american thoracic society documents - samin...

AMERICAN THORACIC SOCIETYDOCUMENTS

Diagnosis and Treatment of Adults with Community-acquiredPneumoniaAn Official Clinical Practice Guideline of the American Thoracic Society andInfectious Diseases Society of AmericaJoshua P. Metlay*, Grant W. Waterer*, Ann C. Long, Antonio Anzueto, Jan Brozek, Kristina Crothers, Laura A. Cooley,Nathan C. Dean, Michael J. Fine, Scott A. Flanders, Marie R. Griffin, Mark L. Metersky, Daniel M. Musher,Marcos I. Restrepo, and Cynthia G. Whitney; on behalf of the American Thoracic Society and Infectious DiseasesSociety of America

THIS OFFICIAL CLINICAL PRACTICE GUIDELINE WAS APPROVED BY THE AMERICAN THORACIC SOCIETY MAY 2019 AND THE INFECTIOUS DISEASES SOCIETY OF AMERICAAUGUST 2019

Background: This document provides evidence-based clinicalpractice guidelines on the management of adult patients withcommunity-acquired pneumonia.

Methods: A multidisciplinary panel conducted pragmaticsystematic reviews of the relevant research and applied Grading ofRecommendations, Assessment, Development, and Evaluationmethodology for clinical recommendations.

Results:The panel addressed 16 specific areas for recommendationsspanning questions of diagnostic testing, determination of site ofcare, selection of initial empiric antibiotic therapy, and subsequent

management decisions. Although some recommendations remainunchanged from the 2007 guideline, the availability of results fromnew therapeutic trials and epidemiological investigations led torevised recommendations for empiric treatment strategies andadditional management decisions.

Conclusions: The panel formulated and provided the rationale forrecommendations on selected diagnostic and treatment strategiesfor adult patients with community-acquired pneumonia.

Keywords: community-acquired pneumonia; pneumonia; patientmanagement

ContentsOverviewIntroductionMethodsRecommendations

Question 1: In Adults with CAP,Should Gram Stain and Cultureof Lower Respiratory SecretionsBe Obtained at the Time ofDiagnosis?

Question 2: In Adults with CAP,Should Blood Cultures BeObtained at the Time of Diagnosis?

Question 3: In Adults with CAP,Should Legionella and

*Co–first authors.

Endorsed by the Society of Infectious Disease Pharmacists July 2019.

ORCID IDs: 0000-0003-2259-6282 (J.P.M.); 0000-0002-7222-8018 (G.W.W.); 0000-0002-7007-588X (A.A.); 0000-0002-3122-0773 (J.B.);0000-0001-9702-0371 (K.C.); 0000-0002-5127-3442 (L.A.C.); 0000-0002-1996-0533 (N.C.D.); 0000-0003-3470-9846 (M.J.F.);0000-0002-8634-4909 (S.A.F.); 0000-0001-7114-7614 (M.R.G.); 0000-0003-1968-1400 (M.L.M.); 0000-0002-7571-066X (D.M.M.);0000-0001-9107-3405 (M.I.R.); 0000-0002-1056-3216 (C.G.W.).

Supported by the American Thoracic Society and Infectious Diseases Society of America.

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the U.S. CDC.

An Executive Summary of this document is available at http://www.atsjournals.org/doi/suppl/10.1164/rccm.201908-1581ST.

You may print one copy of this document at no charge. However, if you require more than one copy, you must place a reprint order. Domestic reprint orders:[email protected]; international reprint orders: [email protected].

This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.

Am J Respir Crit Care Med Vol 200, Iss 7, pp e45–e67, Oct 1, 2019

Copyright © 2019 by the American Thoracic SocietyDOI: 10.1164/rccm.201908-1581ST

Internet address: www.atsjournals.org

American Thoracic Society Documents e45

http://orcid.org/0000-0003-2259-6282http://orcid.org/0000-0002-7222-8018http://orcid.org/0000-0002-7007-588Xhttp://orcid.org/0000-0002-3122-0773http://orcid.org/0000-0001-9702-0371http://orcid.org/0000-0002-5127-3442http://orcid.org/0000-0002-1996-0533http://orcid.org/0000-0003-3470-9846http://orcid.org/0000-0002-8634-4909http://orcid.org/0000-0001-7114-7614http://orcid.org/0000-0003-1968-1400http://orcid.org/0000-0002-7571-066Xhttp://orcid.org/0000-0001-9107-3405http://orcid.org/0000-0002-1056-3216http://www.atsjournals.org/doi/suppl/10.1164/rccm.201908-1581STmailto:[email protected]:[email protected]://www.atsjournals.orghttp://dx.doi.org/10.1164/rccm.201908-1581SThttp://www.atsjournals.org

Pneumococcal Urinary AntigenTesting Be Performed at theTime of Diagnosis?

Question 4: In Adults with CAP,Should a Respiratory Sample BeTested for Influenza Virus at theTime of Diagnosis?

Question 5: In Adults with CAP,Should Serum Procalcitonin plusClinical Judgment versusClinical Judgment Alone BeUsed to Withhold Initiation ofAntibiotic Treatment?

Question 6: Should a ClinicalPrediction Rule for Prognosisplus Clinical Judgment versusClinical Judgment Alone BeUsed to Determine Inpatientversus Outpatient TreatmentLocation for Adults with CAP?

Question 7: Should a ClinicalPrediction Rule for Prognosisplus Clinical Judgment versusClinical Judgment Alone BeUsed to Determine InpatientGeneral Medical versus Higher

Levels of Inpatient TreatmentIntensity (ICU, Step-Down, orTelemetry Unit) for Adults withCAP?

Question 8: In the OutpatientSetting, Which Antibiotics AreRecommended for EmpiricTreatment of CAP in Adults?

Question 9: In the InpatientSetting, Which AntibioticRegimens Are Recommendedfor Empiric Treatment ofCAP in Adults without RiskFactors for MRSA andP. aeruginosa?

Question 10: In the InpatientSetting, Should Patientswith Suspected AspirationPneumonia Receive AdditionalAnaerobic Coverage beyondStandard Empiric Treatment forCAP?

Question 11: In the InpatientSetting, Should Adults with CAPand Risk Factors for MRSA orP. aeruginosa Be Treated with

Extended-Spectrum AntibioticTherapy Instead of StandardCAP Regimens?

Question 12: In the InpatientSetting, Should Adultswith CAP Be Treated withCorticosteroids?

Question 13: In Adults with CAPWho Test Positive for Influenza,Should the Treatment RegimenInclude Antiviral Therapy?

Question 14: In Adults withCAP Who Test Positive forInfluenza, Should the TreatmentRegimen Include AntibacterialTherapy?

Question 15: In Outpatient andInpatient Adults with CAP WhoAre Improving, What Is theAppropriate Duration ofAntibiotic Treatment?

Question 16: In Adults with CAPWho Are Improving, ShouldFollow-up Chest Imaging BeObtained?

Conclusions

Overview

In the more than 10 years since the lastAmerican Thoracic Society(ATS)/InfectiousDiseases Society of America (IDSA)community-acquired pneumonia (CAP)guideline (1), there have been changes inthe process for guideline development, aswell as generation of new clinical data. ATSand IDSA agreed on moving from thenarrative style of previous documents to theGrading of Recommendations Assessment,Development, and Evaluation (GRADE)format. We thus developed this updatedCAP guideline as a series of questionsanswered from available evidence in an “isoption A better than option B” formatusing the Patient or Population,Intervention, Comparison, Outcome(PICO) framework (2).

Given the expansion in informationrelated to the diagnostic, therapeutic, andmanagement decisions for the care ofpatients with CAP, we have purposelynarrowed the scope of this guideline toaddress decisions from the time of clinicaldiagnosis of pneumonia (i.e., signs andsymptoms of pneumonia with radiographicconfirmation) to completion ofantimicrobial therapy and follow-up chest

imaging. The document does not addresseither the initial clinical diagnostic criteriaor prevention of pneumonia.

CAP is an extraordinarilyheterogeneous illness, both in the rangeof responsible pathogens and the hostresponse. Thus, the PICO questions weidentified for this guideline do not representthe full range of relevant questions about themanagement of CAP but encompass a set ofcore questions identified as high priority bythe panel. In addition, although eachquestion was addressed using systematicreviews of available high-quality studies,the evidence base was often insufficient,emphasizing the continued importance ofclinical judgment and experience in treatingpatients with this illness and the need forcontinued research.

Introduction

This guideline addresses the clinical entity ofpneumonia that is acquired outside of thehospital setting. Although we recognize thatCAP is frequently diagnosed without the useof a chest radiograph, especially in theambulatory setting, we have focused onstudies that used radiographic criteria for

defining CAP, given the known inaccuracyof clinical signs and symptoms alone forCAP diagnosis (3). This guideline focuseson patients in the United States who havenot recently completed foreign travel,especially to regions with emergingrespiratory pathogens. This guideline alsofocuses on adults who do not have animmunocompromising condition, such asinherited or acquired immune deficiency ordrug-induced neutropenia, includingpatients actively receiving cancerchemotherapy, patients infected with HIVwith suppressed CD4 counts, and solidorgan or bone marrow transplantrecipients.

