Good afternoon, everyone, from New York City. My name is Dr. Mike Lodell. It's a pleasure this afternoon to join my colleagues from New York Presbyterian Hospital, Dr. Diane Thompson and Dr. Joel Stein, for this afternoon's course update on stroke rehabilitation. I'm going to be starting out this afternoon with just a few updates on different aspects of hemorrhagic stroke. These are my disclosures, none of which are really relevant to today's topic. Today we're going to cover just a little bit of an update on outcomes after hemorrhagic stroke, as well as seizure prophylaxis and DVT prophylaxis. As a reminder, about 10 to 15% of strokes are hemorrhagic in nature, and although there are multiple etiologies, today the focus is really going to be on intercerebral hemorrhage. These strokes can and should be distinguished from ischemic events. The mortality is different. There are different recovery patterns, and the prognosis is slightly different. I might mention that for a number of reasons, rehab research in the area of intercerebral hemorrhage is actually relatively sparse. As a reminder, the difference in the pathophysiology between ischemic and hemorrhagic strokes, on the far right-hand side, you can see that when an ischemic event, there's a blockage of an artery with a relatively distinct area of brain damage distal to that, depending on which artery is involved. This is in opposition to hemorrhagic event, where there's a rupture or disruption of the vessels with a less predictable extravasation of blood. It is worth noting that for the same volume of ischemic versus hemorrhagic stroke, it's been found that there is a greater degree of cell death and inflammation associated with the hemorrhagic stroke. The initial mortality after hemorrhagic stroke is quite high. At 30 days, it may even be up to 50%. We have traditionally thought of hemorrhagic stroke in general as having an increased mortality, but long-term, thinking that these folks do better in terms of prognosis compared to ischemic stroke. There's at least one study that suggested that a lack of progress in ischemic stroke during that first 24 hours, as measured by no change in the National Institute of Health Stroke Score, is really a fairly accurate engraved prognosis predictor. In terms of survival, there's actually conflicting recent evidence. There's at least one study that has showed no increase in survival over a period of time from the early 90s to 2015. This is in contrast to a recent study by Bejau in France showing actually a 55% increase in survival. I think from a rehab standpoint, however, it's important to note that although the survival did increase to 30 days, they saw a substantially increased percentage of patients that were discharged back home not being independent in their ambulation and not being able to achieve a home discharge. Peter Langhorne has, in a review, concluded that the benefits of organized stroke units are about the same between ischemic and hemorrhagic strokes, and that if a readmission occurs after a hemorrhagic stroke, it's likely infectious in etiology. As I'm sure Dr. Stein is going to chat about at the end of this program, there's differences in the inpatient rehab unit versus SNF patterns, but it's somewhat questionable why this is the case. It perhaps is more facility dependent than perhaps the severity of the stroke or other more demographic or severity characteristics. At baseline admission to inpatient rehab, we tend to think of hemorrhagic strokes as being worse or more severely impaired than ischemic strokes, but seeing a faster gain while they're in rehab and beyond. There's slight differing opinions in the literature. Of the six or seven papers that have been published, this all but two would suggest that yes, individuals with hemorrhagic stroke actually progress through rehab just a little bit faster than those with ischemic strokes. There's at least one study that also suggests that starting rehab within a 48-hour period after the initial event leads to decreased mortality and better function, and that during the time in inpatient rehabilitation, a slightly increase in intensity actually leads to better outcomes. Moving on to seizure prophylaxis. This is a topic that we tend to think of really in the terms of traumatic brain injury, but I think it's important to recall that cerebrovascular disease is the number one cause of seizures in folks that are over 60 years old. The overall incidence of seizures after stroke is about 7%, but that's substantially higher in hemorrhagic stroke, maybe up to 25%. It's thought that seizures may be associated with an increased acute length of stay, perhaps increased mortality, increased time on the ventilator, and a delay in starting rehabilitation therapies. A moment of consideration would certainly reveal that by virtue of the mass effect that is associated with hemorrhagic strokes, certainly the blood in traumatic models as well, iron has been found to be a very epileptic genetic factor, and gliotic scarring is perhaps not unexpected. The seizure rate would be higher. We also got to consider that in a fair number of patients with hemorrhagic stroke, they will have undergone craniotomy as well, which also carries a seizure risk in and of itself, and in certain cases, there actually may be head trauma, either from the seizure itself or from an initial fall with the presentation of the stroke. For individuals that have a seizure following hemorrhagic stroke, within about the first two weeks, there's anywhere from a 33% to 35% chance that they will have a late seizure at some point during the first five years or so. There's some question, as in traumatic brain injury, as to what exactly the risk of having an immediate seizure, in other words, within the first 24 hours of injury might be. Standard of care is certainly in most centers to treat the first seizure, but prophylaxis is quite variable among centers. Recent study by NEDIC looking at the impact of phenytoin in seizure prophylaxis concluded that there were increased complications and decreased functional outcomes at three months compared to no prophylaxis at all. Again, this is also consistent with findings that we see in the traumatic literature. Probably more applicable to our patients with seborrheic hemorrhage that may be on nimodipine for vasospasm prophylaxis, several of the anti-epileptics have been shown to decrease nimodipine levels, perhaps increasing the chances of complications. I think it's fair to say for any number of reasons, including decreased plasticity, there's probably no role for phenytoin or phenobarbital for seizure