Demographics, management, and outcomes of out-of-hospital traumatic cardiac arrest: a retrospective cohort study comparing children and adults

Abstract

Aim

Out-of-hospital traumatic cardiac arrests (TCA) are associated with a poor prognosis, yet limited research focuses on paediatric TCA. This study aimed to compare outcomes following TCA between children and adults.

Methods

We conducted a retrospective cohort study using data from the French cardiac arrest registry (RéAC) between July 2011 and March 2023. We included all patients under 65 years who suffered a TCA managed by a mobile medical team. Patients were categorized as children (<18 years) and adults (18–65 years). The primary endpoint was the 30-day survival, and secondary endpoints were: return of spontaneous circulation (ROSC), survival at hospital admission and survival with a favourable neurological outcome (Cerebral Performance Categories 1–2) at 30 days.

Results

Among 5,030 included patients, 396 were children (median age 13 [IQR 4–16] years; 73.2% male) and 4,634 were adults (median age 39 [IQR 27–51] years; 80.4% male). Paediatric patients had significantly higher rates of ROSC (25.5% vs. 20.6%, p = 0.02), survival to hospital admission (21.2% vs. 14.7%, p < 0.001), and 30-day survival (3.5% vs. 1.6%, p < 0.01). However, the proportion of patients achieving a favourable neurological outcome at 30 days did not differ significantly between groups (0.8% vs. 0.9%, p = 0.80).

Conclusions

Paediatric patients with out-of-hospital TCA demonstrate higher rates of ROSC and survival compared to adults, although neurological outcomes remain poor in both populations. These findings underscore age-related disparities in TCA prognosis and highlight the need for age-specific research in TCA patients.

Introduction

The incidence of out-of-hospital cardiac arrest (OHCA) is estimated at 56 per 100,000 person-years¹ In adults, the majority of OHCA cases are of medical origin, primarily due to cardiac causes, with only 3.9% attributed to trauma.¹ Conversely, in the patients aged 1 to 17 years, trauma is the leading cause of OHCA,² accounting for 21% of OHCA in this population.³

Despite extensive resuscitative efforts, the mortality rate for traumatic cardiac arrest (TCA) in adult patients exceeds 97%, with only one-third of the few survivors achieving a favourable neurological outcome.⁴ A meta-analysis published in 2023 estimated a comparable survival rate of 1.2% in a paediatric population.⁵ However, a systematic review published in 2012 suggested that children might have a higher survival rate but tend to have a poorer neurological outcome at discharge compared to adults.⁶ Nevertheless, this review is predominantly based on case series and studies with a low level of evidence. A cohort study published in 2013 reported comparable findings, with a survival rate of 23.1% in children, 5.7% in adults, and 3.7% in the elderly.⁷ Therefore, whether the prognosis of TCA in children differs from that of the adults remains uncertain.

The National Association of Emergency Medical Support Physicians and the American College of Surgeons Committee on Trauma released guidelines focusing on the withholding and termination of resuscitation in TCA.8, 9 Whether these guidelines apply to children or not remains uncertain, as most of evidence is based on studies involving only adult participants.¹⁰ In 2014, a joint position statement from several medical organizations, including emergency, trauma and paediatric experts, highlighted the need for comprehensive research on paediatric TCA, focusing on long-term neurological and functional outcomes.¹¹

The main objective of this study was to assess differences in outcomes between children and adults following a TCA. The secondary objective was to evaluate whether these outcomes varied according to the age group.

