Neurologically favorable survival was defined as a Cerebral Performance Score of 1 or 2 at the time of discharge. Error bars indicate 95% CI. CPR indicates cardiopulmonary resuscitation.
aComparison between conventional CPR and other groups. There was no significant difference between compression-only CPR and no bystander CPR.
eTable. Univariate arrest characteristics and bivariable association with type of bystander CPR.
eFigure 1. Significant interactions for survival to hospital discharge comparison of no bystander CPR with bystander CPR.
eFigure 2. Significant interactions for neurologically favorable survival comparison of no bystander CPR with bystander CPR.
eFigure 3. Significant interactions for survival to hospital discharge comparison of no bystander CPR, compression-only CPR, and conventional CPR.
eFigure 4. Significant interactions for neurologically favorable survival comparison of no bystander CPR, compression-only CPR, and conventional CPR.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Naim MY, Burke RV, McNally BF, et al. Association of Bystander Cardiopulmonary Resuscitation With Overall and Neurologically Favorable Survival After Pediatric Out-of-Hospital Cardiac Arrest in the United States: A Report From the Cardiac Arrest Registry to Enhance Survival Surveillance Registry. JAMA Pediatr. 2017;171(2):133–141. doi:10.1001/jamapediatrics.2016.3643
What are the characteristics and outcomes of bystander cardiopulmonary resuscitation (CPR) and compression-only CPR following pediatric out-of-hospital cardiac arrest?
In this analysis of data from the Cardiac Arrest Registry to Enhance Survival, bystander CPR was provided in 1814 of 3900 instances (46.5%) of out-of-hospital cardiac arrests in children younger than 18 years and was significantly associated with improved overall survival and neurologically favorable survival. Conventional CPR had improved outcomes compared with compression-only CPR.
Public health efforts to improve the provision of CPR in minority communities and increase the use of conventional bystander CPR may improve outcomes for pediatric out-of-hospital cardiac arrests.
There are few data on the prevalence or outcome of bystander cardiopulmonary resuscitation (BCPR) in children 18 years and younger.
To characterize BCPR in pediatric out-of-hospital cardiac arrests (OHCAs).
Design, Setting, and Participants
This analysis of the Cardiac Arrest Registry to Enhance Survival database investigated nontraumatic OHCAs in children 18 years and younger from January 2013 through December 2015.
Bystander CPR, which included conventional CPR and compression-only CPR.
Main Outcomes and Measures
Overall survival and neurologically favorable survival, defined as a Cerebral Performance Category score of 1 or 2, at the time of hospital discharge.
Of the 3900 children younger than 18 years with OHCA, 2317 (59.4%) were infants, 2346 (60.2%) were female, and 3595 (92.2%) had nonshockable rhythms. Bystander CPR was performed on 1814 children (46.5%) and was more common for white children (687 of 1221 [56.3%]) compared with African American children (447 of 1134 [39.4%]) and Hispanic children (197 of 455 [43.3%]) (P < .001). Overall survival and neurologically favorable survival were 11.3% (440 of 3900) and 9.1% (354 of 3900), respectively. On multivariable analysis, BCPR was independently associated with improved overall survival (adjusted proportion, 13.2%; 95% CI, 11.81-14.58; adjusted odds ratio, 1.57; 95% CI, 1.25-1.96) and neurologically favorable survival (adjusted proportion, 10.3%; 95% CI, 9.10-11.54; adjusted odds ratio, 1.50; 95% CI, 1.21-1.98) compared with no BCPR (overall survival: adjusted proportion, 9.5%; 95% CI, 8.28-10.69; neurologically favorable survival: adjusted proportion, 7.59%; 95% CI, 6.50-8.68). For those with data on type of BCPR, 697 of 1411 (49.4%) received conventional CPR and 714 of 1411 (50.6%) received compression-only CPR. On multivariable analysis, only conventional CPR (adjusted proportion, 12.89%; 95% CI, 10.69-15.09; adjusted odds ratio, 2.06; 95% CI, 1.51-2.79) was associated with improved neurologically favorable survival compared with no BCPR (adjusted proportion, 9.59%; 95% CI, 6.45-8.61). There was a significant interaction of BCPR with age. Among infants, conventional BCPR was associated with improved overall survival and neurologically favorable survival while compression-only CPR had similar outcomes to no BCPR.
