Simulating cardiac arrest events to evaluate novel emergency response systems

This paper presents a simulation model approach to predict improvements in survival by new out-of-hospital cardiac arrest response systems. Poor cardiac arrest survival rates have motivated the exploration of new response system concepts to augment EMS systems, including citizen responders dispatched by a cell phone app, and the use of drones to deliver an AED to a cardiac arrest location. With few existing studies, the system effectiveness remains largely unknown. A predictive model was developed to understand the improvement these systems may have on cardiac arrest survival. The model uses a geospatial Monte Carlo sampling approach to simulate the random locations of cardiac arrests and the responding agents. The model predicts the response time of EMS, mobile dispatched responders, and drone AED delivery, based on the distance traveled and the mode of transit, while accounting for additional non-transit system factors. A logistic regression model is utilized to translate response times for CPR and defibrillation to a likelihood of survival. The model was developed to simulate and compare multiple response system concepts. The paper presents a case study to demonstrate the model’s utility.