Does out-of-hospital cardiac arrest survival differ by EMS agency type in the U.S.? Insights from the CARES registry

Abstract

Background

There are limited data on whether out-of-hospital cardiac arrest (OHCA) survival differs by emergency medical service (EMS) agency type in the U.S.

Methods

Within the Cardiac Arrest Registry to Enhance Survival, we surveyed participating EMS agencies with ≥20 OHCA annually during 2015–2019. Agencies were categorized as fire-based, private, government-based, or other (including hospital-based). Using hierarchical logistic regression, we computed the risk-standardized survival rate (RSSR) to hospital admission and discharge for each EMS agency and examined for differences in RSSR across agency types.

Results

Of 577 eligible EMS agencies, 470 (81.5 %) completed the survey. Overall, 40.0 % of agencies were fire-based, 35.0 % were private, 17.3 % were government-based, and 7.7 % were other. Mean agency-level RSSR to hospital admission was 27.8 % ± 3.6 % and was highest in fire-based EMS agencies (28.9 %) and lowest in agencies that were private or other (26.8 %; P < 0.001 for comparison across all groups). Mean agency-level RSSR to hospital discharge was 10.1 % ± 1.8 % and was highest in fire-based agencies (10.3 %) and lowest in agencies that were private or other (9.7 %; P < 0.003 across all groups). Fire-based agencies were more likely to have higher numbers of paramedics, lower annual number of dispatches per paramedic, more hours of OHCA training during orientation, and shorter arrival times on the scene and transport times to the hospital.

Conclusions

In the U.S., OHCAs attended by fire-based agencies were associated with modestly higher rates of survival to hospital admission compared to OHCAs attended by other EMS agency types. Several resuscitation practices differed by EMS agency type.

Introduction

Each year, more than 350,000 individuals experience a non-traumatic out-of-hospital cardiac arrest (OHCA) in the U.S.1, 2 Survival to hospital discharge after EMS treated OHCA remains at approximately 10%.3, 4 Although emergency medical service (EMS) agencies are the cornerstone of prehospital response to OHCA, there is substantial variation in OHCA survival across agencies.⁵ One possible reason for this variation may be structural, as there are a range of different EMS staffing models (e.g., fire-based, private, government). However, to date, data on OHCA survival rates by EMS agency type in the U.S. are lacking. Moreover, whether there are differences in resuscitation training or practices by EMS agency type is unknown. To address this gap in knowledge, we surveyed EMS agencies participating in a national U.S. registry and examined whether resuscitation practices and OHCA survival differed by EMS agency type.

Methods

Data sources

We used the national Cardiac Arrest Registry to Enhance Survival (CARES) registry, a prospective, multicenter, observational registry of OHCA in the U.S. with a catchment area of over 173 million residents, representing approximately 53 % of the U.S. population. The design of the registry has been previously described.6, 7 Briefly, all patients with a confirmed OHCA and for whom resuscitation is attempted are identified and followed by EMS agencies. Standardized international Utstein definitions are used for uniformity in clinical variables and outcomes across EMS agencies.⁸

In 2022, we surveyed 577 EMS agencies in CARES with ≥20 OHCAs annually during 2015 to 2019. Details of the survey design have been previously described.⁹ A total of 470 (81.5 %) EMS agencies completed the survey. The CARES registry then linked survey responses from EMS agencies to CARES data and provided a merged deidentified dataset for analytical purposes for this study question, which was pre-specified. This study was approved by Saint Luke’s Hospital’s institutional review board, which waived the requirement for informed consent as the study involved deidentified data.

Independent variable and study outcomes

For this study, we categorized agencies by EMS agency type, based on their survey responses as fire department, private (including profit and nonprofit entities), government-based, or other (including hospital-based). Fire-based agencies generally were two-tiered BLS and ALS systems, whereas the other EMS agency types were ALS only. In the event that multiple EMS agencies responded, we assumed the primary agency reported the OHCA event. The primary outcome was survival to hospital admission for OHCA, and the secondary outcome was survival to discharge.

Statistical analysis

We used descriptive statistics to compare survey responses on resuscitation training and practices by EMS agency types, with Student t-tests for continuous variables and chi-square tests for categorical variables. Survey questions of particular interest included annual cardiac arrest dispatch run burden per paramedic, number of hours of OHCA training during orientation, and times of arrival at the scene and transport to the hospital.

We then compared OHCA survival rates by EMS agency type. To account for differences in patient case-mix across EMS agencies, risk-standardized survival rates (RSSR) to hospital admission were calculated for each agency using multivariable hierarchical logistic regression, with EMS agency modeled as a random effect and the following as fixed effects: age, sex, location of arrest (home, workplace, street/highway, industrial building, recreational facility, or other), whether the arrest was witnessed, initial cardiac arrest rhythm, 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]).¹⁰ Using the agency-specific estimates (i.e., random intercepts) from the hierarchical model, we calculated RSSR to hospital admission for each EMS agency by multiplying the ratio of each agency’s predicted to expected number of survivors to hospital admission with the overall unadjusted rate.11, 12 We further adjusted for EMS response times in the model to determine if survival differences were attenuated. We then examined for differences in RSSRs across agency types using analysis of variance (ANOVA). These analyses were then repeated for the secondary outcome of survival to hospital discharge.

