Alright. We are going to go ahead and get started. So thank you all for coming to the AAPM&R Data Registry Session. This is on improving patient care through use of the registry and some use case scenarios. So I'm Mark Wong. I'm at the Shirley Renewability Lab. We are a registry participant. I'm also the chair for the Data Registry Steering Committee for the AAPM&R. So we're very proud to sort of give an update as far as where we are for some of the registry data as well as some of the use case scenarios that we're starting to see in terms of the use of the registry. So first off is just sort of the overview. So we're going to go over a little bit about data collection and some of our aggregate data that we've collected to date. And then we're going to go over use case scenarios first involving stroke with Dr. Geis at Brooks as well as with Dr. Harvey at the Shirley Renewability Lab. And then we're going to go to some use case scenarios in low back pain and that will be with Dr. Yang and Dr. Lee. Dr. Yang is at Vanderbilt University and Dr. Lee is at the Shirley Renewability Lab. And then I'll just sort of wrap up with some closing remarks and then obviously some time for questions at the end. Although honestly, if you have questions at any point during the presentation, just go ahead and please raise your hand and we'll see if we can get those answered. As far as our conflict of interest, none of our panelists or presenters today have anything to disclose in regards to this presentation. So just talking about, you know, why are we doing this and why is the AAPMNR getting involved in this area? As you know, the continuity of care, you know, obviously one of the clear issues we have is unclear outcomes, you know, clinicians who are stressed and overworked at their jobs and trying to prove these results and really in this world of data, you know, we're really trying to show the value of rehabilitation through data and what better way to collect it than to sort of take charge and have a registry of our own. We've had some sort of iterations of the registry throughout the years. At first this was more for regulatory compliance. It sort of shifted to more focused on several different areas. In this case the two diagnoses are low back pain and ischemic stroke. So this is sort of a single repository for the registry to really track real world data. So this is more of a, you know, something to keep in mind. This is really a clinical data registry. This is not some sort of research registry like a model system. So very different focus because we're collecting data on individuals that are commonly seen in the practice just with the following diagnoses of stroke and low back pain. So really trying to use this data to really look at several different things, you know, how we can define practice in these settings, really look at how we can move the field forward just through analyzing some of the data, looking at some of the populations, creating some benchmarks based on the participating sites, and really improving patient outcomes. And I think ultimately the focus is on how we can use this data to define, you know, the value of PMNR. Could this eventually lead to practice guidelines, to quality measures that better measure us as physiatrists as opposed to all these primary care measures and other things that we're constantly being forced to utilize to sort of measure things that we don't really, that's not really the focus of our practice. So, you know, in terms of just trying to show the value of the specialty. These are some of our participating sites that are currently involved with the registry. So we obviously are grateful that these sites are up and running and working with us on data collection. So what kind of data are we collecting? I mean, we, if you go to our registry website, there's, you know, you should have the actual data elements. But really it's a combination of two different things and it's really, so first we're collecting data from the electronic health record of the various participating sites that go to the registry. And in addition, probably one of the unique things that we're doing is also collecting patient reported outcomes through surveys or questionnaires that the patients themselves complete to then create our database. And I think one of the important things here is that we are literally the first specialty that is collecting data at a specialty level for patient reported outcomes. And so very important to understand that that's been a big undertaking that makes us different from the other specialties. Again, patient reported outcomes, which is going to be really a big focus in the future in terms of measuring outcome and impact of healthcare professionals. And that has been a big push at CMS, Centers for Medicare and Medicaid Services, to really look at patient reported outcomes with patients and making that part of care and expectations. And so we're, I think we're ahead of the game for most of these specialties in terms of trying to do so. This is some of our aggregate data that has been collected to date with the registry. And keep in mind that this is aggregate across both diagnoses, low back pain and ischemic stroke. You can see by race, majority are white and in the database, which is not unexpected. We also have some that are obviously sort of the race is unknown. As far as ethnicity, we do have some areas in the data where we're trying to get better data collection in the ethnicity spectrum. As you can see, a lot of them don't have the ethnicity defined. But we look at the age, you know, we see that the patients are older in the registry, which makes sense, especially since we're looking at patients with ischemic stroke and some of our low back pain patients, depending upon the site. And then also just showing you the race where we have more females than male in the registry. There are attempts, by the way, in the future, we're gonna be taking a look at other, you know, do we need to take a look at other elements in terms of race or ethnicity as well as gender, given some of the trends today in terms of gender identification. So what is the patient reported outcome measure that we're using? We are using PROMIS, the Patient Reported Outcomes Measurement Information System, which is pretty widely used. You know, you hear about PROMIS-10, PROMIS-29. We're specifically collecting PROMIS-29 within the registry. So you know, has everyone here heard of PROMIS or used PROMIS, you know, in terms of raise your hands? So most people have heard about it, but really, again, the nice part about this is it uses a T-score metric by which 50 is sort of the mean for the reference population and 10 is a standard deviation. So you know, any time you go 10 points above or below, that's one standard deviation. So the higher the score equals more of the concept being measured. So for instance, a score of 50 is average for fatigue, a score of 60 means that they're having more fatigue by one, you know, sort of standard deviation. 70 would be two, that sort of thing. And then again, less if it's 40 or 30 or below. So really, it could be a desirable or undesirable outcome depending upon the measure that you're looking at. And so we can do some filtering. I'm going to show you some examples in our low back pain population. So looking here, I know this is kind of a busy slide, but these are the various elements in low back pain. And on the left-hand side of the screen is really individuals with motor vehicle accidents. And on the right-hand side of the screen is no motor vehicle accidents. I know it's busy, but looking across the top, as you can see, sort of physical functioning and anxiety and depression, you can kind of see that there are slightly higher scores that we notice on the left-hand side with those with motor vehicle collision. And then as we go down, we see further evidence of some of the slightly higher scores across some of the other domains. What's interesting here is the last portion, which is right down here, which is pain intensity and pain interference. If you notice on the motor vehicle collision side, you know, pain interference does go down some, but it kind of creeps back up again. Pain intensity sort of remains fairly high, but you can kind of see over here that it actually seems to decrease over time with