Good morning, or good afternoon, good evening, depending on what time zone you happen to be tuning in to. This is the Academy's talk, How a Physiatric Clinical Data Registry is Changing How We Understand Our Patient Populations and Their Outcomes. We'll be sharing tales from launching the AAPM&R's Registry's Low Back Pain Ischemic Stroke Pilots. Thank you for tuning in today to this fun new platform we're all getting used to and really having a lot of fun with, you as I'm sure as well, all of us. We have a great lineup today. We have Dr. John Lesher from Carolina Spine Institute. Dr. Lesher is spearheading our spine efforts. He's our content expert and our leader in the spine registry, and has really guided us in developing the best practices around spine and guiding our pilot sites and our pilot. Dr. Alan Novick was very graciously volunteered to be our stroke and our neurorehabilitation leader. He'll be helping us. He has been helping us guide the development of our neurorehabilitation pilot and specifically to ischemic stroke. Dr. Novick is at Memorial Hospital in Florida. And of course, we wouldn't be here without the tireless and passionate efforts of Kavitha Narakunda and Beth Radke, our Academy leaders, who have been helping us develop the registry over the last half-decade. And of course, myself, Dr. Michael Hitzakis, I'm the chairman of the Registries Committee, and I've been on this journey probably five or six years now as we have developed and have gone through many iterations to get to the great place we are in right now with the Academy's registry. What we hope you would walk away from today is to really understand the advances in the design, build, and launch of an Academy clinical data registry and its value in healthcare. As you may or may not know, registries have become a vocabulary of modern medicine. More and more specialties every year are embracing clinical registries to provide value to what they do, and you'll be hearing more and more about registries in the next decade and will become common practice in the practice of medicine as we move forward. I've been participating in a councils of medical subspecialties conference in Chicago every year or two. And every time I participate, the numbers almost double the people who are either developing registries or interested in registries. As part of implementation of a registry in clinical practice, workflow changes inevitably occur. We'll be sharing some of these insights and challenges from our pilot sites and their experiences capturing common physiatric data elements for low back pain and ischemic stroke. And part of a registry is collecting patient-reported outcomes, and we'll be discussing the imperative and the usefulness of collecting patient-reported outcomes in a physiatric clinical settings. Then once we have that aggregate data, how do we use it? How do we improve our own clinical practice? How do we advance physiatry's position, both in local and national quality improvement initiatives and in our standing in the medical landscape? We're going to break this talk up into four segments. First we're going to talk, I will talk about why the Academy built a registry, some of the foundations, and how we started with ischemic stroke and low back pain. Dr. Novick will talk a little bit about what the research tells us about the value of registry data, how it impacts clinical practice and the science of our patient conditions. And then both Dr. Novick and Dr. Lesher will dive deep into the research around ischemic stroke and low back pain, what's been done, what registries have currently been initiated, and some of the experiences from our pilot sites in those areas. Then we'll wrap up with a treat at the end. We're going to see the registry interface live because it's ready and it's operational. So we're going to have a video that's going to show just what that registry looks like. In building a registry like this, we start with our foundational beliefs. As physiatrists, we're here for our patients. And so when we build a registry, registry must be first and foremost here for our patients. I like this Institute of Medicine definition, it defines quality of our healthcare as a degree to which health services for individuals and populations increase the likelihood of a desired health outcome and those that are inconsistent with current professional knowledge. As we build this registry, we have to ask basic questions as we create the foundation of a registry, which is, what are the desired outcomes in physiatry? And are our desired outcomes the same as our patients' desired outcomes? And as we develop those, do we actually have the ability to systematically measure them so they're consistent and uniform and meaningful to patients? Can we determine what are the best practices needed to optimize those desired outcomes once we measure them? And then once we develop our best practices, how do we know when we're actually reproducing those best practices day after day in clinic? A registry supports all stakeholders. I mentioned patients in terms of the evidence, improved clinical practice and quality, but it needs to also support us as clinicians, as physiatrists, in our needs and our day-to-day ability to see patients. It can't slow us down. It has to enhance us and improve our ability to develop as professionals. Of course, it has to enhance the specialty. We need to be able to have better knowledge of the disease conditions that we're treating through registry data. And of course, it needs to enhance the academy. We need to understand and use data to advance the value of physiatry in a highly competitive market space. Starting with the patient, why build a registry for patients? Well, a registry allows the patient's care to be focused on outcomes that are important to them. Time and time again, a patient comes in and says, I heard. We have that as our outcome. But when we have registry data, we have more granularity to that outcome. We can say to the patient, you've mentioned all these physical functioning. Let's discuss what's most important to you. And as we do that, we've engaged the patient in their own recovery. Because they can look at that data visit by visit and start to understand which