Physical examination of arteriovenous fistulae by a renal fellow: does it compare favorably to an experienced interventionalist?

Physical examination (PE) has been highlighted to detect vascular access stenosis with high degree of accuracy when performed by an interventional nephrologist (IN) with expertise in physical examination. This study examines the accuracy of PE compared with angiography when performed by a nephrology fellow (NF). It also compares NF results to that of IN. Didactic and hands-on PE training was provided to a renal fellow for 1 month during an interventional nephrology rotation. Forty-five and 142 consecutive cases of arteriovenous fistula dysfunction were examined by the NF and IN, respectively. Preprocedure PE was performed by the NF and IN and the finding secured in a sealed envelope. Angiography from the feeding artery to the right atrium was then performed. The images were reviewed by an independent interventionalist with expertise in endovascular dialysis access procedures and the diagnosis was rendered. The reviewer was blinded to the physical examination. Cohen's Kappa was used as a measurement of the level of agreement beyond chance between the diagnosis made by physical examination and angiography. Outflow stenosis: NF [strong agreement (81%), Kappa value = 0.63]; IN [strong agreement (89%), Kappa score = 0.78]. Inflow stenosis: NF [strong agreement (80%), Kappa value = 0.56]; IN [strong agreement (83%), Kappa score = 0.55]. These differences between NF and IN were not significant. NF performed significantly better than the IN regarding central vein stenosis. NF [strong agreement (79%), Kappa value = 0.44]; IN [weak agreement (11%), Kappa value = 0.17]. An NF can be trained in physical examination and accurately detect and localize stenoses in a great majority of arteriovenous fistulae when compared with an IN. We suggest that nephrology training programs should place more emphasis on this aspect of vascular access education.