Emergency Medical Service Utilization and Response Following COVID-19 Emergency and Stay-at-Home Policies: An Interrupted Time-Series Analysis

Abstract

Objective Examine changing emergency medical services (EMS) utilization and response patterns associated with coronavirus disease 2019 (COVID-19) emergency declaration and stay-at-home orders during the first year of the COVID-19 pandemic. Methods We conducted an uncontrolled interrupted time series analysis of EMS calls (January 1, 2019 - March 1, 2021) in Marin County, California analyzing call volume (All calls, n=46,055); patient refusal of EMS care or transport and patient care resolved on scene (Calls with opportunity for transport; n=37,401); and call severity (Transported calls; n=27,887). Results Pre-COVID-19 (1/1/2019-3/2/2020), EMS transported patients were predominately female (50.6%), 80+ years old (31.6%), and Marin County residents (68.0%). During COVID-19 (3/3/2020-3/1/2021), EMS transported patients were predominately male (52.7%), 35-64 years old (29.8%), and Marin County residents (70.4%). After the first stay-at-home order on 3/17/2020, call volume immediately decreased by 48% (adjusted incidence rate ratio [aIRR]=0.52; 95% CI=0.35,0.79) for children (0-15 years) and 34% for adults 80+ years (aIRR=0.66;95% CI=0.46,0.95). The odds of a transported call being prioritized as severe doubled (adjusted odds ratio [aOR]=2.26; 95% CI=1.11,4.59). Though transport refusals increased by 69% for children after the first order (aOR, 1.69 [95% CI, 1.13-2.52]), immediately following the second order on 12/8/2020, transport refusals decreased by 30% for children but increased 38-40% for adults 35-79 years (aOR=1.40 [95% CI=1.04-1.89] for 35-64 years; 1.38 [95% CI=1.02-1.87] for 65-79 years). Calls resolved on scene by EMS increased after the first order among all ages and after the second order for adults 16-79 years. Conclusions Call volume reduced for children and older adults after the first COVID-19 stay-at-home order. Changes in call severity, patient care refusals, and on-scene care provided by EMS indicated a changing role for EMS during the outbreak.

Keywords: covid-19 outbreak; emergency medical services; emergency utilization; public health policy; stay-at-home orders.

Figure 1. Diagram of Emergency Medical Service (EMS) calls included in each analysis - Marin County, California, January 1, 2019 to March 1, 2021

Model 1 examined trends in the weekly call volume of all EMS calls. Model 2 assessed the likelihood of a call resulting in a transport refusal or being resolved on scene by EMS as compared to transport to a healthcare facility. Model 3 evaluated the likelihood of a call being severe versus non-severe among calls transported by EMS.

Figure 2. Weekly Emergency Medical Service (EMS) call volume and fitted model estimates stratified by age group – Marin County, California, January 1, 2019 to March 1, 2021

Model of fitted average call volume estimates and associated 95% confidence intervals plotted over crude point estimates of weekly EMS call volume by age group. Each fitted line represents the associated age group (see key) in years.

Figure 3. Weekly Emergency Medical Service (EMS) Call Volume adjusted incident rate ratios (aIRR) and 95% confidence intervals during COVID-19 outbreak periods – Marin County, California, January 1, 2019 to March 1, 2021

Incident rate ratios and 95% confidence interval highlighted in grey are statistically significant (p<0.05). ^aEach trend parameter is compared to the trends from the period directly preceding in the incident rate ratio (e.g., the trend from emergency declaration to stay-at-home order compared to the trend pre-emergency declaration). Immediate-level changes compare the week directly before the stay-at-home order went into effect to the week immediately following the order’s implementation.

Figure 4. Transport and severity adjusted odds ratios (aOR) and 95% confidence intervals for Emergency Medical Service (EMS) calls during COVID-19 outbreak periods - Marin County, California, January 1, 2019 to March 1, 2021

Odds ratios and 95% confidence interval highlighted in grey are statistically significant (p<0.05). ^aEach trend parameter is compared to the trends from the period directly preceding in the odds ratio (e.g., the trend from emergency declaration to stay-at-home order compared to the trend pre-emergency declaration). Immediate-level changes compare the day directly before the stay-at-home order went into effect to the day immediately following the order’s implementation. ^bStatistically significant interaction was not observed between age group and any COVID-19 event parameter (p>0.05).