Good morning. We're going to go ahead and get started. My name is Julie Mobade. I am the Product Manager for the LAO Registry. And today we are going to talk about Best Practices for the Registry. And I'm going to introduce April Wojcik. She is the Senior Manager for Cardiovascular Clinical Data and Registry Abstraction at Paralon HCA Healthcare. And she has a wealth of information and we're happy to have her here today. She's going to do her presentation and then you can use the app for questions at the end and we will get to as many as we can. So welcome to April. Thank you. All right. Thank you so much, Julie. We're going to start off with a little poll here. No, it's not on screen. I want to see a raise of hands of who started out in abstraction where you were handed the book and they said, good luck. Yep. Many of us have done that. As a CV nurse I hadn't worked in the cath lab. I worked as a CV nurse recovering open-heart patients. But I was able to learn this. And what better way than to learn the basics. So we're going to start with the basics, the what, the why and all of that. And we'll get started. So we are going to have some polling questions. So if you could, be sure to scan your QR code and we'll be doing the poll for some questions. Today's agenda includes the objectives, some background and landscape, purpose, the what, the why, the basics. And again, it's going to be very basic. But it's also going to tie into the data points during abstraction. This is really focusing on the abstraction of data points and maybe some misunderstandings about those. And then of course some challenges and tips to take away. So I have no disclosures. And the session again is to discuss the basics of LAO and associated data elements. Learning objectives, we want to make sure that this presentation enhances the learner's understanding of the purpose of the actual procedure itself, the registry and the basics of the devices, as well as common data elements. So let me give you a little bit of background and landscape about where I work. So I work with Paralon and HCA Healthcare Systems. And we deliver industry excellence at a large scale, obviously. We service over 2,000 hospitals, providers and locations across 19 states. Our purpose and values are that we serve and enable those who care for and improve human life in their communities. We embrace change. You know, Paralon and CDRA, which is the Clinical Data Registry Abstraction Group that we're called, the CV service line, we embrace change. When we started it four years ago as centralized standardized abstraction for the enterprise, it was much different than what it is today. We openly collaborate not only with our abstractors, our providers, our coordinators at the facilities, but also the registry. Our landscape includes all the registries that you see listed here. And of this, there are about 100 abstractors on our team. Like I said, that abstract for the enterprise of HCA. We abstract a total of 485 registries. That's a lot. And it's within 140 facilities with 123 cases or charts per year. For LAO we had 75 and 9,000 charts. So when it pertains to LAO, there's 75. So let's talk about the purpose of the LAO procedure itself. It's a minimally invasive procedure that reduces the risk of stroke, right, in patients with AFib by preventing that blood clot from going into the bloodstream. And then of course the registry, the purpose of that is to ensure that that data is collected. It also provides the ability to see outcomes, to see what's trending, to see long-term and short-term safety as well as cost-effectiveness. So let's get into the what and the why. So the procedure, like I said, is designed to help manage the stroke risk in patients with non-valvular AFib. It is a treatment option for those patients who are unable to maintain adequate anticoagulation through medication therapy. So why? Why have the LAO procedure? Well the top reason is stroke prevention, right. AFib is potentially a modifiable risk factor associated with acute strokes. And early therapy and risk stratification can help reduce the risk of AFib-related strokes. And therapy, as we all know, can be a balancing act. You know, is it too high risk to put the patient on medications because they're going to bleed? Or are they going to have a stroke if you don't? So there's that balancing act. And I think LAO helps the providers in determining what is the best plan for the patient with the patient's input, okay. So just basic, what type of devices are out there for LAO or the left atrial appendage closure? You have two primarily. One is a blocking. It blocks that appendage. And that would be the Watchman or the amulet. And then you also have the Lariat which is a little bit different. It's a band or suture that loops around the appendage, closes it off. And then, you know, takes care of the appendage. So again, very basics. And the reason why I put this in here is, not all abstractors came from the cath lab. Not all abstractors know what the procedure is and how it works. Many of us were taken off the floor and put into the abstractor mode and you don't even know what it is. So you've got to learn it. So why not start with the basics? So as you see here, the Watchman and the amulet, you know, they insert a guide wire into the femoral vein. The septum is crossed. The device is deployed. And then they release it into the left atrial appendage. The heart tissue goes over it to then seal that appendage. You can see the difference between the two devices and how they differ. The Lariat basics, it's similar but different. So again, you're going to insert that magnetic wire through the femoral vein. They insert a second wire which then is through the outside of the heart through a needle stick in the chest wall. Again, you know, they're going to attach the suture and advance that across the magnetic wire and tie off that suture around the appendage. And then that appendage dissolves over time. So those are the very basics of the procedure and sort of like how it works. So how does it