Risk Score for Long-Term Survival and Major Adverse Cardiovascular and Cerebrovascular Events After Coronary Artery...

Area of Science:

• Cardiovascular Surgery
• Medical Informatics
• Public Health

Background:

• Coronary artery bypass grafting (CABG) is a critical procedure for managing cardiovascular disease.
• Predicting long-term survival and major adverse cardiovascular and cerebrovascular events (MACCEs) after CABG is essential for patient management and resource allocation.
• Existing risk stratification models may not fully capture all relevant predictors or perform optimally across diverse populations.

Purpose of the Study:

• To develop and validate novel risk scoring models for predicting long-term all-cause mortality and MACCEs following isolated CABG.
• To identify key preoperative predictors of adverse outcomes, including demographic factors.
• To assess the performance of these models through internal and external validation.

Main Methods:

• A retrospective analysis of 4,821 patients undergoing isolated CABG between January 2005 and July 2021 was conducted.
• Double-selection logit least absolute shrinkage and selection operator with adaptive selection was used for variable selection.
• Model performance was evaluated using calibration, accuracy, and area under the curve (AUC) via bootstrap cross-validation and external validation with 665 patients.

Main Results:

• Preoperative risk factors were identified as significant predictors for both all-cause mortality and MACCEs.
• African-American ethnicity emerged as a significant predictor for MACCEs post-CABG.
• Internal validation showed AUCs for mortality ranging from 79.1% to 81.3% and for MACCEs from 72.5% to 75% across 1-5 years.
• External validation yielded AUCs above 70% for both outcomes, demonstrating generalizability.

Conclusions:

• The developed Advanced (AD) Coronary Risk Score effectively predicts long-term mortality and MACCEs after isolated CABG.
• The models demonstrate good discrimination and calibration, supported by robust internal and external validation.
• Incorporating factors like ethnicity enhances the accuracy of risk prediction in CABG patients.