Today's topic is acute kidney injury. The session is called Mitigating AKI, The Struggle is Real. You'll understand what that means when you hear their story. You can read the bio for each of our presenters today. In short, this is Patricia or Patty, Danielle and Jacob. And they are from the Ohio State University Wexner Medical Center. So they are a little rock star team. I understand on good authority, because one of the physicians there sits on the steering committee, that Patty is a little bit of a drill sergeant. I'm sure you guys know nothing about that. What? I'm just telling secrets. Anyway. We like it home. That's right. So I'm delighted to welcome them here today. They're going to share their story about this particular quality improvement initiative. It spans years. And they learned a lot until they finally actually appreciated real success. And I think we can all relate to that. So please take a moment to read their bio. I'm going to have Patty come up. And we welcome your questions at the end. And take it away. First let me just say that this whole team is really a rock star. I've been a nurse for a long time. I've worked with a lot of people, including the NCDR. I've been working with the NCDR, working with Connie Anderson for many, many years. I can't say enough good things. So thank you very much for allowing us to come to present our story. Because this is really a novel to be honest with you. But I hope you learn a lot from it, to be truthful. Let me see if I can do this. We have nothing to disclose. Uh-oh. What's that mean? Where is my PowerPoint? Dawn, where are you? Go back. Back. I hit start. Do I need my files? Oh, yeah. We're a rock star group up here. All right. Let's go back. Where's my slides? They're all the same slides, wherever my slideshow is. Oh, there it is, yes, there it is. So our project was to reduce acute kidney injury, and we have been working on this for quite some time. This is our team up here on the screen, and I will tell you that it impacted two rather large heart hospitals. So this team was made up of physicians, nurses, nurse managers, pharmacists, data analysts, et cetera, and Jacob, who's a medical student that we brought on board later. We have nothing to disclose. So apparently you're supposed to scan this code. We do have some polling questions that we think might be of interest. So take a minute here and scan. And obviously you're going to be able to use this to ask questions too. So our first question is, is this a problem at your institution? Yes or no? Yes. I don't see a clock. Is there a clock counting down? Say it again? You have to make it live. I have to make it live? That's not it because that's part of my slide. Any thoughts about how to get the poll going? Go to the next slide? Oh yeah, the next slide. Oh gosh, I'm so glad you guys have already been through this. Man. And you guys just, you guys, this is a struggle, that's right, this is a struggle. Okay, oh my gosh, are you kidding me? That's probably the reason you're here, right? You're all having this issue with us. Okay, let's ask the next question. I don't know if you know this, but what was your percentile the last time you looked at your report? Where are you sitting? 25th, 50th, 75th, 90th, or you don't even remember what that report looks like? We are, by the way, a very fun group from Ohio State, aren't we, Danielle and Jacob? Yes, we are. Okay. Now we got a five percenter there in the 90th percentile, but it looks like the majority of people are in that 50th to 25th percentile. And some people don't know. That might be a problem. Okay, last question. What strategies do you use at your institution to prevent contrast-associated AKI? And for this one, you can select all or any. Okay, we have a lot of people hydrating before the cath, a lot of hydration after the cath. 37% use a left ventricular end-diastolic pressure hydration protocol, 37% had a lower osmolarity contrast and a decent amount have a contrast threshold. Thank you. So this is us, Ohio State. I'm not going to go into all of the things that we actually are involved in, but I do want you to know we're a seven-hospital system, we're located in Central Ohio, and we're the only academic medical center in Central Ohio. We are U.S. News and World and Magnet-recognized, as well as Forbes recognized us. This is our brand-new tower that we're building, and it will open in 2026. It's going to add an additional 820 beds. It is the largest single project ever undertaken at Ohio State at a cost of more than $2 billion. It's huge. So just to start, you know, the mission statement for the ACC, obviously, is to provide information and knowledge and tools, implement quality initiatives, and support patient-centered outcomes and research. And there's my friend up there with his champagne glass saying, we agree. That's our mission, too. We want to make things better for everybody. So we need the ACC, right? We need this registry to be able to do this. Come on. Okay. Our objectives, we are going to discuss the need for perseverance, and I'm telling you, we worked a long time on this. So we need perseverance, and we need to work across the clinical teams. That sounds like a no-brainer, right? But truthfully, that was crucial to this project. We needed a multidisciplinary team approach to reduce this. And we're going to identify two care opportunities designed to mitigate this AKI injury. So you all know the definition of acute kidney injury. When you're looking at the registry, and by the way, my slide actually should say creatinine equal to or greater than 0.3 increase, or 50% relative increase, or new dialysis within five days of the procedure. And patients excluded from this metric if they're already on dialysis. So why is this important? Well, three really big reasons. If patients develop this, they have a longer length of stay. They have higher costs of care, and even that goes out up to a year for these patients. So this is an important thing. This is our performance. If you look at the very top, we started with this graph from 2017 to 2018. And really, this is where we were. We were worse than the 25th percentile when we first started looking at this. We're like, why are we looking like this? And you're going to hear some discussion from our lead pharmacist here a little bit about things that we were actually doing back then that were research-based, and still we just kept kind of being stuck here. So I'm going to turn it over to Danielle. She's our lead pharmacist, and she's going to kind of walk us through where we started and we'll see the struggle. So thank you, Patty. First thing I'd like to start by saying is when we leave it to Patty to kind of keep us real and let us know where we're really at. So this took me back to when I was in college, and I'd always want to go and look at my grade in comparison to everyone else, and it made me shudder and want to figure out, like, how can we be better? So we're going to go through, as we mentioned, this is almost spanning 10 years of hard work to try and get our AKI rates to where we wanted to, which would be in the top 10 percent or 90th percentile. So back in 2014, one of the initial steps that we took is to have a crew resource tool. We recognize that when we're looking at either acute kidney