Vessel diameter and close surveillance helps predict early patency in native arteriovenous fistulas

Arteriovenous fistulas are the preferred method of vascular access for haemodialysis. Data suggests patency rates can be low and may be related to vessel diameters prior to creation. We use specific size criteria for fistula selection. We aimed to establish patency rates in relation to vessel size and whether other factors affect fistula patency.

Consecutive patients undergoing radiocephalic (RCF) or brachiocephalic (BCF) creation between 2016 and 2018 were analysed. Preoperative arterial and venous diameters were collected. Six-week and six-month primary and secondary patency rates were analysed to establish any impact of vessel size on patency and re-intervention rates between groups. A univariate analysis was performed.

Ninety four RCFs and 101 BCFs were created. Median artery and vein size for RCF were 2.7 and 3.0 mm respectively. For BCF, they were 4.6 and 4.3 mm respectively. At 6-weeks, overall satisfactory patency for RCF and BCF combined was 91.8%. 89.7% demonstrated primary patency; 2.1% secondary patency. At 6-months, overall patency was 78.7%; 58.5% demonstrated primary patency, 20.2% secondary patency. A univariate analysis, for both groups, revealed vein size was a significant predictor of overall satisfactory patency at 6-weeks, with larger veins more likely to remain patent (p = 0.025 RCF, p = 0.007 BCF). However, artery size was not predictive (p = 0.1 RCF, p = 0.5 BCF). At 6-months, neither artery nor vein diameter were predictive in either group. When comparing size of vessel based on fistula type, vessels used to create RCFs were smaller than those for BCFs (p < 0.001). RCFs were more likely to receive endovascular intervention or occlude when compared to BCFs (p = 0.014).

Excellent patency and maturation rates can be achieved using fairly strict vessel size criteria. Vein size might be the more important predictor of early success. RCFs can be challenging due to smaller vessels, but maturation rates can be optimised by close surveillance and aggressive re-intervention.