Validating and applying a model for locating emergency medical vehicles in Tucson, Arizona

This paper deals with the problem of locating emergency medical vehicles in Tucson, Arizona. The model is based on a general service time approximation model for spatially distributed queueing systems. Provisions for unequal vehicle utilizations, stochastic travel times, and multiple call classes are included. The model is tailored for emergency medical systems that experience low vehicle utilizations. Validation of the model using the Tucson system is discussed in detail. The predictive ability of the model is demonstrated by comparing model output with actual system data. The model has been used to evaluate potential changes in the paramedic services provided by the Tucson Emergency Medical System and two cases are discussed.