Conclusions and Relevance
Bystander CPR is associated with improved outcomes in pediatric OHCAs. Improving the provision of BCPR in minority communities and increasing the use of conventional BCPR may improve outcomes for children with OHCA.
More than 5000 children experience an out-of-hospital cardiac arrest (OHCA) every year in the United States.1,2 The outcome of these arrests is poor, with a mortality rate greater than 90%.2,3 An important clinical predictor of survival following adult OHCA is bystander cardiopulmonary resuscitation (BCPR),4 yet most studies reveal that only one-third of children receive BCPR following OHCA.2
Widespread national and statewide efforts have been undertaken to improve BCPR rates in adults and children that include the performance of compression-only CPR (COR) for bystanders who are reluctant to perform rescue breaths.5,6 Although COR appears to be as effective as conventional CPR with chest compressions and rescue breaths for adults with OHCA,5 it may be less effective for children, whose OHCAs are most commonly from asphyxia.7,8 Therefore, the American Heart Association recommends conventional CPR for pediatric cardiac arrest instead of COR. However, if a bystander is unable to perform rescue breathing, the American Heart Association recommends COR, which is preferable to no CPR at all.9,10
The objectives of this study were 2-fold: The first aim was to characterize BCPR in pediatric OHCAs and test the hypothesis that BCPR would be associated with improved overall survival and neurologically favorable survival at hospital discharge from a large cardiac arrest registry in the United States.11 The second aim was to characterize bystander COR and compare survival outcomes following bystander COR with conventional CPR in children.
A secondary data analysis of the Cardiac Arrest Registry to Enhance Survival (CARES) database was conducted. The CARES database was established by the US Centers for Disease Control and Prevention in collaboration with the Department of Emergency Medicine at the Emory University School of Medicine in Atlanta, Georgia. The CARES database includes an overall catchment area of more than 90 million people in 37 states across the United States. The CARES is an emergency medical services–based registry for OHCA and is composed of a limited standard set of data elements from 3 sources: 911 call centers, responding emergency medical service professionals, and receiving hospitals.12 Responding emergency medical service professionals were defined as personnel (ie, emergency medicine technicians and paramedics) who responded to the medical emergency in an official capacity as part of an organized medical response team and were the designated transporters of the patient to the hospital. Detailed information on the design and development of this registry as well as the data elements included in the registry is published elsewhere.11
From January 1, 2013, through December 31, 2015, CARES captured all 911-activated nontraumatic cardiac arrest events, defined as apnea and unresponsiveness in which resuscitation with either CPR or defibrillation was attempted. Children with obvious signs of death (eg, rigor mortis or dependent lividity) or for whom a “do not resuscitate” order was respected were not included. Standardized international Utstein Style definitions for defining clinical variables and outcomes were used to ensure uniformity in reporting.13 Because the etiology of cardiac arrest in children is difficult to determine, especially in cases that result in death, all patients with nontraumatic OHCA were included regardless of presumed etiology, including respiratory, cardiac, drowning, electrocution, or other. Only instances of traumatic OHCA were excluded from the analysis. Analysts of the CARES database confirmed the capture of all cardiac arrests by each community’s 911 center during the data review process. Cerebral Performance Category information was obtained from receiving hospitals. The CARES established a point of contact at each participating hospital and trained the contact on CARES hospital element definitions and the data entry process. The local hospital contact abstracted information from patients’ medical records and entered data. All data were entered using a web-based platform, and an Excel file (Microsoft Corporation) was generated with all cardiac arrest events for the specified date range.
This study was approved by the institutional review boards at both Emory University and The Children’s Hospital of Los Angeles. Given the use of deidentified data, the study was determined to be exempt for review by the Children’s Hospital of Philadelphia Institutional Review Board of The Children’s Hospital of Philadelphia. Because the study used data from a deidentified database, informed consent was waived.
