Diagnostic techniques: Intra-observer, interobserver, interultrasound transducer and intercycle variation in colour Doppler assessment of uterine artery impedance

In this study we assessed the variability of measuring uterine artery pulsatility index by colour Doppler ultrasound using different ultrasound transducers (abdominal 3 MHz, abdominal 5 MHz, endovaginal 5 MHz with and without a full bladder). We then assessed the intra-observer coefficient of variation for the different transducers and interobserver variability for two observers using the endovaginal probe. The intercycle variability and effects of a full bladder on uterine artery blood flow were also assessed. The uterine vessels were most easily identified using the endovaginal transducer. The most reproducible results were obtained using the endovaginal probe. The coefficients of variation obtained for the vaginal route were intra-observer 4.1%, interobserver 11.8% and intercycle 16.5%. The blood flow indices were significantly different between the endovaginal and both transabdominal 5 and 3 MHz probes (P=0.0027 and 0.0005 respectively). There was no significant effect of a full bladder on pulsatility index. This study suggests that transvaginal colour Doppler ultrasound is reproducible and the results obtained from clinical use would be reliable