Abstract
Background:
This study evaluates trends in the utilization of emergency medical services (EMS) in New York City, the “epicenter” of the first “wave” of the coronavirus pandemic. We hypothesize that EMS call volumes decreased overall in New York City during the first year of the pandemic, specifically with respect to trauma/injury calls. Contrarily, we posit that calls for “sick” events increased given pervasive fear of virus transmission.
Materials and Methods:
Retrospective New York City EMS calls data (January 1, 2019 to December 31, 2020) were obtained from the NYC Open Data/EMS Incident Dispatch database. Total EMS calls, trauma/injury calls, and “sick” event calls were collected for New York City and for all five boroughs. Census data for each borough were used to weigh daily EMS calls per 100,000 individuals. Mann-Whitney U tests were used to compare pre-pandemic (2019 to March 2020) versus pandemic (April 2020 to December 2020) EMS call volumes, p = 0.05.
Results:
Median daily EMS calls per 100,000 individuals decreased 21.6% at the start of the pandemic across New York City (pre-pandemic, 3,262 calls; pandemic, 2,556 calls; p < 0.001) and similarly decreased when stratified by borough (all, p < 0.001). Median daily trauma/injury and sick event calls per 100,000 also decreased in New York City and the five boroughs from pre-pandemic to pandemic time periods (all, p < 0.001).
Discussion and Conclusions:
These data reflect an unprecedented window into EMS utilization during an infectious disease pandemic. As decreased EMS utilization for multiple conditions likely reflects delayed or impeded access to care, utilization data have important implications for provision of acute care services during possible future disruptions related to the pandemic.
The emergence of coronavirus disease 2019 (COVID-19) has had unprecedented consequences in the United States as well as globally. These consequences were the result not only of the virus itself, but also of the extraordinary measures taken to curb the spread of COVID-19. To reduce disease transmission, social distancing measures were put into place that paralyzed commerce, disrupted domestic travel and prevented international travel, 1 and exacerbated socioeconomic and healthcare disparities among already vulnerable populations. 2 To date, the complete impact of the pandemic is unknown, and likely exceeds the nearly seven million lives lost due to COVID-19. 3 It has been speculated that the current understanding of the “tragedy” of the pandemic fails to capture the true number of lives lost, because many patients who did not die from the virus succumbed indirectly to an overburdened healthcare system. 4
It was well documented at the start of the pandemic that fear surrounding COVID-19 infection led to disruption of both routine and semi-urgent medical care.5,6 In the setting of the exponential increase of COVID-19 cases nationwide, attempts were made in the United States to conserve hospital beds and other scarce resources as well as to minimize further infection risk to patients. 5 As a result, rates of emergency ambulance service calls, 7 emergency department visits,8–11 and utilization of routine screening and preventive services 8 and ambulatory procedures 12 decreased substantially. Patients with chronic conditions including cancer, 13 congenital heart disease 14 and other cardiac diseases, 15 endocrine disorders such as diabetes mellitus, 16 and more had their care interrupted during the early phases of the pandemic, resulting in delayed evaluation and treatment. 13 Dental, 17 orthopedic,18–20 and ophthalmic care 1 were also disrupted.
Unfortunately, preliminary data have shown that because of this decreased healthcare utilization, there have been increases in emergency department mortality, 11 worse outcomes for serious cardiovascular conditions such as myocardial infarction and heart failure, 15 and increased cancer-related deaths, 21 all attributed to delayed presentation or intervention. Impaired access has been addressed by increased utilization of telemedicine,22–24 but obviously not for emergency situations.
To evaluate quantitatively the effects of the COVID-19 pandemic on emergency medical services in the United States, this study examined patterns of 9-1-1 emergency medical services (EMS) call volumes in New York City during 2019 to 2020. New York City was selected because its EMS system is the largest in the United States, servicing upwards of eight million individuals and fielding approximately 4,000 calls per day.25,26 Additionally, by the end of March 2020, New York City was considered by the media and healthcare agencies to be the “epicenter” of the early pandemic, 27 at one point accounting for approximately 5% of the world's confirmed COVID-19 cases. 27 Given this immense COVID-19 burden, it was hypothesized that EMS call volumes would decrease overall in New York City during the first year of the pandemic. Regarding specific call types, it was posited that whereas trauma/injury calls would decrease because of fears surrounding virus transmission, calls for “sick” events would paradoxically increase during this time.
Materials and Methods
This retrospective database study was deemed exempt by the Institutional Review Board of UC Irvine Health; thus, no informed consent was obtained. New York City EMS calls data for January 1, 2019 through December 31, 2020 were obtained from NYC OpenData's EMS Incident Dispatch database. 28 This is an open-access database that is generated by the EMS Computer Aided Dispatch System and is maintained by the Fire Department of New York City. This database collects information from the time the incident is initiated in the system to the time it is closed. As such, there is initial information and corrected final information regarding call type and call severity, as well as dispatch and response times. 28
Total EMS calls, trauma/injury calls, and sick event calls were collected for all of New York City and for all five boroughs (Bronx, Brooklyn, New York [Manhattan], Queens, and Richmond [Staten Island]). The “total” EMS calls category included calls for a variety of conditions, such as acute onset or worsening of cardiopulmonary, gastrointestinal, gynecologic, or neurologic diseases; trauma/injury calls; drug and alcohol abuse or overdose calls; assault calls; motor vehicle collision calls; sick event calls; and firearm-related calls. Within the trauma/injury call category were calls for extremity amputation, human or animal bites, bleeding or hemorrhage, burns, eye injuries, fractures, unspecified non-critical injuries, and “minor” or “major” injuries. Sick event calls included calls for fever, cough, or rash. The age of the patient was not specified in the database and thus patients of all ages were included in the analysis.
After data collection, U.S. Census data for New York City in its entirety and for each borough were used to weight daily EMS calls per 100,000 individuals. 29 Descriptive statistics were compared. The non-parametric Mann-Whitney U test was used to compare pre-pandemic (2019 to March 2020) versus pandemic (April 2020 to December 2020) EMS call volumes. The end of March was used as the pandemic onset given that this was around the time (within 10 days) when New York City hospitals began to be overwhelmed and New York City came to be considered the COVID-19 epicenter of the first wave. 27 Statistics were performed using IBM SPSS Statistics, version 26 (IBM Corp, Armonk, NY). Statistical significance was set as p = 0.05.
Results
Median (interquartile range, IQR) daily EMS calls for New York City in its entirety decreased at the start of the pandemic (pre-pandemic, 3,262 [IQR, 3,133–3,375] calls; pandemic, 2,556 [IQR, 2,393–2,690] calls; p < 0.001) compared with the pre-pandemic time period. Similarly, New York City median total daily trauma/injury calls (pre-pandemic, 695 [IQR, 656–736] calls; pandemic, 580 [IQR, 492–651] calls; p < 0.001) as well as sick event calls (pre-pandemic, 724 [IQR, 681–774] calls; pandemic, 447 [IQR, 418–479] calls; p < 0.001) decreased during the pandemic (Table 1).
New York City Pre-Pandemic Versus Pandemic Emergency Medical Services Calls Per Day, Median (Minimum, Maximum)
With regard to median daily total EMS calls weighted per 100,000 individuals, New York City (pre-pandemic, 39.1 [IQR, 38–40] calls; pandemic, 31.8 [IQR, 29–34] calls; p < 0.001), Bronx (pre-pandemic, 55.4 [IQR, 52–58] calls; pandemic, 46.3 [IQR, 43–49] calls; p < 0.001), Brooklyn (pre-pandemic, 35.3 [IQR, 34–37] calls; pandemic, 28.4 [IQR, 26–31] calls; p < 0.001), New York (pre-pandemic, 49.0 [IQR, 46–52] calls; pandemic, 37.0 [IQR, 33–40] calls; p < 0.001), Queens (pre-pandemic, 27.9 [IQR, 27–29] calls; pandemic, 23.6 [IQR, 21–26] calls; p < 0.001), and Richmond (pre-pandemic, 28.8 [IQR, 27–31] calls; pandemic, 23.5 [IQR, 21–25] calls; p < 0.001) call rates all decreased from pre-pandemic to the pandemic time period.
Trauma/injury EMS calls also decreased during the pandemic in New York City (pre-pandemic, 8.3 [IQR, 8–9] calls; pandemic, 7.0 [IQR, 6-8] calls; p < 0.001) as well as in each of the five boroughs (all p < 0.001). Finally, sick event calls also decreased across New York City (pre-pandemic, 8.7 [IQR, 8–9] calls; pandemic, 5.4 [IQR, 5–6] calls; p < 0.001) in the initial wave of the COVID-19 pandemic. Sick even calls also decreased in the Bronx (pre-pandemic, 13.3 [IQR, 12-15] calls; pandemic, 8.8 [IQR, 8–10] calls; p < 0.001), Brooklyn (pre-pandemic, 8.2 [IQR, 8–9] calls; pandemic, 4.6 [IQR, 4-5] calls; p < 0.001), New York (pre-pandemic, 10.1 [IQR, 46–52] calls; pandemic, 6.8 [6–8] calls; p < 0.001), Queens (pre-pandemic, 5.8 [5–6] calls; pandemic, 3.6 [IQR, 3–4] calls; p < 0.001), and Staten Island (pre-pandemic, 6.1 [IQR, 5–7] calls; pandemic, 3.8 [IQR, 3–5] calls; p < 0.001) during this time (Table 2).
Daily Pre-Pandemic Versus Pandemic Emergency Medical Services Calls by New York City Borough per 100,000 Individuals, Median (Minimum, Maximum)
Discussion
The true extent of the disruption to healthcare delivery brought on by the COVID-19 pandemic is still being studied, including trends in healthcare utilization and the impact on EMS and hospital systems. Preliminary observations suggest that, although many patient deaths were not directly the result of the virus itself, the overburdened healthcare system and the unwillingness of patients to pursue care because of concerns over virus transmission may have been major confounders. 8 This reluctance to pursue care is further evidenced by the changes in EMS call volumes in New York City and each of its five boroughs during the first year of the pandemic. Not only did total EMS call volume decrease substantially from the pre-pandemic to the pandemic time period, so too did trauma/injury calls and sick event calls.
Across the span of a year, the median daily calls placed to 9-1-1 in New York City decreased by approximately 22%. Re-stated, nearly one-quarter of individuals (or bystanders) who suffered from acute-onset or worsening of a systemic disease or those who sustained traumatic injury (e.g., assault, motor vehicle collision, gunshot wound) or those who experienced a drug or alcohol overdose decided not to call 9-1-1, possibly out of fear of contracting infection with severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the etiologic agent of COVID-19. This is concerning for the public health overall, because other studies have shown that certain patient presentations, such as opioid overdose, 30 mental or psychiatric disturbances, 31 and out-of-hospital cardiac arrest 32 increased for individuals and cities in similar circumstances. Thus, for some such diagnoses, the observed reductions of 9-1-1 call volumes may represent an even larger underestimate of the lives lost during the COVID-19 pandemic. As such, it is likely that in New York City during the first year of the pandemic, the care-seeking behavior of patients was altered dramatically, and likely led to delays in diagnosis that have impacted patients' physical and mental well-being permanently. 33
Four comparable reports have been published to date.7,34–36 Lerner et al. 34 surveyed the National Emergency Medical Services Information System (NEMSIS) database, 37 reporting a 26.1% decrease in nationwide EMS call volumes beginning in the week of March 2–8, 2020 that reached its nadir six weeks later. By contrast to our findings, EMS calls for “possible injury” decreased by only 17% (18.4%–15.3%.) Andrew et al. 7 reviewed the demand on “emergency ambulances” from the COVID-19 pandemic in Victoria, Australia (population, ∼6,000,000) from January 2018 to February 2021, including seven months of lockdown from October 19, 2020, to February 28, 2021. The review included 2,356,326 “0-0-0” (the Australian analogue of the familiar 9-1-1 system) calls. By time-series analysis, COVID-19 lockdown was associated with a 12.6% reduction in weekly call volume (incident rate ratio, IRR) 0.874, 95% confidence interval [CI] 0.811–0.941. However, the reduction was not associated with a change in long-term trend (IRR, 1.000; 95% CI, 0.996–1.003). With respect to injury, the only data reported separately by Andrew et al. 7 regarded “traffic accidents,” which were decreased by 28.9% during the lockdown (IRR, 0.711; 95% CI, 0.558–0.907.)
A report from metropolitan Frankfurt, Germany (population ∼750,000) reported a 23.0% decrease of emergency calls compared with pre-COVID-19 during the comparable time period, but trauma calls were negligible (fewer than three per 100,000 population at all time points surveyed. 35 By contrast, the Lombardy region of Italy (Milan and environs, population ∼10,000,000) reported a 51.2% increase in emergency calls during March to April 2020, but the increase was due overwhelmingly to calls classified as “breathing” or “infective.” 36 Calls for injury were not described specifically.
That trauma/injury calls in New York City declined during the pandemic compared with the pre-pandemic era was unsurprising and consistent with both the study authors' hypothesis and the existing scientific literature. Published articles have discussed the observed decrease in trauma activations in the wake of COVID-19.38,39 It has been theorized that COVID-19–related social distancing and stay-at-home and shelter-in-place orders put in place by governmental fiat altered human social interactions (e.g., restrictions on public and private social gatherings, closure of establishments that serve alcohol), through limiting outdoor activities, decreased mobility (e.g., less vehicular traffic, abandonment of public transportation), and dramatic reductions of overall economic activity, and, in a way that inherently decreased risk for trauma/injury. 28 For instance, motor vehicle collisions were among the many causes of injury to have decreased substantially because of the physical distancing measures. 40 Of note, by more recent comparison there were more than 8,000 motor vehicle collisions in New York City in September 2022 alone, with an unfortunate 3,347 deaths. 41 Thus, decreased numbers of motor vehicle collisions had a substantial impact on EMS call volumes in New York City in the first year (2021) of the pandemic.
There was a significant decrease in sick event calls, including calls for fever, cough, or rash, during the first year of the pandemic in New York City This seems to contradict the increasing incidence of COVID-19 cases in New York City at that time, with a noted peak of 5,402 total confirmed cases per day around early April 2020. 42 Although some people infected with SARS-CoV-2 experienced no symptoms, many others experienced common symptoms including fever, cough, muscle aches, fatigue, sore throat, and more. In severe cases, individuals could be expected to have a high fever, severe cough, and dyspnea. 43 As such, it is interesting that calls for these symptoms decreased in New York City while fears surrounding COVID-19 were heightened in this epicenter of disease. Also interesting is the fact that the decline in EMS utilization for sick event calls is not universal, because another study showed calls for fever increased 14% and cough increased 956% in other countries. 44 These inconsistencies warrant further investigation into New York City's handling of pandemic-related restrictions and distribution of healthcare resources for “sick” individuals as it combated the rapid transmission of the virus.
Because this study utilized retrospective databases and U.S. Census data, it is subject to several limitations, such as coding errors, missing data, and reporting bias. However, with regard to the EMS calls data, NYC OpenData's EMS Incident Dispatch database is automated to collect all information for incoming 9-1-1 calls. There is also an internal audit so that the initial information that is collected is corrected for call type and severity once EMS has arrived on the scene. 28 The final, corrected data were utilized in this study to avoid errors from miscommunication between patient and EMS dispatcher, as well as to avoid errors from lack of patient healthcare literacy, particularly that which is limited from being otherwise occupied by an emergency healthcare crisis. Furthermore, U.S. Census data were used to weight the New York City population for the years 2019 and 2020; however, this national data is collected infrequently. As such, we utilized 2020 Census data for both study years, and thus the EMS call rates for 2019 may be underestimates given the population growth.
Fortunately, however, new data has shown that U.S. population growth has nearly flatlined, and therefore the change from 2019 to 2020 may not be great enough to change our results substantially.
45
Additionally, this population data does not account for the large commuter population in-and-out of New York City. However, we believe that commuter populations are generally healthier and thus less likely to require EMS services. Moreover, the majority (approximately 60%) of EMS calls during this study period were made during non-work hours versus work hours (9:00
Conclusions
This study provides a nearly unprecedented window into EMS utilization during the first year of an infectious disease pandemic in New York City As total EMS utilization, trauma/injury calls, and sick event calls decreased uniformly across New York City as well as within each of the five boroughs, it is likely that many patients experienced delayed or impeded access to care. These EMS utilization data have important implications for provision of acute care services during possible future disruptions related to this or another pandemic or other public health emergency of substantial magnitude.
Footnotes
Authors' Contributions
Conceptualization: Donnelly, Barie, Schubl. Data curation: Donnelly. Methodology: Donnelly, Barie, Schubl. Writing: Donnelly, Barie, Schubl. Formal analysis: Kirby, Chau. Supervision: Barie, Schubl.
Funding Information
No funding was provided for this study.
Author Disclosure Statement
The authors have no conflicts of interest.