interventions are helping them more with their various desired goals. It also reinforces their feeling that we're here to support them. We're asking them what's important to them. And then we're following up with asking what interventions are helping them. And we're following up with surveys that make sure that we know if they're improving or they're not improving. And at the same time, it collects a more comprehensive picture of our patient preferences and any barriers they have to recovery. What about the physiatrist? How does it help us day after day? Well, registry data will help provide a goalpost for us to anchor our treatment decisions. If we decide to do an intervention, a referral, or a procedure, we can then look exactly at how those interventions improve the outcomes of the patient after a next visit. Or after a number of visits, we can see what impact the various things that we do are having on our patient. Functional recovery. In terms of quality practice, who doesn't want more engaged patients? If a patient has information, if they're following their progress, by definition, they're more engaged. Having a more engaged patient makes a more satisfying patient interaction for the physician. And importantly, identifying barriers to recovery. Many times, and early in my practice, I've gone headstrong into interventions, treatments, only to find those patients aren't recovering as fast as I would expect. And then I realize there's a barrier that I haven't identified. Having a baseline study that shows also promised domains of depression, fatigue, anxiety, and sleep allows us to identify barriers up front so we can identify them and treat them early and throughout our care. And we can see whether those barriers are improving with sleep interventions, interventions meant to decrease depression and anxiety and discuss those with our patients. It also helps us as a physician in our own personal career development journey. Audit and feedback allows us to improve our professional practice. If we treat a high proportion of chronic low back pain patients, we can then slice the data by ICD code or groups of patients and see whether things that we learn in our conferences, in our textbooks, allow us to be better physicians to see quarter by quarter whether groups of our patients are doing better as we improve our skill. And as we develop clinical practice guidelines, we can align our medical care around what's already been proven to be effective in improving outcomes using registry data. And as that develops, you can see the obvious next step is to use that process to create a new form of maintenance and certification so we can actually learn and better ourselves and get credit for that in our maintenance and certification process. That's one of our aspirational goals for the registry. And then dreaming a little bit, we're thinking about how can we improve the specialty through quality improvement partnerships. We can have multiple centers that are becoming centers of excellence in a certain area. And we can then compare and learn and develop practice guidelines with multiple centers where we can compare our outcomes with people who may be doing something innovative. And then those quality improvement partnerships allow all of us to rise. And as you all heard, the inevitable coming of value-based medical environment. A value-based medical environment relies on data. And data has to look at outcome, and it has to look at cost. And that's our new world, like it or not. And we can't get there without a registry of data. We will be lost with those specialties who are proving that data. And those specialties, such as orthopedics and neurosurgery, internal medicine, all these groups are looking towards and implementing registries. And we use the registry to help support the value of physiatry and of competitive both organizational and local market ecosystems where we live in practice. For the specialty, we need data to support what we do. We've had mixed experiences with clinical practice guidelines and randomized controlled clinical trials. These are good. RCTs are good for measuring very specific and strained interventions. But the essence of physiatry is a multidisciplinary, multifaceted approach. This is where RCTs are very difficult and impractical, if not very expensive to implement. A registry observes and tracks real-world clinical practice. It collects much of the information that we need to assess patient outcomes and also clinical practice in a generalizable way that can then be used to develop care guidelines out of those practices that can be then modeled, tested, and disseminated, which creates an enhanced QI initiative environment. And as I said earlier, allows us to start to think about how we want maintenance and certification to look in the decades to come for our specialty. And again, we can use these quality measures to then help guide CMS and our insurers on how we want them to assess our performance, not waiting for them to provide us their markers of success, because that's coming. As CMS and other insurers are thinking about ways to assess us, we need to take the ball and get ahead of that and define our own quality measures. You might ask, why did we pick stroke and low back pain? Why not pediatrics? Why not cancer? Well, we did this really by the numbers. We couldn't boil the ocean. We couldn't start with everything. Even though it's our intention to create a registry that serves all segments of our diverse physiatric specialty, we had to start somewhere and we had to start constrained in order to make a practical pilot project to get our process perfected. As I said, we did it by the numbers. Two-thirds of our members identified practice areas, at least five clinical practice areas. And as you can see, the top five or six are musculoskeletal medicine and neurorehabilitation. And so statistically speaking, the most obvious choices were the most common diagnoses. Low back pain predominates musculoskeletal care, and ischemic stroke is straightforward and can predominate neurorehabilitation. We picked those two specialties to create our pilot. But we have a pipeline we've identified to start bringing on different specialties and rounding out those, because understand, this is a pilot, but it's a fully operational pilot. We are actually developing an actual registry with data, with PROs, with outcomes that we're going to have dashboards, and this will be a fully functioning registry in those constrained diagnosis groups. And as we get moving, we will then broaden those to larger diagnoses and clinical areas. So now I'm going to turn it over to Dr. Novick, who's going to tell us a little bit more about what the research tells us, the impact registries have on clinical practice. Thank you, Michael. I think Michael did a wonderful job on explaining our rationale on why we need a registry and what a registry can do. But we wanted to look at the literature a little bit and see, is it really accomplishing what we're expecting? Can registries improve quality? You know, quality improvement is all of our goals, and it's really an application of a defined improvement and an intervention and seeing if that intervention is making a difference. One particular study we identified was a literature review, PubMed, Embase, and Cochrane databases over the period of time from 95 through 2017, looking at different registries where they relied on structural data. There was a comprehensive QI intervention performed, and there was impact on improving clinical or patient-reported outcomes described. They found 21 articles that met that criteria. And of those 21 studies, eight showed statistically significant improvement on a particular outcome directly derived from data from the registry. So that's really an important factor to show that the data collected, if used appropriately, can impact outcome. Now, in that study, they found different facilitators for the quality improvement. Among those facilitators was frequent feedback. In order to make an impact, you had to have high-quality data that was audited and reported frequently to those end users who would then utilize that data for a quality intervention. They found that you had to have strong stakeholder involvement, attitude, and enthusiastic commitment from the physicians, clinical managers, and administrators. You had to have patient involvement, including communication and standardization. And there had to be an appreciation of the importance of the measurements. This is not just to collect data. That's easy to do, but to actually look at it, interpret it, and move forward. The interesting thing is not all of these different registries had a QI expert. Some did, some didn't. But the importance of involving a QI expert really wasn't well-established, which is an important factor when you look at taking to your administration, to the powers that pay the bills, how much more resources are they going to have to allot to make this work? And you really did not need an expert in QI. You just needed the data, which we will collect in the registry. Unfortunately, none of these studies looked at the cost. So the true value, the impact of the value, really couldn't be determined, because we only saw the quality end, but not how much they spent to get to that quality. The registry, all the data, can also serve to create more studies and to guide clinical practice guidelines. When we think about clinical studies, they can be observational or interventional. And really, the registry would fall under sort of an analytic cohort study under observation. You can then take that data, and depending what you see with that data, you could set up specific control studies, some cohort studies that you go moving forward prospectively. So what does a well-designed registry look like, and what can it provide users? Well, a lot of the registries will have similar characteristics to a randomized control trial. There should be prospective data collection, clinical data entry points, and patient-reported outcomes, which really is one of the key points, because we don't have a lot of different methodologies to capture patient-reported outcomes currently. You need a high follow-up rate, quality control, and validation. What are some of the benefits? Well, it can provide level one prognostic evidence, which could be crucial for predictive analytics, or even as we move forward with more artificial intelligence to interpret the data. It can provide level two evidence showing whether the care was effective or not. It can help strong external validity towards whether the treatments you're providing, are they really doing the right thing? Are they improving what you're looking to improve? Having a larger database than just your own individual patients will give confirmatory data. And as I said, it provides real-world data. It's a way to get more information than just your own patient population. Okay, we're going to move on and look at specifically the registry's ischemic stroke data and what we've done so far. So if we look a little bit at patient reported data for stroke, you know, the question really is, what do we do right now and how far have we we've come? Because the clinical outcomes have been studied for quite a long time. We've measured patient function with a variety of different scales, whether it's the Barthel index or modified Rankin scale. We've been doing that since the 1950s. But widespread adoption of patient reported outcomes, really from the patient's viewpoint, we have not been doing a lot. Up to about 75% of stroke patients have been found to have deficits which could complicate or even prevent them from completing a patient reported outcome instrument. But we have seen that proxies report greater levels of impairment than the patient themselves. So we know we can use proxies, but we have to see how that will interfere with the data. If we look at PROMIS, it's one of the patient reported outcome measure tools that we're using for the registry, the neurologic registry, and what we found is there's several studies that have looked at that in the stroke literature. One particular study had a little over 2,100 patients, almost 2,200 patients, where they studied ischemic strokes, including, as well as interstitial hemorrhages, subarachnoid hemorrhages, and TIAs, and they had those patients complete the patient reported outcome measurement information scale, or the PROMIS, as well as the quality