injury risk or bleeding risk, we're not very good at kind of guessing where we're at, and so this resource tool keeps us at the forefront of our mind, and I'll show you a picture of that in a second. We looked at our contrast threshold, so what is the maximum amount of contrast that we really want to try and keep at the forefront of our minds as we're doing these procedures, and then our timeout, which was half of our contrast threshold, and this wasn't the way that we were originally doing it. So when we discussed back then, what timeout do we have for our cath lab, it was often 100 mLs or 200 mLs, and looking at some of our patients that had poor renal function, we had already exceeded the maximum amount of contrast that we would want to give these patients. So it is as easy as one, two, three. So the first screenshot is a pre-cath assessment risk tool that our pre-cath nurses will do, and at the very bottom you'll see we will categorize our patients as low, moderate or high risk. The majority of our inpatients will fall in that moderate to high risk. Our lowest patients are often our outpatient procedures, elective procedures. The next thing that they'll do is they'll calculate their contrast threshold and then half their contrast threshold, and this is our original timeout that we'll do. So for a patient that can only get 40 mLs of contrast because they have a low creatinine clearance, that contrast threshold would be only 20 mLs. This contrast threshold is based off of an article by Guram and colleagues out of Michigan. And the second is a whiteboard that is on the back of all of our doors in our cath lab, and so this information from the IHIS will then be put on that whiteboard so that it's in the forefront of everybody's mind as they're going through the procedure, and it'll have the threshold and the timeout on there. And then last we have our pre-procedure sign out, and this is done at the beginning of the case. And again we'll state our contrast threshold and timeout. We didn't get into data today, but we were able to get our contrast, the amount of contrast our patients received down low like hospitals. And so we knew that we were doing a really good job at that. And so once Patty showed us this data, we're like, we have to think a little bit differently. So in 2018, September of 2018, after we had some discussions with one of our heart failure physicians and Dr. Budulis, who is a director of our cath lab, we decided to look at the Poseidon trial and implement an LVEDP guided hydration protocol. What we were finding is pre-hydration was hard because teams that would not know when their patient was going for a cath didn't want to start hydration because they were worried about the patients getting too much fluids. And so this LVEDP guided protocol, it gave us a specific number to target and be able to give them post-hydration and help people to feel a little bit more comfortable about their volume status and the amount of hydration they were getting. And then in January of 2019, we changed to Iodexanol or Vizipake, which is a low osmolar contrast agent. This was a little bit of a debate because of the cost of it, but when we thought about it, if a patient's having AKI, they're staying longer, ending up with dialysis, these costs may supersede the cost of switching to a lower osmolar contrast agent. So our LVEDP guided protocol, similar in concept to the Poseidon trial. It's not exact, but we have our three different categories, and we have on the right a longer period that they get hydrated if they're staying overnight, and then if they're going home the same day, we'll minimize it to three hours. They'll receive at most 300 mLs per hour of fluids. If their LVEDP is between that 19 to 24 mark, we'll cut that in half, and if it's greater than 25, they can use the 1.5 mLs per kilo per hour with a max of 150 mLs per hour, or they could say no hydration. So the next question was how do we implement this, and we have it in our epic order sets. So after the case, in the post-procedure timeout, we'll have the LVEDP. The first thing that the provider will need to do when they're putting these post-hydration orders, and these are required, so they have to answer this before they can get out of the order set, is to select whether they're inpatient or outpatient. Then they'll choose the option for the patient's weight to make sure that we don't go above that either 150 or 300 mLs per hour, and then they'll choose the option for the correct LVEDP. This actually allows us, from a quality perspective and myself as we're reviewing the orders, to see what the LVEDP is and the options so we can make sure that they're following the protocol. So moving forward, I'm going to hand it back over to Patty for a second. We have lots of great work trying to prevent it, but we could do our best efforts and we still may not be able to prevent AKI, and so Patty brought to our attention that there are risk factors that we need to make sure that we're documenting for these patients. So using the ACC model, there are certain odds ratios depending on some of those comorbidities for the patients, and frankly, as you all know, sometimes physicians don't always document well, do they? That struggle is real sometimes. So what we ended up doing was we kind of went through the charts and we said, you know what, I'm not going to mandate the physician document all of the stuff in the model that predicts AKI or even mortality, but these four things, we need some help. And so we built an EPIC on the post-op notes, some mandatory fields for the physicians to document. And then this is probably the reason Dean thinks I'm a drill sergeant, but the other thing I did was one of the other institutions I worked at, we actually do a quarterly physician report card. So once the NCDR publishes their data, then the physicians get their own individualized report cards that says this is you for each of these metrics. These are all your other buddies. So there's a comparison. And let's face it, they're probably all valedictorians of their class, very competitive, you know, they want to be the very best. So I've had some people ask me, like, do they get mad at me? I go, no, they actually love that data. They want to be the best. So that really helped them quarterly to get a feel for where they were and what kind of things they needed to do to change. I am not good at this. Okay. The post-op note for PCI in EPIC is a hard stop. So once they finish this case and they have to do the post-op note, these are the four things that we told the physicians they needed to do a better job in documenting. First off was the status of the patient going into the procedure. I can read a chart. I think the patient looks pretty stable, but the physician has eyes on that patient. You tell me how stable this patient is, because that is huge in the risk model. If you've not paid attention to the odds ratios, you should. Second thing was heart failure. They were not very good at documenting whether or not the patient was having heart failure, particularly their New York heart class, those kind of things. Whether or not the patient was having cardiovascular instability, and then that frailty score, which was kind of new since 2018. It's like, okay, we need to kind of brush up on that