Objectives: Duplex ultrasonography assessment of superficial forearm veins is frequently used before a hemodialysis arteriovenous fistula (AVF) is created. There is, however, no standardized preoperative duplex ultrasonography protocol. This study assessed B-mode image analysis reproducibility and reproducibility of repeated forearm superficial venous diameter measurements on different days at different venous congestion pressures (VCPs).
Methods: Diameters were determined using B-mode ultrasonography in 10 healthy male volunteers on days 1 and 14 at incremental VCP values (10 to 80 mm Hg). Intra- and interobserver agreement was assessed for B-mode image analysis by calculating interclass correlation coefficients (ICC). Reproducibility of repeated diameter measurements (maximum and minimum diameter at days 1 and 14), cross-sectional area size increase, and shape change due to incremental VCPs were determined by calculating ICC values.
Results: Analysis of intraobserver agreement of B-mode image interpretation yielded ICC values of 0.97 (95% confidence interval [CI], 0.94 to 0.99) and 0.97 (95% CI, 0.96 to 0.99) for determination of maximum and minimum diameters, respectively. Interobserver agreement analysis yielded ICC values of 0.95 (95% CI, 0.92 to 0.97) and 0.96 (95% CI, 0.96 to 0.99) for determination of maximum and minimum diameters, respectively. Reproducibility of repeated diameter measurements on days 1 and 14 improved substantially at incremental VCP values, with best reproducibility at VCPs >40 mm Hg. Repeated determination of cross-sectional area size increase and shape change due to VCP increase from 10 to 80 mm Hg yielded ICC values of 0.49 (95% CI, 0.19 to 1.00) and 0.09 (95% CI, 0.00 to 0.92), respectively. Maximum and minimum diameters as well as cross-sectional area size increased significantly (P < .01) due to VCP increase during both sessions. Cross-sectional area shape changed significantly (P < .01) due to VCP increase during both sessions.
Conclusions: Diameter measurements on B-mode images are largely observer independent. Superficial venous cross-sectional area shape is noncircular, and cross-sectional area size depends on VCP. Both maximum and minimum venous diameters should be determined at VCPs >40 mm Hg to attain the best reproducibility. Further studies are needed to determine whether a standardized preoperative vein mapping protocol can reduce AVF nonmaturation rates.