Abstract
BACKGROUND: The effectiveness of antimicrobial stewardship programs may vary by institutional context. We evaluated whether hospital size modifies the impact of nationally incentivized antimicrobial stewardship programs on clinical and economic outcomes in patients with methicillin-resistant Staphylococcus aureus bloodstream infections. METHODS: This retrospective cohort study analyzed Japanese nationwide claims data from 2018 to 2022. We compared outcomes between adult inpatients with methicillin-resistant Staphylococcus aureus bloodstream infections in large (≥ 500 beds) versus non-large (< 500 beds) hospitals using 1:1 propensity score matching. The primary outcome was time to discharge. We used an interaction term to assess effect modification by hospital size on the association between antimicrobial stewardship programs and outcomes. RESULTS: The matched cohort included 926 patients (463 pairs). At baseline, average daily total antimicrobial costs were significantly higher in large hospitals (rate ratio 1.32, 95% CI 1.16–1.50). Hospital size significantly modified the association between antimicrobial stewardship programs and time to discharge (P for interaction = 0.01). In non-large hospitals, antimicrobial stewardship programs were associated with significantly shorter time to discharge (hazard ratio 0.31, 95% CI 0.13–0.75). In large hospitals, antimicrobial stewardship programs were not associated with time to discharge but were associated with significantly reduced average daily antipseudomonal drug costs (rate ratio 0.46, 95% CI 0.30–0.71). CONCLUSIONS: Hospital size modified the association between antimicrobial stewardship program fee acquisition and outcomes among patients with methicillin-resistant Staphylococcus aureus bloodstream infections. In non-large hospitals, antimicrobial stewardship program fee acquisition was associated with shorter time to discharge, whereas in large hospitals, it was primarily associated with changes in prescribing patterns. These findings suggest that stewardship evaluation and policy should be tailored to institutional characteristics. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s40780-026-00558-2.