Efficacy of Oral β-Lactam versus Non-β-Lactam Treatment of Uncomplicated Cellulitis

BackgroundPreferred therapy for purulent skin and soft tissue infections is incision and drainage, but many infections cannot be drained. Empiric therapies for these infections are ill-defined in the era of community-acquired methicillin-resistant Staphylococcus aureus.MethodsA multicenter retrospective cohort study of outpatients treated for cellulitis was conducted to compare clinical failure rates of oral β-lactam and non-β-lactam treatments. Exclusion criteria included purulent infection requiring incision and drainage, complicated skin and soft tissue infection, chronic ulceration, and intravenous antibiotics. Failure rates were compared using logistic regression to adjust for both covariates associated with failure and a propensity score for β-lactam treatment.ResultsOf 2977 patients, 861 met inclusion criteria and were classified by treatment: β-lactam (n = 631) or non-β-lactam therapy (n = 230). Failure rates were 14.7% versus 17.0% (odds ratio [OR] 0.85, 95% confidence interval [CI], 0.56-1.31) for β-lactam and non-β-lactam therapy, respectively. Failure was associated with: age (P = .02), acute symptom severity (P = .03), animal bites (P = .03), Charlson score >3 (P = .02), and histamine-2 receptor antagonist use (P = .09). Relative efficacy of β-lactam therapy was greater after adjustment for factors associated with failure but remained statistically insignificant (adjusted OR 0.81, 95% CI, 0.53-1.24); adjusted including propensity score covariate (OR 0.71, 95% CI, 0.45-1.13). Discontinuation due to adverse effects differed between β-lactam (0.5%) and non-β-lactam (2.2%) therapies (P = .04).ConclusionThere was no significant difference in clinical failure between β-lactam and non-β-lactam antibiotics for the treatment of uncomplicated cellulitis. Increased discontinuation due to adverse events with non-β-lactam therapy was observed.