Methods

Study design

We conducted a multicentre retrospective cohort study from patients prospectively enrolled in the French nationwide cardiac arrest registry (Registre électronique des Arrêts Cardiaques, RéAC). The results are reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines.¹²

Study setting

The French prehospital emergency medical system (EMS) operates on a two-tiered model.¹³ In cases of OHCA, both a fire brigade (FB) ambulance and a mobile medical team (MMT) are dispatched simultaneously to the scene. A bystander may initiate cardiopulmonary resuscitation (CPR) and, if available, defibrillation using a public-access automated external defibrillator (AED) before the arrival of EMS often under the guidance of the dispatcher. The FB typically arrives first, takes over from any bystander, and continues basic life support (BLS). They are also trained to perform trauma-specific interventions such as external haemorrhage control, spinal motion restriction, or pelvic stabilization. The MMT consists of an emergency physician, a nurse, and an ambulance driver and can be dispatched via either ground or helicopter ambulance. They provide advances life support (ALS), including, among other, advanced airway management, intravenous or intraosseous access, administration of vasopressors, fluid resuscitation, blood transfusion and when necessary, surgical procedures such as thoracostomy at the scene.

Data source and collection

Data were extracted from the French nationwide cardiac arrest registry (Registre électronique des Arrêts Cardiaques, RéAC). The majority (87%, n = 305) of the 350 French prehospital MMTs actively participate in the registry, covering a total population of approximately 60 million inhabitants across 82 of the 96 departments in mainland France. The procedure for data collection and management have been described previously.¹⁴ In summary, participating MMTs use a dedicated case report form based on the Utstein-style resuscitation registry’s template to record a broad range of variables, including patients’ sociodemographic data and comorbidities, OHCA aetiology, initial presentation (rhythm, signs of life, etc.), bystander response, BLS and ALS descriptions, as well as all relevant time points and time intervals. A 30-day in-hospital follow-up is conducted to collect patient outcomes. Subsequently, data are logged in the secure RéAC registry (https://www.registreac.org). Several automatic quality controls (in- and off-line) are performed, and a clinical research associate oversees the identification and resolution of any inconsistencies throughout the entire registry.

Population

All OHCA patients recorded prospectively in the RéAC registry between July 2011 and March 2023 were screened. We included patients who experienced a TCA as per the Utstein-style definition¹⁵ and for whom a MMT was dispatched. We excluded patients if they met one or more of the following criteria: aetiology other than traumatic (medical, asphyxia, overdose, electrocution, drowning) or unknown, absence of MMT management, signs of irreversible death upon MMT arrival (i.e., decay, rigor mortis, fixed lividity), end-of-life patients or those with “do not attempt resuscitation” (DNAR) orders and age ≥ 65 years.

Patients were classified into two age groups: children (0–17 years) and adults (18–64 years). An upper age limit of 65 years was chosen to minimize the impact of comorbidities and the use of certain treatments (e.g., anticoagulants), as these factors could significantly affect survival following TCA, thus making comparisons between adults and children less relevant.

Additionally, a subgroup analysis was conducted with children further divided into ≤ 5 years, 6–11 years and 12–17 years groups according to the Utstein-style recommendations.¹⁶

Outcomes

The primary outcome was survival at 30 days or at ICU discharge, depending on which event occurred first (D30). Secondary outcomes included return of spontaneous circulation (ROSC) at the scene, survival at hospital admission (D0), and survival with a favorable neurological outcome at D30, defined as a cerebral performance category (CPC) of 1 or 2.¹⁷

Statistical analysis

Categorical variables are presented as frequencies (percentages) and were compared using Pearson’s chi-square test or Fisher’s exact test, as appropriate. Continuous variables are presented as medians [interquartile ranges, (IQRs)] and were compared using the Wilcoxon rank-sum test or the Kruskal-Wallis test. Odds ratios (ORs) and 95% confidence intervals (95% CIs) for ROSC, survival at D0, survival at D30, and survival with a favorable neurological outcome at D30 were calculated by univariate logistic regression using a generalized linear model with a binomial distribution. Missing data were not imputed or estimated. A significance level of p < 0.05 was set. All the statistical analyses were performed using R® version 4.4.0 (R Core Team, 2024).

Results

During the study period, a total of 151,658 patients sustained an OHCA and were included in the registry. Among them, 5,030 TCA cases were included in the study after applying the exclusion criteria: 396 in the child group (7.9%) and 4,634 in the adult group (92.1%) (Fig. 1).

