Welcome, everyone, to our Member May session. My name is Jordan Burkhardt. I am the Director of Member Engagement for AAPM&R and I just have a couple of housekeeping notes before we start our session tonight. The views expressed during this session are those of individual presenters and participants and do not necessarily reflect the positions of AAPM&R. AAPM&R is committed to maintaining a respectful, inclusive, and safe environment in accordance with our Code of Conduct and Anti-Harassment Policy available on our website at aapmr.org. All participants are expected to engage professionally and constructively. And then this activity is being recorded and will be made available on the Academy's online learning portal. An email will be sent after this activity with a link to bring you to the recording and the evaluation. For best attendee experience during this activity, please mute your microphone when you're not speaking. To ask a question, please either use the raise your hand feature and unmute if you're called upon or use the chat feature to type your question. I will be reading the questions aloud for the recording and so I will address the questions directly to the panel. And depending on time, we may not be able to field every question, but we will do our best to answer all interested parties. And with that, I am going to hand it over to our esteemed host, Dr. Moon-Yu Oh Park, and I welcome you guys to this session. Thank you very much, Jordan. Good evening, everybody. My name is Moon-Yu Oh Park. I'm Chief Medical Officer at Burke Rehabilitation Hospital, White Plains, New York. I also serve as Chair of Age-Friendly Care in Rehabilitation. It used to be called Geriatric Rehabilitation. I did not like the name, so I requested AAPMR to change the name, Age-Friendly Care in Rehabilitation that was approved. So, that's kind of the legacy community, but just a different name. So, today we have this exciting topic of technology and digital health use in aging population. So, as we know, technology, including AI, is already here in the healthcare space and everywhere else. So, it's our responsibility to understand their use and their roles so that this truly can be helpful for the patients as well as ourselves and other providers. So, this is not going to be a PowerPoint presentation. It's going to be a panel discussion. So, I would like to introduce our panel. There are three outstanding physicians. So, the first speaker is Dr. Lisa Merritt. Here she is on the beach. She's a board-certified physiatrist and a leader in our field. She served on the board of AAPMR. She was actually the founder of the resident section of the AAPMR. Currently, she serves as an executive director of Multicultural Health Institute, where she had multiple high-impact initiatives. One of them was positive aging senior wellness series. So, she has this not only PMNR experience, she's also trained in community and family medicine health. So, multiple experience and training. She is an undergrad of Georgetown University, received her medical degree from Howard University, residency training in family and community medicine in UCSF, and completed her PMNR residency in Baylor College of Medicine in Texas. So, she's also a daughter of an artist and engineer, where she melds the innovation and evidence solutions in approaching the patients and the community and population health issues. That's our first. Second speaker is David Chen. Dr. Chen has extensive experience in medical informatics, technology, and AI application in healthcare. He is currently a teaching faculty of a PMNR residency program in Gaylord Specialty Health in Connecticut. He served as a chief informatics officer in Edith Norris Rogers Memorial Veterans Hospital in Bedford, Massachusetts. He got his medical degree from SUNY Downstate and then received medical informatics master's degree from Harvard Medical School. He also did a fellowship biomedical informatics in Mass General Hospital. Board certified in PMNR and preventive medicine and certified as a specialist in clinical informatics. Our last panelist is Dr. Andrea Taylor. Dr. Taylor is chief medical officer of Burma INC, a medication adherence and dispense system that leverages telemedicine, VR, AR, XR, and AI and machine learning to enhance medication compliance and improve healthcare outcomes for the patient. He is currently on the board of directors for dependable home care, licensed home care agency in the New York state. Dr. Taylor serves as a technical medical advisor to several renewable energy, recycle, clean water, and healthcare ventures at startup phase. Dr. Taylor graduated Stanford University School of Medicine earned his MBA from the University of Michigan Stephen Ross School of Business. So they are our panelists. So I'm going to start with the question and perhaps Jordan, if you can, maybe we'll just go to the questions. So all the panelists, this is a question, all the panelists, why technology, AI are important tools for the future of a healthcare of aging population? Who would like to take that first? Maybe Dr. Taylor. All right. So my experience, you know, operating a, several home healthcare ventures and long-term care ventures. Right now, the, there's not enough workforce in the, to handle the aging population in, in the community. Technology, in addition, technology will also help us improve outcomes, which is greatly needed, especially in, in underserved areas. So, so the, we're hoping that the, the implementation of technology with, with trained people who not a new, new and a new user interpret that technology will improve the outcomes and, and make it a better experience and keep the people in the community longer. Thank you. So Dr. Chen, what about, would you maybe see this like a, why this is important to buy? Maybe you can see it by stakeholder perspectives. So I like to thank you again for the introduction. Can you guys hear me? Okay. Yep. All right. So I like to think of this as it's a huge, it's a huge and moving target when it comes to the use of artificial intelligence and technology and healthcare. And unfortunately, or fortunately, a lot of it's not proven. There's a lot of hype. So I hope in today's discussion, we can kind of get a better understanding of what is hype, what is in the future and what is available today. So to do that I like to think about stakeholders who is going to benefit from these technologies, right? So let's break that down into a couple of key domains. Number one, care coordination is, is kind of what we think of as the glue in the communication process between the various stakeholders. And in this case, I'm thinking the patient, the family and the care providers, right? And I, I'm going to loop in hospital with the care providers all in one section. The other, the other big category to think about is medication management that we'll talk about. Patient safety is the next category. And as you know, gait instability is one of the big tenants that we focus on. So that includes things like fall prevention. The next thing is telehealth applications followed by digital health apps and diagnostics. So let's go back to care coordination. And you can think about four major systems, right? The first that we can think of is