The 'Volume-safe Dialysis Duration' KPI
The following blog explores some simple, utterly naïve and probably crazy thoughts I have had about the development of a 'volume' Key Performance Indicator (KPI) for the day-to-day clinical management of dialysis patients. Dialysis has had a 'solute' KPI for 2+ decades – Kt/V urea – and more of this in a moment, but no simple, easy-to-apply marker for volume status currently exists.
Like a dialysis Don Quixote, I have long tilted against the windmill of Kt/V urea. It has been, in my view, the single greatest 'furphy' ever visited upon dialysis ... NB: if you are unfamiliar with the term 'furphy', you can find it explained at: http://en.wikipedia.org/wiki/Furphy or at http://www.onlymelbourne.com.au/melbourne_details.php?id=10424#.U030f1WSxBk
Again, in my view, generations of nephrologists have been incorrectly sold the lie that urea is in some way a good measure of 'adequacy'. This is despite the fact that urea transport and diffusion kinetics bear little or no relationship to the kinetics of most metabolic wastes. While urea moves very rapidly down a concentration gradient, many of the more important 'toxins' move very slowly down time-dependent gradients.
This has led nephrologists—especially in the US—to be 'furphied' into believing that once a certain level of urea clearance has been reached, it is then OK to "turn off the machine and wheel the next victim in". Well, that is just plainly wrong. I have spent the last decade and a half fighting the almost un-fightable misconception that Kt/V urea is the be-all and end-all of good dialysis. Clearly, it is not, but the belief in the US-demanded dialysis KPI of a Kt/V of 1.3 as the measure of adequacy...indeed the only measure of adequacy has become so ingrained that it has excluded all other rational thought.
Enough of my solute hobbyhorse. Let us push solute clearance to the side for the moment. I have been thinking about a simple way to devise a useful volume KPI so, let us think about 'volume'.
By volume, I mean circulating blood volume and the effect that a dialysis treatment has upon it. This is not the same as dry weight. It is vital that what follows is separated from the concept of determining dry weight. That is a different argument. My suggested volume KPI is only about the next upcoming dialysis treatment, and in this particular patient.
The measurement of the impact of any one dialysis treatment seems often confused with the measurement of dry weight; the assessment of the latter including the evaluation of all manner of the useful and the not-so-useful. As regards dry weight, there are supporters of a variety of bio-impedance options, of IVC ultrasound, of isotope dilution, of BNP, etc., but none have (yet) reached acceptance as an easy bedside application, nor do any show promise of providing an immediate, pre-dialysis, practical and cost neutral application for day-to-day use in each and every patient in each and every dialysis unit. Are dry weight assessment techniques important? Yes. But, are they useful for a daily treatment KPI? Probably not!
A useful 'volume' KPI needs to be utterly simple, easily applied, applicable pre-dialysis and have clear and interpretable meaning for both staff and patient.
I have followed Chris McIntyre's work on organ stun closely – and it is very nice work indeed! The simple concept of 'perfusion stunning' has provided excellent evidence that there is a rate of intravascular volume depletion that is risks tissue hypo-perfusion and target organ stun.
While his first work described myocardial stun as a result of reduced coronary perfusion during dialysis, his group has now shown similar data supporting cerebral stun through reduced carotid and/or vertebral perfusion, and episodic AKI and the loss of residual renal function through reduced residual renal tissue perfusion during dialysis. Indeed, the list of at-risk tissues is, potentially, as long as is organ number. This made me think about a simple, bedside, pre-dialysis measure that might provide a workable, useful volume KPI.
I have coned down onto two concepts:
(1) The plasma refill rate (PRR) is the most 'rubbery' of numbers, as so many factors impact its calculation. However, the few who have attempted to quantify a mean average PRR for a mean average patient—if there one—have zeroed in on a mean maximum PRR of between 5 and 6 ml/kg/hr(1,2). The concept of the plasma refill rate simply says if removal of fluid from the plasma volume (e.g. by dialysis) exceeds the capacity of the extracellular fluid (ECF) – to refill and replace the vascular volume [see earlier blog on 'the waterfall concept'] then the intravascular volume must contract by the difference between the removal and refill rates.
If, for arguments sake, the PRR is 'rounded out' to ~5ml/kg/hr ... and our clinical experience with extended hour vs. centre-based dialysis suggests that this figure is about right, then, in the non-existent 'average' person, intravascular volume contraction will likely occur if the rate of volume removal during dialysis exceeds 5ml/kg/hr.
Some may say that it is naïve, very naïve, to use one figure (like 5ml/kg/hr) for the PRR of all. It is clear that factors like weight, size, gender, capillary permeability, albumin level, inflammatory status, cardiovascular competency, and nutritional status (among others) will all impact on the PRR—hence my previous comment that this is a very 'rubbery' number. But, to be fair, these same factors also impact on bio-impedance, isotope dilution and all the other commonly touted, complex and costly ways to assess volume status and/or dry weight. So, for now, let us just hang onto that number.
(2) The 'stun' factor stems from the work done by Chris McIntyre and his group around organ perfusion and tissue oxygenation(3,4). This work adds beautiful symmetry to the volume 'story', as it suggests that stun mechanisms are increasingly likely to occur if fluid is removed from the intravascular volume at a rate >10ml/kg/hr, and are almost certain to occur at volume removal rates >13ml/kg/hr.
Taking a simplistic view...and I am a very simplistic guy...two magic numbers emerge!
(a) Volume contraction will occur in dialysis if fluid is removed at >5ml/kg/hr.
(b) Organ stunning is risked in dialysis if fluid is removed at >2 x the PRR = >10 ml/kg/hr.
These are two very easy-to-remember numbers, despite the fact that they are underpinned by two highly complex (and debatable) mechanisms.
Now, blend (1) and (2) into a volume KPI:
My simplistic solution to the issue of volume is to extend the dialysis treatment, at least and until the dialysis ultrafiltration rate is less than the lower risk level identified for organ stun: i.e., <10ml/kg/hr.
Though clearly a patient should continue to contribute to volume control, it is neither wise not productive to provoke conflict through imposing unrealistic and commonly unachievable at-home, inter-dialytic intake fluid restrictions, especially when the dialysis itself is often primarily to blame for post-dialysis thirst.
Post dialysis thirst—and the resultant excessive fluid gain that follows before the next dialysis—is commonly the direct result of imposing intra-dialytic UF rates that so rapidly contract the vascular volume during dialysis that post-dialysis thirst is irresistible and excessive inter-dialytic weight gain is unavoidable. This is not the patient at fault. It is we who are at fault.
It is we who have prescribed a brutally fast contraction of vascular volume by scripting an unrealistic time frame for volume removal. Dialysis treatment time must be extended such that the target weight (read fluid) loss is achieved at a removal rate no greater than 10ml/kg/hr. Chris McIntyre's risk stratification marker for myocardial (and other tissue) stun.
The necessary duration for any particular dialysis treatment can be easily predicted before the start of that treatment. (1) the pre-dialysis weight is known. (2) the target post dialysis weight is known, noting that this target weight is not necessarily the same as the ideal dry weight. By in-building a maximum intra-dialytic UFR of no greater than 10 ml/kg/hr, it is simple to calculate, pre-dialysis, (3) how long any one particular treatment time will need to be.
If dialysis sessional duration is used as the single variable KPI to judge safe volume removal, sessional lengthening (or shortening) would then be dictated by a pre-dialysis calculation of the time required for that treatment only to achieve the target post-dialysis weight.
This would avert most of the volume and perfusion-related symptoms of dialysis, i.e., the majority of patient-reported symptoms. Hypotensive episodes, nausea and vomiting from splanchnic hypo-perfusion, cramp, and dialysis fatigue would all abate. Prolonged post-dialysis recovery time would disappear. Patients would rapidly learn and understand the value of time in their dialysis program. Patients would rapidly learn that dialysis duration directly relates directly to weight gain—a relationship that would help reinforce inter-dialytic volume management. Above all, the risks associated with key organ hypo-perfusion would be minimized.
