Early Coronary Angiography and Survival After Out-Of-Hospital Cardiac Arrest

Abstract

Background

While out-of-hospital cardiac arrest is commonly due to acute myocardial infarction, it is unknown whether early coronary angiography is associated with improved survival in these patients.

Methods and Results

Using data from the Cardiac Arrest Registry to Enhance Survival (CARES), we identified 4,029 adult patients admitted to 374 hospitals after successful resuscitation from out-of-hospital cardiac arrest due to ventricular fibrillation, pulseless ventricular tachycardia or unknown shockable rhythm between January 2010 and December 2013. Early coronary angiography (occurring within one calendar day of cardiac arrest) was performed in 1953 (48.5%) patients, of whom 1253 (64.2%) received coronary revascularization. Patients who underwent early coronary angiography were younger (59.9 vs. 62.0 years); more likely to be men (78.1% vs. 64.3%), have a witnessed arrest (84.6% vs. 77.4%), and have ST-elevation myocardial infarction (32.7% vs. 7.9%); and less likely to have known cardiovascular disease (22.8% vs. 35.0%), diabetes (11.0% vs. 17.0%), and renal disease (1.8% vs. 5.8%; P<0.01 for all comparisons). In analysis of 1312 propensity score-matched pairs, early coronary angiography was associated with higher odds of survival to discharge (OR: 1.52, [95% CI 1.28–1.80]), P<0.0001) and survival with favorable neurological outcome (OR 1.47, [95% CI 1.25–1.71], P<0.0001). Further adjustment for coronary revascularization in our models significantly attenuated both ORs, suggesting that revascularization was a key mediator of the survival benefit.

Conclusions

Among initial survivors of out-of-hospital cardiac arrest due to VF or pulseless VT, we found early coronary angiography was associated with higher odds of survival to discharge and favorable neurological outcome.

Out-of-hospital cardiac arrest is common, affecting nearly 300,000 people annually in the U.S., and is associated with low survival rates of <10%.¹ Although acute myocardial infarction (MI) is a frequent cause of cardiac arrests due to ventricular fibrillation (VF) or pulseless ventricular tachycardia (VT), the effectiveness of a strategy of early coronary angiography and revascularization is unclear in this setting, as cardiac arrest patients were excluded from clinical trials of early coronary intervention in acute MI.^2–5

Several observational studies have previously reported higher rates of survival in patients who receive early coronary angiography after initial resuscitation from out-of-hospital cardiac arrest.^6–11 However, these studies were limited by single-center and retrospective study designs, often lacked a control group that did not receive coronary angiography, and/or used statistical methods that did not account for indication bias. Nonetheless, quantifying the benefit of early coronary angiography in patients with out-of-hospital cardiac arrest has important implications, given current guidelines recommend regionalizing post-cardiac arrest care in specialized centers that are capable of performing coronary angiography and percutaneous coronary intervention (PCI).¹² Moreover, while professional societies recommended immediate coronary angiography in out-of-hospital cardiac arrest with ST-elevation myocardial infarction (STEMI), and in patients who do not have obvious non-coronary causes but no ST-elevation, it remains unclear whether these have been applied into practice.¹³

To address this gap in knowledge, we leveraged data from the Cardiac Arrest Registry to Enhance Survival (CARES) - a large national prospective registry of out-of-hospital cardiac arrests to examine the association of early coronary angiography (performed within one calendar day of cardiac arrest) and survival. We used a matched propensity score analysis to explicitly account for indication bias. Given that patients with cardiac arrest are at risk for significant neurological disability, we also examined survival with favorable neurological outcome. Finally, we examined whether the benefit of early coronary angiography on our study outcomes was mediated by greater use of revascularization in the early coronary angiography group.

Methods

Data Source

CARES is a large prospective registry of out-of-hospital cardiac arrest in the United States that was established in 2004 as a collaboration between the Emory University School of Medicine and the Center for Disease Control and Prevention. Since inception, CARES has grown to involve over 800 emergency medical service (EMS) agencies and 1,300 hospitals across 23 states, encompassing a catchment area of over 80 million people. Details of CARES’ study population, data collection, data reporting, and EMS and cardiac arrest protocols have been previously reported.^14–16 In brief, CARES collects data on all 9-1-1 activated out-of-hospital-cardiac arrest events of non-traumatic etiology, which involve attempted resuscitation by EMS or first responder units. Data collection in the registry is based on the Utstein template, and includes patient demographics, cardiac arrest characteristics, EMS response, and in-hospital outcomes at enrolling sites.¹⁵ In January 2010, a few optional elements were added to the CARES data collection form, which included whether the patient underwent coronary angiography during hospital stay. The current study was based on data collected since that time.

