Timing of Intervention and Long-Term Outcomes in Mild Congenital Heart Disease: A 35-year Study from the Pediatric Cardiac Care Consortium

Abstract

Background

Mild congenital heart diseases (CHD) such as atrial septal defect (ASD), ventricular septal defect (VSD), patent ductus arteriosus (PDA) and pulmonary valve stenosis (PS) constitute a significant public health problem. Understanding long-term outcomes after interventions for mild CHD is essential to inform lifelong care.

Methods

We queried the Pediatric Cardiac Care Consortium, a multicenter US-based registry of patients with CHD interventions, for patients undergoing surgical or transcatheter procedures for ASD, VSD, PDA and PS; 17,407 patients were included. Statistical methods included survival analysis by Kaplan-Meier plots, multivariable Cox proportional hazards models for risk factors for death and standardized mortality ratios (SMR).

Results

There were 115 in-hospital deaths (0.7%) with the remaining patients included in the post-discharge. The 35-year post-discharge survival ranged from 95.5% for PS to 92.1% for VSD. Survival patterns differed by lesion and age at intervention, with intervention at 1–<5 years generally associated with more favorable outcomes. Lower weight-for-age at intervention was associated with increased hazard of death across all shunt lesions but not for PS. Compared with the general population, mortality risk remained elevated for up to 18–20 years after intervention.

Conclusions

Long-term survival after intervention for mild CHD is excellent; yet lesion-specific risks persist for decades. Age and weight-for-age at intervention are associated with survival, together with excess mortality beyond childhood underscore the need for lifelong, lesion-tailored surveillance in this growing population.

Graphical Abstract

Introduction

Mild congenital heart diseases (CHD) including atrial septal defect (ASD), ventricular septal defect (VSD), patent ductus arteriosus (PDA) and pulmonary valve stenosis (PS) account for about one-third of adults with CHD and constitute a growing public health concern.^1–5 While these lesions are often considered cured after intervention, emerging evidence suggests long-term increased mortality remains elevated compared to the general population.^6,7 Most existing studies include patients with significant comorbidities, making it difficult to isolate the contribution of the CHD. Rigorous evaluations of long-term outcomes after intervention in otherwise healthy patients with mild CHD are limited. For example, a Danish National Registry study included mostly unrepaired cases, limiting its relevance to post-intervention outcomes.⁴

To address this gap, we used the Pediatric Cardiac Care Consortium (PCCC),⁸ a large, US-based clinical registry to conduct a retrospective cohort study examining long-term survival, risk factors and causes of death for patients with mild CHD after surgical or transcatheter intervention.

Methods

Deidentified data and methodological details, including SAS codes, are available upon reasonable request to the corresponding author; however, National Death Index (NDI) matched data are subject to federal disclosure restrictions and cannot be shared.

Cohort Selection

The Pediatric Cardiac Care Consortium (PCCC) is a multicenter U.S. registry that prospectively collected standardized data on CHD interventions from 47 centers between 1982 and 2011. Details of the creation, activities, and function of the PCCC have been previously described.^6–13

We identified patients undergoing surgical or transcatheter intervention at age <21 years for mild CHD (ASD, VSD, PDA and PS without right ventricular hypoplasia) at a participating US center between January 1, 1982 and April 15, 2003 [before Health Insurance Portability and Accountability Act (HIPAA) restrictions on identifiers].¹⁴ We included all patients with US residency at the time of intervention and available direct identifiers. Patients with prior cardiac interventions at non-PCCC centers, incomplete/conflicting data, or coexisting severe cardiac defects were excluded. ASDs included patent foramen ovale (PFO), ostium secundum and sinus venosus, if no mitral valve pathology was present. Coronary sinus ASDs and primum ASDs were excluded because of their distinct diagnostic and surgical considerations. All types of VSDs were included (inlet, muscular, paramembranous) if there was no significant atrioventricular or semilunar valve pathology. Preterm infants with isolated PDA intervention at <2.5 kg were excluded. Extracardiac comorbidities (ECC) likely to affect long-term survival were flagged for stratified analysis (Table S1).

