Kidney Week

Abstract: TH-OR112

Arteriovenous Fistulas, Grafts, or Central Venous Catheters as Vascular Access for Older Adults Starting Hemodialysis: A Randomized Controlled Trial

Session Information

• Late-Breaking Research Orals: Dialysis and Transplantation
  October 22, 2026 | Location: Four Seasons Ballroom 2, Convention Center
  Abstract Time: 04:30 PM - 04:42 PM

Category: Dialysis

• 803 Dialysis: Vascular Access

Authors

• Heggen, Boudewijn, Maastricht Universitair Medisch Centrum+, Maastricht, LI, Netherlands

Boudewijn Heggen, MD

• van der Bogt, Koen E.A., Leids Universitair Medisch Centrum, Leiden, ZH, Netherlands

Koen E.A. van der Bogt, MD, PhD

• Dekker, Friedo W., Leids Universitair Medisch Centrum, Leiden, ZH, Netherlands

Friedo W. Dekker, PhD

• de Haan, Michiel Willem, Universiteit Maastricht, Maastricht, LI, Netherlands

Michiel Willem de Haan, MD, PhD

• Hemmelder, Marc H., Erasmus Universiteit Rotterdam, Rotterdam, ZH, Netherlands

Marc H. Hemmelder, MD, PhD

• Loon, Magda Van, Maastricht Universitair Medisch Centrum+, Maastricht, LI, Netherlands

Magda Van Loon, PhD, NP

• Rotmans, Joris I., Leids Universitair Medisch Centrum, Leiden, ZH, Netherlands

Joris I. Rotmans, MD, PhD

• Schurink, Geert Willem H., Maastricht Universitair Medisch Centrum+, Maastricht, LI, Netherlands

Geert Willem H. Schurink

• Tordoir, Jan, Maastricht Universitair Medisch Centrum+, Maastricht, LI, Netherlands

Jan Tordoir, Md, PhD

• Snoeijs, Maarten G., Maastricht Universitair Medisch Centrum+, Maastricht, LI, Netherlands

Maarten G. Snoeijs, MD, PhD

Group or Team Name

• OASIS Study Group

Background

Elderly patients starting hemodialysis have an increased risk of early arteriovenous fistula (AVF) complications due to maturation failure, while the long-term durability benefit may be limited by short life expectancy. We hypothesized that arteriovenous grafts (AVG) or central venous catheters (CVC) would result in better outcomes than AVFs.

Methods

In this multicenter randomized controlled trial, we enrolled patients ≥70 years with life expectancy <2 years (Dusseux risk score ≥4) who were expected to start hemodialysis within 6 months or had started within the previous 6 months with a CVC. Patients had to be eligible for all access types based on ultrasound vessel mapping. Participants were randomized to AVF, AVG, or CVC strategies and followed until study closeout at 1 year after the last enrollment. The primary endpoint was the rate of access-related interventions per patient-year. Safety endpoints were mortality, access-related serious adverse events (SAEs), and vascular access infections. Intervention groups were compared with the AVF group using negative binomial and Cox regression models.

Results

A total of 165 patients were enrolled: 52 AVF, 57 AVG, and 56 CVC. Mean age was 78 years; 27% had started hemodialysis at enrollment. Median follow-up was 2.0 years. The allocated vascular access type was changed in 9 AVF, 24 AVG, and 4 CVC patients during follow-up (P=0.002). The rate of access-related interventions was 1.12/year with AVF, 1.53/year with AVG (P=0.082), and 0.76/year with CVC (P=0.018). Mortality did not differ significantly. The rate of access-related SAEs was 0.08/year with AVF, 0.27/year with AVG (P=0.012), and 0.11/year with CVC (P=0.418). Access-related infection rates did not differ significantly; the overall catheter-related bloodstream infection rate was 0.37/1000 catheter days.

Conclusion

In elderly patients with limited life expectancy, CVCs required fewer access-related interventions than AVFs, without excess mortality or infection. AVGs are more often abandoned and result in more access-related SAEs than AVFs.

Access-related interventions

Funding

• Government Support - Non-U.S.