RESUMO
OBJECTIVE: A well functioning arteriovenous fistula (AVF) is essential for haemodialysis. Despite regular duplex ultrasound (DUS) a significant number of AVFs fail. Tomographic 3D ultrasound (tUS) creates a 3D image of the AVF that can be interpreted by the clinician. DUS, tUS, and fistulograms were compared for the identification and measurement of flow limiting stenosis. METHODS: Patients with AVF dysfunction on routine Transonic surveillance, defined as (1) > 15% reduction in flow on two consecutive occasions, (2) > 30% reduction in flow on one occasion, (3) flow of < 600 mL/sec, (4) presence of recirculation, underwent DUS. AVF tUS imaging was performed prior to fistulography. All fistulograms were reported by the same consultant radiologist and tUS images by the same vascular scientist blinded to the fistulogram results. Maximum diameter reduction in all stenoses were measured using all three imaging techniques. RESULTS: In 97 patients with 101 stenoses, the mean (± standard deviation [SD]) severity of stenosis was 63.0 ± 13.9%, 65.0 ± 11.6%, and 64.8 ± 11.7% for the fistulograms, DUS, and tUS respectively. The mean (± SD) time between ultrasound and fistulography imaging was 15.0 ± 14.5 days. Assuming the fistulogram as the "gold standard", Bland-Altman agreement for DUS was -1.9 ± 15.5% (limit of agreement [LOA] -32.2 - 28.4) compared with -1.7 ± 15.4% (LOA -31.9 - 28.4) for tUS. Median (± interquartile range) time to complete the investigation was 09:00 ± 03:19 minutes for DUS and 03:13 ± 01:56 minutes for tUS (p < .001). CONCLUSION: DUS and tUS were equally accurate at detecting AVF complications but tUS investigation requires less skill and was significantly quicker than DUS.