Study Population

We identified 21,269 patients aged 18 years or older with an out-of-hospital cardiac arrest due to VF, pulseless VT or unknown shockable rhythm between January 1, 2010 and December 31, 2013. Since patients who died in the field were not eligible to receive coronary angiography, we restricted our sample to patients who had sustained return of spontaneous circulation (ROSC) of at least 20 minutes and survived to hospital admission (n = 8,134). Given that data on coronary angiography was an optional data element, we excluded 3668 patients from sites that did not report data on performance of coronary angiography. We also excluded patients treated at hospitals that did not perform coronary angiography on any cardiac arrest patient and were also confirmed to not have a cardiac catheterization laboratory using the American Hospital Association 2010 dataset (n=65 patients at 29 hospitals). This was done in order to ensure that coronary angiography was available at all study hospitals. Finally, since we were interested in examining whether coronary angiography performed within one day of cardiac arrest was associated with improved survival, we also excluded patients that were missing data on the date of coronary angiography (n = 370) or survival (n = 2). Our final study population comprised 4,029 patients with out-of-hospital cardiac arrest due to shockable rhythms at 374 hospitals (Figure 1).

Figure 1.

Study cohort

VF: Ventricular Fibrillation; VT: Ventricular Tachycardia;

ROSC: Return of Spontaneous Circulation;

Study Variables

The primary exposure variable was performance of early coronary angiography - defined as coronary angiography performed within 1 calendar day of the cardiac arrest.

Patient-level data included subject age (categorized as 18–54, 55–64, 65–74, ≥75), sex, race (white, black, other or unknown), initial cardiac arrest rhythm (VF, VT, unknown shockable rhythm), location of arrest (home, public location, healthcare facility), whether cardiac arrest was witnessed by someone other than a first responder or the EMS provider, provision of bystander cardiopulmonary resuscitation (CPR), use of an automated external defibrillator (AED), and initiation of hypothermia in the field or in the hospital. Information was also available regarding additional data elements - placement of an advanced airway in the field, medical co-morbidities (heart disease, diabetes, hypertension, stroke, hyperlipidemia, renal disease, respiratory disease and cancer), and whether the initial diagnosis was STEMI. However, reporting on these latter variables as well as race was optional and therefore not consistently reported from all sites. We used indicator variables for missing data.

Study Outcomes

Our primary outcome was survival to discharge. Our secondary outcome was survival to discharge with favorable neurological outcome. Neurological status at discharge was assessed using the previously described and validated cerebral performance category (CPC) score.¹⁷ A CPC score of 1 denotes mild or no neurologic disability, 2 moderate neurologic disability, 3 severe neurologic disability, 4 coma or vegetative state, and 5 brain death. Based on prior work, which found prognostic post-discharge differences in survival between patients with a discharge CPC score of 1 vs. 2, we defined favorable neurological outcome as a CPC score of 1 (i.e., none or mild neurological disability).¹⁸

Statistical Analysis

We compared baseline characteristics according to whether or not patients underwent early coronary angiography using the two-sample t-test for continuous variables and the chi-square test for categorical variables. We also compared unadjusted rates of survival to discharge and favorable neurological outcome between the 2 groups, and calculated unadjusted odds ratios (OR).

To account for indication bias, we performed a matched 1:1 propensity score analysis to examine the association between early coronary angiography and survival. To accomplish this, we first constructed a non-parsimonious multivariable logistic regression model to determine each patient’s propensity of receiving early coronary angiography. All covariates listed in the Study Variables section were included in the model. Next, we used a greedy matching algorithm (i.e., without replacement) to match patients who did and did not receive early coronary angiography, using a caliper width that was less than 20% of the standard deviation of the logit of the propensity score.¹⁹ To ensure that both study groups were well balanced after propensity matching, we calculated a standardized difference for each covariate and compared them before and after propensity matching. Standardized differences measure the difference in the means between two groups expressed in units of standard deviation and help assess the degree of imbalance of a variable between two compared groups. Unlike p-values, standardized differences are not influenced by sample size.²⁰ A standardized difference of <10% suggests successful matching on propensity scores. Finally, we used the Cochran-Mantel-Haenszel test to examine the association between early coronary angiography and our study outcomes of survival to discharge and favorable neurological outcome.

