Characteristics, management, and outcomes of perioperative cardiac arrest in children undergoing noncardiac surgery procedures: a scoping review

Graphical abstract

Graphical Abstract

Abstract

Introduction

The primary goal of this scoping review was to create a comprehensive synthesis of the characteristics, management, and outcomes of perioperative CA events in children undergoing noncardiac surgery.

Methods

We conducted electronic searches of PubMed, Embase, and Scopus in February 2024, with an update in June 2025. Inclusion and exclusion criteria were defined a priori. We included studies in children (age <18 years) undergoing noncardiac surgery, experiencing CA while under the care of anesthesia personnel or in immediate postoperative phases, with at least one pre-defined survival, neurofunctional, quality of life, or healthcare utilization outcome ascertained. Two authors reviewed citations independently with a third author solving conflicts.

Results

Of 801 unique citations, 18 studies met inclusion criteria. There were 17 retrospective cohort studies, 1 cross sectional study, and no interventional studies. Median number of CA events per study was 21 (interquartile range, 13–27). Survival to hospital discharge ranged from 30% to 91% overall, with survival of 46% in the U.S.-based study. Younger age, higher American Society of Anesthesiologists score, and emergent procedures were associated with higher CA event occurrence. No study reported data on CA management, including quality of resuscitation, medication use, or defibrillation use. No studies obtained neurofunctional outcomes using standardized measures. No studies addressed long-term survival, neurofunctional, healthcare utilization, or quality of life outcomes after hospital discharge.

Conclusion

Studies on pediatric perioperative CA are heterogenous, with modest sample sizes, and variability in definitions and reporting of outcomes. A large knowledge gap remains in CA management, neurologic consequences, and long-term outcomes.

Introduction

Pediatric cardiac arrest (CA) is the most serious and feared adverse event in the perioperative period. The incidence of perioperative CA in children for all procedure types has been estimated at 2.9–25.4 per 10,000 anesthetics globally.1, 2, 3 Despite a significant evolution in the delivery of pediatric anesthesia, perioperative CA continues to occur and represents a critical point of failure with potentially fatal consequences.

Recent systematic reviews have evaluated the global incidence and mortality in pediatric perioperative CA, helping to better define the epidemiology of these events.2, 3 However, knowledge of the incidence of CA does not adequately prepare anesthesiologists for recognizing which patients are at highest risk for an arrest and for mortality afterwards, and to know how to optimally manage CA when it occurs. Furthermore, many studies do not distinguish between CA in children undergoing noncardiac versus cardiac surgeries. Pediatric cardiac surgery patients have different physiologies, higher incidence of CA, increased odds of death after CA, and different provider teams, and, thus, CA in cardiac surgery patients deserve recognition as a separate entity from those in children undergoing noncardiac surgery.4, 5, 6, 7 A comprehensive synthesis of the characteristics, management, and outcomes of CA events in children undergoing noncardiac surgery is currently lacking.

The primary objective of this scoping review was to summarize the literature on outcomes related to perioperative CA in children undergoing noncardiac surgery. Secondary goals include describing characteristics of the patients and events, and illustrating the current literature on how they are being managed. Synthesis will allow us to define what is known about the nature of perioperative CA and characterize the remaining gaps, setting a foundation for future research and, ultimately, interventions to improve these outcomes.

Methods

A protocol was designed (Supplementary File S1) and electronic searches of PubMed, Embase, and Scopus were conducted February 28, 2024, with an update on June 26, 2025. A combination of medical subject heading terms was used to capture concepts of perioperative CA in children with noncardiac surgery, as detailed in Supplementary File S1. No limits on year of publication were imposed. Studies were included if they met the following criteria: the study population consisted of children <18 years of age or was a mixed pediatric and adult population with the ability to separate data for children, the population underwent noncardiac surgical procedures or was a mixed cardiac and noncardiac surgery population with the ability to separate data for patients who underwent noncardiac surgery, subjects experienced CA in the perioperative period (defined as during care by anesthesia personnel or in the immediate postoperative phase), and the study evaluated at least one outcome of interest. Patients with heart disease undergoing noncardiac surgery were included. Outcomes of interest included return of spontaneous circulation, return of circulation with use of extracorporeal membrane oxygenation (ECMO), post-arrest ECMO utilization, survival to 24 h, survival to hospital discharge, survival to other time point, length of mechanical ventilation, length of intensive care unit stay, length of hospital stay, new or progressive multiple organ dysfunction, post-arrest functional status assessment, neurocognitive functional status, health related quality of life, or new morbidity or technology dependence after the event. Articles published in languages other than English for which eligibility could not be determined were excluded. Studies were excluded if they reported CA incidence without an outcome of interest, used an animal-only design, lacked original data, were a case report or series, or had a sample size of fewer than 10 pediatric CA cases. If multiple studies reported data from the same population during the same study period, only the study presenting the most comprehensive outcome information was extracted. Reporting was done following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for scoping reviews (Supplementary File S1).⁸

