A. A macrolide (azithromycin, clarithromycin, or
pneumococcal bacteremia is present, the organism would be
erythromycin) (strong recommendation; level I
penicillin resistant [196]. Other studies have shown that re-
evidence)
peated use of fluoroquinolones predicts an increased risk of
B. Doxycycline (weak recommendation; level III
infection with fluoroquinolone-resistant pneumococci [195,
evidence)
197]. Whether this risk applies equally to all fluoroquinolones
16. Presence of comorbidities, such as chronic heart, lung,
or is more of a concern for less active antipneumococcal agents
liver, or renal disease; diabetes mellitus; alcoholism; ma-
(levofloxacin and ciprofloxacin) than for more active agents
lignancies; asplenia; immunosuppressing conditions or
(moxifloxacin and gemifloxacin) is uncertain [190, 197, 198].
use of immunosuppressing drugs; use of antimicrobials
Recommendations for the use of highly active agents in pa-
within the previous 3 months (in which case an alter-
tients at risk for infection with DRSP is, therefore, based only
native from a different class should be selected); or other
in part on efficacy considerations; it is also based on a desire
risks for DRSP infection:
to prevent more resistance from emerging by employing the
A. A respiratory fluoroquinolone (moxifloxacin, gem-
most potent regimen possible. Although increasing the doses ifloxacin, or levofloxacin [750 mg]) (strong rec-
of certain agents (penicillins, cephalosporins, levofloxacin) may ommendation; level I evidence)
lead to adequate outcomes in the majority of cases, switching B. A b-lactam plus a macrolide (strong recommen-
to more potent agents may lead to stabilization or even an dation; level I evidence) (High-dose amoxicillin [e.g.,
overall decrease in resistance rates [179, 180]. 1 g 3 times daily] or amoxicillin-clavulanate [2 g 2
CA-MRSA. Recently, an increasing incidence of pneumo- times daily] is preferred; alternatives include cef-
nia due to CA-MRSA has been observed [199, 200]. CA-MRSA triaxone, cefpodoxime, and cefuroxime [500 mg 2
appears in 2 patterns: the typical hospital-acquired strain [80] times daily]; doxycycline [level II evidence] is an
and, recently, strains that are epidemiologically, genotypically, alternative to the macrolide.)
and phenotypically distinct from hospital-acquired strains [201, 17. In regions with a high rate (125%) of infection with
202]. Many of the former may represent HCAP, because these high-level (MIC, ⭓16 mg/mL) macrolide-resistant S.
earlier studies did not differentiate this group from typical CAP. pneumoniae, consider the use of alternative agents listed
The latter are resistant to fewer antimicrobials than are hospital- above in recommendation 16 for any patient, including
acquired MRSA strains and often contain a novel type IV those without comorbidities. (Moderate recommenda-
SCCmec gene. In addition, most contain the gene for Panton- tion; level III evidence.)
Valentine leukocidin [200, 202], a toxin associated with clinical
features of necrotizing pneumonia, shock, and respiratory fail- The most common pathogens identified from recent studies
ure, as well as formation of abscesses and empyemas. The large of mild (ambulatory) CAP were S. pneumoniae, M. pneumoniae,
majority of cases published to date have been skin infections C. pneumoniae, and H. influenzae [177, 205]. Mycoplasma in-
in children. In a large study of CA-MRSA in 3 communities, fection was most common among patients !50 years of age
2% of CA-MRSA infections were pneumonia [203]. However, without significant comorbid conditions or abnormal vital
pneumonia in both adults [204] and children has been re- signs, whereas S. pneumoniae was the most common pathogen
ported, often associated with preceding influenza. This strain among older patients and among those with significant un-
should also be suspected in patients who present with cavitary derlying disease. Hemophilus infection was found in 5%—
infiltrates without risk factors for anaerobic aspiration pneu- mostly in patients with comorbidities. The importance of ther-

apy for Mycoplasma infection and Chlamydophila infection in icillin (amoxicillin [1 g 3 times daily] or amoxicillin-clavulanate
mild CAP has been the subject of debate, because many in- [2 g 2 times daily]) should target 193% of S. pneumoniae and
fections are self-limiting [206, 207]. Nevertheless, studies from is the preferred b-lactam. Ceftriaxone is an alternative to high-
the 1960s of children indicate that treatment of mild M. pneu- dose amoxicillin when parenteral therapy is feasible. Selected
moniae CAP reduces the morbidity of pneumonia and shortens oral cephalosporins (cefpodoxime and cefuroxime) can be used
the duration of symptoms [208]. The evidence to support spe- as alternatives [210], but these are less active in vitro than high-
cific treatment of these microorganisms in adults is lacking. dose amoxicillin or ceftriaxone. Agents in the same class as the
Macrolides have long been commonly prescribed for treat- patient had been receiving previously should not be used to
ment of outpatients with CAP in the United States, because of treat patients with recent antibiotic exposure.
their activity against S. pneumoniae and the atypical pathogens. Telithromycin is the first of the ketolide antibiotics, derived

This class includes the erythromycin-type agents (including dir- from the macrolide family, and is active against S. pneumoniae
ithromycin), clarithromycin, and the azalide azithromycin. Al- that is resistant to other antimicrobials commonly used for CAP
though the least expensive, erythromycin is not often used now, (including penicillin, macrolides, and fluoroquinolones). Sev-
because of gastrointestinal intolerance and lack of activity eral CAP trials suggest that telithromycin is equivalent to com-
against H. influenzae. Because of H. influenzae, azithromycin parators (including amoxicillin, clarithromycin, and trovaflox-
is preferred for outpatients with comorbidities such as COPD. acin) [211–214]. There have also been recent postmarketing
Numerous randomized clinical trials have documented the reports of life-threatening hepatotoxicity [215]. At present, the
efficacy of clarithromycin and azithromycin as monotherapy committee is awaiting further evaluation of the safety of this
for outpatient CAP, although several studies have demonstrated drug by the FDA before making its final recommendation.
that clinical failure can occur with a resistant isolate. When Inpatient, non-ICU treatment. The following regimens are
such patients were hospitalized and treated with a b-lactam and recommended for hospital ward treatment.
a macrolide, however, all survived and generally recovered with-
18. A respiratory fluoroquinolone (strong recommendation;
out significant complications [188, 189]. Most of these patients
level I evidence)
had risk factors for which therapy with a macrolide alone is
19. A b-lactam plus a macrolide (strong recommendation;
not recommended in the present guidelines. Thus, for patients
level I evidence) (Preferred b-lactam agents include ce-
with a significant risk of DRSP infection, monotherapy with a
fotaxime, ceftriaxone, and ampicillin; ertapenem for se-
macrolide is not recommended. Doxycycline is included as a
lected patients; with doxycycline [level III evidence] as an
cost-effective alternative on the basis of in vitro data indicating
alternative to the macrolide. A respiratory fluoroquino-
effectiveness equivalent to that of erythromycin for pneumo-
lone should be used for penicillin-allergic patients.)
coccal isolates.
The use of fluoroquinolones to treat ambulatory patients The recommendations of combination treatment with a b-
with CAP without comorbid conditions, risk factors for DRSP, lactam plus a macrolide or monotherapy with a fluoroquino-
or recent antimicrobial use is discouraged because of concern lone were based on retrospective studies demonstrating a sig-
that widespread use may lead to the development of fluoro- nificant reduction in mortality compared with that associated
quinolone resistance [185]. However, the fraction of total flu- with administration of a cephalosporin alone [216–219]. Mul-
oroquinolone use specifically for CAP is extremely small and tiple prospective randomized trials have demonstrated that ei-
unlikely to lead to increased resistance by itself. More con- ther regimen results in high cure rates. The major discrimi-
cerning is a recent study suggesting that many outpatients given nating factor between the 2 regimens is the patient’s prior
a fluoroquinolone may not have even required an antibiotic, antibiotic exposure (within the past 3 months).
that the dose and duration of treatment were often incorrect, Preferred b-lactams are those effective against S. pneumoniae
and that another agent often should have been used as first- and other common, nonatypical pathogens without being
line therapy. This usage pattern may promote the rapid de- overly broad spectrum. In January 2002, the Clinical Laboratory
velopment of resistance to fluoroquinolones [209]. Standards Institute (formerly the NCCLS) increased the MIC
Comorbidities or recent antimicrobial therapy increase the breakpoints for cefotaxime and ceftriaxone for nonmeningeal
likelihood of infection with DRSP and enteric gram-negative S. pneumoniae infections. These new breakpoints acknowledge
bacteria. For such patients, recommended empirical therapeutic that nonmeningeal infections caused by strains formerly con-
options include (1) a respiratory fluoroquinolone (moxiflox- sidered to be intermediately susceptible, or even resistant, can
acin, gemifloxacin, or levofloxacin [750 mg daily]) or (2) combe treated successfully with usual doses of these b-lactams [112,
bination therapy with a b-lactam effective against S. pneumon- 186, 220].
iae plus a macrolide (doxycycline as an alternative). On the Two randomized, double-blind studies showed ertapenem to
basis of present pharmacodynamic principles, high-dose amox- be equivalent to ceftriaxone [221, 222]. It also has excellent

activity against anaerobic organisms, DRSP, and most Enter- a fluoroquinolone alone resulted in a trend toward inferior
obacteriaceae species (including extended-spectrum b-lacta- outcome [229]. Because septic shock and mechanical ventila-
mase producers, but not P. aeruginosa). Ertapenem may be tion are the clearest reasons for ICU admission, the majority
useful in treating patients with risks for infection with these of ICU patients would still require combination therapy. ICU
pathogens and for patients who have recently received antibiotic patients are routinely excluded from other trials; therefore, rec-
therapy. However, clinical experience with this agent is limited. ommendations are extrapolated from nonsevere cases, in con-
Other “antipneumococcal, antipseudomonal” b-lactam agents junction with case series and retrospective analyses of cohorts
are appropriate when resistant pathogens, such as Pseudomonas, with severe CAP.
are likely to be present. Doxycycline can be used as an alter- For all patients admitted to the ICU, coverage for S. pneu-
native to a macrolide on the basis of scant data for treatment moniae and Legionella species should be ensured [78, 230] by

of Legionella infections [171, 223, 224]. using a potent antipneumococcal b-lactam and either a mac-
Two randomized, double-blind studies of adults hospitalized rolide or a fluoroquinolone. Therapy with a respiratory fluo-
for CAP have demonstrated that parenteral azithromycin alone roquinolone alone is not established for severe CAP [229], and,
was as effective, with improved tolerability, as intravenous ce- if the patient has concomitant pneumococcal meningitis, the
furoxime, with or without intravenous erythromycin [225, efficacy of fluoroquinolone monotherapy is uncertain. In ad-
226]. In another study, mortality and readmission rates were dition, 2 prospective observational studies [231, 232] and 3
similar, but the mean LOS was shorter among patients receiving retrospective analyses [233–235] have found that combination
azithromycin alone than among those receiving other guide- therapy for bacteremic pneumococcal pneumonia is associated
line-recommended therapy [227]. None of the 10 patients with with lower mortality than monotherapy. The mechanism of
erythromycin-resistant S. pneumoniae infections died or was this benefit is unclear but was principally found in the patients
transferred to the ICU, including 6 who received azithromycin with the most severe illness and has not been demonstrated in
alone. Another study showed that those receiving a macrolide nonbacteremic pneumococcal CAP studies. Therefore, com-
alone had the lowest 30-day mortality but were the least ill bination empirical therapy is recommended for at least 48 h
[219]. Such patients were younger and were more likely to be or until results of diagnostic tests are known.
in lower-risk groups. In critically ill patients with CAP, a large number of micro-
These studies suggest that therapy with azithromycin alone organisms other than S. pneumoniae and Legionella species
can be considered for carefully selected patients with CAP with must be considered. A review of 9 studies that included 890
nonsevere disease (patients admitted primarily for reasons other patients with CAP who were admitted to the ICU demonstrates
than CAP) and no risk factors for infection with DRSP or gram- that the most common pathogens in the ICU population were
negative pathogens. However, the emergence of high rates of (in descending order of frequency) S. pneumoniae, Legionella
macrolide resistance in many areas of the country suggests that species, H. influenzae, Enterobacteriaceae species, S. aureus, and
this therapy cannot be routinely recommended. Initial therapy Pseudomonas species [171]. The atypical pathogens responsible
should be given intravenously for most admitted patients, but for severe CAP may vary over time but can account collectively
some without risk factors for severe pneumonia could receive for ⭓20% of severe pneumonia episodes. The dominant atyp-
oral therapy, especially with highly bioavailable agents such as ical pathogen in severe CAP is Legionella [230], but some di-
fluoroquinolones. When an intravenous b-lactam is combined agnostic bias probably accounts for this finding [78].
with coverage for atypical pathogens, oral therapy with a mac- The recommended standard empirical regimen should rou-
rolide or doxycycline is appropriate for selected patients with- tinely cover the 3 most common pathogens that cause severe
out severe pneumonia risk factors [228]. CAP, all of the atypical pathogens, and most of the relevant
Inpatient, ICU treatment. The following regimen is the Enterobacteriaceae species. Treatment of MRSA or P. aeruginosa
minimal recommended treatment for patients admitted to the infection is the main reason to modify the standard empirical
ICU. regimen. The following are additions or modifications to the
basic empirical regimen recommended above if these pathogens
20. A b-lactam (cefotaxime, ceftriaxone, or ampicillin-sul- are suspected.
bactam) plus either azithromycin (level II evidence) or
a fluoroquinolone (level I evidence) (strong recommen- 21. For Pseudomonas infection, use an antipneumococcal,
dation) (For penicillin-allergic patients, a respiratory flu- antipseudomonal b-lactam (piperacillin-tazobactam, ce-
oroquinolone and aztreonam are recommended.) fepime, imipenem, or meropenem) plus either cipro-
floxacin or levofloxacin (750-mg dose)
A single randomized controlled trial of treatment for severe or
CAP is available. Patients with shock were excluded; however, the above b-lactam plus an aminoglycoside and
among the patients with mechanical ventilation, treatment with azithromycin

or tibiotic therapy [80] would suggest that empirical coverage
the above b-lactam plus an aminoglycoside and an should be considered when CA-MRSA is a concern. The most
antipneumococcal fluoroquinolone. (For penicillin-al- effective therapy has yet to be defined. The majority of CA-MRSA
lergic patients, substitute aztreonam for the above b- strains are more susceptible in vitro to non–b-lactam antimi-
lactam.) crobials, including trimethoprim-sulfamethoxazole (TMP-SMX)
(Moderate recommendation; level III evidence.) and fluoroquinolones, than are hospital-acquired strains. Pre-
vious experience with TMP-SMX in other types of severe infec-
Pseudomonal CAP requires combination treatment to pre-
tions (endocarditis and septic thrombophlebitis) suggests that
vent inappropriate initial therapy, just as Pseudomonas noso-
TMP-SMX is inferior to vancomycin [239]. Further experience
comial pneumonia does [131]. Once susceptibilities are known,
and study of the adequacy of TMP-SMX for CA-MRSA CAP is

treatment can be adjusted accordingly. Alternative regimens are
clearly needed. Vancomycin has never been specifically studied
provided for patients who may have recently received an oral
for CAP, and linezolid has been found to be better than ceftriax-
fluoroquinolone, in whom the aminoglycoside-containing reg-
one for bacteremic S. pneumoniae in a nonblinded study [240]
imen would be preferred. A consistent Gram stain of tracheal
and superior to vancomycin in retrospective analysis of studies
aspirate, sputum, or blood is the best indication for Pseudo-
involving nosocomial MRSA pneumonia [241]. Newer agents for
monas coverage. Other, easier-to-treat gram-negative micro-
MRSA have recently become available, and others are anticipated.
organisms may ultimately be proven to be the causative path-
Of the presently available agents, daptomycin should not be used
ogen, but empirical coverage of Pseudomonas species until
for CAP, and no data on pneumonia are available for tigecycline.
culture results are known is least likely to be associated with
A concern with CA-MRSA is necrotizing pneumonia asso-
inappropriate therapy. Other clinical risk factors for infection
ciated with production of Panton-Valentine leukocidin and
with Pseudomonas species include structural lung diseases, such
other toxins. Vancomycin clearly does not decrease toxin pro-
as bronchiectasis, or repeated exacerbations of severe COPD
duction, and the effect of TMP-SMX and fluoroquinolones on
leading to frequent steroid and/or antibiotic use, as well as prior
toxin production is unclear. Addition of clindamycin or use of
antibiotic therapy [131]. These patients do not necessarily re-
linezolid, both of which have been shown to affect toxin pro-
quire ICU admission for CAP [236], so Pseudomonas infection
duction in a laboratory setting [242], may warrant their con-
remains a concern for them even if they are only hospitalized
sideration for treatment of these necrotizing pneumonias [204].
on a general ward. The major risk factor for infection with
Unfortunately, the emergence of resistance during therapy with
other serious gram-negative pathogens, such as Klebsiella pneu-
clindamycin has been reported (especially in erythromycin-
moniae or Acinetobacter species, is chronic alcoholism.
resistant strains), and vancomycin would still be needed for
bacterial killing.
22. For CA-MRSA infection, add vancomycin or linezolid.
(Moderate recommendation; level III evidence.) Pathogens Suspected on the Basis of Epidemiologic
Considerations
The best indicator of S. aureus infection is the presence of
Clinicians should be aware of epidemiologic conditions and/
gram-positive cocci in clusters in a tracheal aspirate or in an
adequate sputum sample. The same findings on preliminary or risk factors that may suggest that alternative or specific ad-
results of blood cultures are not as reliable, because of the ditional antibiotics should be considered. These conditions and
significant risk of contamination [95]. Clinical risk factors for specific pathogens, with preferred treatment, are listed in tables
S. aureus CAP include end-stage renal disease, injection drug 8 and 9.
abuse, prior influenza, and prior antibiotic therapy (especially
with fluoroquinolones [237]). Pathogen-Directed Therapy
For methicillin-sensitive S. aureus, the empirical combina-
tion therapy recommended above, which includes a b-lactam 23. Once the etiology of CAP has been identified on the
and sometimes a respiratory fluoroquinolone, should be ade- basis of reliable microbiological methods, antimicrobial
quate until susceptibility results are available and specific ther- therapy should be directed at that pathogen. (Moderate
apy with a penicillinase-resistant semisynthetic penicillin or recommendation; level III evidence.)
first-generation cephalosporin can be initiated. Both also offer
additional coverage for DRSP. Neither linezolid [241] nor van- Treatment options may be simplified (table 9) if the etiologic
comycin [238] is an optimal drug for methicillin-sensitive S. agent is established or strongly suspected. Diagnostic proce-
aureus. dures that identify a specific etiology within 24–72 h can still
Although methicillin-resistant strains of S. aureus are still the be useful for guiding continued therapy. This information is
minority, the excess mortality associated with inappropriate an- often available at the time of consideration for a switch from

Table 9. Recommended antimicrobial therapy for specific pathogens.
Organism Preferred antimicrobial(s) Alternative antimicrobial(s)

