Modeling framework and reference model for data-driven simulation of patient flows in emergency departments.

Discrete event simulation (DES) of patient flows is essential for the design of emergency departments (EDs), because it enables exploration of scenarios without disrupting real systems. Although DES is beneficial, it is limited by high modeling requirements. In simulations, each scenario to be analyzed must be individually modeled, which consumes significant time and resources. Moreover, models can rapidly become obsolete. This can be caused by changes in patient population, such as an increase in the number of cases during flu season and other seasonal variations. This study explored how to efficiently support simulation modeling from conceptual modeling until specific programmed models to overcome these challenges, thereby enabling consideration of real patient populations in ED simulations. A modeling framework for healthcare simulation and reference model were developed for managing patient population data, generating simulation data using Monte Carlo simulation, and conducting simulation experiments using data-driven simulation techniques. The framework supports conceptual modeling using a detailed set of steps and modeling views, thereby allowing to separate modeling concerns throughout the process. It also integrates an instantiation mechanism to build specific programmed models based on a reference model. A real-life study of an ED demonstrated the developed approach based on two illustrative examples.