…

Amanda Rabinowitz, PhD, and Mike Jones, PhD

Developing Apps for Mobile Rehabilitation

Along with communications, entertainment and interpersonal relationships, modern technology, smartphones and devices such as Siri and Alexa are now at the heart of some revolutionary changes in therapy and rehabilitation, allowing healthcare providers to guide and monitor their patient's progress. In this episode of MossRehab Conversations, Amanda Rabinowitz, PhD, director of the Brain Injury Neuropsychology Laboratory at the Moss Rehabilitation Research Institute talks with Mike Jones, PhD, VP of clinical research and assistive technology at the Shepherd Center in Atlanta about the relatively new field of mRehab.

Read the Transcript

Here is a transcript of our conversation with Drs. Rabinowitz and Jones:

Along with communications, entertainment and interpersonal relationships, modern technology, smartphones and devices such as Siri and Alexa are now at the heart of some revolutionary changes in therapy and rehabilitation, allowing healthcare providers to guide and monitor their patient's progress. In this episode of MossRehab Conversations, Amanda Rabinowitz, PhD, director of the Brain Injury Neuropsychology Laboratory at the Moss Rehabilitation Research Institute talks with Mike Jones, PhD, about the relatively new field of mRehab. Dr. Jones is VP of clinical research and assistive technology at the Shepherd Center in Atlanta.

Rabinowitz: Mike, you and I came to know each other in part, because we're working on this project together that you're the principal investigator for, and that's a grant from the National Institute of Disability, Independent Living, and Rehabilitation Research to study what you're calling mobile rehabilitation or mRehab. Can you start by telling us more about what mRehab is and why it's so important?

Jones: As with most everything in life, you know, the digital revolution has made a tremendous impact on how we do virtually all aspects of life. And part of that is the delivery of mobile health or digital health services to everyone, including people with disabilities. We've been taking that same concept using information and communication technologies to support delivery of rehabilitation and other services to individuals in the home and in the community. So when we say mRehab, we're kind of distinguishing that from the rest of mHealth, if you think of mHealth, it's primarily an individual user who may be using a mobile app on their phone for self-management, personal improvement, health, fitness, diet, exercise, those sorts of things. Whereas mRehab generally involves a clinician being involved as part of that as well. So, the most common scenario would be a, say a physical therapist that is seeing a patient in an outpatient clinic. They would send them home between their clinic visits, with home-based exercises for them to do. And so, when we think about mRehab, it's using tools that can support that home or community-based rehabilitation.

Rabinowitz: I can imagine that that could take a lot of different forms, the types of technology that are used in mRehab. Talk a little bit about some of the different types of technologies and how they're being used to extend that reach of rehabilitation.

Jones: It probably makes sense to start by defining what we refer to as ICT or Information Communication Technologies. You know, that basically is referring to all of the networks, devices, operating systems that we use to communicate with the digital world. Best way to think of it, I think a useful way to characterize it is a convergence of traditional audio-visual networks, so TV and radio. Telecommunications networks, or telephones and computer networks. And probably the of that is the iPhone. You know, the smartphone that includes all three of those. So when we talk about ICT, we're really referring to all of those technologies and how they might be used to deliver services, information, support to all of us, but also certainly including people with disabilities. With respect to mRehab, if you extend that to include things like remote sensors, the internet of things, where now with this vast ICT network and increasingly fast communication speeds, we're now at a point where you can have devices talking to each other. So, it becomes possible to, for example, actually monitor someone at home or have, if you will, the cloud monitors that individual send feedback, you know, other ways that we can take advantage of that to support someone living independently in the community or at home.

Rabinowitz: It sounds like there's just a lot of different possibilities for how mRehab could be implemented, whether it's by a smartphone or other devices that it could mean a lot of different things.

Jones: It certainly can, yeah. It's not limited to any single technology. And, you know, again, we're trying to exploit or take advantage of the emergence of all sorts of technologies, including artificial intelligence, machine learning, how this could be brought to bear in terms of making sense of the vast amount of data that might be collected. If you think about the internet of things, if someone is at home exercising or just engaging in day-to-day activities, and we're able to monitor that level of activity, well that's generating a tremendous amount of information, so much that the clinician just does not have time to really sort through that and make sense of it. So that's where machine learning can come into play to help set up algorithms that are monitoring that data, looking for trends, identifying triggers that might trigger an alert to notify the clinician if there's a change in status or whatever in the home.