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

The 470 EMS agencies that completed the survey formed the study cohort, and these agencies responded to 181,707 OHCAs during the study period. One agency did not respond to the question of EMS agency type, so only 469 could be categorized. Overall, 188 (40.0 %) were fire-based, 164 (35.0 %) were private, 81 (17.3 %) were government-based, and 36 (7.7 %) were categorized as other.

Differences in resuscitation training and practices by type of EMS agency are summarized in Table 1. Fire-based agencies had a higher number of paramedics and lower annual dispatches per paramedic compared with other EMS agency types (209 annual dispatch runs per paramedic on staff, as compared to private [956] and government-based [435] agencies). Fire-based agencies also spent more time on OHCA training during orientation (10.4 h) as compared with private (2.4 h) and government-based (5.6 h) agencies (P < 0.001). Fire-based agencies had shorter times of arrival at the scene and shorter transport times to the hospital. Of note, fire-based agencies more frequently used mechanical CPR and impedance threshold devices than other agencies, although randomized trials have not found a benefit with either.13, 14

Table 1.

Resuscitation characteristics and practices by EMS agency type. Agencies were categorized as fire-based, private agencies, government-based (e.g., city or county), or other (includes hospital-based agencies).

	EMS agency type
	Fire	Private	Gov’t	Other	P-value
	n = 188	n = 164	n = 81	n = 36
No. of Paramedics at EMS Agency	121 ± 255	47 ± 65	70 ± 63	91 ± 94	<0.001
Annual no. of Dispatches Attended by a Paramedic	209 ± 16	956 ± 2372	435 ± 438	507 ± 822	<0.001
Hours of Cardiac Arrest Training During Orientation	10.4 ± 30	3.4 ± 5.3	5.6 ± 5.6	6.8 ± 12.9	0.011
Telephone Instructions for CPR to Bystanders					<0.001
None of the time	22 (11.8 %)	7 (4.3 %)	6 (7.4 %)	2 (5.6 %)
<25 % of the time	3 (1.6 %)	7 (4.3 %)	6 (7.4 %)	2 (5.6 %)
26–50 % of the time	18 (9.6 %)	10 (6.1 %)	4 (4.9 %)	4 (11.1 %)
51–75 % of the time	17 (9.1 %)	16 (9.8 %)	12 (14.8 %)	8 (22.2 %)
76–100 % of the time	76 (40.6 %)	54 (32.9 %)	39 (48.1 %)	11 (30.6 %)
Unknown	51 (27.3 %)	70 (42.7 %)	14 (17.3 %)	9 (25.0 %)
Use of a Mechanical CPR Device During OHCA	171 (91.0 %)	124 (75.6 %)	67 (82.7 %)	25 (69.4 %)	<0.001
Always Use Mechanical CPR Device During OHCA	135 (71.8 %)	67 (40.9 %)	51 (63.0 %)	14 (38.9 %)	<0.001
Impedance Threshold Device Use for Simulation Training	51 (27.1 %)	16 (9.8 %)	5 (6.2 %)	6 (16.7 %)	<0.001
Mean EMS Response Time From Dispatch to Scene of OHCA					<0.001
<4 min	14 (9.7 %)	1 (0.8 %)	1 (1.4 %)	0 (0.0 %)
4–6 min	88 (60.7 %)	26 (21.1 %)	6 (8.3 %)	4 (15.4 %)
7–9 min	34 (23.4 %)	75 (61.0 %)	45 (62.5 %)	15 (57.7 %)
10–12 min	7 (4.8 %)	15 (12.2 %)	16 (22.2 %)	6 (23.1 %)
13–15 min	2 (1.4 %)	6 (4.9 %)	3 (4.2 %)	0 (0.0 %)
>15 min	0 (0.0 %)	0 (0.0 %)	1 (1.4 %)	1 (3.8 %)
Mean Transport Time From OHCA Scene to Receiving Hospital					<0.001
<4 min	4 (2.1 %)	1 (0.6 %)	0 (0.0 %)	0 (0.0 %)
4–6 min	27 (14.4 %)	12 (7.3 %)	1 (1.2 %)	0 (0.0 %)
7–9 min	55 (29.3 %)	36 (22.0 %)	8 (9.9 %)	7 (19.4 %)
10–12 min	49 (26.1 %)	60 (36.6 %)	28 (34.6 %)	11 (30.6 %)
13–15 min	21 (11.2 %)	20 (12.2 %)	17 (21.0 %)	6 (16.7 %)
>15 min	32 (17.0 %)	35 (21.3 %)	27 (33.3 %)	12 (33.3 %)

Abbreviations: CPR, cardiopulmonary resuscitation; EMS, emergency medical service, OHCA, out-of-hospital cardiac arrest.