Each patient was assigned a single primary diagnosis using a previously published severity hierarchy (VSD>PS>ASD>PDA).⁷ When more than one lesion was present, classification followed the most severe diagnosis.

Ascertainment of Outcomes

Eligible patients were matched to the National Death Index (NDI) through 2020, as previously described.^6,13 Follow-up began at discharge after the index procedure and ended at death or December 31, 2020. Underlying and contributing causes of death (COD) were obtained from NDI-Plus using International Classification of Diseases (ICD) codes.^6,13

Statistical Analyses

Continuous variables are reported as median with interquartile range (IQR: Q1, Q3) and categorical variables as counts and percentages. Weight-for-age z-scores (WAZ) at surgery were calculated using WHO Child Growth Standards (WHO Anthro, version 3.2.2) and categorized as “underweight” (≤−2), “normal” (>−2 to <2) or “overweight” (≥2).¹⁵ Chi-Square tests were used for group comparisons.

Time-to-event was analyzed using Kaplan-Meier methods with log-rank tests and Šidák correction for multiple comparisons. Standardized mortality ratios (SMRs) were calculated and reported with 95% confidence intervals (CI) using CDC WONDER data and were stratified by primary diagnosis and age at intervention.^6,7 SMRs were calculated as observed to expected death ratios based on age-, sex-, and calendar year-matched general population U.S rates. SMR trends over time were evaluated using Poisson regression with the log of expected deaths included as an offset term and years after intervention were modeled continuously to assess linear trends. In secondary analyses, SMRs were calculated within age groups at surgery for each lesion in 3-year intervals, and temporal trends were assessed by modeling the interval midpoint as a continuous variable.

Univariable and multivariable Cox proportional hazards (PH) models were used to evaluate associations between patient characteristics and mortality. Center was included as a random effect. PH assumptions were evaluated using Schoenfeld residuals and when violated, models were stratified by the variable that violated the PH assumption to account for non-proportionality. Covariates in the multivariable models were selected based on clinical relevance and/or a univariate p-value <0.20. Lesion-specific models were developed, as well as for left-to-right shunts (ASD, VSD, PDA) and left heart volume overload (VSD, PDA) lesions. Sub-analyses for transcatheter interventions, included interventions after 1987–1989 (dependent by lesion), when such procedures were first recorded in the registry. Given the minimal degree of missing data in the PCCC cohort (e.g., WAZ was missing in 1.7% of observations), multiple imputation was not performed in the multivariable analyses, as complete-case analysis was considered unlikely to meaningfully bias estimates or materially affect precision.

Causes of death (COD) were classified as congenital heart disease (CHD)-related, cardiovascular disease (CVD)-related, and non-CHD/non-CVD. For CVD deaths, additional subcategories were summarized as previously described.⁶ Both underlying and multiple COD were evaluated. Fine-Gray competing risk models accounted for non-CHD/non-CVD deaths as competing events.

Statistical analyses were conducted using SAS 9.4 (SAS Institute, Cary, NC). Statistical significance was assigned as two-sided p<0.05 with adjustment for multiple comparisons when appropriate.

This study was approved by the Emory University Institutional Review Board with waiver of consent for patients enrolled before April 15, 2003.

Results

Cohort description

We identified 17,407 patients who met inclusion criteria. Survival to discharge was 99.3% with in-hospital mortality 1.4% for VSD, 0.7% for PDA and PS and 0.2% for ASD. After excluding 115 in-hospital deaths, 17,252 were included in long-term analyses (Figure 1).

Figure 1.

STROBE-style flow diagram showing selection of study cohort. ASD, atrial septal defect; PDA, patent ductus arteriosus; PS, pulmonary valve stenosis; VSD, ventricular septal defect; PCCC, pediatric cardiac care consortium; CHD, congenital heart disease.

Overall, 57.9% of patients were female and 73.6 % underwent surgical intervention (Table 1). Median age at first intervention was 2.4 years, older for ASD (4.4 years) and younger for VSD (0.8 years). Median follow-up was 24.7 years (IQR: 21.5, 28.7). At intervention, 72.8% had normal WAZ; underweight status was most frequent among those with VSD (48.7%). Chromosomal abnormalities were identified in 8.7% of patients.

