Introduction

Cardiac arrest in pediatric populations remains a critical global health concern due to its high mortality and profound morbidity among survivors [1]. Children present unique physiological challenges during both the pre‐arrest and resuscitation periods, resulting in varied etiologies and outcomes for in‐hospital cardiac arrest (IHCA) [2–4]. Recent data suggest that while the incidence of IHCA in children may be lower than in adults, the mortality rate can be exceedingly high, especially in resource‐limited settings [5].

Advances in pediatric critical care, including more sophisticated monitoring and rapid‐response systems, have contributed to improved survival in certain centers [6–8]. However, this benefit is not uniformly shared across all healthcare settings. In resource‐rich hospitals, dedicated pediatric rapid‐response teams (RRTs) and standardized resuscitation protocols have significantly reduced in‐hospital mortality, whereas facilities with limited pediatric expertise continue to face higher mortality rates [9].

In low- and middle-income countries, limited critical care resources, inconsistent practices, and delayed detection and resuscitation further complicate pediatric IHCA care [10–12]. These factors contribute to poor outcomes for critically ill children, with delays in diagnosis and high mortality rates [13–15]. While single‐center reports can shed light on specific case mixes or best‐practice environments, nationwide data better capture the heterogeneity in healthcare quality, resource allocation, and patient complexity. Findings from such an analysis can help policymakers and clinicians identify the most vulnerable pediatric groups and the critical points during hospitalization that require improvements in recognition, prevention, and resuscitation.

Accordingly, this study leverages nationwide data to quantify incidence, mortality trends, and long-term outcomes. Moreover, we identified associated factors for post-discharge death after pediatric IHCA. Understanding these patterns is essential for designing targeted interventions, optimizing resource utilization, and ultimately improving pediatric IHCA outcomes in Thailand and comparable healthcare systems around the world.

Materials and methods

Results

Incidence of IHCA

Between January 2015 and December 2022, there were between 1,321,335 and 1,845,969 pediatric admissions per year in Thailand, totaling 13,119,641 admissions. A flow diagram illustrating the number of pediatric admissions assessed, the identification of IHCA cases, exclusions based on coding criteria, and the final cohorts included in each analysis has been added as Fig 1. Across this eight‐year span, a total of 20,590 IHCA admissions were identified, corresponding to an overall incidence of 1.57 per 1,000 admissions. Year‐by‐year analysis showed that the incidence was highest in 2015 and 2016 (1.8 per 1,000 admissions), then gradually declined to 1.2 per 1,000 in 2022 (p < 0.001). Over the entire study, the number of IHCA‐related mortality per 1,000 admissions decreased from 1.2 in 2015 to 0.8 in 2022 (p < 0.001) (Table 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Incidence and mortality rate of in-hospital cardiac arrest among Thai children hospitalized during 2015-2022.

https://doi.org/10.1371/journal.pone.0341430.t001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flow diagram summarizes total pediatric hospitalized with in-hospital cardiac arrests, in-hospital mortality, and post-discharge mortality between January 2015 and December 2022.

https://doi.org/10.1371/journal.pone.0341430.g001

Patient characteristics

Demographic and clinical characteristics of the 20,590 IHCA admissions are shown in Table 2. Most patients were male (58.4%, n = 12,022). Neonatal younger than 1 month accounted for nearly half of all IHCA cases (40.9%, n = 8,429), and they had a comparatively higher survival rate (48.9%) compared to older children (p < 0.001). Regional distribution was also significantly different between survivors and non-survivors (p < 0.001), with the Northeast region of Thailand having the largest proportion of IHCA events (30.4%, n = 6,255). Over two-thirds of IHCA cases (66.8%, n = 13,761) occurred in tertiary-level hospitals, and this distribution had a significant association with outcome (p < 0.001) (Table 2).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Characteristics of Thai children experiencing in-hospital cardiac arrest, categorized by discharge status.

https://doi.org/10.1371/journal.pone.0341430.t002

Long-term survival after hospital discharge

Long-term survival data following discharge were obtained from 7,253 cases, indicating that 2,149 (29.6%) children subsequently died post-discharge. Kaplan-Meier curves illustrated that neonates (0 to under 1 month of age) had the most favorable post-discharge survival probabilities, while older pediatric populations, particularly those aged 12 to under 18 years, exhibited lower survival trajectories (Fig 2). At the end of the follow-up period, there were 2,149 deaths among 7,253 survivors at hospital discharge (29.6%). Neonates had mean survival time of 94.6 months (95% CI 93.3–95.9), which was significantly longer than children aged 1- < 12 months (62.4 months; 95%CI 59.3–65.6), 1- < 5 years (65.8 months; 95% CI 62.3–69.3), 5- < 12 years (68.1 months; 95% CI 64.2–72.0), or 12- < 18 years (65.6 months; 95% CI 62.4–68.7). Children with metabolic acidosis and hypoglycemia had substantially lower mean survival times post‐discharge than those without these conditions, mean survival time 45.8 (95% CI 41.4–50.4; p < 0.001) and 42.2 months (95% CI 34.4–50.1; p < 0.001), respectively (Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Mean Survival Time Following Hospital Discharge for In-Hospital Cardiac Arrest in Thai Children.