prophylaxis. This is Engelman, 2019, a meta-analysis that showed not only was there no effect of the use of anti-epileptics in intercerebral hemorrhage on the incidence of seizures, there was no effect also in the overall quality of life. This is a systematic review showing that the use of prophylactic anti-epileptic drugs was not associated with improved short or long-term outcomes after intercerebral hemorrhage. Interestingly enough, and I think in any number of other acquired ideologies of prior brain injury we've seen this as well, there's been a progressive decrease in the use of phenytoin over the last several years. But as this graph shows in the upper portion, that's been more than offset by an increase in the use of levotiracetam or Keppra. It's very interesting. I don't understand, particularly our neurologic and neurosurgical colleagues, there is a overall prevailing thought that something, anything, is better than nothing and why it's appropriate to default to prophylaxing these patients and waiting for studies to show that it doesn't work as opposed to not initiating the prophylaxis in the first place and waiting for studies that actually show that it does work. You might say, well, perhaps it's good news that at least we're using Keppra. Perhaps not. As you can see, this is a recent study showing that there actually are cognitive quality of life decrements that are seen using levotiracetam as prophylaxis versus nothing at all. These are, to sum up, these are two recent guidelines. The AHA guidelines recommends that there is no role for seizure prophylaxis following intercerebral hemorrhage and likewise from the European Stroke Organization as well. So in summary, it appears that seizure prophylaxis in this population does not work and it probably hurts outcome and probably is best for us not to do it. And finally, DBT prophylaxis after hemorrhagic stroke. This is an area that's a little bit more controversial perhaps than seizure prophylaxis. PE can occur up to 5%. Some studies even 10% following intercerebral hemorrhage in the first month or so. This is probably four times higher than the rate that we see in ischemic stroke. The at-risk factors, certainly immobility and hemiparesis as well as other medical issues. There's some evidence to suggest that the incidence of DBT is increasing over the period from 2004 to 2013. If there's one thing that the studies currently do agree on, it's that graduated compression stockings or tethos are not effective in the prevention of DBT. A recent survey by Cherian indicated among the acute care services, this is probably going to be the neurology or neurosurgery services, about 90% of prophylaxis was mechanical using pneumatic compression devices. This is very common across the country. There is a cost associated with this, but certainly mobilization is more difficult with these devices as well. They found substantially less chemoprophylaxis early, about 8%. Obviously this is an acute consideration in this population. When there's an underlying intercerebral hemorrhage, the use of anticoagulation is something that needs to be looked at very carefully. I would suggest there's a difference in that presentation in the acute care versus the rehab unit. By definition, somebody is going to be on one of our inpatient rehabilitation units because they have decreased mobility and are therefore, almost by definition, at a high risk for DBT on the unit. I might mention just briefly, I'm not talking about those individuals who had an intercerebral hemorrhage because they were on anticoagulation, atrial fibrillation, cardiac valve replacement. That is a whole other topic of which there really is not general consensus on how to manage. The worry about exacerbating bleeding with chemoprophylaxis, we're certainly worried that it would increase the rate of disability and certainly mortality. The neurosurgery literature, not necessarily specific to intercerebral hemorrhage, is rather clear that there is a small risk of hemorrhage, non-statistically significant risk of hemorrhage with DBT prophylaxis chemically, but the drop in rates of DBT-PE warrant the small trade-off. Again, increases in bleeding, not statistically significant. It's a little bit controversial when chemoprophylaxis can be started, but one suggestion by the AHA is once bleeding has stopped within the first one to four days after presentation, probably low molecular weight heparin or unfractionated heparin can safely be used. In those individuals who actually do have a diagnosed DBT or PE, being capable of filter or full coagulation is probably indicated. Again, the AHA guidelines are that graduated compression does not work, that early intermittent pneumatic compression should be started in one to four days following the event, assuming there's no further bleeding, that chemoprophylaxis can be initiated. The European stroke guidelines on this topic are a little bit more cautious. They agree in terms of the use of graduated compression stockings and the appropriateness of early intermittent pneumatic compression, but they feel there's insufficient data to warrant recommendations on the use of chemoprophylaxis. In summary, intermittent pneumatic compressions from day one, low molecular weight or unfractionated heparin from day one to four, this would be appropriate on the rehab unit since somebody's probably going to be immobile, this is something that we should consider doing. So in summary, patients with hemorrhagic stroke are more disabled admission, tend to be more disabled in admission in patient rehab and may make faster gains once they're in beyond. Likely there's no role for seizure prophylaxis in this population, perhaps even with the newer agents, and the rehab team should consider advocating to our neurosurgical or neurologic colleagues to wean the anti-epileptic agents if present. DBT prophylaxis using intermittent pneumatic compression initially followed by low molecular weight heparin if immobile is appropriate, and this probably applies to most of our patients in patient rehab. I would conclude by saying the broader inclusion of the patients with intercerebral hemorrhage is something I think we're going to see a greater emphasis on in future clinical trials. Often this population is excluded, and I think we're going to need to be making a greater justification for exactly why this population is excluded in the trials if it happens. It's now my great pleasure to introduce Dr. Diane Thompson, who is going to continue this afternoon's discussion with lifestyle interventions for stroke survivors. Thank you. All right, thank