Table 1.

Patient Characteristics

	Overall	ASD	VSD	PDA	PS
Number of patients, n	N=17292	N=6178	N=4253	N=4968	N=1893
Surgery *	13197 (76.3%)	5504 (89.1%)	4247 (99.9%)	3087 (62.1%)	359 (19.0%)
Catheterization *	4095 (23.7%)	674 (10.9%)	6 (0.1%)	1881 (37.9%)	1534 (81.0%)
Median age at intervention in years, (IQR: Q1, Q3)*	2.4 (0.8, 5.3)	4.4 (2.6, 8.4)	0.8 (0.4, 2.3)	1.9 (0.9, 4.5)	1.0 (0.1, 3.7)
Age category at first intervention, n (%) *
0–<1 years	5137 (29.7%)	332 (5.4%)	2441 (57.4%)	1415 (28.5%)	949 (50.1%)
1–<5 years	7484 (43.3%)	3171 (51.3%)	1275 (30.0%)	2451 (49.3%)	587 (31.0%)
5–<21 years	4671 (27.0%)	2675 (43.3%)	537 (12.6%)	1102 (22.2%)	357 (18.9%)
Median follow-up in years, (IQR: Q1, Q3)	24.7(21.5, 28.7)	24.8 (21.5, 28.6)	24.7 (21.5, 28.9)	24.7 (21.5, 28.7)	24.8 (21.6, 28.6)
Female sex, n (%) *	10006 (57.9%)	3684 (59.6%)	2112 (49.7%)	3210 (64.6%)	1000 (52.8%)
WAZ at intervention, n (%) *
Z≤−2	3879 (22.4%)	799 (12.9%)	2038 (47.9%)	845 (17.0%)	197 (10.4%)
Z≥2	528 (3.1%)	186 (3.0%)	68 (1.6%)	148 (3.0%)	126 (6.7%)
−2< Z<2	12593 (72.8%)	5094 (82.5%)	2072 (48.7%)	3911 (78.7%)	1516 (80.1%)
Missing	292 (1.7%)	99 (1.6%)	75 (1.8%)	64 (1.3%)	54 (2.9%)
Era of intervention, n (%) *
Early (1982–1989)	2211 (12.8%)	781 (12.6%)	611 (14.4%)	641 (12.9%)	178 (9.4%)
Middle (1990–1999)	10637 (61.5%)	3801 (61.5%)	2603 (61.2%)	3009 (60.6%)	1224 (64.7%)
Late (2000–2009)	4444 (25.7%)	1596 (25.8%)	1039 (24.4%)	1318 (26.5%)	491 (25.9%)
Chromosomal abnormalities, n (%) *	1499 (8.7%)	333 (5.4%)	771 (18.1%)	372 (7.5%)	23 (1.2%)
Extracardiac comorbidities, n (%) *	2679 (15.5%)	642 (10.4%)	1054 (24.8%)	779 (15.7%)	204 (10.8%)

Numbers in parenthesis indicate % by columns, except if otherwise indicated.

^*Indicates statistical significance at p<0.001 by Chi-square test. WAZ, weight-for-age z-score; ASD, atrial septal defect; VSD, ventricular septal defect; PDA, patent ductus arteriosus; PS, pulmonary valve stenosis.

Survival Analysis

Cumulative 35-year survival for all mild CHD was 94.1% (Figure 2A, Table S2). Survival was lower for patients with left-to-right shunts than PS (p=0.001) (Figure 2B–C) and for those with ECC (p<0.001) (Figure 2D). Survival did not differ across ASD subtypes (Figure S1).

Figure 2A-D.

Kaplan-Meier survival curves after intervention for mild CHD. A) Survival in the full cohort (curve, left axis) and deaths (bars, right axis). B) Survival stratified by primary diagnosis group (curve and 95%CI bands). C) Survival stratified by left-right shunt lesions vs PS (curve and 95%CI bands). D) Survival stratified by extracardiac comorbidities (curve and 95%CI bands). ASD, atrial septal defect; PDA, patent ductus arteriosus; PS, pulmonary stenosis; VSD, ventricular septal defect; ECC, extracardiac comorbidities.