Identified citations were imported into the Covidence reference manager (Veritas Health Innovation, Melbourne, VIC, Australia) and duplicates were removed. Two investigators (A.S.N. and L.H.) reviewed all citations independently. Conflicts were resolved by a third reviewer (M.M.B). Template evidence tables were created prior to extraction with fields to capture study characteristics, patient characteristics, procedure characteristics, event characteristics, incidence, and outcomes. Extraction was performed by one investigator (A.S.N.) and reviewed by a second investigator (L.H.). Evidence tables were created for included studies (Supplementary File S2). Synthesis entailed summarizing the counts, percentages, and odds ratios with confidence intervals described by individual studies. Tables were created with the counts of articles that addressed different characteristics. No additional analysis was performed on the data as this review was strictly descriptive.

Results

Study flow

The search strategy yielded 867 unique citations, with 707 citations removed based on title and abstract screening and 141 studies removed based on full text review (Fig. 1). Two studies reported data from the same population during the same study period. Only the most recently published of the two, which had the most comprehensive outcome information, was extracted.9, 10 Ultimately, 18 articles were included in the scoping review.6, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26

Fig. 1.

Study search flowchart.

Description of included studies

Characteristics of included studies are summarized in Table 1 and main findings are illustrated in Fig. 2. Of 18 included studies, 16 (89%) were performed in the general pediatric noncardiac surgery population, whereas 2 (11%) were in unique pediatric sub-populations. Specifically, one study was in patients with adverse metabolic or musculoskeletal reactions to anesthesia concerning for malignant hyperthermia²⁵ and the other study was in children undergoing spinal deformity correction.²⁶ In total, there were 17 retrospective cohort studies (94%), 1 cross-sectional study (6%), and no interventional studies. Twelve articles (67%) were single center studies, and the remaining were multicenter studies, 3 (17%) of which utilized established multicenter registries. Dates of publication ranged from 1953 to 2024, with CA event dates ranging from 1924 to 2022. Five studies (28%) included only the intraoperative period, 11 (61%) included the intraoperative and immediate postoperative time (the longest being up to 48 h post-anesthesia), and 2 (11%) did not report their definition of the perioperative period. Six studies (33%) included non-operating room locations in their analyses. The median number of CA events included in these studies was 21 with an interquartile range of 13 to 27. Two articles (11%) had greater than 50 CA events,21, 22 of which only 1 (6%) had greater than 100 CA events included.²¹

Table 1.

Study characteristics.

