Observed versus predicted mortality after isolated tricuspid valve surgery

Area of Science:

• Cardiac Surgery
• Risk Stratification
• Tricuspid Valve Surgery

Background:

• Evaluating the performance of established risk scores in specific surgical contexts is crucial for accurate patient outcomes prediction.
• Isolated tricuspid valve surgery presents unique challenges in risk assessment.

Purpose of the Study:

• To compare the predictive accuracy of the Clinical Risk Score (CRS) and the European System for Cardiac Operative Risk Evaluation (EuroSCORE)-II for isolated tricuspid surgery.
• To assess the ability of these scores to predict both short-term (30-day) and long-term mortality.

Main Methods:

• A cohort of 383 patients undergoing isolated tricuspid surgery was analyzed.
• Receiver operating characteristic (ROC) analysis was employed to evaluate the discriminative ability of each score.
• Multivariate analysis was performed to identify predictors of late cardiac death.

Main Results:

• The Clinical Risk Score (CRS) showed a higher area under the curve (0.7) compared to EuroSCORE-II (0.6) for predicting 30-day mortality.
• EuroSCORE-II tended to underestimate mortality (expected/observed ratio: 0.46-0.6).
• The CRS score was identified as a significant predictor of late cardiac death (p=0.005).

Conclusions:

• Both CRS and EuroSCORE-II can provide a range for expected mortality in tricuspid surgery.
• The CRS score appears more suitable for predicting late survival following isolated tricuspid surgery.