a little bit. So in EPIC, this is what it looks like for the patient. There is status of the patient here. And if the physician hovers over the choices, the actual definition will pop up for them. The NCDR definition will pop up. So there's heart failure. If they say yes, then these things pop up. The New York heart class. Is this new? What type of heart failure is this? The cardiovascular instability. And then what is that frailty score? And this is what I see on the back end. Like at the bottom of the post-op note is the picture. This is where I put the stint in, blah, blah, blah. But at the very bottom, very clearly for any abstractor is the status of the patient, whether or not they have heart failure, the type, et cetera. And if they're having any cardiovascular instability. Danielle? So this is often what I feel like when we're getting these numbers come through. One time I remember Patty sending us something, and I should never look at my phone walking up the stairs, and our number had gone up again. And I told her, thanks for making me almost trip up the stairs. So time's ticking. We implemented the LVDP guided protocol on the changed iodexinol, and then we wanted one of our pharmacy residents took it on as a research project to see, did it make a difference? How are we doing? Did our numbers go down? And then at the same time, we started realizing when some of these results came out, that we need to start doing some nursing chart audits, nursing education, and pull in some of our nursing staff a little bit more closely, and you'll see why in a second. So our research project looked at the timeframe before we changed the LVDP guided protocol, so it was just standard hydration, I guess you would call it, or hospital protocol, even though we didn't really necessarily have one. And then the transition to the lower osmolar contrast agent, and we had four different groups. So not going to go into these numbers on here in detail, but the two things that I want to draw your attention to, we have the four different groups, and then we have the amount of hydration that's ordered, and remember that was driven by our EPIC order sets, and then the amount of fluid that was charted in our electronic medical record. And we noticed when we saw these that there was a very big discrepancy. So it's median and then interquartile range, and so the amount that's ordered should be a lot, and the amount that we were finding charted in a lot of these patients often could even be as low as zero. So that left us scratching our head, and how are we going to even tell if our protocol's working? And then no surprise, our results showed that there was really no difference in the rates of AKI for these patients, and this is not risk-adjusted rates, but again, what's going on here? And left us kind of wondering. There was such a big difference between the amount of fluids that was ordered. We thought we had a great way to kind of drive that in comparison to the amount that was charted. It left us with uncertainty if our protocol was even working, and it took a lot of time and effort to put these in, and we thought we really would make a difference. And then we had to start discussions to make sure that we improve the process of administration of these orders. If it's not documented, the assumption is that it's not being given. And so we set out and kind of looked at our multidisciplinary team. We had quality at the forefront, as we've described, myself working with the interventional cardiologists and the cath lab staff, and then our information technology team to build the order sets, but we really wanted to kind of get into the data and see what's going on here. We wanted to pull in our nursing colleagues to find out why are we not seeing any of this hydration charted in the medication administration record, and then I'm going to hand it over in a few seconds, or we will in a few seconds, to Jacob, our medical student that really helped us to kind of figure out what's going on. So I'm going to hand it back to Patty, who's frustrated by this point. Okay, so this was our problem statement. We all know that we were at the 25th to 50th percentile. After we reduced the volume of contrast, we changed the type of contrast, we put in that LVEDP-guided protocol, but it wasn't really being fully implemented. We had the order set, but patients really weren't getting hydrated. So we set our goal to be greater than 75 percent at least, or equal to, by quarter 3 of 2023, and we really needed to determine why the staff wasn't following the orders. What the heck is going on here? And what were the causal factors? What are the barriers to reaching this? So you see my little ship hitting a big iceberg here, right? This was the Poseidon study that we followed to make that order set, but then this is really how I felt. We were just hitting an iceberg over and over again. So this is me. Every time I open the NCDR report for the longest time, I'm like, why do these numbers not look any different? Do you guys feel like that? Come on. Come on. I saw your numbers at the beginning of the survey. You feel exactly that same way. So come on. Why did we focus on this? Well we're a high volume procedure hospital for this, for PCI, so this was really important. We really didn't want our patients to have acute kidney injury. We had a high rate, and then we just couldn't figure out, like, why did nothing we put into place really change this data? What was going on? So after the study that was done by pharmacy, Ohio State actually does something where they coordinate with the medical students to try to teach them about quality. What a concept. They're going to be attendings. Yes, let's teach them. So in May we had a medical student quality project that started, and what happens is a quality manager like myself will pick a project. The medical student applies, so they have to be interviewed and apply. They're assigned to a project. Unfortunately Jacob got me, and then they have to attend some classes where we actually show them. These are some quality and process improvement tools. This is how we do this, and then you're going to be working, unfortunately, with me. They're going to do a lit search and review the data. They'll have to actually create some sort of a data abstraction tool, and then eventually work with one of the data analysts in our department to analyze the data. So I'm going to hand it over to Jacob. He's the calm one. All right. Hi, I'm Jacob. I have the pleasure of working with Patty and Danielle, and it was really great to work with them on this project. So this was the summer, last summer, May of 2023. So I'll kind of talk about the kind of order to my project. Essentially, I wanted to examine the hydration process from the time that patients get back from the floors and really examine why were patients not getting their fluids on time, why were they not getting the full amount of fluids, and why were they sometimes missing fluids altogether, as kind of Danielle showed in her study. So the first thing I wanted to do was search through the literature for relevant risk factors, so looking through to see what kind of risk factors there were for AKI. Things like hypertension, diabetes, pre-procedure, kidney function, different medications administered