patients who do not have a motor vehicle collision. So some interesting things that are coming out in some of these filters. Not unexpected. I mean, it sort of seems like something that sounds obvious, but the fact that we're able to just go ahead and do a selection by filters, either actually on the right-hand side. This is literally screenshots from the actual dashboard from the registry, so you can go in and actually toggle the filters on your own data in terms of each individual site. The way it works is individual sites will have access to their own data as well as a national benchmark across all sites. And so the nice part about this is you can go in and analyze your own data to take a look at what differences you may see based on the filters on this right-hand side, in which case we see it highlighted in green. That's the motor vehicle accident sort of filter that we're looking at. And then just sort of a deeper dive with the physical functioning aspect, as you can see how it kind of trends downward in terms of the physical functioning, with the average score of about 37 overall, although really it's lower when you look at just the 12-month follow-up where it's starting to increase in physical function, where it's actually increasing to the 42 range. So you get a difference there at the end of 12 months between this 30s versus the 40s in terms of that promised score over time when we're looking at motor vehicle collisions. So showing like some differences there, and this is just some of the power of what we can do in terms of the filtering that's already automatically available to the clinicians at their site. And what's nice about this too is that, just going back to the slide, sorry, let me just go back here again, but nice thing about the registry is that an individual clinician can actually filter their own data or they can filter at the institutional or practice level in terms of what data they want to look at. So they can look at just their own patients, they can look at the whole system, they can average it. Usually there's sort of these black and blue lines that actually are sort of indication of the national benchmark, meaning all the participation in sites versus their current site or how they do as an individual. So something that the clinicians or the site actually has access to. And we look at sort of the T-score examples with low back pain, and this is physical function and pain interference. So you can see at baseline where they start is the circles. So most of them will start in this sort of moderate physical function and impairment, and same thing with sort of moderate pain interference. And then later at the 12-month follow-up they are improving from moderate to mild in terms of the physical impairment piece as well as the pain interference is also improving. So you can see there is some improvement over time with these individuals based on some of the survey data that we have. So during this year we have worked on creating medication reports, and we're starting to build out procedure reports, meaning bio-procedures, especially for low back pain for future use in terms of the registry for the actual filters that can be done on demand. And actually, so Dr. Yang is going to go over actually some of the medication data in his presentation so you're going to see some of the filtering on that. So these are some of the exciting things we're doing. We're also trying to work on solutions to also improve our response rates, potentially looking at other, exploring other options for participants in terms of getting their data into the system. So we'll now move into the use case scenarios. We'll start off with Dr. Geis from Brooks Rehabilitation. Great. Thank you. Hi, I'm Carolyn Geis, and I'm presenting on behalf of Brooks Rehab. For those of you that aren't familiar with Brooks, we're located in Jacksonville in Daytona Beach, Florida. We're a post-acute continuum of care, so we have about 240 inpatient rehab beds. Stroke makes up about 20% of that, and so we take care of about 1,000 stroke patients a year. So it'll help you to kind of put into context what I'm going to present you through given what we're doing. We are relatively new to the registry, and so I'm going to try to present to you information that you can gain from the registry kind of in the early phases of adoption of the registry. So our PRO workflow for patients that have had ischemic stroke that have been with us for at least three days, they're eligible for the survey, and that survey is conducted in person as part of the discharge process. If a patient can't or won't finish the survey prior to discharge, then they can have an emailed survey that they can complete after discharge. So this is our aggregate data, and so I wanted to kind of show you, you'll notice that compared to the slide that was presented earlier, there's not data at 30 days or 60 days or 90 days yet. So this is our baseline data that we collected as patients are discharging. And I think for us, some of the things that I'll point out, the blue score is our score and then the benchmark score is the black dot for each of those metrics. So for us, I think that our interpretation of our baseline data was that when we look at benchmarking against the other facilities in the registry, we felt pretty reassured in multiple areas, especially where our scores were exceeding the benchmark of anxiety, depression, social functioning. So not a lot of information here except that we kind of were reassured that we're benchmarking pretty well against the other facilities. I really wanted to focus on how the data can be helpful by filtering the data at this point early on. So these are the areas that are available for filtering in the stroke registry. So race, ethnicity, age, and then diagnosis. And I expanded this diagnostic category just to show you, and this is just a partial list, but you can really filter by etiology of stroke, whether thrombotic, embolic. You can filter by location of stroke, side of stroke. So there's a lot of variation that you can really pull out here diagnostically if there's a particular group of patients that you're interested in looking at. And so what I did was I wanted to show you a couple ways that we filtered the data and kind of started to interpret some of that data. So this is a comparison by age, and so we have on the right, this is our under 65 group and this is our greater than 65 years old. And so we filtered that here in the clinical age, so this is 19 to 64, and then this of course is greater than 65. And so some of the things that I'll just point out that were interesting, we noticed that as we would expect physical function and social roles and activity were higher, showed higher scores in our less than 65 and a little bit lower in our greater than 65 year old group. One thing that I was interested in looking at was depression and anxiety, looks pretty similar in the two groups, which I think is reassuring and gives our stroke rehab team information that they're addressing mental health issues in that group pretty robustly, which we know. Little bit of difference in the fatigue and sleep disturbances between our younger folks and our older folks, and so not really different, but I think enough different that it probably warrants us keeping an eye on that across time. The next slide further filters that information by age, and so now we're looking at both groups are under 65, but we've split it out. This is right MCA stroke, here's your ICD-10 code, and then this is under 65, and then left MCA stroke. And interestingly here, what I'll point out is the difference between pain interference and pain intensity in our right MCA strokes versus our left MCA strokes, pretty significant difference there. The other thing that's interesting is that we look here at physical function, and we see a difference there. And so I think the question when we looked at this filtering is, is there any correlation between pain interference and pain intensity and this metric of physical function between these two groups, and should we dig into that and look at that a little bit more in detail? So then I'll show you, this is our filtering by age over 65, and then comparing again, right MCA stroke to left MCA stroke, and so I'll point a couple things out here. Here what jumps out is