It's my absolute pleasure to introduce today one of my friends, my colleagues, my mentor, Dr. Allison Grazioli. So Allison is co-trained in both nephrology and critical care. She currently is the medical director of our cardiothoracic intensive care here at the University of Maryland. And today her talk is going to be about renal replacement therapy in the intensive care unit and the challenges that come with it. So Dr. Grazioli, I'm very happy to have you here. Thanks for joining us. Thank you, Andy. I appreciate it. Thank you for the invitation. So like Andy said, I'm going to be talking about renal replacement therapy in the ICU and some of the challenges. This is unfortunately can't be a comprehensive talk. I would need several more hours, but I'm going to review.
Is that moving forward? I'm going to review the most common forms that we deal with in the ICU every day. All of the forms of renal placement therapy are listed here. So the most common one is CRT. Other forms that we'll see in the ICU are intermittent hemodialysis in more stable patients. Less commonly, we'll see peritoneal dialysis. And then some kind of intermediate modalities in terms of clearance between IHD and CRT are prolonged intermittent renal placement therapy and SLED, slow, low efficiency daily dialysis, which I'm not going to talk about in the interest of time, but are definitely useful modes to know about.
What I'm going to really focus on is mainly CRRT and some challenges in particular patient populations. So CRRT is a blanket term for continuous renal replacement therapy. There are several modes that you can use on CRRT. The clearance modes are CVVH, or continuous venous hemofiltration, CVVHD, or continuous venous hemofiltration. CVVHDF is kind of a combination of CVVH and CVHD, continuous venous hemofiltration. And then there are modes where you don't get a particularly robust amount of clearance, but you will be able to remove fluid in slow continuous ultrafiltration or SCUF.
So how exactly are we able to achieve clearance using modalities in CRT and IHD for that matter? So both intermittent hemodialysis and CBBHD, the slow continuous dialysis mode, will use diffusion to clean the blood, whereas the hemofiltration mode, such as CBBHD, will use convection. So, I mean, this is very simple, but just to kind of briefly review and put it in context of what's going on on your dialysis machine. So in both intermittent hemodialysis and CBBHD, the blood flows through the filter and it's exposed to a membrane which has different pore sizes. In the blood, there are things that are in higher concentration compared to on the opposite side of the membrane in dialysis modes. The dialysate is there, which obviously low in things that we want to clear, such as urea and creatinine. But things that are in the dialysate, which we do on purpose, such as bicarbonate, are at higher concentrations than in the blood. And so by simple diffusion, assuming that the membrane has a pore big enough to allow a given molecule to pass, things that are high concentration in the blood that we don't want in there will move into the dialysate and things that are in the dialysate.
and higher concentrations in the blood will diffuse into the blood compartment. So that's how intermittent hemodialysis and CBVHD achieves clearance. So how hemofiltration modes work is via convection. And so again, from the door, you wouldn't necessarily know which mode your patient was on unless you looked really carefully or saw the listed mode on the machine, because the filter's the same, the lines are the same. Again, the blood flows through, exposed to the filter. But instead of...
using diffusion what happens in uh in infiltrations mode is there's a convective force that's applied so from positive to negative across the membrane which will help push out plasma water and by solvent drag you'll pull solute out with it um and we know that i will kind of go into how you target um your your clearance um rate but air average clearance rates are two to three liters per hour in hemofiltration modes. And if we didn't replace that volume, we would obviously have a very hemodynamic unstable patient. So we end up replacing a lot of the volume that we will remove by convection. And we call it the solution set of dialysis replacement filter directly into the blood. And I'll illustrate that. All modes of dialysis, renal replacement therapy, I should say, have some very small component of clearance via adsorption, i.e. proteins will kind of stick to the filter membrane, but it's kind of minor. I mean, it decreases over time as the filter kind of gets bound up by blood components.
So really quickly, just to show you what's going on as you pass blood through a filter. So this is a very standard filter. So the blood will kind of flow in here when you're moving it from your patient. And in here, it will then be distributed into hundreds of thousands of these little tiny. If you were to cut the filter in half, they almost look like little hairs. They're microscopically hollow. And the blood flows through those. And those little hollow tubes are your filter membrane. They have pores. They allow it to diffuse out. And if you were to cut it in cross-section, these little red dots are essentially the blood flowing through those little microscopically hollow tubes.
So the blood will flow through and then it'll go out the other end. And if you're in a dialysis mode, there will be a dialysis port. You'll put your dialysate in. And then what will happen is dialysate will be bathing these little hollow tubes. And so that allows the dialysate to be exposed to the other side of the membrane instead of a diffusion gradient. And it flows countercurrent to the blood flow. That's done on purpose to maximize your diffusion gradient. And then it will kind of go out the effluent port. Now, if you're in a convective mode, there won't be dialysate here. What will happen is the blood will flow through and this white space will be empty. And what will happen is as the convective force is applied, plasma water will drain out from inside these hollow tubes with solute dragged along with it. And then that will drain out the effluent port.
So those are the main differences just to illustrate and just kind of illustrate them again. There we go. Here, if you're in CBVH, hemofiltration mode, again, blood flows through the filter, through the convective force, your plasma water and solute will come out here. It will drain out your effluent port. And what this effluent volume is, it's your replacement volume. And that replacement volume dictates how much clearance you're going to get per hour. And I'll kind of go over that in a minute. It's also included in this volume is the net amount of fluid you want to remove. So your net ultrafiltration volume, you're not going to replace that. And also the volume that you have added to your prescription that is going through your pre-blood pump. And I'll kind of talk about what that is coming up.
Now, your replacement fluid volume, again, is on the magnitude of two to three liters per hour, depending on your patient's weight. And if you weren't going to replace that, your patient would be very hemodynamically unstable. So we know we have to replace that. And so we take our replacement fluid and we put it directly into the blood. And we can either put it directly into the blood pre-filter or post-filter. Most often, it's a combination of percentages, pre-filter and a percentage of post-filter. And I'll talk about why we would do different combinations pre or post.
So the pre-blood pump, not all CRT machines have this. The Prisma Flex, which I think is widely used, does. It was kind of created with this theoretical reason that what it is, is if you take a bag of crystallite or some type of fluid, it has a separate blood pump here. What it does is it draws from the bag and it puts the fluid at whatever rate you say, mixing with the blood, kind of close to the access point. So then the fluid will mix with the blood all throughout the whole length of the bloodline before it gets in the filter. And what it's supposed to be doing is it's supposed to allow this fluid, which is obviously pouring urea, to elute urea out from inside the red blood cells to have it come out into the plasma so that it cannot be cleared.