Antibiotic recommendations for theempiric treatment of CAP are based onselecting agents effective against the majortreatable bacterial causes of CAP.Traditionally, these bacterial pathogensinclude Streptococcus pneumoniae,Haemophilus influenzae, Mycoplasmapneumoniae, Staphylococcus aureus,Legionella species, Chlamydia pneumoniae,and Moraxella catarrhalis. The microbialetiology of CAP is changing, particularlywith the widespread introduction of thepneumococcal conjugate vaccine, and thereis increased recognition of the role of viral

AMERICAN THORACIC SOCIETY DOCUMENTS

e46 American Journal of Respiratory and Critical Care Medicine Volume 200 Number 7 | October 1 2019

pathogens. The online supplement containsa more detailed discussion of CAPmicrobiology. As bacterial pathogens oftencoexist with viruses and there is no currentdiagnostic test accurate enough or fastenough to determine that CAP is due solelyto a virus at the time of presentation (seebelow), our recommendations are toinitially treat empirically for possiblebacterial infection or coinfection. Inaddition, the emergence of multidrug-resistant pathogens, including methicillin-resistant S. aureus (MRSA) andPseudomonas aeruginosa, requires separaterecommendations when the risk ofeach of these pathogens is elevated. Weacknowledge that other multidrug-resistantEnterobacteriaceae can cause CAP,including organisms producing extended-spectrum b-lactamase, but we do notdiscuss them separately because they aremuch less common and are effectivelycovered by the strategies presented forP. aeruginosa. Therefore, throughout thisdocument when discussing P. aeruginosawe are also referring to other similarmultiresistant gram-negative bacteria.

We have maintained the conventionof separate recommendations on thebasis of the severity of illness. Althoughhistorically site of care (outpatient,inpatient general ward, or ICU) has servedas a severity surrogate, decisions about siteof care may be based on considerationsother than severity and can vary widelybetween hospitals and practice sites.We have therefore chosen to use theIDSA/ATS CAP severity criteria that havebeen validated and define severe CAP aspresent in patients with either one majorcriterion or three or more minor criteria.(Table 1)

This guideline reaffirms manyrecommendations from the 2007 statement.However, new evidence and a new processhave led to significant changes, which aresummarized in Table 2.

Methods

The guideline development methodologyand how conflict of interest was managedare presented in the online supplement. Inbrief, the list of PICO questions was finalizedbased on a prioritization of the mostimportant management decisions balancedagainst the decision to reduce the overalllength of the document and total number of

recommendations to maximize readabilityand usability. We followed the GRADEstandards for evaluating the evidence foreach PICO and assigned a quality ofevidence rating of high, moderate, low, orvery low. On the basis of the quality ofevidence, recommendations were assignedas strong or conditional. In some cases,strong recommendations were made in thesetting of low or very low quality of evidencein accordance with the GRADE rules forwhen such recommendations are allowable(e.g., when the consequences of therecommendation were high, such aspreventing harm or saving life). In all othercases, recommendations that were based onlow or very low quality of evidence and notbelieved to represent standards of care werelabeled as conditional recommendations.Statements in favor of strongrecommendations begin with the words“We recommend . . .”; statements in favorof conditional recommendations beginwith the words “We suggest . . . .”Although we specified pairwise PICOquestions for all antibiotic options in theoutpatient and inpatient settings, wesummarized the recommendations usinglists of treatment options, in no preferredorder, rather than retain the PICO formatfor this section.

Recommendations

Question 1: In Adults with CAP,Should Gram Stain and Culture ofLower Respiratory Secretions BeObtained at the Time of Diagnosis?

Recommendation. We recommend notobtaining sputum Gram stain and cultureroutinely in adults with CAPmanaged in theoutpatient setting (strong recommendation,very low quality of evidence).

We recommend obtaining pretreatmentGram stain and culture of respiratorysecretions in adults with CAP managed inthe hospital setting who:

1. are classified as severe CAP (see Table 1),especially if they are intubated (strongrecommendation, very low quality ofevidence); or

2.a. are being empirically treated for

MRSA or P. aeruginosa (strongrecommendation, very low qualityof evidence); or

b. were previously infected with MRSAor P. aeruginosa, especially thosewith prior respiratory tract infection(conditional recommendation, verylow quality of evidence); or

c. were hospitalized and receivedparenteral antibiotics, whether during thehospitalization event or not, in the last90 days (conditional recommendation,very low quality of evidence).

Summary of the evidence. Argumentsfor trying to determine the etiology of CAPare that 1) a resistant pathogen may beidentified; 2) therapy may be narrowed; 3)some pathogens, such as Legionella, havepublic health implications; 4) therapy maybe adjusted when patients fail initialtherapy; and 5) the constantly changingepidemiology of CAP requires ongoingevaluation.

These arguments stand in contrast tothe lack of high-quality evidencedemonstrating that routine diagnostictesting improves individual patientoutcomes. Studies that specifically evaluatedthe use of sputum Gram stain and culturealone (4–7), or in combination with othermicrobiological testing (8–11), also did notdemonstrate better patient outcomes.

The overall poor yield of sputumevaluation for detecting organisms causingCAP limits its impact on management andpatient outcomes. Obtaining a valid sputumspecimen can be challenging because ofpatient-related characteristics (12–17).Performance characteristics of testing alsovary by organism, receipt of priorantibiotics, and setting. For example, inpatients with bacteremic pneumococcalpneumonia who have not receivedantibiotics, microscopic examination andculture of a good-quality sputum sampledetects pneumococci in 86% of cases (18).

Rationale for the recommendation. Inbalancing the lack of evidence supportingroutine sputum culture with the desire forimproved antimicrobial stewardship, thecommittee voted to continue the stance ofprevious guidelines in recommendingneither for nor against routinely obtainingsputum Gram stain and culture in all adultswith CAP managed in the hospital setting.Whether to culture patients or not should bedetermined by individual clinicians on thebasis of clinical presentation, localetiological considerations, and localantimicrobial stewardship processes.

The committee identified twosituations in which we recommend sputum



Gram stain and culture: in hospitalizedpatients with severe CAP, and when strongrisk factors for MRSA and P. aeruginosa areidentified, unless local etiological data havealready shown these pathogens are veryinfrequently identified in patients with CAP.Patients who have severe CAP requiringintubation should have lower respiratory tractsamples, such as endotracheal aspirates, sentfor Gram stain and culture promptly afterintubation, particularly as these patients maybe more likely to have pneumonia due toMRSA or P. aeruginosa, and endotrachealaspirates have a better yield of microbiologicalorganisms than sputum culture (19).

We recommend obtaining sputum forGram stain and culture in situations whenrisk factors for MRSA or P. aeruginosa arepresent, both when initial empiric therapyis expanded to cover these pathogens andwhen it is not expanded. In the former case,negative microbiological test results may beused to deescalate therapy, and in the lattercase, positive microbiological test resultsmay be used to adjust therapy. As discussedlater, although there are numerous studiesidentifying individual risk factors for MRSAand P. aeruginosa, many of theseassociations are weak and vary across sites.

The most consistently strong risk factor toconsider is prior infection with eitherMRSA or P. aeruginosa. In addition,hospitalization and treatment withparenteral antibiotics in the last 90 days isassociated with an increased risk of thesepathogens, and so we recommendsputum culture in this situation. Theserecommendations are not based on high-grade evidence but reflect the committee’sdesire to improve antibiotic use as well asimprove clinicians’ understanding of theirlocal pathogen prevalences and resistancepatterns, which we believe are key toselecting appropriate empiric antibiotictherapy.

Research needed in this area. Rapid,cost-effective, sensitive, and specificdiagnostic tests to identify organismscausing CAP have potential to improveroutine care by supporting the use oftargeted therapy, especially when thereare risk factors for antibiotic-resistantpathogens. All new diagnostic tests shouldbe assessed in high-quality research studiesthat address the impact of testing strategieson treatment decisions and patientoutcomes.

Question 2: In Adults with CAP,Should Blood Cultures Be Obtained atthe Time of Diagnosis?

Recommendation. We recommend notobtaining blood cultures in adults with CAPmanaged in the outpatient setting (strongrecommendation, very low quality ofevidence).

We suggest not routinely obtaining bloodcultures in adults with CAP managed in thehospital setting (conditional recommendation,very low quality of evidence).

We recommend obtainingpretreatment blood cultures in adults withCAP managed in the hospital setting who:

1. are classified as severe CAP (see Table 1)(strong recommendation, very lowquality of evidence); or

2.a. are being empirically treated for

MRSA or P. aeruginosa (strongrecommendation, very low quality ofevidence); or

b. were previously infected with MRSAor P. aeruginosa, especially thosewith prior respiratory tract infection(conditional recommendation, verylow quality of evidence); or

c. were hospitalized and receivedparenteral antibiotics, whether duringthe hospitalization event or not,in the last 90 days (conditionalrecommendation, very low quality ofevidence).

Summary of the evidence. There are nohigh-quality studies that specificallycompared patient outcomes with andwithout blood culture testing. One largeobservational study found lower mortalityfor hospitalized patients associated withobtaining blood cultures at the time ofadmission (20). Three subsequent (smaller)observational studies found similarassociations between in-hospital mortalityand having blood cultures within 24 hoursof admission, but the results were notstatistically significant (8, 21, 22).

The yield of blood cultures in mostseries of adults with nonsevere CAP is low,ranging from 2% (outpatients) to 9%(inpatients) (14, 21, 23, 24); furthermore,blood cultures rarely result in anappropriate change in empiric therapy (25),and blood specimens that include skincontaminants can generate false-positivetest results. Growth of organisms such ascoagulase-negative staphylococci, which arenot recognized as CAP pathogens (26), maylead to inappropriate antimicrobial use thatincreases the risk for adverse drug effects.A study of adults hospitalized with CAPfound blood cultures were associated witha significant increase in length of stay andduration of antibiotic therapy (27). Given theobservational nature of these studies, it isunknown whether the associations found withblood cultures and patient outcomes werecausal or due to unmeasured confoundingfactors, including severity of illness.

Rationale for the recommendation.Although additional diagnostic informationcould improve the quality of treatmentdecisions, support for routine collection ofblood cultures is reduced by the low quality ofstudies demonstrating clinical benefit. Routinelyobtaining blood cultures may generate false-positive results that lead to unnecessaryantibiotic use and increased length of stay.

In severe CAP, delay in covering less-common pathogens can have seriousconsequences. Therefore, the potential benefitof blood cultures is much larger when resultscan be returned within 24 to 48 hours.

