Objective In the present study, we aimed to establish a novel score to predict long-term mortality of non-ST-segment elevation acute coronary syndrome (NSTE-ACS) patients who underwent percutaneous coronary intervention (PCI).
Methods A total of 2,174 NSTE-ACS patients from the CORFCHD-ZZ study were enrolled as the derivation cohort. The validation cohort including 1,808 NSTE-ACS patients were from the CORFCHD-PCI study. Receiver operating characteristic analysis and area under the curve (AUC) evaluation were used to select the candidate variables. The model performance was validated internally and externally. The primary outcome was cardiac mortality (CM). We also explored the model performance for all-cause mortality (ACM).
Results Initially, 28 risk factors were selected and ranked according to their AUC values. Finally, we selected age, N-terminal pro-B-type natriuretic peptide, and creatinine to develop a novel prediction model named “ABC” model. The ABC model had a high discriminatory ability for both CM (C-index: 0.774, p < 0.001) and ACM (C-index: 0.758, p < 0.001) in the derivation cohort. In the validation cohort, the C-index of CM was 0.802 (p < 0.001) and that of ACM was 0.797 (p < 0.001), which suggested good discrimination. In addition, this model had adequate calibration in both the derivation and validation cohorts. Furthermore, the ABC score outperformed the GRACE score to predict mortality in NSTE-ACS patients who underwent PCI.
Conclusion In the present study, we developed and validated a novel model to predict mortality in patients with NSTE-ACS who underwent PCI. This model can be used as a credible tool for risk assessment and management of NSTE-ACS after PCI.