fatigue over here, 61.25, and then fatigue over here, 50.14, so pretty big difference there. Again, pretty big differences between pain interference and pain intensity in the right MCA strokes versus the left MCA strokes. And then again, this impact here, or difference in physical function, and now we see it also in social roles and activity. So I think for us, when we're looking at this type of filtering, we're really thinking about, should we look at fatigue in our right MCA strokes, pain interference as possibly areas to gather more detail and understand a little bit better, so that we can then figure out how to help patients in those areas. So from an early implementation standpoint, some lessons learned. So I mentioned it's really helpful with our own internal benchmarking, and it's a benchmark that's patient generated, which is one piece of the puzzle, in addition to PRESCANI and our own internal metrics for return to acute, it's just another piece of the puzzle that we can use to pull things together. The ability to filter the PRO data by demographic and diagnostic categories helps to potentially identify areas for further improvement. One thing I'll say is that data collection is really challenging, so it's important to have a strategy around this, and so as we've gone through the process, some of the things we've learned is that that IT collaboration, and probably more importantly, that IT monitoring, so you're keeping track of where your IT folks are with things, is really important, like regularly scheduled monitoring of your IT process. Super important to have physician champion or physician champions. Our inpatient rehab discharge planners have been really helpful in the process to make sure we're capturing those surveys at discharge. And then the outpatient follow-up is a little challenging, a little different, but educating those physicians and those teams on the outpatient side, again, has been really beneficial. I think that's my last slide, so I'll turn it over to Dr. Harvey. Hi everyone. Richard Harvey, I'm from the Shirley Reinability Lab and I'm going to carry on with the discussion of stroke, data for stroke patients. Shirley Reinability Lab's been involved with the registry longer than Brooks, so we have a little bit more data over time. So this gives us the opportunity to look to even look at a more narrow window which is at the patient level. So I'm going to actually share cases of individual patients looking at their data, which goes to the level where you can make decisions about your patient based on the data that you see from their own reported outcomes. And it gives you an opportunity to open up conversation in your clinic visit about particular aspects of the patient's life that seems evident based on their own personal reported outcomes. First of all, we collect data. Every site has a different strategy for collecting data. The way we do it at Shirley Rein is that our inpatient nursing staff, we have nurse coordinators on each unit. They take responsibility to assure that we collect the patient reported outcomes just a few days before discharge. So they sit down with the patient and go through the survey and collect it. If the patient has aphasia, they may arrange with the speech pathologist to complete the data with facilitated conversation. Now if the patient doesn't complete the survey at discharge, it will be emailed to them as well. And the patients will receive emails at 30 and 90 days as well. We had trouble collecting the 30 and 90 day outcomes. Not surprisingly, I'm sure all of you have received, you will probably receive a survey from the hotel you're staying at after you return home. And if you're like me, you will just delete it and not complete it. Well, I am sure that patients do that as well. What we've done is we've put in our EMR, we've actually put a flag so that when the patient's in outpatient therapy, when the therapist opens up the patient's medical record, there'll be this sign that comes up that says, please remind the patient to complete their registry data that they received in their email. So that reminder has helped. So we do have some data on follow-up with some of our patients. So let's go and actually talk about some cases here. I'm going to give you four cases. The first is a 72-year-old male. He had a previous history of subarachnoid hemorrhage back in 2021, for which he recovered well, although I am not sure exactly if he had residual impairment. In August of 2022, he presented to acute hospital with right-sided weakness, and the cranial imaging showed a left pontine infarction, as well as mild chronic periventricular white matter changes. He did not have any thrombolytics provided nor endovascular therapy because of his history of subarachnoid hemorrhage, and they did not find a large vessel occlusion. He was then transferred to inpatient rehabilitation at the Shirley Reinability Lab, where he stayed for 21 days. And at the time of discharge, he was independent in bed mobility transfers, dressing, and toileting, touch assist for ambulation with a rolling walker, and tub and shower transfers, and was on a regular diet with Finliquid. So not bad, not bad. Needed a little assistance at home. And then participated in outpatient therapy until November of 2022. So this is some of the patient report outcome data that we have on him for the discharge, at the time of discharge, 30 days and 90 days. So we'll go into some detail here. So if you're going to visit this patient in follow-up, you can pull this data up and take a look at it before you even go into the room. And what you can see is that the physical function score, interestingly at 90 days, is lower, which suggests that this patient's finding that their physical function has declined. Now maybe it's because therapy ended and maybe they've actually had a decline, but this is patient report outcome, so it may be a perception of how well they're functioning. In some cases, I would wonder, is it possible that this patient is starting to realize that this is not a short-term issue, but is a long-term issue that they're going to deal with, and now the impact of that physical functioning is more significant to them. It could be all sorts of things. It gives you an opportunity to open up conversation with the patient about this topic in the clinic visit. Anxiety is also increased, still below the mean, but increased from the baseline, which may also go hand-in-hand with that perception that I thought this was going to be a short-term problem, but it's clearly a longer-term problem, and that can raise some anxiety. There's an element of depression, probably a little bit higher than normal. It's pretty consistent, and fatigue is increased as well. Sleep has improved over time. Social has been variable, and then pain interference, again, has increased at the 90-day reported level, so again, what could this be? Another concept that might put this picture together is spasticity getting worse, and because of that, physical functioning is worse. Anxiety may be worse, and maybe some pain associated with it, so all sorts of things go in my head with this person. I can walk in the room now ready to start asking questions and targeting some of these areas to see what led to these changes in your reported outcomes. Next case is a 75-year-old male in June of 2022. He presented with right-sided weakness and difficulty with speaking. Cranial imaging showed an acute left MCA infarct and mild bilateral white matter changes that were chronic. This patient did receive thrombolytics, but not endovascular therapy because there was only a distal M3 occlusion. This patient then spent 14 days in inpatient rehabilitation, and at time of discharge was independent in all mobility, but continued to have language problems with fluent and required prompts, and comprehension was most effective with written support. This patient was on a regular diet with thin liquids. The patient then participated in outpatient therapy until August of 2022, so mobility was not bad in this patient, but language continued to be a longer-term problem. So here's the patient reported outcomes at time of discharge, 30 days and 90 days, and we'll look at each of these individually. You can see that the function score is fairly high actually above the mean, and that fits with