Interobserver and interequipment variability of echo-Doppler examination of the portal vein: effect of a cooperative training program.

The aim of this study was to assess the interobserver, interequipment, and time-dependent variabilities of echo-Doppler measurements of portal blood flow velocity (PBV), portal vein diameters (PVDs) and their derived parameters, portal blood flow (PBF), and congestion index (CI) in cirrhotic patients. The influence of a cooperative training program of the operators on the reproductibility of the results was also investigated. The echo-Doppler parameters were independently measured in 15 patients by four skilled operators, using four echo-Doppler machines (Acuson, ATL, Hitachi-Esaote, Toshiba. Eight of the 15 patients were restudied after 15 days by the same operators using only one machine. Significantly different values of PBV, PBF, and CI were obtained. PBV variance was equipment-related (32%) and operator-related in a smaller portion (5%). No systematic effect related to the time of investigation was found. After training to define a precise protocol, new measurements were performed by four operators on 8 different patients. No significant differences were found among the operators for any of the parameters and the 95% confidence limits (CL) and coefficients of variation (CV) of PBV showed a marked decrease (CL from +/- 26.4% to +/- 15.6%). These results indicate that (1) a significant systematic variability exists between Doppler measurements with different equipment; (2) there is no significant time-dependent systematic variability of Doppler measurements; and (3) a cooperative training program reduces the interobserver variability for direct measurements, such as PBV.