An iterative estimation and validation procedure for specification of semi-Markov models with application to hospital patient flow.

This article presents a methodology to identify and specify a continuous time semi-Markov model of population flow within a network of service facilities. An iterative procedure of state space definition, population disaggregation, and parameter estimation leads to the specification of a model which satisfies the underlying semi-Markov assumptions. We also present a test of the impact of occupancy upon realizations of population flows. The procedure is applied to data describing the movement of obstetric patients in a large university teaching hospital. We use the model to predict length-of-stay distributions. Finally, we compare these results with those that would have been obtained without the procedure, and show the modified model to be superior.