pertain to our data elements, right? So remember I said it was a balancing act between risk of stroke versus bleeding or the procedure, the LAO. CHADS VASC is one of those data points that is used a lot. And it's also used with the National Coverage Determination ensuring that the procedure is being performed on the correct patients, on the right population of patients so to speak. So it's an acronym. As you see here, each letter stands for a condition or a risk factor that has associated points and that max point value is 9. This score helps drive decisions, treatment plans as to whether or not there's a need for anticoagulation in the AFib patient or not. This is a little example of what I created for our CDRA team. You know, a lot of times people don't understand what the CHADS VASC is. You know, providers, clinicians, people who are living in this world, but as the abstractor you may not know it. So again, the C stands for CHF. And with that, if you have a condition that meets the definition, you would get one point for that. Hypertension would give you one point. Your age greater than or equal to 75 is going to give you 2. Diabetes would give you 1. Stroke is going to give you 2. Your vascular disease is 1. Age of 65-74 would give you 1. And then of course your sex, females would get one point, with males 0. And again, the total max score that you could get would be 9. So based on these conditions or risk factors, it helps the providers determine what risk you're at or your patients are at for a stroke in the AFib patient. So here it just gives you a brief summary of point totals, low, high, intermediate and whether or not aspirin or oral anticoagulation is required. And again, in order for the National Coverage Determination to be met, which is set forth by CMS, the Centers for Medicare and Medicaid, the patient must have a CHADS VASC score of equal or greater than 3, or a CHADS score of 2, which is a shortened version of the CHADS VASC. With that, our team when they abstract, they had a lot of questions. How do we do it? When do we know when to code what the provider's telling us? Maybe the provider's not telling us everything. Maybe they're only giving us a number. Maybe they're giving us everything. There's extra things. What do we do? So we started with a flow chart, right. We made this tool for the group. Did the physician clearly document what risk factors were used when determining that CHADS VASC score? And if that answer is, yes, you would abstract that CHADS VASC score as outlined specifically by that physician, right. But what do you do when there's extra things that he didn't mention in those specifics? Do you capture them? Do you not? Well, first you would want to clarify that with your provider. And if you're unable to do so, you would only code in the abstraction those items that they listed in your CHADS VASC score. Now let's say we have a different scenario. They didn't provide us with what actual elements were determined for that CHADS VASC. At this point you want to ask yourself, does it meet the definition based on the data dictionary, right? No, it does not? Then you don't code it. If it does, then you're going to abstract that based on the NCDR definitions. And I provided a quick overview of that in this grid. And this is a tool that we use and give to our LAO abstractors at large so then they have that workflow. So we're going to start with our questions. So we have a 75-year-old male admitted with LAO procedure and noted to have a history of hypertension, uncontrolled diabetes, left ventricular EF of 35% noted on an ECHO three weeks ago. Physician documents CHADS VASC score of 3. Based on the documentation provided, how would the CHADS VASC section be coded? Try to determine what history the provider used to determine the CHADS VASC of 3. Capture conditions meaning the definition or leave blank. And here is the area of the data collection form that you would be completing. And we're going to go ahead and get ready for our first poll. Sorry, but I have to wait 15 seconds, otherwise it doesn't — okay. So the group has decided that it was a B. And just to refresh what the information was, the answer is B, capture the conditions meeting the definition. And the rationale behind that is the provider did not specify what conditions were used in determining that score. So therefore you would capture anything meeting the definition per the data dictionary. And this is what it would look like if you were to capture it on the paper form or the data collection form. Question 2, 65-year-old female admitted with LAO procedure noted to have a history of hypertension, diabetes, and stroke. Physician documents CHADS VASC of 5, age, sex, hypertension, and stroke. Based on this documentation of the CHADS VASC score section, how would you code it? Again, try to determine what history the provider used to get to the 5. Capture only the conditions the provider states. Capture all conditions meeting the definition or leave blank. And again, here's your data collection form and your time to poll. We have a little music. And I don't sing. All right. So 58% capture only the conditions the provider states. Let's see what we have there. And it is correct, B, capture only those conditions the provider noted. And again, the rationale is, you know, there are other conditions. But because he's very specific in what he chose, that is what you are to extract. All right. And again, this is what it would look like. So we talked about stroke risk. Let's go into the HASBLED. Okay. So I'm thinking data collection form. I'm going through it. This would be like your next section that you're going to be abstracting for. Again, it's an acronym. It's going to assist the provider. Instead of predicting your risk of stroke, it's going to predict clinical relevant bleeds. And again, the max score is 9. So it's very similar to your CHADS-VAS, but it is a little different. And it does identify potential risk factors that may be reversible or modifiable. So again, I use the same setup, you know, what each