Abstract
Several studies have reported racial/ethnic variation in out-of-hospital cardiac arrest (OOHCA) characteristics, which engendered varying conclusions. We performed a systematic review and meta-analysed the evidence for differences in OOHCA survival when considering the patient’s race and/or ethnicity.
We searched Medline and EMBASE databases up to and including 1 Oct 2011 for studies investigating racial/ethnic differences in OOHCA characteristics, supplemented by manual searches of bibliographies of relevant studies. We selected studies of any relevant design that measured OOHCA characteristics and stratified them by ethnic group. Two independent reviewers extracted information on the study population, including: race and/or ethnicity, location, age and OOHCA variables as per the Utsein template. We performed a meta-analysis of the studies comparing the black and white patients.
1701 potentially relevant articles were identified in our systematic search. Of these, 22 articles describing original studies were reviewed after fulfilling our inclusion criteria. Although 19 studies (18 within the United States (US)) compared the black and white population, only 15 fulfilled our quality assessment criteria and were meta-analysed. Compared to white patients, black patients were less likely to receive bystander cardiopulmonary resuscitation (OR = 0.66, 95%CI = 0.55–0.78), have a witnessed arrest (OR = 0.77, 95%CI = 0.72–0.83) or have an initial ventricular fibrillation/ventricular tachycardia arrest rhythm (OR = 0.66, 95%CI = 0.58–0.76). Black patients had lower rates of survival following hospital admission (OR = 0.59, 95%CI = 0.48–0.72) and discharge (OR = 0.74, 95%CI = 0.61–0.90).
Our work highlights the significant discrepancy in OOHCA characteristics and patient survival in relation to the patient’s race, with the black population faring less well across all stages. Most studies compared black and white populations within the US, so research elsewhere and with other ethnic groups is needed. This review exposes an inequality that demands urgent action.
Introduction
Out-of-hospital cardiac arrest (OOHCA) is a leading cause of cardiovascular mortality1,2 with significant variation in survival following successful initial resuscitation.3–5 Factors increasing survival following OOHCA include: younger age, ventricular fibrillation/ventricular tachycardia (VF/VT) as the initial cardiac arrest rhythm, having a witnessed cardiac arrest, receiving bystander cardiopulmonary resuscitation (CPR) and short emergency medical response times.6–8 This is emphasised in a recent consensus statement issued by the American Heart Association (AHA).9
Differences occur between racial or ethnic groups in the frequency of cardiovascular diseases;10 however, few studies have examined cardiac arrest characteristics and survival by racial or ethnic groups.11–13 Studies do show that there are ethnic differences in OOHCA characteristics pertinent to survival, including bystander CPR,11,14–16 witnessed arrest11,17–18 and the presence of an initial shockable rhythm.11,12,17,19 On the other hand, other studies show there are no racial differences in these OOHCA characteristics.12,15,16,20 To our knowledge, these studies have not been systematically reviewed.
Our aims were to review and summarise the literature pertaining to the association of race or ethnicity with OOHCA characteristics.
Methods
Databases and sources
We searched the Medline (1948 – 1 October 2011) and EMBASE (1980 – 1 October 2011) databases (via Ovid SP). Reference lists of all relevant studies were scanned to identify any further studies. Both searches and data extractions were independently performed by study authors A Shah and K Shah.
Search keywords and terms
Our primary search included the following keywords: heart arrest, cardiac arrest, out of hospital cardiac arrest, and ventricular fibrillation or ventricular tachycardia combined with the keywords ethnicity or race. We also searched for the subterms blacks, Asians, Chinese, race and religion (see online Appendix A).
Inclusion and exclusion criteria
Studies of any relevant design that were performed on adults >18 years of age were included, provided that they presented original data. Race or ethnicity had to be recorded and be included as a stratified variable when analysing data. A study was included in our retrospective study if it measured any of the following OOHCA characteristics: bystander CPR, witnessed arrest, initial VF/VT arrest rhythm, response times or survival at hospital admission and/or following discharge. Our search was global, with no language restrictions. We included only published articles but excluded abstracts, as the latter would not have undergone peer review.
Selection of articles and extraction of data
After titles and abstracts were screened for relevance, full-text versions were obtained and authors were contacted if more data clarification was required or if papers were not accessible. One study21 used data from the same database as a previously-published study,11 therefore it was excluded. For each study included, the following information was recorded when possible: study population including race and/or ethnicity data, location, subject age, bystander CPR, initial arrest rhythm (VF/VT), response times, witnessed arrest and survival with respect to hospital admission and/or discharge. The variables chosen were derived from those recommended in the Utstein template.22,23
Data synthesis and statistical analysis
Most studies investigated black and white racial/ethnic populations, and pooling and synthesis of data was seldom possible for other ethnic groups. The studies comparing black and white populations were meta-analysed. Many of these studies were adjusted for confounding factors (within a regression model), but had different endpoints. Initially, we used the actual cohort numbers reported or derived from relevant studies to generate non-adjusted, pooled odds-ratios (ORs) for each variable. When studies provided adjusted ORs, after controlling for confounders, we used these in place of the actual cohort numbers to calculate our pooled adjusted OR, whenever possible. We found that data reporting for response times varied significantly and was not suitable for meta-analysis.
Heterogeneity was examined using the standard chi2 test. The pooled effect was calculated using a random effects model, with analysis performed using Review Manager Version 5.1 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2011) and Comprehensive Meta-analysis (Version 2.0, Biostat Inc, 2006).
Study quality, bias and sensitivity analysis
Each study included for meta-analysis was assessed for quality (see online Appendix B). The Utstein template was used in 11 of the 15 studies to define OOHCA variables, providing consistency in outcome measures. Eleven studies did not report on how ethnicity was classified. Five studies measured all of the OOHCA variables that we studied in this meta-analysis. We constructed funnel plots to assesspublication bias (see online Appendix C). We performed sensitivity analyses by including studies that scored eight and above, in terms of study quality, in a further meta-analysis (see online Appendix B).
Terminology and concepts relating to ethnicity, race and OOHCA
We followed the principles of the concepts and terminology in Bhopal’s glossary;24 however, in our text we provide the terminology as used by the original authors. Readers can access interpretations in the glossary, for example that the terms white and Caucasian are synonyms.
Our use of terms complies with the Utstein templates (unless otherwise stated), initially published in 199122 and updated in 2004.23 ‘Bystander CPR’ denotes CPR attempts made by a bystander prior to arrival of the emergency services, ‘witnessed arrest’ are heart attacks seen by a witness before arrival of the paramedics and ‘initial rhythm’ denotes the first recorded rhythm, post arrest. Reported response times varied between studies, but were specified when describing each study.
Results
A total of 1701 articles were screened using the title and/or abstracts. We identified and reviewed 40 papers of potential relevance. A further 18 papers were excluded, as they did not contain race nor ethnicity as a variable, or did not provide data by race or ethnicity.13,25–40 Four studies appeared twice and were excluded as duplicates.11,13,21,41 Two studies were excluded because they investigated OOHCA in a paediatric population.42,43 Another four papers were identified from other’s citations and methods.18,44–46 One relevant study was identified via word of mouth from a colleague.47 Thus, a total of 22 studies met our inclusion criteria.4,11,12,14–20,41,44–54 (online Appendix D). Of these, 19 studies compared black and white populations: 15 of them fulfilled our quality assessment criteria and were meta-analysed.
Table 1 summarises location, period of data collection, the method used to define race or ethnicity, the total numbers in each racial or ethnic group, the study design as well as the primary objective of each study.
Summary of contextual details of studies: publication, location, timing, study design, definition of race or ethnicity concepts/terminology, sample size, data source and primary objective
| a) Author, Year b) Location c) Period of data collection | a) Basis of sample b) Data source | a) Method used to define race/ethnicity b) Race/ethnicity categories used by authors | Total number and number in each racial/ethnic group | Study design | Primary objective of study |
|---|---|---|---|---|---|
| Studies Within US | |||||
| a) Wilcox-Gok44 b) NJ, US c) 1985 | a) OOHCA cases b) Medical Intensive Care Unit forms | a) Not specified b) Black and white | Total – 1879 Black – 190 White – 1659 Other – 30 | Cohort study | To determine the role of CPR and advanced life support in improving OOHCA survival. |
| a) Becker et al., 199311 b) Chicago, IL, US c) 01 Jan 1987 – 31 Dec 1988 | a) OOHCA cases b) Data from EMS and hospital records | a) Not specified55 b) Black and white | Total – 6451 Black – 2910 White – 3207 Other – 334 | Cohort study | Comparing racial differences in the incidence and characteristics of OOHCA. |
| a) Cowie et al., 199341 b) Seattle, WA, US c) 30 May 1984 – 31 Jul 1986 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 977 Black – 117 White – 860 | Cohort study | Comparing OOHCA characteristics in black and white populations. |
| a) Brookoff et al., 199415 b) Memphis, TN, US c) 01 Mar 1989 – 05 Jun 1992 | a) OOHCA cases b) Data from prospective trial of first responder defibrillation56 | a) Not specified b) Black and white | Total – 1068 Black– 603 White – 465 | Cohort study | Comparing bystander CPR in OOHCA patients and ethnicity. |
| a) Sayegh et al., 199916 b) MI, US c) 1991 – 1996 | a) OOHCA cases b) EMS data registry | a) Vital statistics and hospital records b) Black and white | Total – 1317 Black – 378§§ White – 939 | Cohort study | Assessing whether race or socioeconomic status affects survival at hospital discharge, following OOHCA. |
| a) Chu et al., 199812 b) MI, US c) 01 Jan 1991 – 31 Dec 1994 | a) OOHCA cases b) EMS data registry and hospital medical records | a) Not specified b) Black and white | Total – 1690 Black – 223 White – 1467 | Cohort study | Determining if race, when controlled for socioeconomic status, is an independent risk factor for survival at hospital discharge, following OOHCA. |
| a) Lindholm et al., 199850 b) Kansas City, KS, US c) 01 Jan 1992 – 31 Dec/1994 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 832 (numbers by ethnic group not stated) | Cohort study | Assessing the effect of the return of spontaneous circulation on survival at hospital discharge, against other predictors of survival, following OOHCA. |
| a) Sweeney et al., 199852 * b) NC, US c) 1992 – 1995 | a) OOHCA cases b) EMS and fire department records | a) Not specified b) Black and white | Total – 243 (numbers by ethnic group not stated) | Trial and Cohort study | Assessing whether use of automated external defibrillators in addition to CPR improves survival to hospital discharge. |
| a) Groeneveld et al., 200348 b) US c) 1990 – 1999 | a) Hospitalised VF/cardiac arrest cases b) Medicare sources | a) Self-reported b) Black and white | Total – 5948 Black – 519 White – 5429 | Cohort study | Assessing racial disparity in the utilization of life-saving procedures, post cardiac arrest, and whether procedure-rate differences affect long-term survival. |
| a) Polentini et al., 200546 b) Milwaukee, WI, US c) 1992 – 2002 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 9170 (numbers by ethnic group not stated) | Cohort study | Investigating changes in incidence of each cardiac arrest rhythm between 1992 and 2002, and rates of survival at hospital admission and discharge for VF/VT arrests. |
