Short-term mortality and revisions to promote maturation after arteriovenous fistula creation

Objective: Arteriovenous fistula (AVF) for hemodialysis access is traditionally considered superior to grafts due to infection resistance and purported improved patency. However, challenges to AVF maturation and limited patient survival may reduce AVF benefits. The objective of this study is to identify factors associated with risk of AVF requiring revision before maturation and/or mortality within 2 years of creation. Methods: We performed a retrospective review of 250 AVFs created between May 2017 and November 2020 at a single institution. Maturation was defined as the date the surgeon deemed the AVF ready for use or the patient successfully used the AVF for dialysis. The Risk Analysis Index was used to calculate frailty. The primary outcome was a composite of endovascular/surgical revision to promote maturation and/or mortality within 2 years of AVF creation (REVDEAD). The primary outcome was categorized as met if the patient required a revision to promote maturation or if the patient experienced mortality within 2 years of AVF creation, or if both occurred. REVDEAD was compared with those who did not meet the primary outcome and will be referred to as NOREVDEAD. Results: Survival at 2 years after AVF creation was 82%, and 54 (22%) patients underwent AVF revision. Of those, 31 (59%) patients progressed to AVF maturation. Of the 250 AVFs, 91 (36%) met the primary outcome of REVDEAD and 159 (64%) did not (NOREVDEAD). There was no difference between the REVDEAD and NOREVDEAD groups in age (P = .18), sex (P = .75), White race (P = .97), Hispanic ethnicity (P = .62), obesity (P = .76), coronary artery disease (P = .07), congestive heart failure (P = .29), diabetes mellitus (P = .78), chronic obstructive pulmonary disease (P = .10), dialysis status (P = .63), hypertension (P = .32), peripheral arterial disease (P = .34), or dysrhythmia (P = .13). There was no difference between the groups in the forearm vs the upper arm location of AVF (P = .42) or the vein diameter (P = .58). Forearm access, as opposed to upper arm AVF creation, was associated with higher rate of revision before maturation (P = .05). More patients in REVDEAD were frail or very frail (60% vs 48%, P = .05). Of the AVFs that matured, maturation required longer time in REVDEAD at 110.0 +/- 9.1 days vs 78.8 +/- 5.6 days (mean +/- standard deviation) (P = .003). Adjusted for the vein diameter and the forearm vs the upper arm, frailty increased the odds of REVDEAD by 1.9 (95% confidence interval: 1.1, 3.3). Conclusions: Frail patients who underwent AVF were significantly more likely to die within 2 years of AVF creation with no significant association between frailty and the need for revisions to promote maturation. Forearm AVFs were more likely to require revisions; in patients who are frail, with a high likelihood of 2-year mortality, graft may be more appropriate than AVF. If AVF is being considered in a frail patient, upper arm AVFs should be prioritized over forearm AVFs.