Kellie Wo1, Byron J Morrison1, Russell N Harada2. 1. Department of Surgery, Pali Momi Medical Center, Aiea, HI. 2. Department of Surgery, Pali Momi Medical Center, Aiea, HI. Electronic address: docduffer@aol.com.
Abstract
BACKGROUND: The purpose of this study was to determine the best color-flow duplex ultrasound (US) criteria for the detection of significant stenoses in arteriovenous fistulas (AVFs) by comparing US with the angiographic gold standard. METHODS: Included in this retrospective study were all patients who had both US scans and angiograms of their AVFs performed from December 2008 to July 2015. Comparisons were made between the preoperative US peak systolic velocity (PSV) measurements and the angiographic images looking at the percent diameter reduction in the stenosis. The PSV ratio between the area of stenosis and the normal segment on the inflow side of the stenosis (systolic velocity ratio [SVR]) was also compared with the angiograms. Sensitivity and positive predictive values (PPV) with 95% confidence intervals were calculated. RESULTS: This study included 780 cases of stenotic accesses (47 brachiobasilic, 361 brachiocephalic, and 372 radiocephalic). PSVs of ≥500 cm/s predicted a 50% or greater stenosis with a sensitivity of 89% (95% confidence interval 87-91) and PPV of 99% (99-100). At lower PSVs, sensitivity and PPV remained high; however, there were an increased number of false positive cases compared with those at 500 cm/sec or greater. The SVR was unreliable in predicting the degree of stenosis in AVF. CONCLUSION: PSVs of ≥500 cm/sec are generally reliable in predicting stenosis of 50% or greater in AVFs.
BACKGROUND: The purpose of this study was to determine the best color-flow duplex ultrasound (US) criteria for the detection of significant stenoses in arteriovenous fistulas (AVFs) by comparing US with the angiographic gold standard. METHODS: Included in this retrospective study were all patients who had both US scans and angiograms of their AVFs performed from December 2008 to July 2015. Comparisons were made between the preoperative US peak systolic velocity (PSV) measurements and the angiographic images looking at the percent diameter reduction in the stenosis. The PSV ratio between the area of stenosis and the normal segment on the inflow side of the stenosis (systolic velocity ratio [SVR]) was also compared with the angiograms. Sensitivity and positive predictive values (PPV) with 95% confidence intervals were calculated. RESULTS: This study included 780 cases of stenotic accesses (47 brachiobasilic, 361 brachiocephalic, and 372 radiocephalic). PSVs of ≥500 cm/s predicted a 50% or greater stenosis with a sensitivity of 89% (95% confidence interval 87-91) and PPV of 99% (99-100). At lower PSVs, sensitivity and PPV remained high; however, there were an increased number of false positive cases compared with those at 500 cm/sec or greater. The SVR was unreliable in predicting the degree of stenosis in AVF. CONCLUSION: PSVs of ≥500 cm/sec are generally reliable in predicting stenosis of 50% or greater in AVFs.