Retrospective cohort study of inappropriate piperacillin-tazobactam use for lower respiratory tract and skin and soft tissue infections: Opportunities for antimicrobial stewardship

• We studied use of piperacillin-tazobactam for 2 common infections.
• This broad-spectrum agent is overused for treatment of mild or moderate infections.
• Duration of antibiotic therapy for studied infections was also frequently excessive.
• Antimicrobial stewardship interventions have potential to reduce antibiotic overuse.

BackgroundPatients with skin and skin structure infections (SSTIs) and lower respiratory tract infections (LRTIs) are frequently prescribed piperacillin-tazobactam (TZP) on hospital admission. Inappropriate broad-spectrum coverage may be associated with patient harm, excess expenditure, and escalating rates of antimicrobial resistance.MethodsPatients who received empirical TZP for a diagnosis of LRTI or SSTI from January 1-June 30, 2012, were identified retrospectively. Clinical and antimicrobial data were systematically collected from electronic hospital information systems. Using published guidelines, microbiologic results, and individual clinical responses, the appropriateness of TZP use was assessed. Drug utilization after potential standard audit of therapy on day 3 was also evaluated.ResultsWe reviewed 60 patients with SSTI and 169 patients with LRTI. Inappropriate empirical TZP therapy was found in 41.7% in those with SSTI, and a further 15% had inappropriate continuation of therapy. In LRTI patients, 38.3% received inappropriate empirical TZP, and 10.3% of the treatment courses were continued inappropriately. Community-acquired pneumonia was the most frequent diagnosis where TZP was used inappropriately (96%). A day 3 audit of therapy may have saved 256 days of TZP.ConclusionIn our institution, inappropriate empirical TZP is common for community-onset infections of mild to moderate severity. A prospective audit and feedback program may be a strategy to reduce inappropriate use of TZP as empirical therapy.