Resuscitation Practices at Emergency Medical Service Agencies Working in Black and Hispanic Versus White Catchment Areas in the United States.
May 30, 2025DOI: 10.1161/circoutcomes.124.011799Journal ArticleAHA
Written by: Paul S ChanSaket GirotraKhadijah BreathettKimberly C DukesJessica SperlingChristina M PachecoKevin F KennedyComilla SassonBryan McNallyHeather Schacht ReisingerMarina Del Rios
General Emergency Medicine · CU Anschutz Medical Campus
INTRODUCTION
Emergency medical service (EMS) agencies are the foundation for prehospital response to out-of-hospital cardiac arrest (OHCA). EMS personnel are the first to arrive on the scene in more than half of all OHCAs;1 initiate advanced life support including defibrillation, airway management, and administration of medications; and transport OHCA patients, if needed. Although EMS agencies typically use standardized protocols to respond to OHCA, survival rates are lower for agencies that work in Black and Hispanic communities, even after adjusting for their lower rates of bystander cardiopulmonary resuscitation (CPR).2Given lower OHCA survival rates at EMS agencies working in Black/Hispanic communities, determining whether dispatcher, first responder, and EMS practices differ when compared with agencies working in White communities may yield insights for interventions to reduce racial disparities in OHCA survival. Yet, EMS agency-level data regarding resuscitation training, practices, and policies for treating OHCA are not routinely collected in national registries for OHCA. We recently conducted a detailed survey to assess how dispatch, first responder actions, and EMS resuscitation practices differ among agencies participating in the Cardiac Arrest Registry to Enhance Survival (CARES), a national OHCA registry in the U.S., and linked the survey data with registry data on OHCA survival. These efforts represent the first large-scale assessment of EMS practices linked to agency-level outcomes for OHCA.Accordingly, we leveraged this recent survey and compared dispatch, first responder, and resuscitation practices between EMS agencies with majority Black/Hispanic and majority White catchment areas. We evaluated whether practice differences between these two EMS agency groups explained agency-level differences in OHCA survival.
METHODS
The study adhered to the STROBE reporting guidelines for observational studies. It was approved by Saint Luke’s Hospital’s institutional review board, which waived the requirement for informed consent as the study involved de-identified data. The data that support the findings of this study are available upon request and approval by the CARES registry.Established by the Centers for Disease Control and Prevention and Emory University, CARES is a prospective, multicenter registry of OHCA patients in the U.S. Currently, CARES has a catchment area of over 173 million residents, representing 53% of the U.S. population. The design of the registry has been previously described.3, 4 Briefly, all patients with a confirmed OHCA and for whom resuscitation is attempted (CPR by EMS or defibrillation at any time) are identified and enrolled by EMS agencies. The registry includes variables collected from 9-1-1 dispatch centers, EMS agencies (prehospital variables), and receiving hospitals (hospital outcomes). To ensure standardization of clinical variables and clinical outcomes, data collection is based on Utstein definitions to ensure uniform reporting across sites.5
Based on our prior work,6 our initial cohort for the EMS survey included 764 agencies participating in CARES during 2015–2019 with a mean annual volume of at least 10 OHCA cases. From this sample, we excluded 145 agencies for whom site-specific contact information was unavailable, or another EMS agency had acquired the agency. We also excluded 42 agencies that did not participate continuously in CARES throughout the study period. Therefore, a total of 577 EMS agencies were eligible for the survey in this study.
Details regarding the survey design, implementation, and response rate have been previously described.7 Briefly, the online survey was developed by a multi-disciplinary team of emergency medicine, internal medicine, and cardiology physicians, paramedics, and qualitative researchers. The survey was administered between September 28, 2022, and February 28, 2023, and included 67 items across four domains - agency structure; characteristics of dispatch and first responders; EMS treatment of OHCA, data review and quality improvement activities; and community factors affecting bystander response to OHCA. Each EMS agency identified the person most qualified to answer questions related to resuscitation training, protocols and policies, and first responder and dispatcher response; this individual was most often the EMS agency director. To achieve a high response rate, we sent weekly email reminders, called EMS agency contacts on two occasions, conducted drawings of free automated external defibrillators to generate excitement about survey completion, and leveraged state CARES coordinators to also contact agencies.
