Giá trị của mô hình RF‑CL trong dự báo tắc nghẽn động mạch vành trên chụp cắt lớp vi tính động mạch vành ở bệnh nhân đau thắt ngực ổn định

Abstract

This cross-sectional study included 331 patients with stable angina who underwent coronary computed tomography angiography (CCTA) to evaluate the discrimination and calibration of the risk factor–weighted clinical likelihood (RF‑CL) model in predicting obstructive coronary artery disease (CAD). The median age was 65.5 years old (IQR: 57.7 – 73), and 61.9% were male. The prevalence of obstructive CAD was 72.8%, reflecting a relatively high-risk population referred for CCTA in routine clinical practice. The proportion of obstructive CAD increased across RF‑CL categories of very low (≤ 5%), low (>5 – 15%), and intermediate (>15 – 50%) risk (47.4%, 64.2%, and 86.9%, respectively). RF‑CL showed moderate discrimination with an area under the ROC curve of 0.746 (95% CI: 0.686 – 0.805; p < 0.001). At the 5% RF‑CL threshold, sensitivity and specificity were 88.8% and 33.3%, respectively, whereas at the 15% threshold they were 60.6% and 75.6%. The model demonstrated good calibration (Hosmer–Lemeshow p = 0.317). Compared with the very‑low‑risk group, the low‑ and intermediate‑risk groups had approximately 3.7‑fold and 7.4‑fold higher odds of obstructive CAD. These findings suggest that the RF‑CL model may be useful for the discrimination and risk stratification of obstructive CAD in patients with stable angina undergoing CCTA; however, further validation in larger and multicenter studies is warranted.