letter, the HASBLED stands for. The H is for hypertension uncontrolled greater than 160 systolic. And you'd get one point for that. Your abnormal renal or liver function, now please note it does say or. If you have just one, it's going to be one point. If you have both of them, it's going to be two. Your S is for your stroke, which will be your one point. B is for bleeding history or predisposition for that bleeding. And you're going to get one point there. Labile INR is going to be the one point. Your E is elderly age of greater than or equal to 65 in which you get one. And then drugs and alcohol. This is aspirin and NSAID concomitant medications, not illicit drugs. So you know, are you on that concomitant antiplatelet or NSAID, you would get the one. If you had alcohol use with that as well, it would be two. Again, all of these are based on the definitions set forth in the data dictionary. So the data dictionary is your best friend. So is the FAQs. I'm just going to put it out there. So again, max score of 9. So let's go ahead and do a quick question, 65-year-old for LAO procedure noted to have a history of non-valvular AFib, hypertension, uncontrolled diabetes, takes an aspirin 81-milligram daily. Has BLED score of 3. Based on the documentation provided, what would be captured as a yes in the Has BLED section? Your choices are hypertension, bleeding, antiplatelets, NSAIDs A and B, I mean A and C, A and D, A, B and C. And again, here's what your data collection form would look like, the section that you're coding for. And then we're going to poll you. Okay. So we're a little bit all over the place. So we've got A and C at 42%, A, B and C at 36%. So let's see, what does it say? So we're A and C, right. A and C is our hypertension and our antiplatelets. Now the rationale behind this is, there is an FAQ 25489 that states to capture any dosage of an aspirin as an antiplatelet in Sequence 41.10. Back in the day when I first started LAO this was not the case, right. So this was sort of a trick question. But again with the FAQ, you do capture any dose of aspirin as an antiplatelet now. And this is what it would look like. All right. So now that we've gotten some of our risk factors done, let's talk about shared decision-making, right. The shared approach, it seeks your patient's participation as a coordinator, a physician, a clinician. Shared decision-making is very important. It's also required by the National Coverage Determination that Medicare set forth. It's part of a requirement in order to perform the procedure that the patient must participate in this. It helps your patients explore and compare treatment options, assesses the patient's values and preferences, and helps the patient reach a decision and evaluate the patient's decisions as well. It improves that quality of care. Patient satisfaction is increased because the patient's informed. They know just because it's a procedure doesn't mean that it's right for them, right. Maybe they don't want that. Maybe they do. Maybe they're tired of being on Coumadin or Warfarin and the INRs, the PTN INRs all the time. Maybe the LAO is the answer. So it also helps improve patient adherence to treatment recommendations. So as we all know, the LAO procedure has follow-ups. And it has multiple, right. We have three follow-ups that are required in the registry. But actually CMS tracks that patient for a little bit longer. I don't know if you knew that or not, but they do. So when this patient is having an LAO procedure, you want to make sure that they're very aware that they have to continue to follow-up and evaluate that device and how they're doing. So shared decision-making should include that. So let's talk about the differences between informed consent versus shared decision-making. So informed consent is a brief description of the planned procedure. It's just a description of what they're planning to do with risks and benefits. It's basically obtaining the permission of that patient to proceed with the procedure or the surgery, right. Whereas shared decision-making is different. It is a discussion that is formal between the patient and the providers about the different treatment options. And it's the patient and the clinician who are contributing to the decision-making process and agree on the plan, right. Let's not confuse them as being the same because they are not. So when you are abstracting and it says in the record that informed consent was obtained, that doesn't count as shared decision-making. It must be very clear. And there are different criteria for that. So when abstracting the Sequence 14732, a formal interaction must be noted. It must be documented in the medical record and it must be documented prior to procedure. It's documented that it occurred after? It doesn't count. You would have to code it as a no. You can have some smart phrases that are used in your electronic health record by a non-interventional provider, smart phrases. And if they did say that and it is, you know, like a shared decision-making tool was used, you can code yes. But it has to be very specific. And then of course that statement that the provider did a shared decision-making encounter occurred by the non-interventional provider using an evidence-based tool in the patient with non-valvular AFib, you could code yes. So let's go into our case scenario. Question 4, 65-year-old female admitted for LAO. Risk and benefits discussed. Patient agrees to proceed with the procedure and signs the consent form. HMP notes that the NICE.org tool was used by the primary care physician who's a non-interventional provider prior to admission. Based on this documentation how would you code 14732, shared decision-making? Would you say no, yes, or leave blank? And this is the area of the data collection form that you would be completing. And we'll go into the question. Okay. Wow. We got a 78% for yes and a 22% for no. So let's take a look at it again. And the answer is, we would code yes. Shared decision-making was performed and we could answer that question yes. The rationale is