INCREMENTAL EVALUATION OF EXERCISE CAPACITY AS AN INDEPENDENT PREDICTOR OF CORONARY-ARTERY DISEASE PRESENCE AND EXTENT

To determine the independent incremental value of exercise capacity (M ETS) concerning the presence and extent of coronary artery disease, we analyzed data from 800 patients with suspected coronary disease who u nderwent both exercise testing and coronary angiography. We performed logistic regression analysis of clinical and exercise test data with:a n incremental design to mimic the usual flow of data acquisition. Sepa rate analyses were performed concerning coronary disease presence (gre ater than or equal to 1 vessel with a greater than or equal to 50% les ion) and extent (three-vessel/left main disease). Diagnostic accuracy was determined by calculating receiver operating characteristic (ROC) curve areas. When considered alone, METS was a significant predictor o f both presence and extent of disease. Multivariate analysis revealed that METS was an independent predictor of disease extent but not prese nce. However, comparison of Hoc curve areas failed to show any loss of accuracy when METS was removed from the coronary disease extent analy sis. Despite the strong univariate relationship between exercise capac ity and coronary disease presence and extent and the independence of e xercise capacity as a predictor of coronary disease extent, the lack o f any additional incremental accuracy attributed to its consideration virtually cancels its value as a diagnostic variable for assessing bot h coronary disease presence and extent. (AM HEART J 1994;127:32-8.)