https://doi.org/10.1371/journal.pone.0341430.t003

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Cumulative survival rates in months for patients after in-hospital cardiac arrest using Kaplan-Meier curve.

https://doi.org/10.1371/journal.pone.0341430.g002

Multivariable analyses identified age ≥ 1 month as a significant independent predictor of post-discharge mortality. In comparison to neonates, children aged 1 to <5 years exhibited aHR of 3.07 (95% CI 2.65–3.56; p < 0.001), while those aged 5 to <12 years demonstrated an aHR of 3.17 (95% CI 2.71–3.71; p < 0.001), and adolescents aged 12 to <18 years showed an aHR of 3.37 (95% CI 2.94–3.87; p < 0.001). Among the various associated diagnoses of cardiac arrest, hypoglycemia was identified as a strong predictor of mortality, with an aHR of 1.54 (95% CI 1.25–1.89; p < 0.001), followed closely by metabolic acidosis, which had an aHR of 1.50 (95% CI 1.32–1.71; p < 0.001). Regarding organ dysfunction, DIC was associated with an aHR of 1.51 (95% CI 1.24–1.85; p < 0.001), whereas acute liver failure associated with an aHR of 1.42 (1.10–1.84; p = 0.008). Importantly, anoxic brain damage also demonstrated a significant association with an aHR of 1.23 (95% CI 1.05–1.44; p = 0.009) (Table 4).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Factors associated with mortality after discharge from in-hospital cardiac arrest (N = 7,253 cases).

https://doi.org/10.1371/journal.pone.0341430.t004

Discussion

Our comprehensive eight-year study of a large population found that the overall incidence of IHCA was 1.57 per 1,000 admissions, with an in-hospital mortality rate of 62.7%. These findings highlight the significant clinical burden and the urgent need for continued efforts to reduce mortality in hospitalized children. In our nationwide dataset, both the incidence and mortality of pediatric IHCA declined over time. Although our database did not capture the drivers of this trend, possible contributors include improved pediatric critical care capacity, earlier recognition of at-risk patients, and broader resuscitation training. Future studies linking administrative data with process-of-care measures will be needed to confirm these explanations. In our cohort, survival to hospital discharge was 37.3%, which is notably similar to previously reported from multi‐center registries in high‐income countries, but our observed mortality rates are higher than those in some resource‐rich settings, where survival to hospital discharge has been reported to be 39.8% [18]. This similarity suggests that, despite differences in healthcare structure and resource distribution, short-term survival outcomes following pediatric IHCA in Thailand are broadly comparable to those reported internationally.

Clear age-related variation in post-discharge mortality was observed in this study. Neonates younger than one month had the lowest long-term mortality and were therefore used as the reference group. In contrast, infants between 1 and 12 months of age showed a substantially higher risk of death after discharge, with adjusted hazard ratios closely approximating those seen in older pediatric age groups. This pattern indicates that the apparent survival advantage is primarily limited to the neonatal period rather than encompassing infancy as a whole. Favorable neonatal outcomes may be attributable to the highly structured nature of neonatal intensive care, including continuous physiologic surveillance and standardized approaches to metabolic and hemodynamic management [19,20]. Conversely, the higher mortality among infants and older children may reflect greater heterogeneity in arrest causes, increasing clinical complexity, and a higher burden of underlying conditions, all of which have been linked to poorer outcomes following pediatric cardiac arrest [21].

A striking finding from our results is the elevated risk associated with metabolic abnormalities, particularly metabolic acidosis and hypoglycemia. Both conditions are strong indicators of suboptimal organ function and can represent underlying metabolic or endocrine issues [22]. These results aligned with prior work show that metabolic derangements can precipitate or prolong circulatory collapse [23]. Early identification and aggressive correction of these abnormalities could reduce IHCA incidence and improve survival [1].

