A Data-driven queueing approach for capacity planning of a burn center

Abstract

A data-driven queueing model is developed to analyze burn patient flow at a regional Burn Center in the US. Empirical data is used to identify and fit arrival and service time distributions, which are then incorporated into infinite-server queueing models. Unlike previous simulation-based studies, our approach yields closed-form expressions for key performance metrics and allows rapid evaluation of planning scenarios. The model is validated against observed system data and used to predict occupancy, coverage, census, and the probability of full capacity under varying patient arrival rates and bed allocations. We also examine the impact of reducing unnecessary transfers of minor burn cases. The proposed approach provides an analytically tractable and empirically grounded tool for planning and policy evaluation in specialized healthcare settings.