The impact of socioeconomic status on paediatric out-of-hospital cardiac arrest incidence and outcomes in Western Australia: a population-based cohort study Authors: Sarah Ann Harris, Stephen Ball, David Majewski, Jason Belcher, Jason Acworth, Judith Finn Published: 2026 ABSTRACT Graphical abstract Introduction Paediatric out-of-hospital cardiac arrest [pOHCA] is a rare but devastating event, with poor survival outcomes.1, 2 Compared with adults, the annual incidence of OHCA in children is markedly lower,2, 3 with paediatric rates reported between 3.7 and 15.4 per 100,000 population,4, 5, 6, 7 compared with 62.3–107.4 per 100,000 in adults.8, 9, 10 Incidence is strongly age-dependent, with infants (<1 year) consistently experiencing markedly higher incidence rates (11.5–47.9 per 100,000) than older children (1.3–6.6 per 100,000).5, 11, 12, 13 Despite advances in paediatric resuscitation and post-arrest care,14, 15, 16 overall outcomes following pOHCA remain poor.1, 17 Return of spontaneous circulation [ROSC] is typically reported in fewer than one-quarter of cases,11, 18 while survival to hospital discharge or 30 days is generally below 10%,3, 11, 18, 19 although estimates vary by study and setting. Socioeconomic status [SES] is a well-established determinant of OHCA risk,20, 21, 22, 23 and survival24, 25, 26 in adults, with greater socioeconomic disadvantage associated with higher incidence and poorer outcomes.20, 21, 22, 23, 24, 25, 26 However, in pOHCA the role of SES has not been well characterised at the population level.1, 27 A recent systematic review identified socioeconomic and racial disparities across several components of the pOHCA chain of survival, including bystander cardiopulmonary resuscitation [bCPR]11, 28, 29, 30 and access to defibrillation,29, 31, 32 but highlighted the need for further research examining SES in relation to pOHCA incidence.1 Most studies examining socioeconomic disparities in pOHCA have been conducted in the United States27, 29, 31, 32, 33, 34 or Asia,28, 30, 34 with the majority utilising neighbourhood-level SES measures assigned at the location of arrest.28, 29, 30, 31, 32, 33 Where pOHCA risk has been examined, analyses have largely relied on race or ethnicity as proxies for socioeconomic disadvantage1, 35, 36 or modelled relative risk per incremental change in area-level income and poverty measures rather than estimating SES-specific incidence rates.37 More recently, population-based studies from the United States27 and Canada38 have reported socioeconomic gradients in pOHCA incidence. The United States registry study27 assigned SES at the census tract level based on the location of arrest, while Idrees et al., derived SES using the patient’s residential address.38 Although these studies provide important evidence that socioeconomic disadvantage is associated with pOHCA risk, methodological differences in SES attribution and study design remain. Furthermore, SES-specific incidence stratified by paediatric age group has not been examined. Consequently, it remains unclear whether socioeconomic gradients in pOHCA incidence differ across paediatric age groups when SES is assigned using residential address, which may better reflect the underlying socioeconomic context of the child. This study aimed to quantify socioeconomic gradients in pOHCA incidence using a 10-year, population-based cohort from Western Australia [WA], and to assess whether these gradients differed across paediatric age groups. Secondary objectives were to examine differences in patient characteristics, prehospital details, and survival outcomes using residential address-based SES assignment. Materials and methods Study design This is a population-based, retrospective cohort study of pOHCA cases attended by St John WA [SJ-WA] emergency medical services [EMS] in Western Australia [WA] between 1st January 2015 and 31st December 2024. Patients were included if they were: (i) aged <15 years (consistent with the Australian Bureau of Statistic [ABS] definition of adults ≥15 years)39; and (ii) residents of WA. All pOHCA cases were included, irrespective of whether resuscitation was attempted, or of medical or non-medical aetiology. Setting WA has a land area of 2.53 million km2,40 with a population of 2.7 million and almost 80% residing in the greater Perth metropolitan