Further underpinning this conclusion, DIC increased long‐term mortality risk. This is consistent with literature that pinpoints DIC as a reflection of severe systemic stress, such as sepsis, trauma, or advanced organ failure [24,25]. Our findings also parallel studies demonstrating strong associations between multi‐organ dysfunction and negative outcomes in pediatric critical care [26–28]. Early recognition of these complications in clinical practice remains crucial; however, in our registry-based analysis we could not determine whether they occurred before, during, or after the cardiac arrest event, and they should therefore be interpreted as markers of severe illness among children who experienced IHCA rather than confirmed post-arrest complications. Our study revealed that several characteristics are significantly associated with survival following in-hospital cardiac arrest. Metabolic acidosis was identified as an important factor associated with increased post-discharge mortality consistent with previous research emphasizing its role in deteriorating prognosis by aggravating cardiac dysfunction and lowering the efficacy of resuscitation attempts [29]. Similarly, hypoglycemia emerged as a significant factor consistent with previous research indicating that low glucose levels severely impede neurological recovery and are associated with poor neurological recovery and increased mortality [30]. In terms of organ dysfunction, acute liver failure was significant, supporting previous research showing that severe hepatic involvement is associated with poor outcomes in critically ill children [31,32]. Disseminated intravascular coagulation and anoxic brain damage were also identified as significant negative prognostic factors, consistent with extensive research demonstrating that coagulation disorders and brain injury are strongly associated with poorer outcomes in this population [33,34]. In our dataset, however, these diagnoses should be interpreted as indicators of severe systemic illness among children with IHCA, because the registry does not provide the timing of diagnostic codes. Therefore, we cannot determine whether these organ dysfunctions developed before, during, or after the cardiac arrest, and we are unable to infer a definite causal or temporal sequence despite their strong association with post-discharge mortality.

Comparisons of our findings with international data must be considered within the context of Thailand’s healthcare system, which is relatively well‐developed but still exhibits variability in pediatric critical care capacity among provincial and tertiary hospitals. Several large hospitals have Pediatric Intensive Care Units (PICUs) staffed by trained intensivists, while smaller provincial hospitals may lack such expertise. The higher proportion of IHCA events in tertiary settings (66.8%) underscores that these centers likely receive more severely ill children who are transferred for specialized care.

Advances in early detection, efficient intervention, and high-quality resuscitation have contributed to better survival rates over the decade [8]. Standardizing pediatric early warning scores (PEWS) and rapid-response team activation criteria could expedite the detection of clinical deterioration [35]. In addition, a nationwide push to enhance metabolic monitoring, particularly vigilant glucose and acid‐base assessments, could reduce mortality in at‐risk children [36,37]. Enhanced training in pediatric advanced life support (PALS) for multidisciplinary hospital staff may further improve resuscitation outcomes and limit secondary injuries that contribute to long‐term mortality [38,39].

One of the most striking findings of this study is that nearly 30% of children who survived to hospital discharge died during the follow-up period. Although data are limited, evidence from long-term follow-up studies and systematic overviews suggests that pediatric cardiac arrest survivors often require ongoing care due to dynamic neurologic, cognitive, and functional needs, and that gaps in coordinated follow-up care are common [40]. This outcome highlights the need for robust follow‐up services, including specialized outpatient pediatric cardiology or critical care clinics, alongside community‐level support systems. Many survivors experience new morbidities and functional decline [41]. The need for comprehensive follow-up services is apparent, as research indicates elevated rates of patients being lost to follow-up and gaps in care [42,43]. However, standardized follow-up protocols are lacking, highlighting the need for further research and guideline development.

Although this nationwide analysis benefits from a large, population-based sample, several limitations should be considered. The study relied on administrative data from the National Health Security Office, which limits the depth and granularity of available clinical information and introduces the potential for misclassification or incomplete coding of comorbidities and complications. Because the registry lacks timestamps, we were unable to determine whether certain conditions preceded the cardiac arrest or developed afterward, and some complications may therefore reflect preexisting illness rather than post-arrest organ dysfunction. Detailed resuscitation characteristics including chest compression quality, medication administration, defibrillation, extracorporeal CPR, and post–cardiac arrest care were not available, precluding assessment of resuscitation practices, staff training, or adherence to standardized Utstein reporting elements. Similarly, hospital-level characteristics such as urban or rural location, teaching status, ICU availability, and interdisciplinary resuscitation training could not be evaluated. Information on treatment limitation decisions, including do-not-resuscitate orders, as well as socioeconomic factors and post-discharge care, was also unavailable and may have influenced long-term mortality. Functional and neurological outcomes were not captured, limiting interpretation of survival quality. These constraints underscore the need for future prospective, clinically detailed registries in Thailand that integrate patient comorbidities, hospital resources, resuscitation processes, and long-term neurodevelopmental outcomes to more comprehensively inform quality improvement efforts and survivorship care following pediatric in-hospital cardiac arrest.

Conclusion

This nationwide study quantified the incidence and mortality of pediatric in-hospital cardiac arrest in Thailand, showing a gradual decline in both over the eight-year period. We described long-term outcomes, with nearly one-third of children who survived to discharge dying during follow-up. During a median follow-up of 67 months, we identified metabolic derangements and organ dysfunction as independent factors associated with post-discharge mortality. These findings directly address our study objectives and provide a foundation for future strategies to improve outcomes in this high-risk population.

Acknowledgments

We would like to acknowledge National Health Security Office for data source.