RESUMO
BACKGROUND: At our institution, empirical vancomycin is overused in children with suspected bacterial community-acquired infections (CAIs) admitted to the PICU because of high community rates of methicillin-resistant Staphylococcus aureus (MRSA). Our goal was to reduce unnecessary vancomycin use for CAIs in the PICU. METHODS: Empirical PICU vancomycin indications for suspected CAIs were developed by using epidemiological risk factors for MRSA. We aimed to reduce empirical PICU vancomycin use in CAIs by 30%. After retrospectively testing, the indications were implemented and monthly PICU empirical vancomycin use during baseline (May 2017-April 2018) and postintervention (May 2018-July 2019) periods. Education was provided to PICU providers, vancomycin indications were posted, and the antibiotic order set was revised. Statistical process control methods tracked improvement over time. Proven S aureus infections for which vancomycin was not empirically prescribed and linezolid or clindamycin use were balancing measures. RESULTS: We identified 1620 PICU patients with suspected bacterial CAIs. Empirical vancomycin decreased from a baseline of 73% to 45%, a 38% relative reduction. No patient not prescribed empirical vancomycin later required the addition of vancomycin or other MRSA-targeted antibiotics. There was no change in nephrotoxicity or in the balancing measures. CONCLUSIONS: Development of clear and concise recommendations, combined with clinician education and decision support via an order set, was an effective and safe strategy to reduce PICU vancomycin use. Retrospective validation of the recommendations with local data were key to obtaining PICU clinician buy in.