Interobserver variability in aortoiliac and femoropopliteal duplex scanning

PURPOSE: The interobserver variability of aortoiliac and femoropopliteal duplex scanning in peripheral arterial occlusive disease was assessed. METHODS: Two experienced, independent vascular technologists investigated in random order 61 consecutive patients sent to the vascular laboratory for investigation of the aortoiliac or femoropopliteal arteries. In each aortoiliac vessel segment, the proximal, mid, and distal peak systolic velocities (PSVs) were measured, and corresponding PSV ratios were calculated. The superficial femoral artery was divided in 10-cm segments with a roll-centimeter taped from the upper patella margin. Interclass correlation coefficients (ICCs) were calculated as a means of appreciating the measurement variability of the PSV ratio values as a continuous variable and the Kappa value for the PSV ratio categories of less than 2.0, 2.0 through 3.0, more than 3.0, and occlusions. RESULTS: The overall ICC and Kappa values were 0.72 (95% CI, 0.63-0.79) and 0.53 for the aortoiliac tract and 0.85 (95% CI, 0.79-0.89) and 0.73 for the femoropopliteal tract. Agreement in the PSV ratio categories was 85% in the aortoiliac and 87% in the femoropopliteal tract. Interobserver variation increased markedly with increasing PSV ratio. In the PSV ratio category between 2.0 to 3.0, indicating a borderline stenosis, a substantial disagreement was found (aortoiliac, 1 of 8 agreement; femoropopliteal, 2 of 8 agreement). CONCLUSION: A moderate interobserver agreement was found in the duplex investigation of the aortoiliac and femoropopliteal arteries. One should be aware of this in clinical decision making, especially in cases of borderline stenoses. In these cases, repetition of the measurement or additional diagnostics is advocated