A MILP-based decision support system for an emergency hospital pharmacy

Purpose This study aims to examine a two-echelon, multi-product, multi-period pharmaceutical supply chain in Tunisia, involving a public hospital pharmacy and an emergency pharmacy. The emergency pharmacy’s inventory is represented using a mixed integer linear programming (MILP) model, aiming to minimize costs of ordering, holding and handling shortages or overstock. Design/methodology/approach With an estimated cost benefit of 18%−30%, the MILP model successfully lowers the expense of managing an emergency pharmacy’s drug inventory. Using economies of scale in scenarios with higher demand demonstrates the model’s scalability and cost-effectiveness. Stockout penalties, however, limit cost reductions, underscoring its adaptability to changing demand and cost conditions. A decision support system (DSS) is introduced to optimize drug inventory management in the hospital emergency room, ensuring pharmaceutical availability and reducing operating expenses through mathematical optimization. The suggested DSS is a prime example of how interactive tools and operational research may improve healthcare logistics while striking a balance between essential service requirements and cost effectiveness. Findings With an estimated cost benefit of 18%−30%, the MILP model successfully lowers the expense of managing an emergency pharmacy’s drug inventory. DSS is introduced to optimize drug inventory management in the hospital emergency room, ensuring pharmaceutical availability and reducing operating expenses through mathematical optimization. The suggested DSS is a prime example of how interactive tools and operational research may improve healthcare logistics while striking a balance between essential service requirements and cost effectiveness. Originality/value The DSS incorporates three critical functional components to optimize medication management. First, the use of the MILP, exact approach, to determine the optimal reorder point and the ordered quantity for different medication classes. Second, comprehensive performance indicators provide real-time monitoring of: stock status relative to safety thresholds, days of remaining coverage based on consumption patterns and detailed cost analytics of inventory holding versus procurement. Third, the system demonstrates robust workflow integration, seamlessly interfacing with existing hospital operations through: bidirectional data exchange with pharmacy management systems and support for two-echelon distribution networks spanning hospital pharmacy and emergency pharmacy. This tripartite architecture ensures both theoretical rigor in inventory optimization and practical adaptability to clinical workflows.