Prevalence of pulmonary arterial hypertension associated with congenital heart disease following defect correction: a systematic review and meta-analysis?

Importance Reported pulmonary arterial hypertension (PAH) or pulmonary hypertension (PH) after congenital heart defect repair is clinically important, but its frequency and ascertainment remain uncertain. Objective In this study we sought to estimate the prevalence of pulmonary arterial hypertension (PAH) following defect correction in patients with congenital heart disease (CHD). Data sources We systematically searched PubMed, Embase and the Cochrane Library from 1 January 1970 to 31 August 2023. Full database-specific strategies are provided in Supplementary Appendix 1. Study selection Eligible longitudinal observational studies reported PAH or PH at least 3 months after complete congenital heart defect repair. Two reviewers independently screened records and extracted data. Data extraction and synthesis All studies meeting eligibility criteria were included irrespective of observed prevalence. Random-effects meta-analysis used the Freeman–Tukey double-arcsine transformation and DerSimonian–Laird estimator. Prespecified reviewer-driven sensitivity analyses examined outcome ascertainment, follow-up, risk of bias, and the former post hoc 52-study restriction. Main outcomes and measures The prevalence of PAH in patients following CHD defect correction. Results Fifty-eight studies including 7,044 participants were analysed. The random-effects average prevalence of reported post-repair PAH or PH was 11.4% (95% CI 7.7–15.6; I² = 95.7%; 95% prediction interval 0–51.5). In 15 studies requiring RHC confirmation, prevalence was 7.3% (95% CI 2.2–14.7); in 20 echocardiography-defined studies it was 20.6% (95% CI 11.7–31.1). Adult, paediatric and mixed-age estimates were 18.5%, 8.7% and 7.5%, respectively. The former post hoc restricted 52-study analysis yielded 7.6% (95% CI 4.8–10.9). Higher study-level preoperative PH prevalence remained associated with higher post-repair prevalence in a parsimonious model. Conclusion and relevance Reported post-repair PAH or PH affects a clinically important minority, but the pooled value varies substantially by population and ascertainment and is not a universal risk estimate. Standardised haemodynamic definitions and longitudinal studies distinguishing persistent from incident disease are needed.