Article, Cardiology

Disparities in bystander CPR provision and survival from out-of-hospital cardiac arrest according to neighborhood ethnicity

a b s t r a c t

Study objective: We aimed to determine if there are differences in bystander cardiopulmonary resuscitation (BCPR) provision and survival to hospital discharge from out-of-hospital cardiac arrest (OHCA) occurring in Hispanic neighborhoods in Arizona.

Methods: We analyzed a prospectively collected, statewide Utstein-compliant OHCA database between January 1, 2010, and December 31, 2012. Cases of OHCA were geocoded to determine their census tract of event location, and their neighborhood main ethnicity was assigned using census data. Neighborhoods were classified as “Hispanic” or “non-Hispanic white” when the percentage of residents in the census tract was 80% or more.

Results: Among the 6637 geocoded adult OHCA victims during the study period, 4821 cases were included in this analysis, after excluding 1816 cases due to incident location, traumatic cause, or because the arrest occurred after emergency medical service arrival. In OHCAs occurring at Hispanic neighborhoods as compared with non- Hispanic white neighborhoods, the provision of BCPR (28.6% vs 43.8%; P b .001) and initially monitored shockable rhythm (17.3% vs 25.7%; P b .006) was significantly less frequent. Survival to hospital discharge was significantly lower in Hispanic neighborhoods than in non-Hispanic white neighborhoods (4.9% vs 10.8%; P =

.013). The adjusted odds ratio (OR) of Hispanic neighborhood for BCPR provision (OR, 0.62; 95% confidence interval, 0.44-0.89) was lower as compared with non-Hispanic white neighborhoods.

Conclusions: In Arizona, OHCA patients in Hispanic neighborhoods received BCPR less frequently and had a lower survival to hospital discharge rate than those in non-Hispanic white neighborhoods. Public health efforts to attenuate this disparity are needed.

(C) 2014 The Authors. This is an open access article under the CC BY-NC-ND license


? This study was presented at the American Heart Association Resuscitation Science

Symposium, November 2013, Dallas, TX.

?? There was no grant or other financial support for this study.

? Conflicts of interest: Drs Bobrow and Spaite disclose that the University of Arizona

receives support from the Medtronic Foundation involving community-based trans- lation of resuscitation science.

?? Author contributions: S.M. and B.J.B. conceived and designed the study. B.J.B. and

D.W.S. supervised the conduct of the study and data collection. B.J.B. and U.S. managed the data, including quality control. S.M., U.S., W.K., M.K., and B.J.B. provided statistical advice on study design and analyzed the data. S.M. drafted the manuscript. B.J.B., T.F.V., and C.S. contributed substantially to its drafting and editing. S.M. takes responsibility for the article as a whole.

* Corresponding author at: Department of Emergency Medicine, Korea University Ansan Hospital, 123 Jeokgeumro, Danwongu, Ansansi, Gyeonggido 425-707, Korea. Tel.: + 82 31

412 5380; fax: + 82 31 412 5315.

E-mail address: [email protected] (S. Moon).


Out-of-hospital cardiac arrest (OHCA) is a leading cause of death and has been regarded as an important public health issue [1]. Various efforts have been made to establish effective community response systems for OHCA, but there are still large differences in the survival rate between communities [2-4]. Provision of bystander cardiopulmonary resuscitation (BCPR) strongly influences the outcome after OHCA [1,5- 8]. In prior studies, the probability of BCPR provision was different according to the race or ethnicity [9-12]. Community responses to OHCA differed by neighborhood demographics, and it has been shown that OHCA patients in black neighborhoods were less likely to receive BCPR compared with those in white neighborhoods [13,14].

0735-6757/(C) 2014 The Authors. This is an open access article under the CC BY-NC-ND license (

In Arizona, 30.2% of the population are Hispanic [15]. Hispanics are the fastest growing segment of the US population and are estimated to account for 29% of the population by the year 2050 [16]. To implement Public health measures for OHCA, it is important to better understand how population demographics are associated with BCPR provision.