the formal tool was used, which was the NICE.org tool. And it was performed by a non-interventional cardiologist or provider. And it was used. And it was documented prior to the procedure. So it meets all the criteria of a shared decision-making tool. And this is how it would look like on the data collection form whenever it was abstracted. All right. So AFib termination. This is the next section that you're, as I said, we're going through that data collection form. I'm thinking about it. AFib termination can be one of the hardest things if you're not an EP nurse to really get a grasp on and understand. There's differences of course. So you have that pharmacological AFib termination, Sequence 4415. Here listed are some common medications. I'm not going to read them to you, but these are the most commonly used. And they can facilitate electrocardioversion. So they could be used in conjunction. Then you have your direct current cardioversion, which is another type of your AFib termination, which is the delivery of the electrical shock with synchronized QRS complexes to avoid that V-fib. But we're going to get a little bit deeper into prior ablation strategies. So I was thinking about this. How could I do this? I could do a bunch of pictures and just, we'll talk about it. So why not make it a little animated? So I'm going to start off with, you can see here that these are your pulmonary veins. And the first example we're going to go over is WACA, the Wide Area Circumferential Ablation. What is it? For those of you who are not real big EP experts, I wasn't. It is where they take that ablation catheter and they make the large ablation around both of your pulmonary veins as a pair and it's wide. This is a type of pulmonary vein isolation, right. So this would be your WACA. The next one that I want to talk to you is just plain pulmonary vein isolation, which you may only have one pulmonary vein, but you can see it's individual of each of those pulmonary veins. Now they can go ahead and add linear lesions to that to help mitigate that AFib, right. And that would be adding the orange lines on the roof and then I believe the mitral area to help with the ablation. So this would be a pulmonary vein isolation with linear lesions. And then finally we have the pulmonary vein isolation with complex fractionated electrograms which adds target areas detected by mapping systems that may represent unusual electrical patterns. This is what it would look like. I never had it explained to me this way, but it makes sense to me. And I'm all about working with our team of abstractors and getting them to understand and visualize and understand what it is they're actually trying to capture. Some other ones include the pulse-filled ablation, the convergent procedure as well as cryoablation. You sort of get the gist of what they do with each of these. The convergent procedure is really like a hybrid approach where it consists of epicardial followed by endocardial radiofrequency ablation. And also the cryoablation uses the freezing technology to where it involves cooling that's being released into the catheter balloon to freeze and ablate the tissue. So when you're abstracting this, you want to capture any occurrence between birth and the procedure. Capture both successful and unsuccessful attempts. Pharmacological cardioversions must be specific documentation in the medical record stating the purpose that it was for a cardioversion. Can't assume it. And then here are some examples, a couple examples of when that would be a yes. A bolus in the ED was given of cardizem due to AFib-RVR. Amirodarone was started with a bolus secondary to AFib or need to cardiovert. So these are some examples here. Direct current cardioversions performed after the patient enters the room. This is a big one. Are not collected in this sequence number. Because remember, we're in the history section. What happened before they got into the room? Once they go into the room, if there is any type of cardioversion, it's not captured in this section of the data collection form. And again, AV node ablation is not captured in 4410 or 4425. Although it is used to treat your AFib by controlling your rate, it is not ablating the rhythm. It's just controlling the rate. So we are directed to not capture it in these sequences when AV node ablation is performed. And then finally, if the strategy is not clear, seek provider clarification, right. But that's not always possible. If you can't seek, if you don't get the answers that you need, then you would be directed or guided to leave that sequence blank, excuse me. So let's do a scenario, 89-year-old male admitted with LAO procedure noted to have a history of hypertension, diabetes, stroke, prior history of AFib ablation to include pulmonary vein isolation of both sets of the right and left veins with adjunct of linear ablation of the roof and mitral lines and CAFE. Based on this documentation provided, how would you code the prior ablation strategies to be captured? We have WACA, PVI, empiric left atrial linear lines, complex, we have ACD, BCD, or leave blank? And this is the area of the data collection form that we are abstracting. And we got our polling. It's only 15 seconds, but you'd think it's a long time. All right. So we've got B, C, and D. And let's see where we're at. That is correct. And this is what it would look like on your data collection form where you're going to do your attempt at atrial fibrillation termination, a yes. Pharmacologicals, a no. DC cardiovisions, a no. Catheter ablations, a yes. We would leave the date blank because we don't know the date. I didn't provide that information. So we would leave it blank. And then of course the prior ablation strategies you would select all that apply, which would be the pulmonary vein isolation, empiric left atrial linear, and the CAFEs. So moving a little bit further into the data collection form, we have LAO occlusion indication sequence 14837. We're to code this based on the documentation provided in the medical record as to why the procedure