Limits of reproducibility of cross-sectional echocardiographic measurement of left ventricular ejection fraction.

The total intrinsic variability of echocardiographically determined ejection fraction was evaluated in this study. Cross-sectional left ventricular echocardiograms were made in 14 normal volunteers (9 men, 5 women), ages 31 to 41 years (mean 36 years) with an interval period of about 11 months. We used the parasternal long-axis view, the short-axis view at the papillary muscle level and the four-chamber view from the apex. M-mode recordings were made from the parasternal long-axis view. Simpson's rule and area-length formulas were used to calculate left ventricular volumes and the ejection fraction. From the methods under investigation, the Teichholz technique and the bullet method showed the best reproducibility of the ejection fraction with a coefficient of variation of 4 +/- 3% and 6 +/- 4%, respectively (mean +/- SD). The methods of Baran, and the approach using slices, showed rather large coefficients of variation, 14 +/- 11% and 9 +/- 6%, respectively.