October 12: Today on the Conference channel, it’s an Interview in Action with David Rhew, MD, Global CMO & VP of Healthcare (WCB) at Microsoft. What's possible with AI now that wasn't possible three years ago? What does the rapid advent of AI technology entail for healthcare professionals? What novel possibilities are emerging with the application of deep learning and neural networks? How is the rudimentary task of note-taking being transformed by AI, and to what extent are clinicians adopting these automations? We also probe the seismic shift in data harmonization and centralization and its implications for clinical research. Lastly, we investigate the reservations clinicians may harbor regarding AI adoption and discuss how these technologies could be deemed a tool to enhance their capabilities.
Alex’s Lemonade Stand: Foundation for Childhood Cancer Donate
This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.
Welcome to This Week Health Conference. My name is Bill Russell. I'm a former CIO for a 16 hospital system and creator of This Week Health, a set of channels and events dedicated to leveraging the power of community to propel healthcare forward. Today we have an interview in action from the Fall Conferences on the West Coast.
Here we go.
All right, here we are at the health conference. We're doing an interview in action and someone who I've wanted to interview for a long time, Dr. David Rue, the global chief medical officer for Microsoft. Thank you. Pleasure to be here. I'm looking forward to the discussion. We're talking there's, there's so many things going on in healthcare right now.
The pace, so I'm a former CIO for a 16 hospital system, but the pace at which we're moving right now is really unbelievable. I was just talking to Doug King from Northwestern Medicine and we were sort of comparing notes of. Just how fast AI is moving in healthcare. It seems to be at almost an unprecedented pace.
I
agree. We're definitely seeing a lot of excitement. There's huge opportunity. But there's also a lot of problems that we have in healthcare. So it's a great opportunity for us to apply the AI and try to solve those problems.
What's possible? I mean, it's, it's interesting because we've had AI for a long time, right?
We've been doing OCR and NLP, and we've been doing all sorts of algorithm based clinical decision support for a long time. What's possible now that wasn't possible, say, three years ago?
So one of the real exciting things was when we started thinking about, in deep learning, the neural networks the ability to anticipate that next word or that next concept.
What we had realized was that when we had very large networks of information, like the internet, We could uncover knowledge. So that was where we started realizing generative AI has a huge opportunity. But what made it really interesting was when we put that natural language interface in front of it.
So, any person, could be a healthcare professional, could be a consumer, could ask and query the GPT and be able to get a response back in a natural language way that they could understand it. And that allowed us to democratize that knowledge so that anyone, anywhere could Start doing things that they couldn't have envisioned.
Not just pulling information and extracting it, but summarizing information. Being able to present it in ways that they couldn't have envisioned, like in code. That has just changed the whole concept of what's possible.
Yeah, and it just seems like, it seems like it was Thanksgiving of last year, all of a sudden, everybody, you're going to a party and everybody's talking about this.
And I think about the use cases within healthcare. I think about the ICU nurse who has to pull together all these notes and make sense of them and create a a common working framework for everybody who's going to be working on that patient. Those are the kinds of things that GPT an open AI kind of solution does really well.
It summarizes all this information and puts it in a, an understandable form.
That's right. I like to think of it in terms of three things. One. Being able to pull information from different sources that often times it takes a human to do. Maybe spend minutes or a little bit longer to do that type of work.
So, that's a very manual process. Synthesizing it and organizing it in such a way that we can actually understand it. And then presenting it back in that format that is necessary. In some cases, it's parsing it out and putting it into discrete fields. Other cases, it's actually creating a clinical note.
Other cases, it might actually be trying to reorganize this into a format that could be presented back to a patient. I mean, this is just incredible, now that we have the ability to do all three of those things very seamlessly with technology that is at our disposal. And multiple languages. Multiple languages, at your grade level, and with empathy.
I want to talk about, we'll start with clinicians, then I want to go to patients, and then I want to talk about adoption. How do we, in both of those cases, but let's talk about clinicians. So, Microsoft Nuance came together a little while ago. We're now changing the paradigm. It used to be when you went to Nuance, you ended up with a fair number of transcriptionists now that you were essentially hiring scribes.
But now we're not doing that anymore. So DAX Copilot rebranded is really is using AI to Give almost that same experience.
We believe that the AI technology is going to remove a lot of that administrative workload that somebody has to do Traditionally, it's been clinicians because it was very complicated to understand and required somebody who understood that context to be able to put that into the node or Put that into the field and we now realize that this can be a great assistant So that's why we call it co pilot because it really is an assistant to that individual doing these very complicated tasks.
We're going to come back to that co pilot design concept because I think when we talk about adoption, it's really interesting. On the other side of this wall, we heard about bringing all the data together. And what's possible. Now we, in healthcare, we've brought all the data together before, but it was never harmonized.
And 80 percent of it was sort of locked up in unstructured nodes. Are we starting to unlock that, that data?
Absolutely. That's one of the real exciting things that we're talking about. Being able to take advantage of the fact that we have now an ability to bring Different types of data sets together.
Electronic health record, imaging, genomics remote patient monitoring data. Put that in a way that, in one lake, that allows us to be able to have information readily at hand. To be able to curate that, to be able to run the AI off of that. And what we struggle with today is the fact that oftentimes we have limited data sets.
And we are only able to make information available related to a specific domain. Maybe just images. But without context. You can't really generate the AI and the insights that are necessary. So we need to have that combination of multimodal data in combination with the AI to make that truly effective.
what does this mean for researchers? Oh,
it's incredible. For researchers, there's so many great applications. First of all, having the platform that allows us to be able to pull all this information together. Use the ability to have OMOP analytics and be able to take the observational data and apply that on top of it, convert that to FHIR as well.
It changes the whole paradigm of clinical information, research information, and all the ways that we can now manage that. We have the ability with generative AI to start thinking about how we change the research process. A lot of researchers, it's a very manual process. They're extracting information from different sources, they are generating hypotheses, they're learning and pivoting based on that.
What we have an ability to do is be able to pull information from different sources. Could be published literature, could be other trials. Be able to organize and help the clinicians and the researchers now think ahead, generate new hypotheses. And then from there, be able to move quickly and faster than they've ever done
before. 📍 📍 We'll get back to our show in just a minute. Having a child with cancer is one of the most painful and difficult situations a family can face. In 2023, to celebrate five years of This Week Health, we have partnered with Alex's Lemonade Stand all year long with a goal of raising 50, 000 from our community.
We've already achieved that goal and we've exceeded that goal by 5, 000, so we're up over 55, 000 for the year. We want to blow through that number. We ask you to join us. Hit our website in the top right hand column. You're going to see a logo for the lemonade stand. Go ahead and click on that to give today.
We believe in the generosity of our community and we thank you in advance. Now back to our show. 📍
before.
You know, I, I, I do want to talk about adoption and we recently had an AI webinar and we had Stanford U C S D and U N C. These are three, they're out there on the cutting edge of at least using generative AI in the E H R and those kind of things. , one of the participants talked about the fact that they're doing the notes.
Everybody's doing the notes where auto generates the notes. And they said in 75 percent of the cases, the doctor just goes in and deletes it. It may have been accurate, it may have been, but it's still a comfort. Like how do we, how are we going to bring people along? How are we going to make them comfortable?
that the technology can support them and make their job easier. Yeah,
and I don't anticipate these technologies being used 100 percent of the time by every individual, but if we're able to alleviate some of that administrative workload in certain use cases and make their lives easier. That's what we're really looking for.
And so it's not about everyone has to use the AI technology. It's about, is it helpful? And if so, let's continue to use it and make it better. Well,
the interesting thing in the same breath, it was of the people who are using it. It's saving them close to 30 minutes a day. Oh, it's incredible. And you're sitting there going, well, you would think that they would talk to the other physicians who are just leading it.
But it's change management, right? I mean, every, you have to hear from your peers. It's like, hey, why are you getting out of work so early? And it's like, well, I've got this new technology that's really helping me. Maybe I should give it a go. And I think a lot of it has to do with just socializing and making sure people understand that this is working and that I should try it.
So
you were you were just speaking with the team from Duke. And it struck me, I was talking to some people the other day and I said, how would you feel? If you sent a note to your clinician and you got a response back and at the bottom it said this note was generated by fill in the blank, some AI algorithm, how would you feel about that?
Uh, Well
as long as there was an opportunity for me to know that this was helping the clinician to respond back. And it had the information that I needed. I personally would be fine with that. But, you know, I guess the question is if you are given that understanding that, you needed information and your doctor or your clinician needed some assistance to pull that and summarize it, that'd be fine.
It was interesting, the people I was talking to said, look as long as it says, hey, this was reviewed by my physician, Yeah. and, but it was generated in this way, they're like, look, I don't care if they use a computer, I don't care if they use a calculator, I don't care if they use... I expect them to use the tools that make them the most effective that they possibly can be.
So, recently Satya was on the stage at UGM. Which is no small deal. I don't remember ever seeing another partner up there on that stage. Where do you see that partnership, that relationship going in terms of some of the platform that Microsoft is showcasing here and the things that obviously EPIC has access to so many clinicians.
One of the things that we believe is that the technology needs to be workflow compatible. And when you think about clinical workflows EPIC and EHRs are a very important part of that process. And now that EPIC is embracing the AI technologies and building it directly into their systems to make them better.
I think they made several announcements not only about DEX DEX Express or DEX Copilot as being one, but they talked about the auto node generation for the Inbox, the ability for them to do, apply, generate AI for the Analytics and the Slicer Dicer. There's a whole series of areas that they're looking to build AI.
It's going to make the experience a lot better, more efficient for those using the EPIC system. We're excited to partner with EPIC to improve that experience for everyone using the system that they have.
I want to, so, Satya on stage talked about the perfect machine. Were you at the, at UGM? No, I missed that one.
The perfect machine, he's probably talked about it elsewhere, but natural language front end, reasoning engine, co pilot design construct, and he really emphasized that co pilot design construct And I think they talked a little bit about cars, we're okay as long as there's someone sitting in the seat.
We're still not comfortable the car driving itself with like everybody in the backseat. Talk about how important, you even named DAX CoPilot. I assume there's a reason. Absolutely. It
is a firm belief that we have that now with the natural language interface and the powerful reasoning engine in the back end that we have the capability to do so many important things.
But it's in the context of helping the individual do their job better, which is the copilot. And that's the reason why there was a change from Dex Express to Dex Copilot.
This is gonna be my final question. Are we at the Star Trek moment where we're gonna walk into an exam room and say Hey computer, give me the vitals for this patient.
The last time they were in, It's all going to come back. I mean, we're at that point, aren't we?
We have the capabilities to do that, yes. And the question is really about refining it, making sure that we put in the processes to enable people to take full advantage of it. That means making it workflow compatible, making sure that we put in governance around all AI, and making sure that people know about it.
David, I want to thank you for your time.
Yeah, thank you. It's been a pleasure.
Another great interview. I want to thank everybody who spent time with us at the conference. I love hearing from people on the front lines. It is phenomenal that you shared your wisdom and experience with the community and we greatly appreciate it. We also want to thank our channel sponsors who are investing in our mission to develop the next generation of health leaders.
They are CDW, Rubrik, Sectra, and Trellix. Thanks for listening. That's all for now.