Reproducibility of ductus venosus Doppler flow measurements at 11-14 weeks of gestation

Objectives To assess the intra- and interobserver repeatabilities of fetal ductus venosus Doppler measurements at 11-14 weeks of gestation. Design Flow velocity waveforms were recorded transabdominally. Intraobserver repeatability was studied in 22 fetuses in whom four repeated measurements were performed by the same observer. Interobserver repeatability was assessed in 54 fetuses in each of whom two observers performed two repeated measurements. The pulsatility index for veins (PIV), peak velocity during ventricular systole (S wave), peak velocity during atrial contraction (A wave) and time-averaged maximum velocity (TAMV) were recorded. Reproducibility of the Doppler measurements was analyzed by calculating repeatability coefficient, coefficient of variation (CV), intraclass correlation coefficient (ICC), mean differences and their limits of agreement. Cohen's k-coefficient was used for categorical data. Results Intraobserver repeatability was good with a CV of 10%, 13%, 22% and 13% and an ICC of 0.85, 0.94, 0.94 and 0.95, respectively, for PIV S wave, A wave and TAMV. Interobserver repeatability was also good, and no bias between the observers was detected. A CV of 8.8%, 14%, 27% and 15% and an ICC of 0.86, 0.84, 0.87 and 0.84, respectively, for PIV, S wave, A wave and TAMV were obtained. There was 100% agreement for the detection of normal/abnormal blood flow. Conclusions Intra- and interobserver repeatability of all parameters was acceptable, allowing for the detection of moderate to large changes in Doppler measurements.