Development and Validation of a Prediction Model for Acute Coronary Occlusion in Patients With Non‐ST‐Segment Elevation Myocardial Infarction

Background Acute coronary occlusion (ACO) occurs in approximately one‐quarter of patients with non‐ST‐segment elevation myocardial infarction (NSTEMI) and is associated with adverse outcomes. This study aimed to develop and validate a prediction model for ACO. Methods Consecutive patients admitted with NSTEMI between January 2016 and June 2023 who underwent coronary angiography were enrolled. Demographic, electrocardiographic, laboratory, and echocardiographic data were collected. Patients admitted during 2016−2020 were assigned to the training cohort ( n = 670), and those admitted during 2021−2023 were assigned to the validation cohort ( n = 334). Predictors were selected using least absolute shrinkage and selection operator regression, and a nomogram was constructed based on multivariable logistic regression. Model performance was assessed by receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). Results Among 1,004 patients, 262 (26.1%) had ACO. Seven predictors, including smoking history, ST‐segment depression ≥ 1 mm, regional wall motion abnormality, left ventricular ejection fraction, admission cardiac troponin I, interleukin‐6, and high‐sensitivity C‐reactive protein, were incorporated into the model. The area under the ROC curve was 0.731 (95% CI, 0.688−0.773) in the training cohort and 0.715 (95% CI, 0.654−0.776) in the validation cohort. Calibration curves showed good agreement between predicted and observed probabilities, and DCA demonstrated favorable clinical utility. Conclusion This study developed and validated a multidimensional prediction model for ACO in patients with NSTEMI. The model showed acceptable performance and may aid early identification before coronary angiography, although reliance on IL‐6, which is not routinely available in all centers, may limit its generalizability.