Our aim was to determine if there are differences in BCPR provision and survival to hospital discharge from OHCAs occurring in Hispanic neighborhoods compared with non-Hispanic white neighborhoods in Arizona.


Study design and selection of participants

This is an observational cohort study using our statewide Utstein- compliant OHCA database between January 2010 and December 2012. The study population included consecutive adults (aged >= 18 years) with OHCA. Cases were excluded from the analysis if resuscitation was not initiated, the patient had a do not resuscitate order, the arrest was witnessed by emergency medical services (EMS) personnel, the arrest occurred at a long-term care or health care facility, or the cause of the arrest was trauma. We also excluded arrests that occurred at the airport, in jail, or in government buildings because we thought that these areas would not be affected by neighborhood characteristics.


Arizona has approximately 6.6 million residents and is composed of 15 counties with demographics varying from urban to rural areas [15]. Arizona’s population demographics are 57.1% non-Hispanic white alone (63% nationwide), 4.5% black or African American alone (13.1% nationwide), 5.3% American Indian and Alaska Native alone (1.2% nationwide), 3.1% Asian alone (5.1% nationwide), and 30.2%

Hispanic (16.9% nationwide) [15].

In 2005, there were 30 EMS agencies participating in the state- sponsored OHCA quality improvement program: the Save Hearts in Arizona Registry and Education (SHARE) program, which has been described elsewhere [17,18]. Participation has increased each year of the program, and by 2010, 104 EMS agencies (serving approximately 80% of the population) reported their OHCA data to the SHARE program.

Because OHCA has been designated as a major Public health problem in Arizona and the goal of this program is quality improvement, the data collected were exempt from the Health Insurance Portability and Accountability Act. Permission to publish the deidentified data was obtained from the Arizona Department of Health Services Human Subjects Review Board and the University of Arizona Institutional Review Board.

Data collection and measurement

We obtained patient characteristics and outcomes from the SHARE database as well as information about whether BCPR was provided. Characteristic variables included age, sex, whether the arrest was witnessed, initial rhythm, and location of arrest (private or public). The obtained incident addresses were geocoded using Centrus Desktop geocoder, version 4.0 (Pitney Bowes, Boulder, CO), to determine their census tracts. For each OHCA case, we determined the main ethnicity based on the population composition in census tracts of the 2010 census data [19]. We classified neighborhoods as “Hispanic” when the Hispanic population was more than 80% or “non- Hispanic white” when the non-Hispanic white population was above 80% [13]. If the proportion of either Hispanic population or non- Hispanic white population was not more than 80%, the neighborhood main ethnicity was classified as “integrated.” In addition, poverty rate of neighborhoods was determined in each OHCA case according to the Census Bureau-released 5-year estimates for small areas data, which

were based on American Community Survey data collected from 2005 through 2009 [19].

Outcome measures

The primary outcome measures were the differences in the rate of BCPR provision and survival to hospital discharge according to the neighborhood main ethnicity.

Data analysis

For the purpose of this report, data were abstracted by the SHARE program data manager who was not blinded to the study question. The data were then transported from Microsoft Access for Windows (Microsoft, Redmond, WA) into the Stata 12.1 (StataCorp, College Station, TX) statistical package program for the statistical analysis. Subjects who could not be geocoded due to incomplete address information were excluded from the analysis. The rate of BCPR provision and survival to hospital discharge in each group according to neighborhood main ethnicity were compared using the ?2 test. A multivariate logistic regression analysis was performed to determine the association between BCPR provision and survival to hospital discharge with neighborhood main ethnicity while adjusting for confounders: age, sex, if arrests were witnessed or not, location of arrest, call to EMS arrival time, initially monitored rhythm (for survival to discharge), and neighborhood poverty rate. The goodness- of-fit for regression model was assessed by Hosmer-Lemeshow test.


