Multi-Objective Distributed Operating Room Scheduling in Collaborative Hospital Networks: Balancing Operating Costs, Patient Deterioration, and Patient Utility

Collaborative hospital networks are hospital groups that collaborate to provide healthcare services and share medical resources. They are an important part of the accessibility and efficiency of healthcare delivery. Nevertheless, existing research pays limited attention to multi-hospital scenarios. Furthermore, most studies adopt optimization objectives from the perspective of healthcare providers rather than prioritizing patient-centered service quality. Accordingly, this paper investigates the multi-objective distributed operating room scheduling problem (MDORSP), with the objectives of simultaneously minimizing operating costs and patient deterioration and maximizing patient utility. A multi-objective mixed-integer linear programming model is formulated for the MDORSP. Within this model, to embed patient-centered healthcare considerations, heterogeneous patient preferences for surgeons of varying professional qualifications are quantified based on random utility theory. To tackle this NP-hard MDORSP, a problem-tailored multi-objective evolutionary algorithm is developed. It incorporates priority-based chromosome representation, a heuristic initialization strategy, custom decoding and repair mechanisms, and a refined selection strategy. Extensive computational experiments show that the proposed algorithm obtains higher HV values than TLMOVNS on all benchmark instances and surpasses CPLEX in 86.7% of instances, with its average computational runtime being 78.6% and 85.9% less than TLMOVNS and CPLEX, respectively. The findings underscore the effectiveness of collaborative resource coordination in reconciling operational efficiency and patient-centered service quality.