Incremental Dialysis: Toward a Nuanced Approach
Much has been written, though agreement remains elusive, about when to start dialysis ( 1, 2 ). The Ideal Trial from ANZ—now a decade old—has provided the most certain of all uncertain guides 3, but no trials have been done to guide the most vexing question of all: once dialysis is deemed necessary, how should it then start?
Recently, the concept of incremental dialysis has become quite a hot topic, ( 4, 5, 6 ) despite that there are, as yet, no large trial data to prove or disprove its value. And, if a dialysis start is to be incremental, what does “incremental” mean, and what should be the increment(s)?
Incremental dialysis is built around the premise that a step-wise introduction of dialysis may help preserve residual renal function. In turn, preservation of residual renal function in haemodialysis patients is likely to be highly beneficial—at is has been proven to be in peritoneal dialysis patients—with even low levels of native renal function benefitting the removal of some of the more complex toxins: p-cresol, middle molecules, the indoles and guanadines, etc.
But back to the question “what should the increment be?” This author has long espoused time (t) and frequency (f) as the two most powerful levers in the dialysis prescription. But, the natural nemesis of t and f has always been the powerful resistive psychology of chair time…especially in centre-based dialysis programs where the patient catch-cry—“I don’t want to spend a minute longer on that machine”—is commonly encountered.
Incremental dialysis, as it has currently been used in the literature, is a composite of a mean lower f (1-2 sessions/week), a shorter ‘t’ (reduced hours/session), or some combination of the two ( 4, 5 ).
I fully agree that it clearly makes no sense at all to enforce immediate “full dialysis”—whatever that means—right from the get go. It makes far more sense to blend dialysis in gently, as residual renal function dwindles through the slowly progressive latter phases of CKD5. But, at what point along the CKD5 path should this ‘blend’ begin, and what factors should create this seamless blend?
While dialysis t and f indisputedly exert the most powerful influence on the efficiency of dialysis, the psychology of chair time is profound.
With human nature as it is, once starting with a low f and/or a short t any subsequent attempt to incrementally increase hours/session and/or sessional frequency will be inevitably met by all but the most malleable and dialysis-savvy patients with dispute, argument, angst, and a flat “no.” Reasoned explanation is unlikely to win the day. Patients will claim to be—or be—passably well. After all, they are likely to still have some residual native renal function—and refuse or resist any up-ramping of their schedule.
Suggested regimens for incremental programs in the literature have centered around 1-2 dialysis sessions weekly, those sessions also being shorter, perhaps 1.5 – 2.5 hours duration, until a further decline in urine output and excretory function forces an upscaling of frequency or time (or both), towards an ultimate “full” maintenance dialysis regime—whatever that means.
So, what science might the professional use to underpin the reason for serial t and/or f upgrades? Unfortunately, this is very boggy ground indeed, for our capacity to measure low-level residual function is neither simple, nor precise.
Timed blood and urine samples are the current best option, but should this be for:
A mathematically-averaged mean of both urea and creatinine clearance
Cystatin C excretion
Any, or all? And, what of the role of urine output—large, small or none at all—in the all-important aspect of volume management?
The literature still fails to agree on the best measure of RRF itself, let alone the ideal time interval between RRF testing:
Once a month
Once each 3 months
Each 6 months?
The lack of data in this space would make it very difficult to justify to reluctant patients why an expansion of chair time or treatment frequency was necessary at this specific point in time. Based, they might ask, on exactly what black and white science and proof?
There is a clearly commonsense argument for a rolling dialysis schedule; i.e., dialysis done every second day without an artificial “weekend” long break). But, the 3 sessions/week concept has long been firmly embedded in dialysis lore. People approaching dialysis through the final phases of CKD “know” this to be so, and funding bodies are not likely to embrace an extra treatment for all, every 2 weeks, with gratitude. So, for this blog, and for now, I shall (reluctantly) leave alternate day treatments to one side.
Similarly, expected chair time clearly differs across jurisdictions [US expectations are for 3.0-4.0 hours/session while the Australian and New Zealand expectation is for 4.5-5 hours/session]. Thus, chair time is a jurisdictional “expectation.” For all the arguments supporting longer dialysis treatments, US agencies seem unlikely to dramatically up-regulate sessional duration any time soon, despite Mel Hodges admirable idea to pay by the hour of dialysis, not by the session 7. But, for now, I will leave this to one side, too.
I am aware that my comments in this blog apply as much, or perhaps more so, to the majority who access centre-based care rather than to the luckier few who use home dialysis, although residual renal function is crucial to both. But, for now, I will simply consider dialysis as it is, and not as it should be!
That said, and depending on the jurisdiction, it is a current fact that most centre-based patients will ultimately end up receiving 3 sessions per week, with the hours per session dictated by local practice (between 2.5 and 5.0 hours). Let’s accept and not argue this further, for now.
Let us return to the question of how best to blend dialysis in slowly with conservative care as CKD5 progresses, much as a chef does not pour all the milk into a flour mixture at the start lest the blend fails to smoothly integrate, and goes lumpy. Indeed, there are several other ways to blend.