All children 18 years and younger with nontraumatic OHCA submitted to the registry during the study period were eligible for inclusion. Arrests that occurred in medical facilities or nursing homes, traumatic arrests, and 911 responder–witnessed arrests were excluded.
Patient characteristics obtained from the database included age, sex, race/ethnicity, bystander witness status, arrest location, rhythm type, and automated external defibrillator (AED) use. Race/ethnicity data were evaluated to determine whether CPR provision was influenced by race/ethnicity. Race/ethnicity was assigned as considered by the child, family, or 911 responder. The race/ethnicity category included white, African American, Hispanic/Latino, and other (eg, American Indian/Alaskan, Asian, and Native Hawaiian/Pacific). The primary outcomes of interest were overall survival and neurologically favorable survival, defined as a Cerebral Performance Category score of 1 (no neurologic disability) or 2 (moderate disability) at the time of discharge.14 Unfavorable neurologic outcome was defined as a Cerebral Performance Category score of 3 (severe disability) or 4 (coma or vegetative state) or death. Secondary outcomes included survival after BCPR with COR compared with conventional BCPR. Bystander CPR was defined as CPR administered by a layperson (ie, a lay family member, layperson, or layperson with medical training not part of the organized emergency medical response). Conventional CPR was defined as CPR with chest compressions and rescue breaths; COR was defined as CPR with chest compressions only.
Descriptive analyses report overall child and arrest characteristics, characteristics for arrests that had BCPR vs no bystander CPR, and characteristics for arrests that had COR and conventional CPR. Bivariable statistics were used to describe the percentage of arrests with survival to hospital discharge and neurologically favorable survival. Logistic regression models with robust standard errors were used to examine the associations of the predictors with the probability of survival to hospital discharge and neurologically favorable survival. The analysis was adjusted for potential confounders, including age, sex, arrest witness status, location of the arrest, rhythm type, and AED use. Further analyses examined whether BCPR modified the results by all other covariates via interactions. Results are expressed as odds ratios and standardized marginal probabilities with 95% CIs. Significance was set at P < .05, and all associations were determined through 2-sided testing. Analyses were performed using SAS version 9.3 (SAS Institute) and STATA version 10.3 (StataCorp).
A total of 3900 children younger than 18 years with OHCA were evaluated (Table 1). Most cardiac arrests occurred in infants (59.4% [2317 of 3900]), and arrests occurred more frequently in girls (60.2% [2346 of 3900]) than in boys. Patients who were white (1221 of 3900 [31.3%]) composed the largest racial group, followed by African American (1134 of 3900 [29.1%]) and Hispanic (455 of 3900 [11.7%]) patients. Most arrests occurred at home/residence (83.7% [3263 of 3900]), were unwitnessed (72.2% [2817 of 3900]), and had nonshockable rhythms (ie, asystole or pulseless electrical activity; 92.2% [3595 of 3900]). Most arrests did not receive AED application by a bystander or first responder prior to emergency medical service arrival (82.6% [3222 of 3900]). The overall survival rate was 11.3% (440 of 3900), and the neurologically favorable survival rate was 9.1% (354 of 3900).
Bystander CPR was provided in 46.5% (1814 of 3900) of arrests, most commonly by a family member (67.4% [1222 of 1814]), followed by a layperson (22.9% [416 of 1814]) and layperson with medical training (9.7% [176 of 1814]). Bystander CPR was more common for white (687 of 1221 [56.3%]) compared with African American (447 of 1134 [39.4%]) and Hispanic (197 of 455 [43.3%]) children (P < .001). Bystander CPR was also more common in girls, witnessed arrests, nonhome/public arrests, arrests with a shockable rhythm, and arrests in which an AED was used.
