Comparison of exercise algorithms for diagnosis of coronary artery disease

The following previously proposed exercise algorithms were evaluated: (a) the cumulative area of ST-segment depression during exercise, (b) discriminant analysis of various exercise variables and (c) heart-rate-adjusted ST-segment amplitude changes. The study population comprised 345 males without a history of myocardial infarction. Prevalence of coronary artery disease was 51%. All had a normal ECG at rest. Frank-lead ECG were computer processed during symptom-limited bicycle ergometry. Discriminant analysis and heart-rate-adjusted ST-segment amplitude changes proved to have excellent diagnostic characteristics: sensitivity amounted 80%; specificity to 90%. Both methods seem well suited for diagnostic applications in clinical practice.<
>