Development and validation of a risk model for long-term mortality after percutaneous coronary intervention: The IDEA-BIO Study

Objectives: We aimed to develop a model to predict long-term mortality after percutaneous coronary intervention (PCI), to aid in selecting patients with sufficient life expectancy to benefit from bioabsorbable scaffolds. Background: Clinical trials are currently designed to demonstrate superiority of bioabsorbable scaffolds over metal devices up to 5 years after implantation. Methods: From 2000 to 2011, 19.532 consecutive patients underwent PCI in a tertiary referral hospital. Patients were randomly (2:1) divided into a training (N = 13,090) and validation (N = 6,442) set. Cox regression was used to identify determinants of long-term mortality in the training set and used to develop a risk model. Model performance was studied in the training and validation dataset. Results: Median age was 63 years (IQR 54–72) and 72% were men. Median follow-up was 3.6 years (interquartile range [IQR] 2.4-6.8). The ratio elective vs. non-elective PCIs was 42/58. During 88,620 patient-years of follow-up, 3,156 deaths occurred, implying an incidence rate of 35.6 per 1,000. Estimated 5-year mortality was 12.9%.Regression analysis revealed age, body mass index, diabetes mellitus, renal insufficiency, prior myocardial infarction, PCI indication, lesion location, number of diseased vessels and cardiogenic shock at presentation as determinants of mortality. The long-term risk model showed good discrimination in the training and validation sets (c-indices 0.76 and 0.74), whereas calibration was appropriate. Conclusions: A simple risk model, containing 9 baseline clinical and angiographic variables effectively predicts long-term mortality after PCI and may possibly be used to select suitable patients for bioabsorbable scaffolds.