Also, since hospitals may differ in their tendency to offer coronary angiography to cardiac arrest survivors, we examined whether the benefit of early angiography on survival was mediated by greater use of coronary revascularization in the early coronary angiography group. We constructed a multivariable logistic regression to examine the association between early coronary angiography and our study outcomes while adjusting for all variables listed in the Study Variables section. We then examined whether the strength of the association between early coronary angiography and our study outcomes was attenuated after additionally adjusting for receipt of coronary revascularization (PCI or coronary artery bypass graft [CABG] surgery).

To determine whether the association between early coronary angiography and survival outcomes was only confined to STEMI patients, we repeated our analysis after restricting our cohort to patients who did not have have STEMI. We re-estimated propensity scores for this cohort by constructing a separate multivariable logistic regression model and conducted a similar matched propensity score analysis as described above.

In order to explore the role of unmeasured confounders, we performed additional sensitivity analysis. We assumed an unmeasured confounder to be associated with survival with an odds ratio of 1.5, and a prevalence of 10% in the reference (delayed or no coronary angiography) group. We used the method of Lin et al²¹, to calculate the prevalence of X in the exposed cohort that would make the effect of early coronary angiography on survival non-significant. We then repeated the above calculations assuming an odds ratio of 2.0, and 3.0, respectively between X and survival. We also repeated the above analysis for survival with favorable neurological outcome.

All statistical analysis was performed using SAS 9.3 [SAS System for Windows, version 9.3. Cary, NC: SAS Institute. 2002–2010]. Graphs were generated using GraphPad Prism version 6.0d for Mac OS X [GraphPad Software, La Jolla, California, USA, www.graphpad.com]. The institutional review board at University of Iowa approved the study protocol and waived the requirement for informed consent.

Results

Among 4,029 patients successfully resuscitated from an out-of-hospital cardiac arrest due to VF or pulseless VT included in our study, early coronary angiography was performed in 1,953 (48.5%). Patients in the early angiography group were younger (59.9 years vs. 62.0 years), more likely to be male (78.1% vs. 64.3%), have a witnessed cardiac arrest (84.6% vs. 77.4%) and have an initial diagnosis of STEMI (32.7% vs. 7.9%). In contrast, patients with early coronary angiography were less likely to receive an advanced airway in the field (57.1% vs. 68.1%), be treated with therapeutic hypothermia in the hospital (56.4% vs. 60.7%) or have underlying heart disease, diabetes, stroke, and renal disease (P<0.01 for all comparisons, Table 1).

Table 1.

Baseline characteristics of study patients

	Early Coronary Angiography
Characteristic	Yes	No	P value
N	1953	2076
Age, yrs (mean ± SD)	59.9 ± 12.1	62.0 ± 16.4	<0.0001
Age, yrs			<0.0001
18–54	648 (33.2)	608 (29.3)
55–64	637 (32.6)	504 (24.3)
65–74	441 (22.6)	467 (22.5)
≥75	227 (11.6)	497 (23.9)
Female	428 (21.9)	741 (35.7)	<0.0001
Race			<0.0001
White	1088 (55.7)	1104 (53.2)
Black	182 (9.3)	297 (14.3)
Other	86 (4.4)	91 (4.4)
Unknown	597 (30.6)	584 (28.1)
Location of arrest			<0.0001
Home	1054 (54.0)	1249 (60.2)
Public	794 (40.7)	612 (29.5)
Healthcare facility	105 (5.4)	215 (10.4)
Arrest witnessed			<0.0001
Yes	1653 (84.6)	1605 (77.3)
No	300 (15.4)	470 (22.6)
Unknown	0 (0.0)	1 (0.1)
Bystander CPR	924 (47.3)	1035 (49.9)	0.11
AED used	643 (32.9)	718 (34.6)	0.27
Initial cardiac arrest rhythm			0.98
VF	1347 (69.0)	1426 (68.7)
VT	105 (5.4)	112 (5.4)
Unknown shockable rhythm	501 (25.7)	538 (25.9)
Hypothermia care in field	595 (30.5)	658 (31.7)	0.40
Hypothermia care in hospital			0.02
Yes	1102 (56.4)	1261 (60.7)
No	813 (41.6)	784 (37.8)
Unknown	38 (2.0)	31 (1.5)
Advanced airway placed in field			<0.0001
Yes	1115 (57.1)	1414 (68.1)
No	741 (37.9)	546 (26.3)
Unknown	97 (5.0)	116 (5.6)
STEMI			<0.0001
Yes	639 (32.7)	163 (7.9)
No	361 (18.5)	655 (31.6)
Unknown	953 (48.8)	1258 (60.6)
Co-morbidities
Heart disease	445 (22.8)	727 (35.0)	<0.0001
Diabetes	214 (11.0)	353 (17.0)	<0.0001
Hypertension	471 (24.1)	521 (25.1)	0.002
Hyperlipidemia	128 (6.6)	93 (4.5)	<0.0001
Renal disease	35 (1.8)	121 (5.8)	<0.0001
Respiratory disease	81 (4.2)	179 (8.6)	<0.0001
Stroke	25 (1.3)	85 (4.1)	<0.0001
Cancer	43 (2.2)	83 (4.0)	<0.0001
Unknown	703 (36.0)	641 (30.9)

