Developing and verifying a multivariate model to predict the survival probability after coronary artery bypass grafting in patients with coronary atherosclerosis based on the MIMIC-III database

Background: Coronary atherosclerosis is one of the main cardiovascular diseases affecting the global population. Coronary artery bypass grafting (CABG) is commonly used to improve the survival probability of patients with coronary atherosclerosis. However, the prognosis of patients after CABG remains unclear. Objectives: We aimed to construct a novel nomogram comprising readily available indicators to predict the 1-, 2-, and 3-year survival rates after CABG in patients with coronary atherosclerosis. Methods: We utilized the Medical Information Mart for Intensive Care III (MIMIC-III) database for the study. The calibration plot, concordance index (C-index), net reclassification index (NRI), integrated discrimination improvement (IDI), and area under the receiver operating characteristic curve (AUROC) were used to evaluate the performance of the model, and to compare the nomogram with the Sequential Organ Failure Assessment (SOFA) score and Simplified Acute Physiology Score II (SAPS II) in order to illustrate the clinical effectiveness of the model. Results: The multivariate Cox regression model showed that age, marital status, body mass index, creatinine, platelet count, red cell distribution width, heart rate, intensive-care unit stay time, and Elixhauser Comorbidity Index were risk factors. The C-indexes of the nomogram exceeded 0.75, and its NRI and IDI were both higher than 0. The AUROCs were larger for the nomogram than for the SAPS II and SOFA score. Conclusion: Our new nomogram is a personalized tool that helps clinicians choose treatment options and predict the long-term prognosis of patients.