The rationale for the recommendationfor blood cultures in the setting of riskfactors for MRSA and P. aeruginosa is thesame as for sputum culture.

Table 1. 2007 Infectious DiseasesSociety of America/American ThoracicSociety Criteria for Defining SevereCommunity-acquired Pneumonia

Validated definition includes either onemajor criterion or three or moreminor criteria

Minor criteriaRespiratory rate>30 breaths/minPaO2/FIO2 ratio< 250Multilobar infiltratesConfusion/disorientationUremia (blood urea nitrogenlevel>20 mg/dl)

Leukopenia* (white blood cellcount,4,000 cells/ml)

Thrombocytopenia (plateletcount,100,000/ml)

Hypothermia (core temperature,368C)Hypotension requiring aggressive fluidresuscitation

Major criteriaSeptic shock with need forvasopressors

Respiratory failure requiring mechanicalventilation

*Due to infection alone (i.e., not chemotherapyinduced).



Question 3: In Adults with CAP,Should Legionella and PneumococcalUrinary Antigen Testing Be Performedat the Time of Diagnosis?

Recommendation. We suggest notroutinely testing urine for pneumococcalantigen in adults with CAP (conditionalrecommendation, low quality ofevidence), except in adults with severeCAP (conditional recommendation, lowquality of evidence).

We suggest not routinely testing urinefor Legionella antigen in adults with CAP(conditional recommendation, low qualityof evidence), except

1. in cases where indicated byepidemiological factors, such asassociation with a Legionella outbreakor recent travel (conditionalrecommendation, low quality ofevidence); or

2. in adults with severe CAP (see Table 1)(conditional recommendation, lowquality of evidence).

We suggest testing for Legionellaurinary antigen and collecting lowerrespiratory tract secretions for Legionellaculture on selective media or Legionellanucleic acid amplification testing inadults with severe CAP (conditionalrecommendation, low quality of evidence).

Summary of the evidence. Falguera andcolleagues (28) randomized 177 patients topathogen-directed treatment (targetedtreatment) on the basis of results of urinaryantigen testing for S. pneumoniae andLegionella versus empirical guideline-directed treatment. Of the 88 patients in thetargeted treatment arm, 25% had a positiveurinary antigen test and received pathogen-directed therapy. There were no statisticaldifferences in death, clinical relapse, ICUadmission, length of hospitalization, or

length of antibiotic treatment (28). Asecond trial of 262 patients includeda broader range of microbiological testing(sputum and blood cultures) and onlyLegionella urinary antigen testing, butpatients receiving pathogen-directedtherapy had similar clinical outcomes topatients receiving empirical, guideline-directed therapy, including mortality, ratesof clinical failure, and length ofhospitalization (10).

One observational study evaluated costand antibiotic selection in patients duringtwo time periods, with and withoutpneumococcal urinary antigen testing, butfound no differences during the twotime periods (29). In contrast, otherobservational studies that have evaluatedthe impact of prior CAP guidelineconcordance (including initial diagnostictesting with urinary antigen tests andblood cultures, along with site of care

Table 2. Differences between the 2019 and 2007 American Thoracic Society/Infectious Diseases Society of AmericaCommunity-acquired Pneumonia Guidelines

Recommendation 2007 ATS/IDSA Guideline 2019 ATS/IDSA Guideline

Sputum culture Primarily recommended in patients withsevere disease

Now recommended in patients with severedisease as well as in all inpatientsempirically treated for MRSA orPseudomonas aeruginosa

Blood culture Primarily recommended in patients withsevere disease

Now recommended in patients with severedisease as well as in all inpatientsempirically treated for MRSA or P.aeruginosa

Macrolide monotherapy Strong recommendation for outpatients Conditional recommendation for outpatientsbased on resistance levels

Use of procalcitonin Not covered Not recommended to determine need forinitial antibacterial therapy

Use of corticosteroids Not covered Recommended not to use. May be consideredin patients with refractory septic shock

Use of healthcare-associated pneumoniacategory

Accepted as introduced in the 2005ATS/IDSA hospital-acquired andventilator-associated pneumoniaguidelines

Recommend abandoning this categorization.Emphasis on local epidemiology andvalidated risk factors to determine need forMRSA or P. aeruginosa coverage.Increased emphasis on deescalation oftreatment if cultures are negative

Standard empiric therapy for severe CAP b-Lactam/macrolide andb-lactam/fluoroquinolone combinationsgiven equal weighting

Both accepted but stronger evidence in favorof b-lactam/macrolide combination

Routine use of follow-up chest imaging Not addressed Recommended not to obtain. Patients may beeligible for lung cancer screening, whichshould be performed as clinically indicated

Definition of abbreviations: ATS=American Thoracic Society; CAP=community-acquired pneumonia; IDSA= Infectious Diseases Society of America;MRSA=methicillin-resistant Staphylococcus aureus.



stratification and guideline-concordanttherapy) have reported reduced mortalityfor patients receiving prior CAP guideline-concordant care, including diagnostictesting. Costantini and colleagues reporteda 57% statistically significant reduced oddsof in-hospital mortality for patientsreceiving pneumococcal and Legionellaurinary antigen testing compared withpatients not tested, adjusting for baselinedemographic and clinical differences (27).Uematsu and colleagues reported 25%reduced odds of 30-day mortality in patientsreceiving urinary antigen tests but no impacton length of hospitalizations (7). However,neither study distinguished whether themortality benefits attributed to testing werea direct consequence of the test results ora marker of other improved processes of care.

Rationale for the recommendation.Randomized trials have failed to identifya benefit for urinary antigen testing for S.pneumoniae and Legionella. Concern hasalso been raised that narrowing therapy inresponse to positive urinary antigen testscould lead to increased risk of clinicalrelapse (28). In large observational studies,these diagnostic tests have been associatedwith reduction in mortality; therefore, werecommend testing in patients with severe

disease. An increase in Legionella infectionsin the United States in the past decadehighlights the importance of this diagnosisespecially among severely ill patients,particularly in the setting of potentialoutbreaks due to a common source, althoughmost cases are not associated with a knownoutbreak and remain sporadic (30, 31).

Research needed in this area. Newernucleic acid amplification systems forsputum, urine, and blood are being developedand require rigorous testing to assess theimpact on treatment decisions and clinicaloutcomes for patients with CAP, as well asthe public health benefit in terms ofprevention of additional cases and informingprimary prevention strategies. In particular,we acknowledge the emergence of rapid, low-cost genomic sequence detection assays thathave the potential to greatly improvepathogen-directed therapy and therebyimprove antimicrobial stewardship.

Question 4: In Adults with CAP,Should a Respiratory Sample BeTested for Influenza Virus at the Timeof Diagnosis?

Recommendation. When influenza virusesare circulating in the community, we

recommend testing for influenza witha rapid influenza molecular assay(i.e., influenza nucleic acid amplificationtest), which is preferred over a rapidinfluenza diagnostic test (i.e., antigen test)(strong recommendation, moderate qualityof evidence).

Summary of the evidence. Rapidinfluenza tests have become increasinglyavailable, moving from earlier antigen-baseddetection tests to nucleic acid amplificationtests. We were unable to identify any studiesthat evaluated the impact of influenzatesting on outcomes in adults with CAP.In contrast, a substantial literature hasevaluated the importance of influenzatesting in the general population, specificallyamong patients with influenza-like illness(32). Our recommendations for influenzatesting in adults with CAP are consistentwith testing recommendations for thebroader population of adults with suspectedinfluenza, as summarized in the recentIDSA Influenza Clinical Practice Guideline(33).

Rationale for the recommendation. Thebenefits of antiviral therapy support testingof patients during periods of high influenzaactivity. During periods of low influenzaactivity, testing can be considered butmay not be routinely performed. Of note,this testing recommendation has boththerapeutic and infection-controlimplications in the hospital setting. Updatedinfluenza testing recommendations arealso available on the CDC website(https://www.cdc.gov/flu/professionals/diagnosis/index.htm).

Question 5: In Adults with CAP,Should Serum Procalcitonin plusClinical Judgment versus ClinicalJudgment Alone Be Used to WithholdInitiation of Antibiotic Treatment?

Recommendation. We recommend thatempiric antibiotic therapy should beinitiated in adults with clinically suspectedand radiographically confirmed CAPregardless of initial serum procalcitoninlevel (strong recommendation, moderatequality of evidence).

Summary of the evidence. Severalstudies have assessed the ability ofprocalcitonin to distinguish acuterespiratory infections due to pneumonia(which are of viral or bacterial etiology)from acute bronchitis or upper respiratorytract infections (which are almost

Table 3. Initial Treatment Strategies for Outpatients with Community-acquiredPneumonia

Standard Regimen

No comorbidities or risk factors for MRSAor Pseudomonas aeruginosa*

Amoxicillin ordoxycycline ormacrolide (if local pneumococcalresistance is ,25%)†

With comorbidities‡ Combination therapy withamoxicillin/clavulanate or cephalosporinANDmacrolide or doxycyclinex

ORmonotherapy with respiratoryfluoroquinolonejj

Definition of abbreviations: ER=extended release; MRSA=methicillin-resistant Staphylococcusaureus.*Risk factors include prior respiratory isolation of MRSA or P. aeruginosa or recent hospitalizationAND receipt of parenteral antibiotics (in the last 90 d).†Amoxicillin 1 g three times daily, doxycycline 100 mg twice daily, azithromycin 500 mg on first daythen 250 mg daily, clarithromycin 500 mg twice daily, or clarithromycin ER 1,000 mg daily.‡Comorbidities include chronic heart, lung, liver, or renal disease; diabetes mellitus; alcoholism;malignancy; or asplenia.xAmoxicillin/clavulanate 500 mg/125 mg three times daily, amoxicillin/clavulanate 875 mg/125 mgtwice daily, 2,000 mg/125 mg twice daily, cefpodoxime 200 mg twice daily, or cefuroxime 500 mgtwice daily; AND azithromycin 500 mg on first day then 250 mg daily, clarithromycin 500 mg twicedaily, clarithromycin ER 1,000 mg daily, or doxycycline 100 mg twice daily.jjLevofloxacin 750 mg daily, moxifloxacin 400 mg daily, or gemifloxacin 320 mg daily.



https://www.cdc.gov/flu/professionals/diagnosis/index.htmhttps://www.cdc.gov/flu/professionals/diagnosis/index.htm

Tab

le4.

