Direct duplex scanning parameters in the diagnosis of renal artery stenosis: a study to validate and optimize cut-off points.

OBJECTIVE To test the efficiency of the direct duplex scanning parameters in the diagnosis of renal artery stenosis (RAS), and verify whether or not the cut-off points recommended by medical literature are the most appropriate means to distinguish lesion severity. METHODS Prospective study, including 62 patients with RAS, submitted to a duplex ultrasound scan and selective arteriography. The peak systolic velocity (PSV) and the renal-aortic ratio (RAR) were measured. Statistical analysis included the ROC (receiver operating characteristic curve), unpaired students t-test, sensitivity, specificity, positive and negative predictive values and accuracy. RESULTS The arteriography revealed RAS 0-59% in 31 arteries (24%); RAS 60-99% in 91 arteries (72%) and 5 occlusions (4%). ROC analysis demonstrated that the lesion detection efficiency of PSV and RAR were similar with areas below the curves of 0.96 and 0.95, respectively. Considering the cut-off points recommended by medical literature, PSV of 180 cm/sec presented a sensitivity of 100% and specificity of 81%, while the RAR of 3.5 presented a sensitivity of only 79%, with a specificity of 93%. These parameters were analyzed in conjunction (direct criteria) revealing a sensitivity of 79% and specificity of 97%. The optimized cut-off points were: PSV of 189 cm/sec and RAR of 2.6, demonstrating sensitivity and specificity rates of 100%, 87%, 96% and 87% respectively. CONCLUSION The individual use of optimized peak systolic velocity (PSV) was the most efficient parameter in the detection and grading of RAS.