Reproducibility of Doppler echocardiographic measurements of left ventricular diastolic function.

The present investigation was undertaken in 12 normal subjects to determine the magnitude of technical and biologic variability of six previously validated Doppler indexes of diastolic function. Variability due to the reader was small for each of the six Doppler indexes. Variability due to the technician was larger than for the reader, and day-to-day variability was larger than variability due to either the reader or technician for the great majority of the comparisons (i.e., 21 of 24). Four Doppler indexes assessing early diastolic events (isovolumic relaxation, duration of the early diastolic peak of flow-velocity, rate of decrease of flow-velocity in early diastole, and maximal early diastolic flow-velocity) did not show statistically significant changes due to day-to-day variability. The two Doppler indexes assessing late diastolic events [maximal late diastolic (atrial) flow-velocity, and the ratio between maximal early and late flow-velocity] showed greater change on a day-to-day basis (P less than 0.05). In conclusion, Doppler indexes that represent a measure of isovolumic relaxation and the early filling phases of diastole showed small variability. Indexes that measure the late filling phase of diastole, such as maximal late (atrial) diastolic flow-velocity and the ratio between early and late diastolic flow-velocities, showed relatively large day-to-day variability; therefore, serial changes in the atrial contribution to ventricular filling identified by these indexes should probably be interpreted with some caution.