When To Refer For Dialysis Access?

Despite the knowledge that AV fistulas and grafts give superior outcomes when compared to dialysis catheters, there remain a large chunk of patients who begin dialysis without a fistula or graft. Conversely, there are also examples of "unnecessary procedures"--individuals with advanced CKD who get an AV fistula placed, but never use it either because their CKD never progresses to ESRD or because they die before reaching dialysis. These facts underline the difficulty nephrologists have with accurately predicting the progression of CKD to ESRD.

How should one decide precisely WHEN to refer for dialysis access? KDOQI does have guidelines--they state that individuals with CKD should have a fistula placed once the GFR is less than 25, the creatinine is greater than 4 mg/dL, or there is the anticipated need of dialysis within 1 year--but these are largely based on expert opinion.

One paper which seeks to address this issue in a more quantiative manner is a 2007 KI paper by O'Hare et al. Briefly, investigators looked at a large retrospective cohort of patients with an eGFR less than 25 calculated the ratio of unnecessary to necessary surgeries. An unnecessary surgery, for instance, was one in which a fistula or graft was placed but dialysis was not required within the 1-year period--either because their CKD remained stable or because they died before reaching ESRD. In all the hypothetical scenarios they examined, older patients would have been more likely than younger patients to receive unnecessary procedures. For instance, if all patients were given a fistula at the time of cohort entry (eGFR less than 25), the ratio of unnecessary to necessary procedures would have been 5:1 for patients aged 85-100 years, but only 0.5:1 for those aged 18-44 years.

The bottom line: the decision as to timing of dialysis access placement should probably take into account age: young patients with a low GFR are more likely to eventually reach ESRD because of their longer predicted lifespan, and thus a case could be made to be more aggressive with access patient in this group. Obviously, every patient is unique and the nephrologist must take into account multiple issues (e.g., rapidity of CKD progression, other comorbidities, etc) in making this important decision.