Haemodialysis UF Volume and UF Rate are NOT the Same
I have detected some confusion between ultrafiltration (UF) volume and UF rate. These are NOT synonymous—they mean entirely different things.
While the UF volume has been with us for as long as there has been dialysis, recent interest in the UF rate appears to have introduced some uncertainties about exactly what the difference is.
I have been told (correctly or incorrectly) that Fresenius North America has recently introduced a recommendation that the UF rate should be kept below 13 ml/kg/hour. Unfortunately, this half-step in the right direction may have simply added to this confusion.
Dialysis removes two main things:
- Solutes – dissolved electrolytes (salts) and molecules of waste (toxins) made by the body’s daily functioning
- Fluid (water). Convective positive or negative pressures that are applied to either side of the dialysis membrane largely determine the amount of water that is removed. By adjusting/altering these pressures, more (or less) water can be removed.
The purpose of this blog is to set the ledger straight! So, here goes…
The UF volume is the amount of water that must be removed in a single treatment to return a patient to his or her target (or base) weight. The ultrafiltration volume is commonly expressed in terms of weight (where 1 litre = 1 kilogram).
The UF rate is the speed at which that volume is removed. (Here is a handy, free calculator we built to help you calculate the UF rate).
But, wait a moment: how do we know how much water to “ultrafilter”?
As all who are involved in dialysis know, most patients rapidly gain weight (read water; remember 1 kilogram = 1 litre) between the end of one dialysis and the start of the next. In those lucky few who retain a urine output as a result of residual renal function, this gain can be minimal. However, in time most patients will lose all native kidney function and pass little or no urine. As a result, all fluid consumed between treatments —the inter-dialytic period—will need to be removed at the next treatment to return the patient to their base (NB: I personally prefer the term ‘target’) weight.
While we would like this target end-dialysis weight to be what is often described as “dry weight,” in reality, it rarely is. True dry weight is what a body would weigh IF the volume of water in each of the body’s three fluid primary compartments (cellular, extracellular, and intravascular) were to be ideal. Unfortunately, the exact measurement of dry weight remains a holy grail. Although there are all sorts of methods to try and determine it, dry weight is a notional goal rather than an actual, definable value.
So, we do our best—and, more often than not, our best is a very bad best —to guess at what we think dry weight should be… and as it really isn’t dry weight, we call it “base” or “target weight,” as our best approximation of true dry weight. Maybe, one day, bioimpedance, biochemistry or other biological solutions to the “dry weight” conundrum will be found, but for now, the aimed-for post dialysis weight (the “target weight”) remains just a best guess.
So, why all that talk about weight?
Well, as every dialysis patient knows, they will be weighed at the start of dialysis. The difference between this weight and the post dialysis target weight will then be calculated. Most commonly, the predialysis weight is greater than the target weight. This difference (in kilograms) equals the volume (in litres) that must be removed during the dialysis run by the process called ultrafiltration.
The amount to be removed [gain 2 kg = remove 2 litres; gain 3 kg = remove 3 litres; ‘gain’ 4 kg = remove 4 litres … etc.] is the UF volume. Although small adjustments may be made to this volume to account for saline flushed back at the end of dialysis, or fluid consumed during dialysis (e.g. a cup of tea), in general the UF volume equates the litres (or kg) gained in the interdialytic period.
But…and here’s the crux of this blog…the UF volume is NOT the same as the UF rate! The UF rate is a speed, not a volume, and refers to the volume of water that must be removed in any given time!
This means that:
- If there are 2 litres of water to remove (UF volume) and the dialysis run is 2 hours, the speed of removal—UF rate—will be 1 litre per hour.
- If there are 4 litres of water to remove (UF volume) and the dialysis run is 2 hours, the speed of removal (UF rate) will be 2 litres per hour.
The UF rate is governed by two factors:
- The volume that must be removed (the UF volume).
- The time (or sessional duration) allowed for that removal.
So, the UF volume = litres, but the UF rate = litres per hour.
To add one piece of complexity to this simple distinction, strong data has shown that if water is removed too fast and the circulating blood volume is contacted too quickly, organ perfusion pressures drop. In turn, this risks organ ischaemia and compromises organ oxygenation.
This has led to the concept, advanced by Jennifer Flythe et al1, that there is a maximum rate at which water can be removed. If this rate is exceeded, organ “stun” and cardiovascular morbidity and mortality are at heightened risk.
The UF rate is dependent on a third factor: it is not just the volume that has to be removed and the time allowed for its removal, but the size of the person being water-depleted… i.e. the persons’ body weight…or, from data that Emily See from our service has generated and reported at various meetings in the last year or so, body surface area.
Thus…the UF rate is better expressed in mL/Kg/hour.
Note that the “Kg” in this equation is the patient’s target post dialysis weight. Ideally it should be the patient’s “dry weight,” but dry weight is a notional number, while the target weight is real.
I have argued—especially with my American colleagues—over what a safe UF rate might be. I argue that a maximum UF rate should be no greater than 10 ml/kg/hr. I note that Fresenius (USA) has recently advised a maximum rate of 13 ml/k/hr.
I disagree! Just take a look at Flythe’s graph—data drawn from the HEMO study data.1 What point on that graph says “safe”? I rest my case. I know, if I were a patient, the ultrafiltration rate that I would want!
As one single example of how this works, imagine a patient whose target weight is 100 kg and who has gained 5 kilograms. S/he must lose 5 litres to return to target weight:
- A 3-hour dialysis would mean removing 5 litres (= 5000 mL) in 3 hours = 1,666 ml/hour = 1666 ÷ 100 kg or 16.6 mL/Kg/hr. That would do irreparable damage to the heart!
- If the same patient had 4 hours of dialysis: 5000 mL to remove ÷ 4 hrs ÷ 100 kg target weight ---> 12.5 mL/Kg/hr.
- Do a 5-hour dialysis and the ultrafiltration rate drops to 5000 ÷ 5 ÷100 = 10 mL/Kg/hr (and only just “safe”).
- Better would be 6 hours with an ultrafiltration rate of 8.3 mL/Kg/hr.
For comparison, our Geelong data calculates our mean unit UF rate across 150 centre-based patients to be 7.95 ± 3.11 mL/Kg/hr, while our median unit-wide UF rate is 7.73 ml/kg/hr. NB: these data exclude our 50 patients at home on extended hour and high frequency nocturnal dialysis. What, for interest, is your unit’s mean and median UF rate?
So, finally, we come to “what to do”
There are only two ways to alter (i.e. lower) the UF rate!
- Have less water to remove in any given time
- Take more time to remove the same amount of water
We all now how hard it is for dialysis patients to restrict their water intake. We encourage them, cajole them, some even bully them, but, at the end of the day, limiting water intake is just not possible for some (or most) patients. Before you scold them, try it yourself! Try to limit, day after day after day, to the sorts of limits you seek to impose on your patients.
So, if water intake limitation has its limits, only one course remains: Dialysis sessional duration must be longer. And, that durational extension must be sufficient to ensure that the UF rate is no greater than (in my view) 10 ml/kg/hour.
The issue is how to achieve an ultrafiltration rate ≤ 10 ml/kg/hr?
In Geelong, for almost all patients, we do. How? Well, we do longer dialysis.
End of story.
Flythe JE et al. Kidney Int. 2011 Jan; 79(2):250-7↩