Hello and welcome to the ACC Quality Summit. Today we will be discussing how quality improvement is connecting with our clinical team at the Ohio State Wexner Medical Center. I am Patricia Blake, the Senior Cardiovascular Quality Manager, and I will be discussing several projects along with Sandy Campbell, who is the Program Manager for EP and Vascular. Today I'll be discussing how we use the ACC NCDR registries to carry out some process improvement opportunities by reviewing some of that data. The data helped our teams to identify these areas, particularly in three different areas over the last few years. We use the Action Registry, which is now the chest pain slash MI registry, to improve our door to ECG time for STEMI patients. We have used the data in our cath PCI registry to reduce bleeding, to improve acute kidney injury, and also to improve the risk-adjusted mortality metric. And in the EP registry we've used the data to reduce lead dislodgements, improve documentation of the required LAAO elements, and to reduce pneumothorax. These are our objectives for today's discussion. We will identify members for our multidisciplinary quality teams, we will discuss team building strategies, we will review the NCDR resources that helped to jumpstart some of our quality initiatives, and we will provide examples of completed process improvement projects. So of course I'm an Ohio State Buckeye fan, so I had to actually put this slide in here, but really this would work no matter where you worked. You have to have team members who are highly motivated, definitely involved, and all aiming towards the same goal, working together side-by-side. I like this picture because as you can see they're a close-knit group, they're all striving for the very best, and they all want to be number one. So of course anybody who touches your patient in this initiative should be included as part of your team. So what is the importance of having a multidisciplinary approach? People on the team that are involved bring a representation of different viewpoints. Everybody has their different strengths depending on what their discipline is. It really is okay to disagree, but the strength of this is that those involved are going to look for evidence-based medicine to support any protocol that you might want to build or put into place. When everybody is involved in the team, there's buy-in across the health care system, and like Ohio State and probably like many of the systems or hospitals you work in now, you may be part of a system. So having buy-in across the system to standardize the care is important. You definitely need administrators that help remove barriers as well so you can be successful. There's a reason you want to engage stakeholders. For one thing, that means that there's balanced representation amongst all groups. So for example, at our mortality and morbidity conferences where we talk about complications and so forth, that used to only involve physicians, but many years ago our leader decided that we needed other people at the table, nursing, pharmacy, quality staff, so that we really get all angles of what happened in the case and how we can make improvements. That group really sets a lot of goals for the PCI team, and those goals are tracked all year. You need stakeholder understanding and acceptance of roles. You need somebody that can facilitate the discussions that's neutral. This person really should have a culture of non-judgment. They should know how to create and support an organizational culture of quality improvement. The stakeholders, it's important for them to be connected. We want the care to be standardized so that no matter where any staff person or physician is working, the process is always going to be the same. And you need sustained stakeholder engagements. So there's updated monthly team meetings where we talk about the progress that we're making on our PI projects, and of course we take it to our leadership team meetings. So I'd like to take a few minutes just to talk about some team building strategies that's been successful for me. I learned a long time ago that everybody looks at a project through their own lenses, and so it's important for you to understand the what's in it for me or the WFM. That's an old term. I'm an old nurse. But for physicians, they of course focus on the quality of care, but they definitely focus on how much time. If I'm going to ask them to do something in a team meeting, the first thing they're going to think is, how much time is this going to take out of my day to accomplish this? I am well aware in this current day of electronic health records that they count the number of clicks, and I bet you've heard that too. They also want to know if what we're asking them to do is evidence-based. For nurses, the strategy with them, pretty much as always, is this the right thing to do for the patient? Are we providing quality of care? What if this were my parent or my family member? And from a quality department, we review the data and we try to assist in the creation of a better process with the teams so that everybody can meet the end objective. So I like this quote by Edward Deming, in God we trust, all others must bring data. If you know anything at all about Deming, he worked with Walter Schuart to develop the