All numbers in table represent number (percent) unless otherwise specified

P values derived using two sample t test for continuous age, and Chi-square test for all other variables;

AED: Automatic External Defibrillator; CPR: Cardio-pulmonary Resuscitation; SD: standard deviation; STEMI: ST-Elevation Myocardial Infarction; VF: Ventricular Fibrillation; VT: Ventricular Tachycardia.

Overall, 2,718 (67.5%) patients survived to hospital discharge, and 1,968 (48.8%) survived with a favorable neurological outcome. Compared with patients who did not, patients who underwent early coronary angiography had a higher unadjusted rate of survival to discharge (76.0% vs. 59.4%; OR 2.16, 95% CI 1.89 – 2.47, P<0.0001) as well as survival with favorable neurological outcome (58.7% vs. 39.6%; OR 2.18, 95% CI 1.92 – 2.47, P<0.0001) (Tables 2, 3).

Table 2.

Summary of study outcomes

	Early Coronary Angiography
Outcome	Yes	No	P value
N	1953	2076
Survival to discharge	1484 (76.0)	1234 (59.4)	<0.0001
Neurological outcome			<0.0001
Good cerebral performance	1147 (58.7)	821 (39.6)
Moderate disability	246 (12.6)	261 (12.6)
Severe disability	67 (3.4)	94 (4.5)
Coma, vegetative state	22 (1.1)	54 (2.6)
Unknown	2 (0.1)	4 (0.2)

All numbers in table represent number (percent) unless otherwise specified

Table 3.

Unadjusted and adjusted association of early coronary angiography with survival to discharge and favorable neurologic outcome

Cohort	Survival to Discharge		Favorable Neurological Outcome
	OR	95% CI	OR	95% CI
Unadjusted (Before Matching)# (N = 4,029)	2.16	1.89 – 2.47	2.18	1.92 – 2.47
Propensity Matched* (N = 2,624) – all patients	1.52	1.28 – 1.80	1.47	1.25 – 1.71
Propensity Matched for patients identified as not having a STEMI * (N = 620)	1.29	0.87 – 1.90	1.60	1.14 – 2.26

^#Odds Ratios and P values derived using logistic regression;

^*Odds Ratios derived using Cochran-Mantel-Haenszel Statistics stratifying by matched pairs; P values derived using McNemar’s test.

OR: Odds Ratio; CI: Confidence interval; STEMI: ST-Elevation myocardial infarction.

To account for indication bias, we estimated the propensity of receiving early coronary angiography for each patient using a multivariable logistic regression model (c-statistic = 0.746; Appendix Table 1). A total of 2,624 patients (1,312 in each group) were successfully matched using propensity scores. Compared to patients who were matched, patients who were unmatched were older, more likely to be women, of black race, have STEMI and a greater burden of co-morbidities but less likely to have a witnessed arrest and receive hypothermia (Appendix Table 2). Among matched patients, we confirmed that matching was successful in achieving covariate balance between the two groups as demonstrated by standardized differences of < 10% for all covariates (Figure 2 and Appendix Table 3). In propensity-matched analysis, early coronary angiography remained strongly associated with higher odds of survival to discharge (OR 1.52, 95% CI 1.28 – 1.80, P<0.0001) as well as survival with favorable neurological outcome (OR 1.47, 95% CI 1.25 – 1.71; P<0.0001).

Figure 2.