Source, Year	Funding	Study period, Country, No. of centers, Registry name	Center type	Study design	Sample size of CA	Definition of perioperative	Non-OR locations included
Snyder (1953)11	NR	1924–1952, U.S., Multicenter (2)	NR, Academic	RCS	30	Intraoperative	Not included (1 case of CA during EKG excluded because no anesthetic administered)
Rackow (1961)12	NR	1947–1956, U.S., Single Center	Adult and pediatric, Academic	RCS	23*	Intraoperative	Not included
Greenberg (1962)13 (Overlap of study population with Greenberg 1965 study.)	NR	1953–1960, U.S., Single Center	NR	RCS	17	NR	NR
Greenberg (1965)14 (Overlap of study population with Greenberg 1962 study.)	NR	1957–1962, U.S., Single Center	NR	RCS	14*	NR	Not included
Flick (2007)6	NR	1988–2005, U.S., Single Center	NR, Academic	RCS	26	Entry into OR to PACU discharge or ICU admission	Radiology procedures including MRI and CT, interventional procedures including cardiac catheterization and tracheoesophageal echocardiography
Ahmed (2009)†,10 (Same population and study period as Ahmed 2008. Overlap of study population with Khoso 2021 study.)	NR	1992–2006, Pakistan, Single Center	Adult and pediatric, Academic	RCS	10	Induction to PACU discharge or ICU admission	Radiology procedures
Bharti (2009)15	NR	2003–2008, India, Single Center	Pediatric, Academic	RCS	27	While in continuous care of anesthesia team	Not included
Bunchungmongkol (2009)16	NR	2003–2004, Thailand, Multicenter (20) Registry: Thai Anesthesia Incidents Study	Multiple types	RCS	38*	Intraoperative and within 24 h postoperative	Cardiac catheterization
Gonzalez (2014)17	NR	2005–2010, Brazil, Single Center	Adult and pediatric, Academic	RCS	18*	Intraoperative or PACU	NR
Hohn (2019)18	NR	2008–2016, Germany, Single Center	Adult and pediatric, Academic	RCS	11*	Intraoperative and within 24 h postoperative	Not included
Jansen (2021)19	NR	2008–2018, Germany, Single Center	NR	RCS	18 (18 events in 17 unique patients)	Induction through 60 min after completion of anesthesia; included anesthesia that was induced in ICU	NR
Khoso (2021)20 (Overlap of study population with Ahmed 2009 study.)	NR	1992–2016, Pakistan, Single Center	Adult and pediatric, Academic	RCS	13*	Intraoperative	Not included
Foz (2023)21	NR	2019–2020, U.S., Multicenter (148) Registry: National Surgical Quality Improvement Program Pediatric Database	NR	RCS	269	Intraoperative and during recovering from anesthesia, within 48 h of the procedure	Not included
Rattana-Arpa (2023)22	NR	2014–2019, Thailand, Single Center	Adult and pediatric, Academic	RCS	63	Intraoperative and within 24 h postoperative	Cardiac catheterization and other non-OR anesthesia (exact locations NR)
Nizeyimana (2024)23	NR	2015, Rwanda, Multicenter (3)	Adult and pediatric, Academic	CSS	12 (12 events in 11 unique patients)	Intraoperative and within 24 h postoperative	Not included
Morais (2025)24	NR	2005–2022, Brazil, Single Center	Adult and pediatric, Academic	RCS	27*	Intraoperative	Cardiac catheterization, diagnostic and interventional radiology, radiotherapy, intensive care unit, or other location in hospital requiring care of an anesthetist
Specific subpopulations
Larach (1997)25 Subpopulation: Patient with adverse metabolic or musculoskeletal reaction to anesthesia (reported to malignant hyperthermia registries)	NR	1990–1993, Multi-country, Multicenter Registry: North American Malignant Hyperthermia Registry and Malignant Hyperthermia Association of the United States Hotline	NR	RCS	25	Induction through 24 h after anesthesia	Cardiac catheterization
Menga (2015)26 Subpopulation: Spinal deformity correction	NR	2004–2014, U.S., Multicenter (2)	Pediatric, Academic	RCS	11	Intraoperative	Not included

CA = cardiac arrest; NR = not reported; RCS = retrospective cohort study; CSS = cross-sectional study; OR = operating room; PACU = post anesthesia care unit; ICU = intensive care unit; MRI = magnetic resonance imaging; CT = computed tomography.

^†Same study population and study period as earlier article by Ahmed et al. (2008).⁹ Data was extracted from Ahmed et al. (2009)¹⁰ given the more comprehensive reported outcomes.

^*Original study with cardiac surgery and/or adult patients included. Numbers specific to pediatric non-cardiac patients extracted as able.

Fig. 2.

Illustration of main findings.