you, Dr. O'Dell, and good afternoon, everyone. As mentioned, I will be, my name is Dr. Diane Thompson, and I will be talking to you about lifestyle medicine as a tool, a potential tool in secondary stroke prevention. So in terms of the outline, I will be covering a little bit about the need for secondary stroke prevention. I will also give you an overview of lifestyle medicine and share how the behaviors that are targeted in lifestyle medicine work very well with the risk factors that we see in stroke recurrence. I'll also zone in on two strategies in lifestyle medicine, particularly the exercise prescription and the nutrition prescription, and I will share with you some resources if you are interested in learning more about lifestyle medicine. In terms of disclosure, I have written two books on lifestyle medicine, and I do have a nonprofit that teaches lifestyle medicine education to people of color to help address health disparities. So in terms of secondary stroke prevention, we know that the statistics shows about 800,000 stroke per year. About 23.4% of those patients are recurrence, and often that's when we as physiatrists will see those patients. Often we don't necessarily see them before the first stroke, right? As these patients move away from the time of the first stroke, the cumulative recurrence risk actually increases. Stroke in itself carries a high rate of disability, and when we talk about recurrent stroke, the rate of disability is even higher, and so is the rate of mortality. So we as physicians, it's really important that we see this need to help reduce future risk of events in our patients. When researchers look at the risk factors that are associated with stroke, around 80% are often said to be preventable of these strokes. In some communities, up to 90%, right? And these strokes are associated with about 10 modifiable risk factors that we see over and over again. Hypertension is a very common one and very commonly seen in the black population. So as we see, these are some of the identified modifiable risk factors in stroke, and we see them in recurrent stroke. If you take a look, the first several of them fall under the category of lifestyle. Lifestyle and chronic disorders. And often in lifestyle medicine, we say these are lifestyle-related diseases. We also see included behaviors such as smoking and alcohol, and even in some cases, illicit drug use. Okay so when we look at how the approach for the treatment of secondary stroke prevention it falls under several categories. So the first part is lifestyle. It looks at education, it looks at weight management, addressing sedentary lifestyle and again those behaviors of smoking or alcohol. We also see addressing chronic diseases and again many times lifestyle intervention is seen necessary in addressing these diseases. Along with that is antiplatelet therapy and also some procedures. We are going to focus on the left side where we're talking about lifestyle, the things that we can actually address when those patients come in to see us. One of the models that's proposed for addressing secondary stroke prevention include a combination of things. So it's a multimodal approach where education, targeted education, not just saying to someone eat less and exercise more but targeted education that is aimed at the level of the patient because everybody will need different approach right. Behavior change theory because your patient that doesn't think they're overweight is very different and has very different needs from someone who has been trying to lose weight and probably is now in relapse. Self-management, we find that when patients are involved in their care they are willing to monitor themselves, monitor their intake of certain foods or their the amount of exercise they do. They tend to do much better and person-centered care where this patient again is a active participant in their care so you're not just handing them a prescription and they go along you they're actually very involved in their care. When these things are combined patients tend to do very well. And so lifestyle medicine might be one of these tools that we ought to look at in terms of addressing secondary stroke prevention and why? Well lifestyle medicine is evidence-based lifestyle intervention which is different from just lifestyle information that's on the internet right so it's evidence-based and the idea is it's a belief that changing unhealthy behavior is foundational to medical care so that if a patient has high blood pressure you don't simply hand them a prescription for the high blood pressure and they go about their way that you actually get to the root of it and help to identify unhealthy behaviors that led to that blood pressure or whatever other issue you're you're addressing. It is seen as a primary therapeutic modality and very importantly it involves the patient, the physician, an interdisciplinary team and the community. If you are trying to get your patients to eat better for example and they live in a community where there's no healthy foods well that's going to be a problem so you really have to take everything into consideration. But lifestyle medicine targets certain lifestyle behaviors and according to the American College of Lifestyle Medicine if you take a look you'll see improved nutrition. This is a focus on hopeful plant-based nutrition, physical activity, the idea that people should be active and addressing other lifestyle issues like stress and sleep, addressing those habits that we see as risk factors for stroke, tobacco, alcohol and even addressing social network. We find that that is really important in people's recovery so that many of the targeted behaviors for lifestyle medicine are also the behaviors that we would want to target when we talk about secondary stroke prevention and their risk factors that we want to modify. In lifestyle medicine the physician will prescribe the lifestyle change so that physician has to be knowledgeable about nutrition and exercise and all the other lifestyle targeted behaviors. That physician will engage an interdisciplinary team and for physiatrists we're very comfortable generally with dealing with a team so this works very well. You need to be able to access resources for the patient. For example I am not very good at smoking cessation education but if I identify that my patient needs that I know exactly where to send them. The idea that you have the information as a physician so you can educate, you can assess the level of knowledge that you need to share with the patient, provide support, coach that patient to behavior change and that's