Reinstating and augmenting prospective audit and feedback for carbapenem stewardship: A quasi-experimental interrupted time-series study

BACKGROUND: At a tertiary hospital in Korea, carbapenem stewardship was suspended in March 2024 during a nationwide medical strike, leading to increased carbapenem use. In December 2024, stewardship activities were re-initiated using prospective audit and feedback (PAF), and in January 2025, an augmented PAF strategy incorporating ward round-based feedback was introduced. METHODS: We conducted a quasi-experimental study at a 2400-bed tertiary care hospital in Seoul, South Korea. An interrupted time-series (ITS) analysis evaluated the effects of resuming carbapenem stewardship using PAF on monthly carbapenem days of therapy (DOT) per 1000 patient-days. Prescriber adherence was compared between standard PAF in December 2024 and augmented PAF from January to August 2025. Adherence was defined as discontinuation or modification of carbapenem therapy within 7 days of the initial "not recommended" feedback. RESULTS: A total of 896,495 patient-days were included. Following resumption of carbapenem stewardship using PAF in December 2024, carbapenem DOT showed an immediate level reduction (-17.7 DOT per 1000 patient-days; 95% CI, -26.2 to -9.25; p < 0.001) and a significant slope change (-2.89 per month; 95% CI, -3.84 to -1.93; p < 0.001), with an overall 35.7% reduction compared with the counterfactual projection. Prescriber adherence was higher during augmented PAF than standard PAF (58.4% [97/166] vs. 25.0% [6/24]; risk ratio, 2.34; 95% CI, 1.16-4.73; p = 0.0036). CONCLUSION: Resumption of carbapenem stewardship using PAF after suspension was associated with a substantial reduction in carbapenem use. Adherence appeared higher during augmented PAF incorporating ward round-based follow-up feedback than during standard PAF, suggesting the potential value of active, follow-up-driven stewardship interventions.