| a) Galea et al., 200717 b) New York, NY, US c) 01 Apr 2002 – 31 Mar 2003 | a) OOHCA cases b) Data collection from on-scene first responders prospectively | a) Not specified b) Hispanic, black and white | Total – 4053 Hispanic – 636 Black – 1257 White – 1908 Other – 252 | Cohort study | Comparing incidence of OOHCA and 30-day patient survival following hospital discharge in three racial/ethnic groups. |
| a) Fairbanks et al., 200720 b) Rochester, NY, US c) 01 Jan 1998 – 21 Dec 2001 | a) OOHCA cases b) EMS registry data | a) Not specified b) Black and white | Total – 539 Black – 190 White – 274 Other – 28 Unknown race – 47 | Cohort study | To identify OOHCA factors that affect patient survival to 1 yr. |
| a) Liu et al., 20084† b) Milwaukee, WI, US c) 1995 – 2005 | a) OOHCA cases b) EMS data registry | a) Not specified b) African American and white | Total – 1702 African American – 405 White – 1188 Other – 109 | Cohort study | To determine variability in rates of survival to discharge in patients that were successfully resuscitated following OOHCA. |
| a) Vadeboncoeur et al., 200853 b) AZ, US c) Nov 2004 – Nov 2006 | a) OOHCA cases b) EMS data registry | a) Patient care report and hospital data b) Hispanic and non–Hispanic | Total – 1379 Hispanic – 273 Non–Hispanic – 1106 | Cohort study | Comparing bystander CPR in Hispanic versus non-Hispanic groups. |
| a) Benson et al., 200918 b) Los Angeles,CA, US c) 2000 – 2001 | a) OOHCA cases b) Subset analysis from CARE LA study57 | a) If classified as Latino origin on the 2000 or 2001 CDSMF (California Death Statistical Master File) and if classified as African American, then coded as such regardless of Latino origin. b) Latino, African American and Caucasian | Total – 814 Latino – 154 African American – 225 Caucasian – 435 | Cohort study | Comparing bystander CPR rates in an urban population by ethnicity. |
| a) Teodorescu et al., 201019 b) OR, US c) 01 Feb 2002 – 31 Jan 2007 | a) OOHCA cases EMS registry data and Medical Examiner’s report | a) Not specified58–61 b) Asian, Hispanic, black and white | Total – 1187 Asian – 33 Hispanic – 24 Black – 84 White – 975 Other – 15 Missing – 56 | Cohort study | Evaluation of correlates of PEA compared to VF/VT. |
| a) Merchant et al., 201154† b) US c) Jan 2000 – Sep 2007 | a) OOHCA > 66 yrs b) Medicare data | a) Self- reported (at time of Medicare enrolment) b) Black and white | Total – 68, 115 Black – 7942 White – 60, 173 | Cohort study | Determining whether the racial composition of US hospitals accounts for racial differences in survival rates for patients hospitalised for OOHCA |
| a) McNally et al., 201114§ b) Multiple states, US c) 01 Oct 2005 – 31 Dec 2010 | a) OOHCA cases b) CARES registry62 | a) Not specified b) Black/African American, Hispanic/Latino and white | Total – 20,785 Hispanic/Latino - 1494 Black/African American – 7588 White – 10,989 Other - 714 | Cohort Study | Reporting summary data from an OOHCA surveillance registry in the US |
| a) Wilde et al., 201147 b) IL, US c) 01 Jan 1996 – 31 Dec 2004 | a) OOHCA Cases b) Prehospital database | a) Race coded by paramedic at scene of arrest b) Black and white | Total – 3869 Black – 353 White – 3516 | Case series | Determining pre-hospital differences between blacks and whites experiencing OOHCA and to ascertain which factors are responsible for any such differences. |
| Studies outside the US | |||||
| a) Lim et al., 200249 b) Singapore c) Nov 2001 – Jan 2002 | a) OOHCA cases b) ED admission data and radio log sheets | a) Not specified b) Indian, Malay and Chinese | Total – 93 Indian – 9 Malay – 11 Chinese – 69 Other –4 | Cohort | Evaluating characteristics and outcome of OOHCA patients presenting to the ED, and examining factors that influence the decision to prolong or abort resuscitation. |
| a) Hamaad et al., 200645 b) Birmingham, UK c) 01 Jan 2002 – 31 Dec 2002 | a) Sudden, unexpected deaths b) ED admission data | a) Direct patient records and enquiries b) Indo–Asian, Afro-Caribbean and Caucasian | Total – 126 Afro Caribbean – 10 Caucasian – 90 Indo Asian – 26 | Cohort study | To report ethnic differences in sudden cardiac death prevalence. |
| a) Shah et al., 201051 b) London, UK c) 01 Apr 2003 – 31 Mar 2007 | a) OOHCA cases b) EMS data registry | a) Self- reported b) South Asian and white | Total – 3161 South Asian – 183 White – 1995 Other – 983 | Cohort study | Comparing OOHCA characteristics in South Asian and white populations. |
| a) Author, Year b) Location c) Period of data collection | a) Basis of sample b) Data source | a) Method used to define race/ethnicity b) Race/ethnicity categories used by authors | Total number and number in each racial/ethnic group | Study design | Primary objective of study |
|---|---|---|---|---|---|
| Studies Within US | |||||
| a) Wilcox-Gok44 b) NJ, US c) 1985 | a) OOHCA cases b) Medical Intensive Care Unit forms | a) Not specified b) Black and white | Total – 1879 Black – 190 White – 1659 Other – 30 | Cohort study | To determine the role of CPR and advanced life support in improving OOHCA survival. |
| a) Becker et al., 199311 b) Chicago, IL, US c) 01 Jan 1987 – 31 Dec 1988 | a) OOHCA cases b) Data from EMS and hospital records | a) Not specified55 b) Black and white | Total – 6451 Black – 2910 White – 3207 Other – 334 | Cohort study | Comparing racial differences in the incidence and characteristics of OOHCA. |
| a) Cowie et al., 199341 b) Seattle, WA, US c) 30 May 1984 – 31 Jul 1986 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 977 Black – 117 White – 860 | Cohort study | Comparing OOHCA characteristics in black and white populations. |
| a) Brookoff et al., 199415 b) Memphis, TN, US c) 01 Mar 1989 – 05 Jun 1992 | a) OOHCA cases b) Data from prospective trial of first responder defibrillation56 | a) Not specified b) Black and white | Total – 1068 Black– 603 White – 465 | Cohort study | Comparing bystander CPR in OOHCA patients and ethnicity. |
| a) Sayegh et al., 199916 b) MI, US c) 1991 – 1996 | a) OOHCA cases b) EMS data registry | a) Vital statistics and hospital records b) Black and white | Total – 1317 Black – 378§§ White – 939 | Cohort study | Assessing whether race or socioeconomic status affects survival at hospital discharge, following OOHCA. |
| a) Chu et al., 199812 b) MI, US c) 01 Jan 1991 – 31 Dec 1994 | a) OOHCA cases b) EMS data registry and hospital medical records | a) Not specified b) Black and white | Total – 1690 Black – 223 White – 1467 | Cohort study | Determining if race, when controlled for socioeconomic status, is an independent risk factor for survival at hospital discharge, following OOHCA. |
| a) Lindholm et al., 199850 b) Kansas City, KS, US c) 01 Jan 1992 – 31 Dec/1994 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 832 (numbers by ethnic group not stated) | Cohort study | Assessing the effect of the return of spontaneous circulation on survival at hospital discharge, against other predictors of survival, following OOHCA. |
| a) Sweeney et al., 199852 * b) NC, US c) 1992 – 1995 | a) OOHCA cases b) EMS and fire department records | a) Not specified b) Black and white | Total – 243 (numbers by ethnic group not stated) | Trial and Cohort study | Assessing whether use of automated external defibrillators in addition to CPR improves survival to hospital discharge. |
| a) Groeneveld et al., 200348 b) US c) 1990 – 1999 | a) Hospitalised VF/cardiac arrest cases b) Medicare sources | a) Self-reported b) Black and white | Total – 5948 Black – 519 White – 5429 | Cohort study | Assessing racial disparity in the utilization of life-saving procedures, post cardiac arrest, and whether procedure-rate differences affect long-term survival. |
| a) Polentini et al., 200546 b) Milwaukee, WI, US c) 1992 – 2002 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 9170 (numbers by ethnic group not stated) | Cohort study | Investigating changes in incidence of each cardiac arrest rhythm between 1992 and 2002, and rates of survival at hospital admission and discharge for VF/VT arrests. |
| a) Galea et al., 200717 b) New York, NY, US c) 01 Apr 2002 – 31 Mar 2003 | a) OOHCA cases b) Data collection from on-scene first responders prospectively | a) Not specified b) Hispanic, black and white | Total – 4053 Hispanic – 636 Black – 1257 White – 1908 Other – 252 | Cohort study | Comparing incidence of OOHCA and 30-day patient survival following hospital discharge in three racial/ethnic groups. |
| a) Fairbanks et al., 200720 b) Rochester, NY, US c) 01 Jan 1998 – 21 Dec 2001 | a) OOHCA cases b) EMS registry data | a) Not specified b) Black and white | Total – 539 Black – 190 White – 274 Other – 28 Unknown race – 47 | Cohort study | To identify OOHCA factors that affect patient survival to 1 yr. |
| a) Liu et al., 20084† b) Milwaukee, WI, US c) 1995 – 2005 | a) OOHCA cases b) EMS data registry | a) Not specified b) African American and white | Total – 1702 African American – 405 White – 1188 Other – 109 | Cohort study | To determine variability in rates of survival to discharge in patients that were successfully resuscitated following OOHCA. |
| a) Vadeboncoeur et al., 200853 b) AZ, US c) Nov 2004 – Nov 2006 | a) OOHCA cases b) EMS data registry | a) Patient care report and hospital data b) Hispanic and non–Hispanic | Total – 1379 Hispanic – 273 Non–Hispanic – 1106 | Cohort study | Comparing bystander CPR in Hispanic versus non-Hispanic groups. |
| a) Benson et al., 200918 b) Los Angeles,CA, US c) 2000 – 2001 | a) OOHCA cases b) Subset analysis from CARE LA study57 | a) If classified as Latino origin on the 2000 or 2001 CDSMF (California Death Statistical Master File) and if classified as African American, then coded as such regardless of Latino origin. b) Latino, African American and Caucasian | Total – 814 Latino – 154 African American – 225 Caucasian – 435 | Cohort study | Comparing bystander CPR rates in an urban population by ethnicity. |
| a) Teodorescu et al., 201019 b) OR, US c) 01 Feb 2002 – 31 Jan 2007 | a) OOHCA cases EMS registry data and Medical Examiner’s report | a) Not specified58–61 b) Asian, Hispanic, black and white | Total – 1187 Asian – 33 Hispanic – 24 Black – 84 White – 975 Other – 15 Missing – 56 | Cohort study | Evaluation of correlates of PEA compared to VF/VT. |
| a) Merchant et al., 201154† b) US c) Jan 2000 – Sep 2007 | a) OOHCA > 66 yrs b) Medicare data | a) Self- reported (at time of Medicare enrolment) b) Black and white | Total – 68, 115 Black – 7942 White – 60, 173 | Cohort study | Determining whether the racial composition of US hospitals accounts for racial differences in survival rates for patients hospitalised for OOHCA |
| a) McNally et al., 201114§ b) Multiple states, US c) 01 Oct 2005 – 31 Dec 2010 | a) OOHCA cases b) CARES registry62 | a) Not specified b) Black/African American, Hispanic/Latino and white | Total – 20,785 Hispanic/Latino - 1494 Black/African American – 7588 White – 10,989 Other - 714 | Cohort Study | Reporting summary data from an OOHCA surveillance registry in the US |
| a) Wilde et al., 201147 b) IL, US c) 01 Jan 1996 – 31 Dec 2004 | a) OOHCA Cases b) Prehospital database | a) Race coded by paramedic at scene of arrest b) Black and white | Total – 3869 Black – 353 White – 3516 | Case series | Determining pre-hospital differences between blacks and whites experiencing OOHCA and to ascertain which factors are responsible for any such differences. |
| Studies outside the US | |||||
| a) Lim et al., 200249 b) Singapore c) Nov 2001 – Jan 2002 | a) OOHCA cases b) ED admission data and radio log sheets | a) Not specified b) Indian, Malay and Chinese | Total – 93 Indian – 9 Malay – 11 Chinese – 69 Other –4 | Cohort | Evaluating characteristics and outcome of OOHCA patients presenting to the ED, and examining factors that influence the decision to prolong or abort resuscitation. |
| a) Hamaad et al., 200645 b) Birmingham, UK c) 01 Jan 2002 – 31 Dec 2002 | a) Sudden, unexpected deaths b) ED admission data | a) Direct patient records and enquiries b) Indo–Asian, Afro-Caribbean and Caucasian | Total – 126 Afro Caribbean – 10 Caucasian – 90 Indo Asian – 26 | Cohort study | To report ethnic differences in sudden cardiac death prevalence. |
| a) Shah et al., 201051 b) London, UK c) 01 Apr 2003 – 31 Mar 2007 | a) OOHCA cases b) EMS data registry | a) Self- reported b) South Asian and white | Total – 3161 South Asian – 183 White – 1995 Other – 983 | Cohort study | Comparing OOHCA characteristics in South Asian and white populations. |
Cohort comprised of patients having bystander-witnessed OOHCA, †Cohort comprised of patients following hospitalization post OOHCA, §§Cohort numbers derived from percentages quoted in the study, §Study also included patients <18 yrs old
Abbreviations: CARES: Cardiac Arrest Registry to Enhance Survival, CPR: Cardiopulmonary resuscitation, ED: Emergency department, EMS: Emergency medical services, OOHCA: Out-of-hospital cardiac arrest; PEA: Pulseless electrical activity; VF/VT: Ventricular fibrillation/ventricular tachycardia.
