This prospective four-center study evaluated a preoperative risk model in 538 patients aged ≥16 years undergoing congenital cardiac surgery. In-hospital death, ventilation exceeding seven days, or new dialysis occurred in 23 patients (4.3%). External validation in 383 complete cases showed discrimination of AUC 0.729 (95% confidence interval 0.602–0.855), but the original model systematically overestimated risk. Bootstrap-based model updating yielded an AUC of 0.799 without procedural complexity and 0.857 with the Adult Congenital Heart Surgery procedural mortality score, suggesting potential value from combining patient and procedural factors.
The original model distinguished risk but overestimated absolute event probabilities. Updated models incorporating procedural complexity are promising for perioperative risk assessment, rather than evidence of improved clinical outcomes.
Study limitation: Only 23 composite events occurred, and updated-model performance was assessed through bootstrap-based updating rather than a separate external cohort.
AI-generated from the published abstract with automated source checks. For research discovery; consult the original paper before applying findings.