Development and Validation of a Noninvasive Diagnostic Model for Obstructive Coronary Artery Disease
CHEN Wei
ZHOU Le
HUANG Zhiming
FANG Fang
Abstract：Objective:To construct a noninvasive diagnostic model for obstructive coronary artery disease(OCAD)by combining novel markers and risk factors to predict the occurrence of OCAD.Methods:A total of 700 hospitalized patients without coronary artery disease who were performed invasive coronary artery examination from January 2022 to December 2023 and selected as the model training set.A total of 300 inpatients without coronary artery disease were performed invasive coronary artery examination in Shanghai Shibei Hospital from January 2022 to December 2023 and selected as the model validation set.According to the results of invasive coronary artery examination,the model training set was divided into OCAD group and non-OCAD group.Single factor and Logistic regression analysis were used in the model training set to screen for independent factors for predicting OCAD,and an OCAD non-invasive diagnosis model was constructed.The diagnostic efficacy was then verified using the model validation set.Results:An OCAD non-invasive diagnostic model,OCAD-1,was established based on age,chest pain symptoms,duration of diabetes,left ventricular ejection fraction(LVEF),estimated glomerular filtration rate(eGFR),total cholesterol(TC),triglycerides(TG),high-density lipoprotein cholesterol(HDL-C),low-density lipoprotein cholesterol(LDL-C),lipoprotein(a)[Lp(a)],homocysteine(Hcy),high-sensitivity C-reactive protein(hs-CRP),and interleukin-6(IL-6).The area under the curve(AUC)of OCAD-1 was 0.901,and the cutoff value of the model was 0.31,sensitivity was 85.93％,specificity was 80.42％,accuracy was 82.47％,positive predictive value was 59.24％.Using the model with a threshold of ≤0.31 as the criterion to exclude OCAD,80.42％of non-OCAD patients could be identified,and the negative predictive value was 95.13％.The AUC values of age,chest pain symptoms,diabetes course,LVEF,eGFR,TC,TG,HDL-C,LDL-C,Lp(a),Hcy,hs-CRP,and IL-6 for the diagnosis of OCAD were significantly lower than those of the OCAD-1 model,with statistical significance(P＜0.001).Moreover,when the OCAD-1 model was applied to the model validation set,the AUC for OCAD diagnosis was 0.873,and there was no significant difference in diagnostic efficiency between the two groups,indicating that the model was reproducible.Conclusions:The non-invasive diagnostic model OCAD-1 shows high sensitivity,accuracy,and repeatability in the prediction of OCAD,and can predict the occurrence of OCAD to some extent.
Keywords：obstructive coronary artery diseasediagnostic modelnon-invasivepredictive value
Publication Date：2025-04-25
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：6( 1154-1159 )