of life and neurologic disorders executive function. The eight domains of PROMIS were impacted similarly regardless of the different type of neurologic insult. But one of the things that was found is that female patients, younger age patients, and when the stroke was less than 90 days, they tended to have a worse score in multiple of the domains. Physical function, satisfaction with social role, and executive function domains tended to be the most impact. So when you think about potential intervention, you know, if we're finding that this particular subpopulation of the neurologic patients have more problems with satisfaction with social role, more problems with executive function, then perhaps we may need to have a different intervention with more neuropsychology services for those patients. The data that we get can then help direct the care. I'm looking at the PROMIS again for stroke. We had evaluation of the performance of patient reported outcome measurements, the PROMIS physical function scale, and compared that to a validated stroke impact scale 16, or the SIS-16, in ischemic stroke patients seen in an ambulatory cerebrovascular clinic. Almost 2,000 patients were studied with ischemic strokes, and what they found was the use of the PROMIS tool to obtain electronic patient reported functional status in the ambulatory setting was feasible. You know, we can set up registries, we can set up all the tools, but if the patient can't actually physically do it, or if the flow in an outpatient setting doesn't work, then we're not going to get good data. So this study showed that it was feasible, and that the PROMIS physical function scale was a good option for measurement of physical function in ischemic stroke. The PROMIS physical functions scale had a similar test characteristics as the SIS-16, but it had less, it was less burdensome to the patient, and had a minimal sealing effect. So what it showed was that the PROMIS was a good tool and could be used. When we look at our particular registry, we're looking at a lot of different data points that we're going to be collecting, or have already started collecting. We're looking at some of the demographic information, you know, patient's name, obviously, gender, height, weight, ethnicity, race, and then we're going to be looking at the patient reported outcome, including the PROMIS-29. We're going to look at several different characteristics, level of education, for example, employment status, marital status, alcohol use, caregiver status, to try to see what commonalities our patients have as we go forward with their treatment, with what's going to help drive, what different characteristics will predict certain outcomes, or predict the need for certain interventions. We have inclusion criteria, which include new patients to an inpatient rehabilitation facility setting. Patient has to be 18 years old or older, and has to have a diagnosis of an ischemic stroke. Ideally, as we go forward, we want to expand not only with ischemic strokes, but ultimately, once we we see how well it works with ischemic strokes, we want to expand, but we also would like to expand to other different levels of care. We chose inpatient rehabilitation facilities because it was, one, a big bulk of what we do, but two, most of the rehabs have EMRs, which would support the collection of this data. As a specialty, we certainly want to look at subacute rehab facilities as well, but right now, many don't have an EMR, and it would be quite a burden to collect that information. So, that's one of the sort of the longer-term goals that we'll be thinking about. We have follow-up protocol measuring the function as measured on the IRF patient assessment instrument within the first three days of the patient stay and within the last three days of the patient stay, and then we'll have some registry insights built in where we can analyze some of the data and the trends. So, what kind of insights do we have from the early adapters? Well, some of our programs have said, as a new academically oriented department of PMNR, they felt that it was important to join the registry so that we can use this powerful tool to observe the course of ischemic stroke in our inpatient rehabilitation unit and to assure that we are providing the best health care quality and safety possible. You know, I think we all have that feeling that we think we're doing the right thing, and with inpatient rehab, particularly with stroke, well, we've kind of been doing it the same way for a long time, but are there better options out there? Are there different interventions we can do towards making the ultimate outcome even better? This particular organization felt they were comparing the effectiveness of different treatments, evaluate different approaches, and they felt they could monitor safety through the registry by doing some of the data analysis. They felt it was important to contribute to the development and validation of evidence-based medicine within our specialty, as we all kind of need to continue to advance with the changes and challenges of a constantly evolving health care system. Gathering and sharing data is an excellent way to contribute, and I certainly think that that is a wonderful thought, because if I contribute my patients and down the road they contribute their patients, we just have a much more robust set of data to really start making some very valuable insights. More insights from early adopters, looking at trends and outcomes and having access to national database, looking at how we can best provide care. They felt it was exciting that they needed to be able to offer patients the strategies and interventions that are most likely to promote successful outcomes. Workflow changes will be difficult, which probably can be an understatement, but we've designed it trying to have as much of the data pulled directly from the EMRs to try to limit that burden, but the challenges really should be worth the reward of the data and help us guide the care. So with that, I'll turn it over to Dr. Lesher to talk a little bit about our low back pain data experience. Thanks, Alan. Thanks for passing the registry torch into the spine realm. I just wanted to discuss a little bit about the history