Streptococcus pneumoniae
Penicillin nonresistant; MIC !2 mg/mL Penicillin G, amoxicillin Macrolide, cephalosporins (oral [cefpodox-
ime, cefprozil, cefuroxime, cefdinir, cefdi-
toren] or parenteral [cefuroxime, ceftriax-
one, cefotaxime]), clindamycin,
doxycyline, respiratory fluoroquinolonea
Penicillin resistant; MIC ⭓2 mg/mL Agents chosen on the basis of susceptibil- Vancomycin, linezolid, high-dose amoxicillin
ity, including cefotaxime, ceftriaxone, (3 g/day with penicillin MIC ⭐4 mg/mL)

fluoroquinolone
Haemophilus influenzae
Non–b-lactamase producing Amoxicillin Fluoroquinolone, doxycycline, azithromycin,
clarithromycinb
b-Lactamase producing Second- or third-generation cephalosporin, Fluoroquinolone, doxycycline, azithromycin,
b
amoxicillin-clavulanate clarithromycin
Mycoplasma pneumoniae/Chlamydophila Macrolide, a tetracycline Fluoroquinolone
pneumoniae
Legionella species Fluoroquinolone, azithromycin Doxycyline
Chlamydophila psittaci A tetracycline Macrolide
Coxiella burnetii A tetracycline Macrolide
Francisella tularensis Doxycycline Gentamicin, streptomycin
Yersinisa pestis Streptomycin, gentamicin Doxycyline, fluoroquinolone
Bacillus anthracis (inhalation) Ciprofloxacin, levofloxacin, doxycycline Other fluoroquinolones; b-lactam, if
(usually with second agent) susceptible; rifampin; clindamycin;
chloramphenicol
d
Enterobacteriaceae Third-generation cephalosporin, carbape- b-Lactam/b-lactamase inhibitor,
nemc (drug of choice if extended-spec- fluoroquinolone
trum b-lactamase producer)
e
Pseudomonas aeruginosa Antipseudomonal b-lactam plus (ciproflox- Aminoglycoside plus (ciprofloxacin or
f f
acin or levofloxacin or aminoglycoside) levofloxacin )
Burkholderia pseudomallei Carbapenem, ceftazadime Fluoroquinolone, TMP-SMX
Acinetobacter species Carbapenem Cephalosporin-aminoglycoside, ampicillin-
sulbactam, colistin
Staphylococcus aureus
Methicillin susceptible Antistaphylococcal penicilling Cefazolin, clindamycin
Methicillin resistant Vancomycin or linezolid TMP-SMX
Bordetella pertussis Macrolide TMP-SMX
Anaerobe (aspiration) b-Lactam/b-lactamase inhibitor,d Carbapenem
clindamycin
Influenza virus Oseltamivir or zanamivir
Mycobacterium tuberculosis Isoniazid plus rifampin plus ethambutol Refer to [243] for specific
plus pyrazinamide recommendations
Coccidioides species For uncomplicated infection in a normal Amphotericin B
host, no therapy generally recom-
mended; for therapy, itraconazole,
fluconazole
Histoplasmosis Itraconazole Amphotericin B
Blastomycosis Itraconazole Amphotericin B
NOTE. Choices should be modified on the basis of susceptibility test results and advice from local specialists. Refer to local references for appropriate
doses. ATS, American Thoracic Society; CDC, Centers for Disease Control and Prevention; IDSA, Infectious Diseases Society of America; TMP-SMX,
trimethoprim-sulfamethoxazole.
a
Levofloxacin, moxifloxacin, gemifloxacin (not a first-line choice for penicillin susceptible strains); ciprofloxacin is appropriate for Legionella and most
gram-negative bacilli (including H. influenza).
b
Azithromycin is more active in vitro than clarithromycin for H. influenza.
c
Imipenem-cilastatin, meropenem, ertapenem.
d
Piperacillin-tazobactam for gram-negative bacilli, ticarcillin-clavulanate, ampicillin-sulbactam or amoxicillin-clavulanate.
e
Ticarcillin, piperacillin, ceftazidime, cefepime, aztreonam, imipenem, meropenem.
f
750 mg daily.
g
Nafcillin, oxacillin flucloxacillin.
parenteral to oral therapy and may be used to direct specific pitalization, antiviral treatment seems reasonable from an in-
oral antimicrobial choices. If, for example, an appropriate cul- fection-control standpoint alone.
ture reveals penicillin-susceptible S. pneumoniae, a narrow- Because of its broad influenza spectrum, low risk of resistance
spectrum agent (such as penicillin or amoxicillin) may be used. emergence, and lack of bronchospasm risk, oseltamivir is an
This will, hopefully, reduce the selective pressure for resistance. appropriate choice for hospitalized patients. The neuraminidase
The major issue with pathogen-specific therapy is manage- inhibitors are effective against both influenza A and B viruses,
ment of bacteremic S. pneumoniae CAP. The implications of whereas the M2 inhibitors, amantadine, and rimantadine are
the observational finding that dual therapy was associated with active only against influenza A [251]. In addition, viruses re-
reduced mortality in bacteremic pneumococcal pneumonia cently circulating in the United States and Canada are often
[231–235] are uncertain. One explanation for the reduced mor- resistant to the M2 inhibitors on the basis of antiviral testing

tality may be the presence of undiagnosed coinfection with an [252, 253]. Therefore, neither amantadine nor rimantadine
atypical pathogen; although reported to occur in 18%–38% of should be used for treatment or chemoprophylaxis of influenza
CAP cases in some studies [73, 175], much lower rates of A in the United States until susceptibility to these antiviral
undiagnosed coinfection are found in general [171] and spe- medications has been reestablished among circulating influenza
cifically in severe cases [78]. An alternative explanation is the A viruses [249].
immunomodulatory effects of macrolides [244, 245]. It is im- Early treatment of influenza in ambulatory adults with in-
portant to note that these studies evaluated the effects of initial haled zanamivir or oral oseltamivir appears to reduce the like-
empirical therapy before the results of blood cultures were lihood of lower respiratory tract complications [254–256]. The
known and did not examine effects of pathogen-specific therapy use of influenza antiviral medications appears to reduce the
after the results of blood cultures were available. The benefit likelihood of respiratory tract complications, as reflected by
of combination therapy was also most pronounced in the more reduced usage rates of antibacterial agents in ambulatory pa-
severely ill patients [233, 234]. Therefore, discontinuation of tients with influenza. Although clearly important in outpatient
combination therapy after results of cultures are known is most pneumonia, this experience may also apply to patients hospi-
likely safe in non-ICU patients. talized primarily for influenza.
Parenteral acyclovir is indicated for treatment of varicella-
zoster virus infection [257] or herpes simplex virus pneumonia.
24. Early treatment (within 48 h of onset of symptoms) with No antiviral treatment of proven value is available for other
oseltamivir or zanamivir is recommended for influenza viral pneumonias—that is, parainfluenza virus, RSV, adenovi-
A. (Strong recommendation; level I evidence.) rus, metapneumovirus, the SARS agent, or hantavirus. For all
25. Use of oseltamivir and zanamivir is not recommended patients with viral pneumonias, a high clinical suspicion of
for patients with uncomplicated influenza with symp- bacterial superinfection should be maintained.
toms for 148 h (level I evidence), but these drugs may
be used to reduce viral shedding in hospitalized patients Pandemic influenza.
or for influenza pneumonia. (Moderate recommenda-
tion; level III evidence.) 26. Patients with an illness compatible with influenza and
with known exposure to poultry in areas with previous
Studies that demonstrate that treatment of influenza is ef- H5N1 infection should be tested for H5N1 infection.
fective only if instituted within 48 h of the onset of symptoms (Moderate recommendation; level III evidence.)
have been performed only in uncomplicated cases [246–249]. 27. In patients with suspected H5N1 infection, droplet pre-
The impact of such treatment on patients who are hospitalized cautions and careful routine infection control measures
with influenza pneumonia or a bacterial pneumonia compli- should be used until an H5N1 infection is ruled out.
cating influenza is unclear. In hospitalized adults with influenza, (Moderate recommendation; level III evidence.)
a minority of whom had radiographically documented pneu- 28. Patients with suspected H5N1 infection should be
monia, no obvious benefit was found in one retrospective study treated with oseltamivir (level II evidence) and antibac-
of amantadine treatment [250]. Treatment of antigen- or cul- terial agents targeting S. pneumoniae and S. aureus, the
ture-positive patients with influenza with antivirals in addition most common causes of secondary bacterial pneumonia
to antibiotics is warranted, even if the radiographic infiltrate is in patients with influenza (level III evidence). (Moderate
caused by a subsequent bacterial superinfection. Because of the recommendation.)
longer period of persistent positivity after infection, the ap-
propriate treatment for patients diagnosed with only 1 of the Recent human infections caused by avian influenza A
rapid diagnostic tests is unclear. Because such patients often (H5N1) in Vietnam, Thailand, Cambodia, China, Indonesia,
have recoverable virus (median duration of 4 days) after hos- Egypt, and Turkey raise the possibility of a pandemic in the

near future. The severity of H5N1 infection in humans distin- and zanamivir and resistant to the adamantanes (amantidine
guishes it from that caused by routine seasonal influenza. Re- and rimantidine) [262, 263]. The current recommendation is
spiratory failure requiring hospitalization and intensive care has for a 5-day course of treatment at the standard dosage of 75
been seen in the majority of the 1140 recognized cases, and mg 2 times daily. In addition, droplet precautions should be
mortality is ∼50% [258, 259]. If a pandemic occurs, deaths will used for patients with suspected H5N1 influenza, and they
result from primary influenza pneumonia with or without sec- should be placed in respiratory isolation until that etiology is
ondary bacterial pneumonia. This section highlights issues for ruled out. Health care personnel should wear N-95 (or higher)
consideration, recognizing that treatment recommendations respirators during medical procedures that have a high likeli-
will likely change as the pandemic progresses. More specific hood of generating infectious respiratory aerosols.
guidance can be found on the IDSA, ATS, CDC, and WHO Bacterial superinfections, particularly pneumonia, are im-

Web sites as the key features of the pandemic become clearer. portant complications of influenza pneumonia. The bacterial
Additional guidance is available at http://www.pandemicflu.gov. etiologies of CAP after influenza infection have included S.
The WHO has delineated 6 phases of an influenza pandemic, pneumoniae, S. aureus, H. influenzae, and group A streptococci.
defined by increasing levels of risk and public health response Legionella, Chlamydophila, and Mycoplasma species are not im-
[260]. During the current pandemic alert phase (phase 3: cases portant causes of secondary bacterial pneumonia after influ-
of novel influenza infection without sustained person-to-person enza. Appropriate agents would therefore include cefotaxime,
transmission), testing should be focused on confirming all sus- ceftriaxone, and respiratory fluoroquinolones. Treatment with
pected cases in areas where H5N1 infection has been docu- vancomycin, linezolid, or other agents directed against CA-
mented in poultry and on detecting the arrival of the pandemic MRSA should be limited to patients with confirmed infection
strain in unaffected countries. Early clinical features of H5N1 or a compatible clinical presentation (shock and necrotizing
infection include persistent fever, cough, and respiratory dif- pneumonia). Because shortages of antibacterials and antivirals
ficulty progressing over 3–5 days, as well as lymphopenia on are anticipated during a pandemic, the appropriate use of di-
admission to the hospital [258, 259, 261]. Exposure to sick and agnostic tests will be even more important to help target an-
dying poultry in an area with known or suspected H5N1 activity tibacterial therapy whenever possible, especially for patients
has been reported by most patients, although the recognition admitted to the hospital.
of poultry outbreaks has sometimes followed the recognition
of human cases [261]. Time to First Antibiotic Dose
Rapid bedside tests to detect influenza A have been used as
screening tools for avian influenza in some settings. Throat 29. For patients admitted through the ED, the first antibiotic
swabs tested by RT-PCR have been the most sensitive for con- dose should be administered while still in the ED. (Mod-
firming H5N1 infection to date, but nasopharyngeal swabs, erate recommendation; level III evidence.)
washes, and aspirates; BAL fluid; lung and other tissues; and
stool have yielded positive results by RT-PCR and viral culture Time to first antibiotic dose for CAP has recently received
with varying sensitivity. Convalescent-phase serum can be significant attention from a quality-of-care perspective. This
tested by microneutralization for antibodies to H5 antigen in emphasis is based on 2 retrospective studies of Medicare ben-
a small number of international reference laboratories. Speci- eficiaries that demonstrated statistically significantly lower mor-
mens from suspected cases of H5N1 infection should be sent tality among patients who received early antibiotic therapy [109,
to public health laboratories with appropriate biocontainment 264]. The initial study suggested a breakpoint of 8 h [264],
facilities; the case should be discussed with health department whereas the subsequent analysis found that 4 h was associated
officials to arrange the transfer of specimens and to initiate an with lower mortality [109]. Studies that document the time to
epidemiologic evaluation. During later phases of an ongoing first antibiotic dose do not consistently demonstrate this dif-
pandemic, testing may be necessary for many more patients, ference, although none had as large a patient population. Most
so that appropriate treatment and infection control decisions importantly, prospective trials of care by protocol have not
can be made, and to assist in defining the extent of the pan- demonstrated a survival benefit to increasing the percentage of
demic. Recommendations for such testing will evolve on the patients with CAP who receive antibiotics within the first 4–8
basis of the features of the pandemic, and guidance should be h [22, 65]. Early antibiotic administration does not appear to
sought from the CDC and WHO Web sites (http://www.cdc.gov shorten the time to clinical stability, either [265], although time
and http://www.who.int). of first dose does appear to correlate with LOS [266, 267]. A
Patients with confirmed or suspected H5N1 influenza should problem of internal consistency is also present, because, in both
be treated with oseltamivir. Most H5N1 isolates since 2004 have studies [109, 264], patients who received antibiotics in the first
been susceptible to the neuraminidase inhibitors oseltamivir 2 h after presentation actually did worse than those who re-

ceived antibiotics 2–4 h after presentation. For these and other Table 10. Criteria for clinical stability.
reasons, the committee did not feel that a specific time window
for delivery of the first antibiotic dose should be recommended. Temperature ⭐37.8C
Heart rate ⭐100 beats/min
However, the committee does feel that therapy should be ad-
Respiratory rate ⭐24 breaths/min
ministered as soon as possible after the diagnosis is considered
Systolic blood pressure ⭓90 mm Hg
likely.
Arterial oxygen saturation ⭓90% or pO2 ⭓60 mm Hg on room air
Conversely, a delay in antibiotic therapy has adverse conse- Ability to maintain oral intakea
quences in many infections. For critically ill, hemodynamically Normal mental status
a
unstable patients, early antibiotic therapy should be encouraged,

NOTE. Criteria are from [268, 274, 294]. pO2, oxygen partial pressure.
although no prospective data support this recommendation. De-

a
Important for discharge or oral switch decision but not necessarily for
lay in beginning antibiotic treatment during the transition from determination of nonresponse.
the ED is not uncommon. Especially with the frequent use of
once-daily antibiotics for CAP, timing and communication issues
at a predetermined time, regardless of the clinical response
may result in patients not receiving antibiotics for 18 h after
[269]. One study population with nonsevere illness was ran-
hospital admission. The committee felt that the best and most
domized to receive either oral therapy alone or intravenous
practical resolution to this issue was that the initial dose be given
therapy, with the switch occurring after 72 h without fever. The
in the ED [22].
study population with severe illness was randomized to receive
Data from the Medicare database indicated that antibiotic
either intravenous therapy with a switch to oral therapy after
treatment before hospital admission was also associated with
2 days or a full 10-day course of intravenous antibiotics. Time
lower mortality [109]. Given that there are even more concerns
to resolution of symptoms for the patients with nonsevere ill-
regarding timing of the first dose of antibiotic when the patient
ness was similar with either regimen. Among patients with more
is directly admitted to a busy inpatient unit, provision of the
severe illness, the rapid switch to oral therapy had the same
first dose in the physician’s office may be best if the recom-
rate of treatment failure and the same time to resolution of
mended oral or intramuscular antibiotics are available in the
symptoms as prolonged intravenous therapy. The rapid-switch
office.
group required fewer inpatient days (6 vs. 11), although this
was likely partially a result of the protocol, but the patients
Switch from Intravenous to Oral Therapy also had fewer adverse events.
The need to keep patients in the hospital once clinical sta-
30. Patients should be switched from intravenous to oral bility is achieved has been questioned, even though physicians
therapy when they are hemodynamically stable and im- commonly choose to observe patients receiving oral therapy
proving clinically, are able to ingest medications, and for ⭓1 day. Even in the presence of pneumococcal bacteremia,
have a normally functioning gastrointestinal tract. a switch to oral therapy can be safely done once clinical stability
(Strong recommendation; level II evidence.) is achieved and prolonged intravenous therapy is not needed
31. Patients should be discharged as soon as they are clin- [270]. Such patients generally take longer (approximately half
ically stable, have no other active medical problems, and a day) to become clinically stable than do nonbacteremic pa-
have a safe environment for continued care. Inpatient tients. The benefits of in-hospital observation after a switch to
observation while receiving oral therapy is not necessary. oral therapy are limited and add to the cost of care [32].
(Moderate recommendation; level II evidence.) Discharge should be considered when the patient is a can-
didate for oral therapy and when there is no need to treat any
With the use of a potent, highly bioavailable antibiotic, the comorbid illness, no need for further diagnostic testing, and
ability to eat and drink is the major consideration for switching no unmet social needs [32, 271, 272]. Although it is clear that
from intravenous to oral antibiotic therapy for non-ICU pa- clinically stable patients can be safely switched to oral therapy
tients. Initially, Ramirez et al. [268] defined a set of criteria for and discharged, the need to wait for all of the features of clinical
an early switch from intravenous to oral therapy (table 10). In stability to be present before a patient is discharged is uncertain.
general, as many as two-thirds of all patients have clinical im- For example, not all investigators have found it necessary to
provement and meet criteria for a therapy switch in the first have the white blood cell count improve. Using the definition
3 days, and most non-ICU patients meet these criteria by day for clinical stability in table 10, Halm et al. [273] found that
7. 19.1% of 680 patients were discharged from the hospital with
Subsequent studies have suggested that even more liberal ⭓1 instability. Death or readmission occurred in 10.5% of pa-
criteria are adequate for the switch to oral therapy. An alter- tients with no instability on discharge, in 13.7% of patients
native approach is to change from intravenous to oral therapy with 1 instability, and in 46.2% with ⭓2 instabilities. In general,

