Evidence-based medicine: empiric antibiotic therapy in community-acquired pneumonia.

A number of national society guidelines exist for empiric management of community-acquired pneumonia but these are, to a large extent, not evidence-based, but based on clinical experience, in vitro data, pragmatism and common sense. Many randomized controlled trials of antibiotic therapy in community-acquired pneumonia have been conducted, but most of these have been powered to demonstrate equivalent efficacy of new treatments in comparison with conventional antimicrobial therapy. Development of new antibiotics has been driven by the emergence of penicillin-resistant Streptococcus pneumoniae, but so far there is no hard evidence that beta-lactam therapy fails in community-acquired pneumonia, at least with the higher doses of penicillins that are commonly used in hospital practice. Nonetheless, newer antibiotics have been deployed including macrolides and quinolones, and have demonstrated equivalent (and in some cases, marginally improved) efficacy to older antibiotic treatments in randomized control trials. A number of studies have shown that it is possible to stratify patients according to severity of illness, to in-patient or out-patient management protocols. These have been validated and refined.