Association of Frailty with Postoperative Survival and Outcomes Following Hemodialysis Vascular Access Creation.

Background: The literature suggests that for patients to experience the purported advantages of an arteriovenous fistula (AVF) over arteriovenous graft (AVG), a minimum survival of 18 months is required. With the vascular access guideline shift away from "Fistula First" toward shared decision making, patient survival after vascular access creation is a major factor to consider in optimal access selection. The objective of this study is to examine outcomes of vascular access in patients with short survival and factors associated with short survival, including frailty.

Methods: We performed a retrospective review of 200 access procedures performed between August 2018 and November 2020 at a single institution. Maturation was defined as the date when the surgeon deemed the access ready to be used for dialysis. A modified Risk Analysis Index (RAI) score was used to calculate frailty.

Results: Within 3 years after access creation, 55 (27.5%) patients were recorded as dead (mortality within 3 years of access creation [3YMORT]). In the 3YMORT group, 5 did not follow-up with the surgeon prior to death and 22/34 (65%) of AVF versus 15/16 (94%) of AVGs were deemed mature prior to death (P = 0.03). Of the accesses that matured, the median days to maturation for AVF was 69 (interquartile range [IQR] 53, 87) versus 28 (IQR 18, 32) for AVG (P < 0.001). Patients in the 3YMORT group were older (70.6 vs. 63.4, P = 0.004) and had a lower body mass index (24.8 vs. 27.4, P = 0.03). Patients in the 3YMORT group had higher prevalence of dysrhythmia (35% vs. 15%, P = 0.002), chronic obstructive pulmonary disorder (20% vs. 10%, P = 0.048) and dialysis dependence at the time of access creation (91% vs. 75%, P = 0.01). There was no significant difference in sex, white race, Hispanic ethnicity, coronary artery disease, congestive heart failure, previous coronary artery bypass graft or percutaneous coronary intervention, diabetes, hypertension, and peripheral arterial disease between the 2 groups. The 3YMORT group had a significantly higher prevalence of frailty (78% vs. 49%, P = 0.0002). Patients categorized as frail by the RAI had a significantly higher risk of 3YMORT (odds ratio [OR] 3.74, 95% confidence interval [CI] 1.82-7.66) compared to nonfrail patients. Patients categorized as very frail by the RAI had an even higher risk of 3YMORT (OR 4.20, 95% CI 1.95-9.05), compared to nonfrail patients.

Conclusions: Patients with short life expectancy after vascular access creation may have high rates of AVF nonmaturation and longer time to maturation. Factors associated with high risk of mortality within 3 years of vascular access creation correlate well with factors included in the RAI frailty score. Patients who are frail or very frail may be appropriate candidates for AVG creation over AVF considering their high risk for short life expectancy.