Statistical Process Control. That eventually became the Plan, Do, Check, Act, or PDCA, that we sometimes use in medicine today. We need data. The data needs to be shared at the leadership level, and in our hospital we have the committee called the Cardiovascular Quality Committee. However, one thing that I do, that maybe some of you do as well, is I always try to share my data prior to the meeting for two reasons. One, I want to make sure that people looking at the data believe that it's accurate, and secondly, so that people are not surprised by the data. You don't want to be in a meeting where an administrator asks a nurse manager or physician about the data and they're not prepared to answer the questions. So I try to share the information ahead of the meeting. We send out physician report cards quarterly, so every time there's an NCDR report, the physicians get their own report cards, and then we take the information to our monthly operation and quality team meetings. So anytime there's data that's sent out by the NCDR, we review it. We also look for any monthly trends, and the reason we in quality do this, we feel like we're the conscious of the team. We are there to communicate any negative trends, and sometimes good trends, but definitely any red flags. The ACC NCDR over the years has provided some campaigns that we really feel have been critical to improving patient outcomes. We, along with many of you, belong to the Door-to-Balloon initiative, which obviously reduced mortality rates for STEMI patients. Surviving MI, the Hospital-to-Home, CUN7, those were campaigns that helped us to streamline the discharge-to-home and automatically set up an appointment with the physician within the first week. Patient navigator programs, we have both AMI and heart failure navigators at our institution, and more recently, reducing the risk for PCI bleeding. I really enjoy the PCI bleed listserv, where anybody can ask a question to the group on an email, and then different hospitals from across the country share or provide what best practice is at their institution. Last summer, there was a question raised about how to improve the radial approach by physicians, how to get physicians more engaged at their hospital. And so we were asked in the summer of June last year if we would present a webinar about how we got our physicians to be engaged in the radial approach. We do use an ePRISM risk tool that's integrated in our electronic health record. We use EPIC, and so when the pre-cath nurses are completing information about the patient, there's an automatic, the tool automatically generates whether the patient is low, moderate, or high risk for bleeding. And of course, we look through the toolkit and utilize or steal, shamelessly, anything anybody puts in there to adapt at our organization as well. So I'd like to take a few minutes to highlight three projects that we've done in our organization that have actually impacted three different areas. And the first one I'm going to talk about actually originated from the action which became the chest pain MI registry. These projects definitely were supported by committees that consisted of physicians, nurse managers, nurses, quality team, pharmacists, techs, EPIC clinical applications team, as well as imaging informatics team. The first example we will discuss today is how we use the chest pain MI, formerly the action registry, in order to improve our door to ECG time. And the national benchmark, as we all know, is to try to obtain the ECG from arrival to actually showing it to the physician within 10 minutes. And initially, many years ago, we were only meeting this metric pretty consistently 60% of the time. So our leadership met with us and raised a question at one of the meetings. Can we take a look at that? How can we improve that? At the same time that we had that meeting with the leadership of our hospital, OSU's emergency department had actually doubled in size. It is now the size of two football fields and that happened in 2016. And the process at that time, which is the process for a lot of hospitals, is that once you obtain the ECG, you have to find a physician to show it to and they have to acknowledge it by either dating, timing, initialing it, or something like that. The emergency department at OSU does about 255,000 ECGs per year. We did a time study, a baseline time study, prior to our intervention, which actually indicated that staff spent 9.5 hours of every 24 hours just walking to find a physician. And we recognized that this actually could potentially be increasing our wait times in the emergency department as well. One of the tools that we used in our team was a flowchart. I'm not sure how many of you are familiar with that, but on the left-hand side of the chart would be my customers. The patient, registration, triage nurses, PCAs, physicians, etc. And this should be a step-by-step process where the patient comes in and you do this, followed by this, followed by this. The flowchart would help to identify if there might be any decisions along that step-by-step process. Decisions are highlighted in the flowchart by diamonds. So anywhere where you see a diamond, somebody had to think and make a decision. For example, we were walking down the hallway and we were gonna find the physician and there