area.41 In 2021, 11% of the WA population were aged 0–14 years (n = 505,783; 51% male).41 In WA, SJ-WA is the primary provider of road based EMS.42 All pOHCA patients are transported to the closest hospital in rural areas or to the primary children’s hospital emergency department [ED] within the Perth metropolitan area, unless resuscitative efforts are ceased in the field by paramedics (or not commenced), due to futility or pre-existing not-for-resuscitation orders. Paediatric advanced life support is administered by all metropolitan crews and paramedic-staffed rural centres.43, 44 Data source The WA OHCA database is maintained by the Prehospital, Resuscitation and Emergency Care Research Unit [PRECRU] at Curtin University, WA. Data are extracted from SJ-WA electronic patient care records [ePCR] and computer-aided dispatch [CAD] information for all EMS-attended OHCAs. Outcome measures The primary outcome was pOHCA incidence. Crude and age-standardised incidence rates [ASIRs] were calculated over the 10-year study period. Annual population denominators were derived from the 2016 ABS Census for 2015–2019, and the 2021 Census for 2020–2024,45, 46 stratified by age and SES tertile. ASIRs were directly standardised to the 2001 Australian Standard Population47 using single-year age-specific incidence rates (0–14 years), which were subsequently aggregated to the study-defined age groups. All incidence rates were reported per 100,000 population per year. Secondary outcomes were 30-day survival, determined through manual look-up in the WA Death Registry,48 for patients not recorded as deceased by SJ-WA EMS personnel, and return of spontaneous circulation [ROSC] on arrival at the ED, as documented in the WA OHCA database. Socioeconomic status OHCA cases were geocoded using the geographical coordinates of each patient’s residential address. For deceased patients with missing residential address data, supplementary information was obtained through linkage with the WA Death Registry.48 Geocoded cases were imported into ArcGIS Desktop v10 to assign area-level SES based on residential location. Further methodological details have been described previously.23 SES was assigned using Statistical Area 1 [SA1] units, as defined by the ABS,45, 46 representing the smallest geographic scale at which SES data are available. SA1 boundary files corresponding to the 2016 and 2021 census were applied based on arrest date (2016 SES for events to 31 December 2019; 2021 SES applied thereafter). SES was assigned to each OHCA case using the ABS Index of Relative Socio-Economic Disadvantage (IRSD).49, 50 Each SA1 within WA is assigned a state-based percentile ranking (1–100), with lower values indicating greater relative disadvantage. A benefit of using area-assigned SES is it is assigned at the household level, therefore attributing the same SES classification to all household members, including children aged <15 years. The IRSD is a validated composite index incorporating variables derived from the adults within the household (≥15 years) to determine disadvantage, including income, education, employment status, occupation type, household overcrowding, mortgage stress, and disability status.39, 49, 50 These variables are collected nationally by the ABS as part of the compulsory 5-yearly Australian Census. Additional variables Additional variables included patient demographics (age, sex), arrest characteristics (witness status, bystander-CPR, AED application, aetiology, initial arrest rhythm) and EMS details including EMS response times, EMS intervention (defibrillation), and transportation to hospital. POHCA cases were classified as non-medical (i.e. external causes: hanging, drug overdose, drowning, electrocution, asphyxia (excluding hanging), or other trauma), or medical (i.e. presumed cardiac, anaphylaxis, medical-related respiratory, sudden unexplained death of an infant [SUDI] or other medical causes).15 Definitions To categorise SES status, IRSD SA1 percentiles were collapsed to tertiles (Low SES: 0–<34; Mid SES: ≥34–<67; High SES: ≥67–100), where low SES represents greatest relative disadvantage and high SES represented the least relative disadvantage. Tertiles were used instead of quintiles to ensure analytically stable group sizes given the small number and distribution of pOHCA events. Age groups