OBJECTIVETo identify predictors of disagreement with antimicrobial stewardship prospective audit and feedback recommendations (PAFR) at a free-standing children's hospital.DESIGNRetrospective cohort study of audits performed during the antimicrobial stewardship program (ASP) from March 30, 2015, to April 17, 2017.METHODSThe ASP included audits of antimicrobial use and communicated PAFR to the care team, with follow-up on adherence to recommendations. The primary outcome was disagreement with PAFR. Potential predictors for disagreement, including patient-level, antimicrobial, programmatic, and provider-level factors, were assessed using bivariate and multivariate logistic regression models.RESULTSIn total, 4,727 antimicrobial audits were performed during the study period; 1,323 PAFR (28%) and 187 recommendations (15%) were not followed due to disagreement. Providers were more likely to disagree with PAFR when the patient had a gastrointestinal infection (odds ratio [OR], 5.50; 95% confidence interval [CI], 1.99-15.21), febrile neutropenia (OR, 6.14; 95% CI, 2.08-18.12), skin or soft-tissue infections (OR, 6.16; 95% CI, 1.92-19.77), or had been admitted for 31-90 days at the time of the audit (OR, 2.08; 95% CI, 1.36-3.18). The longer the duration since the attending provider had been trained (ie, the more years of experience), the more likely they were to disagree with PAFR recommendations (OR, 1.02; 95% CI, 1.01-1.04).CONCLUSIONSEvaluation of our program confirmed patient-level predictors of PAFR disagreement and identified additional programmatic and provider-level factors, including years of attending experience. Stewardship interventions focused on specific diagnoses and antimicrobials are unlikely to result in programmatic success unless these factors are also addressed.Infect Control Hosp Epidemiol 2018;806-813.