Summary of contextual details of studies: publication, location, timing, study design, definition of race or ethnicity concepts/terminology, sample size, data source and primary objective
| a) Author, Year b) Location c) Period of data collection | a) Basis of sample b) Data source | a) Method used to define race/ethnicity b) Race/ethnicity categories used by authors | Total number and number in each racial/ethnic group | Study design | Primary objective of study |
|---|---|---|---|---|---|
| Studies Within US | |||||
| a) Wilcox-Gok44 b) NJ, US c) 1985 | a) OOHCA cases b) Medical Intensive Care Unit forms | a) Not specified b) Black and white | Total – 1879 Black – 190 White – 1659 Other – 30 | Cohort study | To determine the role of CPR and advanced life support in improving OOHCA survival. |
| a) Becker et al., 199311 b) Chicago, IL, US c) 01 Jan 1987 – 31 Dec 1988 | a) OOHCA cases b) Data from EMS and hospital records | a) Not specified55 b) Black and white | Total – 6451 Black – 2910 White – 3207 Other – 334 | Cohort study | Comparing racial differences in the incidence and characteristics of OOHCA. |
| a) Cowie et al., 199341 b) Seattle, WA, US c) 30 May 1984 – 31 Jul 1986 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 977 Black – 117 White – 860 | Cohort study | Comparing OOHCA characteristics in black and white populations. |
| a) Brookoff et al., 199415 b) Memphis, TN, US c) 01 Mar 1989 – 05 Jun 1992 | a) OOHCA cases b) Data from prospective trial of first responder defibrillation56 | a) Not specified b) Black and white | Total – 1068 Black– 603 White – 465 | Cohort study | Comparing bystander CPR in OOHCA patients and ethnicity. |
| a) Sayegh et al., 199916 b) MI, US c) 1991 – 1996 | a) OOHCA cases b) EMS data registry | a) Vital statistics and hospital records b) Black and white | Total – 1317 Black – 378§§ White – 939 | Cohort study | Assessing whether race or socioeconomic status affects survival at hospital discharge, following OOHCA. |
| a) Chu et al., 199812 b) MI, US c) 01 Jan 1991 – 31 Dec 1994 | a) OOHCA cases b) EMS data registry and hospital medical records | a) Not specified b) Black and white | Total – 1690 Black – 223 White – 1467 | Cohort study | Determining if race, when controlled for socioeconomic status, is an independent risk factor for survival at hospital discharge, following OOHCA. |
| a) Lindholm et al., 199850 b) Kansas City, KS, US c) 01 Jan 1992 – 31 Dec/1994 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 832 (numbers by ethnic group not stated) | Cohort study | Assessing the effect of the return of spontaneous circulation on survival at hospital discharge, against other predictors of survival, following OOHCA. |
| a) Sweeney et al., 199852 * b) NC, US c) 1992 – 1995 | a) OOHCA cases b) EMS and fire department records | a) Not specified b) Black and white | Total – 243 (numbers by ethnic group not stated) | Trial and Cohort study | Assessing whether use of automated external defibrillators in addition to CPR improves survival to hospital discharge. |
| a) Groeneveld et al., 200348 b) US c) 1990 – 1999 | a) Hospitalised VF/cardiac arrest cases b) Medicare sources | a) Self-reported b) Black and white | Total – 5948 Black – 519 White – 5429 | Cohort study | Assessing racial disparity in the utilization of life-saving procedures, post cardiac arrest, and whether procedure-rate differences affect long-term survival. |
| a) Polentini et al., 200546 b) Milwaukee, WI, US c) 1992 – 2002 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 9170 (numbers by ethnic group not stated) | Cohort study | Investigating changes in incidence of each cardiac arrest rhythm between 1992 and 2002, and rates of survival at hospital admission and discharge for VF/VT arrests. |
| a) Galea et al., 200717 b) New York, NY, US c) 01 Apr 2002 – 31 Mar 2003 | a) OOHCA cases b) Data collection from on-scene first responders prospectively | a) Not specified b) Hispanic, black and white | Total – 4053 Hispanic – 636 Black – 1257 White – 1908 Other – 252 | Cohort study | Comparing incidence of OOHCA and 30-day patient survival following hospital discharge in three racial/ethnic groups. |
| a) Fairbanks et al., 200720 b) Rochester, NY, US c) 01 Jan 1998 – 21 Dec 2001 | a) OOHCA cases b) EMS registry data | a) Not specified b) Black and white | Total – 539 Black – 190 White – 274 Other – 28 Unknown race – 47 | Cohort study | To identify OOHCA factors that affect patient survival to 1 yr. |
| a) Liu et al., 20084† b) Milwaukee, WI, US c) 1995 – 2005 | a) OOHCA cases b) EMS data registry | a) Not specified b) African American and white | Total – 1702 African American – 405 White – 1188 Other – 109 | Cohort study | To determine variability in rates of survival to discharge in patients that were successfully resuscitated following OOHCA. |
| a) Vadeboncoeur et al., 200853 b) AZ, US c) Nov 2004 – Nov 2006 | a) OOHCA cases b) EMS data registry | a) Patient care report and hospital data b) Hispanic and non–Hispanic | Total – 1379 Hispanic – 273 Non–Hispanic – 1106 | Cohort study | Comparing bystander CPR in Hispanic versus non-Hispanic groups. |
| a) Benson et al., 200918 b) Los Angeles,CA, US c) 2000 – 2001 | a) OOHCA cases b) Subset analysis from CARE LA study57 | a) If classified as Latino origin on the 2000 or 2001 CDSMF (California Death Statistical Master File) and if classified as African American, then coded as such regardless of Latino origin. b) Latino, African American and Caucasian | Total – 814 Latino – 154 African American – 225 Caucasian – 435 | Cohort study | Comparing bystander CPR rates in an urban population by ethnicity. |
| a) Teodorescu et al., 201019 b) OR, US c) 01 Feb 2002 – 31 Jan 2007 | a) OOHCA cases EMS registry data and Medical Examiner’s report | a) Not specified58–61 b) Asian, Hispanic, black and white | Total – 1187 Asian – 33 Hispanic – 24 Black – 84 White – 975 Other – 15 Missing – 56 | Cohort study | Evaluation of correlates of PEA compared to VF/VT. |
| a) Merchant et al., 201154† b) US c) Jan 2000 – Sep 2007 | a) OOHCA > 66 yrs b) Medicare data | a) Self- reported (at time of Medicare enrolment) b) Black and white | Total – 68, 115 Black – 7942 White – 60, 173 | Cohort study | Determining whether the racial composition of US hospitals accounts for racial differences in survival rates for patients hospitalised for OOHCA |
| a) McNally et al., 201114§ b) Multiple states, US c) 01 Oct 2005 – 31 Dec 2010 | a) OOHCA cases b) CARES registry62 | a) Not specified b) Black/African American, Hispanic/Latino and white | Total – 20,785 Hispanic/Latino - 1494 Black/African American – 7588 White – 10,989 Other - 714 | Cohort Study | Reporting summary data from an OOHCA surveillance registry in the US |
| a) Wilde et al., 201147 b) IL, US c) 01 Jan 1996 – 31 Dec 2004 | a) OOHCA Cases b) Prehospital database | a) Race coded by paramedic at scene of arrest b) Black and white | Total – 3869 Black – 353 White – 3516 | Case series | Determining pre-hospital differences between blacks and whites experiencing OOHCA and to ascertain which factors are responsible for any such differences. |
| Studies outside the US | |||||
| a) Lim et al., 200249 b) Singapore c) Nov 2001 – Jan 2002 | a) OOHCA cases b) ED admission data and radio log sheets | a) Not specified b) Indian, Malay and Chinese | Total – 93 Indian – 9 Malay – 11 Chinese – 69 Other –4 | Cohort | Evaluating characteristics and outcome of OOHCA patients presenting to the ED, and examining factors that influence the decision to prolong or abort resuscitation. |
| a) Hamaad et al., 200645 b) Birmingham, UK c) 01 Jan 2002 – 31 Dec 2002 | a) Sudden, unexpected deaths b) ED admission data | a) Direct patient records and enquiries b) Indo–Asian, Afro-Caribbean and Caucasian | Total – 126 Afro Caribbean – 10 Caucasian – 90 Indo Asian – 26 | Cohort study | To report ethnic differences in sudden cardiac death prevalence. |
| a) Shah et al., 201051 b) London, UK c) 01 Apr 2003 – 31 Mar 2007 | a) OOHCA cases b) EMS data registry | a) Self- reported b) South Asian and white | Total – 3161 South Asian – 183 White – 1995 Other – 983 | Cohort study | Comparing OOHCA characteristics in South Asian and white populations. |
| a) Author, Year b) Location c) Period of data collection | a) Basis of sample b) Data source | a) Method used to define race/ethnicity b) Race/ethnicity categories used by authors | Total number and number in each racial/ethnic group | Study design | Primary objective of study |
|---|---|---|---|---|---|
| Studies Within US | |||||
| a) Wilcox-Gok44 b) NJ, US c) 1985 | a) OOHCA cases b) Medical Intensive Care Unit forms | a) Not specified b) Black and white | Total – 1879 Black – 190 White – 1659 Other – 30 | Cohort study | To determine the role of CPR and advanced life support in improving OOHCA survival. |
| a) Becker et al., 199311 b) Chicago, IL, US c) 01 Jan 1987 – 31 Dec 1988 | a) OOHCA cases b) Data from EMS and hospital records | a) Not specified55 b) Black and white | Total – 6451 Black – 2910 White – 3207 Other – 334 | Cohort study | Comparing racial differences in the incidence and characteristics of OOHCA. |
| a) Cowie et al., 199341 b) Seattle, WA, US c) 30 May 1984 – 31 Jul 1986 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 977 Black – 117 White – 860 | Cohort study | Comparing OOHCA characteristics in black and white populations. |
| a) Brookoff et al., 199415 b) Memphis, TN, US c) 01 Mar 1989 – 05 Jun 1992 | a) OOHCA cases b) Data from prospective trial of first responder defibrillation56 | a) Not specified b) Black and white | Total – 1068 Black– 603 White – 465 | Cohort study | Comparing bystander CPR in OOHCA patients and ethnicity. |
| a) Sayegh et al., 199916 b) MI, US c) 1991 – 1996 | a) OOHCA cases b) EMS data registry | a) Vital statistics and hospital records b) Black and white | Total – 1317 Black – 378§§ White – 939 | Cohort study | Assessing whether race or socioeconomic status affects survival at hospital discharge, following OOHCA. |
| a) Chu et al., 199812 b) MI, US c) 01 Jan 1991 – 31 Dec 1994 | a) OOHCA cases b) EMS data registry and hospital medical records | a) Not specified b) Black and white | Total – 1690 Black – 223 White – 1467 | Cohort study | Determining if race, when controlled for socioeconomic status, is an independent risk factor for survival at hospital discharge, following OOHCA. |
| a) Lindholm et al., 199850 b) Kansas City, KS, US c) 01 Jan 1992 – 31 Dec/1994 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 832 (numbers by ethnic group not stated) | Cohort study | Assessing the effect of the return of spontaneous circulation on survival at hospital discharge, against other predictors of survival, following OOHCA. |
| a) Sweeney et al., 199852 * b) NC, US c) 1992 – 1995 | a) OOHCA cases b) EMS and fire department records | a) Not specified b) Black and white | Total – 243 (numbers by ethnic group not stated) | Trial and Cohort study | Assessing whether use of automated external defibrillators in addition to CPR improves survival to hospital discharge. |