While incremental dialysis—I personally prefer the term “accretive” dialysis—can clearly be achieved by slowly increasing ‘t’ and/or ‘f’, as I have pointed out, patient resistance and argument against increasing sessional frequency and chair time is likely to be loud and strident. In addition, the complexities that may be encountered by dialysis services in the scheduling of multi-frequency and duration treatments are also significant. But, luckily, many roads lead to Rome.
Dialysis efficiency/efficacy is also influenced by:
The blood flow rate (Qb),
The dialysate flow rate (Qd)
The trans-membrane pressure gradient (TMP)
The ultrafiltration rate (UFR), and by
Dialyser membrane surface area (SA)
It should be possible to effect accretive (incremental) dialysis as residual renal function wanes by an imperceptive background increase in some (or all) of these factors, while delivering – from the start – the jurisdictional long-term intended t (whether it be 2.5 or 5 hours/session) and f (3 weekly sessions), unaltered.
Again, while I fully accept and support the notion of incremental or accretive treatment as native residual function wanes, the psychology of potential patient resistance is likely to be strong. Matched against a low scientific measure for RRF and thus a concrete argument for longer or more frequent sessions, the dialysis unit floor would become a blood-bath of argument.
Staff: “We need to increase your treatments from two to three per week”
Patient: “Why? I feel just fine, and my life has now been adjusted around two sessions”
Staff: “Let’s not start an argument about this—you have to.”
Patient: “No! I don’t feel any different…”
Staff: “We need to increase your hours from 2 to 3 per session as your RRF is dropping and increasing your UFR is unsafe.”
Patient: “I am NOT going to spend one more minute in that chair. Just try and make me!”
There are always alternative paths to winning a war! If dialysis is introduced when and how it is currently jurisdictionally practiced, with t and f prescribed at their long-term goal from the get-go, it is still possible to be smartly “incremental,” but in subtler, less confronting ways.
It should be simple to use the softer background levers of Qb, Qd, TMP, UFR, and SA to achieve the same blend-in result. This approach may, in the end, prove less problematic. If the early dialysis week/months start with:
A Qb of (say) 200-220
A Qd of 500
Iso-volemic or “mini”-volemic dialysis to ensure a subliminal UFR
Dialing a gentler TMP
Attaching a smaller surface area membrane
All of this can be up-adjusted as and when needed—but done so in the background, this approach would ensure incrementality—with its purported benefits—yet without disputation:
The Qb can be slowly increased from 200—>250—>300—>325 mL/min.
The UF rate can be increased as or if the inter-dialytic urine volume (formally measured each 2-3 months) were to fall away.
A larger surface area and/or higher flux dialyser could be introduced.
The platform could be switched seamlessly in the background from haemodialysis to haemodiafiltration.
All this can be done within the standard practice of t and f, and without turning the dialysis floor into a battleground of argument over t and f schedules.
To my mind, this is a discussion that cries out to be had, and an area ripe for a series of well-designed trials by up-and-coming smart nephrologists. In-program nuanced or smart dialysis might be an alternative approach to research in the future.
Meanwhile, the equally crucial issues of (1) Adoption of a rolling schedule (i.e., abolition of the long break), and (2) Introduction of chair time as a Key Performance Indicator (KPI) to permit a gentler, slower UFR 8, are both left for another day.
1. Chen T, Lee VWS, Harris DC. When to initiate dialysis for end-stage kidney disease: evidence and challenges. Med J Aust 2018; 209 (6). doi: 10.5694/mja18.00297↩
2. Tattersall J et al [on behalf of the European Renal Best Practice Advisory Board]. When to start dialysis: updated guidance following publication of the Initiating Dialysis Early and Late (IDEAL) study. Nephrol Dial Transplant. 2011 (26(7) 2018-2016.↩
3. Cooper BA et al. [on behalf of the IDEAL Trial Study Group]. A Randomized, Controlled Trial of Early versus Late Initiation of Dialysis. N Engl J Med 2010; 363:609-619↩
4. Obi Y, Kalantar-Zadeh K. Incremental and Once- to Twice-Weekly Hemodialysis: From Experience to Evidence. KI Reports. 2017. 2(5): 781–784.↩
5. Kalantar-Zadeh K et al. Twice-Weekly and Incremental Hemodialysis Treatment for Initiation of Kidney Replacement Therapy. Am J Kidney Dis. 2014. 64(2): 181–186↩
6. Wong J, Vilar E, Davenport A, Farrington K. Incremental haemodialysis. Nephrol Dial Transplant. 2015. 30(10): 1639–1648↩
7. Hodge M. Letter to Congressman Jim McDermott (Kidney Caucus Co-Chairman), US Congress, Washington DC. ↩
[A KidneyViews blog post]. To be found at: https://homedialysis.org/news-and-research/blog/153-kidney-caucus-letter-repayment-basis
8. Agar JWM. From Blog to Dialysis Chair: Implementing Volume 101. ↩
[A KidneyViews blog post]. To be found at: https://www.homedialysis.org/news-and-research/blog/119-from-blog-to-dialysis-chair-implementing-volume-101