Are cardiac events during exercise therapy for heart failure predictable from the baseline variables?

Abstract

Background: Exercise training (ET) is an emerging therapy for chronic heart failure, but the baseline patient characteristics for predicting cardiac events (CEs) during the course of ET remain unknown. Methods and Results: Of the 111 stable heart failure patients who participated in a 3-month ET program, 6 withdrew from the program for cardiac reasons and 9 had transient interruptions in the program because of CEs. The baseline clinical characteristics of these 15 patients (CE group) and the remaining 96 patients (No-CE group) were compared. Compared with the No-CE group, the CE group had a significantly higher prevalence of pacemaker/implantable cardioverter-defibrillators, larger left ventricular end-diastolic diameter (LVEDDs), lower peak oxygen uptake, greater ventilation drive, and higher plasma brain natriuretic peptide concentration at baseline. Multivariate logistic regression analysis showed that a larger LVEDD was a significant predictor of the occurrence of a transient interruption to or permanent withdrawal from the ET program because of CEs. Receiver operating characteristic curve analysis demonstrated that an LVEDD ≥65 mm had a sensitivity of 93% and specificity of 48% in predicting CEs. Conclusions: Patients with a large LVEDD (≥65mm) at baseline should be monitored carefully during the course of an ET program.