Decision support framework for tactical emergency medical service location planning

Abstract

A well-designed emergency medical services (EMS) system is essential to provide the best possible quality of care to patients in emergencies. Exogenous effects, such as demographic change or extreme weather conditions, lead to a continuous increase in emergency calls. On the other hand, emergency vehicles and qualified personnel are limited. Consequently, there is a continuous need to improve the existing EMS systems in order to optimize the use of the available resources. To assist decision makers in this complex task, we provide a decision support framework at the tactical site planning level. First, we analyze the given historical data to find suitable periods that have similar structures in terms of the number of calls and the time of day. We follow this approach to account for fluctuating demand and time-dependent travel times. A new stochastic, time-dependent, mixed-integer extension of the maximum expected coverage location problem (MEXCLP) is developed that accounts for ambulance site interdependencies. We conduct computational experiments for an existing EMS network structure to optimally locate sites and allocate ambulances. Furthermore, we analyze the benefit of additional ambulances and flexible waiting sites. A discrete-event simulation (DES) is used to evaluate the model solutions and to provide further insights into the trade-off between quality and cost of care for the given model solutions. Results based on anonymized real-world EMS data indicate that flexible waiting sites provide considerable potential savings in the use of ambulances while maintaining a high level of coverage.