something that I like to term AI driven care coordination systems. And what, what does this mean? Well, this means today we have the ability to do predictive analytic platforms that identify high risk patients needing an intervention before something bad happens. So for example, there was a 20 month study at Corwell health spectrum health, which kept around 200 patients from being re-admitted yielding about $5 million in savings, right? And this program use AI predictions and proactive support to address clinical behavior and social needs, right? This report, this report, I believe was also published in 2023 and new England general medicine catalyst. The other, the other section under care coordination is this idea of interdisciplinary team coordination platforms. So technology can bring about these, what we call, I'm sure hospitals have their own digital dashboards and that aggregates patient data for multidisciplinary teams, right? And it really does allow for geriatricians, therapists, and pharmacists and social workers to kind of communicate together. An example of this would be like the PACE model, the program for all inclusive care for elderly. And you guys, Dr. O'Park, you've heard of that? Yeah. Yes. So, so, so that is indeed an example of comprehensive care model integrating both medical and social services. And especially when using that under the digital dashboard, it can facilitate such comprehensive interdisciplinary coordination. Some other ideas, right? One of the things that, and Dr. Andre, you work in home care, right? Yes. Yes. Yes. So you understand the importance of what we call transitional care management tools. Yes. Yeah. These are, these are also available today and they're digital platforms that facilitate discharge planning from day one of hospitalization, right? And the whole purpose of this is to connect hospital teams with community resources to ensure home preparation and follow-up care. And this, of course, the ultimate goal was to reduce re-admission through models like the Coleman Care Transition Intervention and Nyler's Transitioner Care Model. And that was published, I believe, in the Commonwealth Fund in 2024. Are you guys familiar with that, right? The other thing is integrated care provider dashboards. So web and tablet applications to aggregate data from multiple sources, RPM devices, telehealth, medication dispensers, everything at a glance. So these are, these are tools available today that that folk can use. So I like to think of that as kind of like the basis upon which the AI is then built on top, right? Okay. So moving on, medication management is a huge problem. Geriatric polypharmacy is a problem and as average reaction is a problem, right? So how many, how many clinicians, I don't know, or how many folks have the experience where they see a patient and they a, they see a patient, but the patient has seen, you know, four or five other specialists and each specialist piles on some medication and they've been on these medications for 20 years and no one's done any medication reconciliation to actually take people off these medications until something bad happens, right? That is, that is a fairly common thing in rehab that we see. So this is another part in which the AI systems and technologies can help, right? So number one is making sure that folks take the right medication at the right time. So we have smart, this smart pill dispensers and Dr. Andre, you are familiar with that, right? Very. So integrating, so let's finish up on the smart pill dispensers, creating audible record for medication intake for remote monitoring by care teams, electronic. So you have all these gadgets, electronic pill bottles, ingestible sensors, AI camera, face verification systems, et cetera. And this was published in the pharmacy times 2025, but you know what? It's all the same idea. Have some AI algorithm, make sure that folks take their medications. Okay. Automated dispensing systems, right? For electronic medication administration, et cetera. We all, we all have seen that. So that's another example of technology. Now they just tag on this AI thing. All right. And again, polypharmacy management solutions, utilizing AI driven decisions, support tools to review complex medication regimens with automatic red flag, you know, for potentially inappropriate medication combinations, et cetera. And there was a, there was an interesting study using chat GPT, a large language model showing promise in assisting with these deep prescribing decisions. And I think this was a mass general mass general Brigham study in 2024. Now I think I think that covers the medication management section. Does anyone have anything to add to that? Well, the only thing I would suggest is this, when you mentioned the PACE program people have to understand that when you're, you're, you're looking at a PACE program, you're looking at a program that's very, very similar to what, what a lot of people would call a staff HMO model. So everybody, you know, you started talking about a capitated risk model system where everybody has skin in the game is at risk, financial risk. So those kinds of models, you know, when you have the, the providers that working together closely, they work, but it doesn't necessarily mean it's going to work for everybody. I mean, you know, staff HMO staff models didn't work for a lot of other, other people and other groups. And they ended up defaulting to a financial managing model. So that doesn't, you know, that would be my only caveat. The other thing what you were saying about with manage, I mean, the medical models, the medical, I mean, the medical medications a lot of it also has to do with patients don't understand why they're taking the medication. You know, so what we're trying to do is use AI metaverse and virtual reality to, as an instructive tool to not only instruct the patient, but also instruct the caregiver why medication adherence is so important. You know, especially when you're talking about difficult, you know, difficult medication regimens, but those would be my only two caveats. And you did a great job. Yeah. Yeah. So I want to just point out the whole emergence of biosensors is also an interesting aspect to, to marry with that because, you know, I wrote a project 25 years ago using palm pilots to try to do remote blood pressure, blood sugar monitoring, uploading. And now we have that real time. So not only could we look at how we're tracking medication compliance, but also adjustments. If somebody's blood sugars are trending too high or too low, because you're getting under better control, then that's where maybe instead of waiting for a two or three month follow-up, the nurse that's tracking that patient and doing the home health, et cetera, say, oh, we noticed that when you were doing your numbers that you're dropping a little low. This is one of my main things is the challenge of orthostatic hypotension. And that people come in, you can control the stroke, you're dealing with that whole post-stroke environment. You're trying to get people stable. They often have two and three other meds added and they're not, we're rushing them through so fast, we really haven't had a chance to stabilize where they should be at. The other thing