Duplex ultrasound compared with digital subtraction angiography in diagnosing significant stenosis of vascular access

Purpose: This study evaluated the diagnostic accuracy of duplex ultrasound (DUS) using peak systolic velocity (PSV), PSV ratio (PSVR), and volume flow in detecting vascular access stenosis, compared to digital subtraction angiography. Methods: A cohort of 121 patients with arteriovenous (AV) fistulas or grafts were evaluated between January 2013 and September 2013 for clinical indicators of dysfunction, including reduced thrill, weak bruit, decreased dialysis flow, high venous pressure or excessive post-dialysis bleeding using our own collective diagnostic criteria based on previously published parameters. Critical stenosis was defined as PSV >400 cm/s, PSVR >3.5, and volume flow <400 mL/min. Angiography was performed for all hemodynamically significant stenoses diagnosed by DUS. Results: Among 121 patients (mean age, 56.6 ± 17.3 years; 64% men), DUS findings included normal studies (24%), mild stenosis (9.9%), moderate stenosis (15.7%), severe stenosis (40.5%), and total occlusions (9.9%). Stenosis locations were at the anastomotic site (86.3%), the junctional (11.3%), and the venous outflow (2.5%). Critical stenosis was identified in 49 (40.5%) patients. Of 46 patients who underwent angiography, critical stenosis was confirmed in 44. DUS demonstrated a positive predictive value (PPV) of 95.6% for detecting significant stenosis, supporting its reliability in vascular access assessment. Study limitations include its small sample size, retrospective design, and limited generalizability. Furthermore, analysis was limited to positive sonographic findings precluding sensitivity/specificity assessment. Conclusion: DUS is accurate in diagnosing severe vascular access stenosis, with a PPV exceeding 95%. Its non-invasive nature and high predictive value support its use as a first-line imaging modality for vascular access evaluation.