Reproducibility and persistence of abnormal transmitral flow velocity patterns detected by pulsed Doppler ultrasound.

This prospective study examines the reproducibility and persistence of abnormal transmitral flow detected using pulsed Doppler ultrasound on 2 separate occasions between 2 days and 6 weeks apart. The 22 patients included were accepted consecutively from those having an abnormal diastolic flow pattern at initial examination. Abnormal flow velocity patterns were defined as those exhibiting reversal of the ratio of the passive filling velocity (E) and active atrial transport velocity (A). There was no significant difference between the separately recorded values for the ratio of the peak E and A velocities or of the ratio of the planimetered areas beneath each of the velocity waves, with positive correlations for both sets of values (r = 0.68, r = 0.67). Significant positive correlation also existed between the mean rates of acceleration to each of the E and A velocity peaks of the transmitral waveform recorded on separate occasions (r = 0.68, r = 0.95). Interobserver variation for the analysis of hard-copy pulsed Doppler recordings between two trained observers was less than 5% and intraobserver error for recording analysis was less than 2% for both observers. Abnormal transmitral flow velocity patterns persist unchanged in the absence of therapeutic intervention and the acceptably small observer error in the recording and analysis of such flow patterns allows consistent and clinically reliable data to be obtained.