So I wanna first of all thank the organizers for inviting me. It's really an honor to be here, talking about things I really enjoy a lot, most of which is acute kidney injury and other sorts of issues like that. I am dual trained in critical care and nephrology. I work mostly as an intensivist, but I'm also the director of critical care nephrology. As far as disclosures, which won't change from presentation to presentation, none of them are pertinent to this, but I will point out that I am the author of the CRT prescription writing chapter of Up To Date, and I can't avoid talking about CRT and when you may want to start CRT, so I have a little bit of conflict there. Okay, these were the goals and objectives I was given for this talk, which is to define acute renal failure and common risk factors, state the outcomes of patients, understand and describe how we diagnose and manage AKI in the ICU, and describe effective measures to prevent AKI in critically ill patients. Hopefully, I will say a few things that might be humorous and maybe a couple of things that might be controversial. Please save your tomatoes or rotten eggs for afterwards. All right, let's talk about these two things first of all. So let's start with a case, 47-year-old guy, history of diabetes, came in with four days of fever, progressive shortness of breath. In the ER, he had marked respiratory distress. He was hypoxic, tachycardic, a little bit hypotensive and febrile. You can see these were his labs on presentation, his keratinin is already 1.7. We don't have any data to know whether or not this is his baseline or what his pre-morbid keratinin is. He's intubated emergently in a CT chest with multifocal infiltrates and consolidations. He gets admitted to the ICU and overnight, he gets a total of five liters of 0.9% sodium chloride. He's requiring deep sedation dependent on low doses of norepi and he's only made 250 cc's of urine output. And the next day, these are what his labs are. Does this patient have acute kidney injury? Show of hands. Okay, good, we'll cover that again. Does AKI impact the hospital outcomes? How should we manage this AKI patient? You don't need to answer that right now but be thinking about it. And does this patient need dialysis right now? Be thinking about that as well. So first of all, acute renal failure definition. We don't call it acute renal failure anymore, we call it acute kidney injury. We're gonna talk about that but please try to avoid using that term. So what is AKI? AKI is a sudden and rapid loss of kidney function lasting less than or equal to seven days. We use something called the KDGO criteria to sort of stage the severity. This isn't so important in clinical practice necessarily to know exactly what stage your patient is in but patients are classified in research studies oftentimes by the peak that they reach or if we're doing like timing of dialysis trials, we may enroll patients at different stages. Does that make sense? So just sort of having a good concept of what these are but most importantly, you don't need a huge change in increase in creatinine or a decrease in urine output for very long to reach the first stage of this, okay? Now, if it goes on more than seven days but less than 90 days, we now call that acute kidney disease and then once you get out to 90 days, we start talking about chronic kidney disease, okay? So someone who's been on dialysis for two months in your unit, I'm not gonna call them end stage until they have been out of the ICU and are still not recovering and that'll oftentimes be for like 90 days. Okay, how often is this? So these are two studies in hospitalized patients and it affects about 20% of hospitalized patients. The vast majority of it is pretty mild when you look at all comers in the hospital. It increases with age, male gender and CKD and it increases your risk of death during that hospitalization compared to patients without AKI by quite a lot and it increases your cost and your length of stay by quite a bit. This is the biggest epidemiologic study in critically ill patients done as a point prevalence study across the entire world basically on a similar day and we're about to do another version of this but the incidence in the ICUs is about 60%. So think about other diagnoses. There aren't many other diagnoses in your ICUs in which 60% of the patients have. In most ICUs, we don't even have 60% of the patients on the vent, okay? So this is really high and you see it's a little bit different because before the vast majority was pretty mild and it went down in free, like the frequency went down as we went up in severity but in the ICU, the preponderance of patients are more severe. For context, we have about almost six million ICU admissions per year in the United States and we have about 3.3 million cases of AKI. Don't get me started about the fact that we only have 20,000 people supposedly dying of CAUTIs from catheters. So this is obviously a much bigger problem than the catheter-associated UTIs and we need to be paying attention to that. About 15% of patients in your ICUs will require dialysis and again, you have an increased risk of mortality. These are the type of patients that tend to get AKI and you can see basically, it's a who's who of ICU patients, right, okay? And most patients in AKI, the cause of death is something related to sepsis and it's not so much the AKI itself which we'll talk about down the line and other complications like fluid overload. All right, what about risk factors and causes? So these are your sort of modifiable and non-modifiable risk factors but basically, if you're in the ICU, you're sick and you have a risk and that's oftentimes not so modifiable. Hyperglycemia is a curious one. We don't really think about hyperglycemia as a nephrotoxin per se. However, if you look at studies that look at AKI prevention, they clearly focus on controlling glucose and in patients whose glucoses are controlled, there's lower rates of AKI so it's a little bit, it's part of the bundle of AKI prevention is good glucose control. Okay, what about causes? So we have your perfusion-mediated which a lot of people like to use the term pre-renal. I don't like that term because it implies for most people that it's only volume depletion but it's perfusion-mediated problems which include things like volume depletion, impaired cardiac output, intra-abdominal hypertension and compartment syndrome, renal artery stenosis thrombosis, renal artery vasoconstriction and then I've highlighted volume overload and volume hypertension. And venous hypertension which is a huge cause of perfusion-mediated AKI in our ICU patients. Intra-renal causes again and then your urinary catheter obstructions and other obstructions. Please don't discount this because I've seen a lot of problems with like foleys getting clogged or patients who you think are urinating who don't have foleys but in fact, it's actually just overflow incontinence and they're actually obstructed or they're having neurogenic bladder from all the opiates you're giving them. So and then there's these sort of mixed ones which are sepsis and the others. But really in reality, it's usually a nexus of all of these various problems. We can't escape nephrotoxins, everyone's inflamed, we have all these hemodynamic perturbations and I'm gonna skip this since for some reason, the font was super weird on that but suffice to say, a lot of organs can conspire to cause AKI and then AKI can conspire to cause a lot of problems with the organs. Okay, so it's a big issue. All right, let's move on. Understand and oops, sorry about that. Let's talk about understanding and describing the diagnosis and managing AKI. All right, so how do we diagnose AKI? So this is a journey throughout the decades. We've just used creatinine. It doesn't matter where the disease actually is originating, we just look at creatinine. It doesn't matter what the problem is, we look at creatinine, what the issue is, it's creatinine or its severity, it's creatinine, right? So imagine if this is true for cancer, strokes, sepsis, heart disease, MI. I'm gonna talk about a few things that are not necessarily gonna be permeating boards yet but I want you to at least have a sense of what's coming down the pipeline because we really have to be doing this better, okay? All right, so urine output in AKI. So there's this whole question about does urine output actually predict AKI? Does it really mean anything or is it only the creatinine? And the reality is, yes, it does. My pointer is not working on the screen here but what I wanna show you here is in this big table, the boxes are the patients who meet creatinine and urine output criteria for the same stage, okay? But if you look across here and you look at patients, for example, who had no AKI by urine output criteria and no AKI by serum creatinine criteria, they had a certain mortality and very few people ended up on dialysis. But if you met criteria by urine output but not by creatinine, you still had a significant increased risk of death and increased needs for dialysis and that sort of holds true across the whole thing. So urine output really is a very important functional marker of how our kidneys are doing, okay? In the interest of time, I'm gonna skip the details of those two things and happy to talk about it afterwards. But suffice to say, if you meet criteria for urine output-based criteria in the absence of creatinine-based criteria, that's still very important. So how do we continue to enhance our diagnosis? Well, let's remember we have urine studies that we can send. Please get a urinalysis. This is of like the mainstay of diagnosis. We need to understand what's the cause of AKI, not just that the patient has AKI and urinalysis is a mainstay of therapy or a mainstay of diagnosis. If you can't look at the sediment yourself and you're concerned, you can ask one of your nephrology friends to do that. Other urine studies, consider checking a random urine sodium and there's a really very limited role for FINA or fractional excretion of urea. If it's even validated, which is questionable, it's only in patients who are oliguric, okay? All right. So serum creatinine, the other problem is serum creatinine is a lagging indicator. The injury has already happened and it usually has happened one to two days before the creatinine starts to go up because it takes time to produce it. And urine output can also be a lagging indicator. So the key is how can we move forward from like here, where we're noticing that the GFR is going down to earlier when potentially we have a chance to modify stuff before the patient's kidneys are damaged. And furthermore, how do we sort of differentiate between patients who have just functional problems versus structural damage versus both, okay? So what about this idea of subclinical AKI? So what about if you have signs of AKI, but your creatinine never goes up? Your urine output's going down or you have other biomarkers. So we have more and more data in the era of developing biomarkers that shows that even if you don't have creatinine going up, but your biomarker positive, you have an increased risk of bad outcomes and death and dialysis and the other sorts of issues. So that's been shown throughout multiple different studies. So where do we stand now? So if we think about the biomarkers that are coming, and there's two that are available in the United States now, NGAL and something called TIM2-IGFBP7. I don't like to use brand names, but that one's impossible to think about without a brand name and that's called Nefrocheck, okay? So there's two that are FDA approved in the United States. There's gonna be more that are coming. So how are we gonna use these? So we're gonna think about now that we have, instead of just do you have AKI or not, biomarkers are gonna factor into this. So you've got your no AKI when your creatinine and your urine output are negative and your biomarker's negative. Then you're gonna have this sort of subclinical or structural AKI whose creatinine hasn't gone up yet, but biomarkers are very positive. Then you're gonna have functional AKI, which is where your creatinine's going up, but your biomarkers are negative. So this could be a situation like the patient's on an ACE inhibitor or the patient's getting diuresed and the creatinine's going up because you're hemoconcentrating the patient down. You'll be able to sort of differentiate that a little bit better because you'll feel more reassured that there is no signs of structural damage. And then you've got your established AKI, which is your biomarker positive and creatinine positive. And we are currently in the process of revising the definition for AKI, but it will almost certainly now be sort of split a little bit more into whether or not you're biomarker positive or biomarker negative. So even though your hospitals probably don't have access to biomarkers yet, once these definitions start coming out, hospitals are gonna have to start to have access to biomarkers. And we're all gonna have to learn together sort of how we're gonna use these and interpret them and bring them into our clinical practice, okay? So hopefully this is the future. Instead of it being all creatinine, we're gonna be able to target things a little bit better and have profiles and panels that will help answer some of our questions better. But biomarkers are not gonna turn a bad clinician into a good clinician. They're gonna hopefully transform thoughtful care into informed care. Make sense? Okay. A couple of other things from a summary perspective. Again, you have to measure and document the urine output. Do not tolerate systems of care in which urine output, for example, is dumped from a Foley catheter and not measured or documented. It's just unacceptable. UA and microscopy is really, really important. Obviously, you need to serially follow serum creatinine. Consider checking biomarkers when they become more available. In select cases, a kidney ultrasound is good, but you do not need it in almost all cases, okay? And then please verify that the Foley is inserted and working properly because I've had it in the vagina or in other places in which they don't generally make urine. Okay. AKI management. All right, so these are my steps because we, believe it or not, we don't have really great, solid data on this is what you must do in order to manage AKI. So these are sort of what I do, and it's taken from guidelines and expert opinion amongst a number of smart people. But the first thing is we need to determine the type and cause. So please don't just say they have AKI. You really do need to understand, is this AKI secondary to ATN? Or is this perfusion-mediated or whatever? Because we need to treat the underlying cause of the AKI in order for the AKI to get better. Stop and minimize and avoid nephrotoxins, and please dose your medications properly. Determine volume status and management. The goal is to optimize, not maximize volume, right? So on that question about should we just continue to resuscitate with LR, I think it would have, should have said resuscitate with LR if clinically appropriate. And then mitigate and avoid AKI complications. And then finally, think about dialysis. So, and then obviously put a plug in for nephrologists. Okay. So let's talk about this. So I just reminded you these are our four things, okay? We have to understand the cause of this. If you can figure this out without your nephrologist, great. If you need your nephrologist to help you figure out what the cause of this is, that's fine too. But we have to understand what the cause of this is because if it's a perfusion problem and we think it's an ATN issue, we're not gonna be fixing the perfusion problem and shockingly the patient's kidneys aren't gonna be getting better. Does that make sense? Okay. Oh, conversely, if it's ATN, then they functionally have had a nephrectomy, right? So keratin is gonna go up every day until the kidneys start to recover. So we're not gonna really care too much about what the keratin is doing every day. It's just gonna tell us not that the kidneys are worse, it's just gonna tell us that they're not better. Okay, so etiology is really important. All right, what about nephrotoxins? Let's first of all talk about that we need to search for alternatives for many of the medications. We actually do have an alternative and there's a ton of drugs that you may not necessarily realize that are nephrotoxins. We have vacillated back and forth in the last 20 years about whether or not Venk is nephrotoxic. It is, 100%, okay? Venk is nephrotoxic, it's not the diluent. It is the Venk, all right? But there's other drugs that we don't think about, like PPIs. PPIs are increasingly recognized to be nephrotoxic. So if they're not on one at home, they're not GI bleeding and having ulcers in your unit, then maybe they should be on an H2 blocker, for example. Use linazolid instead of Venk, for example, if they have a mercenarmonia. Dose-adjust medications appropriately. I remind you that all the formulas that we look at, like MDRD or the CKD-EPI or the Cockroft-Gault, those are not designed nor validated to estimate creatinine clearance or GFR in the setting of acute kidney injury. We don't have any real great validated tools for this. There are monitoring devices and other things coming down the pipeline that may actually, just like your SAT probe is continuously reading your SAT, you may, in the next five years, have continuous meeting GFRs in your units. That would be super helpful on knowing how to do drugs. All right, what about volume status and complications? Okay, so the death from AKI comes from the complications associated with AKI. This was a fantastic study, really important study out of Harvard in the MIMIC database that just looked and used some very fancy statistics to basically correct for these three problems, fluid overload, hyperkalemia, and acidemia. And when you corrected for those, there wasn't as really significant increased risk of death above no AKI in that FAR group, okay? But when you don't correct for those, then the risk of death goes way, way up, which means it's these three things that are driving the deaths, not the fact that the kidneys aren't working. So once the patient has AKI, you need to minimize these issues. And that's what your focus on rounds needs to be, is I need to avoid these problems, and if I'm having trouble with these issues, that's when we need to think about dialysis. Because in this dataset, dialysis improves survival if it was done for these primary three reasons. Okay, so how do we do fluid management? We need to assess the volume. Objective assessments, please don't use intuition any longer. Objective assessments, volume removal strategies need to be discussed, and then we need to monitor the clinical impact. Mechanism of fluid removal, you've got urine, you've got dialysis, and you've got others, which are like NG tubes burning your patient, bleeding your patient, or causing diarrhea. So these are probably no-goes for most of us. We're not gonna probably plan to burn or bleed our patients. So you either can remove it by urine, or dialysis. And we don't have really any data that says dialysis is better, right? We looked at this in heart failure studies, right? If they're urinating, urine is fine, and you should let them do it, even if you're needing diuretics, okay? All right, so that gets to the next busting of kidney myths. Diuretics are not nephrotoxins, okay? They're not, okay? If the kidney is sick, it usually needs help to excrete salt and water, okay? And I was shocking, right? You think that, shocking. There was a, I don't have time, there's a great study in ICU patients that looked at patterns of diuretic use, okay? And in that study, as creatinine got higher, diuretic use went down. That doesn't make any sense, right? They're not nephrotoxic. If the kidneys can't handle it, just, we need to give them help. If they have CKB5 at home, they're not gonna make enough urine on their own. They're gonna be diuretic dependent, right? Okay, what about dialysis? We're gonna end here. When you consult us, I'm at that blue mark. I don't know which direction these patients are going in, if they're gonna recover, if they're not gonna recover. We've had a ton of trials looking at timing of dialysis. Suffice to say, the three largest randomized controlled trials showed no difference. There was one trial in a single center cardiac surgery patient population that suggested early was better. And importantly, in this KIKI2 trial, later was harmful. So a more delayed strategy was harmful. So this was like delayed versus more delayed. And this was like early versus delayed, okay? So when do we start dialysis? We have to sort of think about it that as we go along in time, our complication rate is gonna go up. We're gonna have more things the dialysis needs to fix, but the dialysis is gonna potentially have a harder time fixing it because you've not just had like the mouse leave the barn, but you've had like all the animals leave the barn, okay? And the benefit of dialysis then potentially goes down. And the harm from starting too early when you don't have any complications is potentially there, but as you go further along, the harm isn't there as much. So what is too early for some patients may be too late for others. And I recognize that this is really very difficult to answer, okay? So think about it like when do we have a difference between gap and demand? And if that gap between capacity and demand is increasing, that's when we need to step in, okay? So these are management steps. Don't forget, we need to be thinking upstream about prevention and we need to be thinking downstream about recovery. I don't really have time to talk about that too much today. Just remember that AKI is common. It can often be prevented. Determine a cause of AKI. Decrease the nephrotoxin burden. Use objective tools for volume assessment. Diuretics can be safely used in AKI and non-AKI to return your patients to uvulemia. And the timing of dialysis in AKI remains elusive. And please, please, please, if they have AKI, they need outpatient nephrology follow-up, okay? We're terrible about that as nephrologists in terms of getting our patients back in. All right, that was a lot, okay? So we need to move on to the next topic.

Acute Kidney Injury

ICU

nephrology

diagnosis

management

CRT