Initial

Trea

tmen

tStrateg

iesforInpatientswith

Com

mun

ity-acq

uiredPne

umon

iabyLe

velo

fSev

erity

andRiskforDrugRes

istanc

e

Stand

ardReg

imen

PriorRes

piratory

Isolation

ofMRSA

PriorRes

piratory

Isolationof

Pse

udom

onas

aeruginos

a

Rec

entHosp

italizationan

dParen

teralAntibiotics

and

Loca

llyValidated

Risk

Factors

forMRSA

Rec

entHosp

italizationan

dParen

teralAntibiotics

and

Loca

llyValidated

Risk

Factors

forP.ae

ruginos

a

Non

seve

reinpatient

pne

umon

ia*

b-Lac

tam

1mac

rolid

e†or

resp

iratory

fluroq

uino

lone

‡AddMRSAco

verage

xan

dob

tain

cultu

res/na

salP

CR

toallow

dee

scalationor

confi

rmationof

need

forco

ntinue

dtherap

y

Addco

verage

forP.ae

rugino

sajj

andob

tain

cultu

resto

allow

dee

scalationor

confi

rmationof

need

forco

ntinue

dtherap

y

Obtain

cultu

resbut

with

hold

MRSAco

verage

unless

cultu

reresu

ltsarepos

itive

.If

rapid

nasa

lPCR

isav

ailable,

with

hold

addition

alem

piric

therap

yag

ains

tMRSAifrapid

testingis

nega

tiveor

add

cove

rage

ifPCR

ispos

itive

andob

tain

cultu

res

Obtain

cultu

resbut

initiate

cove

rage

forP.a

erug

inos

aon

lyifcu

lture

resu

ltsarepos

itive

Sev

ereinpatient

pne

umon

ia*

b-Lac

tam

1mac

rolid

e†or

b-lac

tam

1fluroq

uino

lone

‡AddMRSAco

verage

xan

dob

tain

cultu

res/na

salP

CR

toallow

dee

scalationor

confi

rmationof

need

forco

ntinue

dtherap

y

Addco

verage

forP.ae

rugino

sajj

andob

tain

cultu

resto

allow

dee

scalationor

confi

rmationof

need

forco

ntinue

dtherap

y

AddMRSAco

verage

xan

dob

tain

nasa

lPCR

and

cultu

resto

allow

dee

scalation

orco

nfirm

ationof

need

for

continue

dtherap

y

Addco

verage

forP.ae

rugino

sajj

andob

tain

cultu

resto

allow

dee

scalationor

confi

rmationof

need

forco

ntinue

dtherap

y

Definitionofabbreviatio

ns:

ATS=Americ

anThoracic

Society;CAP=community-acquire

dpneumonia;HAP=hosp

ital-acquire

dpneumonia;IDSA=Infectio

usDiseasesSociety

ofAmeric

a;

MRSA=methicillin-resistantStaphylococcusaureus;

VAP=ventilator-associatedpneumonia.

*Asdefinedby2007ATS/IDSACAPseveritycriteria

guidelines(seeTable

1).

†Ampicillin1su

lbactam

1.5–3

gevery

6hours,cefotaxime1–2

gevery

8hours,ceftria

xone1–2

gdaily,orceftaroline600mgevery

12hours

ANDazithromycin500mgdaily

orclarithromycin

500mgtw

icedaily.

‡Levo

floxacin

750mgdaily

ormoxifloxacin

400mgdaily.

x Perthe2016ATS/IDSAHAP/VAPguidelines:

vancomycin

(15mg/kgevery

12h,adjust

basedonlevels)orlinezo

lid(600mgevery

12h).

jj Perthe2016ATS/IDSAHAP/VAPguidelines:

piperacillin-tazo

bactam

(4.5

gevery

6h),cefepim

e(2

gevery

8h),ceftazidim

e(2

gevery

8h),im

ipenem

(500mgevery

6h),meropenem

(1gevery

8h),oraztreonam

(2gevery

8h).Doesnotincludecoverageforextended-spectrum

b-lactamase–p

roducingEnterobacteria

ceae,whichsh

ould

beconsideredonlyonthe

basisofpatientorlocalmicrobiologicaldata.



exclusively viral in etiology). However, forthe purposes of this guideline, the questionis whether, among patients with clinicallyconfirmed CAP, measurement ofprocalcitonin can distinguish patients withviral versus bacterial etiologies and guide theneed for initial antibiotic therapy. Someinvestigators have suggested thatprocalcitonin levels of 0.25 mg/L indicate a high likelihoodof bacterial pneumonia (34–36). However,a recent study in hospitalized patients withCAP failed to identify a procalcitoninthreshold that discriminated between viraland bacterial pathogens, although higherprocalcitonin strongly correlated withincreased probability of a bacterial infection(37). The reported sensitivity ofprocalcitonin to detect bacterial infectionranges from 38% to 91%, underscoring thatthis test alone cannot be used to justifywithholding antibiotics from patients withCAP (38).

Rationale for the recommendation.Procalcitonin has been used to guideinitiation of antibiotics in patients withlower respiratory infections, but many ofthese studies are not restricted to patientswith radiographically confirmedpneumonia. Some patients with lowprocalcitonin levels have CAP and havebeen safely treated without antibiotics (35),but these represent small subgroups, raisingconcerns about the safety of widely usingsuch a strategy.

Research needed in this area. Given theepidemiological evidence that viruses are animportant cause of CAP, there is a criticalneed to validate the use of current rapidlaboratory tests, including point-of-caretests, to accurately identify situations inwhich antibacterial therapy can be safelywithheld among adults with CAP.

Question 6: Should a ClinicalPrediction Rule for Prognosis plusClinical Judgment versus ClinicalJudgment Alone Be Used toDetermine Inpatient versusOutpatient Treatment Location forAdults with CAP?

Recommendation. In addition to clinicaljudgement, we recommend that cliniciansuse a validated clinical prediction rulefor prognosis, preferentially thePneumonia Severity Index (PSI) (strongrecommendation, moderate quality of

evidence) over the CURB-65 (tool basedon confusion, urea level, respiratory rate,blood pressure, and age >65) (conditionalrecommendation, low quality of evidence),to determine the need for hospitalization inadults diagnosed with CAP.

Summary of the evidence. Both the PSIand CURB-65 were developed as prognosticmodels in immunocompetent patients withpneumonia, using patient demographic andclinical variables from the time of diagnosisto predict 30-day mortality (39, 40). Whencompared with CURB-65, PSI identifieslarger proportions of patients as low riskand has a higher discriminative power inpredicting mortality (41).

Two multicenter, cluster-randomizedtrials demonstrated that use of the PSI safelyincreases the proportion of patients who canbe treated in the outpatient setting (42, 43).These trials and one additional randomizedcontrolled trial (RCT) support the safety ofusing the PSI to guide the initial site oftreatment of patients without worseningmortality or other clinically relevantoutcomes (42–44). Consistent evidencefrom three pre–post interventionstudies and one prospective controlledobservational study support theeffectiveness and safety of using the PSI toguide the initial site of treatment (45–48).

Clinical severity is not the onlyconsideration in determining the need forhospital admission (49, 50). Some patientshave medical and/or psychosocialcontraindications to outpatient therapy,such as inability to maintain oral intake,history of substance abuse, cognitiveimpairment, severe comorbid illnesses, andimpaired functional status.

The PSI may underestimate illnessseverity among younger patients andoversimplify how clinicians interpretcontinuous variables (e.g., all systolic bloodpressures ,90 mm Hg are consideredabnormal, regardless of the patient’sbaseline and actual measurement).Therefore, when used as a decision aid, thePSI should be used in conjunction withclinical judgment.

In comparison to the PSI, there is lessevidence that CURB-65 is effective asa decision aid in guiding the initial site oftreatment. One pre–post, controlledintervention study using an electronicallycalculated version of CURB-65,PaO2/FIO2, 300, absence of pleuraleffusion, and fewer than three minor ATSseverity criteria observed no significant

increase in the use of outpatient treatmentfor adults with CAP (51). A randomizedtrial compared the safety of inpatient versusoutpatient treatment of 49 patients withCURB-65 scores of less than 2 (52) but hadlimited power to detect differences inpatient outcomes; furthermore, outpatienttreatment included daily nursing visits andparenteral antibiotic therapy that istypically restricted to inpatient care.

Rationale for the recommendation. Ourrecommendation to use the PSI as anadjunct to clinical judgment to guide theinitial site of treatment is based onconsistent evidence of the effectiveness andsafety of this approach. Using a safe andeffective decision aid to increase outpatienttreatment of patients with CAP has potentialto decrease unnecessary variability inadmission rates, the high cost of inpatientpneumonia treatment (53, 54), and therisk of hospital-acquired complications.Providing a conditional recommendationto use CURB-65 considers its greatersimplicity of use relative to the PSI despitethe paucity of evidence regarding itseffectiveness or safety.

Research needed in this area. It isimportant to study the effectiveness andsafety of using CURB scores or newprediction rules for prognosis as decisionaids to guide the initial site of treatment forpatients with CAP compared with the PSI.Future studies of prediction rules shouldalso test electronic versions generated in realtime from data routinely recorded in theelectronic medical record and assess theirperformance in patient populationsexcluded from the development of existingprediction rules (55, 56).