during the procedure or before the procedure, so that I could compare kind of how hydration interplays with those variables in a big model at the end. And so then I did a survey collection where I went, well, firstly, I kind of tried to understand the process where I went to the floors after PCI happened and kind of looked at how hydration was given, kind of understand every different aspect to it, and I'll kind of show a model for that on the next slide. And then I wanted to ask nurses specifically, what were the barriers to delivering hydration on time? Why could fluids be missed altogether? And it was really important to do that step because if we don't ask the people administering the fluids, then it's going to be difficult for us to really get, you know, get any true idea of what's going on. And then we did a data pool from the NCDR registry, looking at those variables that I mentioned, and then we did, I did a data collection where I went through charts and kind of looked at hydration statistics that I'm going to show in the next slides. We did a univariate analysis where we saw whether these variables were significant towards the development of AKI within a univariate model and then collected all the variables together and put it into a multivariate model, which I'm going to show. So this is a fishbone analysis. I'm not going to read every single piece here, but basically this shows that there are a lot of moving pieces to hydration. I think when I first walked in, I thought it shouldn't be this hard. We can just get fluids after procedure. And I didn't realize all the moving pieces that are actually incorporated into making this process work. So we have the people, we have physicians that are going to order the hydration, pharmacy verification, training of the nurses, shift handoffs. There's different equipment involved, like our smart pumps, bag availability, and really all these things were things we needed to look at and understand in order to really streamline the process of hydration and understand where exactly our teams were getting caught up. So this is our nursing survey results. So we had 26 nurses fill out the staff that were kind of on the floors. And the big number, the big problem you can see is that there's something where there is an uncertainty or an inability to sign and release held orders. So I'm going to explain this in the next couple of slides. But basically this was a huge thing that is kind of a process within Epic that we needed to modify in order to make sure that hydration was being optimized. There are shift changes during the course of hydration. So if shift changes were happening, staff may terminate hydration early or they may start it late or save it for the next shift. And that could kind of make a difference in whether fluids were delivered on time. There'd be physician delays in ordering hydration. And sometimes there'd be uncertainty in whether to deliver hydration at all because staff were worried about fluid overload within patients. And then could be delays in verifying hydration and then end of procedure transport and availability of staff were kind of some other concerns that were mentioned. So this is kind of summarizing what I talked about. So kind of one thing I mentioned was that there's fear of fluid overload in patients, the order sign and release process, which I'm going to talk about next. One important note is that within our system, the MAR is set for six hours of IV fluids. So if a physician orders fluids at 1 p.m., it's going to be slated to come off at 7 p.m. And if fluids are delayed, so there's delays in getting the patient back from the lab or the nursing staff is busy and can't deliver the fluids exactly on time, the fluids are still going to be slated to come off at 7. So if you only start the fluids at 2.30, that's only going to be around four and a half hours of fluids. So you can see that maybe that would be a concern for us. So we talked with our nursing cath manager to kind of figure out how to improve this process. And basically what she told us is that if patients weren't transferred into the cath lab through Epic, then when orders were placed, they would kind of show this screen for staff. So you can see at the top, there's an area where they'd have to launch the release orders, release orders navigator in order to access the fluid orders that had been placed. And nothing really shows up in signed and held orders. And staff said this made it a lot harder when they're either new staff or they're really busy on the floors to be able to access these orders in a timely manner. And it could lead to delays in getting fluids administered. And sometimes it would be missed altogether just because if patients weren't transferred in the cath lab, it would kind of show this. But if they are transferred within Epic, then it would populate in the right place and really draw staff's attention to it. So I kind of mentioned the chart review I did. And this is the patient population I looked at. So quarter three of 2021 to quarter two of 2022. So we had about 157 patients, 100 that didn't have AKI and 57 that did. And importantly, only 93 of these 157 actually had hydration order undocumented. As Danielle mentioned, sometimes we don't give fluids to patients right after PCI because their fluid status is pretty overloaded. And so obviously, that can lead to AKI, but we have to manage the safety of the patient. In those circumstances. So then we pulled data from the ACC and CR cath PCI and electronic medical record. And then we did a multivariate model data analysis. So these are kind of some of the baseline characteristics. Just going to show this. So with the variables that I ended up picking for my multivariate and univariate study, they're about 35 to 40 based on the literature that we reviewed. These are kind of some of the major ones that are prevalent in the literature. You can see pre-procedure creatinine is a lot higher in those patients that did develop AKI expectedly. So we want to ask, how do these barriers affect AKI rates? So here we have the average percentage of physician-ordered hydration delivered. That's kind of a mouthful. But basically, I wanted to look at if a patient only got three hours of fluids when six hours was ordered, that would be 50% of hydration delivered. And interestingly, we actually didn't see a huge difference in whether patients developed AKI or not, at least within a univariate model, about 85% of the fluids were delivered in patients that didn't develop AKI, and 83% in those that did develop AKI. So we were kind of thinking that that AKI value would be a lot lower. But we did see a pretty big difference in the time from contrast to hydration in those patients that did develop AKI and those that didn't. So you can see the median number. In the 100 patients that didn't develop AKI, there's about 33 minutes from contrast to hydration on average. But in the patients that did develop AKI, you can see there's about 81 minutes. So pretty significant difference in the timing between contrast and hydration. And so that gave us hope that that is an area we could improve and potentially would have an impact in whether patients developed AKI or not. And then this is the amount that factors in. A lot of literature has shown that