Reproducibility of Doppler Ultrasound Measurements

Abstract

Purpose: The reproducibility of Doppler ultrasound measurements was studied in a group of radiologists with different experience of such measurements. Material and Methods: The 12 participating radiologists measured peak systolic velocity (PSV), resistance index (RI), and pulsatility index (PI) from 3 arteries and PSV alone in one artery of a healthy person under standardized circumstances. Each participant repeated the measurements 10 times. Results: Significant (p<0.005) interobserver variation was seen in all of the variables, with an up to 14-fold difference in the PSV values measured and most variation resulting from differences between observers. The reproducibility of RI was better, most variation being within observers. Experience improved the reproducibility of the Doppler US measurements. Significant difference (p<0.05) in Doppler US measurements was seen between the 3 experience groups. The difference was marked between the highly experienced and the medium experienced group when the variances of the measured values of PSV, RI, and PI from all 4 arteries were combined. Conclusion: The reproducibility of Doppler ultrasound measurements is poor. The value of the method is questionable.