Across the study cohort, mean agency-level RSSR to hospital admission for OHCA was 27.8 % ± 3.6 %. However, RSSR to hospital admission was highest in fire-based EMS agencies (28.9 %) and lowest in private and other-based agencies (26.8 %; p < 0.001 for comparison across all groups) (Table 2). Additional adjustment for EMS response times resulted in limited attenuation of survival differences (Supplementary Appendix Table). A similar pattern was seen for the secondary outcome of survival to hospital discharge. Mean agency-level RSSR to hospital discharge was 10.1 % ± 1.8 % and was highest in fire-based EMS agencies (10.3 %) and lowest in agencies that were private or other (9.7 %; P < 0.003 across all groups) (see Table 2).

Table 2.

OHCA survival by EMS agency type. Agency-level rates of survival to hospital admission and hospital discharge were adjusted for patient demographic and cardiac arrest factors using hierarchical regression models.

	EMS agency type
	Fire n = 188	Private n = 164	Gov’t n = 81	Other n = 36	P-value
Risk-standardized survival to admission rate, % ±SD	28.9 ± 3.6	26.8 ± 3.2	27.6 ± 3.5	26.8 ± 4.0	<0.001
Risk-standardized survival rate to hospital discharge, % ±SD	10.3 ± 1.9	9.7 ± 1.7	9.8 ± 1.7	9.8 ± 1.7	0.003

Abbreviations: EMS, emergency medical service; SD, standard deviation.

Discussion

In this study of EMS agencies participating in a large national OHCA registry, we found that OHCA survival differed by EMS agency type, with higher survival at fire-based EMS agencies. We also found differences in some processes of care between types of EMS agencies. Fire-based agencies had more paramedics employed and lower annual dispatch run burden per paramedic on staff, devoted more time to OHCA training during orientation, and had shorter times of arrival at the scene and transport times to the hospital.

To date, few surveys on EMS agency resuscitation practices for OHCA exist in the literature, and until recently, none have linked survey responses by EMS agencies to OHCA survival. One recent study found that certain EMS practices and characteristics (e.g., frequent simulation training, use of CPR feedback devices, specialized transport protocols) were associated with higher EMS agency rates of OHCA survival.¹⁵ Two other studies have found that police response to OHCA varies across communities and may impact OHCA survival.16, 17 To date, an examination of OHCA practices and survival by EMS agency type has not been performed. We found differences in OHCA survival by EMS agency structure (fire-based, private, and government-based agencies), as well as differences in resuscitation training and practices.

In this study, fire-based agencies had more paramedics employed and fewer annual dispatches per paramedic employed. These indicators suggest either lower dispatch volume at fire-based agencies, a higher proportion of paramedics in first responder and EMS vehicles at fire-based agencies, or a combination of both. Unfortunately, the survey did not collect information on how many paramedics, on average, responded to an OHCA to help us better gauge whether higher OHCA survival at fire-based agencies was due to this. It was also notable that fire-based agencies devoted more time to OHCA training during staff orientation. This is important as a modifiable metric such as resuscitation training could be emulated by other EMS agency types.

This study has important limitations that should be taken into consideration when interpreting the data. First, causal inferences derived from these results should be interpreted carefully. Fire-based agencies may care for patients who are less ill than those cared for by other agencies. Although we accounted for differences in our derivation of RSSRs, some unobserved features may not be accounted for. Moreover, the resuscitation training and practices we identified as different across agency types may not be the reason for higher OHCA survival at fire-based agencies. Second, we only looked at EMS agencies in the CARES registry, and our findings may not be generalizable to non-participating agencies in the U.S. Third, arrival times are self-reported and do not reflect time at a patient’s bedside. Finally, we did not have information on funding and resources, the average number of paramedics and emergency medical technicians responding to an OHCA, and other quality improvement initiatives during the study period by EMS agency type.

In conclusion, in the U.S., OHCAs attended by fire-based agencies were associated with modestly higher rates of survival to hospital admission compared to OHCAs attended by other EMS agency types. Certain resuscitation practices related to training, staffing, and response times differed by EMS agency type.

CRediT authorship contribution statement

Janaki L. Nallamothu: Writing – review & editing, Writing – original draft, Conceptualization. Saket Girotra: Writing – review & editing, Supervision. Kevin F. Kennedy: Formal analysis. Paul S. Chan: Writing – review & editing, Supervision.

Funding

• •Drs. Girotra and Chan receive research funding from the National Heart, Lung, and Blood Institute (R01HL160734) and from the American Heart Association for editorial work.
• •Dr. Girotra is also supported by funding from the National Heart, Lung, and Blood Institute (R01HL166305).

None of the above funders had any role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation,

Declaration of competing interest

Dr. Chan receives funding from the American Heart Association and is a consultant for Optum Rx.

Appendix A. Supplementary material

The following are the Supplementary material to this article: