PURPOSE--To describe groups of patients who have obstructive and non-obstructive coronary artery disease, through computadorized exercise stress test. METHODS--The test was done in 121 patients, all male, divided into 3 groups: GN group, 50 patients with normal electrocardiographic response to exercise; GLO group, 40 patients with obstructive coronary artery disease and GNO group, 31 patients with normal coronary arteries, showing one or more of the following entities: intramural coronary traject, coronary tortuosity, slow flow, mitral valve prolapse or left ventricular hypertrophy. GLO and GNO groups presented with abnormal response of the ST segment during exercise.

The quantitative variables registered by computer were particularly analyzed as follows:

STL (point Y depression), slope, index and ST segment integral. The magnitude of ST vector was visually measured and quantified. The statistic study was made through ANOVA and multiples comparison by the Scheffe's method, Fisher's test, quisquare and sensibility, specificity and accuracy calculation. RESULTS--There was a significant statistic difference among the 3 groups relative to slope and index (p < 0.05). The integral variable of ST segment did not allow us to differentiate the GLO and GNO groups. In the association study between the ST vector magnitude and abnormal T loop, there was an increase in sensibility of 15//in the exercise stress test. CONCLUSION--The ST segment slope below zero values, define patients having obstructive disease, and the opposite, non-obstructive disease. Values of ST segment index lower than -2 are linked to obstructive disease and higher than -2 linked to non-obstructive. Values of ST segment lower than -7 microV. s separate individuals with normal exercise stress test from those with ischemic type response. The magnitude of ST vector equal to or lower than 0.20mV define normal vectorcardiographic response to the exercise