Date Published: January 28, 2019
Publisher: Public Library of Science
Author(s): Seonjeong Jeong, Hyunwook Kwon, Jai Won Chang, Min-Ju Kim, Khaliun Ganbold, Youngjin Han, Tae-Won Kwon, Yong-Pil Cho, Frank J. M. F. Dor.
In an incident hemodialysis (HD) population, we aimed to investigate whether arteriovenous fistula (AVF) creation before HD initiation was associated with improved AVF patency compared with AVF creation from a central venous catheter (CVC), and also to compare patient survival between these patients. Between January 2011 and December 2013, 524 incident HD patients with identified first predialysis vascular access with an AVF (pre-HD group, n = 191) or an AVF from a CVC (on-HD group, n = 333) were included and analyzed retrospectively. The study outcome was defined as AVF patency and all-cause mortality (time to death). On Kaplan–Meier survival analysis, primary and secondary AVF patency rates did not differ significantly between the two groups (P = 0.812 and P = 0.586, respectively), although the overall survival rate was significantly higher in the pre-HD group compared with the on-HD group (P = 0.013). On multivariate analysis, well-known patient factors were associated with decreased primary (older age and diabetes mellitus [DM]) and secondary (DM and peripheral arterial occlusive disease) AVF patency, whereas use of a CVC as the initial predialysis access (hazard ratios, 1.84; 95% confidence intervals, 1.20–2.75; P = 0.005) was significantly associated with worse survival in addition to well-known patient factors (older age, diabetes mellitus, and peripheral arterial occlusive disease). Worse survival in the on-HD group was likely confounded by selection bias because of the retrospective nature of our study. Therefore, the observed lower mortality associated with AVF creation before HD initiation is not fully attributable to CVC use, but rather, affected by other patient-level prognostic factors. There were no CVC-related complications in the pre-HD group, whereas 10.2% of CVC-related complications were noted in the on-HD group. In conclusion, among incident HD patients, compared with patients who underwent creation of an AVF from a CVC, initial AVF creation showed similar primary and secondary AVF patency rates, but lower mortality risk. We also observed that an initial CVC use was an independent risk factor associated with worse survival. A fistula-first strategy might be the best option for incident HD patients who are good candidates for AVF creation.
Fistula-first is the general recommendation for all hemodialysis (HD) patients [1–3]. A recent meta-analysis reported that nearly two-thirds of arteriovenous fistulas (AVFs) require the use of a bridging tunneled dialysis catheter (central venous catheter [CVC]) while awaiting maturation, placing patients at increased risk of infection and that approximately 20% of AVFs are abandoned without use . This increased failure rate is associated with increased vascular access (VA)-related complications and procedures . Therefore, AVFs created before HD initiation seem to have improved patency and decreased abandonment compared with those created after HD initiation [4, 6, 7], and the National Kidney Foundation (NKF) recommends that AVFs be created at least 6 months before initiation of HD treatment to allow sufficient time for access creation and evaluation, vein maturation, and, if necessary, maturation-enhancing interventions before cannulation . However, many of the published meta-analyses are insufficiently detailed to perform a subgroup analyses , and, for example, in the elderly HD population, there remains controversy as to whether the fistula-first strategy should be applied [5, 9–12], given operative risks, longer maturation times, and emerging data indicating the lack of a survival benefit compared with CVC or arteriovenous graft (AVG) use in these patients [5, 9]. Moreover, timely creation of an AVF before HD initiation is not always feasible because of the unpredictability of renal failure progression and individual variation in maturation times; premature AVF creation is associated with increased risk of VA-related complications, whereas late AVF creation cannot prevent the need for the use of a CVC.
The study cohort consisted of 524 incident HD patients with identified first predialysis VA creation of an AVF (pre-HD group, n = 191, 36.5%) or an AVF from a CVC (on-HD group, n = 333, 63.5%). There was no mortality or morbidity associated with AVF creation, and there were no CVC-related complications at the time of CVC placement. The baseline characteristics of the study population in relation to the initially created VA are presented in Table 1. There were no significant differences between the pre-HD and on-HD groups in demographics, risk factors, causes of chronic kidney disease, and type of AVF, except that patients in the pre-HD group had a higher prevalence of polycystic kidney disease than those in the on-HD group (P = 0.013). The mean follow-up duration was 45.1 months in the pre-HD group and 43.4 months in the on-HD group, with no significant difference in follow-up duration between the two groups (P = 0.461). During the study period, 33 patients died (17.3%) in the pre-HD group, and 89 died (26.7%) in the on-HD group.