Beta-lactam plus macrolides or beta-lactam alone for community-acquired pneumonia: A systematic review and meta-analysis.

It is unclear whether in the treatment of community-acquired pneumonia (CAP) beta-lactam plus macrolide antibiotics lead to better survival than beta-lactam alone. We report a systematic review and meta-analysis. Trials and observational studies published in English were included, if they provided sufficient data on odds ratio for all-cause mortality for a beta-lactam plus macrolide regimen compared with beta-lactam alone. Two investigators independently searched for eligible articles. Of 514 articles screened, 14 were included: two open-label randomized controlled trials (RCTs) comprising 1975 patients, one non-RCT interventional study comprising 1011 patients and 11 observational studies comprising 33 332 patients. Random-model meta-analysis yielded an odds ratio for all-cause death for beta-lactam plus macrolide compared with beta-lactam alone of 0.80 (95% CI 0.69-0.92, P = 0.002) with substantial heterogeneity (I(2)  = 59%, P for heterogeneity = 0.002). Severity-based subgroup analysis and meta-regression revealed that adding macrolide had a favourable effect on mortality only for severe CAP. Of the two RCTs, one suggested that macrolide plus beta-lactam lead to better outcome compared with beta-lactam alone, while the other did not. Subgrouping based on study design, that is, RCT versus non-RCT, which was almost identical to subgrouping based on severity, revealed substantial inter-subgroup heterogeneity. Compared with beta-lactam alone, beta-lactam plus macrolide may decrease all-cause death only for severe CAP. However, this conclusion is tentative because this was based mainly on observational studies.