616 - Inside Lyrebird at DHF26: The Rapid Growth of AI Tools in Australian Healthcare

Narrator: Hey, if you're interested in digital health but you're not sure where you fit, I made something that you should check out. I'm pretty excited about it too. It's the THT Digital Health Career Toolkit. It's a free five day email series designed to help you on your health tech career journey. There's lots of valuable insights in there, no matter what stage you're at. Head to TalkingHealth Tech Toolkit to get your free access.

Peter Birch: Welcome to the Talking Health Tech Podcast. My m name is Peter Birch. Today on the show I'm joined by Dr. Ray Boyaparte, Chief Clinical Officer at Lybird Health, and Kai Van Lee, Shout CEO Founder as well. G', day, Jens. How you going?

Kai Van Lee: Precision Beat. S. Uh, it's been a while.

Peter Birch: It's been a while, Kai. Ray. We've not had a chat on the podcast yet, so looking forward to diving in and learning a bit more about you and the work you do. Um, but uh, it's great to connect again, Kai. Maybe for those unfamiliar, um, uh, with Lybird, brief overview. But also more importantly, maybe since last time we caught up on the podcast, what's been happening and um, what the flavor is and then we can kind of go from there.

Kai Van Lee: So yeah, live at Health, you know, we effectively, uh, we build AI that helps clinicians treat patients better. Um, and we're on a mission to extend the human health span through that. And so you know, we started out as a kind of ambient medical scribe and then we built so much more on top of that, like a platform essentially. Um, and so yeah, I think we first spoke probably like two years ago or something now and um, you know, a lot's changed. Um, we've, you know, the start of this year, we, we made the Libert um scribe free for all, all, all best practice clinicians. So basically, you know, 80% or so of Australian GPs now have completely free access to a fully Australian integrated medical stripe, which has been so awesome. Um, and to see that impact, you know, we, we've had, you know, just over in the. Over 10,000 GPs in Australia have done a consultation through Lybird, you know, in May already. Right. Um, and so that's, that's, that's, that's growing really exponentially. But um, the ability to give that away to all Australian GPs has just been so awesome. And that's kind of, you know, one headline of what's been going on recently.

Peter Birch: Yeah, and look, we've got Ray here as well. Maybe before I jump in and learn about you and Your background like the, um, uh, building out the team like the. And bringing on Ray. Tell us a bit about that before we.

Kai Van Lee: Yeah, for sure. Like, I think it's, you know, especially like clinical AI is so different to any other piece of technology or product in healthcare.

Dr. Ray Boyaparte: Right.

Kai Van Lee: There's only so many possible blood pressure results or variants that can even be out, but there's literally infinite ways to write a note depending on the clinical setting, the patient context. And so much of this is about the technology.

Peter Birch: Right.

Kai Van Lee: But actually the most important part is embedding clinicians who use the tools and products every single day throughout your whole organization. So Ray's joined as chief clinical officers, uh, obviously on the leadership team. And he's still, you know, I let him introduce himself, but I still a practicing physician to this day and I think that's, it's been so important for us to embed that in the entire company.

Peter Birch: Yeah, excellent. Well, tell us a bit more about yourself, Brad.

Dr. Ray Boyaparte: Yeah, so, uh, it's a pleasure to meet you, Vade. Um, and look, I've come from a bit of a different background. So I'm a practicing clinician, a gastroenterologist at Monash Health and work at the university and do teaching, uh, research still. But I actually uh, have a founder myself. So when I was in med school I founded a company called Med Entry. This company helps people get into medicine and we've sort of become the most popular program in the uk, Ireland, Australia, New Zealand. And so I kind of know what it's like to build something from nothing, which is kind of what this is all about. Uh, but I've got the other aspect which is obviously practicing clinician, gastroenterologist, treating people mainly with chronic diseases, so inflammatory bowel diseases, ulcerative colitis and Crohn's disease. And uh, yeah, and also in academics. I did my doctorate up in Edinburgh researching the sort of ways in which IBD comes about. Um, so a few different sort of angles there. And um, you know, really it was just completely serendipitous. When I met Kai, I wasn't looking for a job in this space at all actually. So we were on a panel together actually at an entrepreneurship thing at my kids school. It happened to be the school that um, Kai went to. Um, and so as an entrepreneur himself, obviously they invited Kai and myself as a parent and we were just talking, ah, at the panel and then we sort of connected afterwards and um, I just sort of thought, look, this is a pretty compelling founder and I sort of can see that. I think in um, you know, the founder has to be a particular type of individual really. I saw in Kai, uh, someone who is really ambitious and really wants to um, think with. And thinks with a lot of clarity, but also has that kind of real curiosity, um, and willingness, willingness to listen to the experts in the space. And so that's kind of where this started and it's sort of gone from there.

Peter Birch: So.

Kai Van Lee: Yeah.

Peter Birch: Oh that's. I mean that's a unique story. The whole school panel. Yeah, stay in school kids. That's. The opportunities are endless. And so tell us a bit more about the chief clinical role that you're in, like the responsibilities there and um, how that influences where Liebert goes.

Dr. Ray Boyaparte: Yeah, I mean I see this space evolving so rapidly. Like I'm seeing it as a user. Right. So in practice myself, I'm seeing all my colleagues and myself use these tools and take it up with incredible alacrity. Like I've not seen this in any kind of tool ever in medicine. Right. And AI is so powerful. That's why it's been taken up so much. It's sort of the magic of it. But I've also felt this really deep um, unease about it as well. Like this idea that as we move from clinical note into some more interesting areas, clinical decision support and other areas that really impact on patient care, we sort of see that this can become, even though it's so powerful, it actually becomes quite risky as well if you don't do it right. And so, you know, the idea of coming on as chief clinical Officer here is that I felt like the space is like just moving so fast. Um, people are using it at such rates and the timing is so critical. Like right now is when we start to determine how this is going to be used into the future. And I thought I've got to be part of it. I was kind of almost semi retired, right Kai. So I was a pulling amount of retirement basically. So Kai's pulled me out and um, I've really enjoyed it so far. And like one of the things I've really enjoyed is that Kai and the other executive team have kind of brought me in and really just listened to what I've said and like my perspective, which has been fascinating. So my role really is to bring that kind of real clinical rigor into everything. Um, in Liebird or the whole DNA of the organization.

Peter Birch: Yeah, you touched on this point that you know, like a, um, uh, these AI, if I was being general in terms of AI tools in maybe in general practice, but anywhere in healthcare really, um, Historically, there's not really been this huge willingness from clinicians and there was a stigma of law. Clinicians don't adopt technology. And what I've learned through the process, they just don't adopt crap technology. Uh, it's like if a tool is useful, um, and if it's going to have a meaningful impact, which means that they can, uh, even with some gaps in terms of not fully understanding how it works, and all these things that in the past we said, well, you know, unless, if I was a clinician, unless I fully understood how it works and I'm not gonna use it, when in reality it's probably just another, like I'm switching out one problem for another. But for these types of tools, I hear of clinicians a lot who are using, uh, them and maybe there's like a whole concept shadow it and just because it's impactful for their role. So I guess, you know, does your role factor in a lot of that? Because there's a different way to kind of frame it off as like, well, how do we keep up with the demand but still maintain clinical safety? Am I going the right direction?

Dr. Ray Boyaparte: Yeah, oh, absolutely, Pete. I think you hit on something important there. So, like, I live through the whole generation which where, like, digital tools were basically hoisted, foisted upon us, right? Like, you know, use this EMR in this way, use this referral system, and it all sounds great. And we were promised the world, right when we went to cave, like, hey, this is going to like, make your life so much easier. And it just did it. It made it more complicated. It made it so that actually we were not speaking with our patients. We were focused on, like, documenting this and that in the right spot. And so I think that's. To your point about, you know, doctors, uh, aren't Luddites, that we. We actually will use the technology if it works. And scribes are a perfect example of that. But, you know, this is the conversations we're having internally. You know, we're sort of these kind of tools and AI in general scaling so quickly. And, um, part of the keynote that we presented was this idea that if it's doing it in this way, like, what is that the right way to do it is just to scale infinitely and just go to wherever, you know, organizations and individuals want, or do we kind of be much more deliberate about it. And I think my approach and Kai's approach and our approach is really to do it in a deliberate way and not just go anywhere and everywhere where the demand, uh, is there, but, like, make sure that we do it in the right way. Because in the end these tools are so powerful, they have significant clinical risk if you don't do in the right way.

