SP586A RETROSPECTIVE OBSERVATIONAL STUDY ON THE OUTCOMES OF SYNTHETIC ARTERIOVENOUS GRAFTS USED FOR HAEMODIALYSIS IN NHS GRAMPIAN

Abstract

Introduction and Aims: Reliable vascular access is essential for long-term haemodialysis. All access types have risks and benefits but there is a well established hierarchy with native arteriovenous fistulae (AVF) the best and temporary central venous catheters the worst. In some patients there may be no option for a native AVF in which case either a tunnelled central venous catheter or synthetic venous graft is required. The use of synthetic arteriovenous grafts varies considerably across UK dialysis centres, based largely on local experience and expertise. NHS Grampian has a consistently low prevalence of central venous catheters for long-term dialysis, and one reason may be a higher use of synthetic arteriovenous grafts. We have investigated the outcome of vascular access grafts used for haemodialysis in our centre. Methods: Data was collected from local databases for all synthetic arteriovenous grafts inserted between 1st January 2008 and 31st December 2012. Patency was recorded up to October 2013. Patients were censored by death, transplantation or reaching the end of the observation period. A graft was only placed if there was no fistula option available according to KDOQI guidelines. During the study period grafts from three manufacturers were used: standard (un-heparinised) polyterafluoroethylene (PTFE), heparin bonded PTFE and a tri-laminate expanded PTFE which allows immediate cannulation (“rapid access graft”). The type of graft was chosen according to surgeon preference and availability. Rapid access grafts were only used to avoid placement of a central venous catheter that would otherwise have been required for immediate dialysis access. Results: 68 Patients had 102 grafts placed during the five year period. 57% of patients were female and 48% were diabetic. 57% had a non-dominant straight upper arm graft while 25% had a similar graft in the dominant arm. 94% of patients were taking an anti-platelet or were anti-coagulated. Median graft survival was two years. No statistical difference in patency was observed between the different graft types (Kaplan Meier survival probability Log rank test p=0.97). After intervention for graft occlusion the median survival was two months, however 20% continued to be used for more than one year. Grafts were lost to infection in two cases. Two patients suffered from steal syndrome after graft placement requiring revision surgery. Conclusions: The performance of synthetic arteriovenous grafts in our study population was similar to that recorded in the best published studies. We were unable to demonstrate the claimed 10% improvement in patency of heparin bonded grafts; however our study was not a prospective study with groups matched and adequately powered. Rapid access grafts performed no worse than other PTFE in a more challenging patient group. Because of the instant needling capability of these grafts, we believe their role could be extended not only to avoid the placement of a central venous catheter but also to reduce the length of time an already placed line is left in situ.