the discharge level of independence in all activities. Anxiety level is below the mean, maybe a little higher than what you would expect for your average person, as well as depression, but consistent across the board, so these haven't changed. The fatigue score's gone down, so this person seems to be finding less fatigue over time. Sleep has improved. Social is a little bit variable. Just like our last patient, there's a variability on how they perceive their own social interactions, and then pain interference is a bit higher, so again, an area to ask questions about. I think overall this person's functioning well. Another question in my mind is this person remains aphasic, so who filled this out? Is this representative of their own reported outcome, or is this a close family member helping them with the survey, and perhaps they're influencing some of the answers? I don't know, but again, opportunities to ask questions about interference of pain. Are they getting out and socially interacting? Our third case is a 73-year-old female. Previous history of right cardioembolic MCA stroke in 2022, and regained functional independence with a straight cane and a rolling walker. So now in February 2023, she presented to acute hospital with a new right-sided weakness, and so this would be on the opposite side, and actually that should be... It's interesting. I think the picture is flipped, so don't look at the picture. I think... So the cranial imaging, so the acute left thalamocapsular infarct associated with the left ICA stenosis. They underwent endovascular therapy with good outcome of a TICI 3. The acute hospital, of course, was complicated by a pulmonary embolism managed with anticoagulation using a direct anticoagulant. 25 days in inpatient rehabilitation at discharge, this patient was touch assist for bathing in the shower, partial assist for bed mobility transfers, anticoagulation with a small base quad cane, and with dressing and substantial assistance for toileting. So of our patients so far, this one has the most physical impairment at discharge from IRF. The patient then participated in outpatient therapy until June of 2023, four months post-stroke. So here is the patient report outcomes. We only have the immediate... I'm sorry, the 30-day outcome and the 90-day. So this patient did not have the data collected at the time of discharge. But what you can see here is that, for the most part, the data looks pretty good, but physical functioning is quite low, as is the social activity of this patient. Looking at physical function, though, even though it's fairly low functioning, it has improved from 30 days to 90 days, and anxiety has reduced. Depression, however, has gone up, as has fatigue. So again, physical functioning doing better. Anxiety is better, but depression, more depression, and more fatigue. Sleep is not as good, which may go hand-in-hand with the depressed mood, although social activity has slightly improved and pain interference is about the same over time. So again, you can see how this gives you an opportunity to target your questioning in your clinical visit with this patient. Our last case is different than the others in that it's a cerebellar stroke. So in July of 2023, this patient presented to the hospital with dizziness, left facial weakness, and slurred speech. Cranial imaging revealed a multifocal stroke in the bilateral cerebellum, right greater than left, and the right pons and the right occipital lobe. The hospital, of course, was notable for chronic type 2 odontoid fracture, for which a cervical collar was applied. This patient spent 12 days in inpatient rehabilitation and was independent in all activities at discharge. Now, though this patient was independent, I suspect that this patient may still have had some ataxia at discharge. They participated in outpatient therapy until September 2023, about two and a half months after stroke. So you can see here the patient-reported outcomes, and going into more detail, you can see that there was a decline in physical function over time. Interestingly, it's possible that although this patient was independent, they may have realized that this ataxia does interfere with their life more than they had originally anticipated. Anxiety is less, depression is less, fatigue is less, so these are all good signs, but their social activities have reduced and their pain interference has gone up. So again, opportunity to talk about pain, origin of that pain, where is that coming from, and to have a more targeted conversation with the patient based on their own self-reported outcomes. So I hope I've shared with you how you can use individual data to get some particular information that the patient themselves have revealed in the patient-reported outcomes, and then utilize it to target your questioning so that you may have a more efficient clinic visit and address issues that are directly important to the patient and their lives. Thanks. I'll turn it over now to Dr. Aaron Yang who will lead us into low back pain. I feel like that was like a Star Trek into the next frontier. Thank you for that. So I'm going to present some cases on low back pain on the outpatient side. Some of this stuff, you know, I do spine intervention, but I think there's a lot of things you can take away from it. And I like to see myself as a very practical guy. And I think if I was in the audience, I'd want to say, what can I take away? How is this practical to my practice? And I think this is something that doesn't have to be just at a large academic institution. What can you take away if you have a small private practice or if you're in a multi-group practice? And so with that, I just wanted to first share how we get some of this information. A lot of it comes from the EHR. And I think it's always intimidating when we initially think, how are you going to get this information when you're busy in clinic? Do you have to spend a lot more time changing up how you're doing things? And this gives the freedom to really just continue your flow, knowing that patients will have the opportunity to fill this out. And so they're getting sent an email from the EHR, which is within that. But I think in the future, you know, you can have QR codes, easy ways. We just have to figure out the challenges of trying to link it to their EMR. But right now, I think it's very low workload on the clinician part to gather this information. I think this very much mirrors what Dr. Wang mentioned about demographics. In terms of our patient population that are in the registry, white, female, over 65, these are predominantly the patients that we're collecting data on right now. And so I think, you know, when I thought about how to present this side practically, I don't think I'm necessarily focusing on, okay, I did this intervention and here's the outcomes. We have plenty of, especially in the interventional spine world, studies on efficacy or effectiveness. And I think more about how is this going to practically help me in a patient visit. And so, again, just briefly going over the details, this was an 84-year-old female who came to my clinic who had left axial back pain. We started off physical therapy and Tylenol. And we proceeded to go through the process of doing a lumbar radiofrequency ablation. Again, I don't think it matters if you're an interventionalist or not, but how has this information helped me? And this patient, as you can see, has filled out outcomes up to six months. And so sort of diving deeper into the information itself, again, the difference between the first slide here and the other slide is really the y-axis and the PROMIS scores. So you could see, okay, I did an intervention. Their physical function is better. Okay, what about their pain? Okay, seems to have improved. But I think a lot of these other components such as sleep, social, anxiety, depression, fatigue, these aren't necessarily things that I'm really focusing on, on follow-up visits, to be honest. If I do a procedure intervention in a busy clinic, I'm seeing them back and saying, hey, how are you doing in terms of your pain? And I think these are a lot of other things that I don't talk about, which Dr. Harvey hit on, is like these are important talking points maybe that instead of just focusing on how's your pain, you can see, well, they report that. So