Carbapenem-Sparing Antimicrobial Stewardship in Internal Medicine: A Prospective Quasi-Experimental Before-and-After Pilot Implementation Study.

Background/Objectives: Carbapenem overuse contributes to antimicrobial resistance, but pragmatic stewardship models for high-complexity Internal Medicine wards remain underreported. We evaluated the feasibility and stewardship impact of a carbapenem-sparing intervention based on time-limited named-patient dispensing, pharmacist-supported reassessment, and mandatory infectious diseases reassessment for continuation beyond seven days, while assessing short-term clinical outcomes as exploratory safety signals. Methods: We conducted a prospective quasi-experimental before-and-after pilot implementation study in two Internal Medicine wards of a university hospital. Consecutive adult inpatients receiving a carbapenem or fluoroquinolone were enrolled during a 2-month control phase (standard care, n = 40) and a subsequent 2-month intervention phase (n = 38). Initial prescription remained at the discretion of the treating physician; targeted antibiotics were dispensed on a named-patient basis for a maximum initial duration of seven days, with pharmacist-supported reassessment on days 3 and 5 and mandatory infectious diseases reassessment for continuation beyond seven days. Results: Antibiotic therapy duration was shorter in the intervention group (median approximately 6 vs. 9 days, p = 0.0001), and antibiotic discontinuation by day 7 was more frequent than in controls (81.6% vs. 45.0%, p = 0.001). Total targeted antibiotic exposure decreased by approximately 40% (mean 5.9 vs. 10.0 Defined Daily Doses (DDD) per patient, p < 0.001), mainly driven by reduced meropenem use. No significant differences were observed in biomarker trajectories, clinical status at day 7, 30-day readmission or relapse, or 90-day mortality. Conclusions: A time-limited dispensing and reassessment model was feasible in Internal Medicine and was associated with lower carbapenem and targeted restricted-antibiotic exposure. No apparent short-term signal of clinical worsening was identified, although the study was not powered to establish safety, clinical non-inferiority, or equivalence. Larger multicentre studies using standardized stewardship metrics are needed.