of spine registries. They've been around now for almost three decades. Not surprisingly, they were first started in Scandinavia, in Sweden, and then in the early 2000s, SpineTango began, which is mostly in greater Europe. The American Academy of Neurological Surgeons started the N2QoD approximately 10 years ago, and that stands for National Neurosurgical Quality Outcome Database. More recently, the neurosurgical group has joined with the American Academy of Orthopedic Surgeons to develop the American Spine Registry that was announced earlier this year, and that's focused essentially mostly on spinal surgeries. The conservative care registry realm is much younger. Again, the efforts to develop conservative care monitoring and registry efforts started in Europe with an offshoot from SpineTango, and around the same time, the Danish Spine Data Registry got started. Our registry is younger, starting in 2017. Additionally, there are some conservative care registries, including the Chiropractic Registry, as well as the North American Spine Society Registry, which is inclusive of spine surgeons and non-surgical care providers. So continuing on this question of the data that we gather with spine registries, does it improve quality? So can we get a good outcome at a cheaper cost? What does the data show us about that when applied to spine? In the surgical realm, there is plenty of robust data. These two studies I'll just talk briefly about, I think, really show how well registry studies can really lead to good randomized control trial studies. So this, the first study wanted to really evaluate, does decompression provide the same outcomes as decompression and spinal fusion in patients with one to two level spinal stenosis with or without a degenerative spondylolisthesis? This was started by the Swedish Spine Group and their registry. And so, again, it started with really large numbers. Over 4,000 patients had decompression, and over 1,100 patients had decompression and fusion. And they followed these folks, or these patients, out to two years. They were looking at the primary outcomes were oswestry disability, the uroqual 5D, and then VAS for back and leg scores. And at two years, both of the groups had similar outcomes, as well as similar reoperation rates. So, again, they were providing similar, in the decompression group, they were providing similar outcomes as fusion, but at much less cost. And so the same spine study group, or the same surgeons in Sweden said, okay, this is really promising data. We want to take it to the next step, or the next level, on the pyramid of research and create a randomized control trial. Again, assessing essentially the same question. Do we need, really, to do a fusion in these patients with one to two level spinal stenosis and with or without a degenerative spondylolisthesis? Now, the randomized control trial was much smaller. There was 120 patients, roughly, in each group. And again, they followed them to two years, and then five years, and essentially found that less is more. And not surprisingly, in the decompression group, there was a shorter length of hospital stay, shorter surgical times, as well as the direct operative costs were significantly less. In this study, it was almost $7,000 less for just decompression. And then, additionally, they followed the patients even over a longer period of time, out to six and a half years, to see if re-operation rates were any different. And they weren't. So there was roughly 20% re-operation rate in both of those groups. And this randomized control trial was ultimately published in the New England Journal of Medicine. So, really, a great example of large volumes of patients showing that a treatment trend is appearing. Less is more. Let's study it more closely and see if we have a similar result. Clearly, it showed this very well. Now, we asked the same question, do the spine registry data improve quality in conservative or non-operative care? Unfortunately, the data is far less robust. And I invite any viewers of this webinar or lecture to parse the data, go on PubMed, see if you find studies that I missed. I did spend quite a bit of time looking to try to really fine-tune or focus on registry-specific studies, looking at conservative care for back pain patients. And I was able to find one. So this study was done by a group of osteopathic physicians in England. And really, what they wanted to see is, is osteopathic care a good and a viable option to treat patients with chronic spine pain? And so, these patients included both low back as well as neck or cervical and thoracic spine pain. And they were mostly, or it was focused mostly on chronic spine pain, lasting over three months. But most of the patients had pain well longer than one year. Involved just over 260 patients. And then, essentially, they did a patient reported outcome studies. Those included the oswestry, the neck disability index, the uroqual, and they also wanted to evaluate just general satisfaction with the treatment. And then determine if there was any adverse events. The groups of patients received either one course or two courses of treatment. And the course was essentially a six-week course of osteopathic care, which included spinal manipulative treatments, in addition to exercise, education, and instruction, as well as just education on living with chronic pain. And then depending on how patients progressed, they would either get one six-week course or possibly two six-week course. And then this was highly physician driven. All of the data was being collected by the physicians at the time of, at the onset of the the treatment, and then at the conclusion of either the first course or the second course. So what did they find? The data was essentially descriptive. It did support the use of osteopaths to deliver conservative care for chronic spine patients. However, the authors did not conclude that they could make a really valuable assessment on the effectiveness of care, because it was just one group study. Essentially, four to eight physicians at one location with short-term follow-up. The estimated follow-up was roughly around four months. Satisfaction with the treatment was very high. There was minimal adverse events. Those were usually just mildly increased pain as a result of the spinal adjustment, which essentially went away by the end of the interval treatment time period. So again, I think this study is great in the sense of showing, hey, these people want to show, does this treatment have an effect on treating spinal pain? But it's just the very beginning here. We need to build on this significantly. So when we kind of look at the results of that study and try to apply it as a way of conservative care treatment for patients with back pain, what can we learn from this, as well as what can the guidelines that we rely on as spine providers tell us? So when you look at the most recent guidelines from the North American Spine Society and focus on SMT, what does it tell us? So when you look at the specific question of how useful is this for chronic spine pain patients, we really don't have great data. So just as a kind of review, the NAS guidelines grade quality of evidence or quality of the recommendation based on A, B, C, and I lettering. So A is there's good evidence, meaning level one studies to back the recommendation up. Grade B is fair evidence. So two, or I'm sorry, evidence of two to three categories. And then C is essentially poor evidence, and those studies are fit the four and five level. I is there's insufficient or incomplete evidence to make essentially a recommendation. So when you look at spinal manipulative therapy for chronic back pain, really it gets a grade recommendation of C, which is essentially poor. And then when it's compared to other treatments, which include medications, modalities, or no treatment, there is not enough evidence to really make a valid conclusion. So again, this is a prime example where we look at the guidelines to help us guide our care, but we don't have the information to provide us with better understanding. So a simple way to further improve this is to get more people enrolled in the registry, looking at treatments like this. This slide really discusses what we're trying to collect with our spine registry. The demographics and characteristics are listed there. We are, our main patient reported outcome is the PROMIS-29, as well as numerical rating scale. Exclusion criteria for these patients are, we're not including patients with prior surgeries, those patients being actively treated for cancer or spinal infection, as well as we're excluding patients with workers' compensation claims or active litigations with regards to their current pain. In addition to PROMIS and numerical rating scores, we're also recording patient satisfaction, ability of return to work if applicable, complications with procedures done with the care, as well as asking patients throughout the time course of follow-up, do they feel that their treatment has been completed? We're following these patients out to a year with interval checkups at six weeks, three months, and six months. We will be relying on our third-party industry, Arbometrics vendor to help us with the personalized reports as well as dashboards so we can get a recurrent or a frequent feedback. And so, as Alan touched on before, registry studies are really dependent on validated patient reported outcomes. And we've decided to follow or employ PROMIS. And this group of outcome measures has been really studied and developed by the NIH since 2004. The positives of this is this allows for flexible testing and computerized adaptive testing. So essentially, questions change or are limited or expanded based on patient responses. The PROMIS-29 is what we will be specifically looking at. And we will be going through a demo later in the talk, looking at the different domains that it will be testing. More recently, PROMIS has been studied more extensively over the past five years with regards to the spine pain population. The University of Utah has a strong group of spine research there who are looking primarily at its use in capturing physical function scores in patients with back pain and comparing them to older or more legacy outcome measures, including the Short-Form 36, as well as the Oswestry Disability Index. And their study essentially showed that the physical function of the PROMIS really scored better than these legacy outcome measures. More recently, in the last couple months, and this was published in the archives of PM&R, the researchers wanted to look at how well does PROMIS hold up with regards to the psychological factors that go into physical functioning. And so they looked at the PROMIS and compared it to the Oswestry and found actually that PROMIS had very similar sensitivities as the Oswestry, and it was far more easier to compete, or pardon me, to complete, similar to what the studies that Dr. Novick discussed with neurorehab. So in addition to having validated outcome measures, we need to be speaking the same diagnostic language when it comes to putting patients in the registry. And as I've been involved more with the registry, this has become more and more important to me. I work in a large multi-specialty spine group where a lot of patients, when they come to my clinic, they already have extensive imaging done because most of the time they will be seeing a lot of the surgeons in my group. When we look at other groups in the registry, they may not have any surgeons on board. They may have no access or minimal access to imaging. And so when a patient comes in complaining of back pain, do we really know, is this discogenic? Is this facetogenic? Is this soft tissue or a combination of those? So what we've employed in our group is that unless we're very certain of a structural diagnosis at the outset, we will put patients in symptomatic subgroups. Those include a low back pain group, a sciatica with or without low back pain group, a neurogenic claudication or stenosis group, and then a radiculopathy group based on motor impairment or weakness. And then there's an other group to capture fractures and scoliosis as well. And so if the patient begins in a symptomatic subgroup, and as we work with them and develop their treatment and further diagnose them along the way, and their structural diagnosis becomes much more clear, then we label them the structural diagnosis so it's captured by our outcomes process. Having this consistent diagnostic labeling or vocabulary is really gonna be crucial to benchmarking and allowing comparisons between groups as well as nationwide. And additionally, as we follow these folks or patients into surgery, we have to be speaking the same language as those patients go into the surgical treatment realm. So insights from early adopters, what are we seeing as we enroll more and more patients? Clearly getting patients enrolled is going to require enthusiastic providers. We can't just rely on giving a patient a tablet, telling them, hey, just fill this out and having them do it in an appropriate way that we can really get a lot of good, useful information. The physicians enrolled in the registry have to explain to patients why we're doing this. It's ultimately to improve care and get patients engaged in the treatments that they're going to be received. The one thing that was really clear to me as we started the registry enrollment is I wasn't enrolling clearly as many patients as I thought. Some of them had comorbidities that made the enrollment process hard. Some of them had surgeries that you really didn't find out about until you examine them and see that they have a surgical scar and say, oh, yes, I did have back surgery. So that's just something to be aware of, as is selection bias. When you look at registry studies, a common complaint is that they do suffer from a selection bias, especially if the physician is doing the active role in enrolling or taking an active role in enrolling patients. So I think a lot when I'm seeing patients and asking them to fill out our baseline outcomes questionnaire, will this patient really take this seriously? I don't really know. And so on blind faith, you just have to do it. You have to get into the habit of doing this really without thinking twice about it. This is not an exact science and it will get honed and fine-tuned as we do it and collaborate more on the process. Additionally, having the ability to collect the data into our EMR as seamlessly as possible and sending it to the registry, getting feedback or getting analytic capability and then sending it back to us is going to be really crucial. Having data questionnaires filled out on paper is an option, but just takes a lot more time than uploading that into the computer or electronic format and then sending it back to us. Getting regular feedback as Dr. Novick touched on previously is really crucial. If the feedback or the data that we're collecting is not being shown to us on a regular basis, this whole process becomes out of sight, out of mind. My busy day is getting ahead of me. I don't have time for this. If we're seeing it weekly or biweekly, it's always at the forefront. We're thinking about it when we see these patients in follow-up. And then finally, when we see these patients in follow-up. Additionally, the patient follow-up collection or data collection is not easy. Low back pain, typically an acute episode that's new onset, never before, resolves within usually three months. So as patients get better, they don't want to be burdened by these types of questionnaires. When you look at the studies on outcomes, they say that ideally, essentially these outcomes or studies are mostly focused on surgical data, that at the 12-month mark, you really want 60 to 80% follow-up rate. The higher, the better. This is really dependent on text messaging and email or electronic communication on getting this done. I will get some of the data in physical clinical evaluations when I see patients back, but this is much smaller percentage. And then this data to kind of give some real life examples is essentially my follow-up data from 2018 up until the end of September. So the slide on the left shows my enrollment of 141 patients, and then my six-week, my three months, and my six-month follow-ups. And so they're pretty abysmal when you look at the completion rates. Knowing that, I knew that it was the first step, but things can only get better. So when we look at it for two years, my enrollment almost nearly doubled, and that has to take into consideration COVID. So for April and May in my group, our non-operative data collection was essentially shut down. And then when you look at follow-ups, well, I did achieve 60% or just over at six weeks, but at the three-month and six-month follow-up appointments, those definitely still need ongoing improvement. And just for September and October, when I get these readouts, it lists the way that these patients are being contacted. So for my follow-up, around 40 to 45% of my patients are filling up follow-up through emails. The other 40 to 45% is being contacted by telephone, and the remaining 10 to 20, I'm capturing actually in my clinic. So again, much room to grow on here. And I'm gonna pass it back to Dr. Hatzikas for our registry demo. Thank you. Thank you, John and Alan, for those really informative and interesting talks. And now I am quite sure that all of you are just dying to see what this registry looks like. So now we're gonna pass off to the video, which is gonna really walk us through what this registry looks like and feels like and how it interacts with our patients. ♪♪♪ The American Academy of Physical Medicine and Rehabilitation, AAPM&R, is the national medical specialty organization representing more than 9,000 physicians who are specialists in physical medicine and rehabilitation. In order to fulfill the vision of the academy and our membership, AAPM&R built a clinical data registry. AAPM&R's registry is a single repository of data which will aid the specialty by guiding efforts to reduce burnout, defend scope of practice, demonstrate value, and provide data to improve patient care. Success of the registry will come from harnessing traditional electronic medical record data with patient-reported outcomes data. Patient-reported outcome, PRO assessments, are critically important tools that are filled out directly by a patient or a proxy to inform both clinicians and the patients themselves of pain, quality of life, goals, and other important metrics. These surveys provide an additional data point to assess whether clinicians are improving the health of patients overall. AAPM&R's registry, governed by an academy-led steering committee, is using PROMIS-29 as the core PRO for the registry. PROMIS-29 is a global tool that can be used across clinical diagnoses and settings. The survey is comprised of 29 questions over eight different domains. Physical function, anxiety, depression, fatigue, sleep disturbance, ability to participate in social roles and activities, pain interference, and pain intensity. Patient-reported outcome surveys are sent to patients that meet specific inclusion-exclusion criteria for low back pain and ischemic stroke. A patient can receive the survey directly via email or they can fill it out in your office. Before answering the PROMIS-29 questions, the survey takes the patient through a number of demographic and characteristic questions that will help the registry supplement data not typically found in an electronic medical record. After completing the first set of questions, the patient will answer PROMIS-29. Most PROMIS-29 domains ask a set of four questions, with pain intensity being the last and final question. All answers follow a similar Likert scale and the patient can easily move through these questions and answer them to the best of their ability. If at any time the patient needs to stop and come back to the survey later, they can choose the Resume Later button at the top of the screen. When they click back on the survey link, it will open up where the patient left off. Once the patient completes the survey, they will receive a thank you message on screen and in their email inbox. Patients will receive follow-up surveys at specific intervals via email. If they choose to use it, all patients will be able to log into their own PRO dashboard to track their progress throughout their follow-up protocol. AAPM&R's registry streamlines the PRO process by capturing data quickly and accurately with results depicted in easy-to-read dashboard reports, allowing participants to compare patient outcomes, professional performance, and care processes against other physicians across the country. Using AAPM&R's registry, physiatrists will learn from each other what treatments are ideal and result in the best outcomes, giving you the data needed to demonstrate your value and improve your care. We're laying the groundwork to develop measures that are more relevant to the clinicians who provide rehabilitation services. We don't believe rehabilitation should just be a process. Rehabilitation should be about improving outcomes and changing the face of the specialty. The only way to do that is by standardizing outcomes, analyzing the data, and proving that what you do is making a difference. AAPM&R's registry is currently recruiting institutions and practices that are committed to quality improvement for data collection. If you are interested in learning more about our registry, please contact registry at aapmr.org. So in summary, we're very excited about the registry. As you can see, this registry is operational. And despite the fact that we're using the word pilot, this is a fully functioning registry. We are collecting data on low back pain, ischemic stroke, and we hope to have some exciting data a year from now in this Academy meeting in a year, wherever we happen to be at that time. In summary, our thoughts for the future, the power of the registry will depend on the number of and the volume of patients that we collect. It will depend on our follow-up ability and also the number of you who decide to participate. Of course, data must really represent the total population over time. Quality of our data must be continually scrutinized and analyzed to make sure it represents not just the diversity, but also represents how we practice. And our patient report outcome measures must show value of our care. Quality of care is hopefully, will give us a better understanding of our disease processes that we treat and really understand the influencers of recovery. And the registry will show benefits to all participants through the data collected. It'll help us in our lobbying efforts to continue to provide the value and the reimbursement for PM&R services. It will help us pave the way for a value-based environment, both in parent negotiation and in CMS outcome tools and all of the ways that we are gonna be acknowledged for the services that we provide. And as we know, data analytic capacities, AI data and analytics are advancing as a science. And we have better, more constrained, more specific data. We can start using those capabilities to draw inferences at our dashboard level, so we can have a better understanding for the care that we provide. And we'll increase specialty awareness amongst payers and patients about the impact what we do has on outcome. And most importantly, it's an environment to help us improve ourselves as professionals, help us improve our skills and to benchmark those skills against outcomes that are important to our patients. And as part of that, to create a maintenance certification model that's relevant and continuously part of our clinical practice. And I know all of you are very excited about this. And at this time in a live meeting, we'd be raising your hands, asking questions. Please come to one of our session booths so we can answer your questions. We'd love to be there and have a discussion about any questions these talks bring up. And if you can't attend one of those, please by all means, reach out to registry at aapmr.org and ask us questions. If you're interested in becoming a pilot site, by all means, reach out and we'll get back to you and discuss the opportunity with you. And of course, I wanna acknowledge that some of the centers that are currently participating, Brooks Rehabilitation, Rusk Rehabilitation NYU, the Shepherd Center, Shirley Ryan Ability Lab, Rhode Island Hospital and Carolina Neurosurgery and Spine where Dr. Lesher is from. We really thank those organizations for stepping up and really being leaders in the registry and participating along with us and helping us guide the future of psychiatry. Thank you very much for attending.

clinical data registry

low back pain

ischemic stroke

patient-reported outcomes

PROMIS-29 questionnaire

treatment decisions

patient engagement

cost savings

data analysis