Predicting long-term mortality after first coronary revascularization: – the Kyoto model –

Area of Science:

• Cardiology
• Medical Statistics
• Health Informatics

Background:

• Mortality prediction is crucial for guiding treatment decisions in patients undergoing coronary revascularization.
• Percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) are common revascularization procedures with varying outcomes.
• Developing accurate predictive models is essential for personalized patient care and risk stratification.

Purpose of the Study:

• To develop and validate the Kyoto model for predicting mortality after PCI and CABG.
• To identify key determinants of mortality in patients undergoing coronary revascularization.
• To provide a tool for individualized decision-making and patient guidance.

Main Methods:

• Analysis of 9,393 patients from the CREDO-Kyoto Registry undergoing their first coronary revascularization.
• Application of Cox regression models to identify mortality predictors for PCI and CABG separately.
• Internal validation using 10-fold cross-validation and assessment with the Hosmer-Lemeshow test and C statistics.

Main Results:

• The Kyoto model identified numerous predictors of mortality, including age, comorbidities (diabetes, kidney disease), and disease characteristics (vessel disease, left main disease).
• The model demonstrated good predictive accuracy, with C statistics greater than 0.78.
• Simulated 3-year mortality rates varied based on patient characteristics and revascularization method, highlighting individualized risk.

Conclusions:

• The Kyoto model serves as a valuable tool for clinicians and patients in managing post-revascularization care.
• It supports adherence to medication and lifestyle modifications.
• The model facilitates individualized decision-making for patients undergoing PCI or CABG.