The Mathematics of Dialysis vs. Two Normal Kidneys
Dialysis is a study of contradictions:
On the one hand, how extraordinary; yet on the other, how meager.
On the one hand, how bright; on the other, how much a shadow.
On one hand, the normally working human kidneys; on the other, our attempts to replace their function.
So, how close does dialysis get to replacing normal kidney function? Let's work through some simple mathematics!
The human heart beats, on average, 70 times a minute. Each beat (or contraction) of the heart ejects, on average, 70 mL of blood. Thus, every minute, 4,900 mL of blood is circulated by the cardiac pump, which makes the cardiac output ~5 litres/minute.
Physiological studies show that the two human kidneys receive (as renal blood flow) between 20% and 25% of cardiac output. Reports vary a little, but are always within that range. Even if we use the lower estimate, i.e. 20%, the two kidneys receive (and filter) a total of ~1 litre of blood/minute.
Therefore, 1 litre x 60 minutes each hour x 24 hours each day x 7 days each week means the kidneys filter just over 10,000 litres of blood/week. Not a bad workload—and don't even try to think how many litres that equals in a normal human lifetime!
How, then, does haemodialysis, the most common and most efficient non-transplant option in renal replacement therapy (RRT), stack up?
To begin, some basic dialysis data are needed before we can understand how dialysis compares to normal kidney function. For these data, I will contrast two regions that, in most respects, are very similar—Australia and New Zealand (ANZ) and the United States (US)—yet where dialysis practice differs quite significantly.
In ANZ, most centre-based dialysis services provide 3 treatments a week, the duration of each being about 5 hours.
In the US, though the same number of treatments are given/week , the duration of each treatment is significantly shorter; commonly ~3.5 hours.
In ANZ, the pump that drives blood through the artificial kidney commonly runs at a speed of between 300-350 ml/minute (say 325 ml/minute).
In the US, to compensate for the shorter treatment duration, blood pumps are often run at speeds of 400 ml/min, or significantly more.
I will not debate here the pro's and con's of pump speed practice. I have addressed this in a previous blog. Suffice it to say that in my view, a high pump speed does not make for a happy vascular access!
Now, let's do the comparative mathematics.
Take ANZ first:
Based on these data, the blood presented weekly to an average dialysis system in centre-based care in ANZ = 325 ml/minute (i.e. 0.325 L) x 60 minutes x 5 hours x 3 treatments/week = 292.5 litres.
Scroll back for a moment and compare: two normal kidneys receive 10,000 litres of blood/week but a dialyser (the artificial kidney) in ANZ gets a measly 300 litres to play with, if it's lucky.
Yet, while that is clearly a stark difference, the comparison in the US is even starker!
The average calculation for a centre-based patient in the US is: 400 ml/minute (i.e. 0.4 L) x 60 minutes x 3.5 hours x 3 treatments/week = 252 litres; a low volume by ANZ standards, even after (my view) the higher access-damaging pump speeds used in the US have been factored in.
To compare these differences in another way, in ANZ only 292/10,000 = 2.93% of the blood presented in a week to two normal kidneys is presented for dialysis. In the US, it is even less: a paltry 2.52% (= 252/10,000).
What about home-based dialysis?
As a comparison, many home patients, especially in ANZ, perform up to 8-9 hours (mean = 8.5 hours) of overnight, while-asleep dialysis for an average of 5 nights/week. Most who undertake these schedules also run access-friendlier pump speeds of 225-250 ml/minute (mean = 0.2375 L). This provides 0.2375 x 60 x 8.5 x 5 = 605 litres/week for dialysis.
In the US, the most commonly used home profile seems to be a 400 ml/minute x 2 hour/treatment x 6 treatments/week program; this is the one that is most often delivered by the NxStage system, a low-flow system that, by its design, offers less efficient solute clearance compared to the ANZ-preferred single-pass systems. Thus, US home patients under this model present 0.4 L x 60 minutes x 2 hours x 6 treatments/week = 288 litres/week to a less efficient system. While this “short daily” home model clearly presents more blood for dialysis per week than does US centre-based practice, it is still less than the average ANZ centre-based volumes and compares even less favourably with the 600+ litres/week that are presented to the more efficient single-pass systems used by many/most ANZ home dialysis patients.
All this goes to show that:
Current dialysis systems are—even at best and wherever they are used—a poor substitute for the “real deal.”
To talk about centre-based dialysis practices as if they are comparable, nation-to-nation, is clearly incorrect.
Similarly, home dialysis practices, country-to-country, are also quite dissimilar, and should not be expected to necessarily have similar efficacy or outcome.
All patients should understand the effect that blood flow rate (pump speed) in litres/minute x minutes/treatment x the number of treatments/week may have on them. By doing their own calculation, patients can begin to explore the macro-influence of time and frequency on their program. Perhaps, then, they may also better see how little their dialysis is affected by pump speed.
This may help explain why my narrative and mantra has always been longer and more frequent, but slower, too.
Now, let's take this a step further, for it is gob-smacking to realise how meager is any/all dialysis, no matter how it might be improved (or diminished) by adjustments to duration, frequency or speed, when it is compared with the amazing power provided by two normal ordinary little kidneys.
Normal kidneys do far, far more than simply filter the vastly greater volume of blood presented to them than that which can presented to a dialysis system - even with the very best combinations of time and frequency. Normal kidneys go much further. They then adjust the filtrate of blood by additional excretion and/or reabsorption within the tubular structures of the kidney... and no dialysis systems beyond those still in the experimental stage yet offer any mimic of tubular function.
And, like the ads on late night TV…”there's more.”
Normal kidneys do other nifty stuff, like making erythropoietin, and converting inactive into active vitamin D; stuff that is far beyond the capability of any current, or any (at least medium-term) future dialysis system; stuff that, in our rudimentary efforts to mimic kidney function, is currently only possible by “giving back” or supplementing in the form of medicines.
Normal kidneys also:
Fine-tune sodium balance
Manufacture and release renin
Manufacture vasodilatory prostaglandins
Internally regulate and balance the vasoconstrictive effects of angiotensin
Use all these mechanisms (and more) to adjust blood volume and blood pressure…
So, when strolling the lakeside paths of Chicago after the 2016 ASN, I was put in mind of two Shakespearean phrases, both slightly “doctored” in their use here:
"What a piece of work is a man, how noble in reason, how infinite in faculty, in form and movement”...
...a phrase that makes us pause to reflect that, in our endeavours to mimic the human kidney through dialysis, our model, even in its most currently efficient form, is...
"...but a walking shadow; a poor player, that struts and frets its hour upon the stage, and then is heard no more...full of sound and fury, but signifying [little]"…
It behooves us to remember, when next we sit and talk with our patients on dialysis, that the treatment we are providing to “replace” their kidneys is really no replacement at all. “Renal replacement” is a misnomer, when it comes to dialysis techniques.
Certainly we should maximise what and where we can—those options being primarily duration and frequency—but we should also pause and take notice that, while we have taken some small first steps in these past 60+ years, there is oh, such a long way to go...