A robust stochastic decision-making model for inventory allocation of surgical supplies to reduce logistics costs in hospitals: A case study

In a hospital, surgical supplies can be stored in multiple locations, each of which has limited space and different associated costs. The locations include central storage, where items are retrieved to build a cart of supplies for each procedure; sterile storage adjacent to the operating rooms; and within the operating rooms themselves. In practice, the decision on allocating items to these locations is often based on the staff’s experience, rather than through optimization methods. In this research, we have identified the costs associated with each location to determine where each item should be stored and in what quantities. These costs include the cost of building the case cart, the cost of returning unused items to storage, the cost of picking items during a procedure, the cost of restocking and the cost of reviewing items to determine what needs to be replenished. Since the number of supplies required to perform a procedure is uncertain, we have developed a robust stochastic mixed-integer programming model to make the inventory allocation decision. The model also enables a hospital to assess the potential cost saving from optimization of the preference cards, which are used by surgeons to specify the requested supplies available on the case carts. The performance of the proposed model is evaluated through a case study. Three alternatives to the current configuration of the system are presented and reduction of inventory expenditure within each alternative is discussed. Finally, sensitivity analyses are performed to determine which cost parameters contribute to the model more significantly and how the model behaves against different levels of risk coefficient.