Standardized differences between cohorts before and after propensity score matching

AED: Automatic External Defibrillator; CPR: Cardio-pulmonary Resuscitation; SD: standard deviation;

STEMI: ST-Elevation Myocardial Infarction;

Before propensity matching, there were significant differences in a number of variables between patients who did and did not receive early coronary angiography (standardized differences > 10%). After propensity matching, the standardized difference for all variables was < 10% suggesting that propensity matching was successful in achieving covariate balance. Variables are presented in the descending order of magnitude of standardized differences prior to propensity matching.

In the overall cohort, 1253 (64.2%) patients underwent coronary revascularization in the early coronary angiography group, compared to 227 (10.9%) patients in the group that did not undergo early coronary angiography. In both groups, patients who underwent coronary revascularization had higher rates of survival to discharge (79.7% vs. 69.2% in the early coronary angiography group; and 92.5% vs. 55.1% in the group that did not receive early coronary angiography) and favorable neurological outcome (61.5% vs. 53.6% in the early coronary angiography group; and 77.5% vs. 34.9% in the group that did not receive early coronary angiography) compared to patients who did not receive coronary revascularization. After additional adjustment of coronary revascularization in a sequential model, we found that the association of early coronary angiography with survival to discharge was markedly attenuated (OR 1.59, 95% CI 1.35–1.86; P<0.0001 to OR 1.12, 95% CI 0.94–1.35; P=0.21). Similarly, the association between early coronary angiography and favorable neurological outcome was attenuated (OR 1.59, 95% CI 1.36 – 1.86; P<0.0001 to OR 1.25, 95% CI 1.05 – 1.49; P=0.01), suggesting that coronary revascularization, in large part, mediated the observed benefit of early coronary revascularization.

We also examined whether the benefit of early coronary angiography was consistent in patients without STEMI. After restricting the analyses to only patients who did not have STEMI on their ECG, a total of 620 patients were successfully matched (310 in each group) after re-estimating propensity scores in this cohort. The association between early coronary angiography and survival to hospital discharge became non-significant, (OR 1.29, 95% CI 0.87 – 1.90; P=0.2). However, survival with favorable neurological outcome remained strongly associated with early coronary angiography (OR 1.60, 95% CI 1.14 – 2.26; P<0.01) (Table 3).

Finally, we examined the role of unmeasured confounders in sensitivity analysis. We found that the association between early coronary angiography and survival to discharge would become non-significant if an unmeasured confounder “X” with an odds ratio of 1.5 with survival was 7 times more prevalent (69%) in the exposed cohort compared to the unexposed cohort (10%). For an odds ratio of 2.0, an unmeasured confounder would need to be 4 times more prevalent (41%), whereas for an odds ratio of 3.0, the unmeasured confounder would need to be 2.5 times more prevalent (27%) in the exposed cohort compared to its prevalence in the non-exposed cohort in order to eliminate the effect of early coronary angiography on survival (Appendix Table 4). The findings were similar with favorable neurological outcome.

Discussion

In a large prospective cohort of patients with out-of-hospital cardiac arrest due to VF and pulseless VT who survived to hospital admission, we found that approximately half the patients underwent coronary angiography within 1 calendar day of cardiac arrest. Early coronary angiography was associated with approximately 50% greater odds of survival to discharge and survival with favorable neurological outcome. The observed benefit of early coronary angiography was largely mediated by coronary revascularization and was consistent in the subgroup of patients who did not have an initial diagnosis of STEMI.

Our results differ from previous observational studies that have examined the association between early coronary angiography and survival after out-of-hospital cardiac arrest in a number of important ways. First, patients who do not receive early coronary angiography are generally sicker compared to patients who receive early coronary angiography, which can lead to indication bias. Indeed, we found that patients in our study who did not receive early coronary angiography were older, had a higher prevalence of co-morbidities and were less likely to have a witnessed cardiac arrest compared to patients who underwent early coronary angiography. Previous studies have not consistently used methods to account for indication bias and therefore may have over-estimated the effect of early coronary angiography on survival.²² In contrast, we used a matched propensity score design to account for indication bias. Second, most of the prior studies were small, retrospective, conducted in single-centers or lacked a standardized definition of cardiac arrest.^7,22,23 Moreover, many of these studies only included patients who were identified as having a STEMI or did not have a control group and therefore were limited in drawing inferences.^10,23,24 In contrast, CARES is the largest registry of out-of-hospital cardiac arrest in the United States that prospectively enrolls all out-of-hospital cardiac arrest patients from participating EMS agencies. The registry is based on the Utstein template with standardized definition of cardiac arrest, clinical variables and study outcomes. As a result, patients enrolled in our study are more likely to be representative of out-of-hospital cardiac arrest victims in the community.

