Internal consistency of echocardiographic estimates of the severity of left ventricular outflow obstruction.

To determine their internal consistency, M-mode and Doppler echocardiography were used to estimate the gradient across the left ventricular outflow tract during 74 evaluations of 50 infants, children, and young adults with congenital valvular (n = 43), subvalvular (n = 6), and supravalvular (n = 1) aortic stenosis. By M-mode the gradient was estimated from the wall-stress formula (left ventricular pressure = 225 x wall thickness/end-systolic diameter) minus systolic blood pressure determined by sphygmomanometry. Doppler (pulsed or continuous wave) methods utilized the Bernoulli formula (gradient = 4 x V2). There was good agreement between the M-mode and Doppler estimates of outflow gradient in most patients (r = 0.69, standard error of the estimate = 26.9). In 46 of 74 comparisons (62%) the two estimates differed by less than 20 mm Hg, and the estimates placed the patient in the same clinical class (mild, moderate, or severe). In 22 patients undergoing cardiac catheterization, there was only a fair correlation between the M-mode (r = 0.50) and Doppler (r = 0.58) gradients and those obtained at catheterization. Each noninvasive technique yielded major overestimates or underestimates of the gradient in several instances. The M-mode and Doppler techniques for estimating the severity of congenital aortic stenosis are complementary. Their combined use minimizes but does not totally eliminate errors of overestimation or underestimation of outflow gradient.