Diagnostic prediction model in subjects with low-risk unstable angina pectoris/non-ST segment Elevation Myocardial Infarction.

Abstract

OBJECTIVE\nThis study aims to construct a prediction model based on non-invasive examination and cardiovascular risk factors, to predict the presence of coronary artery disease (CAD) and its severity in patients with low-risk unstable angina pectoris (UAP)/Non-ST Segment Elevation Myocardial Infarction (NSTEMI).\n\n\nPATIENTS AND METHODS\nThis cross-sectional study aimed to assess the association between non-invasive examinations and cardiovascular risk factors in predicting CAD. Model constructed based on non-invasive assessment and cardiovascular risk factors was compared to coronary angiography, the reference standard.\n\n\nRESULTS\nThis study included 104 patients, comprising 60 men and 44 women, who fulfilled the inclusion criteria. The mean age was 52.3 (6.8) years. Two diagnostic prediction models were constructed after series of analyses. The main model consists of NO, CIMT, history of smoking, and Age-Gender, while the alternative model consists of CIMT, history of smoking, and Age-Gender. The main model has AUC of 74.5% (95% CI: 64.9-84.1), sensitivity of 72.7% (95% CI: 57.2-85.0), specificity 65.0% (95% CI: 51.6 -76.9 for a cut-off point of 74.5. While the alternative model has 69.0% AUC (95% CI: 58.9-79.1), sensitivity of 65.9% (95%: 50.1-79, 5), a specificity of 56.7% (95% CI: 43.2-69.4) for a cut-off point of 69. The main model and the alternative model have similar diagnostic prediction performance based on the ROC comparison test (p = 0.70).\n\n\nCONCLUSIONS\nBased on these results, we conclude that NO, CIMT, smoking history, and age-gender have a value of diagnostic validity in subjects with low-risk UAP/NSTEMI.