patients in higher PSI classes take longer to reach clinical sta- the minimal overall duration of therapy documented to be
bility than do patients in lower risk classes [274]. This finding effective in usual forms of CAP.
may reflect the fact that elderly patients with multiple comor- As is discussed above, most patients become clinically stable
bidities often recover more slowly. Arrangements for appro- within 3–7 days, so longer durations of therapy are rarely nec-
priate follow-up care, including rehabilitation, should therefore essary. Patients with persistent clinical instability are often read-
be initiated early for these patients. mitted to the hospital and may not be candidates for short-
In general, when switching to oral antibiotics, either the same duration therapy. Short-duration therapy may be suboptimal
agent as the intravenous antibiotic or the same drug class for patients with bacteremic S. aureus pneumonia (because of
should be used. Switching to a different class of agents simply the risk of associated endocarditis and deep-seated infection),
because of its high bioavailability (such as a fluoroquinolone) for those with meningitis or endocarditis complicating pneu-

is probably not necessary for a responding patient. For patients monia, and for those infected with other, less common path-
who received intravenous b-lactam–macrolide combination ogens (e.g., Burkholderia pseudomallei or endemic fungi). An
therapy, a switch to a macrolide alone appears to be safe for 8-day course of therapy for nosocomial P. aeruginosa pneu-
those who do not have DRSP or gram-negative enteric path- monia led to relapse more commonly than did a 15-day course
ogens isolated [275]. of therapy [279]. Whether the same results would be applicable
to CAP cases is unclear, but the presence of cavities or other
Duration of Antibiotic Therapy signs of tissue necrosis may warrant prolonged treatment. Stud-
ies of duration of therapy have focused on patients receiving
empirical treatment, and reliable data defining treatment du-
32. Patients with CAP should be treated for a minimum of
ration after an initially ineffective regimen are lacking.
5 days (level I evidence), should be afebrile for 48–72
h, and should have no more than 1 CAP-associated sign
of clinical instability (table 10) before discontinuation OTHER TREATMENT CONSIDERATIONS
of therapy (level II evidence). (Moderate recom-
mendation.) 34. Patients with CAP who have persistent septic shock de-
33. A longer duration of therapy may be needed if initial spite adequate fluid resuscitation should be considered for
therapy was not active against the identified pathogen treatment with drotrecogin alfa activated within 24 h of
or if it was complicated by extrapulmonary infection, admission. (Weak recommendation, level II evidence.)
such as meningitis or endocarditis. (Weak recommen-
dation; level III evidence.) Drotrecogin alfa activated is the first immunomodulatory
therapy approved for severe sepsis. In the United States, the
Most patients with CAP have been treated for 7–10 days or FDA recommended the use of drotrecogin alfa activated for
longer, but few well-controlled studies have evaluated the op- patients at high risk of death. The high-risk criterion suggested
timal duration of therapy for patients with CAP, managed in by the FDA was an Acute Physiologic and Chronic Health
or out of the hospital. Available data on short-course treatment Assessment (APACHE) II score ⭓25, based on a subgroup
do not suggest any difference in outcome with appropriate analysis of the overall study. However, the survival advantage
therapy in either inpatients or outpatients [276]. Duration is (absolute risk reduction, 9.8%) of drotrecogin alfa activated
also difficult to define in a uniform fashion, because some treatment of patients in the CAP subgroup was equivalent to
antibiotics (such as azithromycin) are administered for a short that in the subgroup with APACHE II scores ⭓25 [92, 280,
time yet have a long half-life at respiratory sites of infection. 281]. The greatest reduction in the mortality rate was for S.
In trials of antibiotic therapy for CAP, azithromycin has been pneumoniae infection (relative risk, 0.56; 95% CI, 0.35–0.88)
used for 3–5 days as oral therapy for outpatients, with some [282]. Subsequent data have suggested that the benefit appears
reports of single-dose therapy for patients with atypical path- to be greatest when the treatment is given as early in the hospital
ogen infections [276–278]. Results with azithromycin should admission as possible. In the subgroup with severe CAP caused
not be extrapolated to other drugs with significantly shorter by a pathogen other than S. pneumoniae and treated with ap-
half-lives. The ketolide telithromycin has been used for 5–7 propriate antibiotics, there was no evidence that drotrecogin
days to treat outpatients, including some with pneumococcal alfa activated affected mortality.
bacteremia or PSI classes ⭓III [211]. In a recent study, high- Although the benefit of drotrecogin alfa activated is clearly
dose (750 mg) levofloxacin therapy for 5 days was equally greatest for patients with CAP who have high APACHE II
successful and resulted in more afebrile patients by day 3 than scores, this criterion alone may not be adequate to select ap-
did the 500-mg dose for 7–10 days (49.1% vs. 38.5%; P p propriate patients. An APACHE II score ⭓25 was selected by
.03) [276]. On the basis of these studies, 5 days appears to be a subgroup analysis of the entire study cohort and may not be

similarly calibrated in a CAP-only cohort. Two-organ failure, domized to receive NIV had a 125% absolute risk reduction
the criterion suggested for drotrecogin alfa activated use by the for the need for intubation [114]. The use of NIV may also
European regulatory agency, did not influence the mortality improve intermediate-term mortality. Inability to expectorate
benefit for patients with CAP [92]. may limit the use of NIV [290], but intermittent application
Therefore, in addition to patients with septic shock, other of NIV may allow for its use in patients with productive cough
patients with severe CAP could be considered for treatment unless sputum production is excessive. Prompt recognition of
with drotrecogin alfa activated. Those with sepsis-induced leu- a failed NIV trial is critically important, because most studies
kopenia are at extremely high risk of death and ARDS and are, demonstrate worse outcomes for patients who require intu-
therefore, potential candidates. Conversely, the benefit of dro- bation after a prolonged NIV trial [288, 290]. Within the first
trecogin alfa activated is not as clear when respiratory failure 1–2 h of NIV, failure to improve respiratory rate and oxygen-

is caused more by exacerbation of underlying lung disease ation [114, 289, 290] or failure to decrease carbon dioxide
rather than by the pneumonia itself. Other minor criteria for partial pressure (pCO2) in patients with initial hypercarbia
severe CAP proposed above are similar to organ failure criteria [114] predicts NIV failure and warrants prompt intubation.
used in many sepsis trials. Consideration of treatment with NIV provides no benefit for patients with ARDS [289], which
drotrecogin alfa activated is appropriate, but the strength of may be nearly indistinguishable from CAP among patients with
the recommendation is only level II. bilateral alveolar infiltrates. Patients with CAP who have severe
hypoxemia (PaO2/FiO2 ratio, !150) are also poor candidates
for NIV [290].
35. Hypotensive, fluid-resuscitated patients with severe CAP
should be screened for occult adrenal insufficiency.
(Moderate recommendation; level II evidence.) 37. Low-tidal-volume ventilation (6 cm3/kg of ideal body
weight) should be used for patients undergoing venti-
A large, multicenter trial has suggested that stress-dose (200– lation who have diffuse bilateral pneumonia or ARDS.
300 mg of hydrocortisone per day or equivalent) steroid treat- (Strong recommendation; level I evidence.)
ment improves outcomes of vasopressor-dependent patients
with septic shock who do not have an appropriate cortisol Distinguishing between diffuse bilateral pneumonia and
response to stimulation [283]. Once again, patients with CAP ARDS is difficult, but it may not be an important distinction.
made up a significant fraction of patients entered into the trial. Results of the ARDSNet trial suggest that the use of low-tidal-
In addition, 3 small pilot studies have suggested that there is volume ventilation provides a survival advantage [291]. Pneu-
a benefit to corticosteroid therapy even for patients with severe monia, principally CAP, was the most common cause of ARDS
CAP who are not in shock [284–286]. The small sample size in that trial, and the benefit of the low-tidal-volume ventilatory
and baseline differences between groups compromise the con- strategy appeared to be equivalent in the population with pneu-
clusions. Although the criteria for steroid replacement therapy monia compared with the entire cohort. The absolute risk re-
remain controversial, the frequency of intermittent steroid duction for mortality in the pneumonia cohort was 11%, in-
treatment in patients at risk for severe CAP, such as those with dicating that, in order to avoid 1 death, 9 patients must be
severe COPD, suggests that screening of patients with severe treated [292].
CAP is appropriate with replacement if inadequate cortisol lev- Other aspects of the management of severe sepsis and septic
els are documented. If corticosteroids are used, close attention shock in patients with CAP do not appear to be significantly
to tight glucose control is required [287]. different from those for patients with other sources of infection.
Recommendations for these aspects of care are reviewed else-
where [293].
36. Patients with hypoxemia or respiratory distress should
receive a cautious trial of noninvasive ventilation (NIV) MANAGEMENT OF NONRESPONDING
unless they require immediate intubation because of se- PNEUMONIA
vere hypoxemia (arterial oxygen pressure/fraction of in-
spired oxygen [PaO2/FiO2] ratio, !150) and bilateral al- Because of the limitations of diagnostic testing, the majority of
veolar infiltrates. (Moderate recommendation; level I CAP is still treated empirically. Critical to empirical therapy is
evidence.) an understanding of the management of patients who do not
follow the normal response pattern.
Patients who do not require immediate intubation but who Although difficult to define, nonresponse is not uncommon.
have either hypoxemia or respiratory distress should receive a Overall, 6%–15% of hospitalized patients with CAP do not
trial of NIV [114, 288, 289]. Patients with underlying COPD respond to the initial antibiotic treatment [81, 84, 101, 294].
are most likely to benefit. Patients with CAP who were ran- The incidence of treatment failure among patients with CAP

who are not hospitalized is not well known, because popula- Table 11. Patterns and etiologies of types of failure to respond.
tion-based studies are required. Almirall et al. [295] described
an overall hospitalization rate of 60% in a population-based Failure to improve
study, but the rate of failure among the 30% of patients who Early (!72 h of treatment)
Normal response
initially presented to their primary care physician was not pro-
Delayed
vided. The frequency of prior antibiotic therapy among Med-
Resistant microorganism
icare patients admitted to the hospital with CAP is 24%–40%
Uncovered pathogen
[95, 109], but the percentage who received prior antibiotic Inappropriate by sensitivity
therapy for the acute episode of pneumonia itself versus other Parapneumonic effusion/empyema
indications is unclear. For patients initially admitted to the ICU, Nosocomial superinfection

the risk of failure to respond is already high; as many as 40% Nosocomial pneumonia
will experience deterioration even after initial stabilization in Extrapulmonary
the ICU [101]. Noninfectious
Mortality among nonresponding patients is increased sev- Complication of pneumonia (e.g., BOOP)
eral-fold in comparison with that among responding patients Misdiagnosis: PE, CHF, vasculitis
[296]. Overall mortality rates as high as 49% have been reported Drug fever
Deterioration or progression
for an entire population of nonresponding hospitalized patients
Early (!72 h of treatment)
with CAP [76, 84, 101], and the mortality rate reported in one
Severity of illness at presentation
study of early failure was 27% [81]. APACHE II score was not
Resistant microorganism
the only factor independently associated with mortality in the Uncovered pathogen
nonresponding group, suggesting that the excess mortality may Inappropriate by sensitivity
be due to factors other than severity of illness at presentation Metastatic infection
[101]. Empyema/parapneumonic
Endocarditis, meningitis, arthritis
Definition and classification. Inaccurate diagnosis
PE, aspiration, ARDS
38. The use of a systematic classification of possible causes Vasculitis (e.g., SLE)
of failure to respond, based on time of onset and type Delayed
of failure (table 11), is recommended. (Moderate rec- Nosocomial superinfection
ommendation; level II evidence.) Nosocomial pneumonia
Extrapulmonary
The term “nonresponding pneumonia” is used to define a Exacerbation of comorbid illness
Intercurrent noninfectious disease
situation in which an inadequate clinical response is present
PE
despite antibiotic treatment. Lack of a clear-cut and validated
Myocardial infarction
definition in the literature makes nonresponse difficult to study.
Renal failure
Lack of response also varies according to the site of treatment.
NOTE. ARDS, acute respiratory distress syndrome; BOOP, bronchiolitis
Lack of response in outpatients is very different from that in
obliterans organizing pneumonia; CHF, congestive heart failure; PE, pulmonary
patients admitted to the ICU. The time of evaluation is also embolus; SLE, systemic lupus erythematosis.
important. Persistent fever after the first day of treatment differs
significantly from fever persisting (or recurring) at day 7 of tory failure or hypotension 172 h after initial treatment is often
treatment. related to intercurrent complications, deterioration in under-
Table 11 provides a construct for evaluating nonresponse to lying disease, or development of nosocomial superinfection.
antibiotic treatment of CAP, based on several studies addressing The second pattern is that of persistent or nonresponding
this issue [76, 81, 84, 101]. Two patterns of unacceptable re- pneumonia. Nonresponse can be defined as absence of or delay
sponse are seen in hospitalized patients [101]. The first is pro- in achieving clinical stability, using the criteria in table 10 [274,
gressive pneumonia or actual clinical deterioration, with acute 294]. When these criteria were used, the median time to achieve
respiratory failure requiring ventilatory support and/or septic clinical stability was 3 days for all patients, but a quarter of
shock, usually occurring within the first 72 h of hospital ad- patients took ⭓6 days to meet all of these criteria for stability
mission. As is noted above, as many as 45% of patients with [274]. Stricter definitions for each of the criteria and higher
CAP who ultimately require ICU admission are initially ad- PSI scores were associated with longer times to achieve clinical
mitted to a non-ICU setting and are transferred because of stability. Conversely, subsequent transfer to the ICU after
deterioration [75]. Deterioration and development of respira- achieving this degree of clinical stability occurred in !1% of

cases. A separate multicenter trial demonstrated similar findings further diagnostic testing, and (3) escalation or change in treat-
[297]. Given these results, concern regarding nonresponse ment. Issues regarding hospital admission and ICU transfer are
should be tempered before 72 h of therapy. Antibiotic changes discussed above.
during this period should be considered only for patients with An inadequate host response, rather than inappropriate an-
deterioration or in whom new culture data or epidemiologic tibiotic therapy or unexpected microorganisms, is the most
clues suggest alternative etiologies. common cause of apparent antibiotic failure when guideline-
Finally, nonresolving or slow-resolving pneumonia has been recommended therapy is used. Decisions regarding further di-
used to refer to the conditions of patients who present with agnostic testing and antibiotic change/escalation are intimately
persistence of pulmonary infiltrates 130 days after initial pneu- intertwined and need to be discussed in tandem.
monia-like syndrome [298]. As many as 20% of these patients Information regarding the utility of extensive microbiological

will be found to have diseases other than CAP when carefully testing in cases of nonresponding CAP is mainly retrospective
evaluated [295]. and therefore affected by selection bias. A systematic diagnostic
Two studies have evaluated the risk factors for a lack of approach, which included invasive, noninvasive, and imaging
response in multivariate analyses [81, 84], including those ame- procedures, in a series of nonresponding patients with CAP
nable to medical intervention. Use of fluoroquinolones was obtained a specific diagnosis in 73% [101]. In a different study,
independently associated with a better response in one study mortality among patients with microbiologically guided versus
[84], whereas discordant antimicrobial therapy was associated empirical antibiotic changes was not improved (mortality rate,
with early failure [81]. In table 12, the different risk and pro- 67% vs. 64%, respectively) [76]. However, no antibiotic changes
tective factors and their respective odds ratios are summarized. were based solely on sputum smears, suggesting that invasive
Specific causes that may be responsible for a lack of response cultures or nonculture methods may be needed.
in CAP have been classified by Arancibia et al. [101] (table 11). Mismatch between the susceptibility of a common causative
This classification may be useful for clinicians as a systematic organism, infection with a pathogen not covered by the usual
approach to diagnose the potential causes of nonresponse in empirical regimen, and nosocomial superinfection pneumonia
CAP. Although in the original study only 8 (16%) of 49 cases are major causes of apparent antibiotic failure. Therefore, the
could not be classified [101], a subsequent prospective multi- first response to nonresponse or deterioration is to reevaluate
center trial found that the cause of failure could not be deter- the initial microbiological results. Culture or sensitivity data
mined in 44% [84]. not available at admission may now make the cause of clinical
Management of nonresponding CAP. Nonresponse to an- failure obvious. In addition, a further history of any risk factors
tibiotics in CAP will generally result in ⭓1 of 3 clinical re- for infection with unusual microorganisms (table 8) should be
sponses: (1) transfer of the patient to a higher level of care, (2) taken if not done previously. Viruses are relatively neglected as
Table 12. Factors associated with nonresponding pneumonia.
Overall failurea Early failureb

Risk factor Decreased risk Increased risk Decreased risk Increased risk
Older age (165 years) … … 0.35 …
COPD 0.60 … … …
Liver disease … 2.0 … …
Vaccination 0.3 … … …
Pleural effusion … 2.7 … …
Multilobar infiltrates … 2.1 … 1.81
Cavitation … 4.1 … …
Leukopenia … 3.7 … …
PSI class … 1.3 … 2.75
Legionella pneumonia … … … 2.71
Gram-negative pneumonia … … … 4.34
Fluoroquinolone therapy 0.5 … … …
Concordant therapy … … 0.61 …
Discordant therapy … … … 2.51
NOTE. Data are relative risk values. COPD, chronic obstructive pulmonary disease; PSI, Pneumonia Severity
Index.
a
From [84].
b
From [81].