Patient characteristics

A summary of patient characteristics is provided in Table 2, with details provided in Table S2 (Supplementary File S2). Age summaries of patients experiencing CA were provided in 16 of the 18 studies. CA occurring in children less than 1 year of age ranged from 26% to 83% of all CA events reported by each study.6, 10, 11, 12, 14, 15, 16, 18, 19, 20, 21, 22, 23, 24 When evaluating the 10 studies that only included cases after 1990, children less than 1 year of age consistently accounted for more than 50% of events (range, 54% to 83%).10, 15, 16, 18, 19, 20, 21, 22, 23, 24 Bharti et al. (n = 27) and Foz et al. (n = 269) demonstrated a significant association between age less than 1 year and the development of CA with adjusted odds ratios (ORs) of 2.6 and 3.1, respectively.15, 21 Among the 6 studies in which survival by age category could be ascertained, survival for children less than 1 year of age ranged from 43% to 67%, compared to 20% to 71% for children older than 1 year.6, 14, 15, 22, 23, 24 One of these studies, by Bharti et al. (n = 27), demonstrated a 2.7 increased adjusted odds of death (95% confidence interval [CI], 1.1–5.9) for children less than 1 year of age as compared to older children,¹⁵ whereas another study by Rattana-Arpa et al. (n = 63) showed no significant association (unadjusted OR, 0.7; 95% CI, 0.3–2.0) between age less than 1 year and mortality after CA.²² A third article by Nizeyimana et al., evaluating the 10 patients with available age information, found an association on univariate analysis between age <1 week and “cardiac arrest and/or death” compared to age >1 year to <5 years (OR 8.1; 95% CI, 1.8–37.0), but did not demonstrate a significant association for any other infant age group.²³

Table 2.

Summary of patient, procedure, and anesthesia characteristics.

Characteristics	Number of articles (n = 18)
Patient characteristics
Age breakdown	16 (89)
Sex	10 (56)
Race/ethnicity	2 (11)
ASA physical status	10 (56)
Procedure characteristics
Emergent status	11 (61)
Timing of procedure	3 (17)
Surgery type	16 (89)
Anesthesia characteristics
Provider training level	4 (22)
Cause of arrest (anesthesia-related vs other)	11 (61)

Data presented as n (%). ASA = American Society of Anesthesiologists.

American Society of Anesthesiologists physical status score (ASA score) was reported in 10 studies. Between 50% to 78% of CA events were reported in patients with ASA scores of 1 to 3.6, 10, 18, 19, 20, 21, 22, 23, 24 In all but one study, an ASA score of 3 was the most common.6, 10, 18, 19, 20, 21, 22, 23 Bharti et al. (n = 27) and Foz et al. (n = 269) found a significant association between ASA scores of 3 or higher and the development of CA in adjusted analysis.15, 21 Survival to hospital discharge in patients with ASA score ≤3 was 50% to 69%, compared to survival to hospital discharge of 0% to 47% among those with ASA score >3.6, 10, 22 In studies exploring the association of ASA scores with mortality after CA, Flick et al. (n = 26) and Bharti et al. (n = 27) each found a significant association between higher ASA score and mortality after CA,6, 15 whereas Rattana-Arpa et al. (n = 63) found no significant association.²² An additional study by Nizeyimana et al. evaluated the 11 patients with available ASA scores and reported no significant association between increasing ASA score and “cardiac arrest and/or death” on univariate analysis.²³

Procedure characteristics

A detailed summary of procedure characteristics is provided in Table S2 (Supplementary File S2) and summarized in Table 2. The emergent status of procedures was specified in 11 studies, with procedures being described as emergent in 27% to 64% of cases.6, 10, 15, 16, 17, 20, 23, 24 Three studies categorized emergent and urgent procedures together, accounting for 37% to 50% of the procedures resulting in CA.19, 21, 22 Bharti et al. (n = 27) demonstrated that patients with an emergent procedure had 3.8 times increased odds of developing CA compared to those with non-emergent procedures (adjusted OR, 3.8; 95% CI, 1.9–7.5).¹⁵ Survival after CA for those with emergent procedures ranged from 0% to 45%, compared to 43% to 75% after elective procedures.6, 10, 15, 22 In unadjusted analyses, Flick et al. (n = 26) found no significant association between emergent status and hospital survival and another by Nizeyimana et al. (n = 12) found no significant association between emergent or urgent status and “cardiac arrest and/or death.”6, 23 Conversely, Bharti et al. (n = 27) and Rattana-Arpa et al. (n = 63) each demonstrated a significant association, with an unadjusted OR of mortality of 3.9 for those with emergent-or-urgent status,²² and adjusted OR of 4.9–5.4 for those with emergent or emergent-or-urgent status, respectively.15, 22