Question 7: Should a ClinicalPrediction Rule for Prognosis plusClinical Judgment versus ClinicalJudgment Alone Be Used toDetermine Inpatient General Medicalversus Higher Levels of InpatientTreatment Intensity (ICU, Step-Down, or Telemetry Unit) forAdults with CAP?

Recommendation. We recommend directadmission to an ICU for patients withhypotension requiring vasopressors orrespiratory failure requiring mechanicalventilation (strong recommendation, lowquality of evidence).

For patients not requiring vasopressorsor mechanical ventilator support, we suggest



using the IDSA/ATS 2007 minor severitycriteria (Table 1) together with clinicaljudgment to guide the need for higher levelsof treatment intensity (conditionalrecommendation, low quality of evidence).

Summary of the evidence. The PSI andCURB-65 were not designed to help selectthe level of care needed by a patient who ishospitalized for CAP. Several prognosticmodels have been designed to predict theneed for higher levels of inpatient treatmentintensity using severity-of-illnessparameters based on patient outcomes (ATS2001, IDSA/ATS 2007, SMART-COP, andSCAP score). Studies of prognostic modelshave used different end points, includinginpatient mortality (57, 58), ICU admission(57–59), receipt of intensive respiratory orvasopressor support (59, 60), or ICUadmission plus receipt of a critical therapy(61). In comparative studies, theseprognostic models yield higher overallaccuracy than the PSI or CURB-65 whenusing illness outcomes other than mortality(58, 59, 61).

The 2007 IDSA/ATS CAP guidelinesrecommended a set of two major and nineminor criteria to define severe pneumoniarequiring ICU admission (Table 1). Thesecriteria were based on empirical evidencefrom published studies and expertconsensus. All elements are routinelyavailable in emergency department settingsand are electronically calculable (51, 61).Several groups have validated these criteriain pneumonia cohorts from differentcountries (57–59, 61), with a meta-analysisreporting one major or three minor criteriahad a pooled sensitivity of 84% anda specificity of 78% for predicting ICUadmission (62). Without the major criteria,a threshold of three or more minor criteria(recommended in the 2007 IDSA/ATSguideline) had a pooled sensitivity of 56%and specificity of 91% for predicting ICUadmission (63).

SMART-COP is an alternative,validated prediction rule for identifyingpatients with pneumonia who needvasopressor support and/or mechanicalventilation. The eight SMART-COP criteriaand the nine 2007 IDSA/ATS minor criteriahave five overlapping elements: hypoxia,confusion, respiratory rate, multilobarradiographic opacities, and low systolicblood pressure. SMART-COP had a pooledsensitivity of 79% and specificity of 64% inpredicting ICU admission using a thresholdof three or more criteria but uses albumin,

PaO2, and pH, which are not universallyavailable for real-time clinical decision-making (60). For predicting ICU admission,one comparison reported equivalence of theIDSA/ATS minor criteria and SMART-COP (63) and another reporteda significantly greater performance of theIDSA/ATS minor criteria (61). Norandomized studies have evaluated theeffectiveness or safety of an illness severitytool as a decision aid to guide the intensityof inpatient treatment for patientshospitalized with CAP.

Rationale for the recommendation.Patients transferred to an ICU afteradmission to a hospital ward experiencehigher mortality than those directlyadmitted to the ICU from an emergencydepartment (64–67). This higher mortalitymay in part be attributable to progressivepneumonia, but “mis-triage” of patientswith unrecognized severe pneumonia maybe a contributing factor (64). It seemsunlikely that physician judgment alonewould be equivalent to physician judgmenttogether with a severity tool to guide thesite-of-care decision. We recommend the2007 IDSA/ATS severe CAP criteria overother published scores, because they arecomposed of readily available severityparameters and are more accurate than theother scores described above.

Research needed in this area.Controlled studies are needed to study theeffectiveness and safety of using illnessseverity tools as decision aids to guide theintensity of treatment in adults hospitalizedfor pneumonia.

Question 8: In the Outpatient Setting,Which Antibiotics Are Recommendedfor Empiric Treatment of CAP inAdults?

Recommendation.1. For healthy outpatient adults without

comorbidities listed below or risk factorsfor antibiotic resistant pathogens, werecommend (Table 3):d amoxicillin 1 g three times daily(strong recommendation, moderatequality of evidence), or

d doxycycline 100 mg twice daily(conditional recommendation, lowquality of evidence), or

d a macrolide (azithromycin 500 mg onfirst day then 250 mg daily orclarithromycin 500 mg twice daily or

clarithromycin extended release1,000 mg daily) only in areas withpneumococcal resistance tomacrolides ,25% (conditionalrecommendation, moderate quality ofevidence).

2. For outpatient adults with comorbiditiessuch as chronic heart, lung, liver, orrenal disease; diabetes mellitus;alcoholism; malignancy; or asplenia werecommend (in no particular order ofpreference) (Table 3):d Combination therapy:

B amoxicillin/clavulanate 500 mg/125mg three times daily, or amoxicillin/clavulanate 875 mg/125 mg twicedaily, or 2,000 mg/125 mg twicedaily, or a cephalosporin(cefpodoxime 200 mg twice daily orcefuroxime 500 mg twice daily);AND

B macrolide (azithromycin 500 mgon first day then 250 mg daily,clarithromycin [500 mg twice dailyor extended release 1,000 mg oncedaily]) (strong recommendation,moderate quality of evidence forcombination therapy), or doxycycline100 mg twice daily (conditionalrecommendation, low quality ofevidence for combination therapy);OR

d Monotherapy:B respiratory fluoroquinolone(levofloxacin 750 mg daily,moxifloxacin 400 mg daily, orgemifloxacin 320 mg daily) (strongrecommendation, moderate qualityof evidence).

Summary of the evidence. RCTs ofantibiotic treatment regimens for adultswith CAP provide little evidence of eithersuperiority or equivalence of one antibioticregimen over another, because of smallnumbers and the rare occurrence ofimportant outcomes such as mortalityor treatment failure resulting inhospitalization. Several published trialsincluded comparators that are no longeravailable (e.g., ketolides). This paucity ofdata was noted in a 2014 Cochrane review(68).

We identified 16 relevant RCTscomparing two antibiotic regimens for thetreatment of outpatient CAP (69–84). Meta-analyses of each of these groups of studiesrevealed no differences in relevantoutcomes between any of the comparedregimens. Similar findings have been



reported in a 2008 meta-analysis ofantibiotic treatment for outpatient CAP (85).

The committee also consideredwhether to accept data regarding oralantibiotics given to inpatients with CAP.Webelieved that this evidence, albeit indirect,could be reasonably extended to outpatients,because inpatients are generally higher riskand more severely ill. As observational datasuggest that inpatient and outpatient CAPare due to the same pathogens (69, 71–73,82), except for Legionella and gram-negativebacilli, which are rarely documented inoutpatient settings, it seems reasonable thatan antibiotic regimen that was effective forinpatients would be effective for outpatients.

Studies of high-dose oral amoxicillinhave demonstrated efficacy for inpatientswith CAP (86–88). Similarly, there isevidence supporting amoxicillin-clavulanicacid in outpatient CAP (71, 73) andinpatient CAP (89, 90).

There are limited data regarding oraldoxycycline for pneumonia, mostly involvingsmall numbers of patients (81). Intravenousdoxycycline 100 mg twice daily comparedfavorably to intravenous levofloxacin 500 mgdaily in 65 in patients with CAP (91). In anopen-label randomized trial of intravenousdoxycycline 100 mg twice daily comparedwith standard antibiotics, doxycycline wasassociated with a quicker response and lesschange in antibiotics (92).

Rationale for the recommendation. Giventhe paucity of RCT data in the outpatientsetting, the committee considered allavailable evidence. The data included thefew RCTs of outpatient CAP, observationalstudies, RCTs of inpatient CAP treatment,antimicrobial resistance data fromsurveillance programs, and data regardingantibiotic-related adverse events.

For patients without comorbidities thatincrease the risk for poor outcomes, thepanel recommended amoxicillin 1 g every8 hours or doxycycline 100 mg twice daily.The recommendation for amoxicillin wasbased on several studies that showed efficacyof this regimen for inpatient CAP despitepresumed lack of coverage of this antibioticfor atypical organisms. This treatment alsohas a long track record of safety. Therecommendation for doxycycline was basedon limited clinical trial data, but a broadspectrum of action, including the mostcommon relevant organisms. Some expertsrecommend that the first dose of oraldoxycycline be 200 mg, to achieve adequateserum levels more rapidly. There are no data

assessing whether such an approach isassociated with improved outcomes.

In a departure from the prior CAPguidelines, the panel did not give a strongrecommendation for routine use ofa macrolide antibiotic as monotherapy foroutpatient CAP, even in patients withoutcomorbidities. This was based on studiesof macrolide failures in patients withmacrolide-resistant S. pneumoniae (93, 94),in combination with a macrolide resistancerate of.30% among S. pneumoniae isolatesin the United States, most of which is high-level resistance (95). However, in settingswhere macrolide resistance is documentedto be low and there are contraindicationsto alternative therapies, a macrolide asmonotherapy is a treatment option.

Patients with comorbidities shouldreceive broader-spectrum treatment for tworeasons. First, such patients are likely morevulnerable to poor outcomes if the initialempiric antibiotic regimen is inadequate.Second, many such patients have risk factorsfor antibiotic resistance by virtue of previouscontact with the healthcare systemand/or prior antibiotic exposure (seeRecommendation 10) and are thereforerecommended to receive broader-spectrumtherapy to ensure adequate coverage. Inaddition to H. influenzae andM. catarrhalis(both of which frequently produceb-lactamase), S. aureus and gram-negativebacilli are more common causes of CAP inpatients with comorbidities, such as COPD.