is being performed. It's the value on current procedure. And it's the primary reason for the procedure being performed. And these are our options to select and we're to select all that apply. So when abstracting this sequence, you want to make sure that you're not taking that elevated CHADS FAS score and assuming that it is automatic increased thromboembolic stroke risk as an indication. You cannot automatically discern that. Patient preference must be indicated by the provider, not in a nursing assessment. The provider must state it. Clinical significant bleed risk was added as an option for those patients who have had a prior bleed, but it didn't meet the definition of a major bleed. But again, it must be documented by the provider. And then as always, nothing should be assumed automatic or implied by the abstractor. So in conclusion, you want to make sure that the provider is providing you with that documentation of the indication. If he doesn't provide it, if they don't provide it to you, you need to seek that clarification because we can't assume it. We got a question here, a 70-year-old male admitted for LAO. The physician documents a history of LVEF at 35%, hypertension, persistent non-valvular AFib with a CHADS FAS score of 5 and a labile INR. Nursing assessment, the patient refuses to take Warfarin. Based on the documentation provided, how is sequence 14837 coded? The patient preference, labile INR, increased thromboembolic stroke risk, or all of the above? And this is the area that you would be looking at on the data collection form. And we'll wait our 15 seconds. I am going to say that the NCDR website, FAQs, data collection forms are a wealth of information. And we highly recommend when you're extracting to have it open. All right. So labile INR takes the win it looks like. We do have some that said all of the above. So let's give you some rationale behind that. So again, remember that little caveat that the nursing assessment, we cannot assume that the patient's refusal is the indication for it. Remember, the provider has to state it, your physician. Again, only the indication provided by your provider, which was labile INR. There was nothing else. You can't take the CHADS FAS score of 5 and automatically apply that risk of stroke as an option to code here. So your answer would be the labile INR only. And this is what your data collection form would look like. So this one's my favorite. This was something after numerous contact us through the registry, we got clarification for our team of extractors about the device marginal leak. So as we all know, we're to code the size in millimeters of the residual leak noted at the device margin, right. It's the value on current procedure and there is a range. And we are not to code this data element when the procedure is aborted. You're to leave it blank. I didn't know if you knew that, but it's a little tip. So what we created was a grid for our group. We took all the information that was out there, the data collection dictionary, data dictionary. We took the FAQs. We took our queries that we had sent. And we provided this to our team. So when there is a range provided, a range of 1-2, you're going to take that highest value. And that's based on an FAQ that's out there for everyone. Then if it's documented in the medical record less than a number, then you're going to code the full number below that. So less than 3 would give you a 2, less than 5 would give you a 4 and so on. And again, this is supported by the FAQ that's out there. We did two queries on our own. And we did clarify that when there's documentation of complete seal, no flow, or no leak, could or would we code that as a 0. And we did get that query back. And it did say that yes, we could. Now what it did say on our second query was that if there was those vague descriptors such as appropriate, large, marginal, trivial, significant, not significant, trace, can't assume that to anything, nothing. There's no crosswalk to it. You have to get it clarified with your provider. Or you're going to have to mark it not assessed. And again, this was supported by queries to our team. So this here is a grid that we took all of it and put it together. And they use this as a tool when they abstract. So they're not having to hunt through all these different queries and FAQs and resources. But of course, the Data Dictionary always rules. I'm just going to put that out there. So it is subject to change. But this is currently a tool that we use for our team. So question, 72-year-old female admitted with non-violator AFib with several unsuccessful attempts at AFib ablation. Now with labile INRs, recent ball, op note states that a 25-mm Watchman was deployed successfully with a less than 2-mm residual leak. Based on the documentation provided, how would Sequence 148.48, Device Margin Residual Leak be coded? 2, 1, not assessed, leave blank. And this is the area of the data collection form that you would be coding. And we'll go into the polling. All right. Let's see what we got. Yay, 89%. We said 1 mm. So let's see the why behind it. So remember, they documented less than 2. So we have to select 1 because of the FAQ that is present. Anytime there is a less than number, you're to document that immediate full digit below it. I don't think they would ever say it was a 2.1, but you know, you wouldn't say a less than 2 is a 1.9, right? You're going to go to the whole number. And this is what it would look like on the data collection form. I have one more. So we have a non-valvular AFib with a history of stroke with intolerance to anticoagulation. After entering the room and venous access is gained, the procedure was aborted due to the anatomy of the left atrial appendage being too small. Based on the documentation provided, how would you code Sequence 148, Device Margin Residual Leak? 0. Send a query to the NCDR using their Contact Us feature? Not assessed or leave blank? This is the area we're looking at. And we got 15 seconds. I am going to suggest we need music for these. All right. We have a leave