Is Averaging of Three Waveforms Sufficient to Generate a Truly Representative Value for Pulsed-Wave Doppler Measures of Impedance?

Abstract

We aimed to evaluate the physiological variation in common pulsed-wave Doppler (PWD) indices of impedance to determine the number of waveforms to be averaged to minimise variability to 5%. A single-centre, prospective, cross-sectional cohort study of uncomplicated singleton pregnancies at 20-37 week's gestation. From each patient 100 PWD waveforms were acquired including the umbilical artery (UA), middle cerebral artery (MCA) and uterine arteries (UtAs), with 30 waveforms acquired from the ductus venosus. Each waveform was individually measured using the machine's in-built software in automated mode. The variability was assessed using coefficient of variation. The number of waveforms to be averaged was calculated using the moving average and standard error of mean. From a cohort of 200 pregnancies, a total of 189 were analysed. The pulsatility index (PI) demonstrated greater variability compared with the resistance index (RI) in all vessels studied. A minimum of 14 UA and MCA, and 13 UtA PWD waveforms were required to reduce PI variability to 5%, while only 2 RI waveforms were required for UA, 1 for MCA and 8 for UtAs. The variability shown across all PWD indices and between vessels means that PWD indices results should be interpreted cautiously and averaged over multiple waveforms. Consideration should be given to adoption of RI, as it showed greater stability than PI for maternal-fetal Doppler.