Changes in Emergency Medical Services before and during COVID-19 in the United States, January 2018–December 2020

Abstract Background As a result of the continuing surge of COVID-19, many patients have delayed or missed routine screening and preventive services. Medical conditions, such as coronary heart disease, mental health issues, and substance use disorder, may be identified later, leading to increases in patient morbidity and mortality. Methods The National Emergency Medical Services Information System (NEMSIS) data were used to assess 911 Emergency Medical Services (EMS) activations during 2018–2020. For specific activation types, the percentage of total activations was calculated per week and joinpoint analysis was used to identify changes over time. Results Since March 2020, the number of 911 emergency medical services (EMS) activations has decreased, while the percentages of on-scene death, cardiac arrest, and opioid use/overdose EMS activations were higher than pre-pandemic levels. During the early pandemic period, percentages of total EMS activations increased for on-scene death (from 1.3% to 2.4% during weeks 11–15), cardiac arrest (from 1.3% to 2.2% during weeks 11–15), and opioid use/overdose (from 0.6% to 1.1% during weeks 8–18); the percentages then declined, but remained above pre-pandemic levels through calendar week 52. Conclusions The COVID-19 pandemic has indirect consequences, such as relative increases in EMS activations for cardiac events and opioid use/overdose, possibly linked to disruptions is healthcare access and health-seeking behaviors. Increasing telehealth visits or other opportunities for patient-provider touch points for chronic disease and substance use disorders that emphasize counseling, preventive care, and expanded access to medications can disrupt delayed care-seeking during the pandemic and potentially prevent premature death.

A c c e p t e d M a n u s c r i p t

Background:
The COVID-19 pandemic has had direct and indirect impacts on the healthcare system and on healthcare-seeking behavior. As a result of the continuing surge of COVID-19, many patients have delayed or missed routine screening and preventive services such as annual check-ups (1). The rates of emergency department visits and hospitals admissions for conditions not related to or affected by COVID-19 have also decreased during the start of the pandemic in 2020 compared to pre-pandemic times (2,3). Medical conditions, such as coronary heart disease, mental health issues, and substance use disorder may be identified later than usual given disruptions in availability of health services and changes in healthseeking behaviors, leading to increases in patient morbidity and mortality (4).
Data from 911 Emergency Medical Services (EMS) activations provides a novel source for understanding people's first touch points with the medical system. The information about 911 activations can add insights into resource capacity and limitations. Additionally, EMS response data can complement emergency department (ED) visit data, as it can provide additional information about on-scene deaths, which is not always reflected in ED data.
This study sought to assess changes in the percentages of 911 EMS activations before and during the COVID-19 pandemic for specific medical and behavioral reasons.

Methods:
The National Emergency Medical Services Information System (NEMSIS) is the U.S. national database used to standardize, aggregate, and store EMS point-of-care data from states and territories and represents a convenience sample of participating agencies within each state and territory. Reporting to NEMSIS is voluntary and the number of agencies reporting has A c c e p t e d M a n u s c r i p t increased over time from 9,599 in 2018, to 10,620 in 2019 and 11,257 in 2020. As such, the number of total EMS activations increased by 15% over the study period. To account for the increasing number of agencies and activations, this study examined the percentages of certain activation types (i.e., count of activation type divided by count of all activations) to describe trends over time.
The NEMSIS data was used to assess 911 EMS activations before (January 2018-February 2020) and during (March-December 2020) the COVID-19 pandemic in the United States.
Activations were included if the 911 call resulted in patient contact. Activations for encounters related to cardiac arrest, opioid use/overdose, injury, and mental/behavioral health and activations with an on-scene death disposition were examined. Cardiac arrest activations were identified in the NEMSIS national data if the eArrest.01 -Cardiac Arrest field was marked "yes", meaning there was and "indication of the presence of a cardiac arrest at any time during the EMS event"*. Opioid-related activations were identified if the International Classification of Disease 10 th edition (ICD-10-CM) codes F11 (opioid-related disorders) or T40.0-T40.4 and T40.6 (poisoning by and adverse effects of opioid related drugs) were found in the NEMSIS elements for primary symptom, or other associated symptoms, or provider's primary impression, or secondary impressions (eSituation.09-12).
Injury activations were identified if the NEMSIS field eSituation.02 -Possible Injury was marked "yes", indicating there was an injury related with the EMS event; injury type, location or cause was not indicated in this field (16). Mental/behavioral health activations were identified if eSituation.09-12 had any of the following ICD-10-CM codes: anxiety disorder F41.9, F41.1, R41.82; major depressive disorder F32.9; unspecified mental disorder F99; symptoms and signs involving emotional state R45.89, R45.7, R45.82; or symptoms and M a n u s c r i p t signs involving appearance and behavior R46.2, R46. Lastly, eDisposition.12 -Incident/Patient Disposition was examined to identify on-scene death activations. Any event with "the type of disposition/treatment and/or transport of the patient by the EMS unit" described as "patient dead at scene-no resuscitation attempted (with or without transport)" and "patient dead at scene-resuscitation attempted (with or without transport)" was identified as an on-scene death activation*.
The percentages of EMS activations related to cardiac arrest, opioid use/overdose, injury, and mental/behavioral health and activations with an on-scene death disposition were stable during weeks 1-10 before the pandemic ( Figure 2). In 2020, cardiac arrest and onscene death activations began to increase in week 11 (joinpoint p-values <0.001) ( Table 1).
During weeks 11-15, the percentage of cardiac arrest activations increased from 1.3% to 2.2% (n=7,260 to 9,411) and the percentage of on-scene death activations increased from 1.3% to 2.4% (n=6,782 to 10,213).
A c c e p t e d M a n u s c r i p t Percent activation stratified by calendar week in addition to sex, age group, and race/ethnicity was also examined (Supplement Tables). Opioid-related activations for the 22-60 year age group increased from week 7 and continued above baseline (Supplement Table A). Cardiac arrest, on-scene death disposition, and opioid-related activations increased a similar amount among persons of Black/African American race (cardiac=1.38% to 2.08%; on-scene death=1.29% to 2.24%; opioid=0.61% to 1.17%) and White race (cardiac=1.34% to 2.14%; on-scene deaths=1.26% to 2.31%; opioid=0.58% to 1.10%). Cardiac arrest (1.29% to 2.12%) and on-scene death (1.20% to 2.34%) activations also increased among persons of H/L ethnicity (Supplement Table B). There was little change for Asian and AI/AN populations. Percentages were higher for males than females for all activation types (Supplement Table C).