was a Y in the hallway and the staff stopped and then she decided to go down the left hallway. And I said, why did you do that? And she said, because that physician is nice. Now I'm not trying to make any physicians out there upset, but we know that some days physicians might not be in a great mood and some days they're in a great mood. But what does that do to the workload of the nice physician for one thing? So that was a process that we said, hmm, let's take a closer look at that. We did some baseline measurements back in 2016. Just took a day where we observed 38 patients. We tried to pick the busiest time in the emergency department and there was, in fact, one confirmed STEMI that came through during that time frame. On average, from the time of arrival into the ED to the ECG completion was about 15 minutes. So by taking the patient back to triage, at the time they were typing in the patient demographics and prepping the patient, getting them ready, putting the leads on, and then of course they would run the ECG. If it wasn't good quality, they would run it again. And then once they completed that, then of course they walked to find a physician. Thinking about the staff walking the ECG to the physician made me think, isn't there a better way to deliver this to the physician? So I asked it in a meeting one day only to find out that indeed there's a phone app. So the the ECG can be done by the machine and as soon as it's completed, it actually comes to this application on the phone where the physician can open it up and look at it and digitally or electronically sign it. At the same time, we decided that the staff needed to see a way to know that the physician actually received the ECG, that they opened it, that they signed it electronically, and that there was a way to communicate back to the staff if they wanted them to do something with that patient. So on the left is my phone app and on the right of the screen is what the nurses would actually see on their dashboard. So we took that concept and we actually did a little bit of a pilot for a couple of quarters back in 2016 and this just kind of shows you that we definitely got finally got above the median. We were no longer at the 60% of the time getting the ECG to the physician and so we decided based on that small pilot that this is a process that we should actually develop and support. So during our pilot we noticed a couple of things. We noticed that whenever we had the door to ECG in under 10 minutes we could actually get the door-to-balloon time in under 60 minutes 69% of the time. Now we know the benchmarks 90 but we thought that was a pretty interesting statistic. When we were longer at getting the ECG we only got the door-to-balloon time in under 60 minutes 44% of the time. So you might ask why this is important and this slide was pretty eye-opening to us. If we got the ECG in less than 10 minutes the mortality rate was lower than if we were longer obtaining the ECG. So these are the things we actually worked on in order to change the process in our emergency department. We added bar scan capability to every ECG machine. We no longer wanted them to take time to type in the patient name or any demographics. We added the phone app, but that took some time to get it modified to what the physicians wanted. So believe it or not, it took about two and a half years till the physicians could actually come up with something that they thought would work for them. It is HIPAA secured and the ECG can be sent to any iPhone or iPad. The ECG can be signed digitally by the physicians, so everybody can electronically see a name, date, and time. And then the nurses can actually watch this on their dashboard. They can see the time the ECG was acquired, when it was delivered, when it was signed by the physician, and receive some communication about patient care. One of the things that we also noticed was sometimes because people can take an ECG machine around and just randomly do an ECG on anybody, if they didn't go back into the electronic health record, there was no order matched up with that ECG. So an ECG done in the machine, but not linked to the order in our electronic health record, after 30 days gets discarded. And so that was a little bit of a problem to us and we needed to find a better way to capture that EKG and obviously keep it in the patient record. So here we are in 2020. It's been an odd year, but for this emergency department there's now a new norm. We have 226 users signed up to use the app. The average number of minutes a physician uses that app is about 1.5 minutes, and whenever they're reviewing an ECG it takes them about 40 seconds. They can see the ECG interpretation, they can see ECG measurements, they can look at rhythm strips on their phone app, they can enlarge the ECG right on the phone so they can actually look for any changes, and any STEMI ECG is sent to the app immediately with a very distinct sound. And so when that sound dings on the phone, the physician knows to open it up and look at it. That can actually be opened and viewed and signed anywhere from zero to two minutes. It's really pretty quick. It's obviously a time saver. The staff no longer walks the ECG to the physician. Now let's switch gears and we're going to talk about the cath PCI data registry. We