A Rational Approach to Dialysis Time and Frequency
Good dialysis depends on the combination of both:
- Optimal frequency: to permit small and large solute removal.
- Optimal time: to allow slow, perfusion-friendly rate of volume removal.
The current Kt/Vurea formula has unfortunately encouraged clinicians to increase one of the two numerator variables (K) in an effort to reduce the other (t). This has resulted in the human body being treated like a multiplication table, where increasing one variable and reducing the other yields an unaltered result. Both 2 x 4 and 4 x 2 = 8.
Short hours, low frequency dialysis
This is the current majority reference point for most dialysis patients, and is how current “conventional” dialysis can best be described. Even conventional dialysis—as the term is loosely used in the dialysis literature—is not a single entity:
- In the US, it is commonly 3.0 - 3.5 hours, 3 times a week.
- In Japan and ANZ, it is more commonly 4.5 - 5.0 hours, 3 times a week.
These are unlikely to be fully comparable, either in the efficiency of solute removal, or in the speed of ultrafiltration. Yet, they are frequently conflated and compared as if they were the same in open discussion or in literature reports. Conventional centre-base dialysis – and especially the US conventional model – offers neither sufficient time for full solute removal, nor sufficient frequency for volume control. While an inadequate sessional duration “seduces us” by providing an apparently adequate cleansing of small solutes like urea, the more insidious middle molecules that depend on time for removal are not sufficiently removed.
Further, the rapid ultrafiltration rates required to remove the accumulated interdialytic volume excess are known to promote intradialytic circulatory instability and hypotension, cramp, nausea, vomiting and—more threateningly to survival—“stunning” of vital organs, in particular the heart and brain. Yet, to our shame, this dialysis model is the option most current dialysis patients are offered, and receive.
Long Hours, Low Frequency Dialysis
This is exemplified by most current in-centre nocturnal dialysis regimens. Longer duration dialysis, especially those regimens (commonly nocturnal) that permit between 7 and 9 hours of sessional duration, encourage deep-compartment cleaning (Eloot). Intercompartmental and transcompartmental “deep cleaning” of time-dependent solutes takes exactly that: time, and time cannot be artificially accelerated. So, while from a solute perspective this model works, its failure lies with poor volume management, as the result of inadequate frequency.
A longer interdialytic period will inevitably bias towards a larger interdialytic fluid gain. The more hours without dialysis, the greater the time a patient has to eat, drink, and accumulate excess fluid volume, albeit patients who retain the benefit of a residual urine volume will be at greater advantage under this modality than those who are anuric. Thus, while in theory the longer sessional duration of this model should permit a slower ultrafiltration rate, this theoretical benefit may be effectively cancelled out by the larger accumulated interdialytic volume that requires removal.
High(er) Frequency, Shorter Hours Dialysis
This modality is typified by the (largely US-centric) “short daily hemodialysis” (SDHD) model. Current SDHD broadly favours an increase in treatment frequency to 5 or 6 treatments per week. But, as this model concurrently trades-off a reduction in sessional duration against an increase in frequency, it still does not provide the wisest or most appropriate answer.
- Re solute removal:
- Higher frequency dialysis benefits small solute clearance by more frequently providing a maximal concentration gradient for small solute clearance.
- But…most so-called “middle molecules,” (or PO4 with its hydrophilic skin) are characterised by linear diffusion over time and, rather than moving down a concentration gradient (like urea), exhibit time-dependent removal patterns.
- Thus, while high frequency combined with time-reduced dialysis may deliver better small solute clearance (as measured by Kt/Vurea), high frequency combined with truncated time does not significantly improve larger solute clearance (e.g. β2M, homocysteine, PO4).
- SDHD, like its name, thus falls (all solute) short!
- Re volume control:
- Higher frequency dialysis may reduce the interdialytic fluid accumulation period.
- But…if concurrently combined with a reduction in sessional duration, the shorter available intradialytic time for ultrafiltration will effectively enforce an unchanged (and still excessive) volume removal rate, despite any potential reduction in volume accumulation that might result from the shorter interdialytic period.
- Volume removal above all else needs time, and time is the one universal constant that cannot be abbreviated.
Long Hours and High(er) Frequency Dialysis
This is the model sometimes referred to as “intensive,” or “high intensity,” or “extended hour and frequency” dialysis. It is the model first developed in its modern form by Uldall and Pierratos (Canada) and later spread by Lockridge to the US, and by Agar to ANZ. This model delivers both optimal solute clearance and gentle volume correction:
- Deep-compartmental cleaning is possible, as the long sessional duration allows for time-dependent extraction as well as optimal concentration-gradient driven diffusion.
A lower ultrafiltration rate—and the abolition of the volume-linked symptoms of rapid dialysis—is achieved through both:
- A shorter interdialytic accumulation period, and thus a lesser volume to remove.
- A longer sessional duration in which to remove a lesser-accumulated volume.
Both time and frequency matter in both target domains
What is abundantly clear is that longer and more frequent dialysis makes sense, no matter from what angle dialysis is viewed. Perhaps the “key” variable for solute clearance is sessional duration, though, clearly, an extended interdialytic period will also bias to greater solute accumulation. Perhaps the “key” variable for volume management is sessional frequency, though, clearly, the greater the abbreviation of sessional duration, the higher the ultrafiltration rate will have to be for any given accumulated volume, even if this is lessened by increasing the frequency of sessions.
So, the final question is…what matters more: solutes or volume?
While both matter, I would contend that volume pips solute at the post. Volume is the cardiovascular bête noir. With the possible exception of potassium, morbidity and mortality in dialysis patients is driven by either (or both) an excessive load of volume accumulation, or an excessive speed in volume contraction.
Rationalising this, 'f' beats 't' by a whisker.
Perhaps this is why, when a decade and a half ago, the two giants of 20th century dialysis (Scribner and Oreopoulos) were pondering their proposal for a measure of good dialysis - the Hemodialysis Product (the HDP) - they decided to square 'f' as they arrived at HDP = t x f2 and not simply HDP = t x f.
While squaring 'f' may be overemphasising its importance, giving it equal value to 't' may also be undervaluing 'f'. Perhaps the truth lies somewhere in between? What is certain, in my view, is that time and frequency are the only two factors that really matter. The rest is noise.