Discussion:
Findings from this study of EMS activations emphasize the indirect effects of the COVID-19 pandemic. The proportion of mental/behavioral health-related activations increased initially and then returned to baseline levels; while the proportion of injury-related activations decreased and remained below baseline levels. Most notably, our study results show that the proportion of activations with an on-scene death disposition, cardiac arrest, and opioidrelated activations remained above 2018-2019 baseline levels during all of 2020, supporting previously reported increase of on-scene death activations (5).
In 2020, EMS activations for cardiac arrests closely followed the trend for on-scene death.
Increases in cardiac events have been similarly reported for U.S. emergency room visits (2) A c c e p t e d M a n u s c r i p t Together, these data suggest that cardiac events are increasing while treatment and intervention visits are decreasing. Traditional ambulatory care-seeking behavior has been disrupted and many patients are delaying or deferring necessary care, including preventive care (1,6). However, people with chronic medical conditions like coronary heart disease are at higher risk for poor outcomes without active screening and monitoring by providers (7). to 2019 (8,9). In December 2020, CDC issued a Health Advisory alert describing substantial increases in drug overdose deaths across the U.S. before and during the COVID-19 pandemic (10) EMS activations for opioid use/overdose were elevated above baseline throughout most of 2020. Precautions taken to slow transmission of COVID-19, including physical distancing and stay-at-home orders, may lead to disruptions in access to medications for opioid use disorder (MOUD), naloxone (an opioid overdose reversal drug), recovery support services, and other forms of treatment for substance use disorder (11). In an effort to help mitigate these effects for individuals experiencing substance use disorder, the U.S.
Substance Abuse and Mental Health Service Administration, the U.S. Drug Enforcement A c c e p t e d M a n u s c r i p t Administration, and state regulators approved exemptions to expand take-home doses of methadone and buprenorphine, allowing medications to be initiated and maintained during telemedicine visits (12). Because study results support findings that EMS activations for opioid use/overdose events and mental health events have increased during the pandemic, a finding supported in emergency department visits as well, the suggested shift suggests a change in health-seeking behavior from primary care to emergency care. This change emphasizes the need to integrate care for mental health and substance use disorder screening and prevention services into regular touch points of care during the pandemic (13). Primary care providers are urged to continue to emphasize frequent visits and counseling and routinely discuss the availability of naloxone ¶ , which may improve adherence and outcome for patients at risk for substance use disorder or opioid overdose.
Mental/behavioral health EMS activations were higher than baseline levels in 2020. This finding is supported by results from a nationally-representative web-based survey that showed U.S. adults have experienced higher levels of adverse mental health conditions associated with COVID-19, and that racial/ethnic minority populations, essential workers, and unpaid adult caregivers report increased substance use and elevated suicidal ideation (14). Additionally, higher demand on EMS services on top of COVID-related demands, may also contribute to higher stress and other adverse conditions for EMS workers themselves.
Deferring preventative care has indirect cost implications to both the patient and supportive service providers.
There are limitations that should be considered when interpretating these study results.
First, although NEMSIS data accounts for 911 EMS activations from 11,257 agencies (53% total licensed agencies) in 2020 (15), it does not include information from agencies that do A c c e p t e d M a n u s c r i p t not send their data to NEMSIS. Therefore, results may not be generalizable. Second, the number of total EMS activations varied over time, increasing across the three-year study period as new agencies begin submitting data to NEMSIS and fluctuating throughout the pandemic in 2020, most notably during weeks 10-28 in 2020. The percentage of total activations was used to address the changing denominator and assess trends over time, however, the percentages of total activations during weeks 10-28 of 2020 should be   Medicines Regarding Naloxone. FDA News Release. Released July 23, 2020. Available at: https://www.fda.gov/news-events/press-announcements/fda-requiring-labelingchanges-opioid-pain-medicines-opioid-use-disorder-medicines-regarding . Accessed

2/24/21
A c c e p t e d M a n u s c r i p t M a n u s c r i p t  M a n u s c r i p t A c c e p t e d M a n u s c r i p t M a n u s c r i p t A c c e p t e d M a n u s c r i p t