Antibiotic therapy (AT) is the cornerstone of the management of severe community-acquired pneumonia (CAP). However, the best treatment strategy is far from being established.
To evaluate the impact of different aspects of AT on the outcome of critically ill patients with CAP, we performed a post hoc analysis of all CAP patients enrolled in a prospective, observational, multicentre study.
Of the 502 patients included, 76% received combination therapy, mainly a β-lactam with a macrolide (80%). AT was inappropriate in 16% of all microbiologically documented CAP (n = 177). Hospital and 6 months mortality were 34% and 35%. In adjusted multivariate logistic regression analysis, combination AT with a macrolide was independently associated with a reduction in hospital (OR 0.17, 95%CI 0.06–0.51) and 6 months (OR 0.21, 95%CI 0.07–0.57) mortality. Prolonged AT (>7 days) was associated with a longer ICU (14 vs. 7 days; p < 0.001) and hospital length of stay (LOS) (25 vs. 17 days; p < 0.001).
Combination AT with a macrolide may be the most suitable AT strategy to improve both short and long term outcome of severe CAP patients. AT >7 days had no survival benefit and was associated with a longer LOS.