Arterio-venous fistula: Is it critical for prolonged survival in the over 80's starting haemodialysis?

Abstract

Background: Dialysis in elderly patients (>80-years-old) carries a poor prognosis, but little is known about the most effective vascular access method in this age group. An arteriovenous fistula (AVF) is both time-consuming and initially expensive, requiring surgical insertion. A central venous catheter (CVC) is initially a cheaper alternative, but carries a higher risk of infection. We examined whether vascular access affected 1-year and 2-year mortality in elderly patients commencing haemodialysis. Methods: Initial vascular access, demographic and survival data for elective haemodialysis patients >80-years was collated using regional databases. A cohort of conservatively managed patients was included for comparison. A log-rank test was used to compare survival between groups and a chi-square test was used to compare 1-year and 2-year survival. Results: 167 patients (61% male) were included: CVC (101), AVF (25) and conservative management (41). Mean age (median) of starting haemodialysis (eGFR ≤10mL/min/1.73m2): CVC; 83.4 (2.3) and AVF; 82.3 (1.8). Mean age of conservatively managed patients reaching an eGFR ≤10mL/min/1.73m2 was 85.8 (3.6). Mean (median) survival on dialysis was 2.2 (1.8) years for AVF patients, 2.1 (1.2) for CVC patients, and 1.5 (0.9) for conservatively managed patients (p = 0.107, controlling for age/sex p = 0.519). 1-year and 2-year mortality: AVF (28%/52%); CVC (49%/57%), and conservative management (54%/68%). There was no significant difference between the groups at 1-year (p = 0.108) or 2-years (p = 0.355). Conclusion: These results suggest that there is no significant survival benefit over a 2-year period when comparing vascular access methods. In comparison to conservative management, survival benefit was marginal. The decision of whether and how (choice of their vascular access method) to dialysis the over 80s is multifaceted and requires a tailored, multidisciplinary approach.