a cause of infection in adults but may account for 10%–20% original pathogen and turn attention to other causes of failure.
of cases [299]. Other family members or coworkers may have In addition, a positive pneumococcal antigen test result would
developed viral symptoms in the interval since the patient was also help with interpretation of subsequent sputum/tracheal
admitted, increasing suspicion of this cause. aspirate cultures, which may indicate early superinfection.
The evaluation of nonresponse is severely hampered if a Nonresponse may also be mimicked by concomitant or sub-
microbiological diagnosis was not made on initial presentation. sequent extrapulmonary infection, such as intravascular cath-
If cultures were not obtained, clinical decisions are much more eter, urinary, abdominal, and skin infections, particularly in
difficult than if the adequate cultures were obtained but neg- ICU patients. Appropriate cultures of these sites should be
ative. Risk factors for nonresponse or deterioration (table 12), considered for patients with nonresponse to CAP therapy.
therefore, figure prominently in the list of situations in which In addition to microbiological diagnostic procedures, several

more aggressive initial diagnostic testing is warranted (table 5). other tests appear to be valuable for selected patients with non-
Blood cultures should be repeated for deterioration or pro- response:
gressive pneumonia. Deteriorating patients have many of the • Chest CT. In addition to ruling out pulmonary emboli, a
risk factors for bacteremia, and blood cultures are still high CT scan can disclose other reasons for antibiotic failure,
yield even in the face of prior antibiotic therapy [95]. Positive including pleural effusions, lung abscess, or central airway
blood culture results in the face of what should be adequate obstruction. The pattern of opacities may also suggest al-
antibiotic therapy should increase the suspicion of either an- ternative noninfectious disease, such as bronchiolitis obli-
tibiotic-resistant isolates or metastatic sites, such as endocarditis terans organizing pneumonia.
or arthritis. • Thoracentesis. Empyema and parapneumonic effusions are
Despite the high frequency of infectious pulmonary causes important causes of nonresponse [81, 101], and thoracen-
of nonresponse, the diagnostic utility of respiratory tract cul- tesis should be performed whenever significant pleural fluid
tures is less clear. Caution in the interpretation of sputum or is present.
tracheal aspirate cultures, especially of gram-negative bacilli, is • Bronchoscopy with BAL and transbronchial biopsies. If the
warranted because early colonization, rather than superinfec- differential of nonresponse includes noninfectious pneu-
tion with resistant bacteria, is not uncommon in specimens monia mimics, bronchoscopy will provide more diagnostic
obtained after initiation of antibiotic treatment. Once again, information than routine microbiological cultures. BAL may
the absence of multidrug-resistant pathogens, such as MRSA reveal noninfectious entities, such as pulmonary hemor-
or Pseudomonas, is strong evidence that they are not the cause rhage or acute eosinophilic pneumonia, or hints of infec-
of nonresponse. An etiology was determined by bronchoscopy tious diseases, such as lymphocytic rather than neutrophilic
alveolitis pointing toward virus or Chlamydophila infection.
in 44% of patients with CAP, mainly in those not responding
Transbronchial biopsies can also yield a specific diagnosis.
to therapy [300]. Despite the potential benefit suggested by
Antibiotic management of nonresponse in CAP has not been
these results, and in contrast to ventilator-associated pneu-
studied. The overwhelming majority of cases of apparent non-
monia [301, 302], no randomized study has compared the
response are due to the severity of illness at presentation or a
utility of invasive versus noninvasive strategies in the CAP pop-
delay in treatment response related to host factors. Other than
ulation with nonresponse.
the use of combination therapy for severe bacteremic pneu-
Rapid urinary antigen tests for S. pneumoniae and L. pneumo-
mococcal pneumonia [112, 231, 233, 234], there is no docu-
phila remain positive for days after initiation of antibiotic ther-
mentation that additional antibiotics for early deterioration lead
apy [147, 152] and, therefore, may be high-yield tests in this
to a better outcome. The presence of risk factors for potentially
group. A urinary antigen test result that is positive for L.
untreated microorganisms may warrant temporary empirical
pneumophila has several clinical implications, including that
broadening of the antibiotic regimen until results of diagnostic
coverage for Legionella should be added if not started empir-
tests are available.
ically [81]. This finding may be a partial explanation for the
finding that fluoroquinolones are associated with a lower in-
PREVENTION
cidence of nonresponse [84]. If a patient has persistent fever,
the faster response to fluoroquinolones in Legionella CAP war-
rants consideration of switching coverage from a macrolide 39. All persons ⭓50 years of age, others at risk for influenza
[303]. Stopping the b-lactam component of combination ther- complications, household contacts of high-risk persons,
apy to exclude drug fever is probably also safe [156]. Because and health care workers should receive inactivated in-
one of the major explanations for nonresponse is poor host fluenza vaccine as recommended by the Advisory Com-
immunity rather than incorrect antibiotics, a positive pneu- mittee on Immunization Practices (ACIP), CDC. (Strong
mococcal antigen test result would at least clarify the probable recommendation; level I evidence.)

Table 13. Recommendations for vaccine prevention of community-acquired pneumonia.
Pneumococcal Inactivated Live attenuated

Factor polysaccharide vaccine influenza vaccine influenza vaccine
Route of administration Intramuscular injection Intramuscular injection Intranasal spray
Type of vaccine Bacterial component (polysaccha- Killed virus Live virus
ride capsule)
Recommended groups All persons ⭓65 years of age All persons ⭓50 years of age Healthy persons 5–49 years of
a
age, including health care
providers and household con-
tacts of high-risk persons

High-risk persons 2–64 years of High-risk persons 6 months–49
age years of age
b
Current smokers Household contacts of high-risk
persons
Health care providers
Children 6–23 months of age
Specific high-risk indications for Chronic cardiovascular, pulmo- Chronic cardiovascular or pulmo- Avoid in high-risk persons
vaccination nary, renal, or liver disease nary disease (including
asthma)
Diabetes mellitus Chronic metabolic disease (in-
cluding diabetes mellitus)
Cerebrospinal fluid leaks Renal dysfunction
Alcoholism Hemoglobinopathies
Asplenia Immunocompromising condi-
tions/medications
Immunocompromising condi- Compromised respiratory func-
tions/medications tion or increased aspiration risk
Native Americans and Alaska Pregnancy
natives
Long-term care facility residents Residence in a long-term care
facility
Aspirin therapy in persons ⭐18
years of age
Revaccination schedule One-time revaccination after 5 Annual revaccination Annual revaccination
years for (1) adults ⭓65 years
of age, if the first dose is re-
ceived before age 65 years; (2)
persons with asplenia; and (3)
immunocompromised persons
NOTE. Adapted from the Advisory Committee on Immunization Practices, Centers for Disease Control and Prevention [304].
a
Avoid use in persons with asthma, reactive airways disease, or other chronic disorders of the pulmonary or cardiovascular systems; persons with other
underlying medical conditions, including diabetes, renal dysfunction, and hemoglobinopathies; persons with immunodeficiencies or who receive immunosup-
pressive therapy; children or adolescents receiving salicylates; persons with a history of Guillain-Barré syndrome; and pregnant women.
b
Vaccinating current smokers is recommended by the Pneumonia Guidelines Committee but is not currently an indication for vaccine according to the Advisory
Committee on Immunization Practices statement.
40. The intranasally administered live attenuated vaccine is high-risk concurrent diseases, according to current ACIP
an alternative vaccine formulation for some persons 5– guidelines. (Strong recommendation; level II evidence.)
49 years of age without chronic underlying diseases, in-
cluding immunodeficiency, asthma, or chronic medical Vaccines targeting pneumococcal disease and influenza re-
conditions. (Strong recommendation; level I evidence.) main the mainstay for preventing CAP. Pneumococcal poly-
41. Health care workers in inpatient and outpatient settings saccharide vaccine and inactivated influenza vaccine are rec-
and long-term care facilities should receive annual in- ommended for all older adults and for younger persons with
fluenza immunization. (Strong recommendation; level I medical conditions that place them at high risk for pneumonia
evidence.) morbidity and mortality (table 13) [304, 305]. The new live
42. Pneumococcal polysaccharide vaccine is recommended attenuated influenza vaccine is recommended for healthy per-
for persons ⭓65 years of age and for those with selected sons 5–49 years of age, including health care workers [304].

Postlicensure epidemiologic studies have documented the and winter. (Strong recommendation; level III evidence.)
effectiveness of pneumococcal polysaccharide vaccines for pre-
vention of invasive infection (bacteremia and meningitis) Many people who should receive either influenza or pneu-
among elderly individuals and younger adults with certain mococcal polysaccharide vaccine have not received them. Ac-
chronic medical conditions [306–309]. The overall effectiveness cording to a 2003 survey, only 69% of adults ⭓65 years of age
against invasive pneumococcal disease among persons ⭓65 had received influenza vaccine in the past year, and only 64%
years of age is 44%–75% [306, 308, 310], although efficacy may had ever received pneumococcal polysaccharide vaccine [322].
decrease with advancing age [308]. The effectiveness of the Coverage levels are lower for younger persons with vaccine
vaccine against pneumococcal disease in immunocompromised indications. Among adults 18–64 years of age with diabetes,
persons is less clear, and results of studies evaluating its effec- 49% had received influenza vaccine, and 37% had ever received

tiveness against pneumonia without bacteremia have been pneumococcal vaccine [323]. Studies of vaccine delivery meth-
mixed. The vaccine has been shown to be cost effective for ods indicate that the use of standing orders is the best way to
general populations of adults 50–64 years of age and ⭓65 years improve vaccination coverage in office, hospital, or long-term
of age [311, 312]. A second dose of pneumococcal polysac- care settings [324].
charide vaccine after a ⭓5-year interval has been shown to be Hospitalization of at-risk patients represents an underutilized
safe, with only slightly more local reactions than are seen after opportunity to assess vaccination status and to either provide
the first dose [313]. Because the safety of a third dose has not or recommend immunization. Ideally, patients should be vac-
been demonstrated, current guidelines do not suggest repeated cinated before developing pneumonia; therefore, admissions for
revaccination. The pneumococcal conjugate vaccine is under illnesses other than respiratory tract infections would be an
investigation for use in adults but is currently only licensed for appropriate focus. However, admission for pneumonia is an
use in young children [314, 315]. However, its use in children important trigger for assessing the need for immunization. The
!5 years of age has dramatically reduced invasive pneumococcal actual immunization may be better provided at the time of
bacteremia among adults as well [314, 316]. outpatient follow-up, especially with the emphasis on early dis-
The effectiveness of influenza vaccines depends on host fac- charge of patients with CAP. Patients with an acute fever should
tors and on how closely the antigens in the vaccine are matched not be vaccinated until their fever has resolved. Confusion of
with the circulating strain of influenza. A systematic review a febrile reaction to immunization with recurrent/superinfec-
demonstrates that influenza vaccine effectively prevents pneu- tion pneumonia is a risk. However, immunization at discharge
monia, hospitalization, and death [317, 318]. A recent large for pneumonia is warranted for patients for whom outpatient
observational study of adults ⭓65 years of age found that vac- follow-up is unreliable, and such vaccinations have been safely
cination against influenza was associated with a reduction in given to many patients.
the risk of hospitalization for cardiac disease (19% reduction), The best time for influenza vaccination in North America is
cerebrovascular disease (16%–23% reduction), and pneumonia October and November, although vaccination in December and
or influenza (29%–32% reduction) and a reduction in the risk later is recommended for those who were not vaccinated earlier.
of death from all causes (48%–50% reduction) [319]. In long- Influenza and pneumococcal vaccines can be given at the same
term-care facilities, vaccination of health care workers with time in different arms.
influenza vaccine is an important preventive health measure Chemoprophylaxis can be used as an adjunct to vaccination
[318, 320, 321]. Because the main virulence factors of influenza for prevention and control of influenza. Oseltamivir and zan-
virus, a neuraminidase and hemagglutinin, adapt quickly to amivir are both approved for prophylaxis; amantadine and ri-
selective pressures, new vaccine formulations are created each mantadine have FDA indications for chemoprophylaxis against
year on the basis of the strains expected to be circulating, and influenza A infection, but these agents are currently not rec-
annual revaccination is needed for optimal protection. ommended because of the frequency of resistance among
strains circulating in the United States and Canada [252, 253].
Developing an adequate immune response to the inactivated
43. Vaccination status should be assessed at the time of hos- influenza vaccine takes ∼2 weeks in adults; chemoprophylaxis
pital admission for all patients, especially those with may be useful during this period for those with household
medical illnesses. (Moderate recommendation; level III exposure to influenza, those who live or work in institutions
evidence.) with an influenza outbreak, or those who are at high risk for
44. Vaccination may be performed either at hospital dis- influenza complications in the setting of a community outbreak
charge or during outpatient treatment. (Moderate rec- [325, 326]. Chemoprophylaxis also may be useful for persons
ommendation; level III evidence.) with contraindications to influenza vaccine or as an adjunct to
45. Influenza vaccine should be offered to persons at hospital vaccination for those who may not respond well to influenza
discharge or during outpatient treatment during the fall vaccine (e.g., persons with HIV infection) [325, 326]. The use

of influenza antiviral medications for treatment or chemopro- Public health interventions are important for preventing
phylaxis should not affect the response to the inactivated vac- some forms of pneumonia. Notifying the state or local health
cine. Because it is unknown whether administering influenza department about a condition of interest is the first step to
antiviral medications affects the performance of the new live getting public health professionals involved. Rules and regu-
attenuated intranasal vaccine, this vaccine should not be used lations regarding which diseases are reportable differ between
in conjunction with antiviral agents. states. For pneumonia, most states require reporting for le-
Other types of vaccination can be considered. Pertussis is a gionnaires disease, SARS, and psittacosis, so that an investi-
rare cause of pneumonia itself. However, pneumonia is one of gation can determine whether others may be at risk and whether
the major complications of pertussis. Concern over waning control measures are necessary. For legionnaires disease, re-
immunity has led the ACIP to emphasize adult immunization porting of cases has helped to identify common-source out-

for pertussis [327]. One-time vaccination with the new tetanus breaks caused by environmental contamination [130]. For
toxoid, reduced diphtheria toxoid, and acellular pertussis vac- SARS, close observation and, in some cases, quarantine of close
cine—adsorbed (Tdap) product, ADACEL (Sanofi Pasteur)— contacts have been critical for controlling transmission [330].
is recommended for adults 19–64 years of age. For most adults, In addition, any time avian influenza (H5N1) or a possible
the vaccine should be given in place of their next routine tet- terrorism agent (e.g., plague, tularemia, or anthrax) is being
anus-diphtheria booster; adults with close contact with infants considered as the etiology of pneumonia, the case should be
!12 months of age and health care workers should receive the reported immediately, even before a definitive diagnosis is ob-
vaccine as soon as possible, with an interval as short as 2 years tained. In addition, pneumonia cases that are caused by path-
after their most recent tetanus/diphtheria booster. ogens not thought to be endemic to the area should be reported,
even if those conditions are not typically on the list of reportable
conditions, because control strategies might be possible.
46. Smoking cessation should be a goal for persons hospi-
For other respiratory diseases, episodes that are suspected of
talized with CAP who smoke. (Moderate recommen-
being part of an outbreak or cluster should be reported. For
pneumococcal disease and influenza, outbreaks can occur in
47. Smokers who will not quit should also be vaccinated for
crowded settings of susceptible hosts, such as homeless shelters,
both pneumococcus and influenza. (Weak recommen-
nursing homes, and jails. In these settings, prophylaxis, vac-
cination, and infection control methods are used to control
Smoking is associated with a substantial risk of pneumococcal further transmission [331]. For Mycoplasma, antibiotic pro-
bacteremia; one report showed that smoking was the strongest phylaxis has been used in schools and institutions to control
of multiple risks for invasive pneumococcal disease in immu- outbreaks [332].
nocompetent nonelderly adults [328]. Smoking has also been
identified as a risk for Legionella infection [329]. Smoking ces-
sation should be attempted when smokers are hospitalized; this 49. Respiratory hygiene measures, including the use of hand
is particularly important and relevant when these patients are hygiene and masks or tissues for patients with cough,
hospitalized for pneumonia. Materials for clinicians and patients should be used in outpatient settings and EDs as a means
to assist with smoking cessation are available online from the US to reduce the spread of respiratory infections. (Strong
Surgeon General (http://www.surgeongeneral.gov/tobacco), the recommendation; level III evidence.)
Centers for Disease Control and Prevention (http://www.cdc.gov/
In part because of the emergence of SARS, improved respi-
tobacco), and the American Cancer Society (http://www
ratory hygiene measures (“respiratory hygiene” or “cough et-
.cancer.org). The most successful approaches to quitting include
iquette”) have been promoted as a means for reducing trans-
some combination of nicotine replacement and/or bupropion,
mission of respiratory infections in outpatient clinics and EDs
a method to change habits, and emotional support. Given the
[333]. Key components of respiratory hygiene include en-
increased risk of pneumonia, the committee felt that persons
couraging patients to alert providers when they present for a
unwilling to stop smoking should be given the pneumococcal
visit and have symptoms of a respiratory infection; the use of
polysaccharide vaccine, although this is not currently an ACIP-
hand hygiene measures, such as alcohol-based hand gels; and
recommended indication.
the use of masks or tissues to cover the mouth for patients
with respiratory illnesses. In a survey of the US population, the
48. Cases of pneumonia that are of public health concern use of masks in outpatient settings was viewed as an acceptable
should be reported immediately to the state or local means for reducing the spread of respiratory infections [334].
health department. (Strong recommendation; level III For hospitalized patients, infection control recommendations
evidence.) typically are pathogen specific. For more details on the use of

personal protective equipment and other measures to prevent ated. Trying to increase the number of protected individuals
transmission within health care settings, refer to the Healthcare is a desirable end point and, therefore, a goal worth pur-
Infection Control Practices Advisory Committee [335]. suing. This is particularly true for influenza, because the
vaccine data are more compelling, but it is important to try
SUGGESTED PERFORMANCE INDICATORS to protect against pneumococcal infection as well. Coverage
of 90% of adults ⭓65 years of age should be the target.
Performance indicators are tools to help guideline users mea-
sure both the extent and the effects of implementation of guide- Acknowledgments
lines. Such tools or measures can be indicators of the process
The committee wishes to express its gratitude to Robert Balk, Christian
itself, outcomes, or both. Deviations from the recommenda-
Brun-Buisson, Ali El-Sohl, Alan Fein, Donald E. Low, Constantine Man-