Regimens recommended for patientswith comorbidities include a b-lactam orcephalosporin in combination with eithera macrolide or doxycycline. Thesecombinations should effectively targetmacrolide- and doxycycline-resistantS. pneumoniae (as b-lactam resistance inS. pneumoniae remains less common), inaddition to b-lactamase–producing strainsof H. influenzae, many enteric gram-negative bacilli, most methicillin-susceptible S. aureus, and M. pneumoniaeand C. pneumoniae. The monotherapieslisted also are effective against mostcommon bacterial pathogens.

Both sets of treatment recommendationscontain multiple antibiotic optionswithout specifying a preference order.The choice between these options requiresa risk–benefit assessment for each individualpatient, weighing local epidemiologicaldata against specific risk factors thatincrease the risk of individual choices,such as documented b-lactam or macrolide

allergy, cardiac arrhythmia (macrolides),vascular disease (fluoroquinolones),and history of infection with Clostridiumdifficile. In particular, despite the concernregarding adverse events associated withfluoroquinolones, the panel believed thatfluoroquinolone therapy was justified foradults with comorbidities and CAPmanaged in the outpatient setting.Reasons included the performance offluoroquinolones in numerous studies ofoutpatient CAP (70, 72, 75, 77, 80, 83)and inpatient CAP (see inpatient CAPsection), the very low resistance ratesin common bacterial causes of CAP,their coverage of both typical andatypical organisms, their oral bioavailability,the convenience of monotherapy, and therelative rarity of serious adverse events relatedto their use. However, there have beenincreasing reports of adverse events related tofluoroquinolone use as summarized on theU.S. Food and Drug Administrationwebsite (96).

Of note, we adopt the convention ofprior guidelines to recommend that patientswith recent exposure to one class of antibioticsrecommended above receive treatment withantibiotics from a different class, givenincreased risk for bacterial resistance to theinitial treatment regimen. We also highlightthat although patients with significant riskfactors for CAP due toMRSA or P. aeruginosa(see Recommendation 11) are uncommonlymanaged in the outpatient setting, thesepatients may require antibiotics that includecoverage for these pathogens.

Research needed in this area. There isa need for head-to-head prospective RCTsof outpatient CAP treatment, comparingclinical outcomes, including treatmentfailure, need for subsequent visits,hospitalization, time to return to usualactivities and adverse events. Furthermore,the prevalence of specific pathogens andtheir antimicrobial susceptibility patterns inoutpatients with pneumonia should bemonitored. Newer agents, includinglefamulin and omadacycline, need furthervalidation in the outpatient setting.

Question 9: In the Inpatient Setting,Which Antibiotic Regimens AreRecommended for Empiric Treatmentof CAP in Adults without Risk Factorsfor MRSA and P. aeruginosa?

Recommendation 9.1. In inpatient adultswith nonsevere CAP without risk factors for



MRSA or P. aeruginosa (see Recommendation11), we recommend the following empirictreatment regimens (in no order ofpreference) (Table 4):

d combination therapy with a b-lactam(ampicillin1 sulbactam 1.5–3 g every6 h, cefotaxime 1–2 g every 8 h,ceftriaxone 1–2 g daily, or ceftaroline600 mg every 12 h) and a macrolide(azithromycin 500 mg daily orclarithromycin 500 mg twice daily)(strong recommendation, high quality ofevidence), or

d monotherapy with a respiratoryfluoroquinolone (levofloxacin 750 mgdaily, moxifloxacin 400 mg daily) (strongrecommendation, high quality ofevidence).

A third option for adults with CAP whohave contraindications to both macrolidesand fluoroquinolones is:

d combination therapy with a b-lactam(ampicillin1 sulbactam, cefotaxime,ceftaroline, or ceftriaxone, doses asabove) and doxycycline 100 mg twicedaily (conditional recommendation, lowquality of evidence).Summary of the evidence. Most

randomized controlled studies ofhospitalized adults with CAP comparingb-lactam/macrolide therapy versusfluoroquinolone monotherapy weredesigned as noninferiority trials and hadlimited sample sizes (97–103). These datasuggested that patients treated withb-lactam/macrolide therapy have similarclinical outcomes compared with thosetreated with fluoroquinolone monotherapy.A systematic review of 16 RCTs in 4,809patients found fluoroquinolonemonotherapy resulted in significantly fewerincidences of clinical failure, treatmentdiscontinuation, and diarrhea thanb-lactam/macrolide combination (104).However, mortality rates were low overall,and there were no significant differences inmortality between groups. Anothersystematic review of 20 experimental andobservational studies in adults hospitalizedwith radiographically confirmed CAP,b-lactam plus macrolide combinationtherapy, or fluoroquinolone monotherapywere generally associated with lowermortality than b-lactam monotherapy(105). Therefore, the panel recommendsa b-lactam (ampicillin plus sulbactam,cefotaxime, ceftaroline, or ceftriaxone) plus

macrolide (azithromycin or clarithromycin)or monotherapy with a respiratoryfluoroquinolone (levofloxacin,moxifloxacin) for the management ofinpatients with nonsevere CAP. (Of note,azithromycin but not clarithromycin isavailable in parenteral formulation.) Inchoosing between these two options,clinicians should weigh the risks andbenefits of the drugs, particularly in light ofindividual risk factors, such as a history ofC. difficile infection or risk factors related toU.S. Food and Drug Administrationwarnings (96). The panel recommendsusing doxycycline as an alternative toa macrolide in combination with a b-lactam as a third option in the presence ofdocumented allergies or contraindicationsto macrolides or fluoroquinolones orclinical failure on one of those agents. Ofnote, a newer member of the tetracyclineclass, omadacycline, was recently reportedto be equivalent to moxifloxacin asmonotherapy for adults with nonsevereCAP and is effective in the setting oftetracycline resistance (106). However, asthis is a single published report and thesafety information is less well established,the committee decided to not list this newagent as an alternative to the currentlyrecommended treatment options.

The panel also considered b-lactammonotherapy as an option for inpatientswith nonsevere CAP. An RCT in 580patients with CAP could not rule out thepossibility that b-lactam monotherapy wasinferior to b-lactam/macrolide therapy forinpatients with CAP (107). Nie andcolleagues identified several cohort (n= 4)and retrospective observational (n= 12)studies addressing this question and foundthat b-lactam/macrolide therapy reducedmortality in patients with CAP comparedwith patients treated with b-lactammonotherapy (108). Similarly, Horitaand colleagues demonstrated thatb-lactam/macrolide combinations maydecrease all-cause death, but mainly forpatients with severe CAP (109). Therefore,we suggest that b-lactam monotherapyshould not be routinely used for inpatientswith CAP over fluoroquinolonemonotherapy or b-lactam/macrolidecombination therapy.

Rationale for the recommendation. Assummarized in Table 4, the empiricantibiotic coverage recommendations forpatients hospitalized with CAP remainaligned to cover the most likely pathogens

causing CAP. There is a paucity of RCTs tofavor the recommendation of combinationb-lactam plus macrolide versusmonotherapy with a respiratoryfluoroquinolone versus combined therapywith b-lactam plus doxycycline.

Research needed in this area. There isa need for higher-quality evidence insupport of the use of combination therapywith a b-lactam and doxycycline. Givenconcerns over increasing drug resistance(macrolides) and safety issues (macrolides,fluoroquinolones), there is a need forresearch on new therapeutic agents foradults with CAP including omadacycline(see above) and lefamulin, a newpleuromutilin antibiotic that was recentlydemonstrated to be noninferior tomoxifloxacin in hospitalized adult patientswith CAP (110).

Recommendation 9.2. In inpatientadults with severe CAP (see Table 1)without risk factors for MRSA or P.aeruginosa, we recommend (Table 4) (note,specific agents and doses are the same as9.1):

d a b-lactam plus a macrolide (strongrecommendation, moderate quality ofevidence); or

d a b-lactam plus a respiratoryfluoroquinolone (strongrecommendation, low quality ofevidence).Summary of the evidence. In the

absence of RCTs evaluating therapeuticalternatives in severe CAP, the evidence isfrom observational studies that useddifferent definitions of illness severity toaddress this question. Sligl and colleaguesfound in a meta-analysis of observationalstudies with almost 10,000 critically illpatients with CAP that macrolide-containing therapies (often in combinationwith a b-lactam) were associated witha significant mortality reduction (18%relative risk, 3% absolute risk) comparedwith non–macrolide-containing therapies(111). A mortality benefit from macrolideshas been observed mainly in cohortswith a large number of patients withsevere CAP. In a systematic review,Vardakas and colleagues compared ab-lactam/fluoroquinolone versus ab-lactam/macrolide combination for thetreatment of patients with CAP (112). Theauthors found 17 observational studies andno RCTs addressing this comparison. Thecombination of b-lactam/fluoroquinolone



therapy was associated with highermortality than b-lactam/macrolidecombination therapy, but the overall qualityof the studies was judged to be low,precluding a definitive recommendation(112).

Rationale for the recommendation. Inthe absence of data from clinical trialsdemonstrating the superiority of anyspecific regimen for patients with severeCAP, the committee consideredepidemiological data for severe CAPpathogens and observational studiescomparing different regimens. As a result,we recommend that combination therapywith a b-lactam plus a macrolide or a b-lactam plus a respiratory fluoroquinoloneshould be the treatment of choice forpatients with severe CAP. Bothfluoroquinolone monotherapy and thecombination of b-lactam plusdoxycycline have not been well studied insevere CAP and are not recommended asempiric therapy for adults with severeCAP.

Research needed in this area. Futurewell-designed RCTs should focus ontherapies for patients at highest risk ofdeath with severe pneumonia, as theseare needed to assess the benefits andrisks of combination b-lactam andmacrolide therapy compared withb-lactam and respiratory fluoroquinolonetherapy. Studies of fluoroquinolonemonotherapy in severe CAP are also needed.

Question 10: In the Inpatient Setting,Should Patients with SuspectedAspiration Pneumonia ReceiveAdditional Anaerobic Coveragebeyond Standard Empiric Treatmentfor CAP?

