Fluoroquinolone treatment of community-acquired pneumonia: a meta-analysis.

Abstract

To determine the role of newer fluoroquinolones (FQs) for adults with community-acquired pneumonia (CAP) whose level of illness allows treatment with an oral antibiotic. Meta-analysis of randomized controlled trials comparing a macrolide, beta-lactam, or doxycycline antibiotic with a newer oral FQ for the treatment of CAP. Patients (5118), most of whom were <60 years of age and free of coexisting diseases, were enrolled in 13 studies comparing an oral macrolide or beta-lactam antibiotic with an FQ for the treatment of CAP. No previous study compared doxycycline with an FQ. In the intention-to-treat (ITT) population, no trial demonstrated significant differences between FQs or alternative therapies. Summary estimates showed a statistically significant advantage in favor of the FQs in both the ITT (OR 1.22; 95% CI 1.02 to 1.47; p = 0.03) and evaluable populations (OR 1.37; 95% CI 1.11 to 1.68; p = 0.003). The number needed to treat for an FQ advantage was 33 (95% CI 17 to 362) in the ITT population and 37 (95% CI 22 to 121) in the evaluable population. Treatment failures represented slow symptom resolution; no deaths were reported. The newer oral FQs showed modest therapeutic benefit compared with the studied alternative antibiotics in adults with CAP. Based on the number needed to treat from the ITT population as a measure of treatment effect, clinicians must decide whether treating 33 patients with an FQ to prevent a single therapeutic failure with another studied antibiotic warrants use of an agent from that class for an illness with a generally favorable outcome regardless of antibiotic selection, and at a time when FQ resistance may be increasing.