tions are expected in a proportion of cases, and compliance in thous, Thomas J. Marrie, Joseph F. Plouffe, and David A. Talan, for their
80%–95% of cases is generally appropriate, depending on the thoughtful review of an earlier version of the guidelines.
indicator. Supplement sponsorship. This article was published as part of a sup-
plement entitled “Infectious Diseases Society of America/American Tho-
Four specific performance indicators have been selected for racic Society Consensus Guidelines on the Management of Community-
the CAP guidelines, 3 of which focus on treatment issues and Acquired Pneumonia in Adults,” sponsored by the Infectious Diseases
1 of which deals with prevention: Society of America.
Potential conflicts of interest. L.A.M. has received research funding
• Initial empirical treatment of CAP should be consistent with from Bayer, Chiron, Ortho-McNeil, Oscient, and Pfizer; has served as a
guideline recommendations. Data exist that support the role consultant to Bayer, Cempra, Novexel, Ortho-McNeil, Oscient, Pfizer, San-
of CAP guidelines and that have demonstrated reductions ofi-Aventis, Targanta, and Wyeth; and has served on speakers’ bureaus for
Bayer, Ortho-McNeil, Oscient, Pfizer, and Sanofi-Aventis. R.G.W. has re-
in cost, LOS, and mortality when the guidelines are fol-
ceived research funding from Chiron, Eli Lilly, Pfizer, and Wyeth; has served
lowed. Reasons for deviation from the guidelines should be on the Clinical Evaluation Committee for Johnson and Johnson; has served
clearly documented in the medical record. as a clinical trial participant in studies initiated by Takeda, Biosite, Inverness
• The first treatment dose for patients who are to be admitted Medical Intervention, Johnson and Johnson, and Altana; and has served
as consultant to the Oklahoma Foundation for Medical Quality and the
to the hospital should be given in the ED. Unlike in prior Centers for Medicare and Medicaid Services. J.G.B. serves on the advisory
guidelines, a specific time frame is not being recommended. board of Johnson and Johnson. T.M.F. has received research funding from
Initiation of treatment would be expected within 6–8 h of Binax Incorporated, Ortho-McNeil, Oscient, Pfizer, and Sanofi-Aventis; has
served as a consultant to Bayer, GlaxoSmithKline, Merck, Ortho-McNeil,
presentation whenever the admission diagnosis is likely CAP. Oscient, Pfizer, Sanofi-Aventis, Schering-Plough, and Wyeth; and has served
A rush to treatment without a diagnosis of CAP can, how- on speakers’ bureaus for Abbott, GlaxoSmithKline, Merck, Ortho-McNeil,
ever, result in the inappropriate use of antibiotics with a Oscient, Pfizer, Sanofi-Aventis, Schering-Plough, and Wyeth. N.A.D has
received research support from Altana and Sanofi-Aventis; has served on
concomitant increase in costs, adverse drug events, in-
the advisory boards for Sanofi-Aventis and AstraZeneca; and has served
creased antibiotic selection pressure, and, possibly, increased on the speakers’ bureaus for Pfizer, Schering-Plough, Sanofi-Aventis, and
antibiotic resistance. Consideration should be given to mon- Merck. A.A. has served on the speakers’ bureaus for Altana, Bayer Pharma,
itoring the number of patients who receive empirical an- Boehringer-Ingelheim, Chiron, Elan, GlaxoSmithKline, Ortho-McNeil,
Pfizer, and Sanofi-Aventis; has served as a consultant and on advisory
tibiotics in the ED but are admitted to the hospital without boards for Altana, Bayer Pharma, Boehringer-Ingelheim, Chiron, Elan,
an infectious diagnosis. GlaxoSmithKline, Ortho-McNeil, Pfizer, and Sanofi-Aventis; and has re-
• Mortality data for all patients with CAP admitted to wards, ceived research funding from BART, Bayer Pharma, Boehringer-Ingelheim,
GlaxoSmithKline, and Lilly. M.S.N. serves on the speakers’ bureaus for and
ICUs, or high-level monitoring units should be collected. as a consultant to AstraZeneca, Aventis, Elan, Merck, Ortho-McNeil, Pfizer,
Although tools to predict mortality and severity of illness Schering-Plough, and Wyeth. All other authors: no conflicts.
exist—such as the PSI and CURB-65 criteria, respectively—
none is foolproof. Overall mortality rates for all patients References
with CAP admitted to the hospital, including general med- 1. National Center for Health Statistics. Health, United States, 2006,
ical wards, should be monitored and compared with sever- with chartbook on trends in the health of Americans. Available at:
ity-adjusted norms. In addition, careful attention should be http://www.cdc.gov/nchs/data/hus/hus06.pdf. Accessed 17 January
2007.
paid to the percentage of patients with severe CAP, as de-
2. Jackson ML, Neuzil KM, Thompson WW, et al. The burden of com-
fined in this document, who are admitted initially to a non- munity-acquired pneumonia in seniors: results of a population-based
ICU or a high-level monitoring unit and to their mortality study. Clin Infect Dis 2004; 39:1642–50.
3. Feikin DR, Schuchat A, Kolczak M, et al. Mortality from invasive
rate.
pneumococcal pneumonia in the era of antibiotic resistance,
• It is important to determine what percentage of at-risk pa- 1995–1997. Am J Public Health 2000; 90:223–9.
tients in one’s practice actually receive immunization for 4. Flanders SA, Halm EA. Guidelines for community-acquired pneu-
influenza or pneumococcal infection. Prevention of infec- monia: are they reflected in practice? Treat Respir Med 2004; 3:67–77.
5. Niederman MS, Bass JB, Campbell GD, et al. Guidelines for the initial
tion is clearly more desirable than having to treat established management of adults with community-acquired pneumonia: diag-
infection, but it is clear that target groups are undervaccin- nosis, assessment of severity, and initial antimicrobial therapy. Amer-

ican Thoracic Society. Medical Section of the American Lung Asso- outpatients: an interventional trial. Arch Intern Med 1998; 158:
ciation. Am Rev Respir Dis 1993; 148:1418–26. 1350–6.
6. Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for the 26. Coley CM, Li YH, Medsger AR, et al. Preferences for home vs hospital
management of adults with community-acquired pneumonia: diag- care among low-risk patients with community-acquired pneumonia.
nosis, assessment of severity, antimicrobial therapy, and prevention. Arch Intern Med 1996; 156:1565–71.
Am J Respir Crit Care Med 2001; 163:1730–54. 27. Meehan TP, Weingarten SR, Holmboe ES, et al. A statewide initiative
7. Bartlett JG, Breiman RF, Mandell LA, File TM Jr. Community-ac- to improve the care of hospitalized pneumonia patients: The Con-
quired pneumonia in adults: guidelines for management. The Infec- necticut Pneumonia Pathway Project. Am J Med 2001; 111:203–10.
tious Diseases Society of America. Clin Infect Dis 1998; 26:811–38. 28. Fine MJ, Stone RA, Lave JR, et al. Implementation of an evidence-
8. Bartlett JG, Dowell SF, Mandell LA, File Jr TM, Musher DM, Fine based guideline to reduce duration of intravenous antibiotic therapy
MJ. Practice guidelines for the management of community-acquired and length of stay for patients hospitalized with community-acquired
pneumonia in adults. Infectious Diseases Society of America. Clin pneumonia: a randomized controlled trial. Am J Med 2003; 115:
Infect Dis 2000; 31:347–82. 343–51.

9. Mandell LA, Bartlett JG, Dowell SF, File TM Jr, Musher DM, Whitney 29. Chu LA, Bratzler DW, Lewis RJ, et al. Improving the quality of care
C. Update of practice guidelines for the management of community- for patients with pneumonia in very small hospitals. Arch Intern Med
acquired pneumonia in immunocompetent adults. Clin Infect Dis 2003; 163:326–32.
2003; 37:1405–33. 30. Mundy LM, Leet TL, Darst K, Schnitzler MA, Dunagan WC. Early
10. Guidelines for the management of adults with hospital-acquired, ven- mobilization of patients hospitalized with community-acquired pneu-
tilator-associated, and healthcare-associated pneumonia. Am J Respir monia. Chest 2003; 124:883–9.
Crit Care Med 2005; 171:388–416. 31. Samama MM, Cohen AT, Darmon JY, et al. A comparison of enox-
11. Dean NC, Silver MP, Bateman KA. Frequency of subspecialty phy- aparin with placebo for the prevention of venous thromboembolism
sician care for elderly patients with community-acquired pneumonia. in acutely ill medical patients. Prophylaxis in Medical Patients with
Chest 2000; 117:393–7. Enoxaparin Study Group. N Engl J Med 1999; 341:793–800.
12. Kish MA. Guide to development of practice guidelines. Clin Infect 32. Rhew DC, Riedinger MS, Sandhu M, Bowers C, Greengold N, Wein-
Dis 2001; 32:851–4. garten SR. A prospective, multicenter study of a pneumonia practice
13. Attributes of ATS documents that guide clinical practice. Recom- guideline. Chest 1998; 114:115–9.
mendations of the ATS Clinical Practice Committee. Am J Respir Crit 33. Halm EA, Horowitz C, Silver A, et al. Limited impact of a multicenter
Care Med 1997; 156:2015–25. intervention to improve the quality and efficiency of pneumonia care.
14. Atkins D, Eccles M, Flottorp S, et al. Systems for grading the quality Chest 2004; 126:100–7.
of evidence and the strength of recommendations I: critical appraisal 34. Niederman MS, McCombs JS, Unger AN, Kumar A, Popovian R. The
of existing approaches The GRADE Working Group. BMC Health cost of treating community-acquired pneumonia. Clin Ther 1998; 20:
820–37.
Serv Res 2004; 4:38.
35. Carratala J, Fernandez-Sabe N, Ortega L, et al. Outpatient care com-
15. Wunderink RG. Clinical practice guidelines for the management of
pared with hospitalization for community-acquired pneumonia: a
pneumonia—do they work? New Horiz 1998; 6:75–83.
randomized trial in low-risk patients. Ann Intern Med 2005; 142:
16. Cabana MD, Rand CS, Powe NR, et al. Why don’t physicians follow
165–72.
clinical practice guidelines? A framework for improvement. JAMA
36. Alikhan R, Cohen AT, Combe S, et al. Risk factors for venous throm-
1999; 282:1458–65.
boembolism in hospitalized patients with acute medical illness: anal-
17. Nathwani D, Rubinstein E, Barlow G, Davey P. Do guidelines for
ysis of the MEDENOX Study. Arch Intern Med 2004; 164:963–8.
community-acquired pneumonia improve the cost-effectiveness of
37. McMahon LF Jr, Wolfe RA, Tedeschi PJ. Variation in hospital ad-
hospital care? Clin Infect Dis 2001; 32:728–41.
missions among small areas: a comparison of Maine and Michigan.
18. Menendez R, Ferrando D, Valles JM, Vallterra J. Influence of deviation
Med Care 1989; 27:623–31.
from guidelines on the outcome of community-acquired pneumonia.
38. Fine MJ, Hough LJ, Medsger AR, et al. The hospital admission de-
Chest 2002; 122:612.
cision for patients with community-acquired pneumonia: results from
19. Dean NC, Silver MP, Bateman KA, James B, Hadlock CJ, Hale D. the pneumonia Patient Outcomes Research Team cohort study. Arch
Decreased mortality after implementation of a treatment guideline Intern Med 1997; 157:36–44.
for community-acquired pneumonia. Am J Med 2001; 110:451–7. 39. British Thoracic Society Research Committee. Community-acquired
20. Marrie TJ, Lau CY, Wheeler SL, Wong CJ, Vandervoort MK, Feagan pneumonia in adults in British hospitals in 1982–1983: a survey of
BG. A controlled trial of a critical pathway for treatment of com- aetiology, mortality, prognostic factors, and outcome. Q J Med
munity-acquired pneumonia. CAPITAL Study Investigators. Com- 1987; 62:195–220.
munity-Acquired Pneumonia Intervention Trial Assessing Levoflox- 40. Black ER, Mushlin AI, Griner PF, Suchman AL, James RL, Schoch
acin. JAMA 2000; 283:749–55. DR. Predicting the need for hospitalization of ambulatory patients
21. Capelastegui A, Espana PP, Quintana JM, et al. Improvement of pro- with pneumonia. J Gen Intern Med 1991; 6:394–400.
cess-of-care and outcomes after implementing a guideline for the 41. Daley J, Jencks S, Draper D, Lenhart G, Thomas N, Walker J. Pre-
management of community-acquired pneumonia: a controlled be- dicting hospital-associated mortality for Medicare patients: a method
fore-and-after design study. Clin Infect Dis 2004; 39:955–63. for patients with stroke, pneumonia, acute myocardial infarction, and
22. Benenson R, Magalski A, Cavanaugh S, Williams E. Effects of a pneu- congestive heart failure. JAMA 1988; 260:3617–24.
monia clinical pathway on time to antibiotic treatment, length of stay, 42. Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify
and mortality. Acad Emerg Med 1999; 6:1243–8. low-risk patients with community-acquired pneumonia. N Engl J Med
23. Suchyta MR, Dean NC, Narus S, Hadlock CJ. Effects of a practice 1997; 336:243–50.
guideline for community-acquired pneumonia in an outpatient set- 43. Fine MJ, Smith DN, Singer DE. Hospitalization decision in patients
ting. Am J Med 2001; 110:306–9. with community-acquired pneumonia: a prospective cohort study.
24. Mortensen EM, Restrepo M, Anzueto A, Pugh J. Effects of guideline- Am J Med 1990; 89:713–21.
concordant antimicrobial therapy on mortality among patients with 44. Fine MJ, Singer DE, Hanusa BH, Lave JR, Kapoor WN. Validation
community-acquired pneumonia. Am J Med 2004; 117:726–31. of a pneumonia prognostic index using the MedisGroups Compar-
25. Atlas SJ, Benzer TI, Borowsky LH, et al. Safely increasing the pro- ative Hospital Database. Am J Med 1993; 94:153–9.
portion of patients with community-acquired pneumonia treated as 45. Lim WS, van der Eerden MM, Laing R, et al. Defining community

acquired pneumonia severity on presentation to hospital: an inter- identify low-risk patients with community-acquired pneumonia.
national derivation and validation study. Thorax 2003; 58:377–82. Chest 2000; 118:1339–43.
46. Marrie TJ, Durant H, Yates L. Community-acquired pneumonia re- 68. Roson B, Carratala J, Dorca J, Casanova A, Manresa F, Gudiol F.
quiring hospitalization: 5-year prospective study. Rev Infect Dis Etiology, reasons for hospitalization, risk classes, and outcomes of
1989; 11:586–99. community-acquired pneumonia in patients hospitalized on the basis
47. Ortqvist A, Hedlund J, Grillner L, et al. Aetiology, outcome and of conventional admission criteria. Clin Infect Dis 2001; 33:158–65.
prognostic factors in community-acquired pneumonia requiring hos- 69. El Solh AA, Sikka P, Ramadan F, Davies J. Etiology of severe pneu-
pitalization. Eur Respir J 1990; 3:1105–13. monia in the very elderly. Am J Respir Crit Care Med 2001; 163:
48. Porath A, Schlaeffer F, Lieberman D. Appropriateness of hospitali- 645–51.
zation of patients with community-acquired pneumonia. Ann Emerg 70. Luna CM, Famiglietti A, Absi R, et al. Community-acquired pneu-
Med 1996; 27:176–83. monia: etiology, epidemiology, and outcome at a teaching hospital in
49. Auble TE, Yealy DM, Fine MJ. Assessing prognosis and selecting an Argentina. Chest 2000; 118:1344–54.
initial site of care for adults with community-acquired pneumonia. 71. Garcia-Ordonez MA, Garcia-Jimenez JM, Paez F, et al. Clinical aspects

Infect Dis Clin North Am 1998; 12:741–59, x. and prognostic factors in elderly patients hospitalised for community-
50. Yealy DM, Auble TE, Stone RA, et al. Effect of increasing the intensity acquired pneumonia. Eur J Clin Microbiol Infect Dis 2001; 20:14–9.
of implementing pneumonia guidelines: a randomized, controlled 72. Meehan TP, Chua-Reyes JM, Tate J, et al. Process of care performance,
trial. Ann Intern Med 2005; 143:881–94. patient characteristics, and outcomes in elderly patients hospitalized
51. Neill AM, Martin IR, Weir R, et al. Community acquired pneumonia: with community-acquired or nursing home-acquired pneumonia.
aetiology and usefulness of severity criteria on admission. Thorax Chest 2000; 117:1378–85.
1996; 51:1010–6. 73. Lim WS, Macfarlane JT, Boswell TC, et al. Study of community ac-
52. Dean NC. Use of prognostic scoring and outcome assessment tools quired pneumonia aetiology (SCAPA) in adults admitted to hospital:
in the admission decision for community-acquired pneumonia. Clin implications for management guidelines. Thorax 2001; 56:296–301.
Chest Med 1999; 20:521–9, viii. 74. Leroy O, Santre C, Beuscart C, et al. A five-year study of severe
53. Woodhead M. Assessment of illness severity in community acquired community-acquired pneumonia with emphasis on prognosis in pa-
pneumonia: a useful new prediction tool? Thorax 2003; 58:371–2. tients admitted to an intensive care unit. Intensive Care Med 1995;
54. Capelastegui A, Espana PP, Quintana JM, et al. Validation of a pre- 21:24–31.
dictive rule for the management of community-acquired pneumonia. 75. Ewig S, de Roux A, Bauer T, et al. Validation of predictive rules and
Eur Respir J 2006; 27:151–7. indices of severity for community acquired pneumonia. Thorax
55. Espana PP, Capelastegui A, Quintana JM, et al. A prediction rule to 2004; 59:421–7.
identify allocation of inpatient care in community-acquired pneu- 76. Sanyal S, Smith PR, Saha AC, Gupta S, Berkowitz L, Homel P. Initial
microbiologic studies did not affect outcome in adults hospitalized
monia. Eur Respir J 2003; 21:695–701.
with community-acquired pneumonia. Am J Respir Crit Care Med
56. Aujesky D, Auble TE, Yealy DM, et al. Prospective comparison of
1999; 160:346–8.
three validated prediction rules for prognosis in community-acquired
77. Marik PE. The clinical features of severe community-acquired pneu-
pneumonia. Am J Med 2005; 118:384–92.
monia presenting as septic shock. Norasept II Study Investigators. J
57. Dean NC, Suchyta MR, Bateman KA, Aronsky D, Hadlock CJ. Im-
Crit Care 2000; 15:85–90.
plementation of admission decision support for community-acquired
78. Ruiz M, Ewig S, Torres A, et al. Severe community-acquired pneu-
pneumonia. Chest 2000; 117:1368–77.
monia. Am J Respir Crit Care Med 1999; 160:923–9.
58. Aronsky D, Dean NC. How should we make the admission decision
79. Paganin F, Lilienthal F, Bourdin A, et al. Severe community-acquired
in community-acquired pneumonia? Med Clin North Am 2001; 85:
pneumonia: assessment of microbial aetiology as mortality factor. Eur
1397–411.
Respir J 2004; 24:779–85.
59. Wright AA, Maydom BW. Improving the implementation of com-
80. Kollef MH, Sherman G, Ward S, Fraser VJ. Inadequate antimicrobial
munity-acquired pneumonia guidelines. Intern Med J 2004; 34:507–9.
treatment of infections: a risk factor for hospital mortality among
60. Arnold FW, Ramirez JA, McDonald LC, Xia EL. Hospitalization for critically ill patients. Chest 1999; 115:462–74.
community-acquired pneumonia: the pneumonia severity index vs 81. Roson B, Carratala J, Fernandez-Sabe N, Tubau F, Manresa F, Gudiol
clinical judgment. Chest 2003; 124:121–4. F. Causes and factors associated with early failure in hospitalized
61. Goss CH, Rubenfeld GD, Park DR, Sherbin VL, Goodman MS, Root patients with community-acquired pneumonia. Arch Intern Med
RK. Cost and incidence of social comorbidities in low-risk patients 2004; 164:502–8.
with community-acquired pneumonia admitted to a public hospital. 82. Ewig S, Ruiz M, Mensa J, et al. Severe community-acquired pneu-
Chest 2003; 124:2148–55. monia: assessment of severity criteria. Am J Respir Crit Care Med
62. Riley PD, Aronsky D, Dean NC. Validation of the 2001 American 1998; 158:1102–8.
Thoracic Society criteria for severe community-acquired pneumonia. 83. Mortensen EM, Coley CM, Singer DE, et al. Causes of death for
Crit Care Med 2004; 32:2398–402. patients with community-acquired pneumonia: results from the
63. Angus DC, Marrie TJ, Obrosky DS, et al. Severe community-acquired Pneumonia Patient Outcomes Research Team cohort study. Arch In-
pneumonia: use of intensive care services and evaluation of American tern Med 2002; 162:1059–64.
and British Thoracic Society Diagnostic criteria. Am J Respir Crit 84. Menendez R, Torres A, Zalacain R, et al. Risk factors of treatment
Care Med 2002; 166:717–23. failure in community acquired pneumonia: implications for disease
64. Halm EA, Atlas SJ, Borowsky LH, et al. Understanding physician outcome. Thorax 2004; 59:960–5.
adherence with a pneumonia practice guideline: effects of patient, 85. Perlino CA, Rimland D. Alcoholism, leukopenia, and pneumococcal
system, and physician factors. Arch Intern Med 2000; 160:98–104. sepsis. Am Rev Respir Dis 1985; 132:757–60.
65. Marrie TJ, Wu L. Factors influencing in-hospital mortality in com- 86. Watanakunakorn C, Bailey TA. Adult bacteremic pneumococcal pneu-
munity-acquired pneumonia: a prospective study of patients not ini- monia in a community teaching hospital, 1992–1996: a detailed anal-
tially admitted to the ICU. Chest 2005; 127:1260–70. ysis of 108 cases. Arch Intern Med 1997; 157:1965–71.
66. Metlay JP, Fine MJ. Testing strategies in the initial management of 87. Leroy O, Georges H, Beuscart C, et al. Severe community-acquired
patients with community-acquired pneumonia. Ann Intern Med pneumonia in ICUs: prospective validation of a prognostic score.
2003; 138:109–18. Intensive Care Medicine 1996; 22:1307–14.
67. Marras TK, Gutierrez C, Chan CK. Applying a prediction rule to 88. Feldman C, Smith C, Levy H, Ginsburg P, Miller SD. Klebsielle pneu-

moniae bacteremia at an urban general hospital. J Infect 1990; 20: talized with community-acquired pneumonia. Arch Intern Med
21–31. 2004; 164:637–44.
89. Chen MZ, Hsueh PR, Lee LN, Yu CJ, Yang PC, Luh KT. Severe 110. Fine MJ, Smith MA, Carson CA, et al. Prognosis and outcomes of
community-acquired pneumonia due to Acinetobacter baumannii. patients with community-acquired pneumonia: a meta-analysis.
Chest 2001; 120:1072–7. JAMA 1996; 275:134–41.
90. Querol-Ribelles JM, Tenias JM, Grau E, et al. Plasma d-dimer levels 111. Metersky ML. Is the lateral decubitus radiograph necessary for the
correlate with outcomes in patients with community-acquired pneu- management of a parapneumonic pleural effusion? Chest 2003; 124:
monia. Chest 2004; 126:1087–92. 1129–32.
91. Kuikka A, Syrjanen J, Renkonen OV, Valtonen VV. Pneumococcal 112. Yu VL, Chiou CC, Feldman C, et al. An international prospective
bacteraemia during a recent decade. J Infect 1992; 24:157–68. study of pneumococcal bacteremia: correlation with in vitro resis-
92. Laterre PF, Garber G, Levy H, et al. Severe community-acquired pneu- tance, antibiotics administered, and clinical outcome. Clin Infect Dis
monia as a cause of severe sepsis: data from the PROWESS study. 2003; 37:230–7.
Crit Care Med 2005; 33:952–61. 113. Ruiz M, Ewig S, Marcos MA, et al. Etiology of community-acquired

