Prehospital COVID-19-Related Encounters Predict Future Hospital Utilization

Objective Identify if prehospital patient encounters can predict SARS-CoV-2 (COVID-19) related hospital utilization. Methods EMS data from COVID-19-related prehospital encounters was pulled from NEMSIS systems in Minnesota. This data was plotted against hospital general medical-surgical bed and ICU bed usage during the initial COVID-19 surge and again during a second surge. A validation dataset from 2019 was also utilized. Results There was a total of 6,460 influenza-like-illness calls, and 2,161 COVID-19-specific calls during the studied timeframe. A total of 24,806 medical-surgical bed-days and 20,208 ICU bed-days were analyzed. During initial COVID surge (April-July 2020), EMS encounters best correlated with medical-surgical bed utilization 10 days in the future (r(2) = 0.85, N = 113, p = <0.001), with each encounter correlating with a utilization of 7.1 beds. ICU bed utilization was best predicted 16 days in the future (r(2) = 0.86, N = 107, p = <0.001) with each encounter correlating with the use of 4.5 ICU beds. Similarly strong and clinically significant correlations were found during the second surged during July and August. There was no significant correlation when comparing to a similar dataset using 2019 ILI calls. Conclusion Minnesota prehospital COVID-19-related prehospital encounters are shown to accurately predict hospital bed utilization 1-2 weeks in advance. This was reproducible across two COVID-19 surges. Trends in EMS patient encounters could serve as a valuable data point in predicting COVID-19 surges and their effects on hospital utilization.