Characteristics of study subjects

During the study period, 7404 adult OHCA victims were identified in our SHARE database, and 6637 cases (89.6%) were geocoded with their incident addresses. The remaining 767 victims had insufficient address information. After excluding cases based on the exclusion criteria, 4821 cases were included in this analysis (Figure). Mean age of the study population was 62.5 +- 17.3 years, and 65.0% (3133/4821) were male. There were 215 (4.5%) OHCAs in Hispanic neighborhoods, 1428 (29.6%) in non-Hispanic white neighborhoods, and 3178 (65.9%) in integrated neighborhoods (Table 1). Out-of-hospital cardiac arrest patients in the Hispanic neighborhoods were younger than patients in non-Hispanic white neighborhoods (58.3 vs 67.4 years; P b .001). In OHCAs occurring in Hispanic neighborhoods, call to EMS arrival time was significantly shorter (5.1 +- 3.7 vs 5.8 +- 3.0 minutes; P b .001), and Shockable rhythms (ventricular fibrillation or ventricular tachycardia) were less frequently observed at the scene (17.3% vs 25.7%; P = .006) as compared with non-Hispanic white neighborhood.

Main results

Bystander cardiopulmonary resuscitation was provided in 42.4% (2013/4751) of the total study population, and 10.5% (465/4443) survived to hospital discharge.

Bystander cardiopulmonary resuscitation was provided less frequently to OHCA patients in Hispanic neighborhoods (61/213, 28.6%) than in non-Hispanic white neighborhoods (612/1399, 43.8%) (P b .001). Survival to hospital discharge was significantly lower in Hispanic neighborhoods (9/183, 4.9%) than in non-Hispanic white neighborhoods (143/1325, 10.8%) (P = .013).

In the adjusted model, odds ratio (OR) of Hispanic neighborhood for BCPR provision was significantly lower than non-Hispanic white neighborhood (OR, 0.62; 95% confidence interval [CI], 0.44-0.89). Adjusted OR of Hispanic neighborhood for survival to hospital discharge was 0.50 (95% CI, 0.23-1.08) (Table 2).

Starting population: 7,404 consecutive adult OHCAs

767 arrests with insufficient address

6,637 adult OHCAs (89.6%) with geo-coded incident

Excluded cases:

159 Arrests with traumatic cause 384 Arrests after EMS arrival

1,127 Arrests occurring at long term care facilities or health care facilities

41 Arrests occurring at jail or government buildings

15 Arrests occurring at airport

90 Arrests with lividity, Rigor mortis or DNR

Study population: 4,821 adult non-traumatic OHCAs

Figure. Study population inclusion/exclusion flow chart. DNR, do not resuscitate.


In the United States, the ethnic or racial composition is different according to the region and state. Hispanics are the fastest growing population in the United States, and the state of Arizona has a relatively higher proportion of Hispanics [19]. To the best of our knowledge, this is the first study to report the relationship between OHCA care and outcomes in predominantly Hispanic neighborhoods. Neighborhood effects on various public health issues have been investigated and regarded as having a substantial influence on variations and disparities in the incidence, care, and outcome for other diseases or conditions [20-22]. There have been reports assessing the relationship between the responses to OHCA and neighborhood

characteristics [13,14,20,21].For example, previous studies have found large and unacceptable differences in black vs white main neighborhoods. Importantly, Sasson et al [13] found that patients who had an OHCA in black neighborhoods were less likely to receive BCPR than those in white neighborhoods.

In our previous study of Hispanic BCPR in Arizona [9], we investigated the differences in the rate of BCPR provision based on the victim’s ethnicity, and we found that significantly fewer Hispanic OHCA victims received BCPR compared with non-Hispanic OHCA victims. In the current study, we determined the differences in BCPR provision according to the neighborhood characteristic of ethnicity. Considering the critical role of the First responders in the Chain of survival including early access, early cardiopulmonary resuscitation

Table 1

Demographics, bystander CPR, and survival to hospital discharge according to the neighborhood main ethnicity

Non-Hispanic whitea (n = 1428, 29.6%)

Integratedb (n = 3178, 65.9%)

Hispanicc (n = 215, 4.5%) Pd

Age, mean +- SD (y)