It gives a very, very small amount of clearance. It's really very theoretical. But what we do pretty commonly is use this port, not to augment our urea clearance, but to add things to the blood that we would want, such as anticoagulation. We use the pre-blood pump here for citrate or for heparin. You can use the pre-blood pump as your pre-dilution replacement if you're doing CBV HDF and you want some pre- and post-filter, which I can talk about coming up. And then other fluids where you kind of want to tailor what your pre-mixed ordered replacement fluid is to make sure that you're giving your patient the appropriate concentration of things like sodium. And I'll talk about a case of how you handle hyponatremia.
And one really important point to know is when you set your pre-blood clump replacement rate, the machine knows what that is, and it's automatically going to ultrafiltrate that volume off, so you're not giving your patient that net fluid volume. It's going to get taken off via convection as it goes through the machine. So hemodialysis, again, this is simple. We kind of already explained it. So the blood is going through the same exact filter in and out. And here, now you're not going to take your same bag that you just called replacement fluid, and now you're doing a CVVHD more than a dialysis fluid. You're just going to call it dialsate. You're going to put it in a separate place. So you're going to put it in a port that's going to infuse the dialsate kind of around these fibers. When you look in cross-section, it's going to go countercurrent to the blood flow, and it's going to go out the effluent, same effluent port. And here, actually, made a mistake here. This effluent volume is now your dialsate volume plus your net ultrafiltration volume and your pre-blown volume. But in just continuous venous hemodialysis, there's no replacement volume. You just call it dialsate.
And continuous hemovenous hemodiafiltration is a combination of both. You'll do some clearance spike infection, and you'll replace that volume with replacement fluid, and then the remainder of your prescription you'll clear with dialysis. And then scuff is essentially the exact same filter. You pass blood through the same filter. You apply a smaller convective force to remove fluid on the magnitude of whatever you want, usually 50 cc an hour up to a liter an hour. And then also a lot of people use a pre-blood pump volume here. So what will come out in your effluent is just your net ultrafiltration. Again, you're not going to really clear the blood because it's too low and you're not giving any replacement fluid back. You will have a little clearance with solute drag, but not enough to really change your labs. You can still follow creatinine. You can still follow other markers in the blood because they will rise and follow patients' renal recovery.
So how does clearance by diffusion and convection compare? So if you choose CBVHD or CBVH, what is going on in terms of clearance? So if you look at really small solutes, so anything less than 10,000 Daltons with both diffusion and convection, they're cleared very, very well, I would say equivalently. So things like urea, creatinine, electrolytes, both cleared incredibly well. So it doesn't really matter.
too much. There's some small differences which we'll get into when you're talking about small solutes, which mode you use. Where these curves kind of diverge is when you talk about middle molecules. So anything kind of greater than 10,000 or 15,000 Daltons up into 50,000 Daltons. Now, the membrane pores really don't allow more than 50,000 to 60,000 Daltons to escape. So things like albumin, which is like 67,000 Daltons, should not leave a standard high flux filter. The pores are just not large enough. But things that we can clear with convection where you don't really get much clearance with diffusion are things that clinically we think about, such as myoglobin, so that 17,000 Daltons and certain cytokines can be cleared and other kind of middle molecules.
And so how much of a given molecule is cleared? And if we're talking about hemofiltration, so CVH, there's a term you should kind of just be familiar with called the sieving coefficient. And that's the ratio of a particular solute concentration that comes out in the effluents, whatever you're taking out of the blood, relative to what the concentration was in the plasma. So if it's incredibly well cleared, it's going to make it out very easily out of the filter. And it's going to be the same concentration that it was in the blood. So the sieving coefficient is going to be one. So things that have a sieving coefficient of one, it's kind of like a wide open door that can be cleared very well as they pass through the filter, are small molecules, so urea, creatinine. And don't forget things like that you don't normally think about that we're clearing, like water-soluble vitamins, such as B12, has a sieving coefficient of one.
Things that don't have a high steaming coefficient, which is a good thing, such as albumin has a steaming coefficient of essentially zero because it's too large to be cleared. And then middle molecules like myoglobin have a steaming coefficient of 0.58. This one specifically is 0.58. This kind of changes over time as things kind of get absorbed to your filter and your filter characteristics change. But you can get some good middle molecule clearance with convection. And then lactate, everybody always thinks about lactate clearance. We'll talk about that a little bit more coming up. So that's a very good high steaming coefficient of 1 so it can be cleared. But we'll talk about if that's clinically meaningful or not coming up.
And so which mode is better? Should you use a diffusion mode or convective mode? So we have not been able to prove in the literature that by any definitive clinical outcome that there's an advantage to one technique over another at given levels of intensity. So you can really choose whichever one your institution is more comfortable with. But you can kind of pick depending on what your goals are. Again, the really low molecular solutes are both pretty equally. If you really have a desire to remove middle molecules, then you're going to want to choose a convective mode. Again, we haven't really shown that that really changes major clinical outcomes.
learning point is there's more uh absorption when you use convective modes or filtration modes and that makes sense about you're dragging a plasma water across the memory so there's more opportunity for things to kind of it gets stuck with the force up along the memory and get cleared by absorption so Now I'm going to go through a couple of cases that's going to illustrate challenging situations that we see all the time in the ICU. They are pretty surgery-based, but you guys will see these in any ICU just because these are patients that we've actually had in our cardiac surgery unit. So here's a patient that came in, a 57-year-old female. Her estimated dry weight, so her weight before she came to the hospital was 70 kilograms. She's 5'3 inches tall. She's a history of non-ischemic cardiomyopathy. She suffered from worsening cardiogenic shock, so she was placed on the ECMO before she came to our unit. Her current weight, with all the resuscitation, rose to 80 kilograms by the time she came to us and was in our unit. She had a worsening AKI in the setting of shock, pericannulation. She developed an area. She needs to be started on CRT.