Our goal was to compare EMS agencies by the racial-ethnic makeup of their catchment area for emergency response. EMS agencies serving majority Black/Hispanic catchment areas were defined as those in which the majority (>50%) of residents in their coverage area were of Black race or Hispanic ethnicity. In contrast, EMS agencies serving majority White catchment areas were those in which the majority of residents were either of White or Asian race. We determined a priori to group Black and Hispanic residents into a combined catchment area group based on prior work showing lower rates of OHCA survival for both patient populations8–10 and to group White and Asian patients into a combined catchment area based on prior work2 and as White and Asian patients with OHCA have similar rates of bystander response in the U.S.11
The racial-ethnic composition of the catchment area for each EMS agency was determined as follows. First, each OHCA in CARES was geocoded to a U.S. Census tract using the address where the arrest occurred. Next, neighborhood-level information on the racial/ethnic composition of each Census tract was obtained from the 2019 American Community Survey.12 Since the catchment area of an EMS agency often involves more than one Census tract, we calculated the racial-ethnic composition of the catchment area for each EMS agency by weighting the racial-ethnic composition of each Census tract served by that agency with the proportion of the total OHCA case volume of the EMS agency from that Census tract. For example, consider an agency with 100 OHCAs during the study period: 50 occurred in a Census tract that was 40% Black/Hispanic, and 50 in a Census tract that was 80% Black/Hispanic. Based on these data, the agency would be classified as serving a Black/Hispanic catchment area, with a weighted racial-ethnic composition of 60% Black/Hispanic (see additional example in Supplementary Appendix Table S1). Census tract-level race and ethnicity information was unavailable for 30 (5.2%) of the 577 agencies. For these agencies, we imputed racial-ethnic composition using multiple imputation based on available data, including OHCA case characteristics, racial/ethnic composition of OHCA cases at the agency, and other Census tract-level information. We generated 96 multiply imputed data sets and used the average of the 96 multiply imputed values for each of the 30 agencies. With this imputation approach, we surmised that the unaccounted missing data uncertainty would be minimal given the small percentage of missing data and the high predictivity of the imputation model.
The primary objective of this analysis was to compare various practices between EMS agencies serving majority Black/Hispanic catchment areas and those serving majority White catchment areas, and whether practices differences attenuated differences in OHCA survival between the two EMS agency groups. Resuscitation practices by dispatchers, first responders (police and non-transporting fire units), and EMS were compared between agencies with a majority Black/Hispanic and White catchment area using chi-square tests for categorical variables and t-tests for continuous variables.
We then estimated risk-standardized rates of survival to hospital admission and favorable neurological survival (i.e., survival to discharge without severe neurological disability) between EMS agencies working in majority Black/Hispanic vs. White catchment areas using multivariable hierarchical logistic regression models that adjusted for differences in case-mix across EMS agencies. In this model, EMS agency was included as a random effect, and the following variables were included as fixed effects: age, sex, location of cardiac arrest (home, workplace, street/highway, industrial building, recreational facility, or other), whether the arrest was witnessed, initial cardiac arrest rhythm (shockable vs. non-shockable), etiology of cardiac arrest (presumed cardiac, respiratory, drug overdose, and other), whether bystander CPR was initiated, and urbanicity of the OHCA location (U.S. Census tract classification: urbanized [≥50,000 residents], urban cluster [non-urbanized areas with ≥2500 residents], or rural [<2500 residents]).13 Using the agency-specific estimates (i.e., random intercepts) from the hierarchical model, we then calculated risk-standardized rates of each survival outcome for an EMS agency by multiplying the ratio of each agency’s predicted to expected number of survivors with the overall unadjusted rate.14, 15 The expected number of OHCA survivors at an agency was determined using the average intercept from all sites whereas the predicted number of OHCA survivors was determined using the agency’s individual random effect intercept. The c-statistic of both models were both 0.87, suggesting excellent discrimination.
