Designing a multi-service healthcare network based on the impact of patients’ flow among medical services

Efficient location of medical services is an issue of paramount importance in healthcare strategic planning. In this research, a mathematical model is developed for the location of multi-service health centers, assuming probabilistic demand and service time. Since patients may be shifted to another service after receiving a service by doctors’ order, health system is considered a Jackson queueing network. We assume that patients have “probabilistic choice behavior” and the primary factors contributing to their choice of one center over another are their proximity to the center and the number of medical services offered by the center. The proposed mixed integer nonlinear programming model seeks to minimize the demand weighted total distance travelled by patients among their residential areas and health centers and also among health centers on the one hand, and the weighted sum of undesired deviations from standard arrival rates at service stations on the other hand. The location of health centers as well as the type of services they offer and the number of servers at each service station are the main determinants of the proposed model. Using the proposed model, we can predict patients’ choice patterns and their arrival rates at current or newly provided medical stations. Two heuristics are developed to solve medium and large instances of the proposed model. The computational results obtained from benchmark instances show that the GA-based heuristic is somewhat better than the heuristic based on remove–add-–exchange procedures.