| a) Groeneveld et al., 200348 b) US c) 1990 – 1999 | a) Hospitalised VF/cardiac arrest cases b) Medicare sources | a) Self-reported b) Black and white | Total – 5948 Black – 519 White – 5429 | Cohort study | Assessing racial disparity in the utilization of life-saving procedures, post cardiac arrest, and whether procedure-rate differences affect long-term survival. |
| a) Polentini et al., 200546 b) Milwaukee, WI, US c) 1992 – 2002 | a) OOHCA cases b) EMS data registry | a) Not specified b) Black and white | Total – 9170 (numbers by ethnic group not stated) | Cohort study | Investigating changes in incidence of each cardiac arrest rhythm between 1992 and 2002, and rates of survival at hospital admission and discharge for VF/VT arrests. |
| a) Galea et al., 200717 b) New York, NY, US c) 01 Apr 2002 – 31 Mar 2003 | a) OOHCA cases b) Data collection from on-scene first responders prospectively | a) Not specified b) Hispanic, black and white | Total – 4053 Hispanic – 636 Black – 1257 White – 1908 Other – 252 | Cohort study | Comparing incidence of OOHCA and 30-day patient survival following hospital discharge in three racial/ethnic groups. |
| a) Fairbanks et al., 200720 b) Rochester, NY, US c) 01 Jan 1998 – 21 Dec 2001 | a) OOHCA cases b) EMS registry data | a) Not specified b) Black and white | Total – 539 Black – 190 White – 274 Other – 28 Unknown race – 47 | Cohort study | To identify OOHCA factors that affect patient survival to 1 yr. |
| a) Liu et al., 20084† b) Milwaukee, WI, US c) 1995 – 2005 | a) OOHCA cases b) EMS data registry | a) Not specified b) African American and white | Total – 1702 African American – 405 White – 1188 Other – 109 | Cohort study | To determine variability in rates of survival to discharge in patients that were successfully resuscitated following OOHCA. |
| a) Vadeboncoeur et al., 200853 b) AZ, US c) Nov 2004 – Nov 2006 | a) OOHCA cases b) EMS data registry | a) Patient care report and hospital data b) Hispanic and non–Hispanic | Total – 1379 Hispanic – 273 Non–Hispanic – 1106 | Cohort study | Comparing bystander CPR in Hispanic versus non-Hispanic groups. |
| a) Benson et al., 200918 b) Los Angeles,CA, US c) 2000 – 2001 | a) OOHCA cases b) Subset analysis from CARE LA study57 | a) If classified as Latino origin on the 2000 or 2001 CDSMF (California Death Statistical Master File) and if classified as African American, then coded as such regardless of Latino origin. b) Latino, African American and Caucasian | Total – 814 Latino – 154 African American – 225 Caucasian – 435 | Cohort study | Comparing bystander CPR rates in an urban population by ethnicity. |
| a) Teodorescu et al., 201019 b) OR, US c) 01 Feb 2002 – 31 Jan 2007 | a) OOHCA cases EMS registry data and Medical Examiner’s report | a) Not specified58–61 b) Asian, Hispanic, black and white | Total – 1187 Asian – 33 Hispanic – 24 Black – 84 White – 975 Other – 15 Missing – 56 | Cohort study | Evaluation of correlates of PEA compared to VF/VT. |
| a) Merchant et al., 201154† b) US c) Jan 2000 – Sep 2007 | a) OOHCA > 66 yrs b) Medicare data | a) Self- reported (at time of Medicare enrolment) b) Black and white | Total – 68, 115 Black – 7942 White – 60, 173 | Cohort study | Determining whether the racial composition of US hospitals accounts for racial differences in survival rates for patients hospitalised for OOHCA |
| a) McNally et al., 201114§ b) Multiple states, US c) 01 Oct 2005 – 31 Dec 2010 | a) OOHCA cases b) CARES registry62 | a) Not specified b) Black/African American, Hispanic/Latino and white | Total – 20,785 Hispanic/Latino - 1494 Black/African American – 7588 White – 10,989 Other - 714 | Cohort Study | Reporting summary data from an OOHCA surveillance registry in the US |
| a) Wilde et al., 201147 b) IL, US c) 01 Jan 1996 – 31 Dec 2004 | a) OOHCA Cases b) Prehospital database | a) Race coded by paramedic at scene of arrest b) Black and white | Total – 3869 Black – 353 White – 3516 | Case series | Determining pre-hospital differences between blacks and whites experiencing OOHCA and to ascertain which factors are responsible for any such differences. |
| Studies outside the US | |||||
| a) Lim et al., 200249 b) Singapore c) Nov 2001 – Jan 2002 | a) OOHCA cases b) ED admission data and radio log sheets | a) Not specified b) Indian, Malay and Chinese | Total – 93 Indian – 9 Malay – 11 Chinese – 69 Other –4 | Cohort | Evaluating characteristics and outcome of OOHCA patients presenting to the ED, and examining factors that influence the decision to prolong or abort resuscitation. |
| a) Hamaad et al., 200645 b) Birmingham, UK c) 01 Jan 2002 – 31 Dec 2002 | a) Sudden, unexpected deaths b) ED admission data | a) Direct patient records and enquiries b) Indo–Asian, Afro-Caribbean and Caucasian | Total – 126 Afro Caribbean – 10 Caucasian – 90 Indo Asian – 26 | Cohort study | To report ethnic differences in sudden cardiac death prevalence. |
| a) Shah et al., 201051 b) London, UK c) 01 Apr 2003 – 31 Mar 2007 | a) OOHCA cases b) EMS data registry | a) Self- reported b) South Asian and white | Total – 3161 South Asian – 183 White – 1995 Other – 983 | Cohort study | Comparing OOHCA characteristics in South Asian and white populations. |
Cohort comprised of patients having bystander-witnessed OOHCA, †Cohort comprised of patients following hospitalization post OOHCA, §§Cohort numbers derived from percentages quoted in the study, §Study also included patients <18 yrs old
Abbreviations: CARES: Cardiac Arrest Registry to Enhance Survival, CPR: Cardiopulmonary resuscitation, ED: Emergency department, EMS: Emergency medical services, OOHCA: Out-of-hospital cardiac arrest; PEA: Pulseless electrical activity; VF/VT: Ventricular fibrillation/ventricular tachycardia.
There were 19 studies based in the United States (US) described in Table 1 and 19 articles covering black and white population outcomes. One study compared Afro-Caribbean, Indo-Asian and Caucasian racial/ethnic groups.45 Five studies14,17–19,53 compared Hispanic groups to either black or white racial/ethnic groups. One study compared Chinese, Malay and Indian racial or ethnic groups.49 One study compared South Asian and white populations.51 Most of the studies were cohort studies and the most common data source was Emergency Medical Service (EMS) registries coupled with Emergency Department (ED) admission data. Studies with patient follow-up data have been defined as cohort studies.63
Table 2 summarises the studies that reported on OOHCA by race or ethnicity and our findings are summarised in the text below.
Ethnic variation for out-of-hospital cardiac arrest (OOHCA): age at event, witnessed cardiac arrest, initial arrest rhythm, response times and bystander CPR
| Author | Age, Years(mean), p value or OR | Witnessed Arrest, p value or OR | Initial arrest rhythm of VF/VT p value or OR | Response Time p value or OR | Bystander CPR, p value or OR |
|---|---|---|---|---|---|
| Studies Within the US | |||||
| Becker et al., 199311 | NR/NM | Black – 42% White – 49% p < 0.001 | Black – 17% White – 26% p < 0.001 | Black – 6 min White – 6 min 18-sec difference between black and white patients (p < 0.001). | Black –18% White – 25% p < 0.001 |
| Cowie et al.,199341 | Black – 61.9 White – 68.2 p < 0.01 | Black – 53% White – 61% p < 0.12 | Black – VF 40%, VT 0% White – VF 49%, VT 2% p < 0.09 | Black – 3.4 min White – 3.4 min p > 0.71 | Black – 18% White – 32% p < 0.003 |
| Brookoff et al., 199415 | Black – 61.9 White – 67.4 p = NR | Black – 60% White – 64% p = ns | Black – 44% White – 54% (referent) OR 0.82; 95%CI 0.73–0.93 | No statistically significant difference between blacks and whites (mean times not stated). | Black – 10% White – 21% (referent) OR 0.46 95% CI 0.34–0.61; p < 0.001. |
| Chu et al., 199812 | Black – 62 White – 68 p < 0.001 | Black – 57% White – 61% p = 0.422 | Black – 37% White – 51% p < 0.01 | Black – 5.3 min White – 7.0 min p < 0.001 | Black – 11% White – 20% p = 0.003 |
| Sayegh et al., 199916 | NR/NM | Black – 53% White – 56% p = 0.41 | Black – 43% White – 46% p = 0.286 | ALS ambulance response time <9 min: Black – 88% White – 83% p = 0.039 | Black – 7% White – 10% p = 0.215 |
| Groeneveld et al., 200348 | Black – 76 White – 76 p = ns | NR/NM | NR/NM | NR/NM | NR/NM |
| Polentini et al., 200546 | 0.56 (0.48 – 0.66), whites referent Proportions not reported | ||||
| Fairbanks et al., 200720 | NR/NM | NR/NM | Black – 35% White – 29% p = 0.038 | Black – 5 min White – 5 min p = NR | NR/NM |
| Galea et al., 200717 | NR/NM | Hispanic – 43% Black – 36% White – 42% p < 0.01 | Hispanic – 15% Black – 12% White – 17% p = 0.06 | Hispanic – 5 min Black – 4.7 min White – 4.5 min p < 0.01 | Hispanic – 25% Black – 31% White – 31% p < 0.01 |
| Vadeboncoeur et al., 200853 | Hispanic – 53.2 Non–Hispanic – 64.5 p = 0.005 | Hispanic – 44% Non–Hispanic – 50% p = 0.074 | VF Hispanic – 21% Non-Hispanic – 27% p = 0.036 | Hispanic – 5.1 min Non-Hispanic – 5.5 min p = 0.006 | Bystander CPR Hispanic – 32% Non-Hispanic – 42% p < 0.0001 Lay Bystander CPR Hispanic – 16% Non-Hispanic – 26% p = 0.001 |
| Benson et al., 200918 | Latino – 59.6 African American – 64.7 Caucasian – 69.9 p < 0.05 | Latino – 40% African American – 36% Caucasian – 48% p < 0.05 – (Caucasian vs. African American) p = ns (Caucasian vs. Latino) | Latino – 35% African American – 24% Caucasian – 31% p = ns | NR/NM | Univariate %, OR (95% CI) Latino –13%, 0.48 (0.28–0.80) African American – 13%, 0.47 (0.3–0.74) Caucasian – 24% (referent) Multivariate OR (95% CI) Latino – 0.46, 0.23–0.92 African American – 0.67, 0.38 – 1.16 |
| Teodorescu et al., 201019 | NR/NM | NR/NM | Multivariate Analysis PEA vs. VF/VT, OR (95%CI): Other Race – 1.02, 0.23 – 4.59 Asian – 0.88, 0.2 – 3.98 Hispanic – 0.32, 0.05 – 2.13 Black – 2.64, 1.29 – 5.38 White – 1.0 (referent) PEA vs. Asystole – OR (95% CI): Other Race – 1.10, 0.17 – 7.05 Asian – 0.26, 0.07–1.03 Hispanic – 1.11, 0.15–8.38 Black – 1.3, 0.65–2.63 White – 1.0 (referent) | NR/NM | NR/NM |
| Wilde et al., 201147 | Black – 61.1 White – 68.5 p < 0.001 | NR/NM | NR/NM | Black – 5.3 mins White – 7.3 mins p < 0.001 | Black – 72.5%† White – 80.4% p < 0.001 |
| McNally et al., 201114 | NR/NM | NR/NM | NR/NM | NR/NM | Black – 32.8% White – 33.7% Hispanic – 40.2% p < 0.001 |
| Merchant et al, 201154 | Blacks – 77.4* Whites – 78.1* p=0.01 | NR/NM | VF Blacks – 8%* Whites – 12%* p<0.001 | NR/NM | NR/NM |
| Studies Outside the US | |||||
| Hamaad et al., 200645 | Indo-Asian – 66.0 Afro–Caribbean – 63.7 Caucasian – 68.7 p = ns | NR/NM | VF Indo -Asian – 30% Afro-Caribbean – 30% Caucasian – 45% p = NR | Call to collection time (mins): Caucasian – 28.2 Afro-Caribbean – 31.7 Indo-Asian – 22.6 | NR/NM |
| Shah et al., 201051 | South Asian – 64.6 White – 69.5 p < 0.005 | South Asian −70% White – 62% (referent) OR = 1.1, 95%CI 1.0 – 1.2 | South Asian – 30% White –30% (referent) OR = 1.0, 95% CI 0.7–1.3 | South Asian – 7.46 min White – 7.48 min p = 0.335 | South Asian – 30% (referent) White – 34% OR = 1.2, 95%CI, 0.9–1.7 |
| Author | Age, Years(mean), p value or OR | Witnessed Arrest, p value or OR | Initial arrest rhythm of VF/VT p value or OR | Response Time p value or OR | Bystander CPR, p value or OR |