Fig. 1.

Flow-Chart. RéAC: Registre électronique des Arrêts Cardiaques; DNAR: do not attempt resuscitation.

Baseline characteristics

The median age was 13 [4;16] years in the child group and 39 [27;51] years in the adult group. There were more males in both groups (children 73.2% vs. adults 80.4%), with a higher sex ratio in adults compared to children (p < 0.001). There was no difference in the type of trauma (p = 0.31) with blunt injuries being more common (81.3% vs. 77.9%) than penetrating injuries (18.2% vs. 21.4%). However, trauma mechanisms significantly differed between groups (p < 0.001): children were more frequently involved in road traffic accidents (43.2% vs. 35.2%) but were less likely to experience firearm injuries (3.3% vs. 7.7%) or falls (12.3% vs. 19.8%). The head was the most frequently affected area in children (81.6% vs. 76.1%, p = 0.02), while adults had a higher incidence of limb injuries (17.3% vs. 23.5%, p < 0.01). The proportion of active haemorrhage identified by the on-scene physician was similar in both groups (22.2% vs. 23.5%, p = 0.56). TCA were more frequently witnessed in children (85.6% vs 79.3%, p < 0.01), and CPR was more frequently initiated by bystanders (40.1% vs. 33.4%, p < 0.01). EMS response times did not significantly differ between the two groups (FB: 6 [1;10] vs. 7 [2;11] minutes, p = 0.22; MMT: 15 [8;21] vs 15 [9;21] minutes, p = 0.87). There was no difference in the presence of signs of life (6.3% vs. 5.8%, p = 0.67) or initial rhythm (p = 0.79), with a high prevalence of asystole (87.1% vs. 87.9%) and PEA (9.9% vs. 9.3%) upon MMT arrival. The low-flow duration was longer in children (40 [26–51] vs. 35 [25–47] minutes, p < 0.01), with no difference in the no-flow duration (0 [0;5] minutes in both groups, p = 0.44). Management did not significantly differ in intubation rate (94.7% vs. 94.2%, p = 0.68), epinephrine administration (96.5% vs. 96.3%, p = 0.90), fluid resuscitation (43.9% vs. 38.9%, p = 0.05), blood transfusion (7.3 vs. 5.2%, p = 0.08), and external haemorrhage control (9.1% vs. 10%, p = 0.57). A thoracostomy was performed more frequently in adults than in children (18.4% vs. 23.5%, p = 0.02). (Table 1).

Table 1.

Baseline characteristics.