93. Marik PE, Zaloga GP. Hypothermia and cytokines in septic shock. pneumonia: impact of age, comorbidity, and severity. Am J Respir
Norasept II Study Investigators: North American study of the safety Crit Care Med 1999; 160:397–405.
and efficacy of murine monoclonal antibody to tumor necrosis factor 114. Confalonieri M, Potena A, Carbone G, Porta RD, Tolley EA, Meduri
for the treatment of septic shock. Intensive Care Med 2000; 26:716–21. GU. Acute respiratory failure in patients with severe community-
94. Fernández-Solá J, Torres A, Estruch R, Monforte J, Urbano-Márquez acquired pneumonia: a prospective randomized evaluation of non-
A. High alcohol intake as a risk and prognostic factor for community- invasive ventilation. Am J Respir Crit Care Med 1999; 160:1585–91.
acquired pneumonia. Arch Intern Med 1995; 155:1649–54. 115. Barrett-Connor E. The nonvalue of sputum culture in the diagnosis
95. Metersky ML, Ma A, Bratzler DW, Houck PM. Predicting bacteremia of pneumococcal pneumonia. Am Rev Respir Dis 1971; 103:845–8.
in patients with community-acquired pneumonia. Am J Respir Crit 116. Lentino JR, Lucks DA. Nonvalue of sputum culture in the manage-
Care Med 2004; 169:342–7. ment of lower respiratory tract infections. J Clin Microbiol 1987; 25:
96. Wipf JE, Lipsky BA, Hirschmann JV, et al. Diagnosing pneumonia by 758–62.
physical examination: relevant or relic? Arch Intern Med 1999; 159: 117. Musher DM, Montoya R, Wanahita A. Diagnostic value of micro-
1082–7. scopic examination of Gram-stained sputum and sputum cultures in
97. Mower WR, Sachs C, Nicklin EL, Safa P, Baraff LJ. Effect of routine patients with bacteremic pneumococcal pneumonia. Clin Infect Dis
emergency department triage pulse oximetry screening on medical 2004; 39:165–9.
management. Chest 1995; 108:1297–302. 118. Gleckman R, DeVita J, Hibert D, Pelletier C, Martin R. Sputum gram
98. Levin KP, Hanusa BH, Rotondi A, et al. Arterial blood gas and pulse stain assessment in community-acquired bacteremic pneumonia. J
oximetry in initial management of patients with community-acquired Clin Microbiol 1988; 26:846–9.
pneumonia. J Gen Intern Med 2001; 16:590–8. 119. Reed WW, Byrd GS, Gates RH Jr, Howard RS, Weaver MJ. Sputum
99. Syrjala H, Broas M, Suramo I, Ojala A, Lahde S. High-resolution gram’s stain in community-acquired pneumococcal pneumonia: a
computed tomography for the diagnosis of community-acquired meta-analysis. West J Med 1996; 165:197–204.
pneumonia. Clin Infect Dis 1998; 27:358–63. 120. Garcia-Vazquez E, Marcos MA, Mensa J, et al. Assessment of the
100. Valdivia L, Nix D, Wright M, et al. Coccidioidomycosis as a common usefulness of sputum culture for diagnosis of community-acquired
cause of community-acquired pneumonia. Emerg Infect Dis 2006; pneumonia using the PORT predictive scoring system. Arch Intern
12:958–62. Med 2004; 164:1807–11.
101. Arancibia F, Ewig S, Martinez JA, et al. Antimicrobial treatment fail- 121. Bartlett JG. Diagnostic accuracy of transtracheal aspiration bacteri-
ures in patients with community-acquired pneumonia. Am J Respir ologic studies. Am Rev Respir Dis 1977; 115:777–82.
Crit Care Med 2000; 162:154–60. 122. Bartlett JG, Finegold SM. Bacteriology of expectorated sputum with
102. King MD, Whitney CG, Parekh F, Farley MM. Recurrent invasive quantitative culture and wash technique compared to transtracheal
pneumococcal disease: a population-based assessment. Clin Infect Dis aspirates. Am Rev Respir Dis 1978; 117:1019–27.
2003; 37:1029–36. 123. Jimenez P, Saldias F, Meneses M, Silva ME, Wilson MG, Otth L.
103. US Department of Health and Human Services. Sepsis and CAP: Diagnostic fiberoptic bronchoscopy in patients with community-ac-
partnerships for diagnostics development. RFA no. RFA-AI-04-043. quired pneumonia: comparison between bronchoalveolar lavage and
Available at: http://grants.nih.gov/grants/guide/rfa-files/RFA-AI-04- telescoping plugged catheter cultures. Chest 1993; 103:1023–7.
043.html. Accessed 16 January 2007. 124. Zalacain R, Llorente JL, Gaztelurrutia L, Pijoan JI, Sobradillo V. In-
104. Campbell SG, Marrie TJ, Anstey R, Dickinson G, Ackroyd-Stolarz S. fluence of three factors on the diagnostic effectiveness of transthoracic
The contribution of blood cultures to the clinical management of needle aspiration in pneumonia. Chest 1995; 107:96–100.
adult patients admitted to the hospital with community-acquired 125. Scott JA, Hall AJ. The value and complications of percutaneous trans-
pneumonia: a prospective observational study. Chest 2003; 123:1142. thoracic lung aspiration for the etiologic diagnosis of community-
105. Waterer GW, Wunderink RG. The influence of the severity of com- acquired pneumonia. Chest 1999; 116:1716–32.
munity-acquired pneumonia on the usefulness of blood cultures. Res- 126. Ishida T, Hashimoto T, Arita M, et al. Efficacy of transthoracic needle
pir Med 2001; 95:78–82. aspiration in community-acquired pneumonia. Intern Med 2001; 40:
106. Malcolm C, Marrie TJ. Antibiotic therapy for ambulatory patients 873–7.
with community-acquired pneumonia in an emergency department 127. Bartlett JG. Diagnosis of bacterial infections of the lung. Clin Chest
setting. Arch Intern Med 2003; 163:797–802. Med 1987; 8:119–34.
107. Falsey AR, Hennessey PA, Formica MA, Cox C, Walsh EE. Respiratory 128. Heineman HS, Chawla JK, Lopton WM. Misinformation from spu-
syncytial virus infection in elderly and high-risk adults. N Engl J Med tum cultures without microscopic examination. J Clin Microbiol
2005; 352:1749–59. 1977; 6:518–27.
108. van der Eerden MM, Vlaspolder F, de Graaff CS, et al. Comparison 129. Wimberley N, Faling LJ, Bartlett JG. A fiberoptic bronchoscopy tech-
between pathogen directed antibiotic treatment and empirical broad nique to obtain uncontaminated lower airway secretions for bacterial
spectrum antibiotic treatment in patients with community acquired culture. Am Rev Respir Dis 1979; 119:337–43.
pneumonia: a prospective randomised study. Thorax 2005; 60:672–8. 130. Fields BS, Benson RF, Besser RE. Legionella and Legionnaires’ disease:
109. Houck PM, Bratzler DW, Nsa W, Ma A, Bartlett JG. Timing of an- 25 years of investigation. Clin Microbiol Rev 2002; 15:506–26.
tibiotic administration and outcomes for Medicare patients hospi- 131. Arancibia F, Bauer TT, Ewig S, et al. Community-acquired pneumonia

due to gram-negative bacteria and Pseudomonas aeruginosa: incidence, 150. Murdoch DR, Laing RT, Mills GD, et al. Evaluation of a rapid im-
risk, and prognosis. Arch Intern Med 2002; 162:1849–58. munochromatographic test for detection of Streptococcus pneumoniae
132. Fraser DW, Tsai TR, Orenstein W, et al. Legionnaires’ disease: de- antigen in urine samples from adults with community-acquired pneu-
scription of an epidemic of pneumonia. N Engl J Med 1977; 297: monia. J Clin Microbiol 2001; 39:3495–8.
1189–97. 151. Navarro D, Garcia-Maset L, Gimeno C, Escribano A, Garcia-de-Lomas
133. Cowgill KD, Lucas CE, Benson RF, et al. Recurrence of legionnaires J. Performance of the Binax NOW Streptococcus pneumoniae urinary
disease at a hotel in the United States Virgin Islands over a 20-year antigen assay for diagnosis of pneumonia in children with underlying
period. Clin Infect Dis 2005; 40:1205–7. pulmonary diseases in the absence of acute pneumococcal infection.
134. Jernigan DB, Hofmann J, Cetron MS, et al. Outbreak of Legionnaires’ J Clin Microbiol 2004; 42:4853–5.
disease among cruise ship passengers exposed to a contaminated 152. Murdoch DR, Laing RT, Cook JM. The NOW S. pneumoniae urinary
whirlpool spa. Lancet 1996; 347:494–9. antigen test positivity rate 6 weeks after pneumonia onset and among
135. den Boer JW, Yzerman EP, Schellekens J, et al. A large outbreak of patients with COPD. Clin Infect Dis 2003; 37:153–4.
Legionnaires’ disease at a flower show, the Netherlands, 1999. Emerg 153. Murdoch DR. Diagnosis of Legionella infection. Clin Infect Dis

Infect Dis 2002; 8:37–43. 2003; 36:64–9.
136. Garcia-Fulgueiras A, Navarro C, Fenoll D, et al. Legionnaires’ disease 154. Waterer GW, Baselski VS, Wunderink RG. Legionella and community-
outbreak in Murcia, Spain. Emerg Infect Dis 2003; 9:915–21. acquired pneumonia: a review of current diagnostic tests from a cli-
137. Benkel DH, McClure EM, Woolard D, et al. Outbreak of Legionnaires’ nician’s viewpoint. Am J Med 2001; 110:41–8.
disease associated with a display whirlpool spa. Int J Epidemiol 155. Bates JH, Campbell GD, Barron AL, et al. Microbial etiology of acute
2000; 29:1092–8. pneumonia in hospitalized patients. Chest 1992; 101:1005–12.
138. Helbig JH, Uldum SA, Luck PC, Harrison TG. Detection of Legionella 156. Plouffe JF, Breiman RF, Fields BS, et al. Azithromycin in the treatment
pneumophila antigen in urine samples by the BinaxNOW immu- of Legionella pneumonia requiring hospitalization. Clin Infect Dis
nochromatographic assay and comparison with both Binax Legionella 2003; 37:1475–80.
Urinary Enzyme Immunoassay (EIA) and Biotest Legionella Urin An- 157. Kaiser L, Briones MS, Hayden FG. Performance of virus isolation and
tigen EIA. J Med Microbiol 2001; 50:509–16. Directigen Flu A to detect influenza A virus in experimental human
139. Yzerman EP, den Boer JW, Lettinga KD, Schellekens J, Dankert J, infection. J Clin Virol 1999; 14:191–7.
Peeters M. Sensitivity of three urinary antigen tests associated with 158. Bellei N, Benfica D, Perosa AH, Carlucci R, Barros M, Granato C.
clinical severity in a large outbreak of Legionnaires’ disease in The Evaluation of a rapid test (QuickVue) compared with the shell vial
Netherlands. J Clin Microbiol 2002; 40:3232–6. assay for detection of influenza virus clearance after antiviral treat-
140. Roson B, Fernandez-Sabe N, Carratala J, et al. Contribution of a ment. J Virol Methods 2003; 109:85–8.
urinary antigen assay (Binax NOW) to the early diagnosis of pneu- 159. Landry ML, Cohen S, Ferguson D. Comparison of Binax NOW and
mococcal pneumonia. Clin Infect Dis 2004; 38:222–6. Directigen for rapid detection of influenza A and B. J Clin Virol
141. Ishida T, Hashimoto T, Arita M, Tojo Y, Tachibana H, Jinnai M. A 2004; 31:113–5.
3-year prospective study of a urinary antigen-detection test for Strep- 160. Monto AS, Gravenstein S, Elliott M, Colopy M, Schweinle J. Clinical
tococcus pneumoniae in community-acquired pneumonia: utility and signs and symptoms predicting influenza infection. Arch Intern Med
clinical impact on the reported etiology. J Infect Chemother 2004; 2000; 160:3243–7.
10:359–63. 161. Shetty AK, Treynor E, Hill DW, Gutierrez KM, Warford A, Baron EJ.
142. Stralin K, Kaltoft MS, Konradsen HB, Olcen P, and Holmberg H. Comparison of conventional viral cultures with direct fluorescent
Comparison of two urinary antigen tests for establishment of pneu- antibody stains for diagnosis of community-acquired respiratory virus
mococcal etiology of adult community-acquired pneumonia. J Clin infections in hospitalized children. Pediatr Infect Dis J 2003; 22:
Microbiol 2004; 42:3620–5. 789–94.
143. Marcos MA, Jimenez de Anta MT, De La Bellacasa JP, et al. Rapid 162. Chan KH, Maldeis N, Pope W, et al. Evaluation of the Directigen
urinary antigen test for diagnosis of pneumococcal community-ac- FluA+B test for rapid diagnosis of influenza virus type A and B
quired pneumonia in adults. Eur Respir J 2003; 21:209–14. infections. J Clin Microbiol 2002; 40:1675–80.
144. Rodriguez R, Fancher M, Phelps M, et al. An emergency department- 163. Casiano-Colon AE, Hulbert BB, Mayer TK, Walsh EE, Falsey AR.
based randomized trial of nonbronchoscopic bronchoalveolar lavage Lack of sensitivity of rapid antigen tests for the diagnosis of respiratory
for early pathogen identification in severe community-acquired pneu- syncytial virus infection in adults. J Clin Virol 2003; 28:169–74.
monia. Ann Emerg Med 2001; 38:357–63. 164. Littman AJ, Jackson LA, White E, Thornquist MD, Gaydos CA,
145. van der Eerden MM, Vlaspolder F, de Graaff CS, Groot T, Jansen Vaughan TL. Interlaboratory reliability of microimmunofluorescence
HM, Boersma WG. Value of intensive diagnostic microbiological in- test for measurement of Chlamydia pneumoniae-specific immuno-
vestigation in low- and high-risk patients with community-acquired globulin A and G antibody titers. Clin Diagn Lab Immunol 2004; 11:
pneumonia. Eur J Clin Microbiol Infect Dis 2005; 24:241–9. 615–7.
146. Dominguez J, Gali N, Blanco S, et al. Detection of Streptococcus pneu- 165. Bartlett JG. Diagnostic test for etiologic agents of community-acquired
moniae antigen by a rapid immunochromatographic assay in urine pneumonia. Infect Dis Clin North Am 2004; 18:809–27.
samples. Chest 2001; 119:243–9. 166. Dowell SF, Peeling RW, Boman J, et al. Standardizing Chlamydia
147. Smith MD, Derrington P, Evans R, et al. Rapid diagnosis of bacteremic pneumoniae assays: recommendations from the Centers for Disease
pneumococcal infections in adults by using the Binax NOW Strep- Control and Prevention (USA) and the Laboratory Centre for Disease
tococcus pneumoniae urinary antigen test: a prospective, controlled Control (Canada). Clin Infect Dis 2001; 33:492–503.
clinical evaluation. J Clin Microbiol 2003; 41:2810–3. 167. Templeton KE, Scheltinga SA, van den Eeden WC, Graffelman AW,
148. Benson RF, Tang PW, Fields BS. Evaluation of the Binax and Biotest van den Broek PJ, Claas EC. Improved diagnosis of the etiology of
urinary antigen kits for detection of Legionnaires’ disease due to community-acquired pneumonia with real-time polymerase chain re-
multiple serogroups and species of Legionella. J Clin Microbiol action. Clin Infect Dis 2005; 41:345–51.
2000; 38:2763–5. 168. Mundy LM, Auwaerter PG, Oldach D, et al. Community-acquired
149. Gutierrez F, Masia M, Rodriguez JC, et al. Evaluation of the im- pneumonia: impact of immune status. Am J Respir Crit Care Med
munochromatographic Binax NOW assay for detection of Strepto- 1995; 152:1309–15.
coccus pneumoniae urinary antigen in a prospective study of com- 169. Peiris JS, Chu CM, Cheng VC, et al. Clinical progression and viral
munity-acquired pneumonia in Spain. Clin Infect Dis 2003; 36: load in a community outbreak of coronavirus-associated SARS pneu-
286–92. monia: a prospective study. Lancet 2003; 361:1767–72.