|---|---|---|---|---|---|
| Studies Within the US | |||||
| Becker et al., 199311 | NR/NM | Black – 42% White – 49% p < 0.001 | Black – 17% White – 26% p < 0.001 | Black – 6 min White – 6 min 18-sec difference between black and white patients (p < 0.001). | Black –18% White – 25% p < 0.001 |
| Cowie et al.,199341 | Black – 61.9 White – 68.2 p < 0.01 | Black – 53% White – 61% p < 0.12 | Black – VF 40%, VT 0% White – VF 49%, VT 2% p < 0.09 | Black – 3.4 min White – 3.4 min p > 0.71 | Black – 18% White – 32% p < 0.003 |
| Brookoff et al., 199415 | Black – 61.9 White – 67.4 p = NR | Black – 60% White – 64% p = ns | Black – 44% White – 54% (referent) OR 0.82; 95%CI 0.73–0.93 | No statistically significant difference between blacks and whites (mean times not stated). | Black – 10% White – 21% (referent) OR 0.46 95% CI 0.34–0.61; p < 0.001. |
| Chu et al., 199812 | Black – 62 White – 68 p < 0.001 | Black – 57% White – 61% p = 0.422 | Black – 37% White – 51% p < 0.01 | Black – 5.3 min White – 7.0 min p < 0.001 | Black – 11% White – 20% p = 0.003 |
| Sayegh et al., 199916 | NR/NM | Black – 53% White – 56% p = 0.41 | Black – 43% White – 46% p = 0.286 | ALS ambulance response time <9 min: Black – 88% White – 83% p = 0.039 | Black – 7% White – 10% p = 0.215 |
| Groeneveld et al., 200348 | Black – 76 White – 76 p = ns | NR/NM | NR/NM | NR/NM | NR/NM |
| Polentini et al., 200546 | 0.56 (0.48 – 0.66), whites referent Proportions not reported | ||||
| Fairbanks et al., 200720 | NR/NM | NR/NM | Black – 35% White – 29% p = 0.038 | Black – 5 min White – 5 min p = NR | NR/NM |
| Galea et al., 200717 | NR/NM | Hispanic – 43% Black – 36% White – 42% p < 0.01 | Hispanic – 15% Black – 12% White – 17% p = 0.06 | Hispanic – 5 min Black – 4.7 min White – 4.5 min p < 0.01 | Hispanic – 25% Black – 31% White – 31% p < 0.01 |
| Vadeboncoeur et al., 200853 | Hispanic – 53.2 Non–Hispanic – 64.5 p = 0.005 | Hispanic – 44% Non–Hispanic – 50% p = 0.074 | VF Hispanic – 21% Non-Hispanic – 27% p = 0.036 | Hispanic – 5.1 min Non-Hispanic – 5.5 min p = 0.006 | Bystander CPR Hispanic – 32% Non-Hispanic – 42% p < 0.0001 Lay Bystander CPR Hispanic – 16% Non-Hispanic – 26% p = 0.001 |
| Benson et al., 200918 | Latino – 59.6 African American – 64.7 Caucasian – 69.9 p < 0.05 | Latino – 40% African American – 36% Caucasian – 48% p < 0.05 – (Caucasian vs. African American) p = ns (Caucasian vs. Latino) | Latino – 35% African American – 24% Caucasian – 31% p = ns | NR/NM | Univariate %, OR (95% CI) Latino –13%, 0.48 (0.28–0.80) African American – 13%, 0.47 (0.3–0.74) Caucasian – 24% (referent) Multivariate OR (95% CI) Latino – 0.46, 0.23–0.92 African American – 0.67, 0.38 – 1.16 |
| Teodorescu et al., 201019 | NR/NM | NR/NM | Multivariate Analysis PEA vs. VF/VT, OR (95%CI): Other Race – 1.02, 0.23 – 4.59 Asian – 0.88, 0.2 – 3.98 Hispanic – 0.32, 0.05 – 2.13 Black – 2.64, 1.29 – 5.38 White – 1.0 (referent) PEA vs. Asystole – OR (95% CI): Other Race – 1.10, 0.17 – 7.05 Asian – 0.26, 0.07–1.03 Hispanic – 1.11, 0.15–8.38 Black – 1.3, 0.65–2.63 White – 1.0 (referent) | NR/NM | NR/NM |
| Wilde et al., 201147 | Black – 61.1 White – 68.5 p < 0.001 | NR/NM | NR/NM | Black – 5.3 mins White – 7.3 mins p < 0.001 | Black – 72.5%† White – 80.4% p < 0.001 |
| McNally et al., 201114 | NR/NM | NR/NM | NR/NM | NR/NM | Black – 32.8% White – 33.7% Hispanic – 40.2% p < 0.001 |
| Merchant et al, 201154 | Blacks – 77.4* Whites – 78.1* p=0.01 | NR/NM | VF Blacks – 8%* Whites – 12%* p<0.001 | NR/NM | NR/NM |
| Studies Outside the US | |||||
| Hamaad et al., 200645 | Indo-Asian – 66.0 Afro–Caribbean – 63.7 Caucasian – 68.7 p = ns | NR/NM | VF Indo -Asian – 30% Afro-Caribbean – 30% Caucasian – 45% p = NR | Call to collection time (mins): Caucasian – 28.2 Afro-Caribbean – 31.7 Indo-Asian – 22.6 | NR/NM |
| Shah et al., 201051 | South Asian – 64.6 White – 69.5 p < 0.005 | South Asian −70% White – 62% (referent) OR = 1.1, 95%CI 1.0 – 1.2 | South Asian – 30% White –30% (referent) OR = 1.0, 95% CI 0.7–1.3 | South Asian – 7.46 min White – 7.48 min p = 0.335 | South Asian – 30% (referent) White – 34% OR = 1.2, 95%CI, 0.9–1.7 |
Cohort limited to patients resuscitated from cardiac arrest and subsequently hospitalised
Refers to EMS-provided CPR and not Bystander CPR; Percentages are reported to the nearest whole number
Abbreviations: CI: confidence interval, CPR: cardiopulmonary resuscitation, NM: not measured, NR: not reported, ns: non-significant, OR: odds ratio, VF/VT – ventricular fibrillation/ventricular tachycardia ED: emergency department, EMS: emergency medical services, OOHCA: out-of-hospital cardiac arrest, PEA: pulseless electrical activity, VF/VT ventricular fibrillation/ventricular tachycardia.
Ethnic variation for out-of-hospital cardiac arrest (OOHCA): age at event, witnessed cardiac arrest, initial arrest rhythm, response times and bystander CPR
| Author | Age, Years(mean), p value or OR | Witnessed Arrest, p value or OR | Initial arrest rhythm of VF/VT p value or OR | Response Time p value or OR | Bystander CPR, p value or OR |
|---|---|---|---|---|---|
| Studies Within the US | |||||
| Becker et al., 199311 | NR/NM | Black – 42% White – 49% p < 0.001 | Black – 17% White – 26% p < 0.001 | Black – 6 min White – 6 min 18-sec difference between black and white patients (p < 0.001). | Black –18% White – 25% p < 0.001 |
| Cowie et al.,199341 | Black – 61.9 White – 68.2 p < 0.01 | Black – 53% White – 61% p < 0.12 | Black – VF 40%, VT 0% White – VF 49%, VT 2% p < 0.09 | Black – 3.4 min White – 3.4 min p > 0.71 | Black – 18% White – 32% p < 0.003 |
| Brookoff et al., 199415 | Black – 61.9 White – 67.4 p = NR | Black – 60% White – 64% p = ns | Black – 44% White – 54% (referent) OR 0.82; 95%CI 0.73–0.93 | No statistically significant difference between blacks and whites (mean times not stated). | Black – 10% White – 21% (referent) OR 0.46 95% CI 0.34–0.61; p < 0.001. |
| Chu et al., 199812 | Black – 62 White – 68 p < 0.001 | Black – 57% White – 61% p = 0.422 | Black – 37% White – 51% p < 0.01 | Black – 5.3 min White – 7.0 min p < 0.001 | Black – 11% White – 20% p = 0.003 |
| Sayegh et al., 199916 | NR/NM | Black – 53% White – 56% p = 0.41 | Black – 43% White – 46% p = 0.286 | ALS ambulance response time <9 min: Black – 88% White – 83% p = 0.039 | Black – 7% White – 10% p = 0.215 |
| Groeneveld et al., 200348 | Black – 76 White – 76 p = ns | NR/NM | NR/NM | NR/NM | NR/NM |
| Polentini et al., 200546 | 0.56 (0.48 – 0.66), whites referent Proportions not reported | ||||
| Fairbanks et al., 200720 | NR/NM | NR/NM | Black – 35% White – 29% p = 0.038 | Black – 5 min White – 5 min p = NR | NR/NM |
| Galea et al., 200717 | NR/NM | Hispanic – 43% Black – 36% White – 42% p < 0.01 | Hispanic – 15% Black – 12% White – 17% p = 0.06 | Hispanic – 5 min Black – 4.7 min White – 4.5 min p < 0.01 | Hispanic – 25% Black – 31% White – 31% p < 0.01 |
| Vadeboncoeur et al., 200853 | Hispanic – 53.2 Non–Hispanic – 64.5 p = 0.005 | Hispanic – 44% Non–Hispanic – 50% p = 0.074 | VF Hispanic – 21% Non-Hispanic – 27% p = 0.036 | Hispanic – 5.1 min Non-Hispanic – 5.5 min p = 0.006 | Bystander CPR Hispanic – 32% Non-Hispanic – 42% p < 0.0001 Lay Bystander CPR Hispanic – 16% Non-Hispanic – 26% p = 0.001 |
| Benson et al., 200918 | Latino – 59.6 African American – 64.7 Caucasian – 69.9 p < 0.05 | Latino – 40% African American – 36% Caucasian – 48% p < 0.05 – (Caucasian vs. African American) p = ns (Caucasian vs. Latino) | Latino – 35% African American – 24% Caucasian – 31% p = ns | NR/NM | Univariate %, OR (95% CI) Latino –13%, 0.48 (0.28–0.80) African American – 13%, 0.47 (0.3–0.74) Caucasian – 24% (referent) Multivariate OR (95% CI) Latino – 0.46, 0.23–0.92 African American – 0.67, 0.38 – 1.16 |
| Teodorescu et al., 201019 | NR/NM | NR/NM | Multivariate Analysis PEA vs. VF/VT, OR (95%CI): Other Race – 1.02, 0.23 – 4.59 Asian – 0.88, 0.2 – 3.98 Hispanic – 0.32, 0.05 – 2.13 Black – 2.64, 1.29 – 5.38 White – 1.0 (referent) PEA vs. Asystole – OR (95% CI): Other Race – 1.10, 0.17 – 7.05 Asian – 0.26, 0.07–1.03 Hispanic – 1.11, 0.15–8.38 Black – 1.3, 0.65–2.63 White – 1.0 (referent) | NR/NM | NR/NM |
| Wilde et al., 201147 | Black – 61.1 White – 68.5 p < 0.001 | NR/NM | NR/NM | Black – 5.3 mins White – 7.3 mins p < 0.001 | Black – 72.5%† White – 80.4% p < 0.001 |
| McNally et al., 201114 | NR/NM | NR/NM | NR/NM | NR/NM | Black – 32.8% White – 33.7% Hispanic – 40.2% p < 0.001 |
| Merchant et al, 201154 | Blacks – 77.4* Whites – 78.1* p=0.01 | NR/NM | VF Blacks – 8%* Whites – 12%* p<0.001 | NR/NM | NR/NM |
| Studies Outside the US | |||||
| Hamaad et al., 200645 | Indo-Asian – 66.0 Afro–Caribbean – 63.7 Caucasian – 68.7 p = ns | NR/NM | VF Indo -Asian – 30% Afro-Caribbean – 30% Caucasian – 45% p = NR | Call to collection time (mins): Caucasian – 28.2 Afro-Caribbean – 31.7 Indo-Asian – 22.6 | NR/NM |
| Shah et al., 201051 | South Asian – 64.6 White – 69.5 p < 0.005 | South Asian −70% White – 62% (referent) OR = 1.1, 95%CI 1.0 – 1.2 | South Asian – 30% White –30% (referent) OR = 1.0, 95% CI 0.7–1.3 | South Asian – 7.46 min White – 7.48 min p = 0.335 | South Asian – 30% (referent) White – 34% OR = 1.2, 95%CI, 0.9–1.7 |
| Author | Age, Years(mean), p value or OR | Witnessed Arrest, p value or OR | Initial arrest rhythm of VF/VT p value or OR | Response Time p value or OR | Bystander CPR, p value or OR |
|---|---|---|---|---|---|
| Studies Within the US | |||||
| Becker et al., 199311 | NR/NM | Black – 42% White – 49% p < 0.001 | Black – 17% White – 26% p < 0.001 | Black – 6 min White – 6 min 18-sec difference between black and white patients (p < 0.001). | Black –18% White – 25% p < 0.001 |
| Cowie et al.,199341 | Black – 61.9 White – 68.2 p < 0.01 | Black – 53% White – 61% p < 0.12 | Black – VF 40%, VT 0% White – VF 49%, VT 2% p < 0.09 | Black – 3.4 min White – 3.4 min p > 0.71 | Black – 18% White – 32% p < 0.003 |
| Brookoff et al., 199415 | Black – 61.9 White – 67.4 p = NR | Black – 60% White – 64% p = ns | Black – 44% White – 54% (referent) OR 0.82; 95%CI 0.73–0.93 | No statistically significant difference between blacks and whites (mean times not stated). | Black – 10% White – 21% (referent) OR 0.46 95% CI 0.34–0.61; p < 0.001. |
| Chu et al., 199812 | Black – 62 White – 68 p < 0.001 | Black – 57% White – 61% p = 0.422 | Black – 37% White – 51% p < 0.01 | Black – 5.3 min White – 7.0 min p < 0.001 | Black – 11% White – 20% p = 0.003 |
| Sayegh et al., 199916 | NR/NM | Black – 53% White – 56% p = 0.41 | Black – 43% White – 46% p = 0.286 | ALS ambulance response time <9 min: Black – 88% White – 83% p = 0.039 | Black – 7% White – 10% p = 0.215 |
| Groeneveld et al., 200348 | Black – 76 White – 76 p = ns | NR/NM | NR/NM | NR/NM | NR/NM |
| Polentini et al., 200546 | 0.56 (0.48 – 0.66), whites referent Proportions not reported | ||||
| Fairbanks et al., 200720 | NR/NM | NR/NM | Black – 35% White – 29% p = 0.038 | Black – 5 min White – 5 min p = NR | NR/NM |
| Galea et al., 200717 | NR/NM | Hispanic – 43% Black – 36% White – 42% p < 0.01 | Hispanic – 15% Black – 12% White – 17% p = 0.06 | Hispanic – 5 min Black – 4.7 min White – 4.5 min p < 0.01 | Hispanic – 25% Black – 31% White – 31% p < 0.01 |