the amount of hydration also matters. And the patients that did develop AKI got a lot less hydration on average than those that didn't develop AKI. Some of this may be related to efficiency and protocols, but a lot of this also is likely due to patients receiving less fluids in the first place because their LVEDP status was overloaded. It was slightly overloaded, as Daniel mentioned, and that leads to them getting less fluid for that reason. And so this is kind of the significant predictors of AKI post-PCI with the multivariate model. So we pooled all our variables together and kind of wanted to see which variables within this study were impactful in whether patients developed AKI or not. And so the first variable we looked at was post-procedure hemoglobin. For every point of hemoglobin, this was protective of whether patients developed AKI or not. Importantly, hydration being ordered and documented was significant in this model. So whether patients developed AKI or not was protective. This kind of backed up in the literature, but it's important for us to see this in our own studies and kind of reaffirm to us the importance of hydration being delivered and ordered and documented to be protective for our patients. And within this, the time from contrast hydration and the percentage delivered weren't significant in this multivariate model, but we're currently working on a much larger data set because our data set was admittedly somewhat small to kind of see if there's a role for improving specific hydration metrics, not just hydration as a whole. We're kind of working on that data set right now. And then the age for every year of age a person is, their increased risk of developing AKI, administration of ticagrelor procedurally, a TIMI flow of grade 1 versus a grade of 0. So if a patient had a grade of 1 versus a grade of 0, that would mean there was increased flow and that would be protective of whether patients developed AKI or not. And then, as I kind of mentioned, the pre-procedure creatinine is going to be pretty important for looking at a patient's baseline function and seeing their responsiveness to the contrast. I'll hand it off back to Patti. I'm going to go back one slide because when Jacob and I looked at these risk factors, right, as part of a teaching thing, I said, okay, what can we control, what can't we control? I can control hydration, maybe the medication the patient's getting. There's a lot on here. You can't really change about the patient, correct? So that was a really important thing to kind of take a look at and say, all right, I can only control what we can control. As part of our study, we did say this was a relatively small sample size, but it gave us some pretty good insight into what was actually happening every day in the cath lab as well as on the units when they were trying to recover. We did have some time constraints. I only have the medical student for so long and they have to move on, but he did a really good job and he gave us a lot of information and he's actually working on some data that hopefully we might be able to publish at some point. We always have, you know, resource limitations. This was a big multidisciplinary team, but it took a lot of people, a lot of resources to kind of figure this out. And then we really needed to understand the technology. That whole piece in Epic where it was releasing the orders, signing and holding, releasing the orders, getting it started, the MAR has this stop time on there. How does the smart pump interplay with all of this? There was a whole lot of technology that kind of went on to trying to figure out like what is the cause of this? So our action plan obviously is to optimize the implementation of this hydration protocol and we really need to go back to nursing to explain to them that I know you know what the ejection fraction of this patient is, but the physician did a measurement in the cath lab, the left ventricular end diastolic pressure that tells them whether or not the patient needs fluid or doesn't need fluid. So they really needed to understand that. We did talk about the effective, timely and effective order release process. The one other thing we did that we noticed, it was part of something Jacobs observed. We used to hang a 500cc bag coming out of the cath lab and if the patient needed let's say 650, one of the nurses made a statement to him, well I'm not going to hang another bag because I'd just be wasting the rest of that bag. Wait a minute, they need the rest of that. So we changed that. We said okay, every patient that needs hydration is going to come out of the cath lab with a liter bag, it's not started yet, it's hanging there for two reasons. It's a visual cue for the nurse to start hydration and that larger bag decreased that chance of stopping that IV or that fluid prematurely. We did have to improve communication handoff. Danielle worked with the team, the physicians, to add to the hydration orders, the words, to prevent contrast-associated kidney injury, do this. And then I kind of needed to refine what I was doing I think in my role. I had quarterly meetings with our team, the multidisciplinary team. They saw the data. I gave it to them. I'm assuming they gave it to their peers. They gave it to the nurses. Well, then I realized that maybe it wasn't trickling down. Maybe I needed to do a better job of communicating what our rates are to this big multidisciplinary team. The other thing I put into play was more of a concurrent quality review. So any patient who didn't get six hours of hydration who did result in acute kidney injury, I have this nice little Excel sheet with all the time on here. This is the patient, time IV started out of the cath lab, how much they got, et cetera, et cetera. And I meet with those nurses on that floor to say, what barrier did we have? I'm not trying to be judgmental, but tell me, what happened here that this patient didn't get this full amount? So it's more of a concurrent process so they can remember, oh yeah, I remember that guy the other day. This is what happened. And then we are also looking at a larger patient population so we can just have a little bit better data analysis. This is the post-procedure checklist now. And you see in red not only the amount of contrast that was used, but specifically pointing towards the LVEDP and to hydrate the patient accordingly. And so I'm not going to go through all this. This is just from 2014, 10 years later. This is the process we went through. This is why my hair was standing up like crazy. But we had to do all of this in order to get to a better success rate. And this is where we are. And so, oops, hey, you're not supposed to do that. What happened here? So you can see, I was trying to point over here, here we go. This is the 90th percentile, and we were way above for both of those hospitals. And then right about here we got more people involved. There was a change in the risk model. It went from risk-adjusted to risk-standardized. Not sure how much that played into it. There's Jacob's involvement here, and Danielle and the pharmacy got involved here, and nursing got involved here. But we have been at the 90th percentile for the last six quarters once we kind of tightened everything up, understood what was going on. So we thought that