were defined as infants (<1 year), young children (1–5 years), and older children (6–14 years) broadly consistent with prior pOHCA studies.7, 11, 51 Statistical analysis Descriptive statistics including mean, standard deviation (or median and interquartile range [IQR]) for continuous variables; and frequency and percentage for categorical variables were reported. Inferential comparisons of demographic and arrest characteristics were stratified by SES tertile (low, mid, high) using chi-squared (categorical variables), or Kruskal-Wallis tests (continuous variables). For categorical variables where the expected cell-count was less than 20%, Monte Carlo simulation (10,000 samples, 95% confidence intervals) was performed. Descriptive comparisons for sex and aetiology were further stratified by age group. Incidence was modelled using negative binomial regression with the log of the WA population included as an offset to account for variation in population at risk. SES was specified as the primary exposure, with the low-SES tertile (greatest disadvantage) as the reference category. Age group was included as a categorical variable, and an interaction between SES and age group was fitted to assess effect modification. Year of OHCA was included as an adjustment variable. Incidence rate ratios [IRRs] with 95% confidence intervals were reported. Due to low numbers of survival events, multivariable regression modelling was not feasible. Differences in ROSC and 30-day survival between SES tertiles and age-groups were assessed using chi-squared tests. Statistical significance was set at p < 0.05. Analysis was completed in Microsoft Excel (Microsoft Corporation), IBM SPSS Statistics (Version 30.0.0) and StataCorp (Release 2019). Ethical considerations Ethical approval was granted from Curtin Human Research Ethics Committee (Approval no. HRE2025-0301), as a sub-study to the Western Australian Prehospital Care Record Linkage Project. Approval was also granted by the St John WA Research Governance Committee. Results Patient characteristics A total of 27,723 OHCAs occurred in WA between 2015 and 2024, including 411 paediatric cases (1.5%) in patients aged <15 years who met the study inclusion criteria (Fig. 1). Males accounted for 57.9% of cases (n = 238) (Table 1). Infants comprised 37.5% (n = 154), young children 30.7% (n = 126), and older children 31.9% (n = 131). All patients experienced a single OHCA event during the study period, except one child with two events.Fig. 1Study cohort (n = 411).SES: Socioeconomic Status.Table 1Characteristics of a paediatric OHCA cohort stratified by socioeconomic status.CharacteristicsSocioeconomic statusp-valuesLowMidHighTotalTotal cases (n)20112981411Sex, male, n (%)114 (56.72%)77 (59.69%)47 (58.02%)238 (57.91%)0.867Age Infant (<1 year)86(42.79%)50(38.76%)18(22.22%)154(37.47%)0.007* Young child (1–5 years)63(31.34%)32(24.81%)31(38.27%)126(30.66%) Older child (6–14 years)52(25.87%)47(36.43%)32(39.51%)131(31.87%)Year of OHCA by census period 2015–201999 (49.25%)60 (46.51%)39 (48.15%)198 (48.18%)0.888 2020–2024102 (50.75%)69 (53.49%)42 (51.85%)213 (51.82%)Incidence, per 100,000 person-years Crude+15.018.284.909.030.182 Age-standardised++14.628.145.029.050.001*Incident location, n (%) Private residence166 (82.59%)112 (86.82%)67 (82.72%)345 (83.94%)0.561 Public/other35 (17.41%)17 (13.18%)14 (17.28%)66 (16.06%)Incident region, n (%) Rural/remote56 (27.86%)32 (24.81%)15 (18.52%)103 (25.06%)0.261 Metro145 (72.14%)97 (75.19%)66 (81.48%)308 (74.94%)Witness status, n (% EMS9 (4.48%)9 (6.98%)5 (6.17%)23 (5.60%)0.424 Bystander30 (14.93%)16 (12.40%)17 (20.99%)63 (15.33%) Unwitnessed162 (80.60%)104 (80.62%)59 (72.84%)325 (79.08%)#Bystander CPR (excludes EMS witnessed), n (%145 (75.52%)85 (70.83%)64 (84.21%)294 (75.77%)0.103Initial rhythm, n (% Shockable (VT, VF)5 (2.49%)5 (3.88%)5 (6.17%)15 (3.65%)0.329 (95%CI: 0.319–0.338)^#Bystander AED pads applied for non-EMS witnessed cases, n (%9 (4.69%)<5<517 (4.38%)0.764Aetiology, n (%) Non-medical68 (33.83%)43 (33.33%)30 (37.04%)141 (34.31%)0.843 Medical133 (66.17%)86 (66.67%)51 (62.96%)270 (65.69%)EMS response time Median [IQR] (minutes)8.98[6.82–12.65]9.17[7.03–11.42]9.6[6.98–11.8]9.13[6.92–11.92]0.965EMS Resuscitation Attempted, n (%155 (77.11%)103 (79.84%)64 (79.01%)322 (78.35%)0.830Transport time to hospital, n (%136 (67.66%)87 (67.44%)53 (65.43%)276 (67.15%) Median [IQR] (minutes)9[5.12–13.29]11 [6.88–13.90]10.63 [7.07–13.33]10.13 [5.73–13.55]0.201Died at scene, n (%65 (32.34%)42 (32.56%)28 (34.57%)135 (32.85%)0.934ROSC Any, n (%35 (17.41%)13 (10.08%)7 (8.64%)55 (13.38%)0.061ROSC on hospital arrival, n (%29 (14.43%)12 (9.30%)7 (8.64%)48 (11.68%)0.234Survival to 30-days, n (%17 (8.46%)9 (6.98%)6 (7.41%)32 (7.79%)0.878SES: Socioeconomic Status; ROSC: Return of Spontaneous Circulation; EMS: Emergency Medical Services; IQR: Interquartile Range; VT: Ventricular Tachycardia; VF: Ventricular Fibrillation; AED: automatic external defibrillator.SES tertiles are defined by the IRSD at an SA1 level. Low SES indicates greatest level of disadvantage, to High indicating least disadvantage.Denominator per characteristic total is n = 411 unless otherwise stated.+Crude incidence rates were calculated using the ABS-defined WA population by IRSD tertile for the corresponding census year (2016, 2021). Differences between SES tertiles were assessed using Poisson regression.++Age-standardised incidence rates [ASIR] were calculated using the ABS-defined WA population by IRSD tertile (2016 and 2021 census years) as the exposure population45, 46 and standardised to the 2001 Australian Standard Population. Differences between SES tertiles were assessed using Poisson regression, adjusting for age.#Denominators per SES for non-EMS witnessed. Low n = 192; Middle n = 120; High n = 76; Total n = 388.*p-values significant at the p < 0.05 level. Unless otherwise stated, significance is assessed for differences within characteristic categories across the SES tertiles using Pearson Chi-Square for categorical variables and Kruskal-Wallis for continuous variables.^Chi-Square Monte Carlo simulation (10,000 samples, 95% confidence interval; [95%CI]) to account for small-expected cell counts for categorical variables. Study cohort (n = 411). SES: Socioeconomic Status. Table 1Characteristics of a paediatric OHCA cohort stratified by socioeconomic status.CharacteristicsSocioeconomic statusp-valuesLowMidHighTotalTotal cases (n)20112981411Sex, male, n (%)114 (56.72%)77 (59.69%)47 (58.02%)238 (57.91%)0.867Age Infant (<1 year)86(42.79%)50(38.76%)18(22.22%)154(37.47%)0.007* Young child (1–5 years)63(31.34%)32(24.81%)31(38.27%)126(30.66%) Older child (6–14 years)52(25.87%)47(36.43%)32(39.51%)131(31.87%)Year of OHCA by census period 2015–201999 (49.25%)60 (46.51%)39 (48.15%)198 (48.18%)0.888 2020–2024102 (50.75%)69 (53.49%)42 (51.85%)213 (51.82%)Incidence, per 100,000 person-years Crude+15.018.284.909.030.182 Age-standardised++14.628.145.029.050.001*Incident location, n (%) Private residence166 (82.59%)112 (86.82%)67 (82.72%)345 (83.94%)0.561 Public/other35 (17.41%)17 (13.18%)14 (17.28%)66 (16.06%)Incident region, n (%) Rural/remote56 (27.86%)32 (24.81%)15 (18.52%)103 (25.06%)0.261 Metro145 (72.14%)97 (75.19%)66 (81.48%)308 (74.94%)Witness status, n (% EMS9 (4.48%)9 (6.98%)5 (6.17%)23 (5.60%)0.424 Bystander30 (14.93%)16 (12.40%)17 (20.99%)63 (15.33%) Unwitnessed162 (80.60%)104 (80.62%)59 (72.84%)325 (79.08%)#Bystander CPR (excludes EMS witnessed), n (%145 (75.52%)85 (70.83%)64 (84.21%)294 (75.77%)0.103Initial rhythm, n (% Shockable (VT, VF)5 (2.49%)5 (3.88%)5 (6.17%)15 (3.65%)0.329 (95%CI: 0.319–0.338)^#Bystander AED pads applied for non-EMS witnessed cases, n (%9 (4.69%)<5<517 (4.38%)0.764Aetiology, n (%) Non-medical68 (33.83%)43 (33.33%)30 (37.04%)141 (34.31%)0.843 Medical133 (66.17%)86 (66.67%)51 (62.96%)270 (65.69%)EMS response time Median [IQR] (minutes)8.98[6.82–12.65]9.17[7.03–11.42]9.6[6.98–11.8]9.13[6.92–11.92]0.965EMS Resuscitation Attempted, n (%155 (77.11%)103 (79.84%)64 (79.01%)322 (78.35%)0.830Transport time to hospital, n (%136 (67.66%)87 (67.44%)53 (65.43%)276 (67.15%) Median [IQR] (minutes)9[5.12–13.29]11 [6.88–13.90]10.63 [7.07–13.33]10.13 [5.73–13.55]0.201Died at scene, n (%65 (32.34%)42 (32.56%)28 (34.57%)135 (32.85%)0.934ROSC Any, n (%35 (17.41%)13 (10.08%)7 (8.64%)55 (13.38%)0.061ROSC on hospital arrival, n (%29 (14.43%)12 (9.30%)7 (8.64%)48 (11.68%)0.234Survival to 30-days, n (%17 (8.46%)9 (6.98%)6 (7.41%)32 (7.79%)0.878SES: Socioeconomic Status; ROSC: Return of Spontaneous Circulation; EMS: Emergency Medical Services; IQR: Interquartile Range; VT: Ventricular Tachycardia; VF: Ventricular Fibrillation; AED: automatic external defibrillator.SES tertiles are defined by the IRSD at an SA1 level. Low SES indicates greatest level of disadvantage, to High indicating least disadvantage.Denominator per characteristic total is n = 411 unless otherwise stated.