[A statistical evaluation of the variability in the measurements of the resistive index in kidney transplantation].

Doppler ultrasound (US) is a valuable tool to measure blood flow in the transplanted kidney, but its operator-dependence can greatly affect repeatability and reproducibility of measurements. Aim of this work was to evaluate intraobserver and interobserver variability in measuring the resistive index (RI) in renal transplants. Ten renal transplant recipients were randomly selected among those undergoing follow-up and examined by two operators (FG and LB) with 3.5 MHz and 10 MHz scanheads to assess the variability of RI measurements. Each observer obtained two measurements of the RI with each scanhead within a 10-15 minutes' period. In all, 80 measurements were made, 4 per patient per observer. The statistical analysis included two-tailed Student's t-test for paired data and calculation of repeatability/reproducibility coefficients. Student's t-test analysis demonstrated a statistically significant difference (p = 0.037) between the means of the first and second measurements by FG with the 3.5 MHz scanhead and the first and the second measurements by LB with the same scanhead. Differences between the other means were not statistically significant. Intraobserver variability ranged 0.03 units (or 2.07%) and 0.07 units (or 4.24%), while interobserver variability was 0.04 units with both 3.5 and 10 MHz scanheads, or 3.61 and 3.73%, respectively. Doppler US of renal transplants has statistically quantifiable operator-dependent variability: the possible evidence of statistically significant differences can be minimized by having the same operator make the measurements. However, RI variations ranging 0.02 to 0.04 units should not be considered significant.