blank. We have a not assessed. And the answer is, we're going to leave it blank. And the rationale is because the procedure was aborted. Now it is noted on the data collection form, the paper forms, that you're not to answer this question if it was aborted. I put this in here because this is something that sometimes is overlooked. And I think it's important for abstractors to really keep that in mind that when it's aborted, this is not answered. It is expected to be left blank. So wrapping up those data elements, what challenges and tips do I have for you? Well, with the amount of cases we do and the number of registries, our number one problem is documentation. I don't know who else has it, but we do. I don't care who you are. It's out there. All right. Documentation is always a struggle. You know, what's missing? It's only a partial answer. And then getting it clarified. Multiple electronic health records is one of our problems across our enterprise. Again, we have several facilities. We have everything under the sun that you can think of as an electronic health record. So we mitigated that by providing full access to all of our CV abstractors for our registries. So they have access to all of it. So we don't have to worry about who has what access. We just gave it all to them. All right. We all work for the same team. Next up is the follow-up, follow-up occurring outside of our HCA healthcare system. This is very difficult because we don't have access to other physicians' documentation. So we have to ensure and have that great relationship with our coordinators, our boots-on-the-ground people who are going out there and following up with these patients and gathering that information and getting it scanned into our medical record. It can be labor-intensive. We understand that. But it's what's right for our patients. It's what's needed. And it's what's required. So this is something that is a challenge. But again, we collaborate very well with our coordinators. And then of course the timeliness of obtaining these records. I don't know, but I've experienced it where somebody does a ton of follow-ups and they hold on to that paperwork in the clinic. And then they decide to go and have 50 patients scanned in with follow-ups. Well the follow-up happened three months ago, but I didn't get to see it until today. So what we did was we as an abstraction team decided that we were not going to code the case. We weren't going to go looking for the follow-up until after the window closed. We weren't going to. We were going to wait 14 days. It gives the coordinators time to get what they need to get together at the clinics or from the offices. Get it scanned into medical records. And then it allows us the time and the opportunity to touch that record once instead of having to go in and out. Did they do a follow-up? No, they didn't. Let me check it again. We do it once. Okay. So these are our challenges. Some best practice tips is have a repeatable routine and review your records systematically every time. As an abstractor, I sat in that seat. And if you don't have a routine, you will forget something, you'll miss something. It keeps you efficient. It keeps it consistent. So efficiency is consistency. Next we talk about our abstractors opening all their tools before they begin abstraction. This is the electronic health record. This is their data dictionaries. This is FAQs that may be out there. We have a repository of FAQs that our team sends out to the Contact Us. And we keep it in a repository that is managed by our quality and education team. And if there is a conflict between questions or queries, we then send that to conflicting information to the registry for clarification. But we have this repository that all of our abstractors have access to. And it has all kinds of questions. And again, if there's more than one of the same or similar question and it conflicts, we send that up for clarification. And finally we do have a facility handoff tool so to speak. So because we have so many abstractors, everybody will potentially get cases among different hospitals. We try to keep everybody sort of grouped together. But there's the times where people go on vacation, they go on family medical leave act, whatever the case may be, we may need somebody to cover. So what a better way than having a form that gives you the nuances of that facility, right? It gives you every information. What type of electronic health record? How do they perform their follow-ups? Is it at an HCA facility generally? Or is it out of network? Is it scanned? Is it scanned into the base record? Do they have a separate encounter? All these things are listed. So then as the abstractor I open it up and I know what those nuances are for that facility. Most importantly after they get everything open, we tell them, stop, go directly to your op note. If you don't have an op note, you go to your discharge summary. You need to verify the case is even eligible. Don't spend an hour figuring out your CHADS VASC, your HAZ-BLED, your AFib ablations to only find out that the case is not eligible. So this keeps our team efficient. It keeps them productive. And it keeps them on track and not wasting time doing a whole case. So verify that eligibility straight out the gate. Next submit often, submit frequently. We tell them, submit weekly. Every week at the end of their work week they submit. They also reconcile their dashboard weekly to identify keystroke errors. So our abstractors do have assigned facilities so to speak. And that's their facility that they own to review for data, keystroke errors, all of that. We also have a communication tool with our coordinators that if there's missing information that we need or need clarified, our coordinators will then either speak with the physician, get it scanned into the record. We work closely with them. And then we make those edits. And this is a two-way tool that is our communication tool with the coordinators. We keep our team in the know. We encourage them to attend all of the registry education