+Crude incidence rates were calculated using the ABS-defined WA population by IRSD tertile for the corresponding census year (2016, 2021). Differences between SES tertiles were assessed using Poisson regression.++Age-standardised incidence rates [ASIR] were calculated using the ABS-defined WA population by IRSD tertile (2016 and 2021 census years) as the exposure population45, 46 and standardised to the 2001 Australian Standard Population. Differences between SES tertiles were assessed using Poisson regression, adjusting for age.#Denominators per SES for non-EMS witnessed. Low n = 192; Middle n = 120; High n = 76; Total n = 388.*p-values significant at the p < 0.05 level. Unless otherwise stated, significance is assessed for differences within characteristic categories across the SES tertiles using Pearson Chi-Square for categorical variables and Kruskal-Wallis for continuous variables.^Chi-Square Monte Carlo simulation (10,000 samples, 95% confidence interval; [95%CI]) to account for small-expected cell counts for categorical variables. Characteristics of a paediatric OHCA cohort stratified by socioeconomic status. SES: Socioeconomic Status; ROSC: Return of Spontaneous Circulation; EMS: Emergency Medical Services; IQR: Interquartile Range; VT: Ventricular Tachycardia; VF: Ventricular Fibrillation; AED: automatic external defibrillator. SES tertiles are defined by the IRSD at an SA1 level. Low SES indicates greatest level of disadvantage, to High indicating least disadvantage. Denominator per characteristic total is n = 411 unless otherwise stated. Crude incidence rates were calculated using the ABS-defined WA population by IRSD tertile for the corresponding census year (2016, 2021). Differences between SES tertiles were assessed using Poisson regression. Age-standardised incidence rates [ASIR] were calculated using the ABS-defined WA population by IRSD tertile (2016 and 2021 census years) as the exposure population45, 46 and standardised to the 2001 Australian Standard Population. Differences between SES tertiles were assessed using Poisson regression, adjusting for age. Denominators per SES for non-EMS witnessed. Low n = 192; Middle n = 120; High n = 76; Total n = 388. p-values significant at the p < 0.05 level. Unless otherwise stated, significance is assessed for differences within characteristic categories across the SES tertiles using Pearson Chi-Square for categorical variables and Kruskal-Wallis for continuous variables. Chi-Square Monte Carlo simulation (10,000 samples, 95% confidence interval; [95%CI]) to account for small-expected cell counts for categorical variables. Almost half of all pOHCAs occurred in children residing in low-SES areas (n = 201; 48.9%) (Table 1). Infants represented the largest proportion of cases in the low- and mid-SES groups (n = 86, 42.8%; n = 50, 38.8% respectively), whereas the high-SES group contained proportionally fewer infants (n = 18, 22.2%) and more older children (n = 32, 39.5%) (Table 1). Other demographic, arrest and EMS characteristics were similarly distributed across SES groups (Table 1). Aetiology Most pOHCA events were of medical aetiology (n = 270; 65.79%), with no significant difference in the distribution of non-medical versus medical events by SES (p = 0.843; Table 1). Medical aetiology was most pronounced among infants, where 94.8% of events were medical (n = 146), with SUDI accounting for most infant cases (n = 102, 66.2%). Presumed cardiac aetiology was the most common medical cause among young (n = 52, 41.37%) and older children (n = 48, 36.6%) (Supplementary Table 1). Non-medical events accounted for nearly half of all events in young children (n = 59, 46.8%) and older children (n = 74, 56.5%). Among young children, drowning was the most common non-medical mechanism (n = 29; 23.0%), while in older children, hanging (n = 25, 19.1%) and other trauma (n = 33, 25.2%) were most frequent. Drowning accounted for a higher proportion of all OHCA events among low-SES children compared with mid- and high-SES children (Supplementary Table 1). Incidence For the overall cohort, the