Abstract
BACKGROUND: There is an urgent need to optimize antimicrobial prescribing in the hospitalized setting, driven by the complexity of infection types, the high risk of antimicrobial resistance, and the potentially severe clinical consequences. However, the key determinants of physician prescribing behavior remain inconsistent, and the evidence regarding the effectiveness of interventions is still subject to debate. METHODS: We searched PubMed, Embase, Cochrane Library, and Web of Science up to July 5, 2025, to identify evidence on determinants of and interventions for antibiotic prescribing practices in the inpatient environment. Through a single-arm 3-level model, we quantified the influence of these determinants on prescription behavior. We employed a random-effects model to analyze the effect of interventions on prescription outcomes. Interventions were categorized by behavior change techniques, with the effectiveness rate calculated. RESULTS: A total of 59 studies were included, comprising 20 qualitative and 39 quantitative investigations. The findings indicated that 77.6%, 71.4%, and 64.2% of participants acknowledged the influence of environmental, prescriber, and patient factors, respectively. Interventions were associated with a 21% (RR = 1.21, 95% CI: 1.03-1.42) improvement in rational antimicrobial prescribing. Analysis of behavior change techniques identified "behavior feedback" as the most effective strategy (effectiveness rate = 3.5). CONCLUSION: Our study shows that hospitalized physicians' antibiotic prescribing is shaped by multiple determinants, with contextual and environmental factors most frequently studied. Interventions generally improved prescribing in hospital settings. However, evidence from resource-limited settings remains sparse; rigorous, context-specific studies are needed to optimize prescribing in low- and middle-income countries.