New, Simple Echocardiographic Indexes for the Estimation of Filling Pressure in Patients with Cardiac Disease and Preserved Left Ventricular Ejection Fraction


Address for correspondence and reprint requests: Hisham Dokainish, M.D., F.R.C.P.C., F.A.S.E., F.A.C.C., Associate Professor of Medicine, McMaster University, Director of Echocardiography, Hamilton Health Sciences, 237 Barton St. E., CVSRI #C3 111, Hamilton, ON, Canada L8L 2X2. E-mail:


Background: There are few data on echocardiographic indexes incorporating peak mitral inflow velocity (E), left atrial volume index (LAVi), and pulmonary artery pressure (PAP) for estimation of left ventricular (LV) filling pressure in patients with preserved LV ejection fraction (EF ≥ 50%). Methods: Patients underwent echocardiography ≤20 minutes of cardiac catheterization. Echocardiographic variables were compared to invasively measured LV end-diastolic pressure (LVEDP). Results: Of the 122 patients, 67 (55%) were women, the mean age was 55 ± 9 years, the mean left ventricular ejection fraction (LVEF) was 61 ± 6%, 107 (88%) were hypertensive, and 79 (65%) had significant coronary artery disease at catheterization. E/Ea correlated with LVEDP (R = 0.68, P < 0.0001), compared to PAP (R = 0.53, P < 0.001), peak E velocity (R = 0.48, P < 0.001), and LAVi (R = 0.48, P < 0.001). E/Ea > 12 had 75% sensitivity and 78% specificity for LVEDP ≥ 20 mmHg (area under curve (AUC) = 0.79, P < 0.0001), compared with (PAP + LAVi)/2 > 30 (sensitivity = 72%, specificity = 80%, AUC = 0.84, P < 0.001) and (E + LAVi)/2 > 57 (sensitivity = 73% and specificity = 81%, AUC = 0.82, P < 0.001) (P = NS). E <60 cm/sec had 94% negative, and E>90 cm/sec had 96% positive, predictive value for LVEDP ≥ 20 mmHg. (E + LAVi)/2 added incrementally to E/Ea when E/Ea was in the gray zone. Conclusions: New, simple echocardiographic equations, (E + LAVi)/2 and (PAP + LAVi)/2, have comparable accuracy to E/Ea for LVEDP estimation in patients with cardiac disease and preserved LVEF, and (E + LAVi)/2 added incrementally to E/Ea alone when E/Ea was in the gray zone. Peak E velocity alone had high negative and positive predictive value for elevated LVEDP in this population. These simple echocardiographic variables could be used—in isolation or with E/Ea—in patients with cardiac disease and preserved LVEF for the diagnosis of diastolic heart failure. Echocardiography 2010;27:946-953)