Peter Birch: Yeah, there's something in this whole um, I look at outside of healthcare AI generally Kai, it's like um, we seem to be, and I'm generalizing, but it's kind of like oh, there's more and more ways we can replace humans and displace kind of society and all these kind of, you know, big existential questions and as a collective industry is kind of like how fast can we like disrupt everything? And I sometimes think that point is kind of like the just because he can bit like because in the past we've been like uh, constrained by what the technological capability of, is of tools. But if the capability was wide, then does that mean your scope needs to be as wide in terms of let's just ditch all of the humans, let's just draw the capability and I don't know, we all just hang around and I don't know what we do. But um, does that form part of your thinking as well in terms of uh, as capability expands, how do you almost then narrow down what you do and just as importantly what you don't do 100%.

Kai Van Lee: So rake use is a great example. But as a doctor, as you get more experienced, your scope narrows, not increases. Right. You specialize more and more and more, you know, and I think that's so important about you know, kind of like these tools is, you know, yes, as they get better, there's a breadth that is maybe like enabled by that. But as you go deeper into the stack of care, you know, you're doing referral letters, you're doing our ah, discharge summaries, you're doing all these like billing codes that the depth that's necessitated by that is just so much greater than any increase in kind of capability might enable. So yeah, we, we really believe that like you know, extreme depth is, is just so important. And I think like, you know, we, we've had a lot of examples where it's you know, for example like uh, with indigenous medicine, the nuances of the way it's practiced, it's actually not about how good the model is, it's about your context on that type of medicine and that type of practice. You know, it's not about like, you know, the reason you specialize as a doctor is not because, you know, if you're smart you can go, you know, really broad and you know, if you're not as smart you, you can't, it's, it's, it's so much about the nuances and variances that happen that actually are uh, separate to the intelligence of the model.

Narrator: Yeah, real quick. I get asked a version of this question all the time and it's, I want to work in digital health, but I don't know where to start. And I think the challenge a lot of people have is it's not, they're uh, short on the interest, they get involved, they don't have the clarity on what to do next. Like they've been watching this space for a while or listening to podcasts like

Peter Birch: this one that then you don't know

Narrator: what your next move is. So in an effort to be helpful, I put together something that I call the Digital Health Career Toolkit. It's a free five step email series where we cover mindset, your existing strengths, where the real opportunities are in this industry, how to upskill and how to build a network that actually opens doors for you. So if you've been looking around this space for a while and want to start getting involved in it, head to TalkingHealth Tech Toolkit and get access. It's free and it's easy because it just lands in your inbox each day and people think it's useful. So I hope you like it. That link's in the show notes of this episode too. Talking Health Tech Toolkit.

Peter Birch: Yeah. No, that makes sense. The um, Ray, the like, one thing that obviously you would know is that um, the, the nature of healthcare is like kind of to what Kai was saying that it's your view as well. Like, you know, the, as you get more and more experienced, you get more focused on, on what you do. But the, the role of a clinician, particularly in primary care, but actually anywhere I imagine you're consuming a lot of information. Like there's a lot of inputs that you've got. It's not a very straightforward like I'm going to get this input and then do this particular thing. There's just a lot of like cognitive, like we talk about the cognitive load on clinicians and all these kind of things. Just a little mess. I think in healthcare. And, and I think, you know, one of the, the roles that the right technology plays is put things in boxes and makes it all kind of nice and all that kind of stuff. But is that like, is that what we're trying to do to healthcare is like because we need to be able to put measures on things and code it and all this kind of stuff, but in the End humans are like squishy bags of fluid that, like unpredictable. So how do we find that in between?