case scenarios through our Education and Quality Team as we do IRR and quality control on our reviews of our abstractors because we do review their cases and make sure everybody's on the same page. We want to make sure that it's standardized. So an apple is an apple is an apple across the enterprise. If we come across something that is amiss or we have a lot of people missing the same data point, we do Hot Topics. So we'll do education in communication versus an email. We'll pull together a grid such as like the device marginal leak. That came from a Hot Topic. Instead of just reading them the definition, I'm like, well, let's give them a tool to help them. That was one of those things that came up. We also do quarterly education. And if there's individual education or remediation that's needed for our team, we provide that to them if we see that there's a problem with their quality. Things to avoid, the use of a coding summary. Why? Because it's not permitted to be submitted during an audit. And I'm going to tell you right now, audits are scary. The first time I had my audit I was scared, but I was happy at the same time. Because it confirmed if I was doing my work correct. Because remember, I was in that closet with a book, good luck. The audit confirmed. It validated what I needed to know, what I was doing right. So don't fear audits. They're labor-intensive sometimes, but you learn from them. But don't code from them, from your coding summaries. Extracting from spreadsheets, emails and PDFs that are not part of your permanent medical record. Shared access or share points that are not part of the medical record, we discourage the use of that. We encourage all the information be in the patient's medical record because it's legal. And it also provides a comprehensive chart for the care of that patient. Abstracting on paper, we tell our team, no paper, all work from home 100%. It's a risk of PHI. And it also decreases your efficiency. And then finally, abstracting onto a PDF. Even though it's electronic and you don't have that PHI risk so to speak, you're putting that data onto the data collection in the PDF and then you're putting it into the registry. It increases your chances of a keystroke error, flipping your numbers, and it decreases your efficiency. Just put it straight into the registry. You have it up. Our teams do have multiple monitors. So it does help to where they have the medical record, the data collection tool pulled up in the registry. And then they have all their other resources that if they need to refer to it, they can. And again, it decreases your efficiency and increases your stroke, your production. So I'm April. And I thank you so much, letting you know that I have been in your shoes. I have sat in that closet. I have seen change through the registries. And change is good. And we'll take questions. »» Thank you, April. Can everyone hear me okay? There's a couple of questions for you specifically. So I'm going to try to get to those. What does your communication tool for missing documentation look like and how is it shared? »» Okay. So it is an Excel format. It is a two-way tool. So it's on a shared point. And our coordinators have access to that shared point. We can dictate who can see it and who cannot based on the access to that folder. So our abstractors will put the information onto the missing documentation and items such as a missing cath lab log if you perform your procedure in the cath lab. That's a big one. Your device marginal leak if it's missing, the indication. We have a set drop-down of the most frequently requested items that are missing. So when we first started centralization of abstraction, it was just a free-for-all. Everybody just put everything up on it. So then we decided we talked to the team, got team input, said what is the most frequent. We added it to our drop-down list. That drop-down list is reviewed every year with the team. So again, it's a shared point. It's an Excel that has been vetted through not only leadership, but also the end user. And then they update it. So if the facility updates it, then the abstractor knows because they get notified. You can set your settings. They can make the edit into the record and correct it, or I say into the registry after it's confirmed in the medical record. So we don't abstract from it. It just gives us a tool saying, yes, we've addressed it. It's in the medical record. Abstractor goes out, confirms that yes, it is. This is the data. They put it into the registry. We resubmit. So it's like a rinse and repeat. But that is our tool. And currently right now we do have another tab that has outliers listed. So those outliers we coordinate with our corporate team at the Clinical Service Group at HCA as to the top metrics. We do list them. Like one of them off the top of my head is the FDA indication. Okay. I abstracted for several registries. So I may be confusing it. But if it's there on the primary dashboard, those are the ones that we have listed for the most part. All right. Next question. »» Okay. Well, can you share your IRR template? And how often do you perform an IRR? Who performs them? And along those same lines, do you have a full education team outside of your data abstractors? »» Okay. I caught most of it. So IRR, we have a team that is called our quality and education team. They educate all of our new team members that are coming in. So if you're a new abstractor and you've never abstracted, but you're a nurse or a cath lab tech, you know, we will hire you. We train you if you're a good fit. We have a set roadmap that you follow each day what you do. And a lot of it, not all of it, but there's parts of it that have you take the courses within the NCDR. But we also have created our own, like all these things that you've seen today that we supplement our team with, we have an education platform that it gives them a roadmap of what they need to do. That education usually is about 21 days for the new person. I am going to put a little caveat out there. We have a poster that explains this in