Superiority of visual versus computerized echocardiographic estimation of radionuclide left ventricular ejection fraction.

An optimal method for determining left ventricular ejection fraction (LVEF) by echocardiography should be rapid, reliable, and widely applicable in order to be utilized routinely in a busy clinical laboratory. Most methods reported in the literature are reliable in selected, high-quality echocardiograms. Most require off-line computer analysis and are time-consuming and poorly suited to the routine of a busy laboratory. We compared in a blinded manner several echocardiographic methods of LVEF determination with the ejection fraction obtained by equilibrium radionuclide angiography (ERNA) in 44 patients unselected for image quality. Echocardiographic methods included: [1] cubed M-mode formula; [2] Teichholz M-mode formula; [3] subjective estimation of LVEF from two-dimensional echocardiographic videotape; [4] area-length method in one four-chamber view; [5] average of area-length method in three four-chamber views; [6] average of area-length method in four-chamber and two-chamber views (one beat each); [7] subjective estimation from stored videoloop of four-chamber and two-chamber view. In 30 cases M-mode tracings were available for analysis. In only 23 of the 44 patients were the apical views suitable for quantitative analysis. The ERNA ejection fraction was 44 +/- 17% (mean +/- 1 SD). The best echocardiographic correlation with ERNA ejection fraction in each patient subgroup studied was obtained by method 3. We concluded that subjective analysis of the videotaped study by an experienced cardiologist/echocardiographer provided the best estimation of ERNA ejection fraction. More time-consuming and costly computer techniques yielded a worse estimate.