The pre-dialysis weight is measured and known. The target post-dialysis weight is set and known, pre-dialysis. The volume that must be ultra-filtered is therefore also known, even if an accurate true dry weight is rarely known. It is then dead simple to calculate the required length of a dialysis run to ensure that the UFR for that treatment does not exceed 10ml/kg/hr.
In our unit, we are now collecting prospective data prospectively which will allow the calculation of: (1) the number of patients where dialysis duration would have been lengthened under these guidelines: (2) by what duration.
In addition, we will be able to examine (3) the staffing, rostering and global economic impact of the changes needed should we wish to ensure a 95% compliance of individual and collective dialysis times to a maximum per treatment UFR of <10ml/kg/hr. 'Smart' rostering may then allow the grouping of patients who regularly require longer sessions.
Finally, some patients may qualify for shorter treatment runs—judged on volume criteria alone. Here, solute criteria would play a clear, but secondary, duration role. To strengthen the interplay between solute and volume factors, there is a clear need for a better solute clearance KPI than Kt/V urea; one that considers time-dependant solute clearance, not just small molecular diffusive clearance. Convective therapies (especially HDF) may play an important role here...but that is another story.
A useful spin-off may be that patients—and dialysis professionals—come to better understand the duration inadequacy of most current dialysis programs, with a longer-term benefit to patient symptom relief and reduced stun-related morbidity and mortality.
A simple, simplistic, yet effective Volume-safe Dialysis Duration formula is likely to save many more lives, and make dialysis far more tolerable for far more dialysis patients, than Kt/V urea ever has or will!
Now, for the Volume-safe Dialysis Duration (VsDD) formula:
Volume-Safe Dialysis Duration (VsDD) =
- VOLUME TO REMOVE (mLs)
- 10 x PRE-DIALYSIS WEIGHT (Kg)
- Example # 1
- Weight gain = 2.65 Kg
- Volume to remove = 2650 mL
- Pre-dialysis Weight = 56.75 Kg
- VsDD = 2650/10 x 56.75 = 2650/567.5 = 4.66 hrs
- Example # 2
- Weight gain = 5.2 Kg
- Volume to remove = 5200 mL
- Pre-dialysis Weight = 117 Kg
- VsDD = 5200/10 x 117 = 5200/1170 = 4.44 hrs
- Example # 3
- Weight gain = 3.3 Kg
- Volume to remove = 3300 mL
- Pre-dialysis Weight = 54.25 Kg
- VsDD = 3300/10 x 54.25 = 3300/542.5 = 6.08 hrs
- Example # 4
- Weight gain = 2.25 Kg
- Volume to remove = 2250 mL
- Pre-dialysis Weight = 41.3 Kg
- VsDD = 2250/10 x 41.3 = 2250/413 = 5.45 hrs
- Example # 5
- Weight gain = 0.8 Kg
- Volume to remove = 800 mL
- Pre-dialysis Weight = 68.5 Kg
- VsDD = 800/10 x 68.5 = 800/685 = 1.16 hrs.
NB: in the last example, Kt/V (or, in time, a better measure of solute clearance) would then take over as the duration-defining KPI.
In all circumstances, dialysis duration would be determined by a combination of solute clearance (currently Kt/V or URR – though better solute clearance KPIs are badly needed) with dialysis continuing until both solute and volume KPI data points have been achieved.
If additional fluid needs removal beyond a simple return to the pre-dialysis weight, i.e., if a sequential 'drying down' is required over several dialysis runs, then the additional volume to remove (ml) at each sequential dialysis session is simply added to the numerator.
- Example # 6 [using the same parameters as for Example # 1]:
- Weight gain = 2.65 kg
- Volume to remove = 2650 ml
- Add extra 'dry-down' volume of 250 ml
- Total volume to remove = 2650 + 250 = 2900 ml
- Pre-dialysis Weight = 56.75 kg
- VsDD = 2650 + 250/10 x 56.75= 2900/567.5 = 5.10 hrs
- Kim KE et al. TSAIO. 1970: 16:508
- Chaigon M et al. Hypertension. 1981: 3:327-332
- McIntyre C W et al. CJASN. 2008: 3:19-26
- Burton J O et al. CJASN. 2009: 4:914-920