| Vadeboncoeur et al., 200853 | Hispanic – 53.2 Non–Hispanic – 64.5 p = 0.005 | Hispanic – 44% Non–Hispanic – 50% p = 0.074 | VF Hispanic – 21% Non-Hispanic – 27% p = 0.036 | Hispanic – 5.1 min Non-Hispanic – 5.5 min p = 0.006 | Bystander CPR Hispanic – 32% Non-Hispanic – 42% p < 0.0001 Lay Bystander CPR Hispanic – 16% Non-Hispanic – 26% p = 0.001 |
| Benson et al., 200918 | Latino – 59.6 African American – 64.7 Caucasian – 69.9 p < 0.05 | Latino – 40% African American – 36% Caucasian – 48% p < 0.05 – (Caucasian vs. African American) p = ns (Caucasian vs. Latino) | Latino – 35% African American – 24% Caucasian – 31% p = ns | NR/NM | Univariate %, OR (95% CI) Latino –13%, 0.48 (0.28–0.80) African American – 13%, 0.47 (0.3–0.74) Caucasian – 24% (referent) Multivariate OR (95% CI) Latino – 0.46, 0.23–0.92 African American – 0.67, 0.38 – 1.16 |
| Teodorescu et al., 201019 | NR/NM | NR/NM | Multivariate Analysis PEA vs. VF/VT, OR (95%CI): Other Race – 1.02, 0.23 – 4.59 Asian – 0.88, 0.2 – 3.98 Hispanic – 0.32, 0.05 – 2.13 Black – 2.64, 1.29 – 5.38 White – 1.0 (referent) PEA vs. Asystole – OR (95% CI): Other Race – 1.10, 0.17 – 7.05 Asian – 0.26, 0.07–1.03 Hispanic – 1.11, 0.15–8.38 Black – 1.3, 0.65–2.63 White – 1.0 (referent) | NR/NM | NR/NM |
| Wilde et al., 201147 | Black – 61.1 White – 68.5 p < 0.001 | NR/NM | NR/NM | Black – 5.3 mins White – 7.3 mins p < 0.001 | Black – 72.5%† White – 80.4% p < 0.001 |
| McNally et al., 201114 | NR/NM | NR/NM | NR/NM | NR/NM | Black – 32.8% White – 33.7% Hispanic – 40.2% p < 0.001 |
| Merchant et al, 201154 | Blacks – 77.4* Whites – 78.1* p=0.01 | NR/NM | VF Blacks – 8%* Whites – 12%* p<0.001 | NR/NM | NR/NM |
| Studies Outside the US | |||||
| Hamaad et al., 200645 | Indo-Asian – 66.0 Afro–Caribbean – 63.7 Caucasian – 68.7 p = ns | NR/NM | VF Indo -Asian – 30% Afro-Caribbean – 30% Caucasian – 45% p = NR | Call to collection time (mins): Caucasian – 28.2 Afro-Caribbean – 31.7 Indo-Asian – 22.6 | NR/NM |
| Shah et al., 201051 | South Asian – 64.6 White – 69.5 p < 0.005 | South Asian −70% White – 62% (referent) OR = 1.1, 95%CI 1.0 – 1.2 | South Asian – 30% White –30% (referent) OR = 1.0, 95% CI 0.7–1.3 | South Asian – 7.46 min White – 7.48 min p = 0.335 | South Asian – 30% (referent) White – 34% OR = 1.2, 95%CI, 0.9–1.7 |
Cohort limited to patients resuscitated from cardiac arrest and subsequently hospitalised
Refers to EMS-provided CPR and not Bystander CPR; Percentages are reported to the nearest whole number
Abbreviations: CI: confidence interval, CPR: cardiopulmonary resuscitation, NM: not measured, NR: not reported, ns: non-significant, OR: odds ratio, VF/VT – ventricular fibrillation/ventricular tachycardia ED: emergency department, EMS: emergency medical services, OOHCA: out-of-hospital cardiac arrest, PEA: pulseless electrical activity, VF/VT ventricular fibrillation/ventricular tachycardia.
Age
Ten studies measured racial or ethnic differences by age (Table 2). Racial or ethnic minorities generally suffered from OOHCA at a younger age than whites12,15,18,41,47,54,51 (six of these studies compared blacks and whites12,15,18,41,47,54).
Witnessed arrest
Of nine studies in Table 2 providing witnessed arrest data, three showed that blacks are less likely to have witnessed arrests than whites11,17,18 and Hispanics,17 while six reported no significant differences in witnessed arrests when comparing blacks to whites,12,15,16,41 Hispanics to non-Hispanics,53 and Latinos to whites.18
Initial rhythm
Of 13 studies in Table 2 assessing initial rhythm, seven showed that blacks were less likely to present with VF/VT following OOHCA, compared to whites.11,12,15,17,19,41,45 Four studies noted no racial nor ethnic differences in initial rhythm when comparing whites, African Americans and Latinos,18 whites to blacks,16 whites to South Asians,51 and whites, Hispanics and Asians.19
Response times
Of 11 studies reporting on response times (Table 2), one found that whites had shorter response times compared to blacks and Hispanics,17 whilst three studies reported blacks to have significantly shorter response times compared to whites,12,16,47 and six studies reported no racial or ethnic differences.11,15,20,41,45,51 One study found that Hispanics had a shorter response time than Non-Hispanics.53
Bystander cardiopulmonary resuscitation
Of 11 studies (Table 2) reporting on bystander CPR rates, five showed that blacks were less likely to receive bystander CPR than whites11,12,15,41,14 and one study showed that blacks were less likely to receive bystander CPR than whites, but this effect was eliminated following adjustments for other variables.18 Three studies reported that Hispanics14,53/Latinos18 were less likely to receive bystander CPR as compared to non-Hispanics53 or whites.14,18 Three studies showed no difference in bystander CPR rates when comparing blacks to whites16,17 and South Asians to whites study.51
Survival
Table 3 summarises survival by race or ethnicity across various time frames in 20 studies. Two studies reported there were no racial/ethnic differences in survival at the primary cardiac arrest site50 nor upon ED resuscitation.49
Ethnic variation in survival following out-of-hospital cardiac arrest (OOHCA)
| Author | Survival to Hospital Admission | Survival to Hospital Discharge | Other Survival End Points |
|---|---|---|---|
| Studies Within US | |||
| Wilcox-Gok, 198544 | Black – 12% White –18% p value/OR not stated§ | Black – 6% Whits – 5% p value/OR not stated§ | NR/NM |
| Becker et al., 199311 | Black – 6% White – 11% p < 0.001 | Black – 13% White – 25% p = 0.001 | Overall survival Black – 1% White – 3% p < 0.001 |
| Cowie et al.,199341 | Black (referent) – 17% White – 40% p < 0.001 OR 3.48 (2.08 – 5.82) | Black (referent)– 9% White – 17% p < 0.03 OR 2.15 (1.11 – 4.17) | Logistic regression, OR (95% CI), (Black, referent) Initial resuscitation – 4.05, 2.25–7.28 Hospital discharge without major neurological disability – 2.64, 1.15–6.07. |
| Brookoff et al., 199415 | Black – 22% White – 29% (referent) OR 0.75, 95% CI 0.61–0.93 | Black – 7% White – 9% p = ns | Survival with no neurological deficit Black – 5% White – 6%, p = ns |
| Chu et al., 199812 | Black – 17% White – 20% p value/OR – NR | Black – 6% White – 7% (referent) Multivariate analysis OR 0.931, 95% CI 0.446–1.945 | NR/NM |
| Lindholm et al., 199850 | NR/NM | NR/NM | Pre hospital survival No difference between race† and survival* (OR 0.77, 95% CI, 0.54–1.08, p = 0.13) |
| Sweeney et al., 199852 | NR/NM | No difference between white and black subjects (p = 0.735) Percentage/numbers by ethnicity not provided | NR/NM |
| Sayegh et al., 199916 | NR/NM | Univariate analysis Black 3% White 6% p = 0.034 Multivariate analysis Adjusted OR = 1.39; 95% CI – 0.65 – 2.97 | NR/NM |
| Groeneveld et al., 200348 | NR/NM | NR/NM | Long term survival following discharge Unadjusted life expectancy Black – 1.9 years White – 4.1 years p < 0.0001 Multivariate analyses (White referent) Adjusted for demographic, clinical and hospital factors Age 66–74 – HR 1.3, 95% CI 1.09–1.55, p = 0.003 Age ≥ 75 – HR 0.88, 95% CI 0.74–1.06, p = 0.18 |
| Polentini et al., 200546 | No difference (p value/OR not stated). | No difference (p value/OR not stated). | NR/NM |
| Fairbanks et al., 200720 | NR/NM | NR/NM | Survival to 1 yr Black 4% White 6% p = 0.67 |
| Galea et al., 200717 | Hispanic – 9% Black – 6% White – 11% p < 0.01 | Hispanic – 2% Black –1% White – 3% p < 0.01 | 30 day survival to hospital discharge Univariate analysis Hispanic – 2% Black – 1% White – 3% p < 0.01 Multivariate analysis – no statistical difference in 30-day survival and ethnicity. |
| Vadeboncoeur et al., 200853 | NR/NM | Multivariate analysis: OR adjusted for age, sex, bystander CPR, witnessed arrest, VF, and interval from EMS dispatch to arrival time Hispanic – 8% Non-Hispanic – 7% (Referent) OR 1.2; 95% CI 0.7–2.1 | NR/NM |
| Liu et al., 20084 || | NR/NM | African American – 35% (Referent) White – 30% OR 0.79, 95% CI 0.62 – 1.0; p = 0.053 | NR/NM |
| Benson et al., 200918 | NR/NM | Hispanic – 5% African American – 3% White – 6%, p = ns | NR/NM |
| Wilde et al., 201147 | NR/NM | NR/NM | Pre-hospital survival (ROSC rates) Univariate Analysis Black – 20% White – 26% OR = 0.69, 95% CI 0.53 –0.91; p = 0.007 Multivariate Analysis: adjusted for age, sex, characteristics prior to cardiac arrest, pre-hospital event factors, patient zip code and agency fixed effects OR = 0.71, 95% CI 0.50 –1.01 |
| McNally et al., 201114 | NR/NM | Hispanic/Latino – 9.4% Black/African American – 8% White – 10% p value/OR not stated | NR/NM |
| Merchant et al., 201154|| | NR/NM | Unadjusted survival at discharge White – 33%* Black – 30%* p < 0.001 Note: cohort limited to surviving patients following cardiac arrest and hospitalisation | NR/NM |
| Studies outside US | |||
| Lim et al., 200249 | NR/NM | NR/NM | No difference between racial distribution and survival‡ (p = 0.93) |
| Shah et al., 201051 | South Asian – 23% White – 22% OR = 0.98, 95% CI, 0.98–0.99) | South Asian – 9% White – 9% OR = 0.98, 95% CI, 0.98 – 1.00 | NR/NM |
| Author | Survival to Hospital Admission | Survival to Hospital Discharge | Other Survival End Points |
|---|---|---|---|
| Studies Within US | |||
| Wilcox-Gok, 198544 | Black – 12% White –18% p value/OR not stated§ | Black – 6% Whits – 5% p value/OR not stated§ | NR/NM |
| Becker et al., 199311 | Black – 6% White – 11% p < 0.001 | Black – 13% White – 25% p = 0.001 | Overall survival Black – 1% White – 3% p < 0.001 |
| Cowie et al.,199341 | Black (referent) – 17% White – 40% p < 0.001 OR 3.48 (2.08 – 5.82) | Black (referent)– 9% White – 17% p < 0.03 OR 2.15 (1.11 – 4.17) | Logistic regression, OR (95% CI), (Black, referent) Initial resuscitation – 4.05, 2.25–7.28 Hospital discharge without major neurological disability – 2.64, 1.15–6.07. |
| Brookoff et al., 199415 | Black – 22% White – 29% (referent) OR 0.75, 95% CI 0.61–0.93 | Black – 7% White – 9% p = ns | Survival with no neurological deficit Black – 5% White – 6%, p = ns |
| Chu et al., 199812 | Black – 17% White – 20% p value/OR – NR | Black – 6% White – 7% (referent) Multivariate analysis OR 0.931, 95% CI 0.446–1.945 | NR/NM |
| Lindholm et al., 199850 | NR/NM | NR/NM | Pre hospital survival No difference between race† and survival* (OR 0.77, 95% CI, 0.54–1.08, p = 0.13) |
| Sweeney et al., 199852 | NR/NM | No difference between white and black subjects (p = 0.735) Percentage/numbers by ethnicity not provided | NR/NM |
| Sayegh et al., 199916 | NR/NM | Univariate analysis Black 3% White 6% p = 0.034 Multivariate analysis Adjusted OR = 1.39; 95% CI – 0.65 – 2.97 | NR/NM |
| Groeneveld et al., 200348 | NR/NM | NR/NM | Long term survival following discharge Unadjusted life expectancy Black – 1.9 years White – 4.1 years p < 0.0001 Multivariate analyses (White referent) Adjusted for demographic, clinical and hospital factors Age 66–74 – HR 1.3, 95% CI 1.09–1.55, p = 0.003 Age ≥ 75 – HR 0.88, 95% CI 0.74–1.06, p = 0.18 |