that takes work and that is not how we typically practice medicine right now. So it is a change, it is a new approach to helping our patients get to the root of their problems so they cannot deal with some of these risk factors for recurrent stroke. Since lifestyle is seen as medicine, well in lifestyle medicine it is prescribed. So for example nutrition is prescribed. The basic nutrition prescription and it can get much more complicated depending on the patient because you're offering patient-centered care but typically it uses the acronym TAP so the type of food, the amount and the frequency and you get as specific as you need to be. I've often shared that I had a patient who had no vegetables and actually the only thing he told me he had he ate was french fries. Otherwise there was no no kind of whole food in his diet and we got to the point of where he identified broccoli as what he wanted to eat and we did a prescription for a cup three times a week. His sister was going to prescribe it. When he started being consistent we actually moved to something else and added more. He actually eats vegetables on a daily basis probably not to the level that we wanted or is recommended but it's certainly much better than before. These prescriptions can also take on a positive or a negative take on them so you can add things so eat more of this or remove less of that. In general patients tend to respond more to positive prescriptions so you want to spin it in that way and again depending on your patient you might have to get a little bit more specific. And you know we know that food actually will affect many of these risk factors that our patients have. We know for example the dietary approach to stopping hypertension. This is a diet that was designed to help treat and prevent high blood pressure and has been shown to help with blood pressure control. The same thing with the portfolio diet. The portfolio diet is a vegetarian diet that uses four types of food in a portfolio. These are plant sterile nuts, plant protein and soluble fiber. Actually it's been shown to reduce LDL cholesterol as much as 20%. The diabetes prevention program. This is a program that is now paid for by CMS because it's been so effective and what it did show when it was actually a 27 center randomized clinical trial that showed that with proper lifestyle intervention diet and exercise patients were able to lose weight, keep that weight off but more importantly prevent the progression from pre-diabetes to diabetes. And I said this is a this this has been so successful that it's actually paid for now by by insurance. Exercise is seen as medicine and lifestyle medicine. This is important for our stroke patients. Sometimes they are out of therapy. They're no longer doing therapy and they don't move around much because they're already impaired from the stroke and then the ability sets in so now we just have this cycle. So the idea is that even in between therapy exercise is prescribed. The acronym used is the FIT acronym frequency intensity type and time and you're as specific as you can be and these prescriptions are very specific to your patients. I have patients who for them it is walking the hallway in New York City with their their apartment hallway with their rolling walker and their home attendant twice a day and then maybe we can keep going higher. It's based on what they can tolerate. Kaiser Permanente has a walking prescription and also now we have physicians who use mobile devices. They use an app where that prescription can be titrated based on what patients are doing. And in terms of some of the information out there what we do know is when we're using lifestyle as medicine to prevent stroke recurrence we don't have a lot of randomized control trial. We do have some smaller studies out there that suggest that physical activity and exercise will favorably improve the stroke risk factors that in fact people who moderate exercise at a moderate or high rate tend to lower their risk of stroke incidents or mortality. And in terms of lifestyle intervention itself again somewhat limited in the studies. We still need more studies but we do have information that when lifestyle intervention is used to prevent stroke recurrence or as a secondary stroke prevention patients tend to do better with behavior change when it comes to diet and exercise. It also positively influenced the risk factors of blood pressure cholesterol and weight and patients tend to have better knowledge of stroke and improve quality. We have limited information as to whether or not it prevents recurrence. I want to share this last thing in the last couple of minutes and this is about physician knowledge and barriers. So we know that we have a lot of barriers to lifestyle medicine right now. Okay this health care model that we're in does not support lifestyle medicine right. It is the reimbursement model does not. You're not really paid for counseling. Counseling takes time. If you've ever tried to lose weight yourself you know you fall off the wagon. It's it takes a lot of iterations and so it's not as simple as writing a prescription sending the patient on their way. So it's an investment of time. Patients also sometimes they expect a pill for every ill so they'll walk in and they don't want counseling. They they think that you should give them a prescription for medication because that's how they've been trained. We know that also it requires team involvement. It requires a committed patient right. So some of your patients is going to take a lot of work and so these are some of the barriers that we can't always do something about right away. But something I want us to think about is barriers in terms of we as physicians and other health care professionals. What we know is that physicians tend to offer limited nutritional counseling and this is worldwide even when they've identified that patients need the information. They know based on the diagnosis and what the patient is doing that this information is important but they don't offer it and what's often given as a reason is a gap in knowledge. Physicians feel like I don't know enough about nutrition to provide this counseling. I don't know enough about coaching and counseling patients. I know it's important I don't know how to do it. Another thing is that when you look at a physician's behavior health behavior it actually affects how you provide health care right. So if you are someone who don't practice healthy behaviors you're less likely to provide healthy information for your patients. So your behavior impacts your patient if you exercise if you eat