Timing of procedures was described in 3 studies. Ahmed et al. (n = 10) reported that 70% of CA events occurred during routine working hours (8am to 5 pm) and 30% occurred after hours (5 pm to 8am), with survival of 43% in the routine working hours group and 0% in the after-hours group.¹⁰ Jansen et al. (n = 18) found that 67% of CA events occurred between 7 am and 3 pm, 17% between 3 pm and 10 pm, and 17% between 10 pm and 7am.¹⁹ A third study by Rattana-Arpa et al. (n = 63) reported that 14% of CA events occurred “out of hours,” although specific time intervals were not defined.²² No statistical testing was performed to evaluate the association between timing and development of CA and/or mortality after CA.

Anesthesia and anesthesia team characteristics

A summary of anesthesia characteristics is provided in Table 2, with further details in Table S2 (Supplementary File S2). Provider training level was reported in 4 studies. Jansen et al. (n = 18) simply stated that all providers were specialists in anesthesia with “clinical expertise in pediatric anesthesia.”¹⁹ Studies by Nizeyimana et al. (n = 12) and Ahmed et al. (n = 10) reported that consultant anesthesiologists were present in 45% to 70% of the CA cases, respectively, but neither report pediatric subspecialty trained anesthesiologist presence.10, 23 In the remaining article by Rattana-Arpa et al. (n = 63), subspecialty-trained providers were involved in 62% of the CA events and anesthesiologists with >15 years of experience were involved in 30% of the CA events.²² Among Nizeyimana et al.’s intraoperative CA events (n = 6), 100% resulted in survival when there was a consultant in the room vs 50% survival when personnel other than a consultant was in the room.²³

The role of anesthesia in the cause of the arrest was delineated by 11 studies, although the specific definitions used to deem a CA event to be anesthesia-related were variable. Rackow et al. only included CA events for which anesthesia was the cause or the cause was unknown.¹² In the remaining articles, responsibility for 17% to 73% of CA events was deemed to be associated with anesthesia, whether wholly anesthesia-attributable or partially anesthesia-related.6, 10, 15, 16, 17, 18, 19, 20, 22, 24 Survival after anesthesia-associated CA ranged from 45% to 100% compared to survival after non-anesthesia-associated CA that ranged from 0% to 69%, with these survival outcomes ascertained at hospital discharge (n = 4), at 24 h post-anesthesia (n = 1), and unspecified (n = 1).6, 10, 16, 17, 22, 24

Cardiac arrest management

A summary of CA management is provided in Table 3, with further details provided in Table S2 (Supplementary File S2). Six articles reported details of CA characteristics and CA management. However, two of these articles were published in the mid-1900s and CA management referred to whether cardiac massage or “pounding over the precordium” was used, which is reflective of earlier CA management principles. Here we focus on the four recent articles.11, 13

Table 3.

Summary of cardiac arrest management features.

Cardiacarrest management features	Number of articles (n = 18)
Presentation
Initial rhythm	2 (11)
CPR characteristics
CPR duration	2 (11)
CPR quality metrics	0 (0)
Interventions
Medications administered	0 (0)
Defibrillation use	0 (0)
ECPR/ECMO use	1 (6)

Data presented as n (%). CPR = cardiopulmonary resuscitation; ECPR = extracorporeal cardiopulmonary resuscitation; ECMO = extracorporeal membrane oxygenation

Two articles described the first documented rhythm, with a reported 67% to 82% of CA events starting with bradycardia, 33% starting with pulseless electrical activity, and 18% starting with asystole.16, 19 Two articles described cardiopulmonary resuscitation (CPR) duration. Rattana-Arpa et al. (n = 63), describing a general pediatric noncardiac surgery population, reported a median CPR duration of 6 min (interquartile range, 2–15.3 min).²² In this study, each additional minute of CPR was significantly associated with in-hospital mortality after CA in both univariate (OR 1.07; 95% CI, 1.02–1.13) and multivariate analyses (adjusted OR 1.07; 95% CI, 1.02–1.12).²² The other article by Larach et al. (n = 25), which focused on the sub-population of patients with concerns for malignant hyperthermia, reported a median CPR duration of 42 min (range, 10–296 min).²⁵ No article described the quality of CPR being provided, such as rate of compressions, depth of compressions, length of pauses in CPR, or any other CPR quality metrics. None of the articles explicitly reported the medications used during the event nor if and when defibrillation was implemented.

