Abstract
BACKGROUND: Arteriovenous fistulas (AVFs) are preferred over arteriovenous grafts (AVGs) for haemodialysis, but effectiveness may vary by fistula subtype, access site, and transposition. METHODS: In this systematic review, PubMed, Web of Science, Scopus, and CENTRAL were searched up to 11 January 2025 for comparative studies of AVF versus AVG in adults. Random-effects meta-analyses pooled odds ratios (ORs) with 95% confidence intervals, standardized to 1-year outcomes (2-year effects summarized secondarily). Prespecified subgroups were AVF site (forearm versus upper arm), AVG site (forearm, upper arm, lower limb), AVF type (radiocephalic, brachiocephalic, brachiobasilic, basilic, brachial-brachial), and transposition. Observational studies were appraised with the National Institute of Health tool; randomized clinical trials with RoB 2. Meta-regression evaluated age, diabetes, and hypertension; certainty was appraised with GRADE. RESULTS: Sixty-three studies (357 333 patients: 226 078 AVF; 131 255 AVG) were included. At 1 year, AVF was associated with higher primary (OR = 1.61, 95% confidence interval 1.19 to 2.18), primary-assisted (OR = 1.69, 1.46 to 1.96), and secondary patency (OR = 1.69, 1.18 to 2.44), and lower overall complications (OR = 0.52, 0.34 to 0.78) and mortality (OR = 0.57, 0.34 to 0.98) versus AVG; primary failure and revision did not differ. Benefits were marked for transposed AVFs (primary-assisted OR = 2.07, 1.45 to 2.95; complications OR = 0.62, 0.43 to 0.88) and for upper-arm AVFs in primary patency (OR = 1.56, 1.05 to 2.31). Basilic AVF improved secondary patency (OR = 2.31, 1.17 to 4.57). Infection and thrombosis risks were lower for basilic and brachiobasilic AVF and upper-arm AVF. Meta-regression suggested diabetes modified effects (better secondary patency; fewer revisions), whereas age and hypertension did not. Certainty was generally low to moderate. CONCLUSION: When standardized at 1 year, AVFs outperform AVGs for patency, complications, and mortality, with magnitude varying by site, transposition, and fistula type. Findings support individualized access planning and the role of transposed upper-arm AVFs when feasible. REGISTRATION NUMBER: CRD420251125422 (https://www.crd.york.ac.uk/PROSPERO/home).