170. Revised U.S. surveillance case definition for severe acute respiratory treatment in patients with bacteremia due to erythromycin-resistant
syndrome (SARS) and update on SARS cases—United States and Streptococcus pneumoniae. Clin Infect Dis 2002; 35:556–64.
worldwide, December 2003. MMWR Morb Mortal Wkly Rep 2003; 190. Davidson R, Cavalcanti R, Brunton JL, et al. Resistance to levofloxacin
52:1202–6. and failure of treatment of pneumococcal pneumonia. N Engl J Med
171. File TM. Community-acquired pneumonia. Lancet 2003; 362: 2002; 346:747–50.
1991–2001. 191. Ho PL, Yung RW, Tsang DN, et al. Increasing resistance of Strepto-
172. Fang GD, Fine M, Orloff J, et al. New and emerging etiologies for coccus pneumoniae to fluoroquinolones: results of a Hong Kong mul-
community-acquired pneumonia with implications for therapy: a pro- ticentre study in 2000. J Antimicrob Chemother 2001; 48:659–65.
spective multi-center study of 359 cases. Medicine (Baltimore) 192. Campbell GD Jr, Silberman R. Drug-resistant Streptococcus pneu-
1990; 69:307–16. moniae. Clin Infect Dis 1998; 26:1188–95.
173. Farr BM, Kaiser DL, Harrison BD, Connolly CK. Prediction of mi- 193. Clavo-Sanchez AJ, Giron-Gonzalez JA, Lopez-Prieto D, et al. Mul-
crobial aetiology at admission to hospital for pneumonia from the tivariate analysis of risk factors for infection due to penicillin-resistant
presenting clinical features. British Thoracic Society Pneumonia Re- and multidrug-resistant Streptococcus pneumoniae: a multicenter

search Subcommittee. Thorax 1989; 44:1031–5. study. Clin Infect Dis 1997; 24:1052–9.
174. Marrie TJ, Peeling RW, Fine MJ, Singer DE, Coley CM, Kapoor WN. 194. Vanderkooi OG, Low DE, Green K, Powis JE, McGeer A. Predicting
Ambulatory patients with community-acquired pneumonia: the fre- antimicrobial resistance in invasive pneumococcal infections. Clin In-
quency of atypical agents and clinical course. Am J Med 1996; 101: fect Dis 2005; 40:1288–97.
508–15. 195. Ho PL, Tse WS, Tsang KW, et al. Risk factors for acquisition of
175. Lieberman D, Schlaeffer F, Boldur I, et al. Multiple pathogens in adult levofloxacin-resistant Streptococcus pneumoniae: a case-control study.
patients admitted with community-acquired pneumonia: a one year Clin Infect Dis 2001; 32:701–7.
prospective study of 346 consecutive patients. Thorax 1996; 51: 196. Ruhe JJ, Hasbun R. Streptococcus pneumoniae bacteremia: duration
179–84. of previous antibiotic use and association with penicillin resistance.
176. de Roux A, Marcos MA, Garcia E, et al. Viral community-acquired Clin Infect Dis 2003; 36:1132–8.
pneumonia in nonimmunocompromised adults. Chest 2004; 125: 197. Anderson KB, Tan JS, File TM Jr, DiPersio JR, Willey BM, Low DE.
1343–51. Emergence of levofloxacin-resistant pneumococci in immunocom-
177. Falguera M, Sacristan O, Nogues A, et al. Nonsevere community- promised adults after therapy for community-acquired pneumonia.
acquired pneumonia: correlation between cause and severity or com- Clin Infect Dis 2003; 37:376–81.
orbidity. Arch Intern Med 2001; 161:1866–72. 198. Urban C, Rahman N, Zhao X, et al. Fluoroquinolone-resistant Strep-
178. Sirvent JM, Torres A, El Ebiary M, Castro P, de Batlle J, Bonet A. tococcus pneumoniae associated with levofloxacin therapy. J Infect Dis
Protective effect of intravenously administered cefuroxime against 2001; 184:794–8.
nosocomial pneumonia in patients with structural coma. Am J Respir 199. Gorak EJ, Yamada SM, Brown JD. Community-acquired methicillin-
Crit Care Med 1997; 155:1729–34. resistant Staphylococcus aureus in hospitalized adults and children
without known risk factors. Clin Infect Dis 1999; 29:797–800.
179. Waterer GW, Buckingham SC, Kessler LA, Quasney MW, Wunderink
200. Dufour P, Gillet Y, Bes M, et al. Community-acquired methicillin-
RG. Decreasing b-lactam resistance in Pneumococci from the Mem-
resistant Staphylococcus aureus infections in France: emergence of a
phis region: analysis of 2,152 isolates from 1996 to 2001. Chest
single clone that produces Panton-Valentine leukocidin. Clin Infect
2003; 124:519–25.
Dis 2002; 35:819–24.
180. Chen DK, McGeer A, de Azavedo JC, Low DE. Decreased suscepti-
201. Mongkolrattanothai K, Boyle S, Kahana MD, Daum RS. Severe Staph-
bility of Streptococcus pneumoniae to fluoroquinolones in Canada.
ylococcus aureus infections caused by clonally related community-
Canadian Bacterial Surveillance Network. N Engl J Med 1999; 341:
acquired methicillin-susceptible and methicillin-resistant isolates. Clin
233–9.
Infect Dis 2003; 37:1050–8.
181. Hyde TB, Gay K, Stephens DS, et al. Macrolide resistance among
202. Deresinski S. Methicillin-resistant Staphylococcus aureus: an evolu-
invasive Streptococcus pneumoniae isolates. JAMA 2001; 286:1857–62.
tionary, epidemiologic, and therapeutic odyssey. Clin Infect Dis
182. Perez-Trallero E, Garcia-de-la-Fuente C, Garcia-Rey C, et al. Geo- 2005; 40:562–73.
graphical and ecological analysis of resistance, coresistance, and cou- 203. Fridkin SK, Hageman JC, Morrison M, et al. Methicillin-resistant
pled resistance to antimicrobials in respiratory pathogenic bacteria in Staphylococcus aureus disease in three communities. N Engl J Med
Spain. Antimicrob Agents Chemother 2005; 49:1965–72. 2005; 352:1436–44.
183. Metlay JP. Antibacterial drug resistance: implications for the treatment 204. Micek ST, Dunne M, Kollef MH. Pleuropulmonary complications of
of patients with community-acquired pneumonia. Infect Dis Clin Panton-Valentine leukocidin-positive community-acquired methicil-
North Am 2004; 18:777–90. lin-resistant Staphylococcus aureus: importance of treatment with an-
184. Bauer T, Ewig S, Marcos MA, Schultze-Werninghaus G, Torres A. timicrobials inhibiting exotoxin production. Chest 2005; 128:2732–8.
Streptococcus pneumoniae in community-acquired pneumonia: how 205. Bochud PY, Moser F, Erard P, et al. Community-acquired pneumonia:
important is drug resistance? Med Clin North Am 2001; 85:1367–79. a prospective outpatient study. Medicine (Baltimore) 2001; 80:75–87.
185. Heffelfinger JD, Dowell SF, Jorgensen JH, et al. Management of com- 206. Shefet D, Robenshtok E, Paul M, Leibovici L. Empirical atypical cov-
munity-acquired pneumonia in the era of pneumococcal resistance: erage for inpatients with community-acquired pneumonia: systematic
a report from the Drug-Resistant Streptococcus pneumoniae Thera- review of randomized controlled trials. Arch Intern Med 2005; 165:
peutic Working Group. Arch Intern Med 2000; 160:1399–408. 1992–2000.
186. Pallares R, Capdevila O, Linares J, et al. The effect of cephalosporin 207. Mills GD, Oehley MR, Arrol B. Effectiveness of b-lactam antibiotics
resistance on mortality in adult patients with nonmeningeal systemic compared with antibiotics active against atypical pathogens in non-
pneumococcal infections. Am J Med 2002; 113:120–6. severe community acquired pneumonia: meta-analysis. BMJ 2005;
187. Musher DM, Dowell ME, Shortridge VD, et al. Emergence of mac- 330:456–60.
rolide resistance during treatment of pneumococcal pneumonia. N 208. McCracken GH Jr. Current status of antibiotic treatment for Myco-
Engl J Med 2002; 346:630–1. plasma pneumoniae infections. Pediatr Infect Dis 1986; 5:167–71.
188. Kelley MA, Weber DJ, Gilligan P, Cohen MS. Breakthrough pneu- 209. Lautenbach E, Larosa LA, Kasbekar N, Peng HP, Maniglia RJ, Fishman
mococcal bacteremia in patients being treated with azithromycin and NO. Fluoroquinolone utilization in the emergency departments of
clarithromycin. Clin Infect Dis 2000; 31:1008–11. academic medical centers: prevalence of, and risk factors for, inap-
189. Lonks JR, Garau J, Gomez L, et al. Failure of macrolide antibiotic propriate use. Arch Intern Med 2003; 163:601–5.

210. Pichichero ME. A review of evidence supporting the American Acad- quired pneumonia: a 3 1/2-year experience from a veterans affairs
emy of Pediatrics recommendation for prescribing cephalosporin an- hospital. Arch Intern Med 2003; 163:1718–26.
tibiotics for penicillin-allergic patients. Pediatrics 2005; 115:1048–57. 228. Marras TK, Nopmaneejumruslers C, Chan CK. Efficacy of exclusively
211. Tellier G, Niederman MS, Nusrat R, Patel M, Lavin B. Clinical and oral antibiotic therapy in patients hospitalized with nonsevere com-
bacteriological efficacy and safety of 5 and 7 day regimens of telith- munity-acquired pneumonia: a retrospective study and meta-analysis.
romycin once daily compared with a 10 day regimen of clarithromycin Am J Med 2004; 116:385–93.
twice daily in patients with mild to moderate community-acquired 229. Leroy O, Saux P, Bedos JP, Caulin E. Comparison of levofloxacin and
pneumonia. J Antimicrob Chemother 2004; 54:515–23. cefotaxime combined with ofloxacin for ICU patients with com-
212. Mathers DL, Hassman J, Tellier G. Efficacy and tolerability of once- munity-acquired pneumonia who do not require vasopressors. Chest
daily oral telithromycin compared with clarithromycin for the treat- 2005; 128:172–83.
ment of community-acquired pneumonia in adults. Clin Ther 230. Torres A, Serra-Batlles J, Ferrer A, et al. Severe community-acquired
2004; 26:48–62. pneumonia: epidemiology and prognostic factors. Am Rev Respir Dis
213. Pullman J, Champlin J, Vrooman PS Jr. Efficacy and tolerability of 1991; 144:312–8.

once-daily oral therapy with telithromycin compared with trovaflox- 231. Mufson MA, Stanek RJ. Bacteremic pneumococcal pneumonia in one
acin for the treatment of community-acquired pneumonia in adults. American City: a 20-year longitudinal study, 1978–1997. Am J Med
Int J Clin Pract 2003; 57:377–84. 1999; 107:34S-43S.
214. Hagberg L, Carbon C, van Rensburg DJ, Fogarty C, Dunbar L, Pull- 232. Baddour LM, Yu VL, Klugman KP, et al. Combination antibiotic
man J. Telithromycin in the treatment of community-acquired pneu- therapy lowers mortality among severely ill patients with pneumo-
monia: a pooled analysis. Respir Med 2003; 97:625–33. coccal bacteremia. Am J Respir Crit Care Med 2004; 170:440–4.
215. Clay KD, Hanson JS, Pope SD, Rissmiller RW, Purdum PP III, Banks 233. Waterer GW, Somes GW, Wunderink RG. Monotherapy may be sub-
PM. Brief communication: severe hepatotoxicity of telithromycin: optimal for severe bacteremic pneumococcal pneumonia. Arch Intern
three case reports and literature review. Ann Intern Med 2006; 144: Med 2001; 161:1837–42.
415–20. 234. Martinez JA, Horcajada JP, Almela M, et al. Addition of a macrolide
216. Gleason PP, Meehan TP, Fine JM, Galusha DH, Fine MJ. Associations to a b-lactam–based empirical antibiotic regimen is associated with
between initial antimicrobial therapy and medical outcomes for hos- lower in-hospital mortality for patients with bacteremic pneumococ-
pitalized elderly patients with pneumonia. Arch Intern Med 1999; cal pneumonia. Clin Infect Dis 2003; 36:389–95.
159:2562–72. 235. Weiss K, Low DE, Cortes L, et al. Clinical characteristics at initial
217. Houck PM, MacLehose RF, Niederman MS, Lowery JK. Empiric an- presentation and impact of dual therapy on the outcome of bacteremic
tibiotic therapy and mortality among Medicare pneumonia inpatients Streptococcus pneumoniae pneumonia in adults. Can Respir J 2004;
in 10 western states: 1993, 1995, and 1997. Chest 2001; 119:1420–6. 11:589–93.
218. Dudas V, Hopefl A, Jacobs R, Guglielmo BJ. Antimicrobial selection 236. Sopena N, Sabria M. Multicenter study of hospital-acquired pneu-
for hospitalized patients with presumed community-acquired pneumonia in non-ICU patients. Chest 2005; 127:213–9.
monia: a survey of nonteaching US community hospitals. Ann Phar- 237. Venezia RA, Domaracki BE, Evans AM, Preston KE, Graffunder EM.
macother 2000; 34:446–52. Selection of high-level oxacillin resistance in heteroresistant Staphy-
219. Brown RB, Iannini P, Gross P, Kunkel M. Impact of initial antibiotic lococcus aureus by fluoroquinolone exposure. J Antimicrob Chemo-
choice on clinical outcomes in community-acquired pneumonia: anal- ther 2001; 48:375–81.
ysis of a hospital claims-made database. Chest 2003; 123:1503–11. 238. Gonzalez C, Rubio M, Romero-Vivas J, Gonzalez M, Picazo JJ. Bac-
220. Musher DM, Bartlett JG, Doern GV. A fresh look at the definition teremic pneumonia due to Staphylococcus aureus: a comparison of
of susceptibility of Streptococcus pneumoniae to beta-lactam antibi- disease caused by methicillin-resistant and methicillin-susceptible or-
otics. Arch Intern Med 2001; 161:2538–44. ganisms. Clin Infect Dis 1999; 29:1171–7.
221. Vetter N, Cambronero-Hernandez E, Rohlf J, et al. A prospective, 239. Markowitz N, Quinn EL, Saravolatz LD. Trimethoprim-sulfameth-
randomized, double-blind multicenter comparison of parenteral er- oxazole compared with vancomycin for the treatment of Staphylo-
tapenem and ceftriaxone for the treatment of hospitalized adults with coccus aureus infection. Ann Intern Med 1992; 117:390–8.
community-acquired pneumonia. Clin Ther 2002; 24:1770–85. 240. San Pedro GS, Cammarata SK, Oliphant TH, Todisco T. Linezolid
222. Ortiz-Ruiz G, Vetter N, Isaacs R, Carides A, Woods GL, Friedland I. versus ceftriaxone/cefpodoxime in patients hospitalized for the treat-
Ertapenem versus ceftriaxone for the treatment of community-ac- ment of Streptococcus pneumoniae pneumonia. Scand J Infect Dis
quired pneumonia in adults: combined analysis of two multicentre 2002; 34:720–8.
randomized, double-blind studies. J Antimicrob Chemother 2004; 241. Wunderink RG, Rello J, Cammarata SK, Croos-Dabrera RV, Kollef
53(Suppl 2):ii59-66. MH. Linezolid vs vancomycin: analysis of two double-blind studies
223. Ailani RK, Agastya G, Ailani RK, Mukunda BN, Shekar R. Doxycycline of patients with methicillin-resistant Staphylococcus aureus nosocomial
is a cost-effective therapy for hospitalized patients with community- pneumonia. Chest 2003; 124:1789–97.
acquired pneumonia. Arch Intern Med 1999; 159:266–70. 242. Bernardo K, Pakulat N, Fleer S, et al. Subinhibitory concentrations
224. Ragnar NS. Atypical pneumonia in the Nordic countries: aetiology of linezolid reduce Staphylococcus aureus virulence factor expression.
and clinical results of a trial comparing fleroxacin and doxycycline. Antimicrob Agents Chemother 2004; 48:546–55.
Nordic Atypical Pneumonia Study Group. J Antimicrob Chemother 243. American Thoracic Society, Centers for Disease Control and Preven-
1997; 39:499–508. tion, Infectious Diseases Society of America. American Thoracic So-
225. Vergis EN, Indorf A, File TM Jr, et al. Azithromycin vs cefuroxime ciety/Centers for Disease Control and Prevention/Infectious Diseases
plus erythromycin for empirical treatment of community-acquired Society of America: controlling tuberculosis in the United States. Am
pneumonia in hospitalized patients: a prospective, randomized, mul- J Respir Crit Care Med 2005; 172:1169–227.
ticenter trial. Arch Intern Med 2000; 160:1294–300. 244. Amsden GW. Anti-inflammatory effects of macrolides—an under-
226. Plouffe J, Schwartz DB, Kolokathis A, et al. Clinical efficacy of in- appreciated benefit in the treatment of community-acquired respi-
travenous followed by oral azithromycin monotherapy in hospitalized ratory tract infections and chronic inflammatory pulmonary condi-
patients with community-acquired pneumonia. The Azithromycin In- tions? J Antimicrob Chemother 2005; 55:10–21.
travenous Clinical Trials Group. Antimicrob Agents Chemother 245. Tamaoki J, Kadota J, Takizawa H. Clinical implications of the im-
2000; 44:1796–802. munomodulatory effects of macrolides. Am J Med 2004; 117(Suppl
227. Feldman RB, Rhew DC, Wong JY, Charles RA, Goetz MB. Azithro- 9A):5S-11S.
mycin monotherapy for patients hospitalized with community-ac- 246. Nicholson KG, Aoki FY, Osterhaus AD, et al. Efficacy and safety of

oseltamivir in treatment of acute influenza: a randomised controlled 267. Ziss DR, Stowers A, Feild C. Community-acquired pneumonia: com-
trial. Neuraminidase Inhibitor Flu Treatment Investigator Group. Lan- pliance with centers for Medicare and Medicaid services, national
cet 2000; 355:1845–50. guidelines, and factors associated with outcome. South Med J 2003;
247. Monto AS, Fleming DM, Henry D, et al. Efficacy and safety of the 96:949–59.
neuraminidase inhibitor zanamivir in the treatment of influenza A 268. Ramirez JA, Srinath L, Ahkee S, Huang A, Raff MJ. Early switch from
and B virus infections. J Infect Dis 1999; 180:254–61. intravenous to oral cephalosporins in the treatment of hospitalized
248. Kaiser L, Wat C, Mills T, Mahoney P, Ward P, Hayden F. Impact of patients with community-acquired pneumonia. Arch Intern Med
oseltamivir treatment on influenza-related lower respiratory tract 1995; 155:1273–6.
complications and hospitalizations. Arch Intern Med 2003; 163: 269. Castro-Guardiola A, Viejo-Rodriguez AL, Soler-Simon S, et al. Efficacy
1667–72. and safety of oral and early-switch therapy for community-acquired
249. Jefferson T, Demicheli V, Rivetti D, Jones M, Di Pietrantonj C, Rivetti pneumonia: a randomized controlled trial. Am J Med 2001; 111:
A. Antivirals for influenza in healthy adults: systematic review. Lancet 367–74.
2006; 367:303–13. 270. Ramirez JA, Bordon J. Early switch from intravenous to oral anti-