Baseline characteristics	Children (n = 396)	Adults (n = 4,634)	Missing	p-value
Age, years	13 [4–16]	39 [27–51]	0 (0%)	NA
Sex			2 (0%)	<0.001
Male	290 (73.2%)	3,722 (80.4%)
Female	106 (26.8%)	910 (19.6%)
Location			307 (6.1%)	<0.001
Public area	287 (78.6%)	3,264 (74.9%)
Home/Private place	64 (17.5%)	727 (16.7%)
Work	2 (0.5%)	244 (5.6%)
Nursing centre	2 (0.5%)	19 (0.4%)
Other	10 (2.7%)	104 (2.4%)
Type of trauma			0 (0%)	0.31
Blunt	322 (81.3%)	3,609 (77.9%)
Penetrating	72 (18.2%)	992 (21.4%)
Burn	2 (0.5%)	33 (0.7%)
Mechanism of trauma			83 (1.7%)	<0.001
Firearm	13 (3.3%)	352 (7.7%)
Stab	28 (7.2%)	352 (7.7%)
Motor vehicle accident	169 (43.2%)	1,604 (35.2%)
Fall	48 (12.3%)	901 (19.8%)
Other	133 (34.0%)	1,347 (29.6%)
Anatomic site of injury†
Head	293 (81.6%)	3,168 (76.1%)	508 (10.1%)	0.02
Thoracic	200 (55.7%)	2,495 (59.9%)	508 (10.1%)	0.12
Abdominal	143 (39.8%)	1,553 (37.1%)	508 (10.1%)	0.34
Limbs	62 (17.3%)	975 (23.4%)	508 (10.1%)	<0.01
Haemorrhage	88 (22.2%)	1,090 (23.5%)	0 (0%)	0.56
Witnessed			0 (0%)	<0.01
Bystander	270 (68.2%)	2,745 (59.4%)
EMS (FB or MMT)	69 (17.4%)	919 (19.9%)
Unwitnessed	57 (14.4%)	960 (20.8%)
Bystander CPR	154 (40.1%)	1,488 (33.4%)	188 (3.7%)	<0.01
Bystander AED	2 (0.6%)	15 (0.4%)	570 (11.3%)	0.39
FB response time, minutes	7 [2–11]	6 [1–10]	2,346 (46.6%)	0.22
MMT response time, minutes	15 [9–21]	15 [8–21]	1,060 (21.1%)	0.87
Sign of life upon MMT arrival	25 (6.3%)	268 (5.8%)	293 (5.8%)	0.67
Rhythm upon MMT arrival			339 (6.7%)	0.79
Asystole	318 (87.1%)	3,801 (87.9%)
PEA	36 (9.9%)	404 (9.3%)
VF/VT	6 (1.6%)	80 (1.8%)
Spontaneous circulation	5 (1.4%)	41 (0.9%)
No-flow duration, minutes	0 [0–5]	0 [0–5]	1,481 (29.4%)	0.44
Low-flow duration, minutes	40 [26–51]	35 [25–47]	1,854 (36.9%)	<0.01
Interventions
Intubation	375 (94.7%)	4,332 (94.2%)	0 (0%)	0.68
Epinephrine administration	382 (96.5%)	4,456 (96.3%)	9 (0.2%)	0.90
Fluid replacement	174 (43.9%)	1,802 (38.9%)	0 (0%)	0.05
Blood transfusion	29 (7.3%)	243 (5.2%)	0 (0%)	0.08
External haemorrhage control	36 (9.1%)	463 (10.0%)	0 (0%)	0.57
Thoracostomy (needle or finger)	73 (18.4%)	1,089 (23.5%)	0 (0%)	0.02

^†the same patient may have multiple anatomical sites of injury; EMS: Emergency medical service; FB: fire brigade; MMT: mobile medical team; CPR: cardiopulmonary resuscitation; AED: automatic external defibrillator; PEA: pulseless electrical activity; VF: ventricular fibrillation; VT: ventricular tachycardia.

Comparison of survival outcomes

The proportion of patients with an on-scene ROSC was higher in the child group (25.5% vs. 20.6%, p = 0.02). Similarly, children had a higher rate of survival at D0 (21.2% vs. 14.7%, p < 0.001) and of survival at D30 (3.5% vs. 1.6%, p < 0.01). However, no significant difference was observed in survival with a favourable neurological outcome at D30 (0.8% vs. 0.9%, p = 0.80) (Fig. 2).

Fig. 2.

Comparison of survival outcomes between adults and children. OR: odds ratio; CI 95%: 95% confidence interval; ROSC: return of spontaneous circulation; survival D0: survival at hospital admission; survival D30: survival at 30 days; CPC: cerebral performance category.

Subgroup analyses

Among the 396 children, 118 (30%) were ≤ 5 years old, 66 (17%) were 6–11 years old and 212 (54%) were 12–17 years old. Baseline characteristics between subgroups of age in children are presented in supplementary appendix. (Supplementary Table 1) Children 6–11 years were associated with a higher likelihood of ROSC (OR 2.68 [1.61;4.38], survival at D0 (OR 3.12 [1.83;5.16]) and at D30 (OR 4.92 [1.68;11.47]) compared to adults. Children ≤ 5 years were also associated with a higher rate of survival at D0 (OR 1.81 [1.15;2.76]). Conversely, in adolescents (12–17 years), no differences were observed in any outcomes when compared to adults: ROSC (OR 0.96 [0.66;1.34]), survival at D0 (OR 1.07 [0.72;1.55]) and survival at D30 (OR 1.15 [0.34;2.81]). Survival with a favorable neurological outcome could not be compared in subgroups due to the low number of patients involved (Fig. 3).

Fig. 3.

Comparison of survival outcomes between adults and children by subgroups of age. OR: odds ratio; CI 95%: 95% confidence interval; ROSC: return of spontaneous circulation; survival D0: survival at hospital admission; survival D30: survival at 30 days.

Discussion

This registry-based nationwide cohort study found that children who suffered TCA were associated with higher chance of survival at hospital admission, survival at hospital discharge and survival with favourable neurological outcome at hospital discharge compared to adults.

The characteristics of TCA observed in this study are consistent with previous literature, with a predominance of male patients.⁶ Key prognostic factors described in the literature, such as type of trauma, initial cardiac rhythm, and presence of signs of life were observed at similar frequencies in both adults and children.18, 19 Notably, head trauma was more prevalent in children than in adults, likely due to the larger head-to-body size ratio in children, increasing the likelihood of head injuries.²⁰ Children were more likely to have their TCA witnessed by a bystander and to receive early CPR, whereas TCA in adults more frequently occurred in the presence of EMS personnel, resulting in similar no-flow duration across both groups. However, the low-flow duration was 5-min longer in children compared to adults, which may reflect ethical concerns and the lack of consensus regarding termination-of-resuscitation protocols for paediatric TCA.10, 11

Current guidelines emphasize prioritizing the treatment of reversible causes before initiating CPR to optimize outcomes in TCA.²¹ In this study, ALS interventions such as endotracheal intubation and epinephrine administration were frequently performed, while trauma-specific interventions such as external haemorrhage control, fluid administration, blood transfusion, and needle or finger thoracostomy were underutilized in both child and adult populations.

The observed survival rate to hospital discharge after TCA in children (3.5%) exceeded the 1.2% [0.1%-3.1%] reported in the meta-analysis by Alqudha et al.⁵ but was consistent with more recent cohort studies reporting 30-day survival rates ranging from 3.3% to 3.7%.22, 23, 24 Conversely, the adult survival rate in this study (1.6%) was lower than the 2.8% [2.0%-3.7%] documented in the meta-analysis by Vianen et al.,⁴ likely reflecting differences in the definition of TCA and variations in inclusion/exclusion criteria.

The survival rate in children was therefore twice as high as in adults, which is consistent with the systematic review by Zwingman et al. Interestingly, a better survival was only observed in pre-pubertal children (<12 years) with no significant differences between adolescents and adults. This finding aligns with Shibahashi et al., who noted a significantly lower median age among survivors (8 years) compared to those who died (16 years).²³ However, it contrasts with the results of Alqudah et al., who reported higher survival rates among patients aged 12–16 years (8.8%) compared to those aged 5–11 years (1.9%) or ≤ 4 years (0%).²²

Several factors may contribute to the difference in prognosis between children and adult TCA patients. First, as noted by Donoghue et al., some patients receiving CPR in out-of-hospital settings may not be in cardiac arrest, particularly those presenting with PEA.³ Accurate identification of pulselessness is particularly challenging in children, especially in cases of circulatory shock.²⁵ Second, respiratory causes of arrest, which often respond favourably to ventilation,²⁶ are more common in children than in adults, whose arrests are more frequently associated with hypovolemia, tension pneumothorax, or cardiac tamponade—factors associated with poorer prognosis.²⁷ Indeed, children have a shorter tolerance to hypoxia but exhibit a stronger compensatory response to haemorrhage.²⁸ Their flexible thoracic structures also reduce the likelihood of tension pneumothorax and cardiac tamponade in cases of blunt trauma.²⁹ Third, the effectiveness of external chest compressions may be greater in children due to their more compliant chest walls, an argument previously used to explain better survival in paediatric non-TCA.³⁰ While chest compressions are often considered ineffective in cases of hypovolemia or cardiac tamponade,31, 32, 33 they have been shown to improve prognosis in hypoxic cardiac arrests, which are more common in paediatric TCA.³⁴ Lastly, the lack of consideration for comorbidities, which are generally more prevalent in adults, may partly explain the differences in outcomes.³⁵ The RéAC registry includes comorbidity data but their low completion rate precluded their use as covariates in this study. This limitation is consistent with prehospital settings, where patient comorbidities are often unknown due to patient unresponsiveness and the distress of relatives.³⁶ To minimize the prevalence of pre-existing conditions, we chose to restrict our population to patients under 65 years of age.