well you're more likely to engage your patients on those things. So just something to be aware that as physicians there are things we can do in terms of additional knowledge on lifestyle medicine and how we practice health ourselves. And so in summary stroke is the leading cause of death and disability. We know that there's a risk of recurrence when that risk of recurrence when that recurrence occurs these patients are at higher risk of disability and also mortality so secondary prevention is critical and so we're able to identify these risk factors and address them and lifestyle medicine appears to be a tool to use in addressing these modifiable risk factors. We do know that more studies are needed. I have here a list of resources for you if you are interested in learning about lifestyle medicine about nutrition about coaching and counseling that you can bring this to your patient. This is a list of resources. And so coming up next is going to be Dr. Stein and he's going to be talking about choosing the right level of post-acute care. Great. Thank you Diane. That was great. For those who are posing questions in the chat box I just want to let you know we will be answering those. We're going to leave some time at the end of the of the session for a panel discussion to address any of those questions. So just stand by we will get to those. I'm going to be speaking today about choosing the right level of post-acute care after stroke and focusing on what is the evidence what do we know and how might we decide who should go where. And most of my focus will be on inpatient rehab level of care versus skilled nursing facility. I will touch just briefly on a home versus post-acute facilities but that's a typically less fraught issue. In terms of disclosures I have received some research support from the American Heart Association that's related to some of the work I'll be speaking about and have done consulting work and some research support in unrelated areas that are listed here. So just as a reminder these are the commonly seen levels of care of rehabilitation after stroke. These include inpatient rehab facilities or IRFs or acute rehab skilled facilities also called subacute rehab long-term acute care hospitals which I won't be discussing much today. They account for a pretty small percentage of post-acute care post-stroke and then I'm not going to be speaking as much about home care services or outpatient services today. So let's hone in a little bit on IRF versus skilled nursing facilities. Why does this matter? Why are we so focused on this particular distinction? Well first of all there's a lot of financial considerations here. IRF care is more expensive. The data that I found suggested that it's about $20,000 per admission versus $13,000 for SNF care. This is not for stroke specifically but for all comers and that reflects Medicare payments not what a private insurer might pay. Length of stay on the other hand is shorter in an IRF significantly shorter than a SNF in general and there is a population of patients who come through IRF who substantially go to SNF about six or seven percent of Medicare fee-for-service patients who go to an IRF initially are ultimately discharged to SNF so you get a double payment there from a payer perspective which is obviously challenging. Conversely IRF care may result in better outcomes and that obviously is very important to all of us and it's something that really we need to understand if we're going to intelligently select which patients receive which level of care. There is quite a bit of variation in the utilization of IRF versus SNF and some of that is very difficult to nail down the sources or the causes of. I will say within my own hospital system there are different parts of the system where there's very different utilization levels and the reasons are local. They relate to referral patterns. They relate to geographic availability of IRF beds. They relate to relationships with skilled nursing facilities but there's a lot of variation in the percentage of patients with stroke that are sent to each of these levels of care. There are also a lot of pressures on the payers to try and reduce post-acute care. Historically that's been the payers themselves but now they are pushing some of that to ACOs and other bundled payment systems where in fact our colleagues may be seeking to reduce the cost of care by reducing post-acute utilization and if you look at other areas for example joint replacement a large percentage of the success of the bundled care payment program has been in reducing post-acute care costs not in reducing surgical costs and there is concern that that may impact stroke care in negative ways if we don't understand who should receive which level of care. From an IRF institutional perspective stroke represents the single largest group of patients who were admitted about 20% of the total and very important sort of core population for many IRFs. There are really no clear standards to help people determine who would benefit from which type of care there are of course Medicare criteria that are in use that you that you need to meet to justify IRF care but that doesn't really indicate who is most appropriate it simply reflects a regulatory perspective and I always like to ask the question where would I want my family member to receive stroke rehabilitation if they needed it because I think that is a sort of litmus test for are we doing the right thing for our patients. In terms of the overall financial picture if you look at Medicare fee-for-service program spending on post-acute care which is really where we have the best data that actually has stabilized overall you can see it's about 58 billion dollars a year that's a lot of money to be sure and this is not specific to stroke this is for post-acute care generally but that number has been relatively stable for a number of years which suggests actually that the total costs in inflation adjusted dollars may actually be gradually falling. If you focus in on some of the subcomponents though and in particular IRF costs those have continued to rise and if you look from 2010 to 2018 those costs went up from 6 billion to 8.4 billion representing a compound annually growth rate of about 4% 37% over the interval but about 4% a year on average and that is not a tremendous growth but that is faster than inflation and and therefore is a cause of concern for people who are looking at these overall budgets. There are federal efforts to try and adjudicate this discussion or dispute about which patient should go where and the impact law which was which was passed in 2014 mandated a common measurement system for post-acute care and MedPAC has been pushing the notion of site neutral payments there's a lot of interest in that in CMS the notion being that the same type of patient would get the same payment for post-acute care irrespective of where that care is delivered. Many folks view site neutral payments as essentially fatal for IRFs because the costs per day are so much higher in IRFs than they are in skilled nursing facilities. And therefore if the payments are the same the the IRS won't be able to to deliver the care that they need to within those those budgets. That's the that's the concern. Now in terms of making this decision about where stroke patients should go we can look to the guidelines. These are from the American Heart Association Stroke Rehabilitation Guidelines, which I helped co-author with a large panel of contributors and these were published in 2016, so they're reasonably current and you can see the emphasis first on organized coordinated and interprofessional care which presumably could be provided in a variety of post-acute settings. But there was a recommendation that patients who qualify and have access to IRF care should receive treatment in an IRF based on the interpretation of the literature that the outcomes are generally better. That has generated some controversy and certainly there are those who are advocates of SNF care who feel that that conclusion may not be justified and we'll talk a little bit about the issues of what do you mean by SNF care and what do you mean by IRF care. In terms of the indications for SNF care from the guidelines, you can see that there was a recommendation that this may be useful for patients who can't return directly home or who don't qualify for but who don't qualify for acute rehab or don't have access to acute rehab. And then also some mentioned that some patients who have completed a course of IRF care may still benefit from skilled nursing facility care. So my usual impulse when facing a question, a clinical question of which type of care is better is to conduct a randomized clinical trial. That is our definitive way of addressing these questions in medicine in general. And this is just a generic template for how you would conduct a clinical trial. It's not specific to this population. But why hasn't that been done? Why is that such a difficult thing to contemplate here? Well, first of all, who's going to pay for it? If the incremental cost of acute rehab is seven or ten thousand dollars more per patient and that's may actually be higher than that because that's for Medicare and that's all-comers, not stroke patients who may be higher than average. There's a very substantial cost that you would incur if you took patients who might otherwise get SNF care and send them to acute rehab. That's going to add up pretty fast. The other issue is how do you standardize care in these two settings? Are we talking about typical care in an IRF versus a typical SNF or are we talking about best practices? Are all SNFs the same? No. Are all IRFs the same? They're not the same either. So I think there's an important question about if you were to do this, really, are you looking at the best possible care in that setting or are you looking at more typical care? There is also an ethical concern about taking patients who clinicians feel need a higher level of care such as IRF and randomizing a patient like that to SNF and that is a challenge that could perhaps be in part addressed by focusing on patients who would otherwise go to SNF and randomize some of those to a higher level of care. And then lastly, would patients and families and clinicians agree to a randomization? And that answer is unclear. In terms of answering the question of which has got better outcomes, IRF or SNF care, there is a lot of effort that's gone into observational studies and the sources of data for that include Medicare data, which is made publicly available in a de-identified form, and then some large health care systems such as Kaiser have their own data sets which have been used. This is a brief summary of a study that was done in Kaiser that looked at about 16,000 stroke patients. It was completed a number of years ago. You can see the percentages of patients that went to IRF. They were relatively heavier utilizers of SNF than IRF care and found better survival among patients who went to IRFs than those who went to SNFs. But the data set that they had had no measure of stroke severity and limited clinical data in general. They were using mostly an administrative data set, which raises a lot of concerns about confounders. Were the patients who went to skilled nursing facilities inherently sicker, less likely to benefit from rehabilitation, and that's why they were sent there. And that's a question that's going to come up in essentially all studies here that are addressing this issue. This is looking at survival curves by the highest level, which is to say in most cases the initial level of post-acute care, showing the differences and how much better IRF care was than SNF care in this study. Another study that is widely discussed is the Dobson-Devanza study. This was commissioned by the AMRPA, which is an industry group for IRFs. They issued a report in 2014. It has not, to my knowledge, been published as a peer-reviewed manuscript, but only as a report. They looked at fee-for-service Medicare patients and looked at a broad range of diagnoses, not exclusively stroke, although they did have a neural population. And they did a matching paradigm where they took pairs of patients that looked similar based on the administrative data set that they had and compared outcomes and looked at that at two years. And what they found was that stroke patients had a shorter stay in IRF than SNF. That's really no surprise. What was interesting is that the IRF patients spent about three months more at home over the two-year follow-up period due to fewer recurrent hospitalizations, due to less time in other facilities, and a significant mortality difference favoring IRF patients, as well as a reduction in readmission and ED visits. This is a another study that was using Kaiser data, a much smaller study, and showed that patients who went to IRFs had more improvements in their function using the ANPAC, which looks at mobility and ADLs and cognition. And this is just showing that for each of those three areas. ANPAC has a sub-score for each of those, and you can see that these all showed differences favoring IRF care. And then more recently, this is a study by Hong et al. that was published in 2019 that tried to address the issue of statistical adjustment for these comparisons between IRF and SNF care. We know these patients are not equivalent. We know there are differences. The average age of SNF patients in general is older than the average age of IRF patients. So the question is, how can we best adjust for these differences? And what they did here is they used a variety of different statistical methods that have been developed for making those adjustments. They did multivariate analysis, they did propensity scores, and they used instrumental variable analyses. And I'm not a smart enough in statistics to explain to you the fine points of all those, but the interesting finding was that no matter how they tried to adjust for the differences between these populations, IRF care came out more favorably than SNF care, that it was, these patients did better. No matter how they looked at it without adjustment, they looked at it with these various adjustments, they found consistently that post-acute stroke care had better outcomes. Stroke care had better outcomes in IRF. On the other side of the equation, there's not as much literature suggesting that there's no difference, but this was a report from MedPAC that found that there was, in fact, a minimal difference between the two types of care. Not no difference, but relatively small difference. This was not specific to stroke, however. So, in summary, the problem with observational studies is that our data sets are limited. We don't usually have good clinical information. We don't have an NIH stroke scale. We don't have FIMS scores. These data sets are really not great. And therefore, it's very hard to know with any confidence that the patients are similar or to adjust for differences when they're there. And whenever you're adjusting for baseline differences in population, it's hard to completely trust the results. And this has been summarized by two famous people here, one about the three types of lies, lies, damn lies, and statistics. And then I like the one by Ernest Rutherford, which is that if your experiment needs a statistician, you need a better experiment. And I think that the notion there is that if results aren't obvious to you that really, and you have to manipulate the data that much, that it becomes much less trustworthy. And I think that there's some truth to that in trying to answer this question. So, there's two other strategies we might employ to try and answer this question. One is a so-called bottoms-up approach, which is to look at the patterns of referral to different post-acute care levels and see what factors are driving that to suss those out, as it were. And the other is to ask experts to figure out how they make these decisions. So, this is an attempt to, at a bottoms-up study, this was done. We actually did this prospectively and we collected data, including Barthel index and other clinical variables prospectively for a modest sample of patients in the Northeast. This was published in 2014. And what we found here is that if you look at the Barthel index, which is really as useful as anything else that we looked at, the Barthel for those who don't use it is very similar to the FIM. What you see here is that patients who have high Barthel indexes, so in the far right of this, those patients tend to go home. The patients who have high function don't need post-acute care and they go home. And that is a pretty clear finding. You can see as you go to the left in this, that bottom bar, the orange one of home gets lower and lower and lower. But when you look at the yellow versus the purple, you can see that there's no clear trend there based on severity. That once you get into the moderate or intermediate groups there, it's just hard to see a clear pattern of who goes to SNF versus IRF, because in fact there wasn't one. So, we concluded in that particular analysis that the Barthel index is useful in distinguishing who needs post-acute care versus who can go home, but it wasn't helping us understand who went to SNF versus IRF. Now, more recently, there's been a lot of interest in the use of artificial intelligence and machine learning to try and extract these implicit rules that are governing discharge decisions. These are, the challenges with these are that the rules are often not explicit. They're sort of black box. You can't really understand exactly how the decisions were made and how the computer determined which patients, which characteristics might drive a patient to one place or the other. There are a lot of concerns about AI and machine learning because of its tendency to unintentionally incorporate cultural biases. If there are disparities in care, this approach won't fix them. They will actually sort of institutionalize them to some degree. You have to make sure that these systems don't, in fact, reinforce existing biases in care that may be based on race or ethnicity, but also could be based on geography or other considerations, rural versus urban, that we're at risk for institutionalizing using an AI approach. And the other issue here is that it's assuming that we're doing the right thing now. These are extracting rules from our current practices, but there's no particular reason to believe we're doing this right now. In terms of a top-down approach, this is something that I'm currently working on and I think is an interesting approach to try and figure out what are the factors that should decide who goes to which post-acute level of care. And we convened an expert panel and conducted what's called a Delphi process, which is essentially a multi-stage discussion and voting process. And this is not yet published, but just to give you sort of a preview of what the key factors that this group of, broad group of clinicians and other administrative folks felt were important, not surprisingly, the likelihood of benefiting from an active rehab program was pretty high on the list. And the need for clinicians with specialized rehabilitation skills was also very high on the list. Less important, interestingly, was the likelihood of discharged home, which I think was interesting. And the availability of caregiver support was sort of intermediate. So more to follow on this, it's work in progress, but I think it's another interesting approach. Once a decision support tool might be trialed, we could then see if this matches what experts believe is the best discharge disposition for patients. So I'm going to skip ahead to our conclusions and then I want to leave a few minutes for discussion. I think it's unlikely that there'll be a randomized controlled trial anytime soon on this issue, but it does remain the most definitive way to answer the question if it could be done. Observational data using administrative data sets have not answered this question to anyone's satisfaction and they probably never will. It's not bad to do further studies like that, but I don't think we're ever going to trust them entirely. The patterns of care are not reliable enough to use AI or machine learning approaches from my perspective to develop reliable rules to guide care. And therefore, I think we still have a lot of work to do to really define a method for deciding who goes where that we're all really comfortable with. So we're going to now turn to our Q&A session. I want to thank everybody for their participation thus far. A number of questions have come in through the chat box. I believe one or two have received chat answers, but we'll also discuss those verbally for those who may not be following the chat box. So maybe Dr. O'Dell, you can start. I think there were some questions that came up to you about the use of the discontinuation of DVT prophylaxis while patients are in rehab and when one might consider stopping it while a patient is still on the unit. Yeah, great question. At New York Presbyterian, as I mentioned in the chat box, we use 150 feet ambulation as not a definitive point, but the starting point of that conversation. And as I mentioned, if somebody is walking 150 feet with the assist of two therapists with a dense hemiparesis in that leg, that's probably not the person that you want to stop chemoprophylaxis on. The other hand, somebody that's walking 150 feet with good return of motor recovery and good ankle dorsiflexion and plantar flexion, walking 150 feet with a cane, that may be somebody that you feel a little bit more comfortable with stopping anticoagulation. I don't know, quite frankly, there's a great literature out there giving us definitive guidance on when to stop this stuff. I think to a certain extent, using intuition and reasonableness is the way to go on that. Fair enough. Thank you. Just a small follow-up question that I had, which is just how practical are pneumatic compression boots on the rehab unit? My impression is that they're kind of a disaster there. Agreed. I mean, to be used effectively, these things need to be on pretty much 24-7. Obviously, it's impossible to mobilize somebody three or four hours a day in inpatient rehab with these things on. I also might mention, as somebody who's had surgery and had these things on, you don't sleep very well, they're a complete pain, and often they're taken off. Thank you. Yeah, my concern is often they're also not put back on. People go to the bathroom and then they're- That has been my experience. Absolutely. Yeah. I find them very, very unreliable in the rehab setting. I think people extrapolate from the ICU literature, which just doesn't make much sense to me. Diane, do you want to address the questions that come up to you? One that sort of caught my eye was the issue of cost. It's all well and good to say, eat right, but it's expensive. We subsidize in this country, unfortunately, the wrong foods. What would you say to the issue? Agreed. Right now, the model that we have, of course, it's going to be cheaper to get some chips than it is to get some vegetables. I think that's why it has to be a commitment, and people have to understand that. We sometimes will send patients to get involved in food co-op, where they're able to get food very inexpensively or sometimes free. There are some pantries around that will offer healthier foods. I send patients to farmer's market when they're available. Things like beans. Beans are much cheaper and easier. They last longer than meat, and you'll get the protein. It's a matter of teaching patients how to cook better and cook in a different way, and the food can be affordable. Fair enough. In terms of another question for you, Diane, and this is a follow-up for me, we're often confronted with patients who have elevated lipid profiles, elevated blood pressure. We know that some of these lifestyle interventions that you're talking about can work for those, but they take time. When you're dealing with patients, maybe it's different post-stroke than it is in someone who hasn't had a bad event yet, but how long do you give these things? Do you use these things post-stroke, or do you use them in combination with medication for somebody that has potentially modifiable risk factors with lifestyle interventions? I think typically, because patients are not always consistent, typically you do use medications along with lifestyle. Once they start being consistent and you're seeing improvement, you can start to move away medication. For some patients, you will never be able to move the medications away. Some people will say, look, I want to try lifestyle, and you really should give them that opportunity. If obviously the numbers are such that you're concerned, then you're going to go to medicine and figure it out after. I mean, safety is the priority. Fair enough. I think that makes sense. Especially when you're dealing with someone who's had a stroke and you really want to make sure that you lower their risk immediately, you don't really have the luxury probably of waiting, so you got to do them in parallel. It's important that you just don't use it as an afterthought, that it really should be involved. Can I answer a quick question? I think we have a few seconds. Regarding the Framingham risk factor score, it depends on the patients. I have some patients who will not respond to bad information. You have to look at the patient and decide how best they respond and give them the information accordingly. The apps that are out there, there are quite a few free ones like MyFitnessPal, HealthPal. You find what works and what you really want to focus on. Great. Well, thank you for that. I really want to thank Drs. O'Dell and Thompson for their contributions today. This is a great session and really great questions. I'm sorry that we're out of time. I would say that we're happy to answer questions offline. My email is in my slide deck. I'm sure Drs. Thompson and O'Dell are reachable as well. I apologize that we've reached the end of the session without answering all the questions. Thank you so much for all the people who were watching for their great questions and participation and to both of our speakers today. Thanks for joining. Thank you, everyone. Thank you.

Dr. Mike Lodell

outcomes

seizure prophylaxis

DVT prophylaxis

hemorrhagic stroke

ischemic strokes

seizures

mortality

rehabilitation therapy

DVT