Since the publication of the first vascular access clinical practice guidelines in 1997, the global nephrology community has dedicated significant time and resources toward increasing the prevalence of arteriovenous fistulas and decreasing the prevalence of central venous catheters for hemodialysis. These efforts have been bolstered by observational studies showing an association between catheter use and increased patient morbidity and mortality. To date, however, no randomized comparisons of the outcomes of different forms of vascular access have been conducted. There is mounting evidence that much of the difference in patient outcomes may be explained by patient factors, rather than choice of vascular access.Some have called into question the appropriateness of fistula creation for certain patient populations, such as those with limited life expectancy and those at high risk of fistula-related complications. In this review, we explore the extent to which catheters and fistulas exhibit the characteristics of the "ideal" vascular access and highlight the significant knowledge gaps that exist in the current literature.Further studies, ideally randomized comparisons of different forms of vascular access, are required to better inform shared decision making.Twenty years ago, fistulas were first promoted as the vascular access of choice by the Kidney Disease Outcomes Quality Initiative.1 Since then, the nephrology community has worked to increase fistula prevalence and reduce the use of central venous catheters. These efforts have been supported by initiatives such as Fistula First, 2 clinical practice guidelines from around the world, [3][4][5][6] and in some jurisdictions, by financial incentives which reward dialysis centers with high fistula prevalence and penalize those with high catheter prevalence. 7 In 2007, a modification of the Fistula First Initiative to "Fistula First, Catheter Last" was proposed, re-emphasizing the perceived importance of minimizing catheter use. 8 However, critics of "Fistula First, Catheter Last" have expressed concern that all comparisons between access types have been observational, subject to high degrees of bias, 9,10 and unable to account for other patient factors which likely contribute to poor outcomes seen in patients treated with catheters. 11,12 Others have questioned the utility of fistula creation in elderly patients and in patients at high risk for fistula failure or complications, arguing that such patients may be better served by catheters. 13 Surveys of the nephrology community suggest that many clinicians feel that a catheter is often appropriate and that there is equipoise regarding the ideal form of vascular access in many situations. 14,15 In fact, enrollment recently began for a pilot randomized trial comparing outcomes in incident hemodialysis patients, over the age of 65, who start dialysis with a catheter. Patients will be randomized to creation of a fistula vs continued use of a catheter (clinicaltrials.gov NCT02675569).
16The ideal vascular access has...