were interested in learning a little bit more about the risk adjusted mortality metric for PCI patients. High risk PCI patients are routinely sent to our hospital for PCI. They do have a higher likelihood of mortality, but they might benefit more from having the PCI. We took a look at over 1,100 PCI patients and their predicted risk adjusted mortality scores in 2017. For those who had a mortality, we could not identify any quality of care issues, but one thing we noticed from the registry was that predicted risk of mortality score. And we noticed that some of those patients actually had a low predicted risk. So thinking that was kind of odd, we needed to understand this metric a lot more. Third-party payers, as you know, incentivize hospitals for low mortality scores such as Anthem. Several years ago, I attended the ACC conference and one of the speakers focused on the risk adjusted mortality metric. He referenced this article that you can find in the PCI, cath PCI data registry, where they talk about the comorbids and the odds ratio. It's pretty evident that if the patient's having shock, they have a higher risk of mortality. But there are other things in here that obviously predicted whether or not the patient lived or died and played into that metric. And we really needed to understand and know a little bit more about this. So we took a look at that article and then we took a random sample of 40 charts and we tried to compare the variables that were in the model and whether or not our chart documentation was lacking or inconsistent. I'm sure none of you have that issue. But what we found is that physicians were not always good about saying whether or not the patient had heart failure. If they did have heart failure, they were not very good at saying the New York heart classification. Chronic total occlusion wasn't always documented. Peripheral vascular disease. And the status of the patient going into the procedure was a point of discussion. There was a definite variation in the odds ratio between emergent and urgent. And we decided that the physicians who were looking at the patient, as opposed to the abstractor who was reading the chart, might have a better understanding of how risky that patient was going into the procedure. So of course we met with our multidisciplinary operations and quality team. We took the article and educated the physicians about the variables as well as the odd ratios. We sent physicians scorecards and what we decided to do with the one on mortality was to include at the very bottom of the scorecard a list of those patients, patient name, MRN, and the predicted risk of mortality so the physicians could look up their own patients. You'll see to the right there there's a sample of a physician scorecard that was sent out. So we like to compare to each other. You know physicians are pretty competitive. And you'll see physician F looks like he has a higher rate of mortality. He only had three patients, but at least he had an opportunity to take a look at the predicted risk of mortality and go back to the chart to see if their documentation matched the risk for the patient. So because of that scorecard and that ability to go back and look at their documentation, the physicians got together with our EPIC team and they created a document that would include the most common variables that were lacking in the patient record. We made those variables a hard stop. Included in that hard stop is the ACC definition. I'll show you an example here in a second. We also try to provide some real-time communication back to the physician, particularly if we think that their documentation maybe is not following what the ACC definition says, so that we can kind of rectify that and all be on the same page. This is an example of their PCI post-op note. And what you can see, we didn't put every one of those variables that was in that article. We only put the things that they really didn't do a good job documenting. PCI status is at the top, and what I tried to show you here is that if you hover over any of these buttons there, the definition from the ACC pops up, so the physician can actually read exactly what the ACC means. We did a couple of other things, and at the time that we were developing this and not knowing how this might change with version 5, we added some things that we thought might be included in a future risk of mortality model, like cardiovascular instability. Syntax score is not always documented well, and so we also added that as well. So let's take a look at our outcomes for risk adjuster mortality. In 2017 is when we really started to focus on this metric. OSU is the red bars, and the hospitals at the 50th percentile are the gray bars. So you can see we were higher than hospitals at the 50th percentile, and we took time to educate the physicians and to try to help them to improve their documentation. So in 2018, you can see that our risk adjuster mortality dropped better than hospitals at the 50th percentile, and between that period of time to 2019, we developed our documentation tool with the hard stops, and you can see that that number dropped again. We got our quarter one 2020 report, and you can see we dropped even a little bit more from that. The other thing to think about is that Anthem for financial reimbursement