A national study of autogenous arteriovenous access use and patency in a contemporary hemodialysis population.

OBJECTIVE

The predicted outcomes of autogenous arteriovenous (AV) hemodialysis access creation are predominantly based on historical data; however, both the hemodialysis population and clinical practices have changed significantly during the last decade. This study examined contemporary AV access clinical use and patencies.

METHODS

A multicenter observational cohort study was performed of all new AV accesses created in Scotland in 2015. The primary end point was efficacy assessed by successful AV access use for a minimum of 30 days and primary, primary assisted, and secondary patency at 1 year. Data obtained included all interventions to maintain or to restore patency. Predictors of patency loss including demographics, comorbid conditions, dialysis status, AV access location, duplex ultrasound surveillance, procedures, prior access, and antiplatelets were assessed. Kaplan-Meier and competing risks analyses were performed to estimate the probability of AV access failure. All patients were followed up for at least 1 year or had a censoring event.

RESULTS

A total of 582 AV accesses were created in 537 patients (mean age, 60 [standard deviation, 14] years; 60% men; 42% with diabetes) in nine adult renal centers. Mean follow-up was 11.8 (standard deviation, 7.6) months. By the end of the follow-up, 322 (55.3%) AV accesses were successfully used for dialysis. At 1 year, 48% (95% confidence interval [CI], 44-52) of AV accesses had primary patency, (95% CI, 63-71) had primary assisted patency, and 69% (95% CI, 65-73) had secondary patency. The leading cause of primary patency loss was primary failure (30%). An average of 0.48 intervention per patient-year was required to maintain patency. On multivariable analysis, patency was better for an upper arm than for a forearm AV access (1-year secondary patency of upper arm vs forearm AV accesses, 74% vs 58%). The cumulative hazard and incident functions for AV access failure were 31% (95% CI, 27-35) and 23% (95% CI, 20-27) at 1 year, respectively.

CONCLUSIONS

Despite advances in recent years with preoperative vessel assessment and surveillance, patency rates have not improved, with primary failure remaining the major obstacle. Competing events should be taken into consideration; otherwise, biases may occur with overestimation of the probability of AV access failure.