their mood or depression has not changed. Their fatigue has went up. So I would not have gleaned any of that information. In fact, when I go back to my initial note on this patient, the patient said they're doing fantastic. So I think what am I gleaning from that? Okay, they might have just meant their pain or function, but I didn't get any of this other information. So I think this, again, can bring up a lot of things we're just not focusing on on patients. And as you know, in the spine world with any type of spine condition, it's never a one-time deal. These are patients you have long-term relationships with. And so I think there's a lot of other things that we can talk about of optimizing patient outcomes or as physiatrists looking at the whole patient. And so I think this has been helpful and eye-opening to say, okay, there's other things that we are tracking. And a lot of it, to be honest, is passively. I'm not emailing the patient saying, can you fill this out? These are patient-reported outcomes that patients say how they're feeling. So I think there's a very unique perspective to that. And again, something similar. This is a second patient, 66-year-old female, back and left leg pain, did not undergo any intervention, but just physical therapy, tried some meloxicam, and I think that as you dive in deeper, you could see the physical function. Some of these numbers, they fluctuate, but again, just having that information can help in seeing what trends up or down. And again, just going through social, sleep, pain interference, again, going back to the chart. She's doing better, she's doing much more, but then to be able to visualize this, I think you pick up a lot of discrepancy sometimes. And as you know, even in the pain world, when you do an intervention, we ask for pain scores, and then we try to clarify the pain scores, and the recall is all over the place. And so seeing this actually visualized can see how things are trending better. And as Dr. Huang mentioned, something exciting that we've been able to start to do is also collect medication usage data. And so what you see here is particularly on anti-inflammatories. And so the blue line are patients who have taken anti-inflammatory. There is an orange line that you really can't see that's covered by the black line, but one of those lines is an aggregate data, and the other line is based on Vanderbilt itself. And they really mirror each other because a lot of the data right now we have so far is based on Vanderbilt data. So you could see here that with anti-inflammatories, we're seeing a pretty sharp drop in terms of pain intensity usually within the first six weeks. And you can see their physical function trending up. So if I'm just focusing on those two things, it's interesting because it's unique to our patient population at our institution. We automatically just pharmacology-wise, we're like, well, anti-inflammatories act quickly. But as someone alluded to, I think these are truly what patients are reporting, and we can't just make assumptions because we learned this in medical school. This is how it works, duh. Here's actual information that backs that up. And so these are things that are going to help us as we try to push the field forward of having data, and we can actually look and say, these are patient-reported outcomes that are relevant to our field instead of just relying on pharmacology knowledge from med school or things like that, or efficacy studies on procedural intervention. We can actually track these outcomes, which makes it really special to our field in particular. And then lastly, again, this patient came in, had an acute lumbar compression fracture, some mild radicular pain. We tried a neuropathic medicine, as well as a brief course of hydrocodone. They participated in this survey up to a year. And so you can see, as we dive deeper, you can see the different trends in terms of, again, physical function, anxiety, depression. You can see, for example, depression is bookended by high scores initially, and then at 12 months. But their pain interference improved. And so again, just not trying to get into the weeds of this, a lot of talking points that, again, I think we can often miss if we're just focusing on what we think patients want to tell us. So looking at medication data usage for gabapentinoids, we can see that it's more of a gradual decrease in terms of pain intensity, usually peaking at six months. And you see with anti-inflammatories, you see it peaking at six weeks. Again, some of these things we just intuitively think we know, but it's cool to see it actually laid out like this, and how patients are truly reporting how certain interventions like medications affect them. So in terms of, you know, what am I excited for? I think right now, my biggest hope is to have more institutions, to see more variety of institutions taking part. Because not only can we see how we're doing as an institution, but also looking at how other institutions are doing. And this may lead to collaboration. You know, if I see Shirley Ryan doing some really great things, and I'm wondering, I mean, I think it can also lead to increased communication and more departments working together. We are the field of physiatry. We can share more information together. Again, like how much people can get involved. If you're really passionate about this, you can be at the bedside encouraging all patients to sign up. If you're really busy and you're afraid this can interrupt your clinic, you don't have to, right? And so I think there's different levels of involvement you can have. It's gonna be very interesting, and I'm really excited to start seeing data regarding procedural outcomes. And then, again, it's really exciting because physiatry is doing something like this in terms of the academy and investing and having outcomes specific to our field. Not a lot of institutions or academies are doing this, so we can really lead the charge. Especially in a field where we always complain, well, there's not enough data. We all have an opportunity to really partake in this. And so some other things that I hope in the future we can improve upon from the current process is easily identifying a registry participant. I think it would be really great if a patient walks in, and again, I talked about how the conversation's gonna go, knowing that they are already part of the registry, so then I can look at that data easily. Again, the medication and procedure mapping, I think is gonna be really exciting. And then, again, even if you say you're gonna participate, it's getting the patients to actually participate in the registry. And I think we've been brainstorming lots of different ways to have increased enrollment of patients. And a lot of patients, I think, are very interested in letting the provider know how they're doing. And so I think this is a great opportunity for increased communication that way. So thank you, and I'll bring, have Dr. Lee come up. All right, last section, but I'll share the low back pain registry at the Shirley Ryan Ability Lab. So in terms of our workflow, whenever a new patient checks into our outpatient MSK clinic, if their chief complaint is low back pain, the front desk staff gives them a iPad with the APM and our registry survey already pulled up. And so they're encouraged to kind of fill this out while they're waiting in the waiting area. Even if they're brought back to the clinic room, they're still waiting for us to get to them. And so they're encouraged to fill that out. And ideally, once they have it completed, our nurses or MAs then print it out for us to review prior to going into the patient room. But if they don't get a chance to complete it, then they're emailed the survey like the other cases, and they have seven days to complete. And then follow-up surveys are sent at six weeks, three months, six months, and 12 months. So four cases I will share. So first patient is a 67-year-old male who presented in July of 2022 with recurrent low back pain and left radicular symptoms after a golf outing. It was presumed to be from a left L45 radiculitis from a discogenic etiology. He did also have a history of a disc extrusion. At the baseline data, he was already on ibuprofen and Flexeril, and we prescribed a more focused refresher, McKenzie-based lumbar spine PT with