And what was notable is that her sodium was 120 and had been low several days since her admission, which was six days prior to her cannulation. So the first question is, is what is our initial prescription? Well, she's in Uric, assuming I don't, I didn't put up all the labs or just in the interest of time, but we want to have a standard clearance prescription, right? She's in Uric, she's going to be clearance. So how do we dose CRT for standard purposes? So it's recommended in the literature, you target a dose of 20 to 25 CCs or mLs per kg per hour. It's not really specified if that should be a total actual body weight or adjusted body weight. People do different things. I use total body weight just to...
for things that have the distribution of total body water that I know that I'm kind of getting a clearance of 25 to 25 cc per kg per hour. A lot of people aim higher, which is also very appropriate. So they'll go 30 cc per kg per hour just to allow for interruptions because people come down on CRT all the time. They have to go for studies. You know, there's always interruptions. So aiming for higher is not wrong either. Some people do that for standard practice. And so what's the evidence for this 25 to 30 cc per kg per hour? There's two large studies that came out about 15 years ago now. The ATN study and the renal study, which are randomized control trials. They looked at...
different intensities of IHD and CRT, but just to talk about the CRT, because that's what we're really kind of chatting about now, is they looked at these two different antimicrobial trials, looked at two different levels of clearance. And so in the ATM study, they see me the HDF with a clearance of 35 cc per kg per hour versus a lower clearance of 20 cc per kg per hour. And if you look at the renal study, they looked at a little bit different. They looked at post-dilution CBV HDF at a dose of 40 cc per kg per hour, compared that to the other group that had 25 cc per kg per hour. And in both studies, they failed to check mortality difference with a more intensive CRT. And so this is kind of why we were pretty confident that an appropriate dose is 20 to 25 cc per kg per hour to start for standard clearance purposes. And so now going back to our patient, our patient is 80 kilograms. And so 80 times 25, you know, in this range, we pick 25 to allow for interruptions, you know, just want to be on the higher end. So 25 times 80 is 2,000 mls per hour.
So two liters per hour is going to be her clearance dose. And so now, if you remember, she was pretty significantly hyponatremic. So what considerations do we need for hyponatremia? And so the first question is, is what is our sodium correction goal? And so we all know that anybody who's chronic and severe hyponatremia, we should aim for four to six milliequivalents in 24 hours and never more than eight. And why is that? Well, we know four to six milliequivalents per hour increase in a patient serum sodium will reverse the most severe manifestations of hyponatremia. And also we kind of, even though eight is our max cutoff, we aim for four to six. And that's because...
Our actual sodium correction is often much greater than intended. We all know this on the floor. We'll do our calculations. We'll follow it perfectly, and our patient will overcorrect, or they won't correct how we expected. And there was this nice paper that came out that looked at hospitalized patients, 62 patients. The majority of them were supervised by nephrology. And if their actual sodium correction was there expected, they would have fallen here at this dotted line. And using all the formulas that we have in MD-Calc, the androcheminitis formula, they follow that to a T, and greater than 72% had NX.
70% had a greater than expected rise. And this happens all the time. And it's really what underlies is our formulas are not perfect. They don't really take into account other ins or losses other than our chosen hypertonic salient order ever using to correct a patient. Another reason for this goal is if you read old textbooks, you'll see that the goal used to be like 10 to 12 was thought to be safe. And this kind of got much more conservative in the most recent years, decades. And that's because while we know most cases of osmotic demyelination, which is a big fear of overcorrection, occur with a correction rate of 10 to 12 mil equivalents in 24 hours, as cases had been reported with much less steep corrections, the goals kind of got more strict than they are what we have today for six mil equivalents, never more than eight. And so just to quickly go into what is exactly osmotic demyelination syndrome and what underlies it.
So I just think it's kind of an interesting pathophysiology. So it involves these star-shaped glial cells, to the best of our knowledge, called astrocytes, which make up the bulk of our intracellular space in our brain. They're five times more numerous than our neurons. All of these processes will kind of connect and modulate other cells in the brain. So they'll connect with endothelial cells. The astrocytes will help regulate the brain barrier integrity. They'll connect with the oligodendrocytes. It'll have a homeostasis role in the maintenance and development of normal myelin. And where they...
where they really function in hyponatremia is they're responsible for regulating brain water and organic osmoly homeostasis. So you know when a patient is hyponatremic, it really matters how chronic they are. And so once they've been chronically hyponatremic for about 48 hours, that's when you really have to worry about not correcting them too rapidly, more so than if they were acutely hyponatremic. That's because the regulatory mechanism in the brain is the astrocyte, which will lower the organic intracellular osmo content. And that can't reverse quickly. So you have to slowly correct them back up to allow the astrocyte to build back up the intracellular organic osmoles to not allow them to have too much of a self-stress. So if you overly, if you rapidly correct, you're going to have a very, very strong stress response. The astrocytes will undergo apoptosis. You'll get disruption of the collegiate barrier and exposure to inflammatory cytokines and microglial activation in the brain. And then the downstream lesions are you get a myelin loss and demyelination.
And so now back to our patients. So how fast will we correct our patients at 24 hours? Can we predict this with a given prescription? And the answer is there are models out there. One that I like is a single pool fixed volume sodium kinetic model, which can help you kind of predict how your prescription will correct your patient. So it has the sodium concentration at a given time. So at 24 hours, that's what this is here. What does your patient start in sodium? So our patient starting in sodium was 120.
What is the concentration of sodium in the dialysate or the replacement fluid? So we mostly all buy standard commercial prescriptions in the actual factor of one here. You'll see that a very standard sodium concentration, which is almost all of them, is 140 ml per liter. And then you subtract that by the initial sodium, so 120. And then it's 1 minus E to the minus D times T. So D is the effect of sodium dialysins or clearance, which is about equal to your rate clearance or small molecules. And so what that is essentially the sum of the dialysate and replacement rate. So in our patient, like I said, our prescription was 2 liters per hour, so you put a 2 here. And then times T, well, we want to know how much they'll correct in 24 hours, so times 24.
And then you divide it by V, and V is the total body water volume of your patient. So use the standard total body water volume estimating formula like the Watson formula. And then I like to use a patient's dry weight. It's recommended to use a patient's dry weight because all of that extra weight that they have in the hospital is mostly probably from resuscitation, so water, just pure water. So use the dry weight to estimate their estimated dry weight total body water, and then you add the extra weight that they've had in the setting of resuscitation. So if you use the formula, you get 32. We give our patients a dry weight with 70 kilograms plus 10, and then you get 42 here. If you do the math, you get 134. So at our standard prescription,
How we will correct our patient using our commercial dialysate is 134. So a way big overcorrection. So we really can't do that. We have to figure out how to mitigate that. And so what are our options? Sorry, my thing is not bouncing. How can we prevent this overcorrection? And what modifications could we make for the hyponatremia? We really have two options to mitigate that rapid rise in your patient serum sodium. So you can decrease your total clearance rate. Remember we said we targeted our standard 25 cc per kg. Well, what if we just have that rate? What would that do our patient serum sodium?
