Comparative accuracy of Doppler echocardiographic methods for clinical stroke volume determination

Numerous Doppler echocardiographic methods to measure stroke volume have been proposed in experimental or clinical studies, but their relative accuracy in patients compared with an invasive reference standard remains uncertain. Accordingly, we compared Doppler with thermodilution stroke volume measurement in 18 hospitalized patients, 16 with an acute manifestation of coronary artery disease and two with chronic cardiomyopathies. Doppler time-velocity integrals were measured by darkest line (modal velocity) and the leading edge (maximal velocity) techniques at the aortic annular plane, the mitral orifice, and the mitral annular plane. Two-dimensional echocardiography was used to measure cross-sectional areas (M-mode-corrected at the mitral orifice). The combination of aortic annular cross-sectional area and the leading edge technique of measuring the time-velocity integral of blood flow at this site provided the most accurate measure of stroke volume (r = 0.87, p less than 0.0001, standard error of estimate = 11 cm3; mean difference from thermodilution = 2.8 ml +/- 8.9 ml, p = NS). It also resulted in the most accurate measurement of cardiac output (r = 0.88, p less than 0.0003; mean difference from thermodilution = 0.11 L/min +/- 0.69 L/min, p = NS). Other methods yielded values that correlated less closely and deviated systematically from thermodilution measurements. We therefore conclude that of the six common methods evaluated, the aortic annular leading edge method measures stroke volume with the best accuracy and is most suitable for clinical application.