Eleven studies were included in the review (n=1,509 patients); one randomised controlled trial (RCT, n=101), one controlled comparative trial (n=116), and nine retrospective studies (n=1,292). Four studies scored seven or more stars in the validity assessment, five scored five or six stars and two scored less than five stars.
There was no significant difference between BBAVF and prosthetic upper limb arteriovenous grafts in primary failure rates (11 studies, n=1,509; OR 0.67, 95% CI 0.41 to 1.09) nor secondary failure rates (10 studies, n=1,439; OR 0.88, 95% CI 0.69 to 1.12). There was evidence of significant statistical heterogeneity for primary failure rates (p<0.00001), but not for secondary failure rates. There was a higher re-intervention rate with prosthetic upper limb arteriovenous grafts compared with BBAVF (1.32 per graft with prosthetic arteriovenous grafts and 0.54 with BBAVF).
Subgroup analysis revealed that BBAVF was associated with a significant reduction in the risk of primary failure (two studies, n=207; OR 0.30, 95% CI 0.15 to 0.58), but not secondary failure compared with prosthetic forearm grafts. There was no evidence of significant statistical heterogeneity. There was no significant difference between BBAVF and prosthetic upper arm grafts in primary and secondary failure rates.