Arteriovenous fistulae vs. arteriovenous grafts: a retrospective review of 1,700 consecutive vascular access cases

J Vasc Access. Oct-Dec 2008;9(4):231-5.


Objective: Vascular access (VA) procedures are rapidly becoming the most prevalent surgery in the United States. It is estimated that there will be over 500,000 VA procedures done this year. Previously, surgeons in the US were attempting many more non-autogenous grafts than autogenous fistulae. In recent years, there has been a great push called 'Fistula First' to promote arteriovenous fistulae (AVF) as the first line of treatment vs. nonautogenous grafts. The goal of this investigation is to determine if too many fistulae are now being performed without attention to specific patient profiles.

Methods: A retrospective review of 1700 consecutive cases was performed by one surgeon at one institution between 1997 and 2005. Patients were categorized by demographics, co-morbidities, previous access procedures, access location, and type of graft. Patency was calculated. Kaplan-Meier, Cox regression and the Log Rank Test were used to analyze data. Access endpoints and complications were also documented.

Results: The study reviewed 1700 procedures. The median age was 52 (60.2% male) with 58.7% fistulae and 41.3% grafts. Median patency time was 10 months, with no statistically significant difference between access types. There was no significant difference in length of patency when comparing upper arm (70.1%), lower arm (24.5%) and thigh (5.4%). Graft infection rate was 9.5% and fistula infection rate was 0.9% (p<0.001). The overall infection rate was 4.5%, and decreased patency significantly (4 vs. 11 months). Thrombosis occurred in 24.7% of grafts and 9.0% of fistulae. Thrombosed grafts had better salvage rates (8 vs. 4 months, p<0.001). The data showed diabetes, HTN and HIV have no overall impact on patency.

Conclusions: AVF and grafts are both useful in providing VA for patients requiring hemodialysis. Our data shows that grafts are equivalent in long-term patency. Therefore, it is apparent in those patients who are not candidates for an AV fistula; an AV graft for VA should be placed.

Publication types

  • Comparative Study

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Arteriovenous Shunt, Surgical* / adverse effects
  • Blood Vessel Prosthesis Implantation / adverse effects
  • Blood Vessel Prosthesis Implantation / instrumentation*
  • Blood Vessel Prosthesis* / adverse effects
  • Female
  • Graft Occlusion, Vascular / etiology
  • Humans
  • Kaplan-Meier Estimate
  • Male
  • Middle Aged
  • Patient Selection
  • Practice Guidelines as Topic
  • Proportional Hazards Models
  • Prosthesis-Related Infections / etiology
  • Renal Dialysis*
  • Reoperation
  • Retrospective Studies
  • Risk Assessment
  • Surgical Wound Infection / etiology
  • Thrombosis / etiology
  • Time Factors
  • Treatment Outcome
  • Vascular Patency*
  • Young Adult