Dialysis Unplugged: When should you start a patient on dialysis?

Photograph by Goupil, 1893, Wellcome Library, London

Just yesterday a senior colleague of mine – a famous physician-scientist at the Brigham and Women's Hospital -- made the following statement that caught my ear: ‘What most people don’t understand is that approximately 80% of medical decision making is not based on evidence. Rather, it is based on experience and the interpretation of limited or no evidence. Guidelines don’t help much because most guidelines are out-dated when they appear and clinicians don’t follow them anyway.”

Still, in nephrology we tend to be almost slavish followers of guidelines and fads. Worse than even that, practices in dialysis – at least in the United States – have been affected by financial incentives and vested interests.

When to start dialysis is a very basic question in nephrology. It has been the subject of much discussion and some controversy. Initiating dialysis had been based mostly on clinical judgement, until a fad supported by limited data and well-meaning guidelines pushed for earlier initiation. Whether vested interests also stimulated the "early start" or "healthy start" movement is conjecture, however.

An earlier post reviews the European Renal Best Practice Advisory Board's (ERBP) dialysis initiation guidelines published in NDT in July 2011, and is available at this link. Steve Rosansky in a recent article takes exception with these newly formulated ERBP guidelines, which he argues are based almost solely on the results of the "flawed" IDEAL study. "One size doesn't fit all". He writes: “the basing of the updated ERBP guidelines for dialysis initiation solely on the IDEAL study is short-sighted. The potential harm of early dialysis initiation must be considered. Unless there are definitive uremic symptoms, closely monitored end-stage renal failure patients should wait to initiate dialysis until they have levels of residual renal function of 5–9 ml/min/1.73 m2, or even lower if asymptomatic.”

Rosansky is echoing the warning by William Blake from over 200 years previously: “one law for the lion and the ox is oppressive”.

Thus, I suggest to you that the recommendation in Nephron Clinical Practice published online December 21, 2011, by Dr. Mustafa Arici, a nephrologist at Hacettepe University Faculty of Medicine, Ankara, Turkey, is muddying the waters further. Dr. Arici proposes “a scoring system which labels end-stage patients into green, yellow or red zones similar to traffic lights where patients may wait for dialysis, prepare for dialysis at their own or physician’s leisure, or as soon as possible, respectively.”  

He argues that a traffic light system is necessary because in a stable and symptom-free patient when someone (either the patient or a student) is asking the exact time dialysis will be started, there is no clear-cut answer.” “A scoring system” Dr. Arici argues “should include a validated GFR level and a set of clinical criteria including age, gender, ethnicity, socioeconomic state, health literacy, underlying disease, referral time, major comorbidities, nutritional state, uremic symptoms, basic laboratory parameters (such as BUN, albumin, hemoglobin, potassium, phosphate, bicarbonate) and, if available, rate of renal function loss and attainment of predialysis education programs.”

On the contrary Dr. Arici, we don’t need another flawed substitute for clinical judgement. Dr. William Osler, a master clinician is quoted as saying: “The practice of medicine is an art” …“The whole art of medicine is in observation, as the old motto goes, [but] to educate the eye to see, the ear to hear, and the finger to feel.”

Making a decision to start dialysis should be based on regularly seeing the patient in your clinic or office, taking a good history, examining the patient for key manifestations of uremia, and analyzing the laboratory data. I often-times tell my patients: "you will be the first to tell me that it is time to start dialysis." When the patient starts to become symptomatic dialysis initiation needs to be considered. Sometimes this would be at an estimated GFR of 6 and at other times it might be at an estimated GFR of 10 mls/min/1.73m2 or even a bit higher. And, as Dr. Rosansky writes: "The decision to start dialysis must be individualized, with input from multidisciplinary predialysis clinics that educate patients regarding the advantages and disadvantages of dialytic therapy for their specific set of physical, social, and psychological circumstances."