Next, risk-standardized rates of survival to hospital admission and favorable neurological survival were compared between EMS agencies with a majority Black/Hispanic and White catchment area using multivariable linear regression. We then assessed whether survival differences between these two EMS agency groups were attenuated after adjusting for differences in resuscitation practices. To avoid model overfitting, only resuscitation practices with an unadjusted p-value of <0.20 in the comparison between EMS agencies with a majority Black/Hispanic and White catchment area were included in the models. Finally, we repeated these multivariable linear regression analyses using CARES data between July 1, 2020, and December 31, 2022, after the emergence of the COVID-19 pandemic.
All analyses were evaluated at a 2-sided significance level of 0.05 and performed with SAS 9.4 (SAS Institute, Cary, NC) and R version 3.6.3.
RESULTS
Of 577 eligible agencies, 470 (81.5%) EMS agencies completed the survey, and these agencies comprised the study cohort. During the study period of 2015–2019, these agencies treated 181,707 OHCAs. Mean age was 62.0 ± 17.1 years, 36.2% were female, 28.8% were non-Hispanic Black and 7.3% were Hispanic (Supplementary Appendix Table S2). Overall, 82.1% of OHCAs occurred at home, 56.2% were bystander witnessed, and 20.6% were due to an initial shockable rhythm.
A total of 47 (10.0%) EMS agencies served a majority Black/Hispanic catchment area, whereas 423 (90.0%) served a majority White catchment area. Agencies with a majority Black/Hispanic catchment area were more likely to be in a mid-sized or large city, but there was no difference in the EMS staffing model (paid, volunteer, or hybrid) or EMS type (e.g., fire-based, hospital-based) (Supplementary Appendix Table S3). When comparing EMS agencies working in majority Black/Hispanic vs. White catchment areas, the vast majority of dispatchers in each group used a standardized protocol for pre-arrival instructions and there were no differences in the provision of dispatcher-assisted CPR instructions (Table 1). Dispatchers were able to direct a caller to the closest automated external defibrillator (AED) in 33.0% of agencies and to direct laypersons to the scene to assist with bystander CPR in 28.5% of agencies, but these rates did not differ between agencies working in majority Black/Hispanic and White catchment areas. However, dispatchers at agencies with majority Black/Hispanic catchment areas were reported to be less likely to recognize a cardiac arrest event during a 9-1-1 call.
Table 2 compares first responder actions between the EMS agency groups. Non-transporting fire units were more likely to respond at agencies with a majority Black/Hispanic vs. White catchment area (89.4% vs. 80.9% responded to >50% of OHCAs, p=0.001). In contrast, police first responders were much less likely to attend to OHCA calls at agencies with a majority Black/Hispanic vs. White catchment area (46.8% vs. 68.9% responded to >50% of OHCAs, p=0.005). Moreover, police first responders were less likely to initiate CPR or apply an AED at agencies with a majority Black/Hispanic catchment area as compared to an agency with a majority White catchment area. Finally, police were much less likely to recognize an OHCA at agencies with a majority Black/Hispanic vs. White catchment area.
Table 3 compares initial and ongoing training in CPR and resuscitation between EMS agencies working in majority Black/Hispanic and White catchment areas. There were no differences between the EMS groups in the number of hours devoted to cardiac arrest training during new staff orientation. However, EMS agencies with majority Black/Hispanic catchment areas were less likely to assess maintenance of CPR competency in their personnel at least once annually (46.5% vs. 65.0%; P=0.02) and there was a trend for lower use of simulation for initial cardiac arrest training at EMS agencies with majority Black/Hispanic catchment areas (78.7% vs. 87.5%, P=0.09), although use of full-scenario simulation did not differ between the two EMS groups. There were no differences between EMS groups regarding ongoing review of CPR quality.
