Autoscribe: How A.I Could Change Your Practice

‍About the Event

Join us in this insightful episode as we explore how AI is transforming healthcare, reducing charting time, and improving the overall patient-provider experience.

Physician and founder, Noah Crampton, will share his expertise and insights on how to navigate this exciting technology as well as speak about his own application, Autoscribe, so don't miss out!

‍

Objectives and Discussions

• How does Autoscribe work?
• What are doctors saying about A.I?
• What is next in A.I
• Being a leader
• Future of healthcare in Canada

‍

More Resources:

LinkedIn: https://cherryhealth.co/linkedin

Facebook: https://www.facebook.com/cherryhealthinc/

Instagram: @cherry.health

Twitter: @cherryhealthinc

‍

Have questions? We want to hear them!

‍

Transcript

Speaker Identification:

[Host]: Alitta Tait

[Speaker]: Noah Crampton

‍

Allita: Hello and welcome to this episode of Cherry Live. I'm Alita and for those of you tuning in for the first time, Cherry Live is brought to you by Cherry Health, Canada's medical network, where healthcare professionals go to find and post jobs and to connect. Cherry Live, of course, is our podcast series where we connect with innovators, entrepreneurs, and professionals of all sorts who are working to move Canada's healthcare system forward. I am very excited today for this episode where we're joined by Dr. Noah Crampton. Noah is a family doctor and researcher at... Toronto Western Family Health and the University of Toronto, and is the co-founder and CEO of Mutuo Health Solutions, which is a startup leading the commercialization of Canada's leading AI ambient scribe for clinicians called Autoscribe. Thank you so much for joining us today.

‍

Noah: Thank you so much, Alina. It's my pleasure.

‍

Allita: So let's dig in. Why don't you tell us a little bit about what brought you to medicine in the first place.

‍

Noah: Absolutely. So, you know, I've always been a sort of like nerd and thinking what's the what's the next thing coming down? How can I be part of that excitement, right? so, you know, I was coming up in my training years really interested about how Medicine was going to change over my lifetime, you know, a 50-year career as a doctor is a long a long time So a lot can happen during that time And I was always keen on thinking about how can I provide the best care to my patients using the new tools that were becoming available. And obviously in the... So I wanted to become a doctor because I loved connecting with humans and providing my knowledge base where I can filter that through clinical judgment for the particular case that the patient is presenting with and providing that whole person care, really getting to know them and make sure they... they feel valued in the care that I'm providing. I think there's a dynamic there that's really special. So that's why I went into medicine to begin with. And then, you know, I always viewed my career as, okay, partly will be patient care directly for sure, but also partly will be about, you know, trying to think about what those innovative tools will be and how we can deploy them as an early adopter kind of person that I am to provide that extra value. So. So that's what happened in medical school, but then I obviously started to focus on the particular issue that I ended up being passionate about.

‍

Allita: Yeah, so tell us a little bit about that. I mean, everyone's really curious about AI and where this is going next, but first tell us about, you know, how and who joined the team and how you guys decided to start Mutuo.

‍

Noah: Yeah, so basically, there was a, immediately when I got into medical school, there was a disconnect between what my life was as a consumer with the like, you know, amazing tools that were starting to come out with smartphones and very amazing powered sort of suggestive searches, etc. versus what was, what I was experiencing, you know, as part of the digital tools I was using in healthcare. They were, they were very... poorly designed, they had very minimal functionality. And like, you know, in my head, I'm like, you know, healthcare is the most important thing. You can't do anything else in your life if you don't have your health. So you'd think that as a society, we would invest the most we can to ensure that we're providing the highest performing tools for that objective of patient care and health. So when I realized that, I thought, okay, there's obviously a reason why it's... structured this way and part of that, you know, was with the poorly designed systems and the initiative work related to all the complex patient care we were doing was also skyrocketing. So on the one hand, the volume of work was going up, but also the tools used to enable that work was very onerous and not friendly to the user. So it just led to the extreme frustration, even at my early stages as a medical student, that I then decided I'm going to orient my career to solving this initiative burden through novel technological tools. So I went through residency, I became a family doctor, and then after that I decided to pursue a Masters of Health Informatics, and that's where my research hat came in. And it was in that research where I decided to really focus on leveraging the power that Toronto is a world leading centre for development of the new AI algorithms that are out there. Jeffrey Hinton is one of the world's most forward thinkers in this space, developed the neural networks that are powering all these tools. And he created with some other institutions, what's called the Vector Institute. And so I was based at a hospital here in Toronto called St. Michael's, doing my masters, and I connected with the leading computer scientists at the Department of Computer Science at the University of Toronto and the Vector Institute. And that's where I met my co-founder. So she was the computer science, I'm the clinician, sort of, you know, systems guy, she's the programmer. And it was between the two of us, we conceived of, okay, we can solve this. We have the expertise to solve clinical documentation using artificial intelligence.

‍

Allita: Yeah, well, that's pretty exciting. And so you and your co-founder embark on this journey and start digging into building a company. Tell us how that was, you know, as a practicing physician and also trying to build a tech startup. How has that been, you know, digging into that whole journey?

‍

Noah: Yeah, so at first this was a research project, right? So we just wanted to prove that we could do this, right? We didn't want to embark on that journey if we weren't able to execute on the actual core task. And thankfully we did. So we published a couple of papers that was at 2018, 2019. And so then, our mentors, both my co-founder, her name is Serena Gervley, her mentor, my mentor, Muhammad Mondani, we're all saying, let's see if we can make a real system impact. So we decided to spin off a startup from that academic research where that startup, Mutual Health, would have exclusive license to the core intellectual property. And then I was encouraged, because obviously by nature, I wasn't necessarily an entrepreneur, but I'm a fast learner, and I'm not afraid of trying new sort of tat, ways of executing on an objective and entrepreneurship to spread an important healthcare innovation is definitely a major way of doing that. So throughout the pandemic, I was obviously practicing clinically, but we really got this AI tool very reliably functional as a software application, did all the privacy, security and legal work. And then as of, you know, basically a year ago, we went into the market with our tool. and we launched and that's when the real commercialization took place, marketing, hustling between different healthcare organizations and we're excited, yeah, we're really growing at a nice clip now. People are definitely seeing the value of Autoscribe.

‍

Allita: Thank you. Well, tell us a little bit about that. Tell us, you know, in your opinion, where does this technology really support physicians? How does it help them practice more effectively?

‍

Noah: Yeah, I mean there's really different core benefits that are meaningful to different physician types. So for instance, if you have the microphone on in the exam room, you're talking to your patient, on the screen you'll see AutoScribe in real time generating the note for you as the doctor. Now that means multiple layers of benefits. So if you're for instance, let's get out of the way, like a lot of doctors, they just wanna. increase their financial compensation, you'll definitely have more efficient visits and you won't have to do your notes after the encounter. Very minimal editing work required, you save so much time, you can then see more patients or roster more patients and you'll be correspondingly, there's a return on investment of 4X in that regard. Then, you know, but sometimes it's not just about money, right? It's about, A, it could be connecting with your patients, you don't have to stare at a computer screen throughout the encounter. You can really notice their nonverbal cues, their body language. That's really important for clinical assessments. There's a, you know, you can get really deeper connections with your patients, which is really rewarding as a clinician. As I said at the beginning, we went into medicine to connect with patients and that really fosters that aspect. And you know, what I really was motivated by and what many are is that liberated time that comes from using autoscribe, first of all, makes better notes than I can do.

‍

Allita: Yeah.

‍

Noah: AI is sometimes smarter than me, which is kind of jarring, but it's fun, you know? But beyond that, you know, there's a burnout element to how much administrative work we're having. It's a well-known problem in healthcare. The volume of patient complexity and the amount of tasks that require effort in the electronic medical record is enormous. So this relieves like a huge component of that administrative burden. And so I can, instead of seeing more patients, I could adjust my wellness by seeing my family and friends and sort of reset myself so that I'm to be the best clinician I can for the next day, right? I have more time to do that. So it really brings back the joy of medicine. That's our tagline.

‍

Allita: Yeah, yeah. No, I like that. I mean, we are constantly having conversations with physicians and clinicians through CHERI. And one of those number one things is just getting some time back. And the burden of that administrative task, the burden of trying to multitask at the same time, of taking the notes, remembering what to jot down, and then being able to treat a patient. So it makes a lot of sense. So tell us how it works. Tell us how Autoscribe actually works.

