Practice model of unit-based clinical pharmacists’ individualized daily antimicrobial use density monitoring report on antimicrobial stewardship in intensive care unit of a tertiary hospital in Guangxi, China: an interrupted time series analysis

BACKGROUND: High antimicrobial exposure and antimicrobial resistance in intensive care units (ICUs) remain major challenges to patient safety. This study evaluated whether a unit-based clinical pharmacist (UBCP) model supported by an individualized daily antimicrobial use density monitoring report (IAUD-RP) could improve antimicrobial stewardship in an ICU. METHODS: This single-center, retrospective, quasi-experimental study used interrupted time series analysis in a 12-bed ICU of a tertiary teaching hospital in Guangxi, China. Adult patients admitted between April 1, 2023, and October 31, 2025, were included. The intervention, initiated in August 2024, consisted of UBCP-led daily ward-wide screening, real-time risk stratification, and targeted pharmacist interventions using the individualized monitoring report. The primary outcome was antimicrobial use measured as defined daily doses (DDDs) per 100 patient-days. Secondary outcomes included average antimicrobial cost per hospitalization, multidrug-resistant organism healthcare-associated infection incidence density, clinical outcomes, and changes in specific antimicrobial agents or classes. RESULTS: A total of 657 patients were included (295 before, 362 after). UBCP recommendations achieved a 91.7% acceptance rate. Interrupted time series analysis showed a significant immediate reduction in antimicrobial use (level change, - 29.0 DDDs/100 patient-days; P = 0.038), following a significant pre-intervention upward trend (+ 2.5 per month; P = 0.005). Interrupted time series analysis showed a significant immediate reduction in average antimicrobial cost per hospitalization (level change, - 8304 CNY; P = 0.035), consistent with the crude reduction (25,568 to 14,926 CNY; P < 0.001). Total antimicrobial consumption decreased by 11.9%. Tigecycline, quinolones, and carbapenems decreased by 52.1%, 39.7%, and 15.8%, respectively, whereas WHO Access-group antibacterial agents increased by 67.8%. After excluding patients with indeterminate ('Others') outcomes, clinical failure was lower post-intervention (13.0% vs. 19.2%; adjusted OR 0.62, 95% CI 0.39-0.99; P = 0.045). CONCLUSION: The UBCP-led IAUD-RP model was associated with a significant and sustained reduction in antimicrobial use density, a directionally favorable change in antimicrobial cost and prescribing pattern, and a directionally favorable but non-confirmatory signal for reduced clinical failure.