A Risk Score for Estimating Coronary Artery Disease in Different-age Patients

Abstract

In this paper we evaluated ECG descriptors for classification or recognition of patients with coronary artery (CA) stenosis applying CARDS algorithm and binary regression methods in different-age patients with coronary artery disease (CAD). 199 patients with stale angina pectoris and without ischaemic ECG changes were investigated (140 male and 59 female). One hundred of them (84 male and 16 female) were younger than 60 years and 99 patients (56 male and 43 female) were 60 years old and older. The patients were divided into two main groups according to CA stenosis degree: stenosis of CA lumen ≥ 50% and stenosis of CA lumen ≤30%. Digital ECG was recorded for all patients (discretisation parameters 12 bit, 2 kHz, recording interval of 10 s, 12 standard leads). ECG parameters of depolarization and repolarization were evaluated by computer software, created at Institute of Cardiology, Kaunas medical university. On the ground of multivariate logistic model we developed a risk score for suspicion of CA stenosis. Sensitivity and specificity of CA 70 lesions ≥ 50% was 90.4 % and 52.3% when risk score was higher than 6., and 76.9 % and 76% when risk score was higher than 7 in patients younger than 60 years. Sensitivity and specificity of CA lesions ≥ 50% was 89.2 % and 65.2% when risk score was higher than 5 in patients 60 years and older. Tabl. 4, bibl. 5 (in English; summaries in English, Russian and Lithuanian).