I wanted to just speak to was the care coordination, my love, my first home. And I think this is great to know that we're right in with that because we used to have to do this sort of manually. I mean, I put together a care coordination program for a large ACO over, it was almost a hundred mile catchment area, multiple facilities and organizations. And I would call it, we're gonna look for the big five in terms of the cross-referencing the top diagnoses that we knew caused the most demand on the system. And that those were like five to 10% of the total patients that we had responsibility for. So we were looking for the diabetes, the congestive heart failure, the respiratory pulmonary people, things like that. And so by identifying those patients and the frequency of visits, even if you don't have AI, so to speak, within whatever system you're in, you can also prioritize. It's a challenge to get the resources directed. That's the beauty of the whole ACO system and what they were looking, going towards, we know what's gonna happen now, is that you had the freedom to remodel the delivery of services and your resources to prioritize. So it made more sense to invest in that than trying to stabilize how many diabetic acidosis ICU patients. So I think this is really exciting stuff. Yes. Thank you. Andrew, you wanna say something? Yeah, no, I was only gonna say in New York, we used to have a long-term care program called the Long-Term Home Health Care Program in oil and body program where what we did, we were able to use money. We had a bucket of money and we would use that money not only for addressing the patient's needs, personal needs and medical needs, but also the social determinants of health. And that made a real big difference in addressing the needs like brittle asthmatics, people with COPD, people with congestive heart failure, et cetera, et cetera. So yeah. But you do touch upon a couple of key points here, right? Number one is tele-remote monitoring under the domain of telehealth and also patient safety, right? Because we were mentioning all these cool devices. And I know in one of our previous conversations, and I don't know if folks on the call know about this, but we did some research onto the smart floors that you guys were talking about. So a couple of things. I'm gonna just quickly briefly mention all these cool gadgets out there. We have smart floors, high-tech flooring, it's made by Japan's Kuroyama Group, or Kuroyama Magic Floor. It stays firm for walking, but it becomes a cushion when someone falls on it. It reduces the impact force by about two thirds compared to regular flooring. And over 560 hospitals and care facilities in Japan have adopted this technology, according to JSTORS in 2025. The other thing that's interesting are these wearable hip airbags, right? So they're lightweight belts or vests with sensors that inflate around the hip during a detected fall. And there's the WOLK hip airbag showed nearly 50% reduction hip fractures rate in a Dutch nursing home. There's also Helight's hip air and tangle belt. Anyway, it's interesting that folks are coming with these interesting gadgets. Others include ambient motion sensors, contact-free radar sensors, like the Wallabot Home, detecting falls without wearing these wearables. Now, more realistically, I'd like to give a mention to smartwatches and these pendants that people wear with fall detection capabilities, like the Apple Watch. And so after detecting such a force, it would call emergency services, et cetera. So that's interesting. So what does technology also come in here? Well, we have a gait analysis and fall prediction algorithm. So we're anticipating to see a lot of these companies come up with these algorithms that they're gonna sell for walking patterns, predicting falls, predicting neurological decline way in advance, right? So that is actually interesting. There are companies out there working on detecting Parkinson's, detecting neuromuscular diseases, just based on voice patterns, right? You were talking about that, Dr. Andre. So that's interesting. Now, I'm gonna quickly just talk a little bit about telehealth after the COVID-19 pandemic. That became a really hot buzzword. People like to do video conferencing. And prior to COVID, I don't think that was the thing. But now we have a hospital at home model, right? Where you have remote monitoring and telehealth combined to deliver hospital level care at home, cutting costs around 30% while reducing complications like infections. And it's been well-established in multiple countries like Australia, UK, and Israel. So that's actually very exciting news. Actually, I think the Mayo Clinic was working on a study on that as well, maybe Andre knows about that too. But basically the future is basically moving people out of hospital beds and having their hospital at home. So we also have various interfacing. Folks think of cell phones or laptops, but there's actually also TV-based telehealth interfaces, especially for seniors, right? Andre, you're nodding to that. Yeah, because we're working on that. What you wanna do is convey the information in whatever form that that particular person wants to accept it in. So a lot of people are not computer savvy or have a tablet or whatnot, but almost everybody has a television. And if you can go through the television, that would be the primary way of getting the healthcare information that they need in the format that they wanted. Yes, so actually, you know what? That is true. These tech companies have indeed developed these telehealth on TV solutions. For example, Cisco and American Well prototyped a device to turn any TV into a telemedicine screen aimed at older patients. Quill Health, I don't know if you guys heard of that, a joint Comcast independence health venture is building a hybrid smartphone TV platform supporting patients through procedures like joint replacements, including delivering care plans and even letting users send wound photos to their doctors from home. So I think this was like 2019 to 2020. So that's interesting stuff. Now, the other thing that we should be interested is this idea of a virtual rehabilitation, right? So a tele-rehab platform that enables guided therapy at home with remote therapists. With the use of wearing device tracking wearables, right? For gait or balance in real time with a large to the therapist when patients struggle. So that is also something we can look forward to because that is quite interesting. So apparently randomized trials have found home-based tele-rehab can be non-inferior to traditional rehab. For example, a 12 week mobile guided rehab for outerly hip replacement patients was statistically no worse in mobility and balance outcomes than standard in-hospital therapy. So patients timed up and go and balance test improvements were equivalent, establishing that virtual rehab can safely substitute for inpatient rehab in such cases, right? And that's the caveat. It's a case-by-case basis, may not work for everybody. Everyone has different demands and complications, et cetera. So these results echo prior studies. Also, there are other studies echoing similar results with knee surgery with similar recovery using tele-supervision. So much to be explored. Yeah, this