67.4 +- 16.5

60.6 +- 17.2

58.3 +- 18.1

b .001

Sex (male), no./total no. (%)

964/1428 (67.5)

2034/3178 (64.0)

134/214 (62.6)


witnessed arrest, no./total no. (%)

509/1411 (36.1)

1027/3143 (32.7)

50/212 (23.6)

b .001

Time to EMS arrival, mean +- SD (min)

5.8 +- 3.0

5.4 +- 2.6

5.1 +- 3.7

b .001

Location of arrest (public), no./total no. (%)

211/1404 (15.0)

447/3118 (14.3)

25/210 (11.9)


Bystander CPR performed, no./total no. (%)

612/1399 (43.8)

1340/3138 (42.7)

61/213 (28.6)

b .001

Initial shockable rhythm, no./total no. (%)

368/1428 (25.7)

670/3178 (21.1)

37/214 (17.3)


Survival to hospital discharge, no./total no. (%)

143/1325 (10.8)

313/2935 (10.7)

9/183 (4.9)


a Non-Hispanic white population is more than 80% in their neighborhood.

b Neither Hispanic population or non-Hispanic white population is more than 80% in their neighborhood.

c Hispanic population is more than 80% in their neighborhood.

d P values were obtained by comparing between non-Hispanic white and Hispanic.

Table 2

Adjusted ORs for bystander CPR and survival to hospital discharge

tion campaigns focusing on the neighborhoods with the highest need should be implemented to increase community BCPR rates [27].

Provision of bystander CPR, OR (95% CI)

Age (y) Sex

0.99 (0.98-0.99)

0.98 (0.97-0.99)





0.92 (0.81-1.04)

0.71 (0.56-0.90)






Bystander CPR

1.32 (1.16-1.50)

1.97 (1.57-2.47)

Survival to discharge, OR (95% CI)

There are some limitations to our study. First, there were missing data in BCPR provision (71/4821) and survival to discharge (378/4821). It is possible that these missing data affected the results of this investigation. Second, we did not control for the possibility that post-cardiac arrest hospital variables, including therapeutic hypothermia, percutaneous coronary intervention, or other postarrest therapies, varied according to the neighborhood ethnicity. Finally, neighborhood ethnicity is an imperfect surrogate for the ethnicity of the BCPR provider and for other unforeseen impacts of ethnicity on the quality of care provided.

Not provided – Reference

Provided – 1.19 (0.95-1.47)

Initially monitored rhythm

Nonshockable – Reference

Shockable – 5.15 (4.12-6.45) Location of arrest





1.88 (1.57-2.23)

1.94 (1.51-2.51)

EMS arrival time in min 1.00 (0.98-1.01) 0.96 (0.93-1.00)

Neighborhood ethnicitya

Non-Hispanic white




1.00 (0.86-1.15)

0.98 (0.76-1.27)


0.62 (0.44-0.89)

0.50 (0.23-1.08)

Neighborhood poverty rate

Lower than state average



Higher than state average

0.74 (0.65-0.85)

1.05 (0.82-1.33)

The goodness-of-fit P values on Hosmer-Lemeshow testing were .38 and .60 for provision of bystander CPR and survival to discharge, respectively.

a Non-Hispanic white: non-Hispanic white population is more than 80% in their neighborhood. Integrated: neither Hispanic population nor non-Hispanic white population is more than 80% in their neighborhood. Hispanic: Hispanic population is more than 80% in their neighborhood.

(CPR), and, possibly, Early defibrillation, understanding neighborhood characteristics and their effect is important for tailored policy making and implementation of measures in a certain community.

We found that rates of initial shockable rhythm were lower in Hispanic neighborhoods despite shorter call to EMS arrival times. This finding is consistent with the lower witnessed arrest rates and lower rates of BCPR in Hispanic neighborhoods.

