Reliability and agreement of human renal and segmental artery hemodynamics measured using Doppler ultrasound

To optimize study design and data interpretation, there is a need to understand the reliability of Doppler ultrasound-derived measures of blood velocity (BV) measured in the renal and segmental arteries. Thus, this study tested the following two hypotheses: 1) renal and segmental artery BV measured over the current standard of three cardiac cycles have good agreement with measurements over nine cardiac cycles ( study 1); and 2) renal and segmental artery BV measurements have relatively poor day-to-day reliability ( study 2). In study 1, there was excellent agreement between measurements over three and nine cardiac cycles for BV in both the renal and segmental arteries, as evidenced by BV measurements that were not statistically different ( P ≥ 0.68), were highly consistent ( r ≥ 0.99, P < 0.01), had a coefficient of variation ≤2.5 ± 1.8%, and 97% (renal artery) and 92% (segmental artery) of the individual differences fell within the 95% limits of agreement. In study 2, there was relatively good day-to-day reliability in renal artery BV as evidenced by no differences between three separate days ( P ≥ 0.30), an intraclass correlation coefficient (ICC) of 0.92 (0.78, 0.98), and 7.4 ± 5.5% coefficient of variation. The day-to-day reliability was relatively poor in the segmental artery with an ICC of 0.77 (0.41, 0.93) and 9.0 ± 5.6% coefficient of variation. These findings support measuring renal and segmental artery hemodynamics over three cardiac cycles and the utility in reporting renal BV across days. However, because of the variation across days, hemodynamic responses in the segmental arteries should be reported as changes from baseline when making comparisons across multiple days. NEW & NOTEWORTHY The present study indicates that Doppler ultrasound-derived measures of renal and segmental artery hemodynamics over three cardiac cycles have excellent agreement with those over nine cardiac cycles. These findings support the current practice of measuring renal and segmental artery blood velocity over three cardiac cycles. This study also demonstrates that there is excellent day-to-day reliability for measures of renal artery blood velocity, which supports reporting absolute values of renal artery blood velocity across days. However, it was also found that the day-to-day reliability of segmental artery measurements is relatively poor. Thus, to account for this variability, we suggest that segmental artery hemodynamics be compared as relative changes from baseline across separate days.