has a target of 1.01 to 1.91. So from 2018, 19, and 20, we actually met that metric and got a financial reimbursement for that quality metric first time. Now I'd like to introduce you to Sandy Campbell, and Sandy is the program manager for our EP program, and she's going to talk to you about the ICD registry. Hello, as Patty mentioned, my name is Sandy Campbell. I am the program manager for EP, and I am happy to present the quality improvement project we undertook several years ago related to our lead dislodgements in our ICD implants. We review data frequently from two sources, the ACC registry, which gives us our procedural information, as well as an internal access database we utilize to track any readmissions that occur within 30 days. Based on those reports, we found that we had a higher incidence of lead dislodgements showing at 1.02 percent as compared to like hospitals or like volume hospitals that reported an 0.4 percent. We put together an interdisciplinary team that included EP doctors, nurses from both the procedure and the prep and recovery area, pharmacists, imaging specialists to take a look at this issue. We studied a large group of patients that went back 27 months and consisted of almost 4,300 leads to take a look at their demographic factors, clinical characteristics, the types of leads that were being implanted, as well as the type of leads that were being dislodged, and the incidence of lead dislodgements both prior to discharge and within that first 30 days post-discharge. We devised an action plan and once it was implemented, we continued to study patients actually to the current date, but for the purpose of this study, we included an additional 17 months or almost 2,400 leads. The action plan we put together included several steps, most of which are procedural. They included a tug test for the RA lead, and I'll go into more depth on that one in the next slide. We also started to fluoro the leads after implant while the patient took a deep breath and coughed, and that would ensure the leads didn't become dislodged with that forceful cough and that the leads were moving during that deep breath along with the myocardium and chest. We saved that image so that we had a comparison to the chest x-ray that the patient had to have done prior to their discharge. We also left the leads connected to the programmer and that allowed us to see that those lead parameters remained consistent during that deep cough. The doctors would implant the RV lead with the tip more septal than apical so that we could ensure better stability of that lead, and then once the patient was discharged to the recovery area, we put the sling on that operative side on that arm immediately where we had been waiting to just prior to discharge. Putting it on immediately kind of gave that patient that cue they needed to not use that arm to let it remain with the elbow below the shoulder height so they weren't putting any additional tension on a newly implanted lead. We found the highest incidence of lead dislodgement was with the RA lead, so the physicians devised a strategy to reduce that. What they did was once the RA lead was implanted using a J-tip stylet, they would remove that J-tip stylet and advance a straight stylet down the lead so that the tip of the straight stylet was right at the curve of the J. That allowed a little counter traction against the tip of the RA lead and allowed them to see if it was firmly anchored or not. This slide clearly shows the impact of our action plan on patient outcomes. The control group before the action plan was implemented has high numbers of lead dislodgement. The intervention group, which is after the action plan was implemented, shows a significant drop in each of those leads with the total lead showing a p-value of 0.014. With those findings, we implemented the action plan and it became our standard of practice or our new protocol. We showed that the lead dislodgement occurrences that were at 1.02% before had dropped to significantly to 0.4% after. The majority of the lead dislodgement events occurred events occurred during the first 24 hours and prior to their discharge, and those were significantly impacted as well as the incidence of readmissions due to lead dislodgement. I am pleased to say that based on our findings, we did publish this large study in JAG in 2017. So in summary, to improve quality you must have accurate reliable data provided to an empowered, enthusiastic, interdisciplinary, innovative team with leadership from the top down approach. Data helps to identify and drive your processes to improve your initiatives, and data communication is ongoing and if possible in real time. We hope that you've enjoyed our presentation today and hope that it provided some insight in how we work at Ohio State in order to improve patient outcomes. I would like to thank Connie Anderson for helping us to pull this presentation together, and I would also like to thank the ACC for requesting us to present. Have a great day! Attached are our references for your review. Attached are our references for your review. Have a great day! I hope you're enjoying the conference.

quality improvement

Ohio State Wexner Medical Center

ACC NCDR registries

patient outcomes

lead dislodgements

multidisciplinary teams

ICD implants