If you plug it into the formula at half the rate in 24 hours, you're still going to get an overcorrection. You're going to go from 120 to 129, not just above your upper level of normal, but you're also not really getting the clearance of other things that you'd want to clear in your member of a patient's AMERIC. So what are our other options? Because this is still going to give you a little bit of overcorrection. So what a lot of people will do is they'll adjust the effect of sodium concentration in the replacement of the dialysate fluid. And I'll show you how people kind of do that. And you should always kind of do this with expert oversight. So again, our sodium concentration in a standard dialysate bag or replacement fluid bag is 140. We order that. If we had the luxury of compounding our continuous replacement therapy fluid, we could lower it, but we don't. We buy it. So it's 140 milligrams per liter. Our goal of sodium in 24 hours, our upper limit of goal correction is 128. So we know these two things.
We can't compound our fluid because we work in hospitals that don't routinely do it. So what we do and what we can often do is we can add water either to the pre-flow pump or peripherally to essentially make the bath effectively something that will not overcorrect our patient. So we know 128 won't overcorrect our patient. That's our normal. We'll bring the sodium up to where the replacement fluid is. So 128.
Again, you could give free water peripheral or D5 water peripherally in a separate line. But remember that that is not going to be taken off by the circuit. So that volume is still going to go into your patient. So if you're really concerned about patients having extra volume, that's not going to be automatically removed. And so if we were to add D5 water to make our dialysate effectively something closer to 128, what would our patient's correction look like? So I plug that into the formula here and it's 125. And that's nice. And within the safe range, if we went up a little bit, we're still not going to violate 128 as easily. So that is a nice option. And again, you can put it through the pre-blood pump.
And so how do we choose how much D5 water we would want to put in the pre-blood pump to get our goal to be 128? So just to kind of make it really simple, this is what we have. We have a bag of replacement fluid that has a sodium concentration of 140 milliequivalents per liter or per 1,000 mLs. So it's 0.14 milliequivalents per liter. What do we want? We want that 128 milliequivalents per liter. So this is what we want.
And so how much T5 per liter of replacement fluid do we have to add to the pre-blood pump to get an effective bath of 128 mql per liter? So just doing the math here, 140 plus over 1000 plus X equals 0.128. The volume you have to add per liter is 93.7. And because our replacement volume was two liters, we have to double that and get 187.5. So I just kind of did a little chart here for you. In my practice, when I do it, and if I'm using CBVH, I'll put all the replacement fluid pre-filter.
And obviously, the free blood pump goes pre-filter just so it all kind of mixes together and there's no difference in dilution factors we'll talk about coming up. And again, make sure you have to multiply that volume of D5 water by your clearance prescription. So if you're doing two liters per hour, you're going to take your number here and multiply by two. If you're doing three liters per hour, the clearance will be multiplied by three. Please make sure you monitor your blood sugar giving your patient D5. You can't really put just pure water in there. The intensity may be just a little too low. So you have to really make sure you're monitoring their blood sugar. They will get hyperglysmic, especially if you have to get higher rates. Caution with using. This is just a general rule when you're giving D5, when using more than 300 cc's an hour of D5. Have a sense of your patient's other intakes. This is just generally true. There are other intake volumes in their osmolality. If they're getting a lot of hypotonic fluid, then maybe...
You don't need this much D5. If they're getting a lot of isotonic fluid, then maybe they are going to maybe still overcorrect. Still need very close sodium monitoring. Nothing is perfect with hyponatremia. Nothing will get you away from monitoring serum sodiums very frequently. And always have somebody overseeing you, an expert nephrologist or somebody who's very skilled at renal placement therapy when you're doing something like this.
And so what if your patient overcorrects? So just very quickly, if they overcorrect, what do we do? So we don't really have great data in humans, but there's some really interesting animal data. And I just like to show this study. So here are rats that they made very chronically hyponatremic and severely hyponatremic. So this is a controlled group of rats that were hyponatremic to 108. And they left them there for days, hyponatremic. And you see their brain pathology is very normal. Then they took another group of rats and they overcorrected them pretty dramatically. So 24 hours, they went from 108 to 137. And all the rats died. Their correction rate was...
30 and 24 hours. They all died and they had these giant demyelinating lesions on their brain pathology. They took this other group of rats and did the same thing, made them hyponatremate, chronically hyponatremate, overcorrected them to the same magnitude. But before they did that, they gave them steroids. And many of them died, but some did live. But when they looked at the pathology, there were still osmotic demyelinating lesions, but they were a little bit less.
The third group of rats they took and they overcorrected them by steep magnitude from 104 to 133 in 12 hours. And then they re-lowered them back down to something that we would have considered within our goal of correction. And then almost all of those rats lived and they did not have asmodic demyelidating lesions. So this animal data kind of shapes our practice and how we treat people. And here's just another thing. This is Evans Blue Dye looking at the blood-brain barrier. And in the control group of rats that were just having a treatment, intact blood-brain barrier, overcorrected ones alone had Evans Blue Dye leaking out so they had reached their blood-brain barrier.
With the steroids, that helps protect the blood-raid barrier, increase in permeability, as did the therapy to be lowered. So what about in people? Again, no robust evidence for therapeutic re-lowering in people, but we do do it. So the therapeutic re-lowering target should be just below, depending on where you are, if you're just within 24 hours and go back down to where you would want to have been within the first 24 hours. But if you're beyond that, you'll want to re-lower back to what your maximal target would have been 48 hours out. So if your maximum target is 8, then 16 mL, you're two days out. So rates that have been published to be considered to be, you can re-lower back down at a rate of 1 mL per liter per hour. And it's said to be justified if the sodium correction, again, is a greater than 8 mL per liter and your patients have dyspractors for osmotic stimulation.
severe hyponatremia. There's a whole bunch of other risk factors which are kind of beyond the scope of this talk. Or if you really over-crack them greater than 10 to 12 milligrams per year, you should think about re-lowering everybody. And again, the efficacy and safety is really not known. I looked in the literature to see what the steepest re-lowering rates were reported. There's just some case reports out there. So they have re-lowered people from 137 to 115, 132 to 120, and 122 to 116. And the magnitudes of over-correction are here, and they're pretty steep. And these patients in these case reports, they're just case reports that not have any ill effects from the over-correction that occurred initially.
And so with your patient on CRT, how do you lower them? Well, one, adjust your CRT serum sodium concentration using your pre-locum to what your sodium goal would be. So let's say you go to 130, adjust your bath to 130, 128, using all the techniques that I told you before. And then you'll have to give the water volume back to bring them back down separately, either enterally with D5 water or a combination. If your patient's not on CRT anymore, you know we have the same two strategies in the literature. One, again, replace that water loss, either enderal D5 or combination, to get them back down. Don't forget about the water loss in their urine. There's a formula that you can use to calculate how much of your patient's urine is free water, so you can kind of use this calculation to give that back in addition to what you've already noticed as the amount of water they'll need to get their same sodium concentration back to your goal. So moving on to the second case.
So 55-year-old, 80-kilogram male who had an aortic dissection, was placed on VA ECMO, proposed his aortic dissection repair, had severe rhabdo, secondary as a side effect of the dissection. It was extensive. CKs were very elevated, developed worsening hyperkalemia postoperatively. Hemodynamics were stable, though he was on low-dose pressers and low-dose epinephrine for inotropy. And so the question is, what are our options for hyperkalemia management using renal placement therapy?
So when you want to talk about potassium clearance on CRT, there are three factors to consider. So one, what is the diolsate or replacement fluid composition of potassium that you're using? Two, what mode are you choosing? And three, what is the dose that you've chosen or i.e. your clearance rate? So when you're talking about your fluid composition, you can get increased potassium clearance if you use a lower potassium bath. And again, we don't have luxury of compounding fluids here. So usually our standard bath is either four mil equivalents per liter or zero mil equivalents per liter. Some pre-made bags do have two mil equivalents per liter.
Now, if you're using CVVH and you have the option of using replacement fluid pre- or post-filter, placing your 0K portion of your replacement prescription, some people use a part of the replacement fluid will be 4K and the other one will be 0K. If you choose to do that, placing that 0K or that lower K concentration fluid post-filter will enhance your clearance compared to placing the 0K or the lower potassium concentration fluid pre-filter for any given replacement rate. And in severe hyperkalemia, some institutions, some large institutions will have protocols to use all zero K, especially if you're using CBVHD, you don't have an option of using some pre, some post. It's all dialysate and it's all one concentration. So that's always the case with CBVHD. But if you're using a convective mode, you can just do it for your post. In severe hyperkalemia, it is appropriate to use all zero K. But just remember, you really should have very frequent potassium checks until you get out of the more dangerous range. And so once your potassium goes down to less than six, then you can switch back over to four K and then maybe space out your potassium checks.
So just to look at ideas of what your commercial replacement fluids look like. So a lot of people say the BGK. What does that mean? So B stands for bicarbonate, G stands for glucose, K stands for potassium. So the ones that we purchase here, you see right after the letter will be the concentration. So this is our zero potassium bath. The very standard bath is the BGK4 slash 2.5. The 2.5 is calcium. So it has four milliequivalents. We don't purchase any two milliequivalent liter bags, but they do exist.
So let's just talk a little bit about the difference of putting fluid pre versus post filter when you're doing a mode like CVVH. What's the difference? Why would you choose some portion pre-filter versus post-filter? And what are the benefits of putting mostly pre-filter versus mostly post-filter? So to look at pre-filter, what are you doing when you give pre-filter solution? Well, you're diluting everything in the blood before it goes through the filter, right? So you're going to dilute your clotting factors, which is great, because if they're diluted, then you're going to have a little bit less of the time for them to interact and form clot. So that is kind when you're talking about trying to prevent clotting. Putting pre-filter fluid will prolong your filter half-life.
The downside of pre-filter replacement fluid is that you do get a decrease in your effective small solute clearance, about 15% to 20%. It kind of varies depending on the molecule, given a dilution factor, right? So if you're going to put your replacement fluid pre-filter, you're going to dilute out the concentration of the thing you want to clear. So there's a dilution factor, and you're going to clear a little bit less of that. Now, to overcome that, and it never completely overcomes it, you can think about increasing your blood flow rate a little bit, and or your total replacement rate. If you increase your total replacement rate by 15% to 20%, you should overcome that to some extent, overcome that decrease in clearance. Now, when you use pre-filter replacement...
fluid, and I like to use a lot, most of mine, I sometimes use 100% zero. I do that because of this other concept that you can have a gain in middle molecule clearance. And why do you have a gain in middle molecule clearance? Well, again, you're diluting out all the things in the blood. And as you hemo concentrate along the filter, You start to get the secondary membrane effect. All those larger molecules will kind of become a hindrance to middle molecule clearance. But if they're diluted out more by pre-filter replacement, you kind of mitigate that secondary membrane effect. You also decrease solute and membrane interactions, which can enhance mass transfer. So I like a lot more pre-filter. Most people do more pre-filter, but I'll do 100% pre-filter just because I think I like this gain in a molecule clearance that you can theoretically get.
So why do you use post-filter? Well, one, fewer reductions in clearance, right? So you're not diluting out the stuff you want to clear. You're adding it all back post-filter. It does unfortunately shorten your filter half-life. You're chemo concentrating along the filter, so all the clodic factors are closer together, especially by the time you get to the end of the filter. So it's almost impossible to really get away with 0% pre-filter and all of your post-filter post-filter. You really will probably clot. It will limit your ultra-filtration by chemo concentration. Ultra-filtration is always by convection, so if you're using CBVH, that's the convection for clearance, but any ultra-filtration that you do, no matter what the mode is.
chemo filtration, and that will further increase your hemoconcentration and your filtration fraction, which I'll talk about coming up. And post-filtral replacement will increase your need for antiphagulants. So the second thing to think about with potassium clearance is the mode that you pick. So all clearance modes, as I said, are great for small solutes such as potassium. So CBVH, CBVHD, CBVHDF are all fine for clearing a small solute.
The hemofiltration modalities, as I just mentioned, you'll have a dilution factor when you use pre-filter replacement fluid component, which will kind of decrease your maximal small solute clearance per unit of replacement fluid, which I just mentioned. This dilution factor is not present with CBVH. You're not diluting the blood before it goes to the filter. With CBVHD, it's all by diffusion. And it's also not present when you're using exclusively post-filter replacement on CBVH or CBVHDF.
The pre-filter replacement fluid dilution factor, again, I mentioned before, can be, though to a small degree, mitigated a little bit if you increase your blood flow rate, right? You're bringing more blood to the amount of replacement fluid you're giving. So the higher the blood flow, the more blood per ml of replacement fluid. So it's a little bit less diluted. And then finally, your dose. So your dose is essentially dictating your clearance rate.
And the removal of potassium on any mode of CRT is limited by your rate of clearance. So how much can you clear on a CRT? It depends on your machine. Next stage is a little bit higher, but the Prisma Flex, which is what I think most of us use, allows up to eight liters per hour of either replacement or dial state clearance. If you increase doses of continuous hemofiltration, unlike continuous hemodialysis, will have an increased filtration fraction, which can lead to filter clotting and increased time off CRT. So if you're going to use a mode like CBVH at a very high clearance rate, your filtration fraction is going to be pretty high and you probably will clot, so you're just going to risk downtime. For small molecule clearance, as I said, it's probably fine, but you just have to think about something like that. And so what do I mean by filtration fraction? So the filtration fraction is the ratio of effluent or your convection rate of filtrate.
to the plasma that's entering the filter. So if all of your blood here is entering the filter, so there's a cello component in the plasma, the ratio of the effluent that comes off, so the more that you take off, the more effluent you take off of a given amount of blood per time, the higher your filtration fraction. And so the filtration fraction equation is the flow of the effluent over the flow of the plasma. And so your effluent is equal to everything that comes off. So your replacement rate, your pre-blood pump volume, because that's cleared by convection, and whatever, you're removing net ultra-filtration. So that's here.
up here in the numerator. And in the denominator is your blood flow rate. And it's one minus hematocrit because you really are only talking about plasma. You want to subtract out the cellular portions of the lower hematocrit, the lower filtration fraction. And then you want to add into the denominator the pre-filter volume because that's going to kind of dilute out your blood and it'll help your filtration fraction to a small degree. It'll decrease it to a small degree. And what your goal is, is you want a filtration fraction that's less than 25% to reduce your clotting risk. So that's always our goal. And I just did this math. You don't have to really go too crazy just to show you the comparison. So if the patient's really, really hyperkalemic, you want to use your PrismaFlex to the maximum. You're going to choose an 8 liter per hour prescription, which is the absolute max. What do your filtration fractions look like? So if you're using CVVH, and here I chose 50-50, so 50% pre-filter, 50% post-filter.
The pre-blood pump on both prescriptions are zero. The ultra-filtration rate is 100. Your blood flow is 300. On the CVVHD prescription, again, the same eight liters per hour of clearance, no pre-blood pump rate, and that's fine. You don't have to use a pre-blood pump. Same ultra-filtration rate and the same blood flow rate. So just doing the filtration fraction calculation with the CVVH mode, because actually your clearance component adds to the filtration fraction, your filtration fraction is 52%, which is a setup for clotting.
with a dialysis mode at the same clearance rate your filtration fraction is less than one percent so it's just not even in the same ballpark in terms of cvvhd is much more protective in terms of your cladding risk it's just the only thing that contributes to your filtration fraction um in this prescription is the ultra filtration sorry i'm having a okay and this is a prisma just you know the machines will calculate your filtration fraction so you don't always have to do the math but just know that the nurse or whoever is running the machine has to enter them accurately or it might not be all that accurate but you can always get that number um with any given prescription on your machine So now back to our patient. So we chose CVVH at double the standard dose because we had a desire for middle molecule clearance. Because remember, I said he had rhabdo. We know that middle molecules are fueled better by CVVH. So we gave a nice robust dose of CVVH, one, to clear potassium, and two, because we felt we wanted to clear some of that myoglobin. So we started CVVH at four liters per hour. We put two liters pre-filter of replacement fluid, two liters post-filter of replacement fluid. We decided to use all zero-K baths. So we were monitoring our potassium pretty frequently.
And then, unfortunately, our patients' rhabdo got worse, compartment got more tense, and they had to have a lower extremity fasciotomy at bedside. And then as what happens with fasciotomy is once you reperfuse all that potassium that was sitting in there that wasn't able to kind of be in circulation, gets back into circulation, and the potassium rose further, and we're still at a robust CBVH rate to 7.4. So we decide we're very nervous that that endogenous load of potassium is only going to keep going back into the blood, or potassium is going to be even more out of control. So we increase our dialysis, our CRRT prescription to 8 liters per hour. We change our mode because we don't want to clot our circuit to CBVHD at 8 liters per hour.
But unfortunately, even that does not really get our potassium into a safe range just because of how dramatic the rhabdo is. And the potassium remains high in the sevens range the following day. So now what are our options? We have a patient on ECMO. The CRT machine is hooked up to ECMO. And we've maxed our clearance rate. And we still have potassium that's pretty high.
So, well, we could add a second CRT circuit. We can't really add a second CRT circuit to your ECMO, but you can put it through a dialysis catheter. Or you can do IHD, which is really recommended with severe hyperkalemia, if possible, is to do IHD. And just, you can do IHD in line with ECMO the same way you connect up your CRT machine. Your perfusionist will take, and the dialysis service will hand the bloodlines to the perfusionist. They can hook it up the same way you hook up CRTs. You can do intermittent dialysis through your ECMO circuit. We've done it here several times.
And so which one of these will offer higher clearance, adding a second CRT circuit via catheter or IHT? And so when given the option, again, in cases of severe life-threatening hypercleaner, you really should consider hemodialysis at least first to get your potassium into a safe range. Then you can follow with CRT just to continue the clearance. And the disadvantage of hemodialysis versus CRT is just that hemodialysis, even at the same blood flow rate and without ultrafiltration, can exacerbate hemodynamic instability. And it's because you clear solute really fast, you're returning solute to your patient that's relatively hypotonic to where they were.