Dr. Ray Boyaparte: Yeah, I mean, it's a good point. I, I talk to my medical students and almost compare medicine to carpentry. Um, because medicine is not about knowing everything, it's about experiencing a lot. So just like carpentry, you're going in there and you're experiencing and that's how you learn. It's like an apprenticeship. Right. Same thing with medicine. You need to experience a lot of things. You need to have gone into a lot of clinical contexts in order to gain that clinical judgment. And so that's what you're talking about there. Which is not like it's really hard to get a digital tool to do that.

Narrator: Right.

Dr. Ray Boyaparte: Because digital tools at the moment at least, have been pigeonholing data into different areas and then asking you to kind of know all of that data. So what I think really a good digital tool would do, AI tool will do, is to help the clinician deal with that cognitive load, know where that data is and surface it at the right time. And that's a challenge. But I think it's something we're up for at Liebert. I think that's something that really we want to pursue. Um, because this serious problems that encounter that happen when you don't have that information in front of you. Yeah, because the EMR is terrible or because actually the EMR is fine, but there's five other EMRs you have to deal with. Right. Um, and they're siloed away. You don't have that data. So, you know, clinical medicine is about forming that judgment. It's about forming that with. Via experience. But, and so that's a really good part of medicine. But, um, but in order for a digital tool to work really well, it's all about presenting the doctor with that information at the right time so that they can make the best judgment possible.

Peter Birch: Yeah. And I'm thinking about how, uh, uh, more and more clinicians are using AI powered tools and there's, you know, more and more pieces of software are either everything I. Or it's like partly influenced by it, but it's all kind of, um, baked into more and more that we do. What stage of this whole adoption curve do you think we are? I mean, it's always a sliding scale, but at what, at what point do we just assume that it's all just encompassing in everything that we do? Are we still at this early adopter phase? Because I still go to my family GP clinic, who doesn't use an AI scribe and it's a fantastic gp. And, um, you know, I don't think he has any plans to do that. And there's plenty of other examples too. But, uh, is it. Are we going to get to a point where, you know, it's more clinically safe to use some of these tools than to not. So it's kind of because they're. So at what point are we at with this whole using tools that. That could improve the efficiency or effectiveness of a clinician?

Kai Van Lee: I think we're still quite early. Right. Like, when you look at it, I think there's been some studies published or like numbers that, you know, half of GPs in Australia are using a stripe. But that's not at all. I don't think that's actually representative. I think it's really.

Dr. Ray Boyaparte: It's.

Kai Van Lee: Maybe they've dabbled in one.

Dr. Ray Boyaparte: Right.

Kai Van Lee: But I think it's. The minority is still using a stripe. And the reality is of the tools is that, well, probably the majority have tried one, but I haven't stuck with it. And so something about the technology, not the doctors, in my opinion, because they're not dying to keep doing paperwork. Like, I guarantee, basically none of them are. Right. And so I think there's still a long, like, so far to go with the technology. For it to just be really like, it should just feel like magic, you know, for the clinician. I. I think there's still a lot of work to be done. So I think we're early in the adoption curve and early in the depth curve as well. Um, I think what you said is really interesting.

Dr. Ray Boyaparte: Right.

Kai Van Lee: Like, what happens when it becomes actually, you're definitively proven it's higher risk to not have a tool like this, like a seatbelt. And I think that's going to come. You know, I think that's really in the future ahead.

Dr. Ray Boyaparte: Yeah. And I've got sort of something to add to that. I think that really the early adopters are the ones that innovate and do use these technologies early. But actually, I think there's this huge group of people that aren't sure. They're just not sure in terms of privacy, in terms of trust. And I think, um, a big part of why I've come into liebird and why I think, um, it's really important to have these clinicians at the executive level is really to make sure that everything in the company and the industry is done in a way where we earn that trust. Yes, we earn it from the clinicians, but we earn it from the patients, we earn it from government and enterprise. Um, and you know we work really hard at libird to make sure that that's a big focus of ours. Um, you know, all of our data for example stored and you know, dealt with in Australia. I think a lot of people worry about that. A lot of people worry about, we know whether our data is used to train models. They aren't, you know, and so this kind of thing is on us to make sure as an industry, we make sure that people are aware of this and that they're comfortable with the technology as they use it. Yeah, yeah.

