Discriminant value of clinical and exercise variables in detecting significant coronary artery disease in asymptomatic men

Abstract

To determine whether clinical or exercise test variables could reliably detect coronary disease in asymptomatic men, several variables were compared with angiographic findings in 225 asymptomatic men. None of the individual clinical or rest electrocardiographic variables were able to detect coronary artery disease. The three individual exercise variables with a high likelihood ratio were: 1) at least 0.3 mV ST depression, 2) persistence of ST depression 6 minutes after exercise, and 3) total duration of exercise of less than 10 minutes. However, because of low sensitivity and predictive value, these single variables were not helpful in identifying individual patients with coronary disease. The combination of any single clinical risk factor and any two of these exercise risk predictors was highly predictive (89%) but relatively insensitive (37%) for detecting any coronary disease. These criteria have a sensitivity of 55% and a predictive value of 84% for the detection of two and three vessel coronary disease. The effectiveness of exercise testing for detecting asymptomatic coronary disease is improved when the group is first screened for the presence of risk factors and additional exercise variables other than ST segment criteria are evaluated.