And that transiently causes a fluid shift out of the blood into the intracellular compartment until there's a complete reequilibration. So that returning of, you know, transiently hypotonic solute can exacerbate some intravascular volume depletion. And just to kind of put into perspective the difference in clearance that you get, and because this is comparing dialysis to dialysis, so intermittent hemodialysis to CVHD, what the differences are in clearance. So when you're using a catheter, I know you guys don't write. hemodialysis prescriptions. But when using a catheter, your hemodialysis blood flow rates through a catheter are pretty similar to what they can be on CVVH. So we usually won't really go much above 350 cc per minute. So the average is like 300 to 350 cc per minute on IHD, which is similar to anybody who writes CVVHD prescriptions. The average blood flow rate using a catheter is about 250 to 350, I guess, at most. So pretty, pretty similar blood flow rates. So where is the clearance difference? The clearance difference is completely in the dialysate rate. So when you write your standard dialysis prescription, you're...
Our dialysate rate is 500, really 600 cc's per minute. And if you put that on the same units as what our CVVHD rate, so that's 30 to 36 liters per hour. Now we write CRT prescriptions. We know that our standard CRT clearance rates end up being on the magnitude depending on our patient's weight of two to three liters per hour. So the volume of the amount of clearance is just dramatically different. So much so that dialysis can clear so well, it's almost never, ever okay to use a zero K dialysate.
concentration on intermittent hemodialysis. You will clear too fast and too much. It's too dangerous. The most we'll do, and this makes a lot of us nervous, is one mill equivalent per liter. Whereas with CBVHD, with somebody who's really hamper clean it, we'll do zero K all the time. It's just because the clearance is so much more dramatically lower than on intermittent hemodialysis.
And that is why we all mainly know the benefits of CRT versus IHD. So the benefits, better hemodynamic stability with CRT, capable of greater clearance over time. The dialysis, intermittent hemodialysis wins the race in the short term. You're on CRT for 24 hours a day. So it's kind of like the tortoise in the hare. The tortoise over time will win the race, but per hour IHD wins. CBVH will manage, or CRT will manage volume just because you can remove volume every single hour for 24 hours, where that's not the case with your standard IHD prescription, which only lasts about four hours. And there's less effect on intracranial pressure. So anybody with intracranial hypertension or brain injury, you should use CBVH. And it's for the same reason when you clear solute, you have transiently hypotonic fluid.
that you're returning. And so in the brain, that's not great because then you have higher tenacity in the cells of your brain than you have to the blood that you just returned. And so the water and the blood is going to go to the water and it's going to shift to the cells in the brain. Even if it's transiently, it's still going to increase your intracranial pressure and that can make a brain injury or definitely exacerbate intracranial hypertension. So that's why we do CVV, HRCRT, I should say, in brain injury versus IHD. And so real quick, as I start to kind of roundup, I'll finish, but I'll do as much as I can. I have one more case, but it's, I think, quicker. What role did...
our renal replacement therapy have in this patient's rhabdo. So what do we do? How do we manage rhabdo and CRT? This comes up a lot, especially in trauma. There's a lot of crushing injuries. There's a lot of rhabdo. And again, convective clearance can remove a portion of myoglobin. So should we factor that into how we approach a patient who needs renal replacement therapy? So again, I said myoglobin is a mineral molecule and it's cleared fairly well. It has a sieving coefficient. It's not quite one. One is the best. It's 0.58. But when they really look at it, this actually drops pretty quickly. In over 24 hours, the sieving coefficient can get low just so all that stuff gets absorbed and will change the amount of mineral molecules that can be cleared as you kind of clog up your filter.
So how much myoglobin is removed? We don't really know exactly. It's obviously dependent on your specification, how much myoglobin they're generating, depending on how bad their injury is. And obviously the filter type and how you're receiving coefficient fares over time will change the amount of clearance and your prescription. So it's not known exactly in any given patient how much myoglobin you're able to clear. So one famous nephrology researcher did some nice studies 23 years ago now looking at CRT with a standard filter and did some analysis and found that he was able to remove up to with his modeling about 25% of the myoglobin load.
over 24 hours. So adamantils are able to study it a little bit better in the sense that they can just kind of infuse a given myoglobin load so you know exactly what your generation is. And they were able to find that with their convective clearance study, they were able to clear about 10 or 15%. So it's a good portion. It's not 0%. It's not 90%, but it's a substantial portion. And we all know that we can clear myoglobin because we see it in our effluent bags. This is from a study where they did high volume filtration over several days on a patient with rhabdo. And it was nice and red. Here's the myoglobin pigments that color. And that cleared in the effluent as the myoglobin cleared. You know, normal effluent is kind of yellowish. That's urea. So that's a normal color.
And so should we start CRT to clear myoglobin or should we start replacement therapy for myoglobin? So the data doesn't really support that. For clinical outcomes, first of all, just to preface this, all the studies are really, really low and terribly low quality. They were able to have been able to show in literature that convective clearance will decrease your myoglobin. As I mentioned, some studies show can decrease the duration of oligary and length of stay, but there's really no.
mortality benefit or really improvement in our core renal recovery outcomes. So there's no real indication to start in the literature, at least guided by, to start a CBVH or renal replacement therapy just to clear myoglobin alone. So the standard teaching is you start renal replacement therapy in a patient with severe abdil for any other reason, you would start dialysis in any patient. But just knowing that they're an exogenous load of things like potassium may be higher than just somebody who is ignorant and doesn't have that. If you're going to really, really be dead set on clearing myoglobin and you do something like hook up two CRT circuits, just remember you're doubling your clearance and you're doubling solute clearance of things that you don't always think about, like amino acids, vitamins, catecholamines, antibiotics, medications, and other compounds. I didn't even talk about other things that we're clearing. That's a whole other talk. But you just have to be mindful that you're clearing things that we don't typically think about.
So the final case, I'll try to get through as much as I can. This 50-year-old, 70-kilogram male, you have 30%. Another type data section in our CSICU, status post-repair, post-operatively does well. Then on post-op day one, developed worsening acidosis, anuria, rising press requirement, had a lactate of greater than 17. His serum bicarb fell to nine. The testing was still okay at 3.5, and his pH was 6.95. We ordered a stat bedside echocardiogram. We got a CTA to look at his repair.
we plan to start crt uh he is volume responsive and so we're giving boluses a fluid albumin um because of his essence we push some bicarbonate uh his pressure requirements continue to increase in a gas that we drew sort of ph of 7.1 and a pco2 of 25. and so the question is we know what is the indication for crt so just to take a step back and to just review um where the acid comes from um and we call it lactic acidosis So if we look at the metabolism of glucose, I think some people seem to think that glycolysis in itself produces the H plus plus lactate minus, right? It's not actually true. So glucose gets metabolized in the cytosol to pyruvate. And if you look at what goes into that, so you make two ATP and you make two hydrogens. However, when pyruvate then gets metabolized through anaerobic or aerobic glycolysis to lactate, those two hydrogens get consumed. So there's really no net hydrogen production when you take glucose and metabolize it to lactate in the cytosol.
Now, if pyruvate did not become lactate and it entered into oxidative phosphorylation in the mitochondria, we all know that you make much more ATP and it goes through the TCA cycle. But that is not always the case when we have some type of metabolic stress and we're undergoing glycolysis. So now just look at this again. So where does the hydrogen come from? How do you get acidotic with lactic acidosis? So what happens is the ATP that's generated from glycolysis will become hydrolyzed.
And so that's that. The hydrolysis of ATP will give you the hydrogen. And if you're not undergoing oxidative phosphorylation, those hydrogen ions would normally be buffered by this process. They're no longer buffered. So it's the fact that you're making ATP and hydrolyzing ATP, and there's no longer this buffering of these hydrogen ions that causes the acidosis. So yes, you're making lactate, but the H plus, the acid part of it comes from the hydrolysis of ATP.
So now the question is always, so your patient has lactic acidosis. I don't like lactic acid or lactate. It just looks bad. We know our patient's in trouble. Should I put them on CRT to clear that lactate? First of all, no. Lactate is a bicarbonate equivalent, and I'll kind of go into that now. What I think people really want to do is they want to stop the acid production, but that isn't really what CRT will do. So just this really interesting study came out in 1997 where this group took patients with stable lactates, normal lactate levels. They were ICU patients, and then they infused a given amount of lactate, so they knew what the lactate load was.
It's a continuous rate. And from that, they were able to calculate the endogenous clearance of lactate and what they were able to clear with their given CRT research. And they started these patients on CRT. Remember, lactate gets cleared really well at a saving coefficient of one. However, compared to what our endogenous clearance is, which is all up here in the scale of 13,000 units are, the lactate was 24.2. So it's like a drop in the bucket of what we endogenously clear. So yes, your CRT circuit will clear lactate, but it's so small compared to what our endogenous clearance is that it doesn't change your serum levels. You can follow lactate as a marker of metabolic stress. You can't say, oh, well, they're lactate cleared because I started CRT. That doesn't.
That doesn't mean anything. You're lactate cleared because your patient's metabolic stress went away or resuscitated them. You're treated the reason that they had the lactate generation to begin with. Again, remember, you don't really want to clear lactate. Lactate is a bicarbonate equivalent. It's going to get metabolized back to bicarbonate. What you really want to do is you want to stop the acid production or clear the acid, but you can't clear the acid. The acid gets buffered. That's why we have all these redundant buffering systems, one of them being our bicarbonate system. So once the acid is produced, it's buffered. It's not floating in the serum. We can't clear it by CRT. So what we're really doing when we're starting CRT is essentially using it to administer more buffer that's been used up by the acid process, the acidosis.
And we know giving bicarbonate and in various different ways pushing it infusions isn't really supported in the literature, but we still do it, especially when your pH is critically low, like less than 7.2 or 7.1 when our patient is really low to getting under 7. And so what are the options available to increase bicarbonate delivery via CRT? I'll go through these really quickly and I'm almost done. One, you can increase your clearance rate. And so how much, if you increase your clearance rate, your prescription, how much additional bicarbonate are you giving? Well, if you look at your bed, what is in our standard CRT dialysine? And again, the compositions are a little different depending on which bag you choose, but the standard bicarbonate concentration is 32 milliequivalents per liter. And if you notice, there's some lactate in it as well. Again, lactate's a bicarbonate equivalent. So you count all of the base equivalents, it's 35 milliequivalents per liter.
Remember, if you're increasing your clearance rate, you're going to, let's say you go from two liters per hour to three liters per hour, you're then removing an extra liter per hour of bicarbonate for plasma. So that will give you additional bicarbonate addition, if you think about it that way. So how high can you go? We talked about this. It depends on your machine. The Prisma can give you eight liters per hour of clearance. If you have the luxury of compounding your own CRT, you can add bicarbonate to your CRT fluid and have a bath that isn't 32 milicolves per liter, it can be 40 milicolves per liter. I know they can do that at the NIH.
Another thing that is done very often is they'll get, we'll give additional bicarbonate if we feel that we have to buffer our patient quicker and more buffer than the 32 mcl per liter, and we'll give additional bicarbonate through the pre-blood pump. And so what we're doing is effectively increasing our bicarbonate bath by adding bicarb to the fluid through the pre-blood pump. And so we know that we do that with our typical sodium serum bicarbonate drip, which is 150 mcl per liter of sodium bicarbonate. We'll do that through the pre-blood pump. So I just did a little bit of math here. So the starting bicarbonate is 32. That's our standard bath. And the volume of a bicarbonate drip that you'll need to increase your...
your bath, two of the concentrations listed here, so 39 or 47 are here. So 50 cc's of a bicarb drip for your pre-blood palm per liter. Then you have to double it if your prescription is two liters, 100 or 150. So the point is here, you can do this. There's really no evidence for it. It will increase your serum bicarb quicker, but please avoid using D5. Some people won't think about it, but they'll be making their patients hyperglycemic, and this is a hypertonic fluid. Just use sodium bicarbonate and sterile water.
If you're going to do this, you can also infuse peripherally. If you infuse peripherally at the same rate as you infuse the pre-blip pump, just remember peripherally is going to be additional volume. The pre-blip pump will take off that volume. And then one thing that I really want to warn people not to do is I've seen people do is they will take your post-filter replacement fluid and not use dialysis or replacement fluid like a standard replacement fluid bag. They will add the sodium bicarb drip as post-filter replacement fluid. Do not do that. Remember.
The sodium bicarbonate bag is just sodium bicarbonate. It doesn't have any of the other stuff that's going to give you a balanced replacement. It doesn't have calcium. It doesn't have glucose. It doesn't have magnesium. It doesn't have potassium. You can get severe derangements. We've had that happen. So never, never, ever do that. And so I think that's it. So those are just a couple of cases where we can come up with some common ICU problems where we have to kind of just be really thoughtful and really understand what we're doing with CRT. There's a zillion other cases that we could do, but just in the interest of time, I just pick those three. If anybody has any other questions, any questions, I'm happy to take them.