Peter Birch: And so um, Kai, what next? Like what's the, where's the ball going in terms of like as you're building, as you mentioned, you know that start with the scribe and kind of build out capability there. What's the, what's the trajectory and scale of this thing?

Kai Van Lee: Yeah, so it was like a couple of different directions. I think like the first thing is like now, you know, as Ray mentioned, we're really clear on where we're serving, right? It's like, you know, we're not trying to do vets, right. We're not trying to go, you know, really serve any conditions in the, in America. You know, we're focused on serving, you know, Australia and maybe it's like the UK as well, um, doctors just better than anyone else. Right. And so for us that's like, I think like so the reason I started Liebird was because I really needed to see a doctor and I couldn't. And I Learned that my GP spent 3 hours a day on his paperwork. Right. I always just keep going back to that mission and the reason of starting. And so when I, when you sit back in a consult with a doctor today, even the one that's using a strive and a power user, this is spending like an hour or more on paperwork. So for us like we, our mission to be since it starts, we can get that to like as close to zero as you can. Right? So this, so for us like over the next two years we're going to turn that number to zero but have the quality and safety of that clinical documentation. Not just on average but actually on the majority of that, you know, overwhelming but you know, 90, um, 9 percentile of that, that actually be safer and higher quality too. Um, and then I think we go to like how do we use this to increase the human health span? You know that's like it sounded like a bit like Brian Johnson at the start. Like, but it was really like that's kind of like, okay, that probably feels like if Liberty achieves its mission, like that's probably the uppermost outcome, right?

Peter Birch: Yeah.

Kai Van Lee: And so for us, I think it's this amazing opportunity now where for us to actually do that. Right. And, and actually improve patient outcomes. And you know, when you look at like healthcare, healthcare is like grossly under consumed on a population level. People like are getting tested at uh, like the latest time possible. Like if you, you know, it's just so, humans are just so under optimized. Right. And it's like really bad. And so you go, wow, there's so much low hanging fruit we can do to just improve outcomes here. And it's all, majority of it is an information and a timing problem, right?

Dr. Ray Boyaparte: Yeah, yeah, yeah. I mean I, I feel this as well. Like you know, we talk about the, the risks of AI and, and, and I feel that like we need to do way, but there's so many risks that are just happening at the moment, you know, because things are not working well. You know, the, the, the doctor who's getting burnt out is a risk to the system. Um, they're a risk to their patient, but it's, they're a risk because they might leave the profession and that's going to not be good for society. I feel that, you know, I do triage for example at um, Monash Health. Um, and when I did triage, uh, I was, you know, you triage patients based on what the information you get presented, uh, from the gp. But if the GP is presenting information that's um, very in one consultation, but actually doesn't have the information from the five other GPs that have been seen, the patient seen, you can get it wrong. And when you're triaging, you can triage it wrong. And that can have consequences like real life consequences. That's one example of just so many in the current healthcare system that is right to get done if it's done really with intention, uh, makes a lot of sense.

Peter Birch: Gents, it's been great to sit down, have a chat, see where things are at with Live Bird and the direction it's going to. So the details will be in the show notes this episode. People can check out, learn more. Always a pleasure. Thanks so much.

Dr. Ray Boyaparte: Thanks so much. Thank you very much. Cheers.

Speaker E: Uh, we hope you enjoyed this episode of Talking Health Tech. Make sure you like and subscribe and share this episode with someone who might find it valuable. For more information and resources about healthcare innovation, visit talkinghealthtech.com.

Narrator: Thanks again for tuning in. If you enjoyed this episode, the best thing you can do is subscribe, drop a review and send it to someone that'll get some value. Check out talkinghealthtech.com for more information.