Innovating with Clinical Intent through Tech-enabled Data Capture

Podcast Transcript

Dr. Alex Rae-Grant:  Neuro Pathways, a Cleveland Clinic podcast for medical professionals exploring the latest research discoveries and clinical advances in the fields of neurology and neurosurgery. Welcome to another episode of Neuro Pathways. I'm your host, Alex Rae-Grant, Neurologist in Cleveland Clinic's Neurological Institute. In an effort to explore the latest advances in neurological practice, today we're talking about innovating with clinical intent through tech-enabled data capture. I'm very pleased to have Dr. Jay Alberts join us for today's conversation. Dr. Alberts is vice-chair of Innovation in Cleveland Clinic's Neurological Institute, and research from the Department of Biomedical Engineering. Jay, welcome to Neuro Pathways.

Dr. Jay Alberts: Thanks Alex. Pleasure to be here.

Dr. Alex Rae-Grant:  Let's start off with an easy question, Jay. Where are you from, and how did your career lead you to Cleveland Clinic?

Dr. Jay Alberts:   I'm originally from a very small town in Iowa, about 1200 people. It's probably a village, we would say. There is no stoplights, a flashing yellow light on the outside of town. I went to a small Division III school, Central College in Pella, Iowa, and then I transferred to Iowa State. It was really when I was at Iowa State that I became interested really in Parkinson's disease.

A Parkinson's patient came into our neuroscience class and said, "I know what I want to do with my hands, but I can't make them do it." I was really struck by that, thinking, that's a real issue, problem for this individual, and that really led me down this path of research and coupling biomedical engineering and neuroscience together.

Dr. Alex Rae-Grant:  Today, we're talking about this particular aspect of your work, data capture technology, and it’s really revolutionary use in healthcare. Let's start kind of at the beginning. Tell us how biomechanics were historically measured and how your efforts have changed with adapting consumer electronics into healthcare, in our case, specifically in neurological practice.

Dr. Jay Alberts:  Sure. One of the very first biomechanics studies was actually done to win a bet. There was a bet between two people. It was related to, when a horse is galloping, is there ever a time when all four hooves are off the ground. Muybridge was contracted to take high-speed photographs of this, and in fact found that when the horse is galloping, there is a point that they are off the ground all four hooves, now, different location than they thought in the past. Nevertheless, that was sort of the first really biomechanics for hire that happened.

And then it really evolved from there, using high-speed photography, 3D motion capture, infrared emitting diodes, et cetera, inertial sensors through accelerometers, gyroscope. Historically, you need a really large space to have a biomechanics lab. Many of those measures were done independent of a large number of patients. You had many studies with 10 or fewer subjects, because it was so difficult to collect the data, analyze the data, and then make something of it. We've really evolved since then to your point, with consumer electronics devices where your phone has an accelerometer and a gyroscope that is more powerful than we had 10 years ago from a laboratory perspective.

Dr. Alex Rae-Grant:  I know you've published and secured a number of grants for work being conducted in MS and Parkinson's disease. Can you talk about some of your projects and latest research findings?

Dr. Jay Alberts:   Sure. So we've done a lot with, really, how can we use these consumer electronics devices to objectively quantify cognitive and motor function? I think the important part here is that we're looking at both cognitive and motor functions simultaneously as opposed to just in a silo, each silo. That's really led us to improve in terms of Parkinson's, the treatment of PD patients, and in optimizing DBS, deep brain stimulation parameters.

So rather than looking at the patient post DBS, just on the cognitive domain, and then sending them across the hall to have the neurologist do a traditional UPDRS, we're now able to use both cognitive and motor functioning simultaneously to help program these patients. That's really an important aspect because, you think about it, activities of daily living have a cognitive component and a motor component, whether you're walking across the street, carrying a cup of coffee, or engaged in a conversation. You're processing the information in terms of, is this car coming, are they going to stop? And then you have to monitor your motor performance as well.

I think that's where we have been able to leverage some of the work we've done with the consumer electronics to bring these types of tools to the clinic or to the provider.

Dr. Alex Rae-Grant:  I know you developed some iPad-based tools that we're using in our clinic and the MS center. Do you want to speak a little bit more about the development of those tools?