is exciting discussion. And then there are some kind of information what I hear is that it's gonna be less and less hospital-based care and the care will be more in the community, right? So that's one of the reason why this technology is gonna be very important and that is related to the cost cutting, right? So these are all very good kind of example. And I wanna ask Dr. Married whether you could give, share a patient story or a case where the technology-based solution offer the meaningful care and create a positive patient experience. Sure. Dr. Shen's covered a lot of some of the things that I was gonna talk on, but I'm gonna share a couple of stories. First, with regards to the watch, this is just a little anecdote. I was visiting a friend recently and we were at a restaurant and he sort of plopped down heavily on the cushions. It was a low couch. And then we were just talking and suddenly his watch kept going off. And then finally he looked at it and was already putting in a call to 911 because he hadn't responded to it. And it had detected that he had had a fall, which he of course canceled. But ironically, he had had a pretty significant fall while cutting some tree limbs on a ladder and fell off the ladder and broke some ribs and the watch did not trigger. So accelerometers, they do work, but all of this technology unfortunately is not fully consistent. So that we have to keep in mind that human and high touch and all of that stuff is still, there's gonna be a place for that. So the first case I wanted to just share, this is my work in Northern California, UC Davis. I work both the inpatient hospital as well as consults and my brain injury clinic and get a lot of interesting stories because we're a catchment for the whole Northwestern area. And so some people live in very rural environments. And this one lady, I love the story. In my TBI clinic that she was walking down the hallway and fell, slipped or something and hit her head on the way down and knocked herself unconscious. And she lives alone in a remote area and her faithful Dalmatian went about trying to revive her, jumping on her chest, looking at her face, et cetera, until he finally resorted to dragging her with his teeth, pulling on her, like a dog does retrieve its young, et cetera, to the bedroom where he knew her phone was at. And in dragging her, she sort of woke up and realized what is going on and was able to use her phone to call for help. And she had had a brain injury, fractured ribs, fractured pelvis, it was pretty bad fall. So the dog saved her life, but it was like she needed the dog to get to the device. So that was case one. The other case that I thought was really interesting was a gentleman who, again, getting up at night, just like we mentioned about people getting up in falls at night, especially nocturia, one of the reasons why we wanna keep those bowel and bladder programs going well and try to help improve that and have the lights and things like that at night. So he fell and knew he had hurt himself pretty badly, was not able to bear weight on his limb, was in pretty bad pain. He was able to call out to his Alexa to turn on the lights and sort of swam himself down the hallway on his back to pull his phone off of the nightstand by the court and remotely undo his digital lock and call for help. So he also had a fractured hip and pelvis and pretty bad, but help was able to get there to assist him. And so these falls pendants, watches, different things can certainly be really helpful to summon assistance, particularly when there's loss of consciousness and the person lives alone. You know, these could be important things. They also have these sleep trackers, they have rings now and things can track different variations in sleep patterns, smart beds. And like you mentioned, the remote room sensors and things like that, that might be helpful for families and even cameras to help track people's activities in the room. I think that the whole risk of falling is such a major one that we deal with in our specialty that the more we can use and help educate patients and their families to minimize that, particularly people that we know live alone and in precarious situations. And even if somebody is there, but in the next room, they may not quickly be able to be aware of something that's happened. I think that the, also the use of sensors for patient education. I have some patients that in one of our diabetes education group that we were doing in our clients, one lady came and she was so excited to share with the group that her insurance, her Medicare, Vantage, whatever had given her an insulin, I'm sorry, a blood sugar monitor that tied to her phone, but she didn't know how to use it. She didn't understand what it meant. And we were all oohing and aahing, everybody was oohing and aahing, but she didn't have any understanding still about what she should eat, what was a good blood sugar and these other things. So the teaching was still needing to be there. And also the monitor wasn't exactly fully consistent. And so making sure that people have a full level of comfort, especially seniors with dealing with technology, they can be very daunting. They may feel hesitant to ask questions and rely on their grandkids to help them to interpret the thing or not bother to use it because they're so intimidated by it and they don't want to be flustered. So we have to keep that in mind that high tech is not always the best thing. High touch is still important. Also let's show one last case of a person who called me late to say that their watch said that their heart rate was 160, should they be worried? And I was like, well, what are you feeling? I'm worried. And so taught them how to take mental heart rate, et cetera. Well, long story short, what was happening was she was going in and out of AFib and then the watch became a way for her to track her heart rate while the system worked its slow way through to get her into cardiology, to get her an actual evaluation, to put on a halter and et cetera, and to eventually get to the ablation for her intermittent atrial flutter and fib. So, you know, those heart tracking things that can be done through the phone, through the watch, they're not entirely 100% accurate, but they certainly can give a good indicator. And if people aren't comfortable knowing how to check their own pulse and those kinds of things, and it can alarm if you set it for certain parameters and make people aware, same thing with blood sugar levels for people who get hypoglycemic. So those are just a few of the case studies that I wanted to share with the group and that I too am excited at the possibilities of these devices and what it might mean for quality of life and for caregivers, as well as family members and patients that we treat. We rely a lot on caregivers, but it can be really nerve wracking. I know my dad had dementia and atypical Parkinson's and number of falls and wandering in the whole nine yards. And had we been able to use some of these devices, we might've been able to keep him home longer for his greater safety. But with two elders, both sick, it became untenable. And so I think about this now and I'm really excited about what it can mean for families as well, because it's