the protocol we wrote didn't appear to influence the rates. But the one takeaway here is if hydration is ordered, it's got to be administered. We did need further data analysis to figure out why we weren't getting full hydration. And then we did that whole documentation piece, a hard stop, where the physicians had to document this is the odds ratio, the patient has these comorbidities, status of the patient. So that was really important to getting an accurate piece of information from the NCDR. And then I actually summarize these cases. We do a mortality and morbidity review. Our team gets together, and every single AKI is reviewed at that meeting and then presented by the fellows and the attendings. And I look for all these variables. Their creatinine pre and post, what kind of medications did they get, how much contrast did they get, what was their LVEDP, when did the procedure end, and then when did the IV start, stop, what was the total volume to that? So again, I said, we improved beyond our goal. Our goal was 75%. And when I saw the 90th percentile for the first time, I think I probably caused Danielle to fall up another set of stairs. The whole team, I'm like, oh my God, oh my God, oh my God. And then we just kept it for six straight consecutive quarters. That is absolutely huge. So there are two opportunities. You have to develop and implement the protocol, implement, make sure the patients really get that full hydration. And by the way, we went back to nursing, we looked over the last two years of data from the NCDR, and we said, guys, not a single patient had heart failure after the procedure. They were so worried that those low ejection fractions, they would develop heart failure. And we said, stop. It's okay. The docs really are measuring this. This is accurate for the patient. Please give the full hydration. What happened there? We'd like to publish. We're kind of talking about that. And then, of course, we wanted to present here. Yay! So. We really appreciate you allowing us to present. I do have a poster over here, so I'll be manning that tomorrow if you have further questions as well. My take-home message is don't give up. Don't ever give up. And obviously, go Bucs. Okay. So I have some questions here. Can you all see the questions in the app? And you can vote them up. I just want to make sure they're visual. What's that? You only see one. Okay. So. Okay. Are you seeing more now? Okay, great. Sorry. I was so busy listening. I was not approving these things. So sorry about that. All right. Let's go to some of these great questions. The first one is, did you say that the hydration in the order sets is mandatory on both the electives and inpatients? Yeah. So we do. It's a required field. And there's no way for them to sign out of the order set without filling out whether the patient needs hydration or not. So there is an option that says no hydration needed. And they can put in a reason if they wanted to. But they cannot sign the order set. It's a big red stop sign unless they fill those out. And that's for elective. It is for the cases that we have that are inpatient. It doesn't matter. It's a PCI order set. So that's a required field. And how did you convince your cath lab physicians to do an LVEDP on all patients? So there are patients that don't get the LVEDP if there's specific reasons that they're just not able to do an aortic stenosis or something like that. But Dr. Bedoulis, I think this is where it really became instrumental to work with him as the cath lab director to kind of answer the questions for the providers that may not want to have an LVEDP done or didn't think that it was appropriate. And I think the one thing is that we've all stood behind any of the different things that we've implemented and been able to speak to it. And I think that's how we got all of the physicians to be on board with it. And then giving them feedback if they don't have that LVEDP available so that they're able to follow up on it and change their practice. Remember I give them a scorecard so they know what their AKI rate is. And what is the role of ticagrelor versus other P2Y12 inhibitors? This is a good question. We were laughing about this because there's a lot of speculation on ticagrelor and rates of AKI. In the PLATO study, it does show an increase in serum creatinine in comparison to clopidogrel. We did venture out to see if that was something we could look at, but our patient population is too small. So I do see this as maybe we were thinking a potential research opportunity to look at something in a larger scale to see if that really does have an impact. Is it just an increase in serum creatinine or are there actual clinical outcomes that are impacted by that? We have not changed our practice because of it, but it is something that continues to show up in all the data that Patty and Jacob show us. I think we were actually talking about this in one of our meetings that possibly, hint, hint, the NCDR would like to do a larger process improvement initiative around this since look at all the people who said this was an issue. So if we had a larger patient population to study ticagrelor, say for example, we might be able to pull out what is the impact of that, if any. It sounds like a research on the data itself because we're capturing it. So put in your RPA, Patty, and get busy. Oh, I better write that down. Okay. What measures were implemented to decrease contrast volume or dose? What was effective? And lastly, congrats on your rankings. Okay. Did you get that? So the contrast volume and amount, it was basically just having it in their mind. The pre-cath nurse has calculated it, so it's just having it up there, and they know what it is, and then having that time out. And what that has resulted in, whether this is actually good or not, is sometimes we'll have staged procedures where we'll stop and then bring the patient back to the lab after they've had some time to get hydration, if they've gotten a large contrast volume in comparison to what they're supposed to. And then we don't always do an LV gram on the patients because that does give a considerable amount of volume. And so if there's a patient that has a very low contrast threshold, we may not do that. So there's some things that we've been able to change based on that. But it's really getting that out there and making them aware of it, instead of using some of those arbitrary cutoffs that we were using previously. And then a scorecard, because they get to see their volume. That's part of the score. Yep. All right. A question about the risk-adjusted or risk-standardized kidney metric. It says the ACC NCDR records acute kidney injury if a creatinine baseline is 1, drops to 0.7, that's a change of 0.3, right, and then rises to 1.1. Is this correct? It is correct. And why do we do this? Nephrology would say it's still a normal kidney function. It is normal. But nephrology would also say it wasn't normal for the patient and you have damaged the kidneys. And the kidneys are reacting to the contrast they got. So it's the concept of that tiny, tiny change represents some kidney injury. It's just like having some troponin positive. You still have myocardial damage with it, right. So it's that idea and it's that focus that the registry has on impacting or improving patient care. This is not the only registry that measures that. In the vascular surgery registry they measure creatinine pre and post. Like what is that creatinine right before the procedure and post? And if there's a rise in creatinine in that particular procedure it's considered kidney injury. So. Thank you. You're welcome. And I think that was one of the education points that we had. Because if you look, some data suggests a 0.5 gram per deciliter increase or 50%. Some it's 25%. There's lots of different definitions. So making sure that our providers were aware of exactly what we were measuring so that they weren't thinking, oh, it didn't go up by 0.5, I'm good. They knew that it was 0.3. Right. All right. Given the advancements we have seen with imaging, how often are you dipping into the LV and how are your rates for those patients you don't use this protocol for? Good question. Yeah. That might be a struggle for us to answer. Yeah. I think the challenge that we have with that right now is since we weren't getting hydration we can't accurately look at that. And then educating them that if they're not doing an LVEDP to like maybe have some reason why because we went back retrospectively and tried to figure that out and we weren't always able to. And then giving them the scorecards so they know that it is important and it may make a difference once we get the hydration, which I think now is six quarters worth of data. We have a little bit more weight instead of numbers not changing and trying to tell them to do something. Honestly, when I review charts there's not a lot of charts that don't have an LVEDP. There are some, but the majority of them do. For PCI order sets with special stops, the comment is they don't get used. What strategies did you use to make that change? In other words, there are some personal order sets that can be used. Maybe it has to do with the EHR. You all are on Epic. So maybe how does that work for you all? Can they circumvent that? Let's ask that. They could. I don't think they do. I don't think they know how. So I'm trying to think of the right way to say that, but they can customize. We do have a process when they're updated to make sure that our IT people help to tell us who's customized it and then get those updated. But I think they know they'll get phone calls if there's orders that are missing or changed and we're watching it closely. So it's just not worth it for them to do it. And I think the results also help to speak to the order set and if they use it they'll get the outcomes that they want. Exactly. It's the right thing to do. So they tend to do it. »» Can you share the concurrent spreadsheets you reviewed with the nurses? »» Oh, I might if I remove the patient identifier, yeah, I can share that. I don't know who sent me that. »» So you're at the poster station, right? »» Yeah, but it's not on the poster. »» No, no. But what I meant is they could meet up with you there. »» Oh yeah, meet up with me at the poster. Good answer. »» Okay. All right. And did the patient population include high acuity such as STEMI's? »» Absolutely. It included everybody, salvage patients, elective patients, it made no difference. And personally I thought when Jacob was going to do his data analysis that, you know, surely the patient's in shock, you know, would be one of the variables. But yeah, everybody was in the mix. »» All right. More on LVEDP. Was it for pre-end post-hydration? And is the source of LVEDP from the cath lab procedure or is it coming from pre-procedure echo? »» The LVEDP is done at the end of the, well, it's done during the case. So I would say we have pretty much abandoned pre-procedure hydration for the most part because we got a lot of resistance to doing it. And it also got confusing because it just say normal saline and what we were finding and why we added that comment to the post-procedure order is that if it just says normal saline people don't know what it's for and they feel inclined to stop it, especially if the patient has PO intake. So we just stick with the LVEDP from the case and then the post-procedure hydration order that now tells them it's to prevent AKI. So hopefully they don't just stop it. »» All right. And I hope I say this one right. It's how did you overcome the barriers to expediting sign and held orders being re-released? »» I think it's just important when you continue to show the data and your reports from the NCDR. So taking that information back to nursing to say, or physicians, look, we need to start this earlier. And when we identify a case, they get something within about the first week of the case. This is the case that happened. This is why it happened. And I think it's that constant feedback of information is the only thing I can think to tell you. And I don't know if you have those kind of resources. I have a couple of abstractors that abstract and I'm kind of like over that registry so I look at the data and then work to do process improvement. So that gives me the leeway to focus on the reports that come out, look at those patients where we didn't meet something, feed it back to the multidisciplinary team or the units themselves. So that's kind of my role is to oversee. »» I can add something. »» Go ahead. »» Oh, here. »» Oh, sorry. Yeah. I think kind of partnering with the nursing cath manager was really helpful for that process specifically because we noticed, as I kind of showed on one of my slides, that the orders were being missed a lot of the times. And I think that might be kind of the issue you're bringing up with the question. And they're kind of now having worked with them and telling them how important that is showing the data. They're auditing that process and 99% of the time they're saying that they are doing the proper steps in Epic to make sure that these orders aren't missed and they're showing up how they should be, which wasn't the case before. They're being missed because they were kind of all over the place. »» That's true. I forgot they were auditing like how often they release the orders. Yeah. Thank you, Jake. »» You can steal this anytime. Are scheduled outpatients instructed to hold specific medications prior to their procedure that may affect kidney function? »» That is a great question because we just discussed this at our M&M about two weeks ago. One of the things that we saw in here is that ARBs are associated possibly with a higher rate of AKI and we're not having patients hold those medications before they come in for outpatients. For inpatients, if there's a pharmacist reviewing the profile, we'll often have them hold the ACE inhibitor or ARB. So I would say that is something that we're trying to refine our process for that and be consistent with what we're telling them to hold and when we're telling them to hold it. »» And do all patients get this fluid therapy treatment or specifically those that have CKD? »» If they're eligible, all patients get this. So if they already are on dialysis and come in, they're supposed to get fluid hydration as well. »» And what are your thoughts on the effects of multi-vessel PCI intervention at the same time, I'm reading this wrong, what are your thoughts on effects of multi-vessel interventions at the same time on AKI, like if somebody has a multi-vessel