detail out there. So be sure to go out there and look at that e-poster. But anyways, the IRR on our existing abstractors, we do 10% of the cases. We have selected to not do 100% of the data elements at this time. Because again, our quality and education team was in the beginning. This is something as our group grows, we're incorporating. We hope to get to there to where we do 100% of all data elements. But what we did was we took the top 100. We all sat down, like what is important? What drives HASBLED, CHAS-VAS, AFib ablation, indications, all these big things, complications. Like last name and first name, I mean that's a gimme, right. It's pretty hard to get that wrong. So we did like the important top 100. So we have a team on our quality group. After the case is abstracted they re-abstract the case, determine what is correct or incorrect. They provide feedback to that abstractor with the rationale as to why the abstractor does have the ability to appeal that decision. And if there is an impasse, then they escalate it to me. Which then if I cannot answer it, I escalate it to Julie. Did I get all the questions in there? »» I think so. Did you mention sharing your template? »» The template is basically an Excel file. I'm going to tell you how I made it. I went to the registry and I pulled down the Excel version of the data points. I made it pretty with the help of our data analyst, who we are so grateful to have. She made it parallel and pretty, so to speak. We selected our top 100 and that's what we used. And then again, we tried to do randomized. So our data analyst will randomize those cases and assign them to the quality team to review. »» How do you get your patient list and how do you manage the follow-ups? »» So patient lists, depending on the registry, it may be automated based on ICD-9 codes or CPT codes. We do have that automation to our data warehouse as an enterprise. So like Cath PCI for example, it's automated. LAO, it's automated as well. But automation is a little bit different with that one. So we do have that automation coming in. In the beginning, we didn't have automation. We had to get lists from the facilities, the coordinators. Or we ran reports based on ICD-10 codes, right. But now it's really automated. And then your other question was? »» And then actually along those same lines, how does the follow-up documentation get captured once you have your? »» Okay. So each facility coordinator has a different process. As big as a corporation as we are in this health system, you would think it would be standardized. But it's not. We have different EHRs. We have different providers who may be HCA related or not. So it's hard for us to say that you have to create an encounter and you have to do this. So what we do is this. If they have a brick-and-mortar valve clinic or a valve clinic coordinator, if they perform the follow-up they can scan it into the base record under outside records. And then our team knows to go into those outside records to look for the paperwork of that follow-up. You know, was one done? Is there an echo in there? Are there labs in there? And they also look to see if there were other admissions within our health system to ensure that there were no complications that maybe they went to a sister facility. Now for those physicians that are not HCA, again we do rely and work closely with our coordinators at the facility because they do have to follow-up with that physician to get that information into the record. And again it's scanned into the base. Now I do have some divisions in the Florida area that would make their own encounter number for their follow-ups. Because they actually had a valve clinic where the patient would come in. And so they would have an encounter. So they have a separate encounter for those follow-ups. And that is like the best, I'm not going to lie. Because you go to it. It's right there. They do it all. But we work with our coordinators as to what is needed, what is important. We educate our coordinators. We educate any key stakeholders that have questions about the LAO Registry and its metrics and data points because, you know, we have to go by the definitions. I hope that answers your question. All right. »» Thank you. So this is just a, not to you, I'm just going to make a general statement. We've had a couple of questions about the LAO Registry being mandatory and when that would end. I just FYI, we monitor that regularly. So far nothing at this point. So when we know, of course you would know. But right now we don't have any indication that it's not going to be mandatory for those procedures, for that procedure. And we've got maybe time for one more question. And we have a couple of our experts in the room. So Fernanda and Karen is back there. I think we can stay a couple of minutes, maybe, if there's other questions that we don't get to. Otherwise we'll do our best to answer. Okay. I'm trying to get through the ones with you, April. And you did answer the IRR about your team who does the separate team. »» It's a quality education team. That's their primary goal. They do a little bit of actual cases themselves to keep in the abstraction mode. But their job is to do the IRR's quality and education. They train our new people or cross-train people. And again our new people, like I said, it's about 21 days from hire to releasing on their own. And we don't release them until they're at 95%. And when we're doing a new person, they get 100% of all the data elements reviewed. And then what we do is we do meet with our divisions and try to give them a quarterly synopsis of how their abstractor's quality is to let them know that we are working with our team to ensure that they are getting, that the facilities are getting accurate data that is standardized, which is so important. Thank you. »» Thank you so much. And I thank you for your attention today.

LAO Registry

Paralon HCA Healthcare

cardiovascular data abstraction

stroke prevention

non-valvular AFib

Watchman device

CHADS-VASc score

informed consent

registry documentation