Development and real-world application of a consumption-based model for projecting the budget impact of high-cost reserve in a University Hospital

Objectives Traditional budget impact analysis models, usually based on predefined populations, often fail to capture the complexity of hospital antibiotic use, which is dynamic and often empirical. To address this gap, we developed and applied an adaptable budget impact analysis model based on real-world consumption data of high-cost reserve antibiotics (HCR-ABX), with the objective of projecting their short-term budget impact in a large university hospital over a 3-year horizon. Methods HCR-ABX with a purchase cost >€100/defined daily doses (DDD) were selected. The model used retrospective hospital pharmacy data (2021-2024) and incorporated national guidelines on multidrug-resistant infections, regulatory/reimbursement updates, and local resistance trends. These inputs were used to define drug-specific annual variation rates applied in the projection scenarios. Future costs were estimated by combining projected consumption (DDD/100 hospitalization days) with the hospital purchase cost per DDD. Three forecast scenarios (base case, lower impact, higher impact) plus a Klebsiella pneumoniae stress-test scenario were modeled. Results Despite approximately stable overall HCR-ABX consumption, a shift toward higher-cost agents was observed. The base-case scenario projected a cumulative expenditure of approximately €10 586 248 over 3 years, with increases driven by meropenem/Vaborbactam and cefiderocol. The lower-impact and higher-impact scenarios estimated €10 062 688 and €11 050 385 €, respectively. The Klebsiella pneumoniae stress-test scenario indicated a total cost of € 11 321 117 (+6.95% vs base-case scenario), emphasizing the potential financial impact of epidemiological shifts. Conclusions The proposed model provides a pragmatic, adaptable framework to assess the hospital budget impact of HCR-ABX and supports antimicrobial stewardship. The approach is replicable and can be refined through future validation against observed data.