Macrolide-Based Regimens and Mortality in Hospitalized Patients With Community-Acquired Pneumonia: A Systematic Review and Meta-analysis

MAJOR ARTICLE Macrolide-Based Regimens and Mortality in Hospitalized Patients With CommunityAcquired Pneumonia: A Systematic Review and Meta-analysis Leyla Asadi,1,a Wendy I. Sligl,1,2,a Dean T. Eurich,3 Isabelle N. Colmers,3 Lisa Tjosvold,4 Thomas J. Marrie,6 and Sumit R. Majumdar5 1 Division of Infectious Diseases, 2Division of Critical Care Medicine, 3Department of Public Health Sciences, School of Public Health, 4John W. Scott Health Sciences Library, and 5Department of Medicine, Faculty of Medicine and Dentistry, University of Alberta, Edmonton; and 6Department of Medicine, Faculty of Medicine, Dalhousie University, Halifax, Nova Scotia, Canada Received 12 February 2012; accepted 26 March 2012; electronically published 16 April 2012. a L. A. and W. I. S. contributed equally to this work. Correspondence: Wendy I. Sligl, MD, MSc, Divisions of Infectious Diseases and Critical Care Medicine, University of Alberta, 3C1.12 Walter Mackenzie Centre, University of Alberta Hospital, 8440-112th St, Edmonton, Alberta, Canada T6G 2B7 (). Clinical Infectious Diseases 2012;55(3):371–80 © The Author 2012. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail: . DOI: 10.1093/cid/cis414 Community acquired pneumonia (CAP), combined with influenza, is the eighth leading cause of death in Canada and the United States and the leading cause of hospitalization due to an infectious disease [1, 2]. For patients hospitalized with CAP, 30-day mortality rates are as high as 23% [1] and annual expenditures are $8–$10 billion in the United States alone [3]. Some previous studies have shown reduced morbidity and mortality in CAP patients treated with macrolide-based Macrolides and Mortality in Pneumonia • CID 2012:55 (1 August) • 371 Background. Macrolides are used to treat pneumonia despite increasing antimicrobial resistance. However, the immunomodulatory properties of macrolides may have a favorable effect on pneumonia outcomes. Therefore, we systematically reviewed all studies of macrolide use and mortality among patients hospitalized with community-acquired pneumonia (CAP). Methods. All randomized control trials (RCTs) and observational studies comparing macrolides to other treatment regimens in adults hospitalized with CAP were identified through electronic databases and gray literature searches. Primary analysis examined any macrolide use and mortality; secondary analysis compared Infectious Diseases Society of America/American Thoracic Society guideline-concordant macrolide/beta-lactam combinations vs respiratory fluoroquinolones. Random effects models were used to generate pooled risk ratios (RRs) and evaluate heterogeneity (I 2). Results. We included 23 studies and 137 574 patients. Overall, macrolide use was associated with a statistically significant mortality reduction compared with nonmacrolide use (3.7% [1738 of 47 071] vs 6.5% [5861 of 90 503]; RR, 0.78; 95% confidence interval [CI], .64–.95; P = .01; I 2 = 85%). There was no survival advantage and heterogeneity was reduced when analyses were restricted to RCTs (4.6% [22 of 479] vs 4.1% [25 of 613]; RR, 1.13; 95% CI, .65–1.98; P = .66; I 2 = 0%) or to patients treated with guideline-concordant antibiotics (macrolide/beta-lactam, 5.3% [297 of 5574] vs respiratory fluoroquinolones, 5.8% [408 of 7050]; RR, 1.17; 95% CI, .91–1.50; P = .22; I 2 = 43%). Conclusions. In hospitalized patients with CAP, macrolide-based regimens were associated with a significant 22% reduction in mortality compared with nonmacrolides; however, this benefit did not extend to patients studied in RCTs or patients that received guideline-concordant antibiotics. Our findings suggest guideline concordance is more important than choice of antibiotic when treating CAP. METHODS Although it was not registered, the protocol for this study was developed according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines [12]. trials registry ClinicalTrials.gov. Finally, we consulted with content experts and contacted authors of studies that might have data appropriate for our analysis. For the latter, we attempted up to 3 contacts with the corresponding (first and senior) author before considering them nonresponsive. Study Selection A checklist was used to assess whether studies met our inclusion criteria for population (hospitalized patients with CAP), exposure (macrolide antibiotic), comparison group (nonmacrolide antibiotic), outcome (mortality, even if it was not a primary or secondary outcome of the included study), and study design (randomized control trials [RCTs] and observational cohort studies). Exclusion criteria eliminated non-English or duplicate reports and studies on outpatients, critically ill patients, immunocompromised patients, or patients identified as having some form of healthcare-associated pneumonia (HCAP). Data Collection Two trained reviewers independently conducted study selection, abstracted data, and assessed the risk of bias (L. A. and W. I. S.). Discrepancies between reviewers were resolved through discussion and consensus; if consensus could not be reached, discrepancies were resolved by S. R. M. Risk of bias was evaluated as low, unclear, or high using the Cochrane risk of bias tool for RCTs [14]. For cohort studies, risk of bias was assessed using a modified version of the Newcastle– Ottawa scale that accorded a maximum of 8 points to each study, with <5 points indicating a high risk of bias [15]. Search Strategy Sources of Heterogeneity An experienced librarian (L. T.) helped us conduct a comprehensive search of the following key electronic biomedical databases from inception through December 2011: Medline, Embase, Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effects, Health Technology Assessments, Cochrane Central Register of Controlled Trials, Science Citation Index Expanded, Conference Proceedings Citation Index—Science, BIOSIS Previews, and Scopus. A modification of the Cochrane highly sensitive search strategy for identifying randomized trials [13] and study design filters from BMJ Clinical Evidence were applied in Medline and Embase. All available years from 1996 onward were searched without language restrictions. The search strategy is provided in the Supplementary Data. In addition to electronic databases, we explored sources of gray literature, including the latest 4 years (2008–2011) of proceedings from the Infectious Diseases Society of America (IDSA), American Thoracic Society (ATS), Interscience Conference on Antimicrobial Agents and Chemotherapy Congress of Microbiology and Infectious Disease, the European Congress of Microbiology and Infectious Diseases, and the clinical Potential sources of heterogeneity were considered a priori. These included study design (RCTs vs observational studies) and source of patient population (large administrative database studies vs clinical (...truncated)