Exploring the System Dynamics of Covid-19 in Emergency Medical Services

Abstract

Emergency Medical Services (EMS) are essential to the healthcare system as they maximize the overall expected survival probability of patients. During the COVID-19 pandemic, peoples’ lifestyle changed and their decisions to seek medical assistance were mixed with fear. Similarly, EMS systems needed to take extra precautions in terms of personal protective equipment protocols. These changes created variability in both demand levels and the response times. In this context, this research presents descriptive and predictive analysis to fully explore the Covid-19 impact on EMS in Lebanon. The descriptive analysis focuses on the changes in call volumes and response times during the COVID-19 pandemic compared to both other countries and previous years. Results show that the number of calls and number of missions dropped yet, the emergency response time was higher and more variable than in previous years. The predictive analysis yielded a model of response times for emergency missions through machine learning, specifically using a random forest algorithm. The value in building a predictive model of response time lies in identifying the most influential predictors of response times such as team utilization, case severity, COVID-19 patients, and roadway distance. Furthermore, the model allows for the identification of the variables that influence response time across different segments of the emergency response process: dispatch, wheeling, and roadway times. As a whole, this work supports EMS operations through the identification of managerial levers that have a direct influence on response time.