Recommendation. We suggest not routinelyadding anaerobic coverage for suspectedaspiration pneumonia unless lung abscess orempyema is suspected (conditionalrecommendation, very low quality ofevidence).

Summary of the evidence. Aspiration isa common event, and as many as half of alladults aspirate during sleep (113). Asa result, the true rate of aspirationpneumonia is difficult to quantify, andthere is no definition that separates patientswith aspiration pneumonia from all othersdiagnosed with pneumonia. Some haveestimated that 5% to 15% of pneumonia

hospitalizations are associated withaspiration (114). Rates are higher inpopulations admitted from nursing homesor extended care facilities (115).

Patients who aspirate gastric contentsare considered to have aspirationpneumonitis. Many of these patients haveresolution of symptoms within 24 to48 hours and require only supportivetreatment, without antibiotics (116).

Studies evaluating the microbiology ofpatients with aspiration pneumonia in the1970s showed high rates of isolation ofanaerobic organisms (117, 118); however,these studies often used trans-trachealsampling and evaluated patients late intheir disease course, two factors that mayhave contributed to a higher likelihood ofidentifying anaerobic organisms (114).Several studies of acute aspiration events inhospitalized patients have suggested thatanaerobic bacteria do not play a major rolein etiology (119–121).

With increasing rates of C. difficileinfections (frequently associated with useof clindamycin), the question of addingempiric anaerobic coverage (clindamycinor b-lactam/b-lactamase inhibitors) inaddition to routine CAP treatment inpatients with suspected aspiration is animportant one. However, there are noclinical trials comparing treatmentregimens with and without anaerobiccoverage for patients hospitalized withsuspected aspiration. Most recent studiesare small, retrospective, and provide onlyobservational data on microbiologicpatterns and treatment regimens forpatients hospitalized with suspectedaspiration pneumonia.

Rationale for the recommendation.Although older studies of patients withaspiration pneumonia showed highisolation rates of anaerobic organisms,more recent studies have shown thatanaerobes are uncommon in patientshospitalized with suspected aspiration(119, 120). Increasing prevalence ofantibiotic-resistant pathogens andcomplications of antibiotic usehighlight the need for a treatmentapproach that avoids unnecessary useof antibiotics.

Research needed in this area. Clinicaltrials evaluating diagnostic and treatmentstrategies in patients with suspectedaspiration are needed, especially in terms ofthe ability to distinguish micro- and macro-aspiration events that lead to lower

respiratory tract infection from those thatdo not result in infection.

Question 11: In the Inpatient Setting,Should Adults with CAP and RiskFactors for MRSA or P. aeruginosa BeTreated with Extended-SpectrumAntibiotic Therapy Instead ofStandard CAP Regimens?

Recommendation. We recommendabandoning use of the prior categorization ofhealthcare-associated pneumonia (HCAP) toguide selection of extended antibiotic coveragein adults with CAP (strong recommendation,moderate quality of evidence).

We recommend clinicians only coverempirically for MRSA or P. aeruginosa inadults with CAP if locally validated riskfactors for either pathogen are present(strong recommendation, moderate qualityof evidence). Empiric treatment options forMRSA include vancomycin (15 mg/kgevery 12 h, adjust based on levels) orlinezolid (600 mg every 12 h). Empirictreatment options for P. aeruginosa includepiperacillin-tazobactam (4.5 g every 6 h),cefepime (2 g every 8 h), ceftazidime (2 gevery 8 h), aztreonam (2 g every 8 h),meropenem (1 g every 8 h), or imipenem(500 mg every 6 h).

If clinicians are currently coveringempirically for MRSA or P. aeruginosa inadults with CAP on the basis of publishedrisk factors but do not have local etiologicaldata, we recommend continuing empiriccoverage while obtaining culture data toestablish if these pathogens are present tojustify continued treatment for thesepathogens after the first few days of empirictreatment (strong recommendation, lowquality of evidence).

Summary of the evidence. HCAP, asa distinct clinical entity warranting uniqueantibiotic treatment, was incorporated intothe 2005 ATS/IDSA guidelines formanagement of hospital-acquired andventilator-associated pneumonia (122).HCAP was defined for those patients whohad any one of several potential risk factorsfor antibiotic-resistant pathogens, includingresidence in a nursing home and otherlong-term care facilities, hospitalization for>2 days in the last 90 days, receipt of homeinfusion therapy, chronic dialysis, homewound care, or a family member witha known antibiotic-resistant pathogen. Theintroduction of HCAP was based on studiesidentifying a higher prevalence of



pathogens that are not susceptible tostandard first-line antibiotic therapy, inparticular MRSA and P. aeruginosa, insome subsets of patients with CAP (123).Since then, many studies havedemonstrated that the factors used todefine HCAP do not predict highprevalence of antibiotic-resistant pathogensin most settings. Moreover, a significantincreased use of broad-spectrumantibiotics (especially vancomycin andantipseudomonal b-lactams) has resulted,without any apparent improvement inpatient outcomes (124–133).

Studies have identified risk factors forantibiotic-resistant pathogens, and in somecases the risk factors are distinct for MRSAversus P. aeruginosa (134–154). However,most of these individual risk factors areweakly associated with these pathogens.The most consistently strong individual riskfactors for respiratory infection with MRSAor P. aeruginosa are prior isolation of theseorganisms, especially from the respiratorytract, and/or recent hospitalization andexposure to parenteral antibiotics (134, 155,156). Therefore, we have highlighted theseindividual risk factors to help guide initialmicrobiological testing and empiriccoverage for these pathogens.

Unfortunately, no validated scoringsystems exist to identify patients withMRSAor P. aeruginosa with sufficiently highpositive predictive value to determine theneed for empiric extended-spectrumantibiotic treatment. Scoring systemdevelopment and validation are complicatedby varying prevalence of MRSA andP. aeruginosa in different study populations.Moreover, no scoring system has beendemonstrated to improve patient outcomesor reduce overuse of broad-spectrumantibiotics.

Although there is limited evidence tosupport the use of a specific set of riskfactors to identify patients with sufficientlyhigh risk of MRSA or P. aeruginosa towarrant extended-spectrum therapy,a stronger evidence base guidesdeescalation of therapy after extended-spectrum therapy is initially prescribed.Although no randomized prospectivestudies have been reported, recentobservational (157) and retrospective(158–161) studies in patients with CAPprovide strong evidence that deescalation ofantibiotic therapy at 48 hours in accordwith microbiological results that do notyield MRSA or P. aeruginosa is safe and

reduces duration of antibiotic treatment,length of hospitalization, and complicationsof broad-spectrum therapy. These results arereinforced by retrospective (162) andprospective and randomized but not blinded(163) studies of patients with severe sepsis,the majority of whom had pneumonia, aswell as by a recent meta-analysis ofdeescalation in adults with sepsis (164).

We propose that clinicians need toobtain local data on whether MRSA orP. aeruginosa is prevalent in patients withCAP and what the risk factors for infectionare at a local (i.e., hospital or catchmentarea) level. We refer to this process as “localvalidation.” This recommendation is basedon the absence of high-quality outcomestudies, the very low prevalence of MRSAor P. aeruginosa in most centers, andsignificant increased use of anti-MRSA andantipseudomonal antibiotics for treatmentof CAP (142, 155, 165). Although weacknowledge that centers may not currentlyhave local prevalence data, adopting therecommendations for culture of sputumand blood when risk factors for MRSA or P.aeruginosa are present will enable cliniciansto generate these local data over time. Werecommend analyzing the frequency ofMRSA or P. aeruginosa as a CAP pathogenrelative to the number of all cases of CAP,not just those for whom cultures are sent.Finally, routine cultures in patientsempirically treated for MRSA or P.aeruginosa allow deescalation to standardCAP therapy if cultures do not reveala drug-resistant pathogen and the patient isclinically improving at 48 hours.

Rationale for the recommendation. Ourapproach to treating inpatient adults withCAP is summarized in Table 4. Ourrecommendation against using the formercategory of HCAP as a basis for selectingextended-spectrum therapy is based onhigh-quality studies of patient outcomes.Although we understand that clinicianswould prefer a simple rule that does notrequire incorporating site-specific data, thecurrent evidence does not permitendorsement of a simple and accurate ruleto determine which patients with CAPshould be covered for MRSA and/or P.aeruginosa. However, the alternativeapproach to MRSA and P. aeruginosa thatwe propose as a replacement is not basedon high-quality studies, because suchstudies do not exist. The lack of adequateoutcome data and marked variationbetween sites in prevalence of MRSA and

P. aeruginosa make generalizing any findingsextremely difficult. We hope that futureresearch will improve our understanding ofthis challenging clinical problem.

Our first principle was to maintain thedistinction between severe and nonseverepneumonia as per prior guidelines, becausethe risk of inadequate empiric antibiotictherapy is much greater in severe CAP. Asnoted previously, severity is defined by thedegree of physiological impairment, asclassified by the IDSA/ATS 2007 criteria.

The second principle was that there issufficient evidence that prior identificationof MRSA or P. aeruginosa in the respiratorytract within the prior year predicts a veryhigh risk of these pathogens beingidentified in patients presenting with CAP(139, 141, 143, 150, 155, 165), and thereforethese were sufficient indications torecommend blood and sputum cultures andempiric therapy for these pathogens inpatients with CAP in addition to coveragefor standard CAP pathogens, withdeescalation at 48 hours if cultures arenegative. We endorse the empiric treatmentrecommendations for MRSA and P.aeruginosa provided by the 2016 ClinicalPractice Guideline from IDSA and ATS forthe management of adults with hospital-acquired and ventilator-associatedpneumonia (166).