Low rates of BCPR are regarded as a significant contributor to the poor survival rate for OHCA patients in certain communities [23]. There have been many large-scale efforts to increase the rate of BCPR provision in Arizona, such as increasing the use of Hands-only CPR for bystanders, changing the traditional CPR educational paradigm by using brief and ultrabrief educational videos, and implementing school-based CPR and automated external defibrillator training in both English and Spanish [23-25]. In addition, to assess for disparities among communities, the importance of systemwide data collection and analysis has been emphasized.

After performing multivariate logistic regression analysis controlling for the confounders including neighborhood poverty rate, the relation- ship between the neighborhood main ethnicity and BCPR provision was still significant. However, the OR of Hispanic neighborhood for survival to hospital discharge was not significant in our adjusted model (0.51; 95% CI, 0.24-1.09). It is possible that this is related to the relatively small number of communities and, thus, OHCA, which met our 80% criteria to define a neighborhood main ethnicity.

Public health policies and interventions aimed at attenuating these disparities according to the neighborhood main ethnicity have to be addressed. The reasons for lower BCPR rates in Hispanic neighbor- hoods are unknown but might be related to various factors including but not limited to access to training, economic barriers precluding CPR training, and language barriers that preclude exposure to community CPR efforts and access to 9-1-1 telephone prearrival CPR instructions that are in English [26]. Better understanding of vulnerable neigh- borhoods and barriers in BCPR provision and training is needed [27]. Furthermore, culturally sensitive, community-specific public educa-


In this study, we found that OHCA victims in Hispanic neighborhoods in Arizona received BCPR less frequently and had lower survival to hospital discharge rates than those in non-Hispanic white neighborhoods. Further investigation to determine the specific causes for these disparities and public health efforts to attenuate them are needed.