‍

Noah: Yeah, so you can use it as a standalone application. You sign up, you'll get a login, you'll go to your browser, you'll just click and launch, and it'll start recording. It'll access your computer's microphone, or you can use it on your smartphone. And it just let it run in the background as you're talking to your patient, because it has the intelligence to, yes, transcribe the actual dialogue, but also in real time generate. the note representative of that dialogue, which we would normally as clinicians have to write from scratch ourselves. So you see that happening in real time. And by the end of the visit, it will structure it in the format of a note template that you as a clinician prefer. You can set that in advance, depending on which tier of service you sign up for. And so, basically on average, because the editing is so minimal, you're saving like two to three hours of charting time per week, which is just a huge amount of time saved. And so, you know, that's sort of like the standard way. Then there's, we have a couple of electronic medical record integrations, EMR integrations, two of Canada's leading EMRs for outpatient care, which is TELUS, PSSuite, and Oscar EMR. So if you have that version, what's nice about that is you just launch it from inside the patient's chart in your... in your EMR and it'll bring that up, that patient's context inside Autoscribe so it already knows which patient it is and you can refer back to it if you want to go back to that Autoscribe generated note and then it'll post directly into the progress note section of that EMR so you don't have to copy paste from this, like you would do from the standalone application. But yeah, I mean it's super easy to learn, super easy to sign up.

‍

Allita: So, I mean, that's great feedback. Tell us what are great context, what are doctors saying about using AI? I mean, there's within the AI world, there's plenty of pushback of, ooh, this is scary. Or, you know, I think a lot of physicians see the words AI and think, oh my gosh, I don't know, that might not be for me. But what kind of feedback are you getting from physicians who are using the tool? Or what are some of those commonly asked concerns that they have and how are those addressed?

‍

Noah: Yeah, and you know, like, that's the benefit of, like, we started this a few years ago now, right, before the chat GPT revolution, everybody should know, everybody probably knows what chat GPT is now, sort of like this out of nowhere, very intelligent, AI that understands language. But you know, there's concerns about like what, like, how did it get the data to understand all that if data keep, you know, continuously being fed into it, and it's going to some large corporation and monetizing that data, you know, in healthcare, we're very, very careful about how we leverage patient level data, right? There's very key regulations you have to adhere to. So the benefit of us starting early is that we have a long record of already thinking about all these issues. So for instance, you know, our data that we've trained our algorithms are very much in-house generated so that there's no concern about that leakage of data to third parties that you didn't consent to. Speaking of consent, we have a very robust, easy to implement patient consent process where the patient is informed about what Autoscribe is doing, the audio coring process associated with it. We have third-party assessments that have done very deep, rigorous assessments of our privacy and security compliance. So we want to minimize all the risks because the benefits of AI are enormous. Like this is just the beginning of trying to capture some of the benefits, which is these administrative task automations. But AI, because it really does understand language at a deep level, it could start doing other tasks. The one thing that we're really focused on as a company right now is minimizing what's called bias and hallucinations. So there's these two major issues that can occur with AI. And again, we've been thinking about this from the beginning, which is which data sets you use to train your algorithms can have biases baked into them. So they would over represent certain patient populations than others. So how do you ensure proper represented in this so that the outputs are always equitable? And the hallucinations is basically sometimes these large language models, which is the type of algorithm that power these algorithms, randomly, like 1 to 2% of the time, will just generate something completely random that has nothing related to what it was prompted to do. So there are ways to restrict that. You can't eliminate them at this time, but it's actually ensuring like you are working through a very rigorous step-by-step process that minimizes the hallucinations as much as possible. But that's all to say like you should never treat AI scribe generated notes as the final note. As a clinician it's your job professionally to ensure that clinical documentation is appropriate. Once in a while you may see hallucination and we ensure in our application a very trustful process of ensuring the editing occurs in a way that the clinician feels good about.

‍

Alitta:Yeah.

‍

Noah: And so for me as a founder with a growing company, those stories is what's going to inspire me to keep addressing and keep going further and growing and hopefully beyond, not just Canada, go internationally because humans are humans, patients are patients all around the world. That impact could be global, then that would be amazing. And what I'm really focused on now is there's a lot of legacy sort of institutions and how the systems are, how different organizations are built within government health systems. So thinking about how can a novel technology tool be adopted into those legacy systems, right? Because otherwise you're just, you're not realistic about what the state of healthcare is. So that's what's really my big focus for now, because I know I have the impactful tool, I just need to get it into real world practice.

‍

Alitta: Yeah, I mean, and we've seen that as one of the biggest challenges in this system is how do you get it in front of doctors? How do you encourage them to try it on for size and see if it impacts them when there's so much going on and so many new technologies and so much, you know, they're already overwhelmed to begin with. So, you know, marketing to them is difficult, getting them to try something new is difficult. All those steps along the journey can be really challenging. Tell us within your organization itself, I mean, you talk about physician burnout and wanting to try and impact them. How do you bring that culture back into your team or into your workplace? And what does that look like for the team that's actually a part of Mutual?

‍

Noah: Yeah, so those stories are super impactful for me, but they're also like every time our team meets. So I have my technical meetings with the people that are actually building the application software. And then I have my business team, and we're always meeting regularly. And everyone gets juiced up. Every week we have new, very positive testimonials that come in and that's just ads fired us to either build that new feature or reach a new segment of healthcare providers that we didn't have reached to before. So the people that we've chosen that are part of our team are very much motivated by that core inspiration. There's many other companies out there that you can make a lot. money quickly. This is this is something that's a long journey. Like it could we could be a hopefully in the future, you know, very economically beneficial company. But but the team here is motivated by providing the relief to physicians to provide the best care. And, and yeah, so we're a growing team of eight. And, you know, if you're interested in joining us on our journey, please reach out to me as well.

‍

Alitta: Absolutely, yeah. And you can find them on Cherry or on their website at Mutual or Autoscribe as well. That's too funny. No, it's, I mean, because I think we hear a lot too about, and I've seen this a lot in my experience of we have this vision for this great impact. But sometimes it's really interesting that disconnect between this vision for this great impact and then, you know, and being able to help people out there. And then the way that sometimes manifests internally in the team is like, go faster, go harder. It's okay, like, you know, the burnout within their own teams in a startup, in the startup world is pretty high. So it's kind of a funny hypocrisy that ends up playing out. But no, that's great to hear that you guys really have been brought together by that mutual vision. And hey, most startups aren't exactly the place you go to make cash quick. That's okay. And so, I mean, for you...

‍

Noah: Yeah.

‍

Alitta: We've talked about AI building companies. Where do you see healthcare in Canada kind of evolving over the next few years? What are you hoping to see so that we can, yes, with through AI, but ultimately, as a physician who's practicing and seeing a lot of these challenges as well, where are you hoping we're heading in Canada?

‍

Noah: Yeah, you know, Canada is actually one of the most in terms of like those legacy sort of system setups and legacy technology tools, like really, really impeding innovation, unfortunately, like that is we are known for that as a country in health care. But we have some core features to our system, which enable which is very, very beneficial. official for a future state of AI. So for instance, you know, every province links their patients to an identifier. They each get a, you know, government-supported healthcare with a healthcare card number that represents the free service that they're going to get whenever they go to a doctor or hospital or whatnot. That actually...enables the tracking of that patient throughout the entire journey wherever they go for healthcare. That data exists where whether they go to one hospital or another hospital, they go up north in a province or south in a province, the data tracking that patient's journey through their healthcare system is there. And all that needs to happen is more predictive algorithms to say, you know what, this patient

‍

Alitta: Oh.

‍

Noah: You know, going to a walk-in clinic a little bit more than they were before, we should check, the family doctor should check in with them to see if there's something that can be actionable before they end up in hospital. You know, there's going to be all these predictive, proactive opportunities because of how Canadian healthcare is structured. When you think of other systems, you know, many, many countries do not have that one identifier following their patients. So, you know, again, Toronto's a world-leading

‍

Alitta: Mm-hmm.

‍

Noah: ...center for AI development. We just got to get that world leading technology into the existing world of healthcare and then like we can really leapfrog from our you know bottom of the pack status as a country in performance for healthcare and leapfrog to the top. I truly believe that.

‍

Alitta: Love that. Well, I think being mindful of time has been a fantastic conversation. Anything else you want to add to the cherry audience that you think is important for them to know about where you all are going?

‍

Noah:Yeah, so, you know, with my researcher hat on, anybody just wants to discuss AI in general, in healthcare, you know, many as various different applications, please reach out to me, happy to just discuss at that level. I would love to demo Autoscribe for you. I think once you see the demo, you will have your mind blown, just like many of our other clients. So you can find us at our website, MutualHealth.com, or you can email me, noah.crampton@MutualHealth.com. And it's gonna be, I promise you, it's an exciting journey if you wanna join us.

‍

Alitta: Amazing. And we'll make sure that those, that contact information is all in the, in the show notes as well, so that anyone can get ahold of you. But you know what, thank you for joining us today. And you know, I think, yeah, we're just at the tip of the iceberg here in terms of what's happening next. So very excited to follow along with your journey and continue to see where this AI takes healthcare in Canada. So thanks again for joining us today.

‍

Noah: Thank you so much for having me.