no new medications. At the six-week visit, he actually reported natural resolution of the pain and admitted that he didn't really proceed with doing the formal PT. And he didn't come to see me afterwards, but he did manage to complete the three-month and six-month surveys. And so just delving deeply into each of the sections. From the first visit to six months later, you can see that physical function and fatigue improves. There's slight worsening of anxiety and depression, but we don't know details of why that would be. The pain interference and social participation also improves. The sleep disturbance initially improves, but then returns back to baseline. Second case is a 37-year-old female who presented in May of 2023 with low back and right leg pain, presumably from right L5S1 radiculitis, again from discogenic etiology. At the baseline visit, she reported already having done some general PT, was taking NSAIDs as needed. At that time, we prescribed more focused McKenzie-based lumbar spine PT. Again, no new medications at that time. At the three-month visit, she reported no significant improvement with PT, so we proceeded with ordering the MRI at that time, followed by a spine injection, a transformal epidural steroid injection, which was done in October of 2023. And we didn't have any additional clinic visits, but she did fill out the 12-month survey, thankfully. So from the first visit to 12 months later, we do see some worsening physical function, anxiety, and fatigue with PT alone initially, but all had improved by the 12-month. Maybe was it due to receiving the injection? I don't know, because she didn't follow up with me at the 12-month, but just interesting data to see, and depression score remained stable. Sleep disturbance, social participation, and pain interference also improved for this patient specifically. Third patient is a 74-year-old male who presented in June of 2022 with low back and bilateral leg pain, presumably from lumbar stenosis with L5S1 radiculitis. Didn't have baseline data or survey filled out, but at the three-month visit, when we looked back, we had prescribed PT, and at that three-month visit, he had reported at least 50% improvement in the back and leg symptoms, but at this visit was reporting new onset of left-sided neck pain, so a separate pain from the low back pain, and so at that time, we suggested adding cervical spine PT. Didn't see him for the six- and the 12-month visit, but he did manage to fill out the survey. So from the three-month to 12-month data, he reports worsening physical function, fatigue, and anxiety scores, but it's unclear was this because of the new onset of neck and arm symptoms, or did the back pain recur, because at the three-month visit, he had reported improvement of the low back pain. Depression scores overall remained stable. There was also worsening social participation and pain interference, while the sleep disturbance score remained stable. Last case is a 62-year-old female who presented June of 2024 with axial left low back pain from lower lumbar facet arthropathy. At the baseline visit, she was already on Tylenol and tramadol PRN from her PCP. She reported having allergies to NSAIDs and had already failed a course of PT, and so we ended up ordering MBBs and RFAs, which were then completed in July of 2024. I didn't see her at the three-month visit, which she did still complete the data, but at the four-month mark, when she did come back, she reported near-resolution of the axial back pain. So from the first visit to three months later, we see that the physical function and fatigue improved. There was slight worsening of anxiety and depression, but she definitely did endorse increased life stressors and work stressors at our initial visit, so that could be a contributing factor. And social participation and pain interference improved. There was some worsening of sleep disturbance, but possibly could this have been related to the increased anxiety and depression. And so just for time's sake, I'm not gonna repeat everything that the other colleagues have mentioned about the utility and value of the registry, but I think one thing to add is sometimes patients return to our clinic and say they don't see any improvement of their symptoms, but it's been helpful to review some objective data with them, like for instance, through physical therapy notes when I say, hey, but it looks like this score improved, your ODI score improved, then they realize, oh, maybe I do have some improvement. So I think this is another tool that I can utilize in clinic if I pull it up real time and say, hey, I'm seeing this trend, to help them track their progress a bit better, or if they're not progressing, open up more discussion about different things that we can try. So I'll give it back to Dr. Cohn. So I think you've seen the examples from some of our sites in terms of how data can be used at both sort of an aggregate level and individual level, and I think that's kind of one of the key things we wanted to really emphasize with the registry is that, again, we call it a clinical data registry. So on the one hand, we wanna make it useful to the clinicians at the patient level. So you're seeing a lot of those with these specific examples. How can this enhance your clinic visit, help you track a patient's outcome and trajectory over time? Are they on track, are they not on track? We saw that first example with Dr. Lee where that patient was actually doing better, didn't actually need the PT, started to continue to have sort of natural resolution, right, physical function, some of those other things were improving. The other patient had had the new onset, had the low back pain, which got better, but then the neck pain got worse, sort of wasn't doing as well with sort of the physical function. So was that sort of getting worse? Sadly, she didn't have further follow-up with the patient, but it would have been interesting to say, all right, well, what happened with that neck pain over time? Did that get worse? Did that really impact some of those scores? But you can see even from the examples of the low back pain versus even in the stroke realm, Dr. Harvey's more specific examples, you know, where you're seeing maybe worsening of anxiety, depression in some areas, or maybe pain, and is pain influencing physical functioning? How you can use these scores in an individual patient level? And then we pull back a little bit with just even Dr. Geis' initial site data. You know, looking at the site and maybe even drilling down, you know, by the various filters that we have existing, you know, are there laterality differences? There seem to be, the right-sided strokes, is this right? Where the ones that are slightly worse are the pain versus the left-sided, did I recall that correctly? So it seems like that is one of those things that we're seeing is like, hmm, there's some differences in some of the data. And how can we look at that, even the outcomes by age, the differentials by age with some of the changes in pain, and taking a look at those things and seeing are there age-related differences? Are there laterality-related differences? You know, and seeing the motor vehicle collision, which was, you know, probably obvious, right? And we think that they might do worse. You know, other covariance interestingly in there, are there litigation and other things involved in there? We actually do have filters based on that too. And the registry has evolved over time. You know, initially we were doing, we were approaching it as more trying to be purist in data collection, you know, excluding motor vehicle collisions, excluding cancer, excluding, you know, litigation. We've actually went back and said, we should include that, this is real life. Real life is, you're seeing patients with these things, with cancer, with motor vehicle accidents, with collisions. And then we looked at stroke, where initially we're just ischemic, and we're gonna pull back and include other stroke types. Because really what it boils down to is you at a clinical level, you know, we wanna be able to use your data, we wanna be able to show data to you that's useful, be able to compare, and allow us to do this. So, you know, down the road, you know, obviously a lot of these diagnoses are strictly by ICD-10, but we can start doing the ICD-10 groupings, like for instance, our medication data, we gathered all our sites, we looked at the sites, how do we wanna do medication classifications, gabapentinoids, you know, opioids, you know, non-steroidals, not just are they on meloxicam, are they on salicoxib, you know, so we're trying to group classes, we're trying to do groupings, to help with sort of analyzing the data, and showing the value of like what potentially interventions can do. And then you can look at the site level data, and even compare it to your national aggregate data, how is your site doing? I find it fascinating looking at different site demographic data, you know, for instance, Shirley Ryan is very similar to Vanderbilt, in the low back pain, in terms of like fascinating, that almost spitting images, in terms of just demographic cohorts, especially with regards to gender and age, very, very, very similar, and now it's like, well, maybe we need to think about, do we need sites that are, you know, we're talking two major urban metropolitan areas, with their sort of tertiary care, what if we look at other sites that potentially might have different demographics? That could actually add to the richness of the registry, so we're thinking about how we can use this data, create sort of trends, and aggregate data, to actually take a look at what are the larger trends, what can drive and improve care for patients? Maybe that can drive clinical guidelines again, drive outcome measures, or sort of quality measures, that actually measure more of what we do, than controlling high blood pressure, and managing their weight, and smoking cessation, which yes, are very important things, but that's not what you do as a clinician in physiatry, we were trying to make a difference in function, and a lot along these lines, so really, one of the goals is we want to obviously next year start publishing some of this aggregate data, and what we see here, we're actually gonna have a couple posters, which are actually already up in the kiosks, but in kiosk 10, with Dr. Armando Messiano, and also Dr. Andrew Gordon, created some posters regarding sort of gender-based results from the overall aggregate data from the low back pain registry, as well as age-based results from the AAPMNR registry with respect to stroke, so those are actually posters that are actually available in the registry, Dr. Yang has also given a presentation, or actually gave one in the research hub, in terms of Vanderbilt, and they're showing the medication data and outcomes, so more to come on that, as far as if you want to schedule a meeting and participation in the registry, obviously the QR code up there, we do have to review contracts, business associate agreements, registry participation, and really just understanding the needs for those that might be interested. You know, we wanna encourage participation, it does require institutional, obviously, collaboration, as well as IT collaboration, and we can help you along the way with that, because we've been doing this with other sites, so we can sort of help with a roadmap, a pathway to help remove some of those barriers as we look into sort of involving more sites with the data registry, and giving you tips, like what you've seen here from Dr. Geis, and clinical champions, IT connections, things like that from Dr. Yang, in terms of like, you know, how can you incorporate it more potentially in your practice, or how you can look at the data, so more to come on that road, we hope to continue to evolve the data, make it more useful, make it more robust over time, as we continue in this journey. And we'll go ahead and ask panels to come up, if you guys have questions, please feel free to come up, and we'll try and get those answered. All right we got a quiet audience here so any questions from folks online or I should say not online in the room because I don't think there's any online live chat for this so yeah. So, we do have some centers we're still working on getting in line, is that correct, Kavitha? I think some of them, we have some of the Encompass sites, right? For ischemic stroke, we're working through some of the data requirements and data mapping for some of our, for these Encompass sites, which is actually going to be large because if we can get Encompass, which has a lot of the inpatient rehab facilities, you know, that's another area that we can tap into for, especially for ischemic stroke, to get some of that data in. Okay. Sorry, I don't want to make people nauseous, so we go way back through some of these slides. If we go back to the beginning here, where it shows you the filters that we have. Every site's going to have filters available to them. Each site is going to have the ability to see their own data, as well as aggregate data. I know some of this is not quite there, so I'll just kind of see. These sort of slides. You'll be able to filter, automatically filters are going to be there for various demographic data, as well as you can filter by diagnosis, you can filter by medications, procedures is coming, we're doing the procedure mapping now, so those are the kind of things you can do. There's sort of different reports that you can run. You can run aggregate demographic data report for your site, you can actually break it down by clinician. So you as an individual clinician can kind of look at your own data, how that compares to your institutional's data, how you compare it to the national data, so it's kind of nice to be able to do that. You can toggle, even if you want to group clinicians, you can group two different clinicians, so there's various ways that each site can toggle their data. You can look at, you know the nice thing is you can look at the actual PROMIS scores, which is this is more PROMIS reporting data, and then really on the other side it's more demographic data reporting that you can have. There is actually different snapshots or dashboards that are available, more demographic focused, more PROMIS focused. There's even PRO response rate focused data, so you can look at your response rate for PROs by the type of survey, your percentage response rate, so those are all things that you have at your disposal, so it's very nice that you actually each site can then drill down to the level that they want, clinician versus obviously whole site wise, and of course the ability to drill down to an individual patient level to see individual patient responses to the surveys in addition to the actual PROMIS score data over time. So those other things that we're showing, the mappings, you can kind of see, you can actually get sort of a map that looks like this for an individual patient for a PROMIS score, so you can kind of look at that as well, so it's very useful. And actually you can even see their individual responses, because there's really in the survey itself the baseline, you're looking, we're not only doing PROMIS, we're doing additional questions that were vetted by both the stroke groups as well as the low back pain groups in terms of what were relevant data points, so for example motor vehicle collisions, presence of litigation, those are extra questions that were asked. On the low back pain side, ischemic stroke side, they're asking about anticoagulants, both they were also asking about medication compliance and some other things, so there are other additional data elements that we're also hoping to map and do into filters as well, so this is all part of the data collection that the patients are actually completing in terms of the surveys. So there's patient collected data, and of which PROMIS is one, plus additional questions and of course EHR collected data, which is a lot of the demographics, the procedures, the ICD-10 diagnosis, other demographic things that you can get from registration and from the electronic health record. But can I look at site data for Vanderbilt or Brooks? So no, so the answer is can you look at other individual site data, the answer is no. So each individual site will only get access to their own data plus the aggregate in terms of just the dark line. Mark, what about, can you clarify, can the patients themselves access their PROMIS? So this is a good question, so Dr. Lee's asking about can the patients access their own data, so at this time not yet, but we are actually potentially looking into some exploring additional modules that will allow patients to access their own PRO data, in addition to allowing it easier to send PROs out. So the hope is as we move forward we're working with the board in terms of finalizing some additional enhancements which include text messaging of surveys to help improve the response rate, giving patients the access to their own data. My own hope is we want better EHR integration, meaning that we want this back into the clinician's hands more easily where you don't have to log into a separate dashboard to get their data, you're actually hopefully, my holy grail is to get it back in so that it's like a little dashboard much like prescription drug monitoring, you know, your little PDMP survey, you know, usually most people have PDMP access within the EHR where, you know, you're not logging into PDMP to check their opioid usage, you can kind of go in, it's already a dashboard on that patient, you pull up John Doe, you're going to see their PDMP data on the opioids prescribed, who prescribed them, what pharmacy they go to, so in this case my hope is to be able to like get, you know, this data, you know, these little shots that were got at a more patient specific level back into your hands to be able to see it on site for your patients, so that's ultimately our holy grail to kind of get there so that eventually we can get to this point where it's much more integrated into the EHR, of course it's always a challenge because everybody's got different EHRs, but, you know, this is where we would like to go or at least a portal where they're doing say, or at least in between where the portal makes it easier for you to see the patient's data, you know, in addition to what's currently available on the dashboard, yes, a couple more questions. So right now, so our current registry vendor is ArborMetrics and at this time what happens is as soon as we get the demographic data from the EHR for a patient that's considered a candidate. So usually what happens is every site usually has a push or a report or a pull shall we say where like it's sending data to the registry to say, John Doe is a registered participant. They should be in low back pain. Here's John Doe's email. Here's the first visit for John Doe. We're going to send out the email based on that. And so the email comes out from the registry itself, not from the sites. We're hoping one of the later integrations is could we integrate that with patient portals at various locations? So lots of things we're starting to work on to try and make it a little bit easier for the EHR. But from the standpoint of line, I thought we came up to the conclusion with 2FA was working, right? Did we come to that conclusion? Yeah. Yeah. Two-factor authentication, which you guys do. So, Arbometrics would send the email out. Of course, that would be requiring that the patient has an email address. So again, so yeah, that's sort of like, yes. Right now the registry sends out that email. We're working on a process now to make that email a little bit more robust, looking more customizable from the sites. So it looks more like, oh, okay, so Vanderbilt's sending out an email, or you know, Brooks Health is sending out a survey email. And hopefully less chance for delete from the user end, you know, all right, you know, tell us your feedback today. So it's like, you know, we get a billion of these. So it's like, it's kind of like trying to reduce that, or at least send it in a text form and to sort of help maybe enhance some of that ability for patients to complete it. But as far as that, yes, sidebar we'll converse later, but I thought we've come with this inclusion of the line of health in terms of that two-factor authentication piece. Yes, there's another question over here too. I think, you know, I think it depends. I mean, I think a lot of cases it could be something we can explore. I think some more conversations with some of our AAPM&R sort of reps, registry reps, which could be this here right now, is one of our representatives, also Beth Radke. So, the two of them are sort of our point people at the AAPM&R to talk more about that. But, yeah, I mean, we've had smaller practices involved, too, you know, in terms of in the registry before, so it can be doable. But, yes, it requires usually champions, someone to help out, an IT sort of resource in there. But, yes, can you do it? Yes, and it is something that can be done. And we have a couple more questions in the back here. Yes. Oh, absolutely. That certainly has the potential out there. No question about that. You know, for our purposes, we're trying to get more data, more responses, get more robust, so we can really get a lot more data points to really then be able to run models like that or to do analyses like that. But absolutely, that's always something you want to be able to do with that data. And that's why you really want to collect it from multiple sites, you know. I think the key here is getting multiple sites, getting multiple points of data, getting that data back in the clinician's hands so it's actionable, getting that data back so the patients see that you're using the data and then they'll want to fill out the surveys more. So this is a chicken-egg kind of process with this. So we want to, you know, get more data, get data back in the hands, have it more, you know, get more realization from providers, from patients, that, yeah, I should fill this out. My clinician's looking at this, so this is important for me to do. It's also important for the specialist to know how we're doing, you know, and also even pointing out things that weren't even there before. Hey, you know, look at what's going on here with depression, what's going on here with physical functioning, what's going on here with pain interference, why are these things getting worse? You know, or better, you know. So I think, you know, these are opportunities that we have as we look at this rich data from PROMIS. And just even just taking a look at PROMIS, because, you know, sites are gonna have many other outcome measures that they might want to utilize too. And for us, for the purpose of the registry, we're trying to find one common, sort of, patient-reported outcome measure to kind of be the level set. And then certainly, you know, obviously customization down the road, but we have to, this is sort of the, we gotta start somewhere, we gotta find a common data set that could be applied, as you can kind of see here. Like it does, PROMIS 29 does apply pretty well. I mean, is it directly measuring what's happening in stroke and low back pain? Not necessarily, but it actually gives you so many more data, rich data points on what's happening with the patient, in addition to what you have in your EHR. You kind of get a better sense of what's happening with these patients, so what kind of difference we can make, and I think that's very important. Oh, absolutely. We have one... Sorry, I'm going to... Dr. Schlieber already asked her. I'm going to bypass him. There's a gentleman in yellow back there. He also had his hand up. Yellow T-shirt. Did you have your hand up, too? You were here. Those are great questions. We can follow up with you later if you talk with Kavitha, but yes, you don't have to be both. Actually, Shirley Ryan's one of the few that has both, but many sites are only single site participation, meaning just ischemic stroke, just low back pain. And down the road, some of the other things we haven't really talked about today, but looking at other diagnoses. One of the big things down the road we're talking about spasticity, next step in the registry, cross diagnosis. Now we're not looking at diagnosis specific, but condition that crosses various things, even to the realm of pediatrics versus adults. So that is probably our next foray. Again, we're still working on these other things. That's just possibly, that's on the roadmap. It's just a question of when. But anyway, so Dr. Saliba, you had a question or a comment. I know we're over. Thank you so much for your time. If you guys have any other further questions or anything please feel free to stop on up, but thank you all for coming.

AAPM&R

Data Registry

patient care

low back pain

ischemic stroke

electronic health records

PROMIS tool

patient outcomes

benchmarking

clinical data

rehabilitation