Background:

Arteriovenous fistula (AVF) is a form of vascular access for haemodialysis. An AVF is normally created at the level of the wrist, but occasionally it is created in the elbow when there is no suitable vessel in the forearm. The most common type of elbow (antecubital) fistula (AFF) is a brachiocephalic fistula, which carries significantly higher risk of steal syndrome (AVF-associated hand ischaemia) than wrist fistulas. More recently, AFF using proximal radial or ulnar artery as inflow has been described and shown to have a lower rate of Steal syndrome than brachiocephalic fistula. This study aims to investigate the incidence of steal syndrome between AFF using brachial artery and that using the proximal radial/ulnar artery as inflow.

Arteriovenous fistulas (AVF) are the safest form of vascular access for long-term haemodialysis in patients with end-stage renal failure. The strategy in creating an AVF in the upper limbs is to start at a distal site and if that fails, to attempt an AVF on a more proximal site i.e. from wrist, forearm to elbow. The most common type of AVF is the radiocephalic AVF at the wrist. A more proximal AVF is often created as a primary procedure when there is poor vasculature in the distal forearm or as a secondary procedure when a wrist fistula has failed. Traditionally, brachiocephalic fistulas (BCF), which involves anastomosing the cephalic vein to the brachial artery, have been the most common type of AVF created in the antecubital fossa at the elbow level. Other common types of antecubital fossa arteriovenous fistula (AFF) are the brachiobasilic (BBF) and brachio-median cubital AVF.

Steal syndrome relates to hand ischaemia associated with AVF creation, and is a major risk of AVF formation. The symptoms of steal syndrome ranges from cold extremities, numbness, hand claudication (pain after exercise), to rest pain and tissue loss. Steal syndrome can also be measured by Digital Brachial Pressure Index. Severe steal syndrome is debilitating, and limb-threatening, and requires surgical revision or ligation of the AVF. This leads to additional surgical risks and loss of dialysis vascular access.

Diabetes and the types of AVF have been found to be independent risk factors for developing steal syndrome following AVF creation1. The highest risk is seen in patients with a proximal AVF i.e. BCF/BBF; up to 50% of patients in some studies, compared to 5-8% in all upper limb AVFs.

An alternative technique that may reduce risk of steal in this group of patients is to anastomose the vein to the radial artery or ulnar artery, distal to the brachial artery bifurcation. This technique, theoretically, will only 'steal' blood from one artery e.g. radial artery if the anastomosis is created on the proximal radial artery so blood flow can therefore be maintained by the ulnar arterial system.

Recent studies have suggested that using the proximal radial or ulnar artery reduced the risk of steal to as low as 0% to 3%. The type of arterial inflow to an AFF is therefore a potentially significant factor in causing steal syndrome. There is, however, no current randomised controlled trial to prove this hypothesis.

The definition of steal syndrome varies greatly in the literature. Some studies have defined steal syndrome as the presence of mild symptoms such as cold hand, while the others reported steal syndrome when it was severe enough to require surgical intervention. This has led to the huge variations in the incidence of steal syndrome being reported and has made comparison difficult between studies. A few scoring systems to describe the severity of steal syndrome have been suggested in previous studies, but none of them has been widely used.

In this study, the difference in the severity of steal between the two intervention groups will be investigated. This will be done using the Hoek score, which was originally used by Hoek et al in 2006 to report steal syndrome associated with the AVFs created in their centre. There was, however, no comparison of scores among the different types of AVF.

Inclusion Criteria:

End stage renal failure Requiring AVF at in the antecubital fossa for haemodialysis

Exclusion Criteria:

When the participant has not received the allocated intervention due to anatomical anomaly, technical issue or any other reason