disease? »» Yeah. We get a lot of multi-vessel disease. They're supposed to be on the protocol if the LVEDP is in the right range. »» But you had also mentioned that you are doing some more staged procedures possibly as a result. »» True. »» We definitely see the staged procedures when we get those patients that have multi-vessel disease and their contrast volumes get up, especially if they have poor renal function. So we've had a lot of discussions around is that the right approach if we go a little bit over what is more detrimental for the patient, staged PCI versus controlling the amount of contrast. And I don't know that we have enough data to specifically say what the difference is. »» Okay, Jacob, this is for you. Can you explain more on the effect of post-procedure hemoglobin on AKI based on your LIT review? »» Yeah. So I think kind of what I saw was that the higher the hemoglobin, I think if the hemoglobin was lower, I think there's a higher concentration of contrast within the fluid. And so that effect could be more nephrotoxic versus if there is less blood loss during the procedure and consequently a higher hemoglobin, that could maybe dilute out the effect of some of the nephrotoxic agent. »» Was there physician pushback on using lower molecular contrast? Our doctors say they can't see the angiogram as clearly with it. »» Well, you know, we involved the kidney specialist, the nephrologist before we made this decision. And I honestly don't remember any pushback whatsoever. »» The only group that I heard say with the imaging would be the vascular interventions. But for our cath interventions our physicians didn't seem to have any concerns with the images and the lower osmolar contrast agent. »» We were primarily using ISOV before that. So I don't know if there's such a small difference between the two that they were already used to it. »» All right. And I missed a question at the top. Did this inversely affect your bleed metric? So did you have hemodilution that impacted looking like there was a 4-gram drop? »» No. I'll be honest with you. I followed the bleeding metric. And to be truthful with you, it's not that we have a lot of access site bleeds. We do tend sometimes to give PAC-RED blood cells. But it didn't drop that hemoglobin like greater than 3. »» Okay. Items you've made mandatory in your post-op note with it being documented after, can you see this for pre-procedure documentation abstractors, example, if heart failure isn't documented until after PCI in the post-op note? Okay. I think what they're going for here is the post-op note that you implemented are things that are actually pre-procedure, right? And the definition is there. But can you explain how you educated your physicians and that they're not being influenced to document differently after the procedure? »» I think what drove that change actually was the risk-adjusted at the time, risk-adjusted mortality rate. And so we pretty much said if you don't document these things, your rate isn't going to be what you want it to be. Like I said before, I didn't have them document all 12 items, just the four that we really didn't seem to do well on. And because I didn't ask for the world, the thing that I've learned about that I've talked to Jacob about, when you're working in process improvement, if I ask a physician to do something, it better take less steps. If it takes more steps for them to do it, they're not going to do it. So I have learned over the years how to deliver my message to get the end result. For nursing it's slightly different. Nurses really, really want to do the right thing for the patient. But if I were to say, hey, look, we might make more money if we do this, they won't ever talk to me again. But it kind of is true. When you're doing quality, it pays off. Patients don't stay as long. That kind of strategy is really true. But delivering the message makes a huge difference in getting people on board. »» I think the other piece of that too is you were having to email after the fact. So once we knew that they could document it up front and it would avoid emails afterwards to correct their work, that also was a huge driver. »» Yeah. They don't get too many emails from me anymore on the front end of like what. But yes, those are pre-procedure respecters. »» Kim Lavin, can you come up to that mic right there? I'm going to read a question. I'm going to answer it. But I might need your help. All right. This is the question. We have a 94% same-day discharge PCI rate. This lowers the total number of patients and thus the AKI denominator population, remember that metric is excluding same-day discharge. So the point of this question is, you know, they're discharging a high rate of patients and overall they feel the metric would better demonstrate quality of care if it included all patients. Now this was a specific question that was addressed by the workgroup that looked at this. And they analyzed in detail the number of same-day discharge and how hospitals perform by including those and not including those. And essentially there is no difference. But there's probably a better way to say that, which is why Kim is standing there. So Kim, can you elaborate a little bit more or just say it in a different way so that »» Basically we did a lot of evaluation of that during the risk model development because that was a concern. However, it just didn't show in the data, right? It didn't say when we looked at high facilities with a higher rate of same-day discharge compared to lower, there was no difference in the overall AKI rates. »» Yeah. And I'll just say that, you know, when we pull together workgroups, we're pulling in physicians from all over the country. And in fact, one of the physicians had emailed us within a year of that workgroup being pulled together and this was their concern for his healthcare system. And he put together a very nice argument for it, et cetera. So we actually extended an invitation to him. And he was on the committee. And when it was all over and he was part of the workgroup that agreed that it was of no value essentially to include same-day discharge but to focus on those patients that are having a post-procedure grant. And he was convinced as well. So that was satisfying. But yes. No, it's a valid question. It's asked frequently of the registry. »» We've had the same discussion. Like if we drew creatinines on everybody and made our denominator bigger, would it make a difference? So that's why I was kind of laughing. That's come up probably three or four times. »» All right. We're almost at time. We have one minute left. So the last question is for Danielle. Is there a different protocol for high-risk patients? »» There's no different protocol. We previously had tried to stratify. The more complex the process gets, the less likely they are to follow it. So we've tried to keep it simple and do the same protocol for everyone. »» All right. Well, thank you very much to this team. They're amazing. »» Thank you.

acute kidney injury

quality improvement

hydration protocol

PCI procedures

LVEDP

nursing education

electronic health record

multidisciplinary team

performance metrics

healthcare quality

contrast procedures

risk assessment

Ohio State University

process barriers

documentation practices

medical center