Back

Regional Variation after Cardiac Surgery in Adults with Congenital Heart Disease in the US

Diaz-Castrillon, C. E.; Morell, V.; Smith, K.; Viegas, M.; Castro-Medina, M.; Sultan, I.; Hoskoppal, A.; Wang, Y.; Nelson, J. S.

2025-12-19 epidemiology
10.64898/2025.12.17.25342525 medRxiv
Show abstract

IntroductionWe describe regional variation in perioperative care among adults with congenital heart disease (ACHD) undergoing cardiac surgery in the United States. MethodsAnalysis using the 2016-2020 Nationwide Inpatient Sample, including admissions of [≥]18 years old patients with congenital heart ICD-10 diagnosis codes and cardiac surgical procedure codes. Outcomes included hospitalization costs, length of stay, high-resource utilization (HRU), and in-hospital mortality. Adjusted multilevel regressions estimated observed-to-expected (O/E) ratios and risk-standardized (RS) outcomes. ResultsAmong 83,800 admissions, the median age was 59 years (IQR 48-67) and the most common diagnosis was congenital aortic insufficiency (AI) with 58.7% (n=49,149). Elective admissions represented 76.5% (n=64,115) admissions; valve replacements were coded in 68.1% (n=57,065) admissions, followed by CABG (22.0%, n=18,420). Regional distribution was 21% West, 20% Northeast, 27% Midwest, and 31% South. Midwest had a higher proportion of valve repairs whereas Northeast had a higher valve replacement procedure. RS-HRU rate raged from 14.8% (95% CI 13.0-16.7) in the Midwest to 20.4% (18.5-22.3) in the South. RS-costs varied from $49,303 (95% CI 47,944-50,665) in the Northeast to $59,899 (95% CI 58,432-61,355) in the West. In-hospital mortality did not significantly differ by region. Sensitivity analysis excluding congenital AI admissions showed increased resource utilization and mortality. ConclusionSubstantial regional variation exists in ACHD resource utilization. Future studies will focus on specific procedure types and hospital characteristics to uncover potential region-specific strategies to optimize ACHD care as healthcare moves toward value-based models aligned with regional demands.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.