A nomogram predicting 30-day mortality in patients undergoing percutaneous coronary intervention

BACKGROUND AND AIMS: Early detection of mortality after percutaneous coronary intervention (PCI) is crucial, whereas most risk prediction models are based on outdated cohorts before the year 2000. This study aimed to establish a nomogram predicting 30-day mortality after PCI. MATERIALS AND METHODS: In total, 10,444 patients undergoing PCI in National Center for Cardiovascular Diseases in China were enrolled to establish a nomogram to predict 30-day mortality after PCI. The nomogram was generated by incorporating parameters selected by logistic regression with the stepwise backward method. RESULTS: Five features were selected to build the nomogram, including age, male sex, cardiac dysfunction, STEMI, and TIMI 0–2 after PCI. The performance of the nomogram was evaluated, and the area under the curves (AUC) was 0.881 (95% CI: 0.8–0.961). Our nomogram exhibited better performance than a previous risk model (AUC = 0.7, 95% CI: 0.586–0.813) established by Brener et al. The survival curve successfully stratified the patients above and below the median score of 4. CONCLUSION: A novel nomogram for predicting 30-day mortality was established in unselected patients undergoing PCI, which may help risk stratification in clinical practice.