One article by Rattana-Arpa et al. (n = 63) reported use of extracorporeal cardiopulmonary resuscitation (ECPR), with this modality used in 6% of all CA events.²² Another article by Larach et al. (n = 25) reported cardiopulmonary bypass being utilized in 8% of CA events.²⁵ Survival outcomes for these cases were not reported.

Outcomes

A summary of CA event outcomes is provided in Table 4, with details provided in Table S3 (Supplementary File S2). All 18 studies included at least one survival (or mortality) measure. The majority utilized survival as the outcome measure, so for the purpose of this synthesis, those reporting mortality were re-adjusted to reflect survival. Survival to hospital discharge of 30% to 91% was reported in 5 articles, all with data from 1980s to present.6, 10, 18, 22, 24 Only one of these studies, by Flick et al., was conducted in the United States, with survival to hospital discharge of 46%.⁶ Return of spontaneous circulation was reported in 3 articles, and was attained in 50% to 89% of CA events.10, 15, 19 Other survival endpoints that were utilized include survival to 6 h (1 study),¹⁵ survival to 24 h (2 studies),16, 22 survival to 48 h (2 studies),15, 20 and survival to 30 days (3 studies).19, 21, 24 Eight of the 18 articles reported on survival without specifying the endpoint of that survival.11, 12, 13, 14, 17, 23, 25, 26

Table 4.

Summary of reported outcomes after cardiac arrest.

Outcome type	Number of articles (n = 18)
Survival
Return of spontaneous circulation	3 (17)
Survival to hospital discharge	5 (28)
Survival endpoint not specified	8 (44)
Length of hospital stay	2 (11)
Discharge destination	1 (6)
Neurologic outcomes
Subjective neurologic outcomes	4 (22)
Objective neurologic outcomes	0 (0)
Long-term outcomes
Neurobehavioral functioning	0 (0)
Healthcare utilization	0 (0)
Quality of life	0 (0)

Data presented as n (%).

Two studies, by Foz et al. and Jansen et al., reported hospital length of stay, with 29% to 78% of patients having a hospital length of stay ≥30 days, respectively.19, 21 The median hospital length of stay was 7 days (interquartile range, 2–23 days).²¹ No articles reported on mechanical ventilation needs, new or progressive multiple organ dysfunction, or length of intensive care unit stay. One article by Foz et al. (n = 269) reported discharge destination, with 32% dying prior to discharge, 36% discharged to home, 1% discharged to rehabilitation, 1% discharged to skilled care or unskilled facility, and 29% still in the hospital at 30 days.²¹

Neurologic outcomes were described in 4 studies (22%), but without using standardized scales. Neurologic outcomes were reported as “central nervous system injury” in 43% of survivors by Greenberg et al.,¹³ “hypoxic brain damage” in 13% of survivors by Khoso et al.,²⁰ “hypoxic ischemic encephalopathy” in 8% of survivors by Rattana-Arpa et al.,²² and a “vegetative state” in 7% of survivors by Larach et al.²⁵ None of these articles defined how these diagnoses were established or the specific neurologic deficits or technology dependence that each patient had.

No article reported long-term outcomes, such as long-term survival, neurobehavioral function, healthcare utilization, or quality of life after recovery from the acute period.

Discussion

This scoping review synthesized the literature on pediatric perioperative CA for children undergoing noncardiac surgery, identifying an array of characteristics that have been studied, as well as the limited information that is known about management during and outcomes after these events. These studies were mostly single center retrospective cohort studies with modest sample sizes (median 21 CA events; interquartile range 13–27). Although research in this area includes events as far back as 1924, 14 of the 18 studies were published after 2000 and report CA events no earlier than 1988.

A number of studies were excluded from this review due to inability to separate cardiac surgery from noncardiac surgery procedures, including three studies published between 2000 and 2010 using the Pediatric Perioperative Cardiac Arrest registry.27, 28, 29 Although it reduced the number of included studies, it was important to understand characteristics, management, and outcomes specific to noncardiac surgery patients. Studies directly comparing pediatric cardiac surgery to noncardiac surgery have shown that the incidence of perioperative CA in the cardiac surgery cohort ranges from 21.9 to 127.3 per 10,000 compared to 1.8 to 4.4 per 10,000 in the noncardiac surgery cohort.5, 6, 7 Furthermore, Flick et al. demonstrated a wide discrepancy in survival among those undergoing cardiac surgery (9%) vs noncardiac surgery (46%), estimating that the odds of death after CA associated with cardiac surgery were 42 times higher (95% CI, 22–80) than the odds of death after CA associated with noncardiac surgery in an adjusted analysis.⁶ Children undergoing cardiac surgery have different physiology and pathogenesis of CA and, as such, likely have different risk factors. Furthermore, they are generally cared for by a specialized perioperative team, separate from the teams that manage patients in the general operating rooms. Here we aim to understand the unique phenotype of children undergoing noncardiac surgery and, thus, excluded articles where that could not be ascertained.

Patient characteristics represent a better described area of perioperative CA. Studies with events after 1990 demonstrate that more than half of all CA events occurred in infants, suggesting this may be a risk factor for CA. This is supported by multivariable logistic regression analyses performed by Foz et al. and Bharti et al. that showed significantly increased odds of development of CA in children less than 1 year of age vs older ages.15, 21 The effect of age on survival seems less clear, however, demonstrating a wide variability in survival among those below vs above 1 year of age. Multivariable analysis by Bharti et al. suggests that age <1 year is significantly associated with mortality (adjusted OR 2.74, 95% CI 1.1–5.9), whereas univariate analysis by Rattana-Arpa et al. showed no significant association of age <1 year with mortality (OR 0.7, 95% CI 0.3–2.0).15, 22 Further studies are needed to better delineate the relationship between age and mortality among children undergoing noncardiac surgery.

ASA scores, used as a surrogate of preoperative health status, were another commonly described characteristic in these studies. Increasing ASA scores were a significant risk factor for CA.15, 21 Higher ASA scores also suggest a trend towards decreased survival (survival of 0–47% for ASA >3 vs 50–69% for ASA ≤3), which was further evidenced by Flick et al. (n = 26) and Bharti et al. (n = 27) who demonstrated a significant association between higher ASA score and mortality after CA.6, 10, 15, 22 Although this association was not demonstrated by Rattana-Arpa et al. (n = 63),²² studies with mixed populations of cardiac and noncardiac surgery patients and mixed populations of children and adults have also demonstrated this association between higher ASA score and mortality.4, 24, 27, 28, 30 It is notable that ASA scores of 1 through 3 made up 50% to 78% of CA event cases.6, 10, 18, 19, 20, 21, 22, 23, 24 This suggests that half or more of children were not ill enough to be a “constant threat to life” or to put them at risk of dying without surgery, but suffered a CA nonetheless.31, 32 However, previously published evidence suggests that ASA scores can be challenging to apply to pediatric patients and there is variability in how anesthesiologists assign scores, which could have affected these results.7, 33, 34, 35

Only 3 studies reported the timing of procedures with CA events and none evaluated the association of timing of procedure with occurrence of CA or survival after CA.10, 19, 22 A study by Zgleszewski et al. that included a mixed cardiac and noncardiac surgery pediatric cohort did not find a significant association between time of day or weekend of the surgical procedure with the development of CA in an unadjusted models.⁷ However, a more recent study by Christensen et al. that also included children undergoing cardiac as well as noncardiac surgery, demonstrated an adjusted OR of mortality after perioperative CA of 2.2 (95% CI 1.2–3.9) if the event occurred on a night or weekend vs 7:00 am–7:00 pm on a weekday.⁴ A similar study in a mixed pediatric and adult cohort also showed this association between non-standard working hours and decreased immediate survival and hospital survival after perioperative CA.³⁰ This evaluation of the relationship between timing and mortality after CA has not been performed within the pediatric noncardiac surgery population yet and warrants further investigation.

