Prevalence And Spectrum of Congenital Heart Disease in Nepal: A Systematic Review and Meta-Analysis

BackgroundThe true population burden of congenital heart disease in Nepal remains uncertain due to the blending of community- and hospital-based data. We aimed to synthesize congenital heart disease prevalence estimates, quantify heterogeneity, and explore the impact of the study setting and design on reported rates. MethodsFollowing PRISMA guidelines, we searched PubMed, Embase, ScienceDirect, and Nepal-specific journals through May 1, 2025 (PROSPERO CRD420251125950). Eligible observational studies reported an extractable numerator of congenital heart disease cases and a denominator within community/school, neonatal (births or consecutive admissions), or hospital/referral (including consecutive echocardiography) frames. Pooled prevalence used binomial GLMM on the logit scale (PLOGIT; random effects; method.tau = ML); fallback was inverse-variance PLOGIT with REML and Hartung-Knapp CIs. We reported 95% CIs, 95% prediction intervals (PIs), and I^2/{tau}^2. The prespecified subgroups were community (CO) vs. hospital/referral (H) and Kathmandu Valley vs. sub-urban/mixed. Lesion outcomes (ASD, VSD) were pooled as shares among CHD and population prevalence; studies with multi-lesion reporting (>100% summed percentages) were excluded from pooling. Exploratory meta- regression (Knapp-Hartung) evaluated log{square}{square}(sample size) ResultsNine studies (N = 79,595; CHD cases = 462) met the inclusion criteria. Overall pooled prevalence was 8.9 per 1,000 (95% CI 3.0 - 26.1; PI 0.2 - 338.9), with I^2 {approx} 99% and {tau}^2 {approx} 2.76 (logit). Hospital/referral series showed higher descriptive prevalence than community (25.64 vs. 3.81 per 1,000), but in meta-regression, the hospital effect attenuated and was not significant after adjusting for log{square}{square}(n) (OR {approx} 1.32, 95% CI 0.21 - 8.15). Larger studies reported a lower prevalence (OR per 10-fold increase {approx} 0.10, 95% CI 0.04 - 0.27). By geography, the pooled prevalence was 5.69/1,000 in Kathmandu Valley versus 15.46/1,000 in suburban/mixed settings. Lesion patterns differed by setting: Atrial Septal Defect share was higher in the community and Ventricular Septal Defect share was higher in hospitals. Population-scale estimates were ASD 1.18/1,000 (CO) vs 4.93/1,000 (H) and VSD 0.38/1,000 (CO) vs 6.68/1,000 (H); all lesion analyses showed wide uncertainty. ConclusionsThe prevalence of Congenital Heart Disease in Nepal is highly context-dependent and dominated by the frame and study size. Hospital/referral data are not generalizable to the population prevalence. Community-based estimates are more informative for planning. Given the extreme heterogeneity and wide PIs, the pooled figures should be interpreted with caution. The risk of bias was moderate to high, and the certainty was very low.