A comparison between colour duplex ultrasonography and arteriography for imaging infrapopliteal arterial lesions

To investigate the agreement between colour duplex ultrasonography and digital subtraction arteriography of the infrapopliteal arteries.

Retrospective, blinded study.

Vascular laboratory and Radiology Department, University Hospital.

The infrapopliteal vasculature was examined in a total of 51 limbs by both colour duplex ultrasound and digital subtraction angiography. By examining all arteries from the distal popliteal to the pedal arteries, a total of 204 individual arterial segments were available for analysis. Each segment was graded as 0-49%, 50-99% diameter reduced or occluded by both modalities. Using ultrasound, classification of stenoses was achieved by observing peak systolic velocity ratios; a doubling of peak systolic velocity indicating a > or = 50% diameter reducing stenosis. Where no Doppler signal could be obtained, the vessel was assumed to be occluded. From angiographic studies, two radiologists separately and blindly assessed the extent of disease for each infrapopliteal artery noting areas of > or = 50% diameter reduction and occlusion. The Kappa statistic was used to examine the level of agreement between angiography and ultrasound as well as between both radiologists.

The Kappa level (95% confidence interval) of agreement between ultrasound and angiographic assessments for distinguishing patent from occluded segments was 0.61 (0.49-0.74) for all segments. The equivalent agreement between radiologists was 0.80 (0.70-0.89). Poorest agreement was observed from ultrasound assessments of the peroneal and tibioperoneal trunk arterial segments.

Since agreement between colour duplex scanning and angiography never fell significantly below levels achieved between two radiologists, we conclude that colour duplex ultrasound can be used to assess infrapopliteal artery patency.