really exhausting to stay up all night, keeping an eye on people who can be pretty erratic and impulsive. Thank you. Thank you for sharing that. And then I can actually share my own experience of using a Fitbit, for example. So when you're doing exercise, typically these kind of devices will tell you, hey, you're earning a zone activity. Your heart rate is moderate or a higher vigorous activity and your heart rate is over a certain range. So one day I was very sick. I actually think I could have a flu or something, so high fever. And then my Fitbit was telling me I was earning a zone activity because of the fever, I had a very high heart rate, but the person thought, not the person, but the Fitbit was thinking that, interpreting that I was exercising. So it's very intuitively, we should know this immediately, but the machine doesn't have that kind of ability. So just this thing in terms of the force, right? So, a lot of things are in the community, we have this ambient, you know, recording or some wearables and so on. But in the hospital settings, it can be a little tricky. So what Dr. Chen mentioned about this smart flowing is a very effective. So when you look at the data, so in Japan, actually, all the hospitals, in store, this floor had zero injurious fall. So it's, and then actually there are some hospitals in the East Coast introducing this in the hospital settings. So I'm expecting actually more and more hospitals will use that. And the other part, at least in our institution, start to use is a toilet sensor. So when the person, the most of the falls in the hospital actually is related to some kind of, you know, blood of our activity. And then when they're in the bathroom and then when they try to stand up, the sensor actually detects it. And then that sends an alarm. So we are in the middle of analyzing that data and patient actually do not always like that because they are alarm, it's like, you know, they are tired of all this alarm, right? So that get used to it. And then your bed is alarming, your chair is alarming. Now I can't even go to the bedroom, the toilet sensor is alarm. So all those things, you know, some of the things are also have to look at, from the patient's perspective as well. Okay, so enough of that part. And let's move on to the, again, you know, I wanna kind of talk about, we had a lot of discussion earlier about the patient medication management. And as we all know, in the inpatient rehab setting, it's medication reconciliation is a mandatory thing. Admission, discharge, we also do in addition to that, we do in the middle of it. When they are in the outpatient setting, Dr. Taylor has a lot of actually experience and real time, the true device experience about this. So could you tell us a little bit about the current status of medication burden of the aging population in the US and then how the technology-based solution can help in terms of medical error as well as compliance? All right, so to give you guys some background, NHANES did a survey of older adults in the community and they found that among older adults, four to five prescriptions per day was the average. There was a study done of older adults receiving home healthcare services and they found that between seven and nine prescriptions of medications per day. So that gives you an idea of how many medications people in the community are taking and trying to keep those medications on track of all those medications is hard. I have healthcare professionals that have hard times keeping that many medications together. So what you really want, you want to look at things like medication dispensers to get the medication adherence and those medication dispensers could be smart pill boxes. That would help you track the adherence, also track, give data-driven alerts to the people who have vested interest in making sure that you are compliant with your medications. The other things you wanna look at, there are advances in EHRs with AI-driven clinical decision-making. I mean, Cerner and IBM Watson are doing a lot of that clinical support and they're also looking to address medication adherence and delivery. Dr. Chen did a really good job on big data analytics and predictive modeling, really, really good for risk stratification and population health. Probably one of the best companies that are doing that is Optimum. The emerging technologies right now, I would say some of the best technologies in the medication delivery and adherence and compliance are gonna be the things that do patient education and patient engagement. And that's gonna be your augmented reality, virtual reality, your metaverse. And the goals there will be to try to understand, get the patient to understand why they're taking the medication and how non-compliance affects their conditions. So the great thing about all of this is that the integration of AI, big data, remote patient monitoring and telemedicine, it creates a huge dynamic ecosystem which is reinforcing the medication adherence and delivery. Yeah, so I have a follow-up question about that. So this education of the patient and the family member so the ones I have seen was a phone app and then you scan the medication bottom and then it kind of shows the pictures and then what is the use of the medication. Is that what you're referring? Actually, no, I'm talking more about telepharmacy where you actually may have a pharmacist with a specifically tailored demographic tailored educational sections for that particular patient. So suppose you have a Chinese person with a 10th grade education level that doesn't speak English, you can tailor that education specifically for that particular patient and give him the information that he needs in a matter that he's gonna understand or whatnot. So that's what I'm talking about and you can do this with the machine learning and AI and stuff like that. Okay, so we look forward to that. Can I add something to that? So there was a study, well, a Michigan hospital's emergency department saw medication reconciliation errors drop by 90% after implementing an AI driven EHR integration system. And this was reported in TechTarget September, 2021. And the point of it is to highlight how automating very tedious condense like thick data, right? When it comes to drug, you know, like adverse events, you know, indications, contradications, et cetera, met history compilations, that is best used by technology, right? And I know Professor Lisa had also mentioned machines are not always right. And there's gonna be errors even by machines. But I think there's a bigger take home point from tonight's discussion, which is we talk about all this technology and advancements and we talk about AI, like it's going to save everything. And the truth is it's probably not because technologies do not solve social problems. And medicine apparently is an art as well. And so the art of medicine is a social thing. So no amount of technology is ever gonna fix that if there's a problem with it. So I think we just have to be very aware of that. So whenever someone, maybe a patient learns of a new interesting tool and they bring it to the clinicians, maybe we got to think about that and have a conversation that, you know, maybe it's a, maybe a good tool, maybe it may not