Dr. Jay Alberts: Yeah, I'd be happy to. If you remember, I said I was from a small town in Iowa. This all started really around concussion. When we looked at our own management of concussion here at the Cleveland Clinic, we saw some gaps in terms of the assessment of patients and continuity of care, because we had athletic trainers who were in the field and they were operating, didn't have access to Epic, et cetera, so standardization was far from optimal.

We developed the concussion app to help standardize documentation and assessment. We really tried to then refocus the language or the discussion around data, what were the outcomes, or how did this patient present from a data perspective, and that could then lead us to the next phase in terms of treatment or care. Previously, our mantra of care for concussion, not just here but everywhere, was, really rest and pray. Come back in a week or 10 days, if you're better, great. If you're not better, rest more and pray harder.

We really took that out of the equation by now having very objective data to characterize motor and cognitive performance. I'm proud of the fact that NFL is using our software, but more proud that my hometown in Iowa is using it, and other places are using it as well. For example, we've got a group of white middle class kids in the middle of farm country using it. We also have another group of minorities in Inner-City Los Angeles using it. On the surface, these two populations couldn't be any different. However, they share one thing in common, and that is, they both lack access to care. I think technology, in this case, is allowing us to bring their level of care up.

We're really sort of raising the bar wall lovely in the playing field so we can have access to all of their data and provide some external guidance to the management of concussion. That's really where it started, and then it evolved into, as you're familiar with, the multiple sclerosis performance test as well as the Parkinson's assessment as well.

Dr. Alex Rae-Grant:  We talked about MS and Parkinson's a bit, now you've mentioned head injury. Are there other areas in neurology that potentially these kinds of approaches could help us out with?

Dr. Jay Alberts:             Yeah, I think... I would say it goes certainly within neurology and maybe beyond. When we start to think about cognitive impairment and motor impairment, they're not unique to just neurological populations. We're certainly looking at cognitive impairments in a primary care setting for those individuals who are 65 and older. I think we've taken some of these tools to them, to the primary care physicians, and it's been helpful in the sense that... And it's changed their model. In the past, they said, "I really didn't want to measure cognitive function, because I can't fix it."
I think neurologists know that you can't fix it, but what do you do? Just like schools have been doing for decades, you can accommodate it. That's a big aspect of what we're trying to do with the primary care providers to help them accommodate and get to the right provider.

Similarly, we know that cognitive dysfunction is a pretty good predictor of who will be readmitted. So, again, we're now using their role in these tests out to our pre-anesthesiology clinic, and those individuals who have lower than one standard deviation from the norm get flagged, and then we can allocate resources to them before they even have surgery, or while they're an inpatient to reduce readmissions. I think there's a really great opportunity where I think neurology can sort of inform a lot of these other areas in terms of assessment, and again, doing something when we can, or at least accommodating the best we can.

Dr. Alex Rae-Grant:  I know most practicing physicians would say they're getting overwhelmed with data these days. Now, we're capturing more data in terms of biometric analysis. Are there things that we can do that you guys are developing in terms of algorithms or artificial intelligence approaches that maybe could help us with the numbers game here?

Dr. Jay Alberts:  Yeah. Absolutely. You hit on it very well. If you look at different surveys, and such, about physician burnout, one of the top reasons is interaction with the EHR and overwhelming data. I think that's why, here at the clinic, we have a very sort of different approach in terms of technology development. We're not just developing technology for the sake of developing technology. It's not the Bluetooth toaster, which exists, it's a real product. All it saves you is a button press, you still have to put the toast in the toaster. We're really developing technology with clinical intent. We spent a lot of time with the providers, and the Melanson are talking about, what are your pain points, and how can technology potentially help them? And then we spent a lot of time with the patients as well to see, is this technology that you can interact with? Is it reliable? Does it provide you with some level of value, and will it provide your physician with value?

We actually published a paper that shows that when using the MSPT app from beginning to end of a case or a patient, you save about 27% of documentation time. What we're trying to do there is really sort of a mini digital neurological exam. We're evaluating their cognitive function, which historically, again, people understand that it's imperative in MS, but it hasn't been evaluated systematically. Again, challenges within the clinical workflow. Upper extremity function using the Nine-Hole Peg Test, or an electronic version of the Nine-Hole Peg Test on the iPad, and then the timed 25 foot-walk as well to characterize gait and some levels of postural stability.

In our work on Parkinson's, it's very similar to MS. What we're really trying to understand is, looking at some of the cognitive aspects of working memory, so we have a digital version of the Trail's Test which we can then subdivide into a cognitive aspect and a motor aspect. Again, very important when we think about motor freezing and now this whole concept of cognitive freezing, looking at functional aspects of gait, doing the clinical version of the Timed "Up and Go," and collecting these biomechanical variables from the iPad or the iPhone. And then we're also doing upper extremity or manual dexterity testing in these individuals.

To me, I'm excited about that because that allows physicians to do what they're trained to do, which is practice medicine, not type and document.

Dr. Alex Rae-Grant:  I can speak to that as well in the clinic, and I'm a multiple sclerosis specialist. We start to look at the longitudinal numbers on cognition, on hand function, gait, depression measures, quality of life. It helps us see where there's an issue and where the patient's falling off where they should be. So it gives us early warning ideas about what we should be intervening on.

Dr. Jay Alberts:  We talk a lot about AI. The value of AI is tremendous. However, when you look at that, you have to think about the quality of data that's going into those types of algorithms. As you know, the general data in the electronic health record isn't as pristine as we would like. That's probably one of the reasons why AI hasn't advanced as rapidly in healthcare as it has other industries. But I think with the data that we're gathering that's very standardized and systematic, we now have, really, by AI standards, a pristine amount of data or pristine type of data. As we increase the size of that, we can actually do those better disease prediction algorithms, treatment algorithms. Again, that gets me excited because, obviously, for the patients of Northeast Ohio, but also for those patients in rural and underserved areas where they don't have a Dr. Rae-Grant sitting next to them, or someone who they can connect with.

Dr. Alex Rae-Grant:  Obviously, these benefits couldn't be realized without some fairly large collaborative effort between your team in the lab and our team in the clinic. Can you tell us how your team works for the practitioners, and the patients, and other stakeholders to make these technology-enabled tools?

Dr. Jay Alberts: Sure. I think the real key is spending a lot of time listening and observing clinical workflow, and understanding that the technology we develop is really secondary here, and maybe even tertiary. The first, obviously, is the patient, and are we going to do something that the patient can do or will do, and if it's providing value for them, and then it's the clinical workflow. As a scientist who never was in the clinic, I had no idea how important the clinical workflow was. And so we developed early tools where like, "Oh, here's a great tool, you should go use it. It's only an extra 15 minutes." As you know, an extra 15 minutes just randomly added on, doesn't work. It's completely disruptive. That's where we really try to educate ourselves on, what are things that cannot be compromised within the clinical workflow, and then where can we potentially use technology in the clinical workflow, again, to facilitate workflow and the gathering of these types of data?

Dr.Alex Rae-Grant:  Are there additional takeaways for clinicians caring for patients, people like myself, particularly about the use of technology and where we're going in medical practice and technology?

Dr. Jay Alberts: Yeah. I think it's an exciting time. Our ability to characterize patients' cognitive performance, motor performance. All of the things that are important in the neurological exam, I think very soon will be able to be captured digitally. I think that's exciting from the amount of data we can gather, and then use that data to create prediction algorithms, and disease prediction algorithms, and progression algorithms.

What I would be excited about if I were you was your ability to impact patients. It's no longer going to be whether you write a paper, or just a patient across from you. I think your ability to impact patients far and wide is going to increase exponentially.

Dr. Alex Rae-Grant:  Well, Jay, this is great stuff. I'm looking forward to further developments of your work. Thanks for joining us today, and we'd like to see how things go in the future.

Dr. Jay Alberts: Absolutely. Happy to come back anytime and let you know.

Dr. Alex Rae-Grant:  This concludes this episode of our Neuro Pathways podcast. You can find additional podcast episodes on our website, clevelandclinic.org/neuropodcast. Subscribe to the Neuro Pathways podcast on iTunes, Google Play, Spotify, SoundCloud, or wherever you get your podcasts. Don't forget, you can access real-time updates from experts in Cleveland Clinic's Neurological Institute on our Consult QD website, consultqd.clevelandclinic.org/neuro, or follow us on Twitter @CleClinicMD. All one word, that's @C-L-EClinicMD on Twitter. Thank you for listening. Please join us again soon.