The major additional risk factors forMRSA and P. aeruginosa identified in theliterature are hospitalization and parenteralantibiotic exposure in the last 90 days(136–138, 140, 142–151, 153). In patientswith recent hospitalization and exposure toparenteral antibiotics, we recommendmicrobiological testing without empiricextended-spectrum therapy for treatment ofnonsevere CAP and microbiological testingwith extended-spectrum empiric therapy inaddition to coverage for standard CAPpathogens for treatment of severe CAP, withdeescalation at 48 hours if cultures arenegative and the patient is improving.

The data supporting rapid MRSA nasaltesting are robust (167, 168), and treatmentfor MRSA pneumonia can generally bewithheld when the nasal swab is negative,especially in nonsevere CAP. However, thepositive predictive value is not as high;therefore, when the nasal swab is positive,coverage for MRSA pneumonia shouldgenerally be initiated, but blood andsputum cultures should be sent and therapydeescalated if cultures are negative.However, this latter strategy of deescalation



in the face of a positive nasal swab will varydepending on the severity of CAP and thelocal prevalence of MRSA as a pathogen.

Question 12: In the Inpatient Setting,Should Adults with CAP Be Treatedwith Corticosteroids?

Recommendation. We recommend notroutinely using corticosteroids inadults with nonsevere CAP (strongrecommendation, high quality of evidence).

We suggest not routinely usingcorticosteroids in adults with severe CAP(conditional recommendation, moderatequality of evidence).

We suggest not routinely usingcorticosteroids in adults with severeinfluenza pneumonia (conditionalrecommendation, low quality of evidence).

We endorse the Surviving SepsisCampaign recommendations on the use ofcorticosteroids in patients with CAP andrefractory septic shock (169).

Summary of the evidence. Tworandomized controlled studies ofcorticosteroids used for treatment of CAPhave shown significant reductions inmortality, length of stay, and/or organfailure. The first study found a largemagnitude of mortality benefit that has notbeen replicated in other studies, raisingconcerns that the results overestimated thetrue effect (170). In the second study, therewere baseline differences in renal functionbetween groups (171). Other RCTs ofcorticosteroids in the treatment of CAPhave not shown significant differences inclinically important endpoints. Differenceshave been observed in the time toresolution of fever and other features ofclinical stability, but these have nottranslated into differences in mortality,length of stay, or organ failure (172, 173).

Some (174, 175), but not all (176, 177),meta-analyses of published corticosteroidstudies have shown a mortality benefit inpatients with severe CAP, although noconsistent definition of disease severity wasused. Side effects of corticosteroids (on theorder of 240 mg of hydrocortisone perday) include significant increases inhyperglycemia requiring therapy andpossible higher secondary infection rates(178, 179). No reported study has shownexcess mortality in the corticosteroid-treated group.

In pneumonia due to influenza, a meta-analysis (180) of predominantly small

retrospective studies suggests that mortalitymay be increased in patients who receivecorticosteroids. This finding might reflectthe importance of innate immunity indefense against influenza as opposed tobacterial pneumonia.

Rationale for the recommendation.There are no data suggesting benefit ofcorticosteroids in patients with nonsevereCAP with respect to mortality or organfailure and only limited data in patients withsevere CAP. The risk of corticosteroids inthe dose range up to 240 mg ofhydrocortisone equivalent per day fora maximum of 7 days is predominantlyhyperglycemia, although rehospitalizationrates may also be higher (176), andmore general concerns about greatercomplications in the following 30 to 90days have been raised (179). At least onelarge trial (clinicaltrials.gov NCT01283009)has been completed but not reported andmay further inform which subgroups ofpatients benefit from steroids. We alsoendorse the Surviving Sepsis Campaignrecommendations on the use of steroids inpatients with septic shock refractory toadequate fluid resuscitation andvasopressor support (169).

Of note, there is no intent that ourrecommendations would override clinicallyappropriate use of steroids for comorbiddiseases, such as chronic obstructivepulmonary disease, asthma, andautoimmune diseases, where corticosteroidsare supported as a component of treatment.

Research needed in this area. Large,multicenter, randomized trials with well-defined inclusion and exclusion criteria andmeasurement of multiple relevant clinicaloutcomes are needed to define the subsets ofpatients (if any) who benefit or arepotentially harmed from corticosteroidtherapy. The trial should also makeextensive efforts to define causativepathogens, to define whether there areclear pathogen-specific indications orcontraindications for corticosteroid therapy(especially disease due to S. pneumoniaeand influenza).

Question 13: In Adults with CAP WhoTest Positive for Influenza, Should theTreatment Regimen Include AntiviralTherapy?

Recommendation. We recommend thatantiinfluenza treatment, such as oseltamivir,be prescribed for adults with CAP who test

positive for influenza in the inpatientsetting, independent of duration of illnessbefore diagnosis (strong recommendation,moderate quality of evidence).

We suggest that antiinfluenzatreatment be prescribed for adults with CAPwho test positive for influenza in theoutpatient setting, independent of durationof illness before diagnosis (conditionalrecommendation, low quality of evidence).

Summary of the evidence. No clinicaltrials have evaluated the effect of treatmentwith antiinfluenza agents in adults withinfluenza pneumonia, and data are lackingon the benefits of using antiinfluenza agentsin the outpatient setting for patients withCAP who test positive for influenza virus.Several observational studies suggest thattreatment with oseltamivir is associatedwith reduced risk of death in patientshospitalized for CAP who test positive forinfluenza virus (181, 182). Treatmentwithin 2 days of symptom onset orhospitalization may result in the bestoutcomes (183, 184), although there may bebenefits up to 4 or 5 days after symptomsbegin (181, 185).

The use of antiinfluenza agents in theoutpatient setting reduces duration ofsymptoms and the likelihood of lowerrespiratory tract complications amongpatients with influenza (186), with mostbenefit if therapy is received within48 hours after onset of symptoms (187).

Rationale for the recommendation.For inpatients, a substantial body ofobservational evidence suggests thatgiving antiinfluenza agents reducesmortality risk in adults with influenzainfection. Although benefits are strongestwhen therapy is started within 48 hours ofsymptom onset, studies also support startinglater (185). These data underlie our strongrecommendation for using antiinfluenzaagents for patients with CAP and influenza inthe inpatient setting, consistent with therecently published IDSA Influenza ClinicalPractice Guideline (33).

Although we did not identify studiesthat specifically evaluated antiinfluenzaagents for treating outpatients with CAPwho test positive for influenza, we make thesame recommendation as for inpatients, onthe basis of the inpatient data and onoutpatient data showing better time toresolution of symptoms and preventionof hospitalization among those withinfluenza but without pneumonia. Ourrecommendations are consistent with the



http://clinicaltrials.gov

IDSA influenza guidelines, which wererecently released (33).

Research needed in this area.Randomized controlled studies are needed tosupport the recommendation for use ofantiinfluenza agents to treat for influenzapneumonia in the outpatient setting. Inparticular, knowing whether therapy isvaluable when started more the 48 hours aftersymptom onset would help guide clinicaldecision-making.

Question 14: In Adults with CAP WhoTest Positive for Influenza, Should theTreatment Regimen IncludeAntibacterial Therapy?

Recommendation. We recommend thatstandard antibacterial treatment be initiallyprescribed for adults with clinical andradiographic evidence of CAP who testpositive for influenza in the inpatientand outpatient settings (strongrecommendation, low quality of evidence).

Summary of the evidence. Bacterialpneumonia can occur concurrently withinfluenza virus infection or present later asa worsening of symptoms in patients who wererecovering from their primary influenza virusinfection. As many as 10% of patientshospitalized for influenza and bacterialpneumonia die as a result of their infection(188). An autopsy series from the 2009 H1N1influenza pandemic found evidence of bacterialcoinfection in about 30% of deaths (189).

S. aureus is one of the most commonbacterial infections associated with influenzapneumonia, followed by S. pneumoniae, H.influenzae, and group A Streptococcus; otherbacteria have also been implicated (188,190–192). Given this spectrum of pathogens,appropriate agents for initial therapy includethe same agents generally recommended forCAP. Risk factors and need for empiriccoverage for MRSA would follow theguidelines included earlier in this document.Rapidly progressive severe pneumonia withMRSA has been described in previouslyhealthy young patients, particularly in thesetting of prior influenza; however, it istypically readily identified in the nares orsputum and should be identified by followingthe recommendations of earlierrecommendations in this guideline.

Rationale for the recommendation. Therecommendation to routinely prescribeantibacterial agents in patients withinfluenza virus infection and pneumoniawas based on evidence suggesting that

bacterial coinfections are a common andserious complication of influenza, as well asthe inability to exclude the presence ofbacterial coinfection in a patient with CAPwho has a positive test for influenza virus.Although low levels of biomarkers such asprocalcitonin decrease the likelihood thatpatients have bacterial infections, thesebiomarkers do not completely rule outbacterial pneumonia in an individual patientwith sufficient accuracy to justify initiallywithholding antibiotic therapy, especiallyamong patients with severe CAP (37, 38, 193).We have provided a strong recommendationbecause of the significant risk of treatmentfailure in delaying appropriate antibacterialtherapy in patients with CAP. However inpatients with CAP, a positive influenza test,no evidence of a bacterial pathogen(including a low procalcitonin level), andearly clinical stability, consideration could begiven to earlier discontinuation of antibiotictreatment at 48 to 72 hours.

Research needed in this area. Randomizedcontrolled studies are needed to establishwhether antibacterial therapy can bestopped at 48 hours for patients withCAP who test positive for influenza andhave no biomarker (e.g., procalcitonin) ormicrobiological evidence of a concurrentbacterial infection.

Question 15: In Outpatient andInpatient Adults with CAP Who AreImproving, What Is the AppropriateDuration of Antibiotic Treatment?

Recommendation. We recommend that theduration of antibiotic therapy should beguided by a validated measure of clinicalstability (resolution of vital signabnormalities [heart rate, respiratory rate,blood pressure, oxygen saturation, andtemperature], ability to eat, and normalmentation), and antibiotic therapy should becontinued until the patient achieves stabilityand for no less than a total of 5 days (strongrec