Association between shockable rhythms and long-term outcome after pediatric out-of-hospital cardiac arrest in Rotterdam, the Netherlands: An 18-year observational study

Introduction: Shockable rhythm following pediatric out-of-hospital cardiac arrest (pOHCA) is consistently associated with hospital and short-term survival. Little is known about the relationship between shockable rhythm and long-term outcomes (>1 year) after pOHCA. The aim was to investigate the association between first documented rhythm and long-term outcomes in a pOHCA cohort over 18 years. Methods: All children aged 1 day 18 years who experienced non-traumatic pOHCA between 2002 2019 and were subsequently admitted to the emergency department (ED) or pediatric intensive care unit (PICU) of Erasmus MC-Sophia Children’s Hospital were included. Data was abstracted retrospectively from patient files, (ground) ambulance and Helicopter Emergency Medical Service (HEMS) records, and follow-uped retrospectively from patient files, (ground) ambulance and Helicopter Emergency Medical Service (HEMS) records, and follow-up clinics. Long-term outcome was determined using a Pediatric Cerebral Performance Category (PCPC) score at the longest available follow-up interval through august 2020. The primary outcome measure was survival with favorable neurologic outcome, defined as PCPC 1 2 or no difference between preand post-arrest PCPC. The association between first documented rhythm and the primary outcome was calculated in a multivariable regression model. Results: 369 children were admitted, nine children were lost to follow-up. Median age at arrest was age 3.4 (IQR 0.8 9.9) years, 63% were male and 14% had a shockable rhythm (66% non-shockable, 20% unknown or return of spontaneous circulation (ROSC) before emergency medical service (EMS) arrival). In adolescents (aged 12 18 years), 39% had shockable rhythm. 142 (39%) of children survived to hospital discharge. On median followup interval of 25 months (IQR 5.1 49.6), 115/142 (81%) of hospital survivors had favorable neurologic outcome. In multivariable analysis, shockable rhythm was associated with survival with favorable long-term neurologic outcome (OR 8.9 [95%CI 3.1 25.9]). * Corresponding author. E-mail address: c.buysse@erasmusmc.nl (C.M.P. Buysse). https://doi.org/10.1016/j.resuscitation.2021.05.015 0300-9572/© 2021 The Authors. Elsevier B.V.This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Conclusion: In children with pOHCA admitted to ED or PICU shockable rhythm had significantly higher odds of survival with long-term favorable neurologic outcome compared to non-shockable rhythm. Survival to hospital discharge after pOHCA was 39% over the 18-year study period. Of survivors to discharge, 81% had favorable long-term (median 25 months, IQR 5.1 49.6) neurologic outcome. Efforts for improving outcome of pOHCA should focus on early recognition and treatment of shockable pOHCA at scene.


Introduction
Pediatric out of hospital cardiac arrest (pOHCA) is uncommon, with incidences ranging from 9.0 to 19.7 per 100,000 person-years. 1À4 Whereas CA in adults is mostly of cardiac origin, in pediatrics it is commonly due to respiratory failure. 5 Survival following pOHCA is poor, especially among infants, 6,7 but increasing due to 'chain-of-survival' improvements. 7À13 Children receive more bystander basic life support (BLS), more automated external defibrillators (AED's) are available and post-return of spontaneous circulation (ROSC) care has improved, despite AED use in children remaining low. 6,7,9,13,14 Shockable rhythms in children seem more common than once thought, 15,16 especially in adolescents (aged 12À18 years) with a prevalence of 19%. 7 The positive association between shockable rhythm and short-term outcomes (ROSC, survival to hospital discharge (SHD) and outcome up to 1 year) has been reported but true long-term follow-up (>1 year after event) is lacking. 6,17 Is increased short-term survival rate after pOHCA associated with more children with severe long-term neurological sequelae due to hypoxic ischemic brain injury 18À20 ? To be able to detect a child's full potential (neurologic) recovery, a statement from the American Heart Association recently recommended one year of follow-up minimally. 21 Literature on outcomes beyond one-year following pOHCA is scarce, often small in sample size, using different and mostly crude measurements and mainly based on data prior to 2008. 17,22À27 Since 2012 the Erasmus MC-Sophia Children's Hospital has a long-term follow-up program including all pOHCA, as part of standard of care, which led to the following subjective observations: 1) the incidence of shockable rhythms increased over time and 2) shockable pOHCA's achieve favorable long-term neurological outcome more frequently compared with non-shockable pOHCA's.
The aim of this study was to investigate the association of first documented cardiac arrest (CA) rhythm on true long-term outcome in non-traumatic pOHCA. We hypothesized that a shockable rhythm was positively associated with survival with long-term favorable neurologic outcome.

Study design
This cohort study was performed at the PICU of the Erasmus MC-Sophia Children's Hospital, a tertiary-care university children's hospital in the Netherlands. The hospital and Helicopter Emergency Medical Service (HEMS) provide health care in the southwest of the Netherlands with approximately five million inhabitants, about 25% of the Dutch population. The Medical Ethics Review Board of the Erasmus MC approved the data collection and gave a waiver for the requirement of informed consent (MEC-2019-0440).

Inclusion criteria
All children aged 24 h to 18 years with non-traumatic pOHCA, admitted to the Erasmus MC-Sophia Children's Hospital (ED or PICU) with or without CPR in progress between January 2002 and August 2019 were included. Arrests in neonates younger than 24 h were excluded as they are generally caused by perinatal asphyxia. CA was defined as the need for chest compressions for at least one minute. Cardiopulmonary resuscitation was defined as 'basic life support', in line with the European Resuscitation Council Guidelines, and if needed, followed by 'advanced pediatric life support' (APLS). 5

Data collection
Existing CPR databases were used to combine CPR data from 2002 until 2019. 23,28 All CPR data were derived from ground ambulance records, HEMS records and hospital health record systems. Because HEMS are always deployed in the Netherlands in (suspected) pOHCA, all HEMS records between 2002 and 2019 were also analysed to get an insight of pre-hospital mortality and potential transport to other hospitals. In some rare cases of conflict between data sources (ground EMS and HEMS) HEMS data was used as golden standard.
Data included: A) basic child characteristics (age, gender, parent's Social Economic Status (SES), pre-existing health status). The SES was calculated using a 'Status Score' divided into tertiles to interpret a 'low status (1)', 'intermediate status (2)' and 'high status (3)'. 29 The 'Status Score' is based on income, education level and unemployment rate by postal code. B) OHCA characteristics (year, location, first documented rhythm (shockable/non-shockable or unknown), witnessed, cause, bystander CPR, use of AED, CPR duration, extracorporeal CPR (ECPR), targeted temperature management, first blood lactate and pH after ROSC or at hospital arrival, regional transport, re-arrest). C) outcome (pre-hospital mortality, ROSC, SHD and neurologic outcome at the longest available follow-up interval).
At the longest available follow-up interval the neurologic outcome was determined using a Pediatric Cerebral Performance Category score (PCPC, ranging from 1 to 6) and a Functional Status Scale score (FSS, ranging from 6 to 30). The PCPC and FSS scores are internationally validated scores for assessing a child's overall cognitive and functional status after critical illness or injury. 30,31 The PCPC and FSS scores were based on one of four possible sources: 1) the prospective longitudinal follow-up outpatient clinic database (2012À2019 cohort). 2) the cross-sectional outcome database (2002À2011 cohort). 23 3) hospital letters from outpatient clinic visits. 4) hospital discharge letters after the pOHCA. Both crosssectional and prospective follow-up databases included validated neurocognitive and daily functioning questionnaires. Hospital letters contained more crude descriptions. The PCPC and FSS were scored by two physicians and one pediatric neurologist independently and in case of disagreement (in less than 5% of cases) agreement was reached through a consensus meeting.

Outcome measures
The primary outcome measure was survival with favorable neurologic outcome at the longest available follow-up interval. Survival with favorable neurologic outcome was defined as a PCPC score of 1À2 or no difference between pre-and post-arrest PCPC, in hospital survivors at the longest available follow-up interval. Unfavorable outcome was defined as: no ROSC, no survival to hospital discharge despite ROSC and PCPC 3À6. Secondary outcome measures were survival and favorable neurological outcome in the group of hospitalsurvivors.
No universal definition of favorable neurologic outcome exists. The PCPC score is mostly based on daily activity and school performance so 'favorable outcome' largely depends on a country's school system. Favorable neurologic outcome has been defined in the literature as PCPC 1À2 as well as PCPC 1À3. 9,21,32 Because in the Netherlands, a high threshold for attending a special needs classroom exists, favorable neurologic outcome was defined as PCPC 1À2.

Statistical analysis
Baseline characteristics and survival outcome were reported using descriptive statistics. Categorical variables were reported as percentages and frequencies, and differences were analyzed with Chi-square test or Fisher's exact test when applicable. Continuous data was presented as median and interquartile ranges (IQR) for skewed data, and mean and standard deviation (SD) for normal distributed data. Differences were tested using an independent sample t-test for continuous data or MannÀWhitney U test dependent on normality.
The associations of first documented rhythm, AED use, bystander BLS, year of event and the post AED guideline change period with long-term neurologic outcome were calculated with a multivariable logistic regression model. The choice of inclusion of covariates was made in three steps. First, the following covariates were considered based on existing literature: age, gender, pre-existing condition (yes or no and related to CPR event or not), SES (1, 2 or 3), event location (private or public), year of event (including before and after the AED guideline change), witnessed arrest (yes or no), bystander CPR (yes or no), bystander AED use (yes or no), CPR duration (in minutes), first documented cardiac arrest rhythm (shockable, non-shockable or unknown), cause of arrest (specific), ECPR (yes or no) and first lactate and pH after ROSC. Second, collinearity analysis to explore correlation between all covariates using a correlation matrix was performed. A cut-off value of >0.7 was used for the exclusion of variables in the model. Third, inclusion of the abovementioned potential confounders in the final models was based on >10% change of the effect estimate in the crude model. These covariates were entered one-by-one in the crude model to see the effect on the effect estimate.
Results are presented as odds ratio (OR) and 95%-confidence interval (CI).
A sensitivity analysis comparing the different definitions of favorable neurologic outcome (PCPC 1À vs PCPC 1À3, or no preand post-arrest difference) was performed. Stratified analysis by age group (below and above 8 years of age; infant; aged <1 year, child; aged 1À11 years and adolescent; aged 12À18 years) was also done. Lastly, a propensity score analysis using 1:1 nearest-neighbor matching of shockable to non-shockable rhythm was performed. The propensity score was estimated using a multivariable logistic regression model including the following variables: gender, age at arrest and year of event. Both groups were tested for association with long-term neurologic outcome using a multivariable logistic regression model.
Our data contained missing values for CPR duration (19%). Other covariates had <10% missing data. Variables were imputed using multiple imputation (n = 5 imputations) function based on the distribution of existing data.
A two-tailed p-value <0.05 was considered statistically significant. All analyses were conducted using SPSS software version 24 (IBM SPSS Statistics for Windows, Armonk, New York, USA).

Child and CA characteristics
The target population consisted of 581 children, of whom 138 (24%) had termination of resuscitation and were pronounced deceased at scene and 74 (13%) were transported to other hospitals by HEMS. Of 369 eligible children admitted to the Erasmus MC-Sophia, 360 were included (9 children, 2%, had missing data). An overview of the inclusion is given in Fig. 1. The basic characteristics are presented in Table 1.

Multivariable analysis
The crude associations were adjusted for witnessed arrest, bystander CPR, age at arrest, year of arrest, first lactate, pre-existing conditions related to arrest and CPR duration. After adjustment, first documented shockable rhythm showed significantly improved odds of favorable outcome compared with non-shockable rhythm, with an OR of 8.9 [95% CI 3.1À25.9] (Table 3)

Supplementary material
Stratified analysis for age are presented in the supplementary material. It proved unfeasible to create a nearest-neighbor propensity matching model (for 1:1 as well as 1 to many matching) because of the age distribution of shockable compared to non-shockable rhythm. The results are therefore not presented. The child and CA characteristics sorted by age group are presented in Supplementary Table S4. In adolescents (aged 12À18 years) the incidence of shockable rhythm was 39%. In the analysis stratified by age group an unknown rhythm was associated with favorable outcome in children <8 years (OR 5.6 [95% CI 3.6À8.8]) and children 8 Table S6).

Discussion
Over an 18-year period and after a median follow-up of 25 months, this retrospective single-center study of pOHCA showed a nine times higher odds of shockable rhythms surviving with long-term favorable neurologic outcome compared to non-shockable rhythm, even after adjustment for confounders. First documented rhythms were 14% shockable (in adolescents, aged 12À18 years, 39%), 66% non-shockable and 20% unknown. SHD after pOHCA was 39%. 81% of hospital survivors achieved long-term favorable neurologic outcome and of all included children 32% survived with favorable neurologic outcome. 17,22,24 Only few studies have true long-term follow-up and are thus comparable with the present study. We will summarize these, beyond case reports or series. 17,22À24,27 The study of Meert et al., a secondary analysis of The Therapeutic Hypothermia after Pediatric Cardiac Arrest Out-of-Hospital (THAP-CA-OH) trial, has comparable methodology as the present study as children were included after OHCA upon admission to hospital. 17 They also found that shockable rhythm was associated with greater 12-month survival and greater 12-month survival with favorable neurobehavioral functioning, assessed using the Vineland Adaptive Behavior Scales.     25,26 They found significant neuropsychological and neurobehavioral deficits in initially comatose pOHCA survivors although they were classified one year post-arrest as having favorable neurologic outcome. In addition they observed 3month outcomes to be predictive of outcomes after 1 year. 33 Van Zellem studying in-and out-of-hospital arrests et al. used different IQ tests, neuropsychological tests and questionnaires, incomparable with the PCPC scoring system. 23 Lopez-Herce et al. found in 95 children (multicenter, 1998À1999), 17% favorable neurologic outcome after one year. 24  What are the implications of the present study? First, shockable rhythm was shown to significantly and relevantly improve odds of true long-term favorable outcome. With favorable outcome defined as PCPC 1À3 the relationship was even stronger. And most notably in children eight years and above, shockable rhythm was statistically significantly associated with favorable outcome with OR 22.7 [11.6À44.8). This can be explained by the relatively high incidence of shockable rhythm in adolescents (aged 12À18 years) (39%). Also young children are less likely to have an AED used during CPR than older children, possibly because arrests are more often occurring at home rather than in public locations where AEDs are  available. In a cohort study from an OHCA registry in Japan, the proportion of adults with a favorable neurological outcome 30 days after event was significantly higher in those who received publicaccess defibrillation than those who did not (845 [37.7%] vs 5676 [22.6%]. 34 Our results might implicate that the efforts for improving outcome of pOHCA should focus on early recognition and treatment of shockable OHCA at scene and the importance of improvements in the chain of survival (e.g. bystander BLS, public access to and use of AED and adequate EMS response). 35,36 Second, a remarkable finding was that 81% of survivors to hospital discharge achieved long-term favorable neurologic outcome beyond 1 year. This could be due to the setting in the Netherlands (e.g. high incidence in AED use and bystander CPR, the availability of HEMS 24/ 7, short transfer time from the scene to the hospital). Another possible explanation could be that in our study cohort the main cause of inhospital mortality after ROSC was WLST (21%), probably due to poor neurologic prognosis. Less WLST could lead to higher survival to discharge numbers, but with more severe neurologically damaged children surviving long-term. Accurate neurological prognostication in a comatose child after OHCA remains challenging and no international pediatric guidelines exist. 21,37,38 Potentially inaccurate prognostication and WLST may bias outcome. 28,37,39,40 Third, the median age at time of follow-up was 6.6 years (IQR 3.4 À13.4), which is relatively young in childhood and thus growing into deficits might not yet be present. Moreover, neurologic outcome was measured by PCPC, which is a crude outcome scale ranging from 1 to 6 (from no disabilities to brain death). It is unknown whether PCPC reflected how these children function in daily life and if it was associated with detailed neuropsychological functioning. In our opinion, it is crucial to identify how these pOHCA survivors will function on different physical and neuropsychological domains when reaching adolescence or young adulthood. Will they be able to live independently and happy, have a job and start a family? The importance to understand the influence of an arrest on long-term education and development as children grow into adulthood seems clear. 21 True long-term follow-up is time and resource consuming, with the potential of losing children to follow-up. 21 Long-term follow-up outpatient clinics have to be set up also beyond the 18 year boundary to support this group in maximizing outcome.
Our study has several limitations. First, it was an observational, retrospective single center study. Secondly, there were missing data due to the incomplete documentation of the CPR-event (e.g. CPR duration), which required imputation in up to 10% of the data. We minimized this potential bias by doing supplemental analyses with and without imputation. Additionally, we were not able to report and correct for some important CPR characteristics (e.g. quality of CPR, post-ROSC care). Finally, our study is not a complete regional or national pOHCA study since only children admitted to our hospital (with or without CPR in progress) were included. This could have led to selection bias by not including those children who died at scene or transferred to another hospital.

Conclusion
Shockable pOHCA had an almost nine times higher odds of long-term favorable neurologic survival compared to non-shockable rhythm, adjusted for confounding. The overall SHD after pOHCA was 39% over the 18-year study period, of which 81% of survivors achieved long-term (median 25 months, IQR 5.1À49.6) favorable neurologic outcome. This indicates the efforts for improving outcome of pOHCA should focus on early recognition and treatment of shockable pOHCA at scene.

CRediT author statement
The contributions of the authors were as follows: M. Albrecht, R. De Jonge and C. Buysse had the original idea for the study. M. Albrecht, as first author, participated in its design, performed the statistical analysis, interpreted the data, drafted and critically revised the article. All co-authors revised the manuscript critically for important intellectual content. C. Buysse provided supervision. All authors read and approved the final article. All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.