Despite having the knowledge that fistulae are better than catheters, many parents and/or patients still select catheters over fistulae in the pediatric population. This may be as a result of a multiple issues, preconceived notions, or unintentional provider bias [5]. The selection of access for initiation of treatment depends largely on what information is provided by caregivers but also relies on preconceived notions about ESRD, parent and caregiver biases, and the age and maturity of the child. When evaluated by vascular surgeons, many of these children already have CVCs in place and have initiated hemodialysis. Regardless, there should be a complete discussion of the benefits of AVF over CVCs, and ample time for making an informed decision should be given.

While anecdotally there are many factors that limit placing AVF in children, few studies have been done to study these barriers. A recent publication by Chand and colleagues describes some of these barriers and identifies communication issues between providers as a major issue, in addition to lack of standardized referral practices for CKD patients, lack of standardization as to whom the patients should be referred to, and finally lack of early communication between surgeons and interventional radiologists and dialysis staff (nephrologists, nurse practitioners, dialysis nurses) regarding problematic fistulae [5]. Similarly, few studies have assessed the psychosocial aspects of decision-making in choosing a form of dialysis access in the pediatric population. As one might imagine, for younger patients the decision is up to the parents and caregivers, which may place a huge burden on them. Once a child is able to express desires and dislikes, even without necessarily completely understanding all the options, the decision is often left up to them or at least made with their preference in mind. In our own practice, some of the reasons given a patient might not want a fistula include fear of needles, inability to wear jewelry at the site of an AVF, inability to participate in sports, ugly appearance of fistulae, and desire for the fistula to remain unseen. Parents also voice concerns with fistula placement which include the uncertainty of knowing what is best for their child despite receiving all the data supporting fistulae over catheters and the hope that a more “permanent” solution is approaching and that the “bridge to transplant” could be accomplished with a central venous catheter. In another survey by Brittinger and colleagues, assessing the pediatric patients’ discomfort with cannulation, 39 % of patients reported no discomfort, 39 % had tolerable discomfort, and 22 % reported great discomfort. Interestingly, 95 % of the participants reported they would prefer not to revert to central venous catheter for access [6].

Larger studies are needed to identify barriers to fistula placement, and even further projects are needed to address these barriers and offer solutions. Despite having the knowledge that autogenous access is better than CVCs, many patients and families still select CVCs over AVFs for access. As providers we use CVCs first in 90 % of the pediatric population, and thus both patients and their caregivers may be resistant to change or to undergo another (more invasive) procedure once a central venous catheter is in place. Efforts to educate families at an earlier stage, before initiation of any treatment, should be pursued. A campaign to place AVF in patients months before they start dialysis is ideal; thus, nephrologists and surgeons should communicate frequently and early about these patients such that AVF surgery can be completed in a timely fashion.

Several studies have investigated the reasons for CVC preference among pediatric patients undergoing hemodialysis. Fadrowski and colleagues retrospectively analyzed a cohort of 1,284 patients from 2001 to 2003 [7]. In this cohort, 755 (59 %) had a central venous catheter. The reasons given for choosing a central venous catheter included “small body size” in 142 (18.8 %), having “maturing” AVF/AVG in 53 (7 %) patients, and a “transplant scheduled” in 83 (10.9 %) of patients. Among these 755 patients, 32.2 % did receive a transplant within the year. In another retrospective cohort study published in 2006 by the same group looking at ESRD patients aged 12–18 years old receiving HD for the year 2000, the authors quantify the increased risks attributable to CVCs compared to patients with arteriovenous fistula [8]. The authors included 418 patients, 41 % of whom had an arteriovenous fistula or graft and 58 % had central venous catheter. Data analysis revealed an increased relative risk among central venous catheter patients with regard to all-cause hospitalization (RR 1.84 CI 1.38–2.44), hospitalizations due to infections (RR 4.74 CI 2.02–11.14), and complications of vascular access (RR 2.72 CI 2.00–3.69).

The durability of CVCs, while improved in recent years with smaller profile catheters, still remains inferior to AVF and averages between 4 and 10 months and in some cases is under 1 month. Several groups have published data on the longevity of CVCs and investigated the reasons they fail [9–11]. Central venous catheter durability ranges from 0 to 62 % at 1 year, and failure is attributed to infection (17 %), thrombosis (33 %), extrusion (5.4 %), and kinking (which is more common in smaller catheters). In general, cuffed catheters carry a lower risk of infection and have a longer durability (months) when compared to non-cuffed catheters.

One of the major long-term complications of central venous catheter placement is central venous stenosis. In an attempt to minimize this complication, NKF-KDOQI has delineated management in the event that a central venous catheter is placed in children [12]. The recommendation lists, in order of preference, the right internal jugular vein, right external jugular vein, left internal and external jugular veins, subclavian veins, femoral veins, and finally translumbar and transhepatic access to the IVC.

The prevalence of central venous stenosis associated with a history of subclavian central venous catheter placement is 25–50 % [3, 13–15]. In a recent case report, the author brings to light the fact that central vein stenosis might be grossly underdiagnosed, and as surgeons we likely are only seeing the cases that are significantly stenotic enough to cause symptoms [15]. In this report, however, it is not only hemodialysis-related access which was identified as a risk factor for developing central venous stenosis but rather the use of both tunneled and non-tunneled dialysis catheters, peripherally inserted central catheters (PICC), as well as other CVCs and ports. The length and duration of the catheter and multiple catheters are two factors most closely associated with developing central venous stenosis. A 2012 retrospective review evaluated failure rates of arteriovenous fistulae in adult patients with a history of ipsilateral vs. contralateral catheters [14]. Their results indicate that while maturation times and primary failure rates were similar in both groups, there was a lower cumulative fistula survival at 2 years in patients with ipsilateral catheters compared to contralateral catheters (54 % vs. 74 %). This result is echoed in other publications in the adult population [3, 15, 16]. This phenomenon however has not been demonstrated in the pediatric population. In a study by Wartman and colleagues, catheter history did not affect patency of arteriovenous fistulae after surgery [17]. Thus if a pediatric patient has had a central venous catheter, this does not become a contraindication for ipsilateral arteriovenous fistula creation, although if central venous stenosis is clinically suspected, it should be ruled out as this could confer long-term complications.

Central venous catheters may be either non-tunneled or tunneled. These can be placed percutaneously under moderate sedation and local anesthetic; however, tunneled catheters in the pediatric population often require general anesthesia to ensure patient compliance.

Major challenges in establishing central venous access for hemodialysis in children are that there are no evidence-based rules for selection of catheter size and that the pre-curved catheters commercially available for children are limited to larger sizes. Larger catheters offer higher volumes during dialysis, but the size of the child and his or her vessels limits the size of the catheter that can be used. Catheters that are smaller mean that the length of dialysis sessions has to be longer with slower flows. A useful formula that has proved safe in selecting catheter size is [Size (Fr) = Age +/− 2]. Taking the age of the child and converting it to the diameter, in French measurement, with adjustments after physical examination of the child to reduce the size if the patient is small for age or increase size if the child is larger than peers the same age. Pre-curved catheters available for the adult populations make percutaneous and subcutaneous tunneling possible; however, these catheters are largely not available in the smaller sizes for the youngest of pediatric ESRD patients. In these patients placing a tunneled central venous catheter usually means forcing the curve during placement and making one or more counter incisions over the access vessel for accurate and precise placement.