Ambulance dispatching and operating room scheduling considering reusable resources in mass-casualty incidents

Mass casualty incidents (MCIs) impose heavy demands on emergency response capabilities. It is essential to schedule limited emergency resources (e.g., ambulances, operating rooms, medical personnel, etc.) efficiently in order to maximize lifesaving capabilities. Most previous studies only consider a single type of emergency resource, which may result in inefficient coordination or bottlenecks between various types of resources. This paper addresses the combined problem of ambulance dispatching and operating room scheduling during an MCI. Specifically, we consider the limited ambulances and emergency operating rooms as reusable resources, more accurately reflecting reality than previous research. A novel mixed integer programming model aims to maximize the number of patients who can undergo surgery before their critical surgery time. A hybrid algorithm combining the Tabu Search and an Adaptive Large Neighborhood Search with five new specific removal operators is proposed for solving large-scale instances. The model is solved by Gurobi, and the results are compared with the proposed TS-ALNS. On small-scale instances, TS-ALNS is comparable to Gurobi, while on large-scale instances, it outperforms other meta-heuristics, including the Tabu search algorithm, the adaptive large neighborhood search algorithm, the simulated annealing algorithm, and the variable neighborhood search algorithm. In addition, the utilization and effectiveness of the five proposed removal operators are analyzed.