Determining the preferred percent-predicted equation for peak oxygen consumption in patients with heart failure.

BACKGROUND Peak oxygen consumption (Vo(2)) is routinely assessed in patients with heart failure undergoing cardiopulmonary exercise testing. The purpose of the present investigation was to determine the prognostic ability of several established peak Vo(2) prediction equations in a large heart failure cohort. METHODS AND RESULTS One thousand one hundred sixty-five subjects (70% males; age, 57.0+/-13.8 years; ischemic etiology, 43%) diagnosed with heart failure underwent cardiopulmonary exercise testing. Percent-predicted peak Vo(2) was calculated according to normative values proposed by Wasserman and Hansen (equation), Jones et al (equation), the Cooper Clinic (below low fitness threshold), a Veteran's Administration male referral data set (4 equations), and the St James Take Heart Project for women (equation). The prognostic significance of percent-predicted Vo(2) values derived from the 2 latter, sex-specific equations were assessed collectively. There were 179 major cardiac events (117 deaths, 44 heart transplantations, and 18 left ventricular assist device implantations) during the 2-year tracking period (annual event rate, 10%). Measured peak Vo(2) and all percent-predicted peak Vo(2) calculations were significant univariate predictors of adverse events (chi(2)> or =31.9, P<0.001) and added prognostic value to ventilatory efficiency (VE/Vco(2) slope), the strongest cardiopulmonary exercise testing predictor of adverse events (chi(2)=150.7, P<0.001), in a multivariate regression. The Wasserman/Hansen prediction equation provided optimal prognostic information. CONCLUSIONS Actual peak Vo(2) and the percent-predicted models included in this analysis all were significant predictors of adverse events. It seems that the percent-predicted peak Vo(2) value derived from the Wasserman/Hansen equations may outperform other expressions of this cardiopulmonary exercise testing variable.