Table 4 compares resuscitation practices in responding to OHCA between the EMS agency groups. Compared with agencies with a majority White catchment area, EMS agencies with a majority Black/Hispanic catchment area were less likely to use waveform capnography (91.5% vs. 99.5%; P<0.001). There were no significant differences between EMS agencies with majority Black/Hispanic and White catchment areas in the use of automated CPR feedback devices, mechanical CPR devices, or prehospital electrocardiograms for suspected ST-segment myocardial infarction. Similarly, EMS response and transport times, as well as the type of hospital to which OHCA patients were transported, did not differ significantly between the two groups. Table 5 compares data review and quality monitoring practices. EMS agencies with a majority Black/Hispanic catchment area were more likely to employ staff for quality improvement (95.7% vs. 75.4%; P<0.001). Otherwise, monitoring of time metrics (time from 9-1-1 call to time of OHCA dispatch and time from dispatch to EMS arrival) were similar between the two EMS agency groups.
Agency-level rates of survival to hospital admission were lower at EMS agencies with a majority Black/Hispanic catchment area as compared to those with a majority White catchment area (26.0% ± 4.0% vs. 28.0% ± 3.5%; P<0.001). Survival differences were attenuated by approximately 30% after adjusting for differences in resuscitation practices, dispatch characteristics and first responder actions (26.6% ± 3.5% vs. 28.0% ± 3.1%, P=0.004; Table 6). Similarly, agency-level rates of favorable neurological survival were lower at EMS agencies with a majority Black/Hispanic vs. White catchment area (6.3% ± 1.9% vs. 8.3% ± 1.7%; P<0.001), and these differences were attenuated after adjustment (6.9% ± 1.5% vs. 8.2% ± 1.5%, P<0.001). When these analyses were repeated using CARES data between mid-2020 and 2022 during the COVID-19 pandemic, there were no longer differences in rates of survival to hospital admission between the two EMS agency groups. However, there were differences in rates of favorable neurological survival between the EMS agency groups (an outcome of greater importance to patients and their families), with a similar pattern of attenuation after adjustment for EMS factors (Supplementary Table S4).
DISCUSSION
Given large disparities in OHCA survival by race and ethnicity in the U.S., we surveyed EMS agencies within the national CARES registry to determine if dispatch, first responder and EMS practices differed between agencies with a majority Black/Hispanic and White catchment area. We found that dispatchers at agencies with a majority Black/Hispanic catchment area were less likely to recognize an OHCA event, whereas police at these agencies were less likely to recognize and respond to an OHCA, start CPR, or deploy an AED. Moreover, EMS agencies with a majority Black/Hispanic catchment area were less likely to assess CPR competency in their personnel at least once annually and use waveform capnography. Notably, adjustment for differences in dispatch, first responder and EMS practices attenuated survival differences between agencies with majority Black/Hispanic and White catchment areas. Collectively, our study provides important insights regarding potential strategies for reducing racial-ethnic disparities in OHCA survival.
Despite the critical role of EMS agencies in the prehospital management of OHCA, little is known about whether dispatch, first responder, and EMS practices differ at agencies working in majority Black/Hispanic vs. White catchment areas. In part, this is because national surveys of EMS agencies have had low response rates16–18 and typically are not linked to survival outcomes. We were able to achieve a high survey response rate and had 470 agencies in the study cohort. Our study therefore extends the EMS literature in two important ways. First, we compared resuscitation practices between EMS agencies with a majority Black/Hispanic and White catchment area to better understand how these factors may vary between the two EMS agency groups. We then evaluated whether measured differences in these factors may, in part, explain disparities in OHCA survival in the U.S. To our knowledge, this comparison has not been performed before since previous EMS surveys have not had a sufficient sample size of agencies in different catchment areas or were not linked to OHCA survival data.
At agencies with a majority Black/Hispanic catchment area, we found that dispatchers less frequently recognized a cardiac arrest event was occurring. Any delays in dispatcher recognition of a cardiac arrest would delay initiation of CPR instructions to bystanders, possibly decreasing survival probability. In Black and Hispanic communities, we also found differences in first responder actions for an OHCA. Specifically, police were much less likely to recognize and respond to OHCA, initiate CPR, and deploy an AED. When police are part of the emergency response to an OHCA, CPR may be initiated sooner and there would likely be a higher number of available personnel to more effectively deliver continuous and effective CPR. Additionally, the arrival of police first responders with vehicles equipped with an AED could decrease defibrillation time for shockable OHCAs. Our findings underscore the recommendations from the Resuscitation Academy, which emphasize training of police first responders to respond to OHCA and initiate CPR, and equipping patrol cars with AEDs, as key strategies to improve OHCA survival.19
We also found differences in EMS training and practices between the two agency groups. CPR competency assessment of EMS personnel on an annual basis was performed less frequently in EMS agencies with majority Black/Hispanic catchment areas, as compared to those with majority White catchment areas. Moreover, waveform capnography, which the International Liaison Committee on Resuscitation recommends using to monitor end-tidal CO2 as a measure of CPR quality,20 was less frequently employed at EMS agencies with majority Black/Hispanic catchment areas. There was also a trend for less use of simulation for initial cardiac arrest training at agencies working in majority Black/Hispanic catchment areas. These practice differences may lead to differences in resuscitation competency and delivery of high-quality CPR.