Rabinowitz: So, it's not only allowing rehab to happen outside of the clinic and extending that reach, but it's also actually enabling clinicians to do things that they haven't been able to do in their own practice, because the systems can really capitalize on data and techniques for processing that data to give us clinically relevant information that we wouldn't otherwise have.

Jones: Exactly. If you look at the traditional outpatient scenario of a physical therapist that might be seeing a patient in the clinic, you know, there are tons of challenges with that, especially for people with disabilities, not least of which is just the logistics of getting to a clinic visit, you know, there are transportation barriers, all sorts of challenges to get to the outpatient clinic. The therapist may send the patient home with exercises or other therapeutic activities to do at home, but they really have no way of knowing whether they do those or not. So there are issues of adherence and engagement as a person actually initiate the activities that have been recommended. And do they do them at a level of intensity or frequency that's to be expected. The only way you can really determine that is based on self-report of the patient when they come back to the next clinic visit. So, some people will go home and overdo it. Some people go home, do nothing. And the ideal patient might go home and do exactly what's requested. We also know that there's a big discrepancy between what the clinician might see in the clinic when that patient visits and what the patient actually does at home. They may engage in the activity at the level you would expect to see in the clinic, but then go home and not do anything. It's a very common phenomenon, particularly in stroke rehab, where the therapist can see in the clinic that they're using their impaired arm, for example, but then when they go home, they don't really use that arm at all. And of course, there's no way to give feedback or change that patient's program until they come back for the next clinic visit and little information to go on, other than self-report in terms of what changes might be necessary. So, if we had a system, and this is what we're working towards, that you're kind of collecting that data of what they're doing at home when they do their exercises or don't do their exercises. And that information is going up into the cloud, and again, with machine learning artificial intelligence, you could set up parameters so that it notifies the clinician if there's a problem, or it just progresses the patient if there's not a problem, so they can move to the next level of activity or intensity in the exercise.

Rabinowitz: Do you have any examples from your work of patients or participants who've benefited from mobile rehab that you could share with listeners?

Jones: Well, I mean the biggest example, I think most compelling cases that transition home and we have quite a lot of data showing that these technologies can be used very effectively to support the transition home after inpatient rehab. You know, I do a lot of cycling, so I've got titanium in a couple of places in my body, and like anybody who's had a rehab experience, certainly not as significant as what many of our patients have, but it is frustrating. You know, you're sent home with exercises in the old days, you might get a bad photocopy of stick figures that are trying to figure out what's going on with the exercise. So, a little better these days, in that there are platforms that allow you to actually download a YouTube video to see how the exercise is to be done. But the frustration is many people want to go much faster than the rate at which the therapist is recommending, or again, there's issues of adherence when you get home. So I think anybody that's been in that situation can appreciate the fact that they would have much more control if you had a system that they could progress at a rate, that again, the clinician has set up the parameters, but they're able to progress at a rate based on their performance, at a pace that's more appropriate for them. So, I think that almost anybody can relate to that.

Rabinowitz: My area of research is traumatic brain injury and I think that these transition periods of care are so important, and they are places where mobile rehab can really fill in gaps in the traditional model. So, in traumatic brain injury, rehabilitation services are really front-loaded, but we know now as more and more research comes out, that there are chronic effects of brain injury that continue to persist for many, many years, and people are left without the same services and supports that they had early on in their recovery. So, I think that transition from inpatient to outpatient care and from the acute phases to the chronic phases is some places where there are current gaps that mRehab is really poised to fill in.

Jones: Yeah, I think so. And I certainly would not look at this as an alternative to clinicians. I think it's a way to supplement or augment or complement what a therapist is able to do. You know, one of our real interests in this area is if you think of the traditional scenario of outpatient visits and the way funding has gone, particularly under the Affordable Care Act, the good news is that rehab is an essential benefit under Obamacare. Kind of the downside of that is that oftentimes that's now become the defined benefit, depending on the level of plan you have. And it's regulated state by state, you may have 10, 20, 30 outpatient visits a year, regardless of the condition. So you could be rehabbing a knee, or you could have a spinal cord injury or a brain injury, and you're limited to that number of visits. So it begs the question of how can we kind of spread out those visits over time and supplement them with technologies that would allow the patient to progress between them in such a way that it's not taking a lot of the clinicians time to monitor that information. Then it could send alerts that say, okay, maybe this person needs to come in a little bit sooner than their next scheduled visit, or maybe we can delay the next scheduled visit and have the system automatically progress into the next level of complexity or whatever in terms of their exercise.