In univariable analyses, surgical approach, male sex, age of <1 year and low WAZ at intervention, left-to-right shunt physiology, as well as presence of chromosomal abnormality or ECC were each associated with higher hazard for post-discharge death (Table S3).

Multivariable Cox PH models (stratified by sex and ECC for PH violations), showed that intervention at 1–<5 years was associated with lower hazard of post-discharge death (aHR 0.62, 95%CI 0.50–0.76, p<0.001), whereas low WAZ (Z ≤ −2) was associated with higher hazard (aHR 1.91, 95%CI 1.58–2.31, p<0.001). All left-to-right shunt physiologies had higher hazard of death than PS (Table 2). Competing-risk analyses accounting for non-CHD/non-CVD death as a competing event yielded consistent results (Table S4).

Table 2.

Multivariable analysis of associations between clinical characteristics and post-discharge death by physiological grouping

With all lesions
	aHR	95% CI	P-value
Age at intervention
5 years – 21 years	REF
0 – < 1 year	0.85	0.68–1.07	0.16
1 year – < 5 years	0.62	0.50–0.76	<0.001
WAZ at intervention
−2 < Z < 2	REF
Z ≤ −2	1.91	1.58–2.31	<0.001
Z ≥ 2	1.34	0.82–2.19	0.24
Era of intervention
2000–2009		REF
1982–1989	1.00	0.75–1.34	0.98
1990–1999	1.13	0.92–1.40	0.25
Chromosomal abnormality (Yes vs No)	1.15	0.88–1.49	0.31
Lesions
PS	REF
Left–to–Rightshunt	1.44	1.05–1.99	0.027
By physiology
Age at intervention
5 years – 21 years		REF
0 – < 1 year	0.86	0.68–1.09	0.20
1 year – < 5 years	0.62	0.50–0.76	<0.001
WAZ at intervention
−2 < Z < 2	REF
Z ≤ −2	1.91	1.57–2.32	<0.001
Z ≥ 2	1.34	0.82–2.19	0.24
Era of intervention
2000–2009	REF
1982–1989	1.01	0.75–1.34	0.98
1990–1999	1.13	0.92–1.40	0.25
Chromosomal abnormality (Yes vs No)	1.15	0.88–1.49	0.31
Lesions
PS	REF
Right heart volume overload (ASD)	1.46	1.03–2.06	0.032
Left heart volume overload (PDA/VSD)	1.44	1.04–1.99	0.027

Cox proportional hazards analysis is stratified by sex and extracardiac comorbidity due to violation of the proportional hazard assumption. In stratified Cox models, hazard ratios for the stratification variable are not estimated and therefore not presented in the tables. WAZ, weight-for-age z-score; ASD, atrial septal defect; VSD, ventricular septal defect; PDA, patent ductus arteriosus; PS, pulmonary valve stenosis.

Lesion-specific models revealed heterogeneity across individual lesions. The survival benefit of intervention at 1–<5 years among shunt lesions was largely driven by ASD (aHR 0.48, 95%CI 0.35–0.64, p<0.001); age at intervention was not associated with hazard for VSD, while intervention before 1 year was associated with higher hazard for PDA (aHR 1.61, 95%CI 1.06–2.44, p=0.024). For PS, intervention before 5 years was associated with lower hazard (Table 3). Low WAZ was associated with increased hazard across all shunt physiologies (ASD, VSD, PDA), but not PS. Excluding PFOs did not change ASD findings (Table S5).

Table 3.

Multivariable analysis of associations between clinical characteristics and post-discharge death by individual lesion

Atrial Septal Defect (ASD)	aHR	95% CI	P-value
Age at intervention
5–<21 years	REF
0–<1 year	0.78	0.48–1.28	0.33
1–<5 years	0.48	0.35–0.64	<0.001
WAZ at intervention
−2<Z< 2	REF
Z≤–2	2.9	2.09–4.03	<0.001
Z≥2	2	1.02–3.92	0.044
Era of intervention
2000–2009	REF
1982–1989	0.65	0.37–1.12	0.12
1990–1999	0.92	0.64–1.33	0.67
Chromosomal abnormality (Yes vs No)	1.13	0.68–1.87	0.63
Ventricular Septal Defect (VSD)
Age at intervention
5–<21 years	REF
0–<1 year	1.09	0.63–1.89	0.75
1–<5 years	1.11	0.64–1.92	0.71
WAZ at intervention
−2<Z< 2	REF
Z≤–2	1.39	1.01–1.90	0.039
Z≥2	2.07	0.76–5.65	0.16
Era of intervention
2000–2009	REF
1982–1989	1.17	0.72–1.89	0.54
1990–1999	1.02	0.69–1.49	0.94
Chromosomal abnormality (Yes vs No)	1.09	0.68–1.76	0.72
Patent ductus arteriosus (PDA)
Age at intervention
5–<21 years	REF
0–<1 year	1.61	1.06–2.44	0.024
1–<5 years	0.76	0.50–1.16	0.2
WAZ at intervention
−2<Z< 2	REF
Z≤–2	2.09	1.50–2.92	<0.001
Z≥2	0.8	0.26–2.52	0.71
Era of intervention
2000–2009	REF
1982–1989	1.13	0.67–1.90	0.65
1990–1999	1.41	0.94–2.07	0.078
Chromosomal abnormality (Yes vs No)	1.32	0.83–2.09	0.24
Pulmonary Valve Stenosis (PS)
Age at intervention
5–<21 years	REF
0–<1 year	0.41	0.20–0.83	0.013
1–<5 years	0.36	0.16–0.81	0.014
WAZ at intervention
−2<Z< 2	REF
Z≤–2	1.69	0.68–4.17	0.26
Z≥2	0.21	0.03–1.76	0.15
Era of intervention
2000–2009	REF
1982–1989	0.71	0.15–3.26	0.66
1990–1999	1.57	0.60–4.09	0.36
Chromosomal abnormality (Yes vs No)	12.29	3.40–44.40	<0.001
Sex (Male vs Female)	2.39	1.19–4.81	0.015
Extracardiac comorbidities (Yes vs No)	2.33	0.95–5.73	0.064

Cox proportional hazards models are stratified by sex and extracardiac comorbidity due to violation of the proportional hazard assumption. In stratified Cox models, hazard ratios for the stratification variable are not estimated and therefore not presented in the tables. WAZ, weight-for-age z-score.

In years with overlapping availability of surgical and transcatheter interventions, unadjusted survival favored transcatheter repair only for PDA (p=0.005) but not for ASD or PS (Figure S2); adjusted models showed no survival differences by intervention approach across lesions (Table S6).

SMRs for shunt lesions exceeded general population rates up to 20 years after intervention, with persistent excess risk beyond 20 years for VSD, whereas SMRs for PS were only modestly elevated all along (Figure 3 A–D). After excluding ECC, excess mortality was attenuated across all lesions and was no longer significant for PS (Figures S3). SMRs declined over time for ASD, PDA, VSD (p-trend<0.01) but not for PS (p-trend=0.92). SMR trends by age at intervention demonstrated lesion-specific patterns (Figure S4) with excess mortality declining for ASD patients repaired at 1–<5 years, across all age groups for PDA and for VSD repaired before 5 years. For PS, repair before 5 years was associated with general population-level mortality, whereas later repair exhibiting persistent excess risk.

Figure 3A-D.

Standardized Mortality Ratios (SMR) (1982–2019) for patients with A) Atrial septal defect (ASD), B) Patent ductus arteriosus (PDA), C) Ventricular septal defect (VSD) and D) Pulmonary valve stenosis (PS). SMRs with 95% confidence intervals are provided for all patients including those with extracardiac comorbidities using age-, sex-, and year-matched CDC WONDER data. SMRs for all patients are higher than the general population until an average of 20 years of age. SMRs may increase again for those with VSD interventions.