In contrast, no difference in neurological outcome were observed between children and adults. A possible explanation could be the higher incidence of head trauma in children compared to adults. Indeed, traumatic brain injury (TBI) is a well-recognized negative prognostic factor in TCA.37, 38 Another potential explanation could be the prolonged resuscitation in children, which could lead to higher survival but more extensive brain damage.³⁹ Finally, the absence of difference in neurological outcomes might be attributed to a lack of statistical power due to the very small number of neurologically intact children in our study.

Strengths and limitations

This study is the first cohort study focused on comparing outcomes between children and adults following TCA. To the best of our knowledge, it also represents the largest cohort of pediatric TCA cases within a physician-led prehospital care system. The RéAC registry, unlike many others, adheres to the Utstein universal template for reporting OHCA, enabling standardized and specific metrics for evaluation.⁴⁰ Additionally, it includes patients who were declared dead on-scene, allowing for a comprehensive analysis of all OHCA cases. This feature contrasts with studies that only include patients transported to the hospital, which can introduce selection bias and overestimate survival rates.⁴¹

However, the study has several limitations. The proportion of missing data in the RéAC registry is high but comparable to that observed in other OHCA registries.⁴⁰ The RéAC registry also lacks data on trauma severity (e.g., injury severity score), which hinders the ability to draw direct comparisons between child and adult populations. Additionally, there is no information regarding in-hospital management, including intensive care or surgical interventions, which could have impacted the outcome at discharge. Finally, this database is specific to France, and the results may not be extrapolated to other prehospital systems particularly those where patients cannot be declared dead on-scene.

Conclusions

Pediatric patients with out-of-hospital TCA demonstrate higher rates of ROSC and survival compared to adults, although neurological outcomes remain poor in both populations. These findings underscore age-related disparities in TCA prognosis and highlight the need for age-specific research in TCA patients.

Ethical approval

The registry received approval from the French national data protection commission (Commission Nationale de l’Informatique et des Libertés, CNIL) [# 910946] and approval from the French advisory committee on information processing in material research in the field of health (Comité Consultatif sur le Traitement de l’Information en Matière de Recherche dans le Domaine de la Santé) [#10.326Ter]. In line with the French legislation on retrospective studies of registry data, patient consent was not required.

Sources of funding

The RéAC registry is supported by the Société Française de Médecine d’Urgence (SFMU), the Fédération Française de Cardiologie (FFC), the Mutuelle Générale de l'Education Nationale (MGEN), the university of Lille, and the institute of health engineering of Lille. The authors declare that the sources of funding had no role in the conduct, analysis, interpretation, or writing of this manuscript.

CRediT authorship contribution statement

Baptiste Morcel: Writing – original draft, Methodology, Formal analysis, Data curation, Conceptualization. Eric Mercier: Writing – review & editing. Guillaume Debaty: Writing – review & editing. Jean-Stéphane David: Writing – original draft. Etienne Javouhey: Writing – original draft. Valentine Baert: Writing – review & editing, Funding acquisition, Data curation. Amaury Gossiome: Writing – review & editing. Francis Desmeules: Writing – review & editing. Alexis Cournoyer: Writing – review & editing. Karim Tazarourte: Writing – review & editing. Axel Benhamed: Writing – original draft, Validation, Supervision, Methodology, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

The following are the Supplementary data to this article: