Comparing the Predictive Power of Preoperative Risk Assessment Tools to Best Predict Major Adverse Cardiac Events in...

Area of Science:

• Cardiology
• Nephrology
• Transplant Surgery

Background:

• Kidney transplant recipients face elevated cardiovascular event risk due to pre-existing conditions like hypertension and end-stage renal disease.
• Accurate preoperative risk stratification is crucial for this high-risk patient population.

Purpose of the Study:

• To compare the predictive accuracy of three risk assessment models for major adverse cardiac events (MACE) in kidney transplant recipients.
• To identify the optimal model for preoperative cardiovascular risk assessment in this population.

Main Methods:

• Comparison of the PORT, RCRI, and Gupta risk assessment models using generalized U-statistics to evaluate the area under the receiver operator curve (AUC).
• Assessment of MACE prediction at 30 days and 1 year post-transplant.
• Exploration of adding novel covariates to the best-performing model to enhance predictive accuracy.

Main Results:

• The PORT model demonstrated superior performance, particularly at 1 year post-transplant, compared to the Gupta model (AUCs 0.650 vs. 0.557).
• At 95% sensitivity, the PORT model showed significantly higher specificity (0.227) than RCRI (0.071) and Gupta (0.08).
• Adding new covariates to the PORT model improved AUC from 0.664 to 0.703, though this enhancement was not statistically significant.

Conclusions:

• The PORT model is the most effective among the evaluated calculators for predicting MACE in kidney transplant patients, especially at clinically relevant sensitivity levels.
• The superior performance of the PORT model is attributed to its inclusion of variables specific to the transplant population.