Causes of Death (COD)

Among 679 deaths, 18.7 % were attributed to CHD and 36.7 % to CVD; 56.0% of deaths (56.0%) were due to non-CHD/non-CVD causes (Table 4). CHD-related deaths were most frequent for VSD (32.4 %) and least for ASD (10.6 %). VSD had higher frequency of heart failure (7.8%) and pulmonary hypertension (14.2%). Arrhythmic deaths were more frequent in patients with VSD and PS (6.4%).

Table 4.

Any causes of death by primary diagnosis

	Total	ASD	VSD	PDA	PS
Deaths, n (%)	N=679	N=208	N=219	N=205	N=47	P-value
CHD	127 (18.7%)	22 (10.6%)	71 (32.4%)	29 (14.1%)	5 (10.6%)	<0.001
CVD	249 (36.7%)	70 (33.7%)	96 (43.8%)	68 (33.2%)	15 (31.9%)	0.073
Endocarditis	4 (0.6%)	1 (0.5%)	2 (0.9%)	1 (0.5%)	0 (0.00%)	0.86
Ischemia	16 (2.4%)	6 (2.9%)	7 (3.2%)	2 (1.0%)	1 (2.1%)	0.45
Pulmonary hypertension	67 (9.9%)	11 (5.3%)	31 (14.2%)	24 (11.7%)	1 (2.1%)	0.009
Cardiac arrest	84 (12.4%)	27 (13.0%)	30 (13.7%)	20 (9.8%)	7 (14.9%)	0.57
Arrhythmia	26 (3.8%)	6 (2.9%)	14 (6.4%)	3 (1.5%)	3 (6.4%)	0.079
Heart failure	47 (6.9%)	12 (5.8%)	17 (7.8%)	15 (7.3%)	3 (6.4%)	0.86
Miscellaneous CVD	3 (0.4%)	0 (0.00%)	1 (0.5%)	2 (1.0%)	0 (0.00%)	0.48
Non-CHD/Non-CVD	380 (56.0%)	130 (62.5%)	96 (43.8%)	124 (60.5%)	30 (63.8%)	<0.001

P-values were calculated by Chi-square or Fisher’s exact test. Causes of death are not age-adjusted. CHD, congenital heart disease; CVD, cardiovascular disease; ASD, atrial septal defect; VSD, ventricular septal defect; PDA, patent ductus arteriosus; PS, pulmonary valve stenosis.

Discussion

Long-term Survival and Comparison to the General Population

In this large cohort, post-discharge survival after intervention for mild CHD was approximately 95% at 35-years. Survival varied by lesion and was lower in patients with ECC or indicators of more severe physiology, including lower weight or younger age at intervention. SMRs demonstrated excess mortality for up to two decades after intervention, most pronounced for VSD and PDA, likely reflecting greater hemodynamic burden from left heart volume overload compared to right heart volume (ASD) or pressure (PS) overload. In contrast, excess risk for PS was largely attributable to extracardiac comorbidity rather than the lesion itself. These findings align with prior European studies reporting long-term excess mortality even in lesions often excluded from adult CHD specialty follow-up.^2,3,16–20

Weight-for-age and Long-term Risk

Although prior studies have linked low weight at surgery to early postoperative mortality,^21,22 our findings indicate that low WAZ also signals long-term vulnerability for left-to-right shunts. This likely reflects more severe hemodynamic burden, delayed referral, and/or suboptimal preoperative management rather than poor nutrition alone. While weight improvement may not fully mitigate risk from severe physiology, timely intervention and preoperative optimization may offer opportunities to improve outcomes. Overweight status at intervention for ASD was also associated with increased hazard, possibly reflecting emerging cardiometabolic risk.

Timing of Intervention, Procedural Complexity and Lesion-specific Considerations

Age at intervention demonstrated lesion-specific associations, reflecting a balance between early relief of physiological burden early and the higher procedural risk associated with younger age. Differences between hazard ratios and SMRs highlight complementary perspectives, distinguishing within-lesion risk from excess mortality relative to the general population.

Indications for intervention and procedural complexity are also important considerations. Earlier intervention for shunt lesions is often prompted by symptoms or hemodynamic compromise, whereas later repair more commonly reflects milder physiological burden and elective indications.²³ Our findings suggest that long-term survival is more strongly associated with underlying lesion physiology and patient characteristics than with procedural complexity. Consistent with this, adjusted analyses demonstrated no significant survival differences between surgical and transcatheter repair for ASD, PDA, or PS.

For ASD, a lesion characterized by right heart volume overload and generally lower procedural complexity, intervention at 1-<5 years of age was associated with both the lowest hazard of death and lowest SMR, suggesting an optimal balance at this age between mitigating physiologic burden and minimizing procedural risk.

For VSD, a lesion with left heart volume overload and higher procedural complexity, age at intervention was not associated with hazard of death, though SMRs were lowest when repair occurred after 5 years of age, consistent with lower baseline disease severity and lower procedural complications.

PS, a lesion characterized by RV pressure overload and low procedural complexity, exhibited a distinct pattern: intervention before 5 years of age was associated with both lower hazard of death and population-level SMRs, consistent with safe early interventions and adverse effects of prolonged right ventricular pressure overload.²⁴

Interpretation was more challenging for PDA, a lesion causing left heart volume overload and with low procedural complexity across ages. Both hazard of death and SMRs were highest among infants with intervention before 1 year of age. This pattern may reflect adverse consequences of exposure to excessive continuous left-to-right shunting or greater baseline vulnerability not fully captured in our dataset.

Strengths and Limitations

This study draws on the PCCC, a large, multicenter registry with rich clinical data,⁸ enabling long-term analyses of survival, mortality risk factors, and COD over 35 years. Linkage with CDC vital statistics and NDI records allowed mortality assessment and comparison with the general population within a unified analytic framework.

Limitations include potential incomplete capture of some diagnoses and comorbidities and the absence of granular data on defect size, shunt magnitude, and clinical status at intervention. The lesion severity hierarchy used reflects convention rather than a strict prognostic scale, though aggregate outcomes as with other large datasets likely approximate average severity.^4,7,19,25,26 Timing of intervention may be confounded by evolving practices and indication bias, an inherent limitation to observational studies; nevertheless, our findings reflect real-world practice patterns.^26–28

Residual lesions, reinterventions, and socioeconomic factors were not captured and extracardiac comorbidity may be incompletely ascertained in this dataset. NDI linkage was unavailable after 2003 due to HIPAA restrictions; however, limited follow-up for later patients would have contributed minimally to long-term estimates. Mortality ascertainment, though high for the PCCC-NDI dataset (88.1%), was incomplete and may result in conservative mortality estimates.¹³ COD assignment based on death certificates may misclassify remote CHD or comorbidities. Finally, as an observational study, causal inference is not possible, and residual confounding cannot be excluded.

Conclusion

In this large cohort with extended follow-up, long-term survival after intervention for mild CHD was excellent, yet lesion- and age-specific risks persisted for decades. Timing of intervention should be individualized based on lesion physiology, patient size, and procedural risk. The PCCC experience from early and intermediate eras of transcatheter interventions supports catheter-based approaches when feasible. Collectively, these findings challenge the perception that mild CHD lesions are “cured” by early intervention and highlight the need for lifelong surveillance given persistent postoperative risk.^29,30

What is Known?

• Surgical and transcatheter interventions for mild congenital heart diseases (CHD) such as ASD, VSD, PDA, and PS are common in the general population.

• Short- and medium-term outcomes after intervention for mild CHD are excellent with apparently normal clinical status and longevity for most patients.

What the Study Adds

• Late survival after intervention for mild CHD is lower compared to the general population for up to two decades after the intervention.

• Long-term risk varies by lesion type and is influenced by underlying physiology, while weight-for-age z-score, comorbidities, and age at intervention show strong associations with survival, highlighting important lesion-specific and patient-specific risk profiles.

Non-standard Abbreviations and Acronyms

ASD

Atrial Septal Defect

CDC

Centers for Disease Control and Prevention

CHD

Congenital Heart Disease

COD

Causes of Death

ECC

Extracardiac Comorbidities

NDI

National Death Index

PCCC

Pediatric Cardiac Care Consortium

PDA

Patent Ductus Arteriosus

PH

Proportional hazards

PS

Pulmonary Valve Stenosis

VSD

Ventricular Septal Defect