be, maybe it's over heights, but just to keep that in mind, I guess. Thank you so much. That's a great point. Yeah. The other thing I wanted to give what the people in the audience know is that 10% of all community dwellers in, you know, in the community that are taking, that are older have adverse drug events. And if you're receiving a home healthcare services that jumps up to 15 to 30%. So these adverse medication events or drug events, they're not, you know, they're, it's, you know, they're significant, you know, they can make a significant difference, especially when it comes down to course. And what I was gonna say about what Dave, follow up with Dave is a lot of it is reimbursement systems and what you reimburse. You know, if you're not reimbursed, if you don't structure them correctly, you're not necessarily gonna enforce the kind of activities and actions that you really, really want that are gonna help address the outcomes. And right now, the way the reimbursement is structured is not doing that, but that's another discussion. That is a social problem. Yes, okay. So we are gonna move on. So we are gonna change a little bit topic to a slightly controversial topic, longevity. So AI is utilized a lot and in terms of the drug development now. So it used to have a huge, like, you know, and to be able to do a randomized clinical trials and all those things. And by utilizing a lot of this, you know, AI models, you know, patient enrollment, the number of patients to be enrolled in the study is becoming smaller and smaller. So I want to ask how will the AI transformation transform potentially the longevity research in the next decade or so, Dr. Chan? That is a very loaded question. Yes. It's also a fast moving. So to kind of simplify things, let's kind of break it down to major categories, at least in my head. So the first category is what we call next generation biomarker discovery and refinement, right? And so these are like multimodal aging clocks, right? And that's using AI to unlock interesting, I guess, correlations between genomics, proteomics, metabolic nomics, et cetera, imaging digital biomarkers from wearables, blah, blah, blah. Others are organ specific aging metrics, right? There's a lot of interest in that. And then finally, real time aging tracking, right? I'm just going to say that all of this is partially there to unverified and speculative and very conceptual. And that is pretty much what it is now. All right, so I'll get back to that. But I think what Dr. O'Park was talking about was AI accelerated drug discovery pipeline, right? That's what you're alluding to before. So we have a little bit more data to this that we could talk about with examples. So that includes that category, some subcategories, right? So we're talking end-to-end AI drug developments. We're talking about combination therapy optimizations, right? We're talking about using AI to gain insight into repurposing drugs at scale. And then we're also talking about personalized dosing algorithms, right? So I'm going to spend most of my time talking about these items because there's much more data on that. But I think the panel's also interested in other things like AI enhanced cellular reprogramming, right? In silico rejuvenation models, digital twins, et cetera. So I'm just going to also say that using AI and nanotech and cellular reprogramming is speculative and it's unproven and there's really nothing right now. So let's get back to the AI drug discovery when there is something to talk about. Okay. So the first thing is you guys have probably read in the news. There's a company called Insilico. So what they did was they use AI-based screening of large drug libraries to identify unexpected longevity benefits of approved drugs, right? So the Insilico TNIK example is a case in point. They use AI to find a known target, TNIK, which could be inhibited to mimic youth-like effects in cells. More broadly, so projects like Drug Age Database and IBM's Watson Health have used algorithms to mine literature and omics data for drug longevity links. So for example, repurposing drugs like Miformin or A-carpals or rapamycin analogs, et cetera, okay? So that's kind of interesting, right guys? But let's talk about something that's even more cool. So there's a drug, ISM001-055, and it is also made by Insilico. And this drug is in clinical trials for idiopathic pulmonary fibrosis, right? Human trials began in 2022. But what's cool about this drug was it was designed by AI, right? Again, this is a TNIK kinase inhibiting TNIK. TNIK has been shown to clear out senescent cells and might act as a senolytic. So in 2023, a study by Insilico's team reported TNIK inhibitors as a potential anti-aging strategy. And therefore, this demonstrates that AI can discover compounds with geroprotective effects. So what are they doing now? Well, they're looking into other types of targeting aging pathways, senolytics, NAD+, boosters, et cetera. So we should expect more of these to come in the next three to five years, or at least that's my thought. And so that is an example of AI-accelerated drug discovery pipelines. Yeah, thank you. Yeah. So we're not gonna, you know, we are rehabilitation specialists. However, there are, like you go to social media or like, you know, JGPT, there are so many things about living long. We don't talk about why we have to live long, but obviously this is a topic and thank you for your insight for that. So we have only five to 10 minutes left in terms of we are gonna open up a Q&A. So the last question is, what are the ethical considerations around the AI and technology use, especially regarding equity, data biases, and patient privacy? I can start with any panelist. Dr. Merritt, if you'd like to. Yeah, I mean, I think the first thing always is truth. Yeah, true. And unfortunately, as we know, AI can hallucinate. I love that term and the ambient AI. What's nice is that, for example, you're in your visit with the patient, it will capture the essence of that visit and create a note for you. with the patient, it will capture the essence of that visit and create a note with the interim history. If you want, you can narrate your physical, we'll put together your problem and your plan and even put the codes down. But you have to still read it and monitor it because it will be a little too helpful sometimes and thinks it knows your style and start making suggestions and put things in there that was not what you said or did. But it will level the playing field in terms of the patient's voice perhaps being heard and captured better by busy clinicians who may not always record that part because they're focusing what they're gonna do in their head. So there's nuance I think we need to speak about, is nuance. And I think there's the ethical obligation as clinicians that whatever we sign our name to, that we read and that we agree is our language and our assessment truly. I think the whole issue of bias, we know that that is going to be an ongoing challenge in terms of the people that develop this, it's humans. And in terms of the way that the things are trained, we'll have to make sure that there is a variety of inputs that it's reflecting our multicultural population, multi-ethnic, multi-genetic, so many things that we need to make sure that we're still at the beginning in terms of it's only regurgitating in essence and re-putting, formatting and re-approximating available information that's out there. So if there's a large body of evidence and we already know a lot of research, we're just still beginning to catch up on equity and doing research on women from a gender standpoint, much less a variety of other groups. So we need to make sure that there's inclusivity to the sources of information that the AI is being trained on and using. When it comes to things like facial recognition software and identity and things like that, it's a huge challenge and a problem for nominated people. And even with pulse oximetry as a technology, we know that studies have shown that there can be a great degree of variability in the accuracy of that. And I deal with that clinically all day long that you cannot just rely on the pulse oximeter. If somebody is tachypneic and they're saying they're short of breath, give them oxygen, okay? Because the pulse oximeter may be in the battery or may just not be reading correctly. And that can happen time and time again. That can be a temperature issue and things like that. So still use your clinical judgment. Be the doctor is my message. I love all the gadgets. I've grown up through all of this stuff and it's cool and it's hip and I get it. And I love all the future stuff that Dr. Shen is putting forth. And I thought we'd be living like the Jetsons. Now we're almost there. We've got like personal drones and stuff, almost like Jetsons, but we're not quite there. We're still feet on the ground. And I was at a meeting in LA and the young man was offering to share a ride together. And he said, well, I'll call a Waymo. And I was like, nah, I don't think so. I'm not ready yet. For the driverless cars, you see them going all around. I want to test that technology just a little bit more before I want to climb into a car that nobody's driving that I can see. So, you know, it's up to our own levels of comfort and a sense of obligation to duty to the patients for their safety and optimal care and excellence always. And as we grow together with, as the field emerges, we need to be part of it. All of us went to the CES Consumer Electronics Show as part of a contingent from the American Academy of Physical Medicine and Rehab, asked experts to come weigh in on digital health issues. And it was exciting, but also there's a lot of things going maybe down the wrong road. We need to go a different way that we can inform, particularly as physiatrists. So we're comfortable with technology. We're on the cutting edge and I'm excited to be part of this and to be part of this panel. I hope we can get some questions and I don't know if there's any from the group. Yes, so I think that I'm going to ask Dr. Taylor also make a comment about the last question. So any ethical considerations, Dr. Taylor? I was going to cede my time to Dr. Chen because I wanted to hear what he had to say because we're running out of time. Yeah, real briefly, Dr. Chen. I would say with the ethical concerns as clinicians, we have to be aware of them and we have to make sure that the technology is not taking us down the rosy path that we don't really want to go under. So that's, I mean, people have to remember what happened with opiate narcotics. I mean, we went down that road and ended up causing a total disaster. So we can't do this with the technology and AI. We just can't do it. So that would be my concern. Those are my concerns. So the data is biased because it just is, right? And I think this could be a very big detriment or it could be an opportunity. And the reason I say this is because you see health is a global, it's a shared common goal, right? It can be used for changing policies. It can be used for making, changing rules in cities and guidance, et cetera. And I think this does touch upon public health. And so we have an opportunity to use AI and let me just dream a little bit. Imagine an open source sharing of data, right? This would make the AI much more equitable in the sense that the data is much more diverse. The data sets much more diverse and is not owned by any one company or anybody who's well-funded. Then it's open to the public, right? And then that means the data is much usable because it's relatable, it's pertinent, to whichever societies which are benefiting from it. So imagine a small town in the middle of nowhere versus a big city. Imagine gaps in terms of wealthy folks versus non-wealthy folks. You're gonna have a disparity there. But this changed, right? Using this AI thing. It does have the ability to level that playing field if we do it right as a society. If we don't, then it's just going to make things worse and perpetuate this inequality even more. Yes, so as a counterpoint, technology will, or it's not will, can potentially make the wealthy healthy. So because there is a some degree, who is gonna pay for it, right? As a CMO, who is gonna pay? Who is gonna pay? If we are gonna give all these things to the patient, who is gonna pay? That's where the problem lies, right? So I would like to kind of end our discussion here and then we would like to open the floor for the questions. Anyone? This is Dr. Friedman. Can you hear me? Yes. Okay. Oh, you froze. Just wanna say that in some sessions regarding fall detection and fall injury protection devices. Yes. I have a prototype device for fall prevention for folks that use a walker. Folks that use walkers often fall when they're not using a walker. And my reminding walker device attaches to any walker and can detect the walker. As the person is actually standing, it reminds them in that real time moment that the walker is there. And the running signal is friendly and reinforcing. And I'd be glad to talk to anyone about it in the future. That's a nifty concept. That's very exciting. It's kind of like the pill dispensers that say, it's time to take your pill. It's time to take your pill. And a little human dimension to it. I like that. That's cool. So what's the name? Is it smart walker? What's the name of the walker, Dr. Friedman? Well, I call it the reminding walker. Smart walker? Or something like that? No, I call it reminding. Reminding. Reminding. Gotcha. Because anybody can forget. We don't wanna be told that the walker is smarter than us. Right. Just wanted to have a gentle reminder. Yeah, so- Everybody needs a reminder now and then. Even you and I. Thank you for sharing that because I saw a similar kind of a concept in one of the hackathon. You know, I went to Princeton to see the hackathon. I was a judge in the patient safety section. And these, like, you know, 20 year old students, they actually designed the smart cane. Very similar concept. And then they had the little cameras, things like that. So clearly that is one of those things that are coming in the pipeline. Yeah, so yes, there was a cane a blind person had that in rooms that are set up, the cane will talk and narrate location, which was really helpful. Yeah, it was connected to GPS. So it would tell you exactly where you were and where you needed to go. It was really cool. Pretty cool. Any other questions? Let's hear from the audience. Did you guys like this talk? Is it interesting? Did you learn something? Maybe Dr. Slova can talk about the academy and the registry. That's our technology data thing that we've been working on for years and years. Well, I have actually a question for you. I am somewhat of a Luddite. Luddites were people who fought against technology because they thought it was going to replace human activity. And do you struggle with acceptance of technology, not only in the elderly population or patients, but also amongst older physicians? Because we know how difficult it is to change culture and change the way you practice. And so do you come up against some resistance to this technology, not only from patients, but also from possibly more senior physicians? I think that's an excellent point. And I want to just say that it really depends on how you introduce it and if it can be seen as a win-win. Like when we did our positive aging program, we had young people doing oral histories with seniors as a start, and then helping them with their devices on how to use their phone or their iPads to get online and do different things like playing together with it. So it was less controversial and off-putting. UC Davis, they have a wonderful IT team where they'll come and do hands-on. Like I had the IT lady come again as we were implementing this AI thing that wasn't really working as well. As for PM&R, we're a little complicated, but we have different ways of doing it than just a straight family medicine and easy soap kind of format. And we both learned, because she was like, oh, well, you know, review system's not essential because you don't go for that. No, review system is very important. That has to stay in. That is, I do ask all those questions. That is part of the examination. So helping to be part of the creation of the templates and to participate in it might help overcome some of that fear and apprehension, but it's a very valid point because until some people began adopting it to see how it actually could be helpful and reduce stress and keystroke time and give ASL more face-to-face time with the patients, there was a bit of resistance. But now that some people are beginning to adopt and they see the benefit and the relative accuracy with just a little bit of tweaking, people are starting to relax a little bit and trust the process. But you're absolutely right. It really has to be individual. It has to be the way it's kind of introduced and oriented to have a level of calm. Not just dump something in here and click on this link or do this QR thing and then just imagine we're supposed to instantly understand it because we don't have antennas in our heads. You know, it's like, yeah, it's a good point. Andre? Yeah, I was gonna say, you know, when you look in the community, initially you have resistance, but if the patients keep using and getting used to the technology and then see the huge benefits and the things they couldn't do before, then you get a huge adoption of whatever the technology or whatever you're offering or whatnot. COVID and the grandkids really got everybody on Zoom, you know, because people wanna stay connected to their family. That was the first amazing step I thought. The problem I have in the community is with the providers. They're a lot more resistant to adopting or changing how they do their workflow than the patients, you know, especially when it's giving them information that they're not even sure what to do with it. That's where I see there's a resistance with the providers or whatnot. Yeah, if it doesn't help make things more efficient, it's gonna be some resistance because people are so overloaded already. They don't wanna click one more box or thing because she was running around, well, I can add this and I can do a smart link over here and then you can, I was like, no, that's like five clicks to do what I could just type in. That's not helpful. Yeah, that's why like me and Dave agree that you've gotta, you know, if you're presenting data to a provider, it's gotta be in a format that they can quickly understand it and have actionable suggestions and stuff like that. Dr. Andre, I would say that there's an old wise saying, if it's not a leverage, it's a liability. So, you know, I agree with James. Sometimes we should not be the first to jump on because it just may not be ready. And I think the problem here is the over hyping of, take for example, AI and healthcare and note writing, et cetera, right? Like the whole problem with note writing is for billing. So again, that's a social issue, right? And we're just adding on extra tech, trying to fix something that doesn't really make sense in the beginning. So a different approach would be, and this is just me hypothetically, this is not, this is probably never gonna happen, but bear with me. If a electronic health correct system should be about medical decision-making, right? So imagine a Wikipedia style in which you only edit and annotate the relevant portions of it. The rest you leave untouched. Does that make sense? Taking the billing out of it, right? Then you would solve a lot of issues. You wouldn't need fancy stuff just to get something done, right? Because a lot of times 90% of it is just getting through the main template and the grudges. So again, that's a social issue. So- I don't know, Dr. Chen, I don't know if I'm hearing you correctly, but I think that it's a very important vehicle for continuity of care, particularly when we're talking about the documentation, the assessment, our physical exam, the history, the plan, the problems, the diagnoses. So instead of- Those are all important to be, and some of that remains static, and you can create templates so that does stay static, but you do have to update your physical. If somebody is getting weaker between the time you saw them and they see them, then that means that that's gonna trigger other decision-making. Yes, but I think in my head, I'm not visualizing it out for everyone, but imagine, so the Wikipedia approach is you only edit whatever has changed, and everything else remains static, but you can see the edit changes through time, right? So yeah, the idea is to save us, to save clinicians from having to rewrite the whole thing over again. That's why the cut and paste is a problem. That's why copy forward is a problem, because people don't wanna rewrite, and then rewrite and annotate a little bit, it's hard. So by not having to do that from the beginning makes a lot more sense, but the billing people won't allow that. So we have only two minutes left. So I think we will table the EHR issue for the next session. And then if we don't have any other questions, I would like to thank all the panelists. Thanks so much for participating and preparing, and also all the participants who were in late night in the East Coast, and then early night in the West Coast, and in the middle. Thank you. Thank you all very much for attending. Have a great evening. Thank you so much. Thank you as well. Bye. Take care. Take care, guys. The future is bright. The bottom line is the future is bright. Future is bright.

technology

AI

healthcare

aging population

geriatric care

machine learning

telemedicine

medication management

data privacy

ethical implications

healthcare outcomes