  1. Travers AH, Rea TD, Bobrow BJ, Edelson DP, Berg RA, Sayre MR, et al. Part 4: CPR Overview: 2010 American Heart Association Guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2010;122:S676-84.
  2. Berdowski J, Berg RA, Tijssen JG, Koster RW. Global incidences of out-of-hospital cardiac arrest and survival rates: Systematic review of 67 prospective studies. Resuscitation 2010;81:1479-87.
  3. Nichol G, Thomas E, Callaway CW, Hedges J, Powell JL, Aufderheide TP, et al. Regional variation in out-of-hospital cardiac arrest incidence and outcome. JAMA 2008;300:1423-31.
  4. Rea TD, Eisenberg MS, Simibaldi G, White RD. Incidence of EMS-treated out-of- hospital cardiac arrest in the United States. Resuscitation 2004;63:17-24.
  5. Bobrow BJ, Spaite DW, Berg RA, Stolz U, Sanders AB, Kern KB, et al. Chest compression-only CPR by lay rescuers and survival from out-of-hospital cardiac arrest. JAMA 2010;304:1447-54.
  6. Holmberg M, Holmberg S, Herlitz J. Effect of bystander cardiopulmonary resuscitation in out-of-hospital cardiac arrest patients in Sweden. Resuscitation 2000;47:59-70.
  7. Swor RA, Jackson RE, Cynar M, Sadler E, Basse E, Boji B, et al. Bystander CPR, ventricular fibrillation, and survival in witnessed, unmonitored out-of-hospital cardiac arrest. Ann Emerg Med 1995;25:780-4.
  8. Sasson C, Rogers MA, Dahl J, Kellermann AL. Predictors of survival from out-of- hospital cardiac arrest: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes 2010;3:63-81.
  9. Vadeboncoeur TF, Richman PB, Darkoh M, Chikani V, Clark L, Bobrow BJ. Bystander cardiopulmonary resuscitation for out-of-hospital cardiac arrest in the Hispanic vs the Non-Hispanic populations. Am J Emerg Med 2008;26:655-60.
  10. Brookoff D, Kellermann AL, Hackman BB, Somes G, Dobyns P. Do blacks get bystander cardiopulmonary resuscitation as often as whites? Ann Emerg Med 1994;24:1147-50.
  11. Becker LB, Han BH, Meyer PM, Wright FA, Rhodes KV, Smith DW, et al. Racial differences in the Incidence of cardiac arrest and subsequent survival. N Engl J Med 1993;329:600-6.
  12. Galea S, Blaney S, Nandi A, Silverman R, Vlahov D, Foltin G, et al. Explaining racial disparities in incidence of and survival from out-of-hospital cardiac arrest. Am J Epidemiol 2007;166:534-43.
  13. Sasson C, Magrid DJ, Chan P, Root ED, McNally BF, Kellermann AL, et al. Association of neighborhood characteristics with bystander-initiated CPR. N Engl J Med 2012;367:1607-15.
  14. Sasson C, Keirns CC, Smith DM, Sayre MR, Macy ML, Meurer WJ, et al. Examining the contextual effects of neighborhood on out-of-hospital cardiac arrest and the provision of bystander cardiopulmonary resuscitation. Resuscitation 2011;82:674-9.
  15. US Census Bureau, US Department of Commerce. Available at: http://quickfacts. . [Accessed June 15, 2013].
  16. PewResearch Hispanic Trends Project, U.S. Population Projections: 2005-2050. Available at: 2005-2050/ . [Accessed September 7, 2013].
  17. Vadeboncoeur T, Bobrow BJ, Clark L, Kern KB, Sanders AB, Berg RA, et al. The Save Hearts in Arizona Registry and Education (SHARE) program: who is performing CPR and where are they doing it? Resuscitation 2007;75:68-75.
  18. Bobrow BJ, Vadeboncoeur TF, Clark L, Chikani V. Estabilishing Arizona’s statewide cardiac arrest reporting and educational network. Prehosp Emerg Care 2008;12:381-7.
  19. US Census Bureau, US Department of Commerce. Available at: http://factfinder2. . [Accessed August 15, 2013.].
  20. Iwashyna TJ, Christakis NA, Becker LB. Neighborhoods matter: a population based study of provision of cardiopulmonary resuscitation. Ann Emerg Med 1999;34:459-68.
  21. Sasson C, Keirns CC, Smith D, Sayre M, Macy M, Meurer W, et al. Small area variations in out-of-hospital cardiac arrest: does the neighborhood matter? Ann Intern Med 2010;153:19-22.
  22. Diez-Roux AV. Investigating neighborhood and area effects on health. Am J Public Health 2001;91:1783-9.
  23. Abella BS, Aufderheide TP, Eigel B, Hickey RW, Longstreth WT, Nadkarni V, et al. Reducing barriers for implementation of bystander-initiated cardiopul- monary resuscitation: a scientific statement from the American Heart Association for healthcare providers, policymakers, and community leaders regarding the effectiveness of cardiopulmonary resuscitation. Circulation 2008;117:704-9.
  24. Sayre MR, Berg RA, Cave DM, Page RL, Potts J, White RD, et al. Hands only (compression-only) cardiopulmonary resuscitation: a call to action for bystander response to adults who experience out-of-hospital sudden cardiac arrest: a science

    advisory for the public from the American Heart Association Emergency Cardiovas- cular Care Committee. Circulation 2008;117:2162-7.

    Cave DM, Aufderheide TP, Beeson J, Ellison A, Gregory A, Hazinski MF, et al. Importance and implementation of training in cardiopulmonary resuscitation and Automated external defibrillation in schools: a science advisory from the American Heart Association. Circulation 2011;123:691-706.

  25. Bradley SM, Fahrenbruch CE, Meischke H, Allen J, Bloomingdale M, Rea TD, et al. Bystander CPR in out-of-hospital cardiac arrest: The role of limited English proficiency. Resuscitation 2011;82:680-4.
  26. Sasson C, Meischke H, Abella BS, Berg RA, Bobrow BJ, Chan PS, et al. Increasing cardiopulmonary resuscitation provision in communities with low bystander cardiopulmonary resuscitation rates: a science advisory from the American Heart Association for healthcare providers, policymakers, public health departments, and community leaders. Circulation 2013;127:132-50.

Leave a Reply

Your email address will not be published. Required fields are marked *