The importance of early coronary angiography is underscored by the fact that acute MI is the most common etiology of out-of-hospital cardiac arrest, especially due to VF and pulseless VT.^23,25 Moreover, previous studies have shown that clinical and electrocardiographic criteria are unreliable in identifying coronary ischemia in cardiac arrest. In a French study of 435 patients with out-of-hospital cardiac arrest all of whom underwent coronary angiography, nearly 70% (301) patients did not have ST-elevation on ECG following resuscitation. However, 58% of such patients had at least 1 significant coronary lesion during coronary angiography (negative predictive value = 42%).¹⁰ Therefore, a strategy of routine coronary angiography has considerable appeal in this patient population, even if they do not have electrocardiographic evidence of a STEMI.

It is noteworthy that 50% of patients with cardiac arrest due to shockable rhythms did not undergo early coronary angiography despite surviving to hospital admission. The low utilization of early coronary angiography in this population may reflect the lack of a strong evidence base supporting improved outcomes, given that cardiac arrest patients were largely excluded from randomized controlled trials of early revascularization in patients with acute MI. Furthermore, treating physicians may be reluctant to refer patients for invasive procedures when prognosis for neurological recovery is unclear. In this study, we found that early coronary angiography was associated with improved outcomes in patients with out-of-hospital cardiac arrest due to shockable rhythms and this benefit was due to greater use of early revascularization. Although we used statistical methods to account for indication bias, causal interpretation is limited due to the observational nature of our study. However, given our study findings and the potential to improve survival in out-of-hospital cardiac arrest victims, there is an urgent need for randomized controlled trials to confirm the benefit of coronary angiography in patients with out-of-hospital cardiac arrest including its timing (immediate [<2 hours], early [<24 hours] or after neurological prognostication), especially in patients who are not identified to have ST-elevation on admission.

Our study findings should be interpreted in light of the following limitations. First, assignment of patients to early coronary angiography was not random, and indication bias would be of concern in such analysis. While we used a matched propensity score analysis to account for indication bias, and achieved balance on measured covariates after matching on propensity scores, potential for residual confounding from unmeasured patient-level variables remains. For example, variables that may influence selection of early coronary angiography as well as survival (such as resuscitation response time, neurological status at the time of admission, presence of shock) were unavailable. Likewise, details of coronary angiography (e.g., percent stenosis, number of vessels treated), and related complications were not available. Further studies will require putting these findings in the context of these other key variables that may play a role in both the decision to perform early coronary angiography and the outcomes for patients. Second, although our matching algorithm was successful in achieving covariate balance, only 65% of our sample was successfully matched. Patients who remained unmatched were sicker compared to matched patients, and therefore our findings may not generalize to such patients. Third, unlike 30-day survival, survival to hospital discharge may be influenced by hospital discharge practices (e.g., discharge to hospice). However, this is less likely to be an issue given that early coronary angiography was also associated with favorable neurological outcome in this study. Fourth, it is possible that the association between early coronary angiography and our study outcomes is confounded by other unmeasured hospital processes of care (e.g., quality of post-resuscitation care) which may have led to improved outcomes. However, we found that the effect of early coronary angiography on improving survival and neurological outcome was largely mediated by greater use of coronary revascularization in this group. Fifth, some variables (e.g., STEMI and co-morbidities) were missing in a large proportion of patients, as sites were not required to report these variables. Moreover, race variable was also missing in a subset of patients, as several EMS agencies are reluctant to report this variable. Variables with missing values were coded as “unknown” in our propensity score model, which may have introduced bias. Finally, our study only included hospitals that performed coronary angiography on at least one cardiac arrest survivor. Whether patients admitted to hospitals lacking cardiac catheterization laboratory should be transferred to hospitals with these facilities was not addressed in this study.

Conclusions

In conclusion, we found that early coronary angiography is associated with higher rates of survival and favorable neurological outcome in patients successfully resuscitated from an out-of-hospital cardiac arrest due to VF or VT. Given the observational nature of the data, further research with randomized controlled trials is needed to confirm this potential benefit.