| Polentini et al., 200546 | No difference (p value/OR not stated). | No difference (p value/OR not stated). | NR/NM |
| Fairbanks et al., 200720 | NR/NM | NR/NM | Survival to 1 yr Black 4% White 6% p = 0.67 |
| Galea et al., 200717 | Hispanic – 9% Black – 6% White – 11% p < 0.01 | Hispanic – 2% Black –1% White – 3% p < 0.01 | 30 day survival to hospital discharge Univariate analysis Hispanic – 2% Black – 1% White – 3% p < 0.01 Multivariate analysis – no statistical difference in 30-day survival and ethnicity. |
| Vadeboncoeur et al., 200853 | NR/NM | Multivariate analysis: OR adjusted for age, sex, bystander CPR, witnessed arrest, VF, and interval from EMS dispatch to arrival time Hispanic – 8% Non-Hispanic – 7% (Referent) OR 1.2; 95% CI 0.7–2.1 | NR/NM |
| Liu et al., 20084 || | NR/NM | African American – 35% (Referent) White – 30% OR 0.79, 95% CI 0.62 – 1.0; p = 0.053 | NR/NM |
| Benson et al., 200918 | NR/NM | Hispanic – 5% African American – 3% White – 6%, p = ns | NR/NM |
| Wilde et al., 201147 | NR/NM | NR/NM | Pre-hospital survival (ROSC rates) Univariate Analysis Black – 20% White – 26% OR = 0.69, 95% CI 0.53 –0.91; p = 0.007 Multivariate Analysis: adjusted for age, sex, characteristics prior to cardiac arrest, pre-hospital event factors, patient zip code and agency fixed effects OR = 0.71, 95% CI 0.50 –1.01 |
| McNally et al., 201114 | NR/NM | Hispanic/Latino – 9.4% Black/African American – 8% White – 10% p value/OR not stated | NR/NM |
| Merchant et al., 201154|| | NR/NM | Unadjusted survival at discharge White – 33%* Black – 30%* p < 0.001 Note: cohort limited to surviving patients following cardiac arrest and hospitalisation | NR/NM |
| Studies outside US | |||
| Lim et al., 200249 | NR/NM | NR/NM | No difference between racial distribution and survival‡ (p = 0.93) |
| Shah et al., 201051 | South Asian – 23% White – 22% OR = 0.98, 95% CI, 0.98–0.99) | South Asian – 9% White – 9% OR = 0.98, 95% CI, 0.98 – 1.00 | NR/NM |
Survival denoted successful resuscitation at the primary cardiac arrest site
Study does not specify which racial group is compared to Black
Survival denoted successful Emergency Department resuscitation
Percentages derived from cohort numbers
Survival at hospital discharge compared by ethnic group in a cohort already admitted to hospital following OOHCA
Percentages are reported to the nearest whole number
Abbreviations: CI: confidence interval, CPR: cardiopulmonary resuscitation, NR: not reported, NM: not measured, ns: not significant, OR: odds ratio, OOHCA: out-of-hospital cardiac arrest, ROSC: return of spontaneous circulation, EMS: Emergency Medical Service, VF: ventricular fibrillation.
Ethnic variation in survival following out-of-hospital cardiac arrest (OOHCA)
| Author | Survival to Hospital Admission | Survival to Hospital Discharge | Other Survival End Points |
|---|---|---|---|
| Studies Within US | |||
| Wilcox-Gok, 198544 | Black – 12% White –18% p value/OR not stated§ | Black – 6% Whits – 5% p value/OR not stated§ | NR/NM |
| Becker et al., 199311 | Black – 6% White – 11% p < 0.001 | Black – 13% White – 25% p = 0.001 | Overall survival Black – 1% White – 3% p < 0.001 |
| Cowie et al.,199341 | Black (referent) – 17% White – 40% p < 0.001 OR 3.48 (2.08 – 5.82) | Black (referent)– 9% White – 17% p < 0.03 OR 2.15 (1.11 – 4.17) | Logistic regression, OR (95% CI), (Black, referent) Initial resuscitation – 4.05, 2.25–7.28 Hospital discharge without major neurological disability – 2.64, 1.15–6.07. |
| Brookoff et al., 199415 | Black – 22% White – 29% (referent) OR 0.75, 95% CI 0.61–0.93 | Black – 7% White – 9% p = ns | Survival with no neurological deficit Black – 5% White – 6%, p = ns |
| Chu et al., 199812 | Black – 17% White – 20% p value/OR – NR | Black – 6% White – 7% (referent) Multivariate analysis OR 0.931, 95% CI 0.446–1.945 | NR/NM |
| Lindholm et al., 199850 | NR/NM | NR/NM | Pre hospital survival No difference between race† and survival* (OR 0.77, 95% CI, 0.54–1.08, p = 0.13) |
| Sweeney et al., 199852 | NR/NM | No difference between white and black subjects (p = 0.735) Percentage/numbers by ethnicity not provided | NR/NM |
| Sayegh et al., 199916 | NR/NM | Univariate analysis Black 3% White 6% p = 0.034 Multivariate analysis Adjusted OR = 1.39; 95% CI – 0.65 – 2.97 | NR/NM |
| Groeneveld et al., 200348 | NR/NM | NR/NM | Long term survival following discharge Unadjusted life expectancy Black – 1.9 years White – 4.1 years p < 0.0001 Multivariate analyses (White referent) Adjusted for demographic, clinical and hospital factors Age 66–74 – HR 1.3, 95% CI 1.09–1.55, p = 0.003 Age ≥ 75 – HR 0.88, 95% CI 0.74–1.06, p = 0.18 |
| Polentini et al., 200546 | No difference (p value/OR not stated). | No difference (p value/OR not stated). | NR/NM |
| Fairbanks et al., 200720 | NR/NM | NR/NM | Survival to 1 yr Black 4% White 6% p = 0.67 |
| Galea et al., 200717 | Hispanic – 9% Black – 6% White – 11% p < 0.01 | Hispanic – 2% Black –1% White – 3% p < 0.01 | 30 day survival to hospital discharge Univariate analysis Hispanic – 2% Black – 1% White – 3% p < 0.01 Multivariate analysis – no statistical difference in 30-day survival and ethnicity. |
| Vadeboncoeur et al., 200853 | NR/NM | Multivariate analysis: OR adjusted for age, sex, bystander CPR, witnessed arrest, VF, and interval from EMS dispatch to arrival time Hispanic – 8% Non-Hispanic – 7% (Referent) OR 1.2; 95% CI 0.7–2.1 | NR/NM |
| Liu et al., 20084 || | NR/NM | African American – 35% (Referent) White – 30% OR 0.79, 95% CI 0.62 – 1.0; p = 0.053 | NR/NM |
| Benson et al., 200918 | NR/NM | Hispanic – 5% African American – 3% White – 6%, p = ns | NR/NM |
| Wilde et al., 201147 | NR/NM | NR/NM | Pre-hospital survival (ROSC rates) Univariate Analysis Black – 20% White – 26% OR = 0.69, 95% CI 0.53 –0.91; p = 0.007 Multivariate Analysis: adjusted for age, sex, characteristics prior to cardiac arrest, pre-hospital event factors, patient zip code and agency fixed effects OR = 0.71, 95% CI 0.50 –1.01 |
| McNally et al., 201114 | NR/NM | Hispanic/Latino – 9.4% Black/African American – 8% White – 10% p value/OR not stated | NR/NM |
| Merchant et al., 201154|| | NR/NM | Unadjusted survival at discharge White – 33%* Black – 30%* p < 0.001 Note: cohort limited to surviving patients following cardiac arrest and hospitalisation | NR/NM |
| Studies outside US | |||
| Lim et al., 200249 | NR/NM | NR/NM | No difference between racial distribution and survival‡ (p = 0.93) |
| Shah et al., 201051 | South Asian – 23% White – 22% OR = 0.98, 95% CI, 0.98–0.99) | South Asian – 9% White – 9% OR = 0.98, 95% CI, 0.98 – 1.00 | NR/NM |
| Author | Survival to Hospital Admission | Survival to Hospital Discharge | Other Survival End Points |
|---|---|---|---|
| Studies Within US | |||
| Wilcox-Gok, 198544 | Black – 12% White –18% p value/OR not stated§ | Black – 6% Whits – 5% p value/OR not stated§ | NR/NM |
| Becker et al., 199311 | Black – 6% White – 11% p < 0.001 | Black – 13% White – 25% p = 0.001 | Overall survival Black – 1% White – 3% p < 0.001 |
| Cowie et al.,199341 | Black (referent) – 17% White – 40% p < 0.001 OR 3.48 (2.08 – 5.82) | Black (referent)– 9% White – 17% p < 0.03 OR 2.15 (1.11 – 4.17) | Logistic regression, OR (95% CI), (Black, referent) Initial resuscitation – 4.05, 2.25–7.28 Hospital discharge without major neurological disability – 2.64, 1.15–6.07. |
| Brookoff et al., 199415 | Black – 22% White – 29% (referent) OR 0.75, 95% CI 0.61–0.93 | Black – 7% White – 9% p = ns | Survival with no neurological deficit Black – 5% White – 6%, p = ns |
| Chu et al., 199812 | Black – 17% White – 20% p value/OR – NR | Black – 6% White – 7% (referent) Multivariate analysis OR 0.931, 95% CI 0.446–1.945 | NR/NM |
| Lindholm et al., 199850 | NR/NM | NR/NM | Pre hospital survival No difference between race† and survival* (OR 0.77, 95% CI, 0.54–1.08, p = 0.13) |
| Sweeney et al., 199852 | NR/NM | No difference between white and black subjects (p = 0.735) Percentage/numbers by ethnicity not provided | NR/NM |
| Sayegh et al., 199916 | NR/NM | Univariate analysis Black 3% White 6% p = 0.034 Multivariate analysis Adjusted OR = 1.39; 95% CI – 0.65 – 2.97 | NR/NM |
| Groeneveld et al., 200348 | NR/NM | NR/NM | Long term survival following discharge Unadjusted life expectancy Black – 1.9 years White – 4.1 years p < 0.0001 Multivariate analyses (White referent) Adjusted for demographic, clinical and hospital factors Age 66–74 – HR 1.3, 95% CI 1.09–1.55, p = 0.003 Age ≥ 75 – HR 0.88, 95% CI 0.74–1.06, p = 0.18 |
| Polentini et al., 200546 | No difference (p value/OR not stated). | No difference (p value/OR not stated). | NR/NM |
| Fairbanks et al., 200720 | NR/NM | NR/NM | Survival to 1 yr Black 4% White 6% p = 0.67 |
| Galea et al., 200717 | Hispanic – 9% Black – 6% White – 11% p < 0.01 | Hispanic – 2% Black –1% White – 3% p < 0.01 | 30 day survival to hospital discharge Univariate analysis Hispanic – 2% Black – 1% White – 3% p < 0.01 Multivariate analysis – no statistical difference in 30-day survival and ethnicity. |
| Vadeboncoeur et al., 200853 | NR/NM | Multivariate analysis: OR adjusted for age, sex, bystander CPR, witnessed arrest, VF, and interval from EMS dispatch to arrival time Hispanic – 8% Non-Hispanic – 7% (Referent) OR 1.2; 95% CI 0.7–2.1 | NR/NM |
| Liu et al., 20084 || | NR/NM | African American – 35% (Referent) White – 30% OR 0.79, 95% CI 0.62 – 1.0; p = 0.053 | NR/NM |
| Benson et al., 200918 | NR/NM | Hispanic – 5% African American – 3% White – 6%, p = ns | NR/NM |
| Wilde et al., 201147 | NR/NM | NR/NM | Pre-hospital survival (ROSC rates) Univariate Analysis Black – 20% White – 26% OR = 0.69, 95% CI 0.53 –0.91; p = 0.007 Multivariate Analysis: adjusted for age, sex, characteristics prior to cardiac arrest, pre-hospital event factors, patient zip code and agency fixed effects OR = 0.71, 95% CI 0.50 –1.01 |
| McNally et al., 201114 | NR/NM | Hispanic/Latino – 9.4% Black/African American – 8% White – 10% p value/OR not stated | NR/NM |
| Merchant et al., 201154|| | NR/NM | Unadjusted survival at discharge White – 33%* Black – 30%* p < 0.001 Note: cohort limited to surviving patients following cardiac arrest and hospitalisation | NR/NM |
| Studies outside US | |||
| Lim et al., 200249 | NR/NM | NR/NM | No difference between racial distribution and survival‡ (p = 0.93) |
| Shah et al., 201051 | South Asian – 23% White – 22% OR = 0.98, 95% CI, 0.98–0.99) | South Asian – 9% White – 9% OR = 0.98, 95% CI, 0.98 – 1.00 | NR/NM |
Survival denoted successful resuscitation at the primary cardiac arrest site
Study does not specify which racial group is compared to Black
Survival denoted successful Emergency Department resuscitation
Percentages derived from cohort numbers
Survival at hospital discharge compared by ethnic group in a cohort already admitted to hospital following OOHCA
Percentages are reported to the nearest whole number
Abbreviations: CI: confidence interval, CPR: cardiopulmonary resuscitation, NR: not reported, NM: not measured, ns: not significant, OR: odds ratio, OOHCA: out-of-hospital cardiac arrest, ROSC: return of spontaneous circulation, EMS: Emergency Medical Service, VF: ventricular fibrillation.