Notably, adjustment for these factors attenuated differences in rates of favorable neurological survival between agencies with majority Black/Hispanic and White catchment areas, suggesting that these practice differences may account, in part, for survival disparities. However, because adjustment for these resuscitation practices accounted for only some of the survival differences, our findings suggest that granular information on other factors are needed to provide additional insights into agency-level differences in OHCA survival. These include potential structural inequities between the EMS agency groups such as funding and resources for training, personnel, and equipment that may affect resuscitation effectiveness, as well as community CPR training that may affect the speed with which bystanders recognize and respond to OHCA. Information on structural differences in neighborhood-level healthcare infrastructure, medical insurance, and access to advanced cardiac care, could inform why there may be differences in comorbidities in OHCA patients between the EMS agency groups. Additional information on barriers to accessing and activating emergency care in Black and Hispanic communities, including language (e.g., delays in communicating with dispatchers or for dispatchers to initiate instructions for bystander CPR), immigration status and fear of calling 9-1-1,21, 22 would provide additional insights into factors contributing to survival differences.
Although EMS agencies with a majority Black/Hispanic catchment area have lower mean survival and likely higher barriers to accessing emergency cardiac arrest care than those with a majority White catchment area, some of these agencies have achieved exceptional survival rates for OHCA in the U.S. In a recent analysis, we found that 12 of 82 (14.6%) EMS agencies with a majority Black/Hispanic catchment area had OHCA survival in the top quartile in the CARES registry.2 Understanding how these 12 agencies overcome barriers for prehospital care to achieve these high survival rates is crucial for defining best practices for EMS and providing additional insights for interventions to reduce disparities in OHCA survival. We are currently conducting qualitative interviews during site visits at some of these EMS agencies, as part of an NHLBI-funded study (Reducing Ethnic-racial Disparities in Cardiac Arrest Survival Outcomes [RED-CASO]; R01HL160734).
Our findings should be interpreted in the context of the following limitations. First, the survey was limited to EMS agencies in CARES, which has a coverage area of ~53% of all U.S. residents. Moreover, we restricted our survey to agencies with an annual case volume of at least 10 OHCA cases. Therefore, our findings may not be representative of non-participating agencies, especially those with low OHCA volume (e.g., rural areas). Nevertheless, to date, this is the only study to compare practices between EMS agencies with majority Black/Hispanic and White catchment areas. Second, the survey was usually completed by the EMS agency director, who was deemed as the individual most knowledgeable about the range of survey questions. Given the logistical difficulty of obtaining survey responses, responses were not independently verified for accuracy. Moreover, responses on fire and police first responders are from the EMS respondent’s perspective. Third, although our model to risk-standardize EMS agency-level survival incorporated many of the Utstein variables, data on certain variables (e.g., patient comorbidities, social determinants of health, implicit and explicit bias) were unavailable and could explain residual survival differences. Finally, it is important to emphasize that we cannot determine a causal relationship between the identified practices and EMS survival due to the observational study design, and these practices may be surrogate measures of other key practices.
In a large U.S. registry, we identified differences in dispatcher, first responder, and EMS practices for OHCA between agencies with majority Black/Hispanic and White catchment areas. These practice differences may partly account for disparities in OHCA survival between the two EMS agency groups.
CS
Dr. Comilla Sasson
Emergency Medicine · General Emergency MedicineUniversity of Colorado Anschutzcomilla