age-standardised incidence of pOHCA was 9.1 per 100,000 children per year. Age-standardised incidence demonstrated a significant monotonic socioeconomic gradient, decreasing from 14.6 per 100,000 children per year in the low-SES group to 5.0 per 100,000 children per year in the high-SES group (p = 0.001) (Table 1). Although crude incidence showed a similar directional trend, age-adjusted estimates are presented given the strong age dependence of pOHCA incidence. Incidence of OHCA was greater in infants (54.3 per 100,000 children per year) than in young children (8.3 per 100,000 children per year) or older children (4.8 per 100,000 children per year) (Table 2). After adjustment for year of OHCA, a clear SES gradient in incidence was observed. There was evidence of interaction between SES and age group (p = 0.015), indicating that the magnitude of the SES association differed across age groups. The magnitude of the SES association was strongest in infancy, where high-SES infants had an 80% lower incidence than low-SES infants (IRR 0.20; 95% CI 0.12–0.33) (Fig. 2). The relative reduction was smaller in young children (IRR 0.44; 95% CI 0.28–0.69) and older children (IRR 0.46; 95% CI 0.31–0.69).Table 2Crude and age-standardised [ASR] rates of pOHCA by SES, with age-specific incidence rates (per 100,000 population per year).Socioeconomic status categoryWA populationaAll OHCA(n = 411)pOHCA cases per census periodIncidence2016202120162021CrudebASRcLow SES129,050138,7149910215.0114.62Mid SES151,182160,48160698.288.14High SES161,767168,73739424.905.02Age groupWA populationapOHCA cases per census periodAge specific incidenced2016202120162021Infant (<1 year)28,60228,139837154.28Young child (1–5 years)152,084152,54158688.27Older child (6–14 years)261,313287,25257744.78SES: Socioeconomic status; pOHCA: Paediatric out-of-hospital cardiac arrest; WA: Western Australia.Low SES indicates greatest level of disadvantage, to High indicating least disadvantage.aABS-defined WA, 0–14 year old population by IRSD tertile or age-group for the corresponding census year (2016, 2021).bCrude incidence rates were calculated using the ABS-defined WA population by IRSD tertile for the corresponding census year (2016, 2021).cAge-standardised incidence rates were calculated using the ABS-defined WA population by IRSD tertile (2016 and 2021 census years) as the exposure population and standardised to the 2001 Australian Standard Population.dAge-specific incidence rates were calculated using WA population denominators from the ABS, stratified by study-defined age groups and corresponding census year (2016 or 2021).Fig. 2Model-predicted pOHCA incidence by socioeconomic status and age group.Predicted incidence (events per 100,000 person-years) for infants (<1 year), young children (1–5 years), and older children (6–14 years), stratified by SES tertiles. Incidence estimates were derived from a negative binomial regression model including an SES × age-group interaction term and adjusted for year of pOHCA, with the log WA population included as an offset. WA population denominators were based on 2016 and 2021 census dates.SES tertiles were defined using the IRSD at the SA1 level, with Low-SES indicating greatest disadvantage and High-SES indicating least disadvantage. Table 2Crude and age-standardised [ASR] rates of pOHCA by SES, with age-specific incidence rates (per 100,000 population per year).Socioeconomic status categoryWA populationaAll OHCA(n = 411)pOHCA cases per census periodIncidence2016202120162021CrudebASRcLow SES129,050138,7149910215.0114.62Mid SES151,182160,48160698.288.14High SES161,767168,73739424.905.02Age groupWA populationapOHCA cases per census periodAge specific incidenced2016202120162021Infant (<1 year)28,60228,139837154.28Young child (1–5 years)152,084152,54158688.27Older child (6–14 years)261,313287,25257744.78SES: Socioeconomic status; pOHCA: Paediatric out-of-hospital cardiac arrest; WA: Western Australia.Low SES indicates greatest level of disadvantage, to High indicating least disadvantage.aABS-defined WA, 0–14 year old population by IRSD tertile or age-group for the corresponding census year (2016, 2021).bCrude incidence rates were calculated using the ABS-defined WA population by IRSD tertile for the corresponding census year (2016, 2021).cAge-standardised incidence rates were