EXERCISE TESTING IN PATIENTS WITH CHEST PAIN AND NORMAL CORONARY-ARTERIES - IMPROVING TEST SPECIFICITY BY USE OF A SIMPLE LOGISTIC MODEL

Despite an exercise electrocardiogram (EGG) positive for ischemia by e stablished criteria, many patients referred for coronary angiography t o evaluate chest pain are found to have angiographically normal corona ry arteries (NCA). Exercise ECG were analyzed from 27 patients with ch est pain and angiographically NCA and 28 patients with chest pain and coronary artery disease (CAD) using univariate and multivariate logist ic regression analysis. We derived the following logistic model for th e legit probability of CAD: 3 + SEX x 4 - METs x 0.7 + STDV5 x 0.8, wh ere SEX = 0 for female and SEX = 1 for male, METs = maximal estimated work load (metabolic equivalents) and STDVS = horizontal or downslopin g ST depression (mm) in V5. A legit probability greater than or equal to 0 identified CAD with a sensitivity of 79% and a specificity of 89% . The model correctly identified 28/36 (78%) patients with CAD, and 7/ 10 (70%) patients with NCA (correct diagnosis 76%; p < 0.02) in a sepa rate random group of 46 unselected patients with positive exercise tes ts undergoing diagnostic coronary angiography.