Survival at hospital admission
Eight studies reported on survival to hospital admission (Table 3). Six studies found that whites were more likely to survive, compared to blacks11,12,15,17,41,44 and Hispanics.17 Two studies reported that there were no racial/ethnic differences in survival at hospital admission, when comparing blacks to whites46 and South Asians to whites.51
Survival at discharge/overall survival
Six studies showed that blacks had a lower survival hospital discharge rate when compared to whites;11,14,16,17,41,54 however, Galea et al.17 found no statistically significant difference in 30-day survival, after adjusting for other variables in a regression model. Sayegh et al.16 showed that blacks were less likely to survive at hospital discharge, compared to whites, in a univariate analysis but not in a multivariate analysis.
Four studies comparing whites and blacks found no statistically significant differences in survival when hospital discharge rates were studied.12,15,46,52 Three studies showed no statistically significant difference in survival when comparing whites, African Americans and Hispanics;18 Hispanics with non-Hispanics;53 and whites with South Asians.51
Meta-analysis
Wilde et al.47 compared EMS-provided CPR, as opposed to bystander CPR, in blacks and whites; therefore, we excluded the study from meta-analysis. Merchant et al.54 reported that blacks were less likely to have an admission diagnosis of VF, compared to whites. However, since this study was reporting an admission rhythm rather than an initial rhythm at the site of cardiac arrest, the study was excluded from meta-analysis. One study provided adjusted estimates for bystander CPR,18 two studies provided adjusted estimates for initial VF/VT arrest19,46 and four provided adjusted estimates for survival at hospital discharge.4,12,16,41 Two studies did not report numerical ORs and were not included in the adjusted pooled estimates.11,46 No studies provided adjusted risk estimates for witnessed arrest and survival to hospital admission. A total of 15 studies remained suitable for meta-analysis.
Black patients were less likely to receive bystander CPR (OR = 0.63, 95% CI = 0.53–0.76), have a witnessed cardiac arrest (OR = 0.77, 95% CI = 0.72–0.83) or have an initial VF/VT arrest rhythm (OR = 0.69, 95% CI = 0.59–0.79), as compared to their white counterparts. Black patients had lower rates of survival at hospital admission (OR = 0.59, 95% CI = 0.48–0.72) and at discharge (OR = 0.72, 95% CI = 0.60–0.86) (Figure 1).
Non-adjusted Forest plots for bystander CPR (a), witnessed arrest (b), initial VF/VT arrest (c), survival at hospital admission (d) and survival at hospital discharge (e). CPR: cardiopulmonary resuscitation; M-H: Mantzel-Haenzel; VF: ventricular fibrillation; VT: ventricular tachycardia.
Of the original 15 studies, six had provided adjusted risk estimates across various variables. One study provided adjusted estimates for bystander CPR,18 one study for initial VF/VT arrest19,46 and four studies for survival at hospital discharge.4,12,16,41 Two studies did not report numerical ORs for survival, so were not included in the adjusted pooled estimates.11,46 Using reported adjusted ORs, where available, we found that black patients were still less likely to receive bystander CPR (OR = 0.66, 95% CI = 0.55–0.78), have an initial VF/VT arrest (OR = 0.66, 95% CI = 0.58–0.76) or survive hospital discharge (OR = 0.74, 95% CI = 0.61–0.90) (Figure 2). Confounding factors adjusted within studies varied significantly; these are summarised in online Appendix E.
Forrest plot showing non-adjusted and adjusted meta-estimates comparing white and black populations, looking at: bystander CPR, witnessed arrest, initial VF/VT arrest, survival at hospital admission and survival at hospital discharge. CPR: cardiopulmonary resuscitation; VF: ventricular fibrillation; VT: ventricular tachycardia. CI: confidence interval; NA: not applicable; OR: odds ratio; VF/VT: ventricular fibrillation/ventricular tachycardia.
Sensitivity analysis
Seven of 15 studies had study quality scores at eight or above. A sensitivity analysis, by excluding the studies scoring less than eight, in terms of study quality, yielded similar results (online Appendix B). The funnel plots (online Appendix C) we generated were symmetrical across all variables investigated, making publication bias unlikely. The majority of studies included in the meta-analysis were carried out across various states in the US (online Appendix F).
Discussion
We identified 22 studies investigating OOHCA characteristics, survival and race or ethnicity that primarily compared black, white and Hispanic populations, mostly in the US. Of these, 15 studies comparing blacks with whites were suitable for meta-analysis. Our primary finding highlighted the significant discrepancy between black and white populations, as black patients were found to be less likely to receive bystander CPR, suffer from an initial VF/VT arrest or have a witnessed cardiac arrest. These results further translated into lower survival rates for black cardiac arrest patients, with respect to survival upon both hospital admission and discharge. Additionally, our review highlighted the lack of studies relating to OOHCA characteristics and survival in Europe's migrant and ethnic minority groups. In the mostly US studies used here, ethnic minority groups suffered from OOHCA at a younger age, as compared to white populations. These differences in age between ethnic groups could potentially reflect the younger underlying population age structure of the ethnic minorities and not simply more morbidity.
Given the presence of ethnic variations which have been known for some time and the increasingly multi-ethnic nature of urban populations, this topic deserves higher priority. This retrospective emphasizes the need for further research in this area, given the advances in both primary and secondary prevention interventions available for myocardial infarctions. Conclusions on other racial or ethnic groups are difficult to make, given the scarcity of studies in these groups. Only one population-based study has investigated the role of ethnicity when comparing South Asians and whites in greater London, UK.51 All studies performed to date indicate that there are important variations that require more and larger studies before firm global ethnicity-related conclusions can be drawn.
To our knowledge, this is the first systematic review and meta-analysis to focus on the effects of race or ethnicity on OOHCA characteristics. Our search strategy was extensive and encompassed all the available published studies that investigate race or ethnicity as a stratifying variable in OOHCA.
Limitations
We identified 15 studies suitable for meta-analysis comparing black and white populations. Several limitations should be considered when interpreting this meta-analysis. First, we found significant heterogeneity across the variables studied, when comparing black and white populations. This may reflect different population structures, varying ethnic proportions, the sample sizes and/or participant characteristics. However, pooled ORs showed consistency across all variables and survival outcome measures in both ethnic groups. The heterogeneity of methods remains an important and difficult issue to control in performing meta-analyses of observational studies.64 Future studies need to be designed with prior agreement to standardisation, to allow comparability.
Secondly, data extraction and analyses were not blinded to the authors; however, our literature screening and data extraction were conducted by two independent investigators, making selection bias unlikely. Meta-analysis of observational studies has limitations with inherent biases, including geographical bias and the degree of heterogeneity of the populations analysed. Despite this, meta-analysis has been increasingly used for statistical synthesis of epidemiological data.64
Thirdly, we did not collect the primary data from the relevant studies, but had we done so we might have been able to adjust the pooled risk estimates utilizing our own regression models. We saw that few studies attempted to adjust for confounders (online Appendix E). Using the adjusted risk estimates, in place of actual cohort numbers, the associations between ethnicity and cardiac arrest variables did not change, when comparing white and black populations (Figure 2). Whether the confounders are the true reason behind ethnic differences in OOHCA variables and survival does need further research.
Given the funnel plot analysis, we were unlikely to have missed out on studies with no differences between black and white populations nor studies where black populations would have fared better. Online Appendix E shows that the population cohorts studied involved most states in the US, making localised geographical bias less likely and the results more relevant for the entirety of the US. Our meta-analysis yielded clear conclusions with relation to black and white populations in the US, adding substantially to the prior fragmented and often contradictory picture.
Conclusion
Our work highlights the significant discrepancy in OOHCA characteristics and survival in relation to race, with black patients less likely to receive bystander CPR, have a witnessed cardiac arrest or an initial VF/VT arrest, which translates into black patients having lower survival rates, when compared to white patients. Out-of-hospital cardiac arrest is an important public health and clinical issue that contributes significantly to cardiovascular mortality and morbidity. The relationship of race or ethnicity to OOHCA is complex, with many potential confounders including socioeconomic status. Although these confounders need to be disentangled, we believe that the differential burden of OOHCA, whatever the causes, needs action, even as research proceeds.
We found a variety of concepts and terminology relating to race or ethnicity and its association with OOHCA characteristics. For this field to progress, especially to aid future systematic review and meta-analysis, more standardisation is essential. Larger studies with more clearly standardised concepts and terminology need to be conducted to allow for effective comparisons between racial or ethnic groups and OOHCA characteristics, especially outside the US and in a wider variety of global populations. Further research is needed to disentangle and explain why these differences exist: For example, are individuals from poorer backgrounds less likely to receive basic life support training and do these poor areas consist of a disproportionate number of minority individuals?
Out-of-hospital cardiac arrest registries and databases could help achieve equitable health outcomes by race or ethnicity, but this would require a concerted, coordinated effort to standardise data collection, analysis and data outputs.65 Currently, given the importance of the problem and the multi-ethnic nature of modern-day urbanised societies in much of the world, we were surprised to find that the scientific literature on the subject of cardiac arrest by ethnic or racial group was so small.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest
None declared.
Comments