Hospital Surgical Volume and Regional Disparities in Congenital Heart Surgery Outcomes: Analysis of Korean National Health Insurance Claims Data, 2002-2021

Background and Objectives: The volume-outcome relationship in congenital heart surgery (CHS) has been widely reported internationally, but systematic nationwide evidence from Korea remains limited. Given the concentration of high-volume centers in the Seoul Capital Area (SCA), we aimed to examine whether hospital surgical volume was associated with short-term mortality and to what extent regional disparities could be explained by differences in surgical volume. Materials and Methods: We conducted a nationwide retrospective cohort study of 31,150 patients who underwent CHS in 91 hospitals in Korea between 2002 and 2021 using National Health Insurance claims data. Hospitals were classified by location (SCA vs. non-SCA). Annual surgical volume was defined using two approaches, (i) above vs. below the overall mean annual volume (17.1 cases per hospital), and (ii) three categories (≤20, 21-40, and >40 cases/year). The primary outcome was 30-day mortality. Multivariable logistic regression adjusted for case mix, including J-STAT category, sex, hospital type, age, prematurity, and low birth weight. Hospital-level variation was further evaluated using generalized linear mixed models with random hospital intercepts, and intraclass correlation coefficients (ICCs) were estimated to quantify between-hospital variation and the explanatory contribution of surgical volume. Results: Overall 30-day mortality was 1.99%, with higher mortality observed in non-SCA hospitals compared with SCA hospitals (3.19% vs. 1.57%). After adjustment, lower annual surgical volume was strongly associated with higher 30-day mortality. Compared with hospitals performing >40 cases/year, the adjusted odds ratios were 4.13 (95% CI, 3.30-5.17) for hospitals performing 21-40 cases/year and 4.95 (95% CI, 3.98-5.95) for those performing ≤20 cases/year. In multilevel analyses, annual surgical volume accounted for 54% of the between-hospital variation in 30-day mortality. Adjustment for surgical volume substantially attenuated the regional disparity, with the odds ratio for non-SCA versus SCA hospitals decreasing from 2.12 (95% CI, 1.80-2.49) to 1.14 (95% CI, 0.95-1.37). Conclusions: A strong volume-outcome relationship exists in congenital heart surgery in Korea, with excess mortality concentrated in low-volume hospitals rather than regional location itself. Regional disparities in outcomes appear largely attributable to the uneven distribution of surgical volume. Strategies focused on service consolidation and strengthened referral to high-volume centers may be effective in reducing inter-hospital variation and improving national outcomes in pediatric cardiac surgery.