250. Kaiser L, Hayden FG. Hospitalizing influenza in adults. Curr Clin biotics in hospitalized patients with bacteremic community-acquired
Top Infect Dis 1999; 19:112–34. Streptococcus pneumoniae pneumonia. Arch Intern Med 2001; 161:
251. Gubareva LV, Kaiser L, Hayden FG. Influenza virus neuraminidase 848–50.
inhibitors. Lancet 2000; 355:827–35. 271. Ramirez JA, Vargas S, Ritter GW, et al. Early switch from intravenous
252. High levels of adamantane resistance among influenza A (H3N2) to oral antibiotics and early hospital discharge: a prospective obser-
viruses and interim guidelines for use of antiviral agents—United vational study of 200 consecutive patients with community-acquired
States, 2005–06 influenza season. MMWR Morb Mortal Wkly Rep pneumonia. Arch Intern Med 1999; 159:2449–54.
2006; 55:44–6. 272. Halm EA, Switzer GE, Mittman BS, Walsh MB, Chang CC, Fine MJ.
253. Bright RA, Shay DK, Shu B, Cox NJ, Klimov AI. Adamantane resis- What factors influence physicians’ decisions to switch from intrave-
tance among influenza A viruses isolated early during the 2005–2006 nous to oral antibiotics for community-acquired pneumonia? J Gen
influenza season in the United States. JAMA 2006; 295:891–4. Intern Med 2001; 16:599–605.
254. Kaiser L, Keene ON, Hammond JM, Elliott M, Hayden FG. Impact 273. Halm EA, Fine MJ, Kapoor WN, Singer DE, Marrie TJ, Siu AL.
of zanamivir on antibiotic use for respiratory events following acute Instability on hospital discharge and the risk of adverse outcomes in
influenza in adolescents and adults. Arch Intern Med 2000; 160: patients with pneumonia. Arch Intern Med 2002; 162:1278–84.
3234–40. 274. Halm EA, Fine MJ, Marrie TJ, et al. Time to clinical stability in patients
255. Treanor JJ, Hayden FG, Vrooman PS, et al. Efficacy and safety of the hospitalized with community-acquired pneumonia: implications for
oral neuraminidase inhibitor oseltamivir in treating acute influenza: practice guidelines. JAMA 1998; 279:1452–7.
a randomized controlled trial. US Oral Neuraminidase Study Group. 275. Zervos M, Mandell LA, Vrooman PS, et al. Comparative efficacies
JAMA 2000; 283:1016–24. and tolerabilities of intravenous azithromycin plus ceftriaxone and
256. Hayden FG, Treanor JJ, Fritz RS, et al. Use of the oral neuraminidase intravenous levofloxacin with step-down oral therapy for hospitalized
inhibitor oseltamivir in experimental human influenza: randomized patients with moderate to severe community-acquired pneumonia.
controlled trials for prevention and treatment. JAMA 1999; 282: Treat Respir Med 2004; 3:329–36.
1240–6. 276. Dunbar LM, Wunderink RG, Habib MP, et al. High-dose, short-course
257. Haake DA, Zakowski PC, Haake DL, Bryson YJ. Early treatment with levofloxacin for community-acquired pneumonia: a new treatment
acyclovir for varicella pneumonia in otherwise healthy adults: ret- paradigm. Clin Infect Dis 2003; 37:752–60.
rospective controlled study and review. Rev Infect Dis 1990; 12: 277. Rizzato G, Montemurro L, Fraioli P, et al. Efficacy of a three day
788–98. course of azithromycin in moderately severe community-acquired
258. Hien TT, de Jong M, Farrar J. Avian influenza—a challenge to global pneumonia. Eur Respir J 1995; 8:398–402.
health care structures. N Engl J Med 2004; 351:2363–5. 278. Schonwald S, Skerk V, Petricevic I, Car V, Majerus-Misic L, Gunjaca
259. Chotpitayasunondh T, Ungchusak K, Hanshaoworakul W, et al. Hu- M. Comparison of three-day and five-day courses of azithromycin in
man disease from influenza A (H5N1), Thailand, 2004. Emerg Infect the treatment of atypical pneumonia. Eur J Clin Microbiol Infect Dis
Dis 2005; 11:201–9. 1991; 10:877–80.
260. World Health Organization. Production of pilot lots of inactivated 279. Chastre J, Wolff M, Fagon JY, et al. Comparison of 8 vs 15 days of
influenza vaccines from reassortants derived from avian influenza antibiotic therapy for ventilator-associated pneumonia in adults: a
viruses: interim biosafety assessment. Available at: http://www.who randomized trial. JAMA 2003; 290:2588–98.
.int/csr/resources/publications/influenza/. Accessed 16 January 2007. 280. Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of re-
261. Beigel JH, Farrar J, Han AM, et al. Avian influenza A (H5N1) infection combinant human activated protein C for severe sepsis. N Engl J Med
in humans. N Engl J Med 2005; 353:1374–85. 2001; 344:699–709.
262. Le QM, Kiso M, Someya K, et al. Avian flu: isolation of drug-resistant 281. Ely EW, Laterre PF, Angus DC, et al. Drotrecogin alfa (activated)
H5N1 virus. Nature 2005; 437:1108. administration across clinically important subgroups of patients with
263. de Jong MD, Tran TT, Truong HK, et al. Oseltamivir resistance during severe sepsis. Crit Care Med 2003; 31:12–9.
treatment of influenza A (H5N1) infection. N Engl J Med 2005; 353: 282. Opal SM, Garber GE, LaRosa SP, et al. Systemic host responses in
2667–72. severe sepsis analyzed by causative microorganism and treatment ef-
264. Meehan TP, Fine MJ, Krumholz HM, et al. Quality of care, process, fects of drotrecogin alfa (activated). Clin Infect Dis 2003; 37:50–8.
and outcomes in elderly patients with pneumonia. JAMA 1997; 278: 283. Annane D, Sebille V, Charpentier C, et al. Effect of treatment with
2080–4. low doses of hydrocortisone and fludrocortisone on mortality in pa-
265. Silber SH, Garrett C, Singh R, et al. Early administration of antibiotics tients with septic shock. JAMA 2002; 288:862–71.
does not shorten time to clinical stability in patients with moderate-to- 284. Torres A, Ewig S, El-Ebiary M, Filella X, Xaubet A. Role of gluco-
severe community-acquired pneumonia. Chest 2003; 124:1798–804. corticoids on inflammatory response in nonimmunosuppressed pa-
266. Battleman DS, Callahan M, Thaler HT. Rapid antibiotic delivery and tients with pneumonia: a pilot study. Eur Respir J 1999; 14:218–20.
appropriate antibiotic selection reduce length of hospital stay of pa- 285. Confalonieri M, Urbino R, Potena A, et al. Hydrocortisone infusion
tients with community-acquired pneumonia: link between quality of for severe community-acquired pneumonia: a preliminary random-
care and resource utilization. Arch Intern Med 2002; 162:682. ized study. Am J Respir Crit Care Med 2005; 171:242–8.

286. Marik P, Kraus P, Bribante J, et al. Hydrocortisone and tumour ne- 306. Butler JC, Breiman RF, Campbell JF, Lipman HB, Broome CV, Fack-
crosis factor in severe community acquired pneumonia. Chest lam RR. Pneumococcal polysaccharide vaccine efficacy: an evaluation
1993; 104:389–92. of current recommendations. JAMA 1993; 270:1826–31.
287. Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin 307. Sims RV, Steinmann WC, McConville JH, King LR, Zwick WC,
therapy in the critically ill patients. N Engl J Med 2001; 345:1359–67. Schwartz JS. The clinical effectiveness of pneumococcal vaccine in the
288. Brochard L, Mancebo J, Wysocki M, et al. Noninvasive ventilation elderly. Ann Intern Med 1988; 108:653–7.
for acute exacerbations of chronic obstructive pulmonary disease. N 308. Shapiro ED, Berg AT, Austrian R, et al. The protective efficacy of
Engl J Med 1995; 333:817–22. polyvalent pneumococcal polysaccharide vaccine. N Engl J Med
289. Ferrer M, Esquinas A, Leon M, Gonzalez G, Alarcon A, Torres A. 1991; 325:1453–60.
Noninvasive ventilation in severe hypoxemic respiratory failure: a 309. Farr BM, Johnston BL, Cobb DK, et al. Preventing pneumococcal
randomized clinical trial. Am J Respir Crit Care Med 2003; 168: bacteremia in patients at risk: results of a matched case-control study.
1438–44. Arch Intern Med 1995; 155:2336–40.
290. Antonelli M, Conti G, Moro ML, et al. Predictors of failure of non- 310. Jackson LA, Neuzil KM, Yu O, et al. Effectiveness of pneumococcal

invasive positive pressure ventilation in patients with acute hypoxemic polysaccharide vaccine in older adults. N Engl J Med 2003; 348:
respiratory failure: a multi-center study. Intensive Care Med 2001; 1747–55.
27:1718–28. 311. Sisk JE, Moskowitz AJ, Whang W, et al. Cost-effectiveness of vacci-
291. Ventilation with lower tidal volumes as compared with traditional nation against pneumococcal bacteremia among elderly people. JAMA
tidal volumes for acute lung injury and the acute respiratory distress 1997; 278:1333–9.
syndrome. The Acute Respiratory Distress Syndrome Network. N Engl 312. Sisk JE, Whang W, Butler JC, Sneller VP, Whitney CG. Cost-effec-
J Med 2000; 342:1301–8. tiveness of vaccination against invasive pneumococcal disease among
292. Eisner MD, Thompson T, Hudson LD, et al. Efficacy of low tidal people 50 through 64 years of age: role of comorbid conditions and
volume ventilation in patients with different clinical risk factors for race. Ann Intern Med 2003; 138:960–8.
acute lung injury and the acute respiratory distress syndrome. Am J 313. Jackson LA, Benson P, Sneller VP, et al. Safety of revaccination with
Respir Crit Care Med 2001; 164:231–6. pneumococcal polysaccharide vaccine. JAMA 1999; 281:243–8.
293. Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign 314. Whitney CG, Farley MM, Hadler J, et al. Decline in invasive pneu-
guidelines for management of severe sepsis and septic shock. Crit mococcal disease after the introduction of protein-polysaccharide
Care Med 2004; 32:858–73. conjugate vaccine. N Engl J Med 2003; 348:1737–46.
294. Menéndez R, Torres A, Rodrı́guez de Castro F, et al. Reaching stability 315. Flannery B, Schrag S, Bennett NM, et al. Impact of childhood vac-
in community-acquired pneumonia: the effects of the severity of dis- cination on racial disparities in invasive Streptococcus pneumoniae
ease, treatment, and the characteristics of patients. Clin Infect Dis infections. JAMA 2004; 291:2197–203.
2004; 39:1783–90. 316. Lexau CA, Lynfield R, Danila R, et al. Changing epidemiology of
295. Almirall J, Bolibar I, Vidal J, et al. Epidemiology of community- invasive pneumococcal disease among older adults in the era of pe-
acquired pneumonia in adults: a population-based study. Eur Respir diatric pneumococcal conjugate vaccine. JAMA 2005; 294:2043–51.
J 2000; 15:757–63. 317. Gross PA, Hermogenes AW, Sacks HS, Lau J, Levandowski RA. The
296. Celis R, Torres A, Gatell JM, Almela M, Rodriguez-Roisin R, Agusti- efficacy of influenza vaccine in elderly persons: a meta-analysis and
Vidal A. Nosocomial pneumonia: a multivariate analysis of risk and review of the literature. Ann Intern Med 1995; 123:518–27.
prognosis. Chest 1988; 93:318–24. 318. Jefferson T, Rivetti D, Rivetti A, Rudin M, Di Pietrantonj C, Demicheli
297. Daifuku R, Movahhed H, Fotheringham N, Bear MB, Nelson S. Time V. Efficacy and effectiveness of influenza vaccines in elderly people:
to resolution of morbidity: an endpoint for assessing the clinical cure a systematic review. Lancet 2005; 366:1165–74.
of community-acquired pneumonia. Respir Med 1996; 90:587–92. 319. Nichol KL, Nordin J, Mullooly J, Lask R, Fillbrandt K, Iwane M.
298. Mittl RL Jr, Schwab RJ, Duchin JS, Goin JE, Albeida SM, Miller WT. Influenza vaccination and reduction in hospitalizations for cardiac
Radiographic resolution of community-acquired pneumonia. Am J disease and stroke among the elderly. N Engl J Med 2003; 348:1322–32.
Respir Crit Care Med 1994; 149:630–5. 320. Carman WF, Elder AG, Wallace LA, et al. Effects of influenza vac-
299. El Solh AA, Pietrantoni C, Bhat A, et al. Microbiology of severe cination of health-care workers on mortality of elderly people in long-
aspiration pneumonia in institutionalized elderly. Am J Respir Crit term care: a randomised controlled trial. Lancet 2000; 355:93–7.
Care Med 2003; 167:1650–4. 321. Potter J, Stott DJ, Roberts MA, et al. Influenza vaccination of health
300. Ortqvist A, Kalin M, Lejdeborn L, Lundberg B. Diagnostic fiberoptic care workers in long-term-care hospitals reduces the mortality of
bronchoscopy and protected brush culture in patients with com- elderly patients. J Infect Dis 1997; 175:1–6.
munity-acquired pneumonia. Chest 1990; 97:576–82. 322. Public health and aging: influenza vaccination coverage among adults
301. Fagon JY, Chastre J, Wolff M, et al. Invasive and noninvasive strategies aged ⭓50 years and pneumococcal vaccination coverage among adults
for management of suspected ventilator-associated pneumonia: a ran- aged ⭓65 years—United States, 2002. MMWR Morb Mortal Wkly
domized trial. Ann Intern Med 2000; 132:621–30. Rep 2003; 52:987–92.
302. Ruiz M, Torres A, Ewig S, et al. Noninvasive versus invasive microbial 323. Influenza and pneumococcal vaccination coverage among persons
investigation in ventilator-associated pneumonia: evaluation of out- aged ⭓65 years and persons aged 18–64 years with diabetes or
come. Am J Respir Crit Care Med 2000; 162:119–25. asthma—United States, 2003. MMWR Morb Mortal Wkly Rep
303. Sabria M, Pedro-Botet ML, Gomez J, et al. Fluoroquinolones vs mac- 2004; 53:1007–12.
rolides in the treatment of Legionnaires disease. Chest 2005; 128: 324. From the Centers for Disease Control and Prevention. Facilitating
1401–5. influenza and pneumococcal vaccination through standing orders
304. Harper SA, Fukuda K, Uyeki TM, Cox NJ, Bridges CB. Prevention programs. JAMA 2003; 289:1238.
and control of influenza: recommendations of the Advisory Com- 325. Monto AS, Pichichero ME, Blanckenberg SJ, et al. Zanamivir pro-
mittee on Immunization Practices (ACIP). MMWR Recomm Rep phylaxis: an effective strategy for the prevention of influenza types A
2005; 54:1–40. and B within households. J Infect Dis 2002; 186:1582–8.
305. Bridges CB, Harper SA, Fukuda K, Uyeki TM, Cox NJ, Singleton JA. 326. Hayden FG, Belshe R, Villanueva C, et al. Management of influenza
Prevention and control of influenza: recommendations of the Ad- in households: a prospective, randomized comparison of oseltamivir
visory Committee on Immunization Practices (ACIP). MMWR Re- treatment with or without postexposure prophylaxis. J Infect Dis
comm Rep 2003; 52(RR-8):1–34. 2004; 189:440–9.

327. Ward JI, Cherry JD, Chang SJ, et al. Efficacy of an acellular pertussis 332. Hyde TB, Gilbert M, Schwartz SB, et al. Azithromycin prophylaxis
vaccine among adolescents and adults. N Engl J Med 2005; 353: during a hospital outbreak of Mycoplasma pneumoniae pneumonia.
1555–63. J Infect Dis 2001; 183:907–12.
328. Nuorti JP, Butler JC, Farley MM, et al. Cigarette smoking and invasive 333. Centers for Disease Control and Prevention (CDC). Respiratory hy-
pneumococcal disease. Active Bacterial Core Surveillance Team. N giene/cough etiquette in health-care settings. Available at: http://
Engl J Med 2000; 342:681–9. www.cdc.gov/flu/professionals/infectioncontrol/resphygiene.htm. Ac-
329. Marston BJ, Lipman HB, Breiman RF. Surveillance for Legionnaires’ cessed 16 January 2007.
disease: risk factors for morbidity and mortality. Arch Intern Med
334. Experiences with influenza-like illness and attitudes regarding influ-
1994; 154:2417–22.
enza prevention—United States, 2003–04 influenza season. MMWR
330. McDonald LC, Simor AE, Su IJ, et al. SARS in healthcare facilities,
Toronto and Taiwan. Emerg Infect Dis 2004; 10:777–81. Morb Mortal Wkly Rep 2004; 53:1156–8.
331. Nuorti JP, Butler JC, Crutcher JM, et al. An outbreak of multidrug- 335. Garner JS. Guideline for isolation precautions in hospitals. The Hos-
resistant pneumococcal pneumonia and bacteremia among unvac- pital Infection Control Practices Advisory Committee. Infect Control

cinated nursing home residents. N Engl J Med 1998; 338:1861–8. Hosp Epidemiol 1996; 17:53–80.

Idsa Cap

Uploaded by

Copyright:

Available Formats

Idsa Cap

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Idsa Cap

Uploaded by

Copyright:

Available Formats

SUPPLEMENT ARTICLE

Infectious Diseases Society of America/American

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

EXECUTIVE SUMMARY priate starting point for consultation by specialists.

1. Locally adapted guidelines should be imple-

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S27

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S28 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

Antibiotic Treatment Inpatient, non-ICU treatment

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S29

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

Pathogen-directed therapy. Duration of antibiotic therapy.

S30 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S31

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S32 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S33

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S34 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S35

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S36 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S37

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S38 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

The need for diagnostic testing to determine the etiology of

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S39

Blood Sputum Legionella Pneumococcal

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

NOTE. NA, not applicable; UAT, urinary antigen test.

S40 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S41

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S42 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S43

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S44 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S45

Condition Commonly encountered pathogen(s)

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S46 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S47

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S48 • CID 2007:44 (Suppl 2) • Mandell et al.

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

IDSA/ATS Guidelines for CAP in Adults • CID 2007:44 (Suppl 2) • S49

Downloaded from https://academic.oup.com/cid/article-abstract/44/Supplement_2/S27/372079 by guest on 15 October 2019

S50 • CID 2007:44 (Suppl 2) • Mandell et al.