For bivariable analysis, survival to hospital discharge was associated with BCPR (BCPR, 14.28%; no BCPR, 8.68%; P < .001) (Table 2 and Table 3) and . Other factors associated with improved survival to hospital discharge included age older than 1 year, female sex, white race/ethnicity, witnessed arrests, nonhome/public arrests, shockable rhythm, and AED use (Table 2 and Table 3). For multivariable analysis, BCPR was independently associated with survival to hospital discharge compared with no BCPR (adjusted proportion, 13.19%; 95% CI, 11.81-14.58 vs 9.48%; 95% CI, 8.28-10.69; adjusted odds ratio [OR], 1.57; 95% CI, 1.25-1.96). Other predictors of survival to hospital discharge included age older than 1 year, female sex, witnessed arrests, nonhome/public arrests, and arrests with a shockable rhythm. Several interactions noted with BCPR included age, sex, race/ethnicity, witnessed arrests, arrest location, rhythm type, and AED use (eFigure 1 in the Supplement). For infants, survival to hospital discharge was not associated with BCPR, although there was a significant association with BCPR in children older than 1 year.
For neurologically favorable survival, BCPR was significantly associated compared with no BCPR in bivariable analysis (BCPR, 11.63%; no BCPR, 6.86%; P < .001). Other factors associated with improved neurologically favorable survival for children included age older than 1 year, female sex, white race/ethnicity, witnessed arrests, nonhome/public location, shockable rhythm, and AED use. The significant association of BCPR with neurologically favorable survival held in the multivariable model compared with no BCPR (adjusted proportion, 10.32%; 95% CI, 9.10-11.54 vs 7.59%; 95% CI, 6.50-8.68; adjusted OR, 1.54; 95% CI, 1.21-1.98). Predictors of neurologically favorable survival in children included age older than 1 year, white race/ethnicity, witnessed arrests, nonhome/public location, and shockable rhythm. Several interactions with BCPR were also noted, including age, sex, race/ethnicity, witness status, arrest location, rhythm type, and AED use (eFigure 2 in the Supplement). For infants, neurologically favorable survival was not associated with BCPR, although there was a significant association of BCPR in children older than 1 year.
Of the 1814 children that received BCPR, the type of CPR was available for 1411 (77.8%). For those events where type of CPR data was available, 49.4% (697 of 1411) received conventional CPR and 50.6% (714 of 1411) received COR (eTable in the Supplement). Children older than 1 year and African American and Hispanic children were more likely to receive COR compared with conventional CPR. There was no difference in the type of BCPR by sex, witnessed arrest, arrest location, rhythm type, and AED use.
An analysis was conducted on the association of type of BCPR with no BCPR and outcomes (Table 4 and Table 5). For bivariable analysis, conventional CPR and COR resulted in higher rates of survival to hospital discharge compared with no BCPR (conventional CPR, 17.07%; COR, 14.15%; no BCPR, 8.68%; P < .001). On multivariable analysis, conventional CPR (adjusted proportion, 16.79%; 95% CI, 14.26-19.31; adjusted OR, 2.23; 95% CI, 1.69-2.95) and COR (adjusted proportion, 12.09%; 95% CI, 9.99-14.19; adjusted OR, 1.14; 95% CI, 1.05-1.97) were independently associated with survival to hospital discharge compared with no BCPR (adjusted proportion, 9.37%; 95% CI, 8.17-10.57). On bivariable analysis, conventional CPR and COR had improved neurologically favorable survival compared with no BCPR (conventional CPR, 13.34%; COR, 11.48%; no BCPR, 6.83%; P < .001). On multivariable analysis, conventional CPR (adjusted proportion, 12.89%; 95% CI, 10.69-15.09; adjusted OR, 2.06; 95% CI, 1.51-2.79) was independently associated with neurologically favorable survival compared with no BCPR (adjusted proportion, 7.54%; 95% CI, 6.45-8.61). There were several important interactions with type of BCPR for survival to hospital discharge and neurologically favorable survival, including age, sex, race/ethnicity, witnessed arrests, arrest location, rhythm type, and AED use (eFigure 3 and eFigure 4 in the Supplement). For infants, conventional BCPR was associated with higher rates of survival to hospital discharge and neurologically favorable survival. However, COR had outcomes that were similar to no BCPR (Figure). For children older than 1 year, both conventional CPR and COR were associated with higher rates of survival to hospital discharge and neurologically favorable survival compared with no BCPR.
To our knowledge, this is the first report focused on the association of BCPR and COR in children from the United States. There are several important findings to highlight. Bystander CPR was provided for 47% of these children after cardiac arrest, which is greater than previously reported.2 Second, BCPR was associated with a higher survival to hospital discharge and neurologically favorable survival than previously reported from the United States.2,3 Third, a racial disparity exists in the provision of BCPR. Fourth, COR and conventional CPR were provided in a similar number of arrests. Fifth, conventional CPR was associated with better outcomes than COR for children overall. Sixth, both conventional CPR and COR were associated with better outcomes than no BCPR for children aged 1 to 18 years. Seventh, although infants composed the largest age group, BCPR was only associated with improved outcomes in infants when ventilations were also performed.
In this study, BCPR was provided in just under half of pediatric OHCAs (46.5%), which is similar to reports from Japan (52%)15 but higher than previous reports from the United States.1,2 The Resuscitation Outcomes Consortium reported a 35% incidence of BCPR in 624 children with OHCA that occurred between 2005 and 2007.2 This increase in BCPR may be secondary to an increase in COR by bystanders unwilling to perform conventional CPR with rescue breaths. This is supported by the present study, which shows that an almost equal number of children received conventional CPR and COR.
This report found a survival to hospital discharge of 11.3%, which is higher than previously reported in the United States.1-3 In a report from the Resuscitation Outcomes Consortium, survival to hospital discharge was 6.4%.2 In a more recent report from the CARES registry evaluating OHCA in 1980 children between 2005 and 2013, 8.2% survived to hospital discharge.3 This improvement is presumably because of the increased BCPR noted in this study as well as improvements in management after cardiac arrest.16,17 The CARES is a voluntary registry, and agencies that choose to participate may have more resources to dedicate to quality improvement initiatives, resulting in higher survival rates than other agencies.18 Greater than 80% of children who survived a cardiac arrest had neurologically favorable survival. Overall, 9.1% had neurologically favorable survival at discharge, substantially higher than the 3.2% and 5.1% favorable neurologic survival reported by Kitamura et al19 and Akahane et al,15 respectively, in Japan.
These data also reveal a racial disparity in the provision of BCPR. White children were much more likely to receive BCPR compared with African American or Hispanic children. A similar racial disparity has been observed in adult studies. Becker et al20 studied 6451 adults with nontraumatic OHCA in the Chicago, Illinois, area between 1987 and 1988 and found that African American adults were less likely to receive BCPR, have a witnessed cardiac arrest, have a shockable rhythm, or be admitted to a hospital. More recently, Sasson et al21 studied 14 225 adults with cardiac arrest from the CARES registry and found that patients who had an OHCA in low-income and predominantly African American neighborhoods were less likely to receive BCPR than those in high-income and predominantly white neighborhoods. A study examining whether race/ethnicity was an independent predictor of survival to hospital discharge when controlled for income observed 1690 adults with OHCA in an affluent population and found that race/ethnicity did not predict adverse outcomes; however, African American individuals were less likely to receive BCPR compared with white individuals (11% vs 20%; P = .002).22 This report suggests a potential target for intervention in African American and Hispanic communities where a public health strategy can be undertaken to enhance BCPR in children.
In the present study, COR was provided as commonly as conventional CPR. Neurologically favorable survival was more likely to occur after conventional CPR. A recent study from Japan23 on the association of dispatcher-assisted CPR on neurological outcomes in children with OHCA had similar findings, with a benefit of conventional CPR over COR on overall survival and neurologically favorable outcome 1 month following OHCA. Another study from Japan by Kitamura et al19 evaluated the association of the type of BCPR in children with OHCA in which one-third of children received conventional CPR and another 17% received COR. For children with arrests from noncardiac causes, conventional CPR had improved neurological outcomes compared with COR; however, in children with arrests from cardiac causes, both conventional and COR were similarly effective.19 The current study was not able to differentiate etiology of cardiac arrest, although conventional CPR was associated with improved overall outcomes compared with COR.
Most pediatric OHCAs occurred in infants, similar to previous studies.2 Most of these arrests were most likely secondary to sudden infant death syndrome,7 which is usually unwitnessed, discovered long after death has occurred. This is in stark contrast to the outcomes of in-hospital cardiac arrests, which are generally witnessed and where infants have the greatest survival rates.24,25 There was no observed benefit of BCPR for infants unless ventilations were performed, which was not the case in older children. For children older than 1 year, there was a significant benefit of bystander COR and conventional CPR compared with no BCPR. As the survival rate among infants remains low and public health campaigns stress the provision of chest compressions over ventilations (a more difficult technique), alternative public health strategies may be needed to improve outcomes in this age group.
This study has several limitations. The data are observational and therefore may be associated with unmeasured confounding. Determining the etiology of a cardiac arrest in the field is problematic, so all nontraumatic etiologies were considered regardless of the presumed initial field etiology that was chosen. However, multiple previous reports describe respiratory disease as the most common etiology of OHCA in children.7,8 Type of BCPR is a supplemental element in CARES and therefore was not available for all arrests because only a subset of agencies chose to enter this data element. In addition, there were missing data for 72 patients, who were excluded from analysis. This study did not include CPR quality data; however, previous studies show that CPR quality in children is poor even when performed by professional rescuers,26,27 and presumably, the provision of low-quality CPR is better than no CPR. There were no data on the duration of CPR or time to CPR. In addition, there is no long-term follow-up of survivors.
Bystander CPR is associated with improved outcomes in children with OHCA. Conventional BCPR is associated with improved outcomes compared with COR, and among infants, there was no benefit of BCPR unless ventilations were provided. Efforts to improve the provision of CPR in minority communities and increasing the use of conventional BCPR may improve outcomes for children with OHCA.
Corresponding Author: Maryam Y. Naim, MD, Division of Cardiac Critical Care, Departments of Anesthesiology, Critical Care Medicine, and Pediatrics, The Children’s Hospital of Philadelphia, Perelman School of Medicine, University of Pennsylvania, S 34th Street and Civic Center Boulevard, Philadelphia, PA 19104 (firstname.lastname@example.org).
Accepted for Publication: September 20, 2016.
Published Online: November 12, 2016. doi:10.1001/jamapediatrics.2016.3643
Author Contributions: Dr Griffis and Mr Song had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Naim, Griffis, Markenson, Bradley, Rossano.
Acquisition, analysis, or interpretation of data: Naim, Burke, McNally, Song, Griffis, Berg, Vellano, Markenson, Rossano.
Drafting of the manuscript: Naim, Song, Griffis, Markenson, Rossano.
Critical revision of the manuscript for important intellectual content: Burke, McNally, Griffis, Berg, Vellano, Markenson, Bradley, Rossano.
Statistical analysis: Burke, Song, Griffis, Markenson, Rossano.
Administrative, technical, or material support: McNally, Vellano, Markenson, Bradley, Rossano.
Conflict of Interest Disclosures: None reported.
Funding/Support: The Cardiac Arrest Registry to Enhance Survival was funded by the US Centers for Disease Control and Prevention from 2004 to 2012. The program is now supported through private funding from the American Red Cross, the American Heart Association, the Medtronic Foundation HeartRescue Program, Physio-Control Corporation, Zoll Corporation, and in-kind support from Emory University. At present, the Centers for Disease Control and Prevention provides technical support and expertise for the program.
Role of the Funder/Sponsor: The study sponsors had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Previous Presentation: This work was presented at the American Heart Association Scientific Sessions 2016; November 12, 2016; New Orleans, Louisiana.
Additional Contributions: We thank the emergency medical services and hospital personnel who have completed the patient care reports for the Cardiac Arrest Registry to Enhance Survival (https://mycares.net/sitepages/map.jsp).