calculated using the ABS-defined WA population by IRSD tertile (2016 and 2021 census years) as the exposure population and standardised to the 2001 Australian Standard Population.dAge-specific incidence rates were calculated using WA population denominators from the ABS, stratified by study-defined age groups and corresponding census year (2016 or 2021). Crude and age-standardised [ASR] rates of pOHCA by SES, with age-specific incidence rates (per 100,000 population per year). SES: Socioeconomic status; pOHCA: Paediatric out-of-hospital cardiac arrest; WA: Western Australia. Low SES indicates greatest level of disadvantage, to High indicating least disadvantage. ABS-defined WA, 0–14 year old population by IRSD tertile or age-group for the corresponding census year (2016, 2021). Crude incidence rates were calculated using the ABS-defined WA population by IRSD tertile for the corresponding census year (2016, 2021). Age-standardised incidence rates were calculated using the ABS-defined WA population by IRSD tertile (2016 and 2021 census years) as the exposure population and standardised to the 2001 Australian Standard Population. Age-specific incidence rates were calculated using WA population denominators from the ABS, stratified by study-defined age groups and corresponding census year (2016 or 2021). Model-predicted pOHCA incidence by socioeconomic status and age group. Predicted incidence (events per 100,000 person-years) for infants (<1 year), young children (1–5 years), and older children (6–14 years), stratified by SES tertiles. Incidence estimates were derived from a negative binomial regression model including an SES × age-group interaction term and adjusted for year of pOHCA, with the log WA population included as an offset. WA population denominators were based on 2016 and 2021 census dates. SES tertiles were defined using the IRSD at the SA1 level, with Low-SES indicating greatest disadvantage and High-SES indicating least disadvantage. Survival Overall, 11.7% of cases achieved ROSC (n = 48) and 7.8% survived to 30 days (n = 32) (Supplementary Table 2). Among EMS-treated cases (n = 328), overall 30-day survival was 9.2%. For all cases, low-SES cases had higher proportions of ROSC (14.4%) and 30-day survival (8.5%) than mid-(9.3%, 7.0%) or high-SES groups respectively (8.6%. 7.4%); but differences were not statistically significant (ROSC χ2 = 2.90, p = 0.23; 30-day χ2 = 0.26, p = 0.88) (Table 1). Subgroup comparisons should be interpreted cautiously given small event numbers (Supplementary Table 2). Thirty-day survival for all cases, was highest in young children (11.1%), compared with infants (5.8%) and older children (6.9%). However, differences were not statistically significant (ROSC χ2 = 2.42, p = 0.30; 30-day χ2 = 2.90, p = 0.23). Discussion This population-based study examined the relationship between SES and pOHCA using residential address-based SES. We observed a clear socioeconomic gradient in pOHCA incidence, with children living in the most socioeconomically disadvantaged areas experiencing nearly threefold higher incidence than those in the least disadvantaged areas, and approximately double that of middle SES areas. Overall pOHCA incidence was similar to prior Australian estimates2, 18 and demonstrated strong age dependence, with the greatest burden observed in infancy.5, 11, 13 Socioeconomic disparities in incidence were evident across age-groups, with higher incidence among more disadvantaged children, consistent with patterns reported in adult20, 22, 23, 52 and paediatric27 OHCA populations. This gradient was most pronounced in infants and less evident in older children, suggesting socioeconomic disadvantage may exert its strongest influence on pOHCA risk early in life. The pronounced SES gradient in infancy may reflect the heightened susceptibility of early life to social and environmental conditions.53, 54, 55, 56 Socioeconomic disadvantage may influence risk through adverse antenatal exposures, differences in access to and engagement with health services, suboptimal living environments, and variation in parental education and health literacy.54, 55, 57 Psychosocial stress and material deprivation may further influence caregiver capacity to recognise and respond to early signs of illness.58 Consistent with previous research, medical aetiology was predominant in infants,59 while the proportion of non-medical pOHCA increased with age.60, 61 This pattern was observed across all SES tertiles, suggesting that socioeconomic differences in incidence are not explained by aetiology alone. No statistically significant differences in bystander CPR (b-CPR) delivery were observed across SES tertiles. While higher rates of bystander intervention in more advantaged populations have been reported previously,28, 29 our findings did not demonstrate a clear socioeconomic gradient in early resuscitation response. The majority of pOHCAs presented with a non-shockable initial rhythm, irrespective of SES. Overall, shockable rhythms occurred in 3.7% of cases, lower than reported in some previous studies.5, 27, 28 In this population-based cohort, 30-day survival following pOHCA was 7.8%, comparable to registry-based studies, reporting survival between 5 and 10%.11, 18, 62 Survival proportions were similar across SES tertiles, with no statistically significant socioeconomic differences observed. In contrast, Gathers et al.,27 reported socioeconomic gradients in survival following pOHCA. Due to the limited number of survivors in the present cohort, adjusted survival modelling was not undertaken. Consistent with prior reports, descriptively, survival varied by age, with the poorest outcomes observed in infants and progressively higher survival in older children.11, 62, 63 Future research examining socioeconomic disparities in pOHCA should prioritise consistent reporting of incidence to improve comparability across studies and better characterise patterns across the SES spectrum. SES-stratified analyses by age are warranted, given established age-related differences in incidence, aetiology and survival. Limitations These findings should be interpreted considering several limitations. Low case numbers are inherent to pOHCA research and (fortunately) reflect the rarity of the condition rather than a limitation unique to this study. While some comparisons could be tested (e.g. incidence by SES and age group), others were not undertaken due to concerns about low statistical power, including adjusted survival modelling. Finally, comparisons with prior studies should be made cautiously, due to heterogeneity in age definitions, study settings (e.g. registry-based versus school or sport-specific cohorts), and inclusion criteria such as whether resuscitation was attempted, all of which influence reported incidence and survival estimates. Conclusion Socioeconomic disadvantage was strongly associated with higher pOHCA incidence, with an overall near threefold difference between children living in the most and least disadvantaged areas. The socioeconomic gradient was most pronounced in infancy, suggesting that early life may represent a period of heightened vulnerability to social determinants of health. These findings emphasise the importance of addressing socioeconomic disadvantage in paediatric prevention strategies and highlight the need for further large-scale studies to clarify mechanisms underpinning SES disparities in pOHCA incidence and outcomes. Declaration of generative AI and AI-assisted technologies in the manuscript preparation process During the preparation of this work the authors used ChatGPT to improve the readability and brevity of the content. After using this tool, the authors reviewed and edited the content as needed and assume full responsibility for the content of the publication. CRediT authorship contribution statement Sarah Ann Harris: Writing – review & editing, Writing – original draft, Visualization, Project administration, Methodology, Formal analysis, Conceptualization. Stephen Ball: Writing – review & editing, Data curation, Conceptualization. David Majewski: Writing – review & editing, Data curation. Jason Belcher: Writing – review & editing, Methodology. Jason Acworth: Writing – review & editing, Data curation. Judith Finn: Writing – review & editing, Project administration, Methodology, Funding acquisition, Data curation. Declaration of competing interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: JF: Receives research funding from SJ-WA, and is on the Editorial Board of the journal Resuscitation. SB JF: Hold adjunct research positions with SJ-WA. JB: Employee of SJ-WA. DM JA SH have no conflict of interest.