PURPOSE: To characterize the accuracy of color-flow duplex ultrasound (DUS) in planning lower extremity revascularization procedures, we prospectively compared operations predicted by means of DUS arterial scanning (DUSAS) and operations predicted by means of conventional angiography (CA) with actual operations performed in 36 patients undergoing 40 vascular reconstructions for critical (grade II/III) lower extremity ischemia. METHODS: All patients were examined with lower extremity DUSAS followed by CA. DUSAS was performed from the aorta to the pedal vessels of the affected extremity. Adequacy of inflow was assessed, and the best distal target vessel with continuous, unobstructed flow was defined. An operative prediction was made and recorded based upon the DUSAS findings, and in a blinded fashion, based upon subsequent CA. The McNemar test for comparing correlated proportions was applied to test for the statistical significance of the difference (P < .05) between correct operations predicted by DUSAS and CA. RESULTS: Of the actual operations performed, 83% were correctly predicted by means of DUSAS (95% CI; range, 77% to 89%). Seven operations were incorrectly predicted with DUSAS. Of the actual operations performed, 90% were correctly predicted by means of CA (95% CI; range, 81% to 99%). Four operations were incorrectly predicted with CA. The McNemar test determined that the difference between correct operations predicted by means of DUSAS and correct operations predicted by means of CA was not statistically significant (P = .50). CONCLUSIONS: With few exceptions, DUSAS can be used to reliably predict infrainguinal reconstruction strategies. Vessels defined as adequate with DUSAS are rarely unfit for bypass. Prospective investigation of lower extremity revascularization based solely upon DUSAS is warranted.
PURPOSE: To characterize the accuracy of color-flow duplex ultrasound (DUS) in planning lower extremity revascularization procedures, we prospectively compared operations predicted by means of DUS arterial scanning (DUSAS) and operations predicted by means of conventional angiography (CA) with actual operations performed in 36 patients undergoing 40 vascular reconstructions for critical (grade II/III) lower extremity ischemia. METHODS: All patients were examined with lower extremity DUSAS followed by CA. DUSAS was performed from the aorta to the pedal vessels of the affected extremity. Adequacy of inflow was assessed, and the best distal target vessel with continuous, unobstructed flow was defined. An operative prediction was made and recorded based upon the DUSAS findings, and in a blinded fashion, based upon subsequent CA. The McNemar test for comparing correlated proportions was applied to test for the statistical significance of the difference (P < .05) between correct operations predicted by DUSAS and CA. RESULTS: Of the actual operations performed, 83% were correctly predicted by means of DUSAS (95% CI; range, 77% to 89%). Seven operations were incorrectly predicted with DUSAS. Of the actual operations performed, 90% were correctly predicted by means of CA (95% CI; range, 81% to 99%). Four operations were incorrectly predicted with CA. The McNemar test determined that the difference between correct operations predicted by means of DUSAS and correct operations predicted by means of CA was not statistically significant (P = .50). CONCLUSIONS: With few exceptions, DUSAS can be used to reliably predict infrainguinal reconstruction strategies. Vessels defined as adequate with DUSAS are rarely unfit for bypass. Prospective investigation of lower extremity revascularization based solely upon DUSAS is warranted.