Rabinowitz: So really kind of maximize the bang for your buck that you get out of that defined benefit and really deploy the clinician's attention in the most efficient and the smartest way possible.

Jones: That's right. That's right. That's what we hope to be able to prove of this by setting up a system that would allow us to compare the conventional approach to a technology augmented approach, if you will, that would be able to get more mileage with the same level of benefit or frequency of visit.

Rabinowitz: Tell us a little more about the NIDRR Rehabilitation Engineering Research Center Program grants opportunity and what you're doing at Shepherd now.

Jones: Yeah, I'd be happy to. If you look back at NIDRR over the years, of course its name has changed among other things, but back when it was the National Institute of Handicap Research in the seventies, they set up four model programs, primarily as service programs, serving the needs of different types of disability groups. So blind, low vision, hearing impairment, deaf, pediatric prosthetics and orthotics, and one on the burgeoning area of computer access. Since then, they've grown to, it varies every five years or so, you know, these were funded on a five-year cycle, but anywhere between 15 and 19 rehab engineering research centers, and they really grew during the George W. Bush era, Steve Tingus was the director of NIDRR at that time and was a technology user himself. And so, he was really a major advocate for the RERC program and that's when it almost doubled in size. As I say, most of these centers, and historically they've been focused on specific disability areas but probably starting in the late nineties they began focused on different aspects of technology. We were fortunate enough in 2001, there was a partnership with Georgia Tech to receive a RERC grant on mobile wireless technologies. And of course, that was when smartphones were just coming out. And that was a great 15-year run on that project. But a lot of the concerns then, and those continue today is the challenges of using these technologies by people with disabilities. So there's a group of RERC's that are focused on what's called ICT access to make sure that these technologies as they emerge and become more and more a central part of our lives, that they're accessible and usable by people with disabilities. Early on, much of our focus was on that very fact, how do we make sure that smartphones with touchscreen interface are going to be usable by somebody who's blind and can't really see the actual display? But more and more they've been focused on functional areas like tele rehab. So, through the years, this will be our third RERC. We had one that's just winding up on community living and the use of technology to support living at home. And then most recently funded center on mobile rehabilitation, which has a great group of partners, of which

you're one of them, reached out to our team a couple of years ago, wanting advice on tech transfer. And I was so taken with your project when this opportunity came up, I immediately called and said, we'd love to get you on board with this. Your use of conversational agents as they're called a kind of Siri's and Alexa's of the world to help support individuals with traumatic brain injuries that are struggling with depression and need to take advantage of positive approaches like behavioral activation as a way to manage depression. I thought it was aningenious idea and was really intrigued with the artificial intelligence machine learning aspects of your project. And you were able to bring to our team, George Collier, who is our big data scientist that's helping with your effort as well as overall collaboration on this home-based, what we're calling the sensor enhanced activity management platform. The other collaborators are Dave Reinkensmeyer who's a brilliant biomedical engineer and professor at University of California, Irvine and Dave has been involved in home-based therapy for many years. So, he was a natural to get involved with this. In fact, a lot of the ideas that ended up in the proposal came from a conversation we had at ACRM I guess, two years ago now, and he'd done a presentation on some of the work he's doing. He set up a company called Flint Rehab that is developing sensory enhanced rehab devices that can be used for home therapy. And I was just really taken with some data he showed on how much better engagement is, folks who when give them a sensor-enhanced device that can monitor what they're actually doing, and then present that in a kind of gamification strategy that motivates them to kind of stick with the program. We've seen two-to-three-fold the expected level of adherence in terms of implementing exercises at home. So Dave's part of the team and then mentioned Flint Rehab, the company that he helped start. And then there's another company PT Pal that has probably the most widely used therapy management platform, both inpatient and outpatient, but the majority of their business is on the outpatient side. Again, it's kind of the latest, greatest way that a therapist working in outpatient can send a patient home with prescribed exercises that the patient has an account, they can look these up when they get home, they can see how to do the exercise, they can document whether or not they've done them. So it's the perfect starting point for what we're trying to build in terms of a merged platform that includes those management components, but then sensor-enhanced devices and exercises that can collect these data that can then help the therapist then drive initiative for that patient.
Rabinowitz: It's a great group that you've assembled, Mike, under this RERC. And I just want to say it's been a thrilling group to be a part of, and we've discovered lots of areas of complementarity that I don't think we even knew at the outset. So, it's been a really great collaboration.

