Objective: To compare two methods of detecting graft stenoses after infrainguinal bypass.
Design: Prospective study.
Setting: Vascular Laboratory, University Hospital.
Materials: 110 infrainguinal graft studies (60 vein, 50 PTFE) in 74 patients were performed prospectively to detect graft-related stenoses.
Chief outcome measures: The diagnostic accuracy of computer assisted impedance analysis was assessed using Quickscan (QS) as the gold standard in the detection of graft-related stenoses. CHIEF RESULTS: QS showed occlusion in 4 grafts (two vein and two PTFE), no stenosis in 86 graft studies and stenoses in 20 studies. All 12 stenoses with a frequency ratio > or = 1.4, were confirmed with intraarterial digital subtraction angiography (IADSA). Eight stenoses with a frequency ratio of 1:3 continued graft surveillance. The median thigh impedance score of vein grafts with QS confirmed stenoses was 0.51 (0.36-0.70) compared with 0.44 (0.30-0.60) for non-stenosed vein grafts (p = 0.015, Mann-Whitney U test). The median thigh impedance score in PTFE graft with QS confirmed stenosis was 0.58 (0.53-0.76) compared with 0.42 (0.28-0.73 for non-stenosed grafts (p = 0.0001). An impedance score > 0.45 has been suggested for detection of "at risk" grafts. Using QS as the gold standard, impedance assessment gave 90% sensitivity, 63% specificity in the thigh; 80%, 52% in the calf and 90%, 46% taking the higher score on calf for thigh data. Taking a QS frequency ratio of 1:4 as indicating a significant stenosis (50% diameter reduction), 11% (12/106) of surveillance studies went on to intervention, that is 12/74 (16%) grafts.
Conclusions: If the higher impedance score derived from either the calf or thigh was used to detect stenoses, 60% (64/106) of graft studies would have been referred for intervention. We believe this high level of intervention is unrealistic and cannot therefore recommend impedance analysis for graft surveillance.