Cardiac Function Stratification Based on Echocardiographic or Clinical Markers of Left Ventricular Filling Pressures Predicts Death and Hospitalization Better Than Stratification by Ventricular Systolic Function Alone

BACKGROUND: A normal left ventricular ejection fraction (LVEF) often underestimates the poor prognosis associated with diastolic dysfunction. METHODS: We compared overall and hospital-free survival according to echocardiographic diastolic function classification (echo class), clinical probability of diastolic dysfunction (clinical class) and LV grades based on biplane LVEF, in 114 subjects followed-up over a median of 47 months. Diastolic function was classified into normal, impaired relaxation, and severe dysfunction (SDD), using a previously validated 3-staged classification. RESULTS: There were 16 deaths and 42 combined end points of death and hospitalization. Although each classification method globally prognosticated survival (P = 0.001, P =0.046, and P = 0.034 by the echo class, clinical class and LVEF grades, respectively), only echo class correctly distinguished three risk levels. Death was not hierarchically predicted by LVEF whereas severe diastolic dysfunction was associated with a hazard ratio by univariate or a multivariate model (that evaluated the effects of age, gender, and LVEF) of 4.31 (P =0.004) or 3.88 (P = 0.03), respectively. Also, a significant separation was found for the combined end points associated with SDD relative to nonsevere echo classes (P = 0.045). Neither clinical risk staging, nor LV grading showed significant separation of the Kaplan-Meier plots between "high risk" versus others combined, and Normal LV grade versus others combined, respectively. Severe diastolic dysfunction trended strongly as an independent predictor of combined end point with multivariate hazard of 2.29 (95% CI 0.99-5.26 P=0.05). CONCLUSION: Stratification of the severity of diastolic dysfunction using comprehensive echocardiographic parameters of systolic and diastolic function is effective at predicting death and hospital-free survival.