Prognostic value of exercise testing in risk stratification of patients with heart failure

Aim: To compare the predictive power of conventional exercise testing (CVET) vs cardiopulmonary exercise testing (CPET). Methods: A cohort study of 1,474 patients with heart failure was analyzed. We assessed variables of CVET and CPET. We used Schiller CS200 equipment with modified Balke protocol ramp. The comparison between groups was performed by correlation and contingency tables. It was considered stochastic significance when P <.05. Results: 80% of the patients were male with an average age of 53 +/- 15 years. The most prevalent causes of heart failure were 65% for ischemic heart disease and 27% for dilated cardiomyopathy. Left ventricle ejection fraction (LVEF) was 34 +/- 10%, the workload were 6.3 +/- 2 (87%) METs and VO2 peak was 21 +/- 7mlO2/kg/min. CPET qualified 9% more patients at high risk compared to the 78% CVET, P <.001. The variables with more percentage at high risk were: 60% chronotropic incompetence and 49% VO2 recovery. The sensitivity, specificity and accuracy CVET for detecting high-risk subjects, compared to the CPET was 90%, 100% and 91% respectively. Weighted kappa between two tests was 0.7 (P <.001). Conclusion: The CVET has adequate sensitivity and specificity to detect patients with heart failure at high risk relative to the CPET. The degree of agreement is significant, but not enough to consider it as a surrogated test. (C) 2015 Instituto Nacional de Cardiologia Ignacio Chavez. Published by Masson Doyma Mexico S.A. All rights reserved.