Placebo-controlled trials of treatments for community-acquired pneumonia: review of the literature and discussion of feasibility and potential value.

Abstract

Community-acquired pneumonia (CAP) has a broad range of clinical manifestations, from apparently self-limited illness to severe disease. This observation raises the question of whether to perform placebo-controlled trials involving patients with mild CAP to determine quantitatively the treatment effect of antibiotic therapy for CAP. A critical analysis of the literature reveals that most clinical trials underestimate the proportion of CAP cases of all severities that are caused by typical bacteria, particularly pneumococci. Studies that use vigorous diagnostic methods consistently report that the majority of CAP cases are caused by pneumococci. Currently, there is little evidence that viruses cause a substantial proportion of CAP cases in adults. Although relevant placebo-controlled trials have not been published in recent years, other evidence indicates that antibiotic therapy is unequivocally effective for pneumococcal pneumonia. The availability of effective therapy for Streptococcus pneumoniae pneumonia presents clinical, ethical, and pragmatic obstacles to the performance of placebo-controlled trials. Furthermore, if placebo-controlled trials were performed for mild CAP, patients with risk factors for pneumococcal pneumonia and patients with predictors of systemic illness would be excluded, thereby resulting in a highly selected population. On the basis of the data reviewed, there is no justification for placebo-controlled trials involving patients with CAP of any degree of severity.