Response time, i.e., the time from dispatch of an ambulance to its arrival at the scene of an emergency, is an important measure of performance in an urban ambulance system. We developed a model that predicts the entire distribution of response time, explicitly accounting for the rate and spatial distribution of demand, variable ambulance velocities, and queueing effects. We tested the model using data sampled from 3,936 ambulance runs in Houston and achieved dose agreement between empirical and predicted distributions of response time. Our use of probability theory to predict response times yielded a model that complements those previously reported for planning and evaluating urban ambulance systems.
By David W. Scott, Lynette E. Factor, and G. Anthony Gorry