March 28: Live from HIMSS, Chris Carmody, CTO & Senior Vice President for UPMC, and Tyler Bryson, Corporate Vice President, Health & Public Sector Industries, US for Microsoft, delve into the innovative partnership between UPMC and Microsoft, focusing on the transformative power of AI in healthcare. Chris Carmody shares insights into how leveraging Microsoft's AI platforms has revolutionized their clinical analytics, