Simulation for triage

This paper aims to develop a framework for real-time simulation-assisted patient triage in emergency departments (EDs). The framework utilizes discrete event simulation (DES) to schedule and plan the treatment order of patients who are waiting in the ED. Triage is utilized to rapidly categorize and prioritize patients according to their injuries or illnesses. The most widely used triage algorithm in the US is called Emergency Severity Index (ESI). With this algorithm, patients are treated based on their acuity level, where ESI level 1 patients are treated first followed by ESI level 2 and so on. A major drawback of the ESI algorithm is that the algorithm assumes no, or ignores the fact of, variation or uncertainty in the system. In this paper, the simulation model helps nurses make better triage decisions by integrating a deterministic schedule, which results from triage, with risk measures that account for system variation and uncertainty. The risk measure in this model is the probability of treating a patient on time. This probability can also be used as a multi-dimensional measure of ED overcrowding by which ambulance diversions or other mitigation actions are considered.