Jones: It has certainly been a discovery process. We had some general ideas, in fact, John Dzivak with PT Pal, we're having a conversation, and he said, you know, we're pretty sure there are different profiles of patients. You got the person, the jock that's going to kind of overdo it. You got the person who maybe because of pain or other issues is going to underdo it, but his interest was being able to profile patients and try to figure out what is it about some people that stick with it and others that don't. And as you know, you're a part of those kind of weekly meetings now, kind of pouring over the data that PT Pal and Flint have and finding really some incredible insights into what's going on at home. And you know, what seemed to be predictors of who's going to stick with exercise. Can you identify, antecedents if you will, that suggest someone might be about to drop off, and it begins pointing to things we can do to keep them with it and make sure that they get the best benefit.

Rabinowitz: I mean, and that's been a really interesting discussion to be a part of because then we think, you know, what can we take that we know from motivational theory and theories of behavior change to intervene when we notice a patient might be on that path that suggests that they might stop adhering and can we intervene before they get there and actually boost their adherence and get them better outcomes and more benefit from their treatment.

Jones: More importantly, can Siri intervene, you know, can we set up the system, so it actually automatically identifies those patterns that suggest someone might be dropping off or not doing optimally and then build in nudges, if you will, to help keep them with their exercise.

Rabinowitz: So what do you see as some of the potential impacts of this RERC grant? Where does this group go from here? What are some of the impacts on clinical outcomes and patient care?

Jones: That's a great idea. I mean, I think the first thing we need to do is demonstrate proof of concept. Right now, as I say, we're very much in the discovery phase, but hopefully in the near future, we'll be able to say, okay, we can identify antecedents that might suggest someone who's going to drop off of their program or stop entirely. And we can intervene to change that, to keep them with their program beyond the expected time at which they would drop off. So I think the first step we've got to prove that concept, having done that, it'd be great to move into what we call proof of product, implement that in a clinical demonstration that as I say, perhaps compares the conventional outpatient therapy approach to an approach where the therapist can set up the parameters of this sensor-enhanced activity management system. And it will progress the patient or send feedback back to the clinician to encourage them to intervene or change the program, etcetera. If we can show with that demonstration, that it is more effective and efficient than the conventional model, then I think we're really getting somewhere. And then what I'd be interested in beyond that is if that works, how readily is it adopted? You know, we've done some preliminary work, surveys and focus groups with clinicians to get their take on this. And I've been frankly surprised at how much interest there is and how much folks are already using, for example, therapy management platforms in their practice. We're seeing about 13% of the folks we've reached out to in a national survey. So they're already using some online management tools for home therapy for their clients. 90% said they love this idea and think it's something that they could certainly implement effectively in their practice. There are concerns about how well it integrates with their existing work processes to make sure that it could be integrated into practice effectively. And part of that is reimbursement. So I think we've got to make the case and perhaps COVID has helped make that case for us, that these strategies can be used effectively, do have benefit and therefore can be reimbursed.

Rabinowitz: So the field might really be ready to start to adopt some of these things. It sounds like and there might be the impetus now to make the reimbursement scheme get on board.

Jones: I'm hoping we're there, we've talked about it for years, but I really think that part of the silver lining of COVID is that maybe we now have that demonstration of the value.

So, you know, we've talked a bit about how you folks at Moss are using conversation agents in your rehab project, using that to promote the use of behavioral activation in your patients with brain injury. I know there are other efforts underway at Moss to, for example, use remote sensor technologies, as well as artificial intelligence in rehab. You know, our overall focus has been on mobile rehabilitation in rehab. I wonder if you could share with us your thoughts about how that might change practice at Moss and what you see the future holds for both your patients and clinicians there at MossRehab.

Rabinowitz: Well, one of the things that obviously has been accelerated by the pandemic as we've discussed is this need to deliver more of our services remotely. And so that's something that we've seen accelerate on the clinical side, in a number of different efforts, but also something that I think has shaped and put to the fore, some of the work that was already happening at the Research Institute, so I can give you some examples of that. One thing that I had been involved in now, going back starting a little over a year ago, was a pilot demonstration project supported by the Brain Injury Association of Pennsylvania that's being funded by the PA Department of Health that was looking at remote delivery of cognitive rehabilitation services. And so Moss clinicians and some other partners in the area, we've been providing them with technology, with iPads and high speed internet that they can give to their patients and actually do deliver all or part of their outpatient cognitive rehabilitation for brain injury remotely. So that is the clinical program. And at the Research Institute I'm leading a program evaluation project to evaluate the efficacy and the feasibility of that intervention. So how difficult is it? What equipment do people tend to need? Do we reduce missed appointments? Are clinicians finding that there's some things that they're able to do better because they're actually reaching the individual in their homes instead of the clinic? So that's a project that was already underway and we're eager to see how those results pan out, because we think that getting more reimbursement support for those kinds of remote services could really open up a lot of possibilities for serving our population. And that's just one example of something that I think Moss has been involved in that I only see broadening in the future. In the Research Institute, a number of my colleagues have been involved in projects that use technology that can be also implemented remotely. So things like virtual reality interventions for motor rehabilitation in the context of stroke and other clinical issues. We have a number of projects that use remote sensor data that we've evaluated using wearable technology to gather actigraphy data, to get a sense of what people's physical activity levels are like. We have projects that are looking at that in individuals with traumatic brain injury. And we've also been involved in evaluating a method for assessing sleep after brain injury that can be implemented remotely. And so we can get a sense of what individuals might be struggling with sleep problems, which are really common after traumatic brain injury. So those are just a few examples of things that are happening on the clinic side or things that are being developed in the Research Institute that we can really see some immediate clinical applications of.

Jones: Fascinating. It sounds like Moss has certainly been at the forefront of using these technologies, certainly during this era of COVID and continue to serve our patients at a distance. I think it'll be exciting for all of us to see where we go from here. Envision a new normal post COVID where this will become part of our standard practice throughout rehab.

Rabinowitz: Yeah, I think you've noted this, Mike, a lot of this is we're adapting on the fly to the circumstances that we're in right now, but there's so much promise to actually increase the accessibility of care. And I think that everybody has recognized that there's transportation barriers, there's financial barriers, there's a lot of things that make it difficult to serve our population in the traditional model where people have to come to the clinic. Home visits have always been a part of rehab, but those can be costly as well. So, this kind of explosion of digital platforms and this pressing need to adapt what we're doing, I think could really result in an improvement in our service delivery model that lasts far beyond the situation that we currently find ourselves in.

Jones: Well, we know necessity can be the mother of invention. I think this shows that it also can be the mother of adoption in terms of the use of these new technologies and rehabs.

Rabinowitz: Do you have any advice or recommendations for people who want to learn more about mRehab?

Jones: It's a burgeoning field, as I say, but I'd certainly encourage people to check out our website, mrehabrerc.org. We've got a start of a bibliography there. And so there are a lot of references to both our work and others', much of that is still conceptual, certainly, but that may be useful to folks that are interested in learning more about this area. For those folks that are in rehab and active in ACRM check out the technology track, the technology networking group. There are a number of presentations have been over the last couple of years and I think a couple planned this year that are focused on mRehab strategies.

Rabinowitz: Well, thank you so much for your time, Mike, it's always, it's always such a delight to talk to you. I learned a lot about mRehab and the other things that are going on. So this is really a treat. Thanks for speaking with us today.

Jones: Well, it's my pleasure, Amanda, and thanks so much for taking time to chat.

That was Amanda Rabinowitz, PhD, director of the Brain Injury Neuropsychology Laboratory at the MossRehab Research Institute and Mike Jones, PhD, VP of clinical research and assistive technology at the Shepherd Center in Atlanta. Go to mossrehab.com for more information on this topic and be sure to check back at our website for more episodes or subscribe to this series wherever you get your podcasts. For MossRehab conversations, I'm Bill Fantini. Thanks for listening.