MORGAS ( MOR tality OR iented G lucose, A ge, S moking, and S tent S core) predicts mortality after percutaneous coronary intervention

BACKGROUND: Accurate prediction of mortality after percutaneous coronary intervention (PCI) remains a clinical challenge. Existing risk scores often lack specificity or require complex variables, limiting bedside applicability. This study introduces MORGAS (MORtality ORiented Glucose, Age, Smoking, and Stent score), a novel predictive model for post-PCI mortality. METHODS: Data from 573 PCI patients in a derivation cohort and 566 patients in an external validation cohort were analysed. Feature selection combined Extreme Gradient Boosting and LASSO regression, identifying four key predictors: age, fasting glucose, smoking status, and number of stents. A logistic regression model incorporating polynomial and interaction terms was developed. Model performance was assessed using area under the curve (AUC), calibration plots, Brier scores, and decision curve analysis. RESULTS: In the derivation cohort, MORGAS achieved an AUC of 0.766, sensitivity 75.6%, specificity 71.2%, and negative predictive value 97.4%. Internal validation confirmed stability (AUC 0.726). External validation showed moderate discrimination (AUC 0.656) with preserved clinical utility across relevant risk thresholds. Calibration analysis indicated slight overestimation at higher predicted risks. CONCLUSIONS: MORGAS is a simple, clinically applicable tool integrating four readily available variables to predict mortality after PCI. Its performance and decision-analytic utility suggest potential for guiding individualised post-PCI care and improving risk stratification in diverse clinical settings.