This scoping review exposes the paucity of information on how perioperative CAs are managed and the role of mechanical circulatory support such as ECPR on outcomes. The American Heart Association has had published guidelines for Pediatric Advanced Life Support since 1986,³⁶ with the most recent guidelines published in 2025.³⁷ Current recommendations emphasize high quality CPR, with minimal interruptions and adequate rate and depth. To identify and address existing gaps in knowledge and clinical performance, further investigation into the management of perioperative CA events is essential. Such efforts are critical to improving care and optimizing outcomes in this high-risk context.

Although every study included in this review documented some measure of survival or mortality, the exact timeframe for what constituted “survival” was variable, highlighting the lack of standardization in reporting and making comparisons among studies challenging. Among studies that utilized the measure of survival to hospital discharge, the results were quite wide, ranging from 30% to 91%.6, 10, 18, 21, 22, 24 Other short-term outcomes, such as length of mechanical ventilation, new or progressive multiple organ dysfunction, unplanned intensive care unit admission, length of intensive care unit stay, length of hospital stay, and discharge destination were each presented sparingly or not at all. Similarly, neurologic outcomes were referenced in only a few articles and with subjective terminology that did not specify the basis of diagnosis or the specifics of the neurological deficits. Understanding neurological outcomes is important and necessary – the quality of survival is just as important as the survival outcomes themselves. Furthermore, none of these published studies addressed long-term outcomes such as neurobehavioral functioning, healthcare utilization, and quality of life. These represent major gaps in the literature.

This scoping review highlights major limitations of the available literature, with significant heterogeneity among the studies included. Variability in the definition of “perioperative,” the types of procedures included, the time periods covered, the countries where studies were conducted, and the measures reported for outcomes contributed to a lack of consistency across studies. Due to the rarity of perioperative CA in pediatrics, most studies reported a low number of events, making it difficult to draw robust conclusions from individual analyses, and highlighting the need for larger, more comprehensive studies in the future. Unfortunately, the heterogeneity of study designs, CPR management reporting, and outcomes reporting would make aggregation and meta-analysis of results challenging for a potential systematic review.

Limitations

We did not set a limit for date of publication so that we can understand the historical context of the evolution of perioperative CA. This led to inclusion of articles as far back as 1953, with events included from as far back as 1924. Results reported in these articles are reflective of anesthetic practices that are no longer in use. These early studies had limited descriptions of CA events, reducing the impact they would have had on the synthesis of results. The choice of age threshold may have reduced the number of included studies. Age less than 18 years was used for inclusion because it is a commonly used threshold between child and adult and is the age for which Pediatric Advanced Life Support guidelines are designed.³⁷ However, this could have potentially led to exclusion of studies from children’s hospitals where patients are cared for into their early 20s. Lastly, a number of studies were excluded from this review due to inability to separate cardiac surgery from noncardiac surgery procedures, as described above. This was necessary, though, as the objective of this review was to synthesize the literature on CA specifically in the noncardiac surgery population.

Conclusion

This scoping review demonstrates the array of studies published on the topic of pediatric perioperative CA associated with noncardiac surgery and emphasizes the lack of standardization in reporting. Moreover, it highlights the current gaps in knowledge on some of the potential risk factors for CA and mortality, CA management quality and interventions, neurologic outcomes, and long-term functional and quality of life outcomes. Further research is essential to establish a more comprehensive understanding of perioperative CA in children undergoing noncardiac surgery – knowledge that will be invaluable for ultimately improving clinical care for these children.

CRediT authorship contribution statement

Anisha S. Nadkarni: Writing – review & editing, Writing – original draft, Visualization, Supervision, Project administration, Methodology, Investigation, Data curation, Conceptualization. Lisa Hwang: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation, Conceptualization. Donald H. Shaffner: Writing – review & editing, Conceptualization. Alejandro V. Garcia: Writing – review & editing, Conceptualization. Katie Lobner: Writing – review & editing, Methodology, Conceptualization. Melania M. Bembea: Writing – review & editing, Visualization, Supervision, Project administration, Methodology, Investigation, Conceptualization.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. A.S.N. received salary support from the NIH Postdoctoral Research Training in Anesthesiology and Critical Care Medicine grant (T32GM075774). M.M.B.’s institution received funding for ECMO-related research from the National Institute of Neurological Disorders and Stroke (R01NS106292).

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 material to this article: