Exploring the Future of Healthcare: AI, Wearables, and Digital Twins with Katie Donohue McMillan, MPH

Ty Hagler: Katie, thank you so much for joining me on the medDesign podcast.

Katie McMillan: I'm glad to be here. Thank you.

Ty Hagler: Yeah I've just been so impressed with your writing and your thinking and as you and I have been getting to know each other, you told me a story about a Grand Rounds presentation you gave at Duke seven years ago. I think that'd be a fun way to maybe introduce the audience to who you are and maybe some of the predictions you gave at that point.

Because I think you talked about some of the gaps in the care for OBGYN providers, and you had set out a vision. Do you mind sharing some of that story? And maybe has any of that vision come to pass?

Katie McMillan: Yeah I have equal amounts of interest and inspiration in both digital health and women's health and I've worked in my career in both of those spaces and so at that time, I was working at Duke University health system and their innovation team and was starting to learn more about this kind of emerging area, which people were calling FemTech at the time, which is like new technology utilized to enable, improve health for women.

At the time, as somebody that was thinking about getting pregnant and was in my reproductive years, I started learning about it from my personal interest standpoint. And then got my boss to approve me to take a month to do an essential, huge market research project to look at what was happening across various subclinical areas within obstetrics and gynecology then I presented my findings to the entire department of over OBGYNs while I was 6 months pregnant, and so it wasn't it wasn't even the littlest bit intimidating.

Ty Hagler: But at that point in time, you were an expert too.

Katie McMillan: Yeah, I actually saw some of in the audience and I was like, this is so weird. It was great. It opened the door for a lot of fun conversations, actually, during my visits.

T he biggest feedback that I got from that was like, and the point of it was essentially your patients are going to start coming to you with these consumer tech products that they are hearing about on Instagram or via their friends or in magazines. And you need to be prepared for these conversations and you need to be aware of them. And you need to know what works and what doesn't work. And the cacophony of feedback was like, how do we know it works? This is still really early stage. And I feel like it was a completely valid question.

And it's an area that I've tried to address. Throughout my career ever since then pretty much and I've done that in a variety of ways from to build up a database. I do a lot of writing. And so I've written various articles that have done deep dives into things like perimenopause hormone tests or pelvic floor at home, pelvic floor therapy, fertility trackers, etc.

One of the things I like to do is find out answers to questions that people have and then share that knowledge with the rest of the world and hope that it does something.

Ty Hagler: And so you set out this vision that there are all these consumer tech products that would be coming into the market. And then it sounds like that's continued to bear fruit. Where it's not like we've slowed down on I don't know, maybe low clearance data feeds. Is that a good way to put that?

Katie McMillan: I would just say there's thousands of companies out there now valued at a combined billions of dollars, I think over 100 billion dollars at this point. Yeah, a lot is a lot has happened.

Ty Hagler: Yeah. And do you feel like there's been a heuristic or a shortcut that's been developed?

Other than the great insightful writing that you've been doing and reviews on this, do you feel like the providers and OBGYNs have built a kind of a better sense of like which of those to trust, or is it still more of a wild west?

Katie McMillan: I'll say yes. And no I think of the things that I love to do is dig into PubMed and look at the peer review journals. And so when I hear about new technology, or people ask me about certain things that's the 1st thing that I go to see. On the website of that company, do they have any studies that they've published? And some of them could be, in early days you don't really have the ability to run a large clinical trial, but they may have done some early stage publications on that are white papers or things like that and that's a good starting point. If I can't find anything there, then I go to peer reviewed journals and see if I can find something there. But I do know that clinicians with their 15 minute time slot with a patient do not have time to run all of those research things, and that is one of my goals, maybe for this year is to actually build that giant database that has the products and has links to the peer review journals and the research and then maybe assigning some sort of quality score for it, so it's an easy glanceable thing for both clinicians to make decisions about things to recommend to their patients, or if patients come to them to say, yes, this is validated or no, it's not. And for consumers to make better choices about how to spend their money when it comes to their health care. But there are some other organizations, so one is called the digital medicine society. They have just recently put together what they're calling the dime seal. Which organization software companies can submit an application for them to review to say, does it meet certain standards for patient engagement, security, clinical validation, science, etc. And then Peterson health technology institute is also doing something similar. They're looking at it more from a payer side of things as far as does it improve. Clinical outcomes and save money. And so they've done some initial reports just around diabetes and hypertension. And so they do more like slices of things. And but neither of them have really tackled women's health.

Ty Hagler: Thank you for those resources. We'll make sure we link to them in the show notes.

So I'm gonna bring up something.

We haven't talked about this, but as you were sharing kind of consumer data feeds the Owlet Infant Monitor.

I'm a early fan of it because I think I dug up maybe their 2009 business model canvas pitch when they were students forming it. And I think I've seen a lot of different groups use business model canvas. And I think they were one of the ones that most effectively used that tool to rapidly iterate on their business model.

And of course, they've become one of the early entrants to kind of consumer tech and health monitoring for infants. I'm curious, you're just, and that's just me from like an innovation practitioner. Like they, they understood the framework. I'm curious from your perspective, like how you've seen, cause they've also, they went down the consumer tech.

Lane first because they didn't want to go through FDA clearance and then have had some hurdles as a consumer product now trying to break into more kind of a FDA approved device as well.

Katie McMillan: I have some serious thoughts about it.

Ty Hagler: Okay. Do you mind sharing those? Because I'm curious.

Katie McMillan: Yeah. They're not the first ones to go through the consumer and then have to backpedal and decide to go down the med device side of things. I did an interview with the CEO of bloom life, which is an at home contraction tracker, which they got shut down by the FDA and then they had to come back and become an approved medical device.

And that's available if anyone wants to read it, but yeah, I have a few friends that are pediatricians and they personally do not love the Owlet because they think it's like a parent panic device they said that newborn sleep for parents and children is already very fraught. And adding something else that the kid could potentially become irritated by, or that is alerting the parents constantly on their phones doesn't necessarily help people like, somehow the human race has marched on for millennia without this. This isn' t really necessary.

I think Owlet tried to do, and I'm going to be curious to see what the outcomes are of it is to have it more be something that is prescribed by physicians, so it's not just accessible to any paranoid parent out there, but rather somebody that maybe you've come home from the NICU or your child has some sort of other medical complications that you're trying to actively monitor. And then in those situations, it may make sense. So I'll be curious to see what are the outcomes they're looking for, are we averting crises here? Are we averting ER visits? But if it's just parents waking up in the middle of the night to make sure that their child is still asleep there's other ways you can do that with just a regular baby monitor where you can hear whether or not they're awake.

Ty Hagler: Do you necessarily need another alarm or, false positives that

It would be alerting and adding to the anxiety of parents versus, trying to offer more of a sense of calm.

Katie McMillan: Yeah. And I'm like, Very interested in the quantified self thing I've had Fitbits and Apple watches. I have an Oura ring now. I think it's interesting, but I think the question is always, what do you do with the data? And what is the value of the data? Can you parse through the data personally and be able to make sense of it without going into a panic?

Ty Hagler: Yeah, which is where the whole regulatory side of this of are you making recommendations with this kind of what kinds of analytics can you draw from this? And is it based in clinical evidence? So we met at the FemTech event that UNC hosted.

Katie McMillan: Yes,

Ty Hagler: Yeah, that was really fun. It's a great event.

Kind of showcase of some of the FemTech companies that are in the triangle area and one of the common areas we've learned is we both come at healthcare from a design thinking and entrepreneurship standpoint.

You mentioned that you had been part of a early design thinking education program that you'd help put together as part of some of your time at Duke, is that right?

Do you mind sharing some of that story?

Katie McMillan: Yeah that was actually my pre-Duke days so after my masters program, which I graduated in 2003, I went to work for a software development company. And when I was there, I met a med student named John O'Donnell. I don't know if you've ever come across him, but he's also in that kind of med tech space in the triangle. And at the time, he was just a 1st year medical student and was interested in health tech. And so we got to know each other. And then I started sharing some concepts around design thinking that I was learning about at work. And he was like, we need to make this more broadly available to people that are coming up through healthcare education.

I decided to teach a class for about six weeks after work. It was like five to seven. And we put out applications for it and students from, you had to just be Enrolled in a current, either undergraduate or graduate program at 1 of the universities nearby. And so we got over 100 applications, I think, for 20 spots, which was crazy. And it ended up really awesome because we had industrial design students from NC state. We had nursing students, pharmacy students, public health students. Business school students, and everyone got to pick a problem they wanted to work on for the course of the week. And so we'd spend the 1st, half of the class kind of talking about the framework and where we were in it.

And then they would apply those lessons to their project that they were working on. And the end we had, everyone presented and we had a big pizza party and we charged next to nothing for the class just to essentially cover the cost of the pizza. Not even our time. It was just like pay for pizza and some champagne at the end of it. And it was really enjoyable.

Ty Hagler: Cool. Do you know if any of those students, any of the projects went on to do anything or more of the education opportunity?

Katie McMillan: So what's really wild is none of those projects really became anything, but over the last year, 3 of those students, like their work has come across my desk in a way that I was not anticipating. I did an article where I was actually talked with a at UNC that was writing these chatbots around reproductive health for women.

And so I was looking at pub med to see, okay, what are the different studies that have been done on how effective this is around increasing health literacy and whatnot. And this one author that I kept seeing over and over again was one of my students.

And so that has become her area of study has been, like, the use of technology to improve reproductive health education.

And so she's done a lot of work on chat bots. And so that was really fun. And I was on another webinar and somebody was like, okay, it was a, about a new software company out there creating an AI tool to help providers quickly triage through their notes and then they also had a patient portal, to try to get information faster, I guess about their healthcare conditions. That was trustworthy and they had the CEO gave most of the talk and then they were like, okay, and now they introduced their chief medical officer to answer questions from the audience. And I was like, Mark, he was one of my students back in the day.

Ty Hagler: wow.

Katie McMillan: So now he's the chief medical officer of a healthcare AI company.

Ty Hagler: Incredible.

Katie McMillan: I was like, man, I really know how to pick them. These people have gone to be more successful than I am. It's been really cool. I think that's one of the fun things about working with students is you never know what seeds you're gonna plant and then see what comes out of them.

Ty Hagler: Yeah it seems like the skill that you've also taught there is like helping them to then further navigate because maybe necessarily the first output of that process. Versus the next sequential iteration of it that is something that ends up somewhere unexpected, but prepares them to be better innovators when kind of lightning strikes, and they have that big opportunity in front of them.

Katie McMillan: Yeah, I think so. I was an Odyssey of the Mind kid growing up

Ty Hagler: Yeah.

Katie McMillan: And are you're familiar with it?

Ty Hagler: Yeah. It was created by an industrial designer.

Katie McMillan: Okay, perfect. Yeah. My dad was a mechanical engineer and he was our coach.

But I feel like just that teaching people, creative problem solving skills no matter where you access it during your life, whether or not it's when you're 7 years old, or when you're 22 years old it's just something that you can continually pull from for the rest of your life when you face different problems in your career.

Ty Hagler: Okay, Odyssey of The Mind, and then history is part of your background, and then Masters of Public Health. And you've woven together a really interesting story for, how your career's developed. I've been fascinated, getting to know your background. I guess just, you recently wrote an article on some of your twists and turns and how you've come into digital health and technology.

How do you see the forest through the trees, as you say?

Katie McMillan: Oh, such a reflective question

I think that my whole career, I've always approached things with an open mind and with questions. And when I was at Duke, my, a lot of people would come to my door and they would say, I want to solve this project or solve this problem with technology. And I would always say tell me more about the problem first. And so then we would figure out what the scale of the problem was. Is it just a you problem? Is it a departmental wide problem? Is it a industry level problem? And then from understanding the scopes of problems, you can then. Size solutions appropriately and consider what the level of effort would be potentially for solutions.

And so I think you guys, you and I are both big fans of starting small

And trialing it and seeing if it works. And if it does, scaling that, or if it doesn't work, trying something else. and that has guided me from creating custom software for people where they would come in and say, I want to build this thing.

And I would say why do we need to build the thing? What are we? What are we trying to address? Who's going to use the thing? Why will they use the thing? Why is it better than what exists right now? and then doing that across every different clinical area. And now I do that with startups as well.

And then I think the other thing that kind of draws on my history background, it's like, we would always look at historical events from a social, political and economic standpoint. And we would say, like, how do all of these things come together that cause this event to happen? And I think the same thing is in health care.

You look at the social you look at patients and you look at doctors and you look at, like, how is what you're proposing going to affect them. From a political standpoint, who needs to buy into it? Are you going to put something on nurses that they don't really have time to do or they don't want to do or they don't find value in it.

Thinking how to garner that political support for things, and then the economic side of things is like, what is it going to cost? What does it cost? Not just from a money side of things, but from a resources and the time side of things as well. I hope that my college professors would be like, it does make sense where she's ended up with her career, I apply some of those things that I learned at a younger age to more real life work versus the historical research that I used to do back when I was 22, and I thought I wanted to be an archaeologist.

Then with the writing that I do, that's looking at industry trends and industry discussions, I think that is just research chops have carried me as well.

Look at facts and data and make sense of it and what's a good source and what's not a good source and help people make better decisions based upon that knowledge.

Ty Hagler: Yeah. Imagine there's a rigor to studying history that just helps because from a, an innovation context is having to supplant a previous solution to a persistent problem.

And understanding that history. And it could be that we just discarded something 50 years ago that didn't fit the changing of times, but there was a political solution or a sociological solution that fit 50 years ago that could be cleaned and learned from as we're now trying to, not necessarily throw the baby out with the bathwater every time we're trying to innovate, but having that historical context sometimes of, what did work in a different context or pull from that.

Katie McMillan: Yeah, why is why are things the way that they are.

I remember when I was at Duke this giant survey went out that was like, to everybody that works there, it was like, what are your pain points here? And I don't know how I got a copy of all the results, but I did. And I remember reading through them, and the one that really stood out to me was like, how pediatric brain cancer patients would have to have their entire head shaved when they were having surgery.

And for little girls this was something that was like, particularly jarring to them and the person was like, this is a problem and it makes little kids sad. And it's makes my job harder to have to see them cry when they're bald. And I, that was 1 thing that jumped out at me and I walked my way down to the department and I was like, can I talk to this person?

And I was like, okay what do you think would make this better? And she was like, if we could just have hair ties, then we could pull the hair back and keep it out of the sterile field before we did the surgery. And then we wouldn't have to have all these sad little kids anymore. And I was like, can someone just order them some hair ties?

And so it was like such a simple solution. And getting to the reason why it was like hair can get in the sterile field. We don't want that to happen. And so but nobody had put that on the list of things that they needed to pay for, so.

Problem solved.

Ty Hagler: Yeah, but like you're it's like maximizing local optimizations, right? Of make sure the sterile field stays clean, so therefore you eliminate all risk of that at the expense of the patient well being. It's just, what a great story, I hadn't heard that story before, but there's like simple interventions like that, Taking a broader view.

That's what a great story.

Katie McMillan: Yeah, but then they had to keep everyone from coming to take the hair ties at work.

Ty Hagler: Oh,

Katie McMillan: They were like, people would come and be like, oh, I forgot my hair tie. And they'd be like, no, these are for the pediatric patients. But I was like, this is outside of my purview. You guys figure out how many you need to buy to keep both yours happy and your patients happy.

Ty Hagler: Right. Exactly. Or just have enough hair ties. So you take care of anybody that needs it in that environment.

Katie McMillan: Yes.

Ty Hagler: Seems like a solvable problem.

Katie McMillan: Absolutely. It's good when you have those sometimes.

Ty Hagler: Exactly. You've mentioned, some of your history background and then how that's informed some of your writing. You're such a strong writer and researcher. It seems like that early grand rounds presentation you gave and that research that you did, it seems like that was a launching point for you.

I'm just curious. You've been staying on, as I see some of the articles that you write.

Seems like you're staying like on top of a lot of the trends in a very fast moving space, but particularly wearable technology and biomarkers, I guess there's like what you're working on now, or maybe even what's capturing your attention these days, I'm just curious.

What are you, I don't know, excited about, you mentioned the aura ring. What's capturing your attention?

Katie McMillan: Yeah I I started back in September writing a piece a week for digital health insights, which is sponsored by chime, which is a big digital health technology organization and sometimes the clip can feel a little bit intimidating, but, it does allow me the space to be able to think about big problems or big things that are happening in the industry. And I really enjoy that. And so like for the biomarkers piece, even before I started writing for them, I did an initial piece that was like a market scan of who out there, what tech companies are building wearables specifically for women and what are they monitoring? And I realized that nobody had pulled together a list of that previously that had said there's the Oura ring and the Fitbit and the Apple watch and there's this new Evie ring and the Ava bracelet and all these different things. And nobody had put together a compendium of these are the top players. This is how much money they have. This is what the devices are. This is the data they collect. And then this is, the price points for all those things. So that was like phase 1 and then, my bigger question that started that research project was, like, what's next? What do we not know about women's health that we could potentially garner from passive data collection in a way that doesn't require people to go to a lab or order tests at home?

And what could we benefit from having 24/7 data collection from? And so then that, that piece sat in my brain, took up real estate in my brain for probably a good six months before I ended up writing the follow up piece, which was like, what does it take to build new digital biomarkers?

And that was when it came out on digital health insights and I was able to connect with a PhD level researcher at Duke who runs an entire lab that looks at that, and they work with industry to help figure out what different signals that they can collect and what this could look like in the future.

And that was really fun to figure out that one of the foremost researchers in this area is actually in my backyard.

An article on digital twins.

Ty Hagler: So wait, what is a digital twin?

Katie McMillan: Okay. Sure. Happy to talk about that. It's a super nascent concept that basically with the right amount of data, you could build a digital model of either a human or a hospital or entire system basically, and then you could run tests against it to see how they would perform under certain scenarios.

And when I wrote the piece, I remember there was an article on this show called the good doctor. Where it's very Hollywood it's this VR heart software where these 2 surgeons are standing in the operating room and this heart is beating in front of them. And then they're like, okay, what if we did this?

No, the patient dies. What if we did this? Okay they live and so they go through all these different scenarios and it's going to be this very complicated surgery and they try to reduce the amount of risk by having a, fully digital model of what that patient's heart anatomy would look like.

And that's basically the concept digital twins is that you could do something like that in the future. Could we reduce the amount of people that participate in clinical trials if we looked at past registrants and their data with their electronic health records to create, various, I don't want to say digital again, but digital patients that exist in a database that then you could try things against. And like I said, it's a pretty new concept, but I thought it was like, a really interesting one and in that article, I looked at what they're doing for it in oncology and cardiac care and chronic disease management.

So you can check it out if you're interested.

Ty Hagler: I will definitely check that out in part, you think about like simulation centers give some of that opportunity to test, say work flows within the hospital to iterate, apply some of the design thinking mindset to the unknown of we're doing something that's never done before, we're introducing change into the system and with digital twins, it sounds like you're able to test, but not in an environment where you're actually putting the patient's life at risk, but you're getting to a safer outcome by being able to do some of this fail fast, succeed sooner methodology, but in a digital context where you can learn from it and before you apply it in the real world.

Katie McMillan: Yeah, and I think some of the more rudimentary versions of this were happening during Covid as hospitals were trying to figure out what would happen if x number of patients got sick and they came into the hospital and what would that mean for their chain and all that kind of stuff? And it's just really cool. Other way that technology could help improve health care.

Ty Hagler: Yeah. It seems like you're venturing into the futurist kind of mode of starting to predict some of the future.

Have you ever considered that moniker?

Katie McMillan: It's interesting. I haven't really thought about it. I would say 50 percent of my writing, I focus on something happening at the intersection of digital health and women's health, and 50 percent are things that are happening more at a macro level. And some of that is in the futuristic space. Right now everyone's really interested in AI and depending on the day. I feel differently about it.

Tried to use a tool recently. And gave it a fictional article concept to see, Hey, write a blog post about this. This would be the introduction 5 bullet points. Give me the conclusion and include all the sources that you would pull the data from and it sent it back.

And I was like. It's not bad could use a little work, but then when I looked up each of the sources, they didn't exist, they were all hallucinations.

Ty Hagler: Yeah,

Katie McMillan: These people didn't the researchers didn't exist, I spent probably way too much time going down this rabbit hole, but I like, I looked at the journals, I looked at PubMed, I tried to find these people on their academic institutional websites, and none of it existed.

And so that made me feel like, really bummed out about

And whether or not consumers are ready to have access to this, because I was like, man, one sad college kid one day is going to like, be stressed at finals, and is going to turn some paper in, and they're going to get totally slammed because they're going to have cited all of these resources that like, were never published.

But then I read another piece this morning by Jen Goldsack on the World Economic Forum website, and it was about not about how AI can be used in actual patient care, but what can we do further back in the process in the public health standpoint, or to be disease averting, potentially using these AI models.

And that gave me A little bit more hope.

Some people are like all in on it. This is the future. And I'm cautiously optimistic, but keeping some thoughts around it.

Ty Hagler: Yeah. Social media did a lot of good for the world, but it wasn't, just a straightforward all roses situation. AI is going to have similar pros and cons. The one that I ran into recently was, if you do find a link from chat GPT it'll add a tracking cookie on the end. So if you then just, put the link into the browser, it says, Oh, this also came from chat GPT that you used this link.

I was like, Oh, they're acting more like a search engine than before of starting to track who's using these different links and whatnot. So it's, they're definitely changing how it's being used and how, how the applications of it, even if it's not hallucinating, it wants to know, wants you to know that, Hey, this link came from chat GPT.

Katie McMillan: That's interesting. Yeah, I did a piece one of my 1st pieces for digital health insights was around whether or not, generative AI will help or hinder patient education and the underlying point for that, essentially, after I did all the research to see, like, how if you Google something, if you say these are my symptoms, what do I have, you get an answer back.

It's probably going to be the most common thing that can come up. And you always have to make sure that the sources that you're pulling from, or that they're speaking to are reputable organizations. All those disclaimers this is not a doctor. You have to talk to a real doctor. and yes, it can be helpful to help people, maybe step off the ledge in the middle of the night when they think I might have a brain tumor, but you actually have a migraine. Ultimately see a doctor for some patient care.

Ty Hagler: Yes. So then as you're looking at the different sources that you're pulling from, for looking for innovation in digital health, what are some of the media sources you turn to as you're reviewing in articles or books that you see coming down the pipe?

Katie McMillan: Digest that I get in my inbox that I feel are pretty trustworthy. And then, those will sometimes spark ideas for concept pieces. I like Mobi Health News. I think they're pretty good reporting source. Rock Health as far as digital health things go. I started getting the Crunchbase emails and the CB Insights also is another one to get like some kind of macro level stuff. And then there's various like sub stack people that I like to follow. Like Chrissy Farr has been a writer. She used to work for CNBC and covered digital health and now she does her own things.

But yeah, I feel like I'm just, I follow also a lot of physicians on LinkedIn and so I see what they're talking about or what they're writing about and then sometimes I'll reach out to them for expert opinions for sources as well when I'm writing pieces.

Ty Hagler: Very good.

If somebody wants to reach out to you either for collaboration or input on where digital health is going for the future. What's the best way to get in touch with you?

Katie McMillan: Probably email katie (at) wellmadehealth.com or they can find me on LinkedIn. I'm under Katie Donahue McMillan, kept my maiden name for SEO purposes. And my website wellmadehealth.com.

Ty Hagler: Okay. Very good. Katie, thank you so much for joining me today. I've learned a lot. We took it in some directions that I've found delightful. So thank you so much for joining me.

Katie McMillan: I'm glad to be here. Thanks for inviting me.

Ty Hagler: Thank you.