The effect of regional distribution of isolation rooms in emergency departments on ambulance travel time during the COVID-19 pandemic

Abstract

Objective The number and distribution of isolation rooms in Korea differ by region. The distribution of isolation beds in emergency departments may have affected ambulance travel time and burden on emergency medical service (EMS) during the COVID-19 pandemic.Methods This retrospective observational study analyzed EMS records in four regions of the Gyeonggi Province, Korea, from January 01, 2019 to December 31, 2020. The main exposure was the number of emergency department isolation rooms in each region. The primary outcome was call-to-return time for the EMS. The interaction effect of the number of regional isolation rooms on the call-to-return time during the COVID-19 pandemic was analyzed using a generalized linear model (GLM) and logistic regression.Results A total of 781,246 cases was included in the analyses. During the COVID-19 pandemic, the call-to-scene time (before 8 minutes vs. after 9 minutes, P<0.05) and call-to-return time (before 46 minutes vs. after 52 minutes, P<0.05) for emergency patients increased significantly compared to before the pandemic. As the number of regional isolation rooms increased, the effect of COVID-19 on the call-to-return time decreased significantly in the multivariable GLM with an interaction term (with 10.14 isolation rooms per million population: adjusted exponential β coefficient [exp(β)], 1.33; with 12.24 isolation rooms per million population: adjusted exp(β), 1.18). As the number of regional isolation rooms increased, the effect of COVID-19 on the call-to-scene time decreased significantly in the multivariable GLM with an interaction term (with 10.14 isolation rooms per million population: adjusted exp(β), 1.20; with 12.24 isolation rooms per million population: adjusted exp(β), 1.09).Conclusion During the pandemic, the increases in call-to-return time and call-to-scene time were smaller in regions with more isolation rooms per population.