Risk Prediction in Multivalve Surgery: TRISCORE vs. EuroSCORE II in Combined Left-Sided Valve Surgery with Concomitant Tricuspid Intervention

All articles of this category (opens in new window) Background: The TRISCORE was developed to predict in-hospital mortality in patients undergoing isolated tricuspid valve surgery and demonstrated superior predictive accuracy compared with the EuroSCORE II. However, no validated risk model exists for patients undergoing combined left-sided valve surgery with concomitant tricuspid intervention, a high-risk surgical cohort. This study aimed to evaluate and compare the predictive performance of EuroSCORE II and TRISCORE in this population. Methods: Between January 2018 and September 2024, 134 consecutive patients undergoing combined left-sided valve surgery, including the tricuspid valve, at a single tertiary center were retrospectively analyzed. Patients with active endocarditis, prior cardiac or transcatheter procedures, or concomitant nonvalve surgery were excluded. In-hospital mortality was the primary endpoint. For each patient, the EuroSCORE II and TRISCORE were calculated. Predictive accuracy was assessed by receiver operating characteristic (ROC) curve analysis, and the area under the curve (AUC) with 95% confidence intervals was determined. Results: Patients had a mean age of 72 ± 7 years, 60% were male, and the mean LVEF was 53%. Most were presented in NYHA classes III to IV (87%), with mitral regurgitation as the predominant lesion (97%). Mitral valve repair or replacement was performed in 126 patients and aortic valve replacement in 23, via sternotomy in 85 and transaxillary access in 49. Median cardiopulmonary bypass and cross-clamp times were 100 minutes (IQR: 86–122) and 76 minutes (IQR: 65–92). In-hospital mortality occurred in 11 patients (8.2%), stroke in 2 (1.5%), re-exploration for bleeding in 16 (11.9%), and reintubation in 7 (5.2%). Median ICU stay was 3 days (IQR: 1–6) and hospital stay 13 days (IQR: 9–16). Risk prediction analysis showed an AUC of 0.740 (95% CI: 0.599–0.881, SE 0.072) for TRISCORE vs. 0.667 (95% CI: 0.497–0.836, SE 0.087) for EuroSCORE II, indicating superior discriminative performance of TRISCORE, though both models demonstrated only moderate predictive power. Conclusion: EuroSCORE II and TRISCORE both provide clinically useful risk stratification in patients undergoing combined left-sided valve surgery, including the tricuspid valve. TRISCORE showed higher predictive accuracy for in-hospital mortality and may better capture the risk of this complex cohort. However, given the moderate overall discrimination of both scores, further refinement or integration with additional clinical parameters appears warranted to improve the precision of risk assessment in multivalve surgery. Publication History Article published online: 17 February 2026 © 2026. Thieme. All rights reserved. Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany