Multivariate analysis and individualized nomogram construction for predicting radial artery occlusion risk after...

To identify independent risk factors for radial artery occlusion (RAO) after coronary angiography (CAG) and percutaneous coronary intervention (PCI), and to develop risk prediction models for CAG, PCI, and the overall population.

This retrospective study included 781 patients undergoing CAG or PCI. RAO occurrence was recorded. Baseline characteristics, intraoperative factors, and laboratory indicators were collected. Variables were screened using univariate logistic regression and LASSO regression. Independent risk factors were identified via multivariate logistic regression to develop nomogram prediction models. Model performance was evaluated using receiver operating characteristic curves, calibration curves, and decision curve analysis.

The overall RAO incidence was 8.32%, significantly higher in the CAG group than the PCI group (11.26% vs. 5.64%, P=0.007). Independent risk factors for RAO in the CAG group were female, elevated creatine kinase, multiple punctures, radial artery spasm, and puncture site pain, with BMI ≥25 kg/m² as a protective factor; in the PCI group, heart failure, multiple punctures, and puncture site pain, with elevated creatinine as a protective factor; in the overall population, smoking, heart failure, high CHA₂DS₂VASc score, multiple punctures, radial artery spasm, and puncture site pain, with BMI ≥25 kg/m² and elevated creatinine as protective factors. The prediction models demonstrated good discriminatory ability in each group (AUCs: 0.979, 0.921, and 0.951, respectively) and clinical utility.

Risk prediction models for RAO established in patients undergoing CAG, PCI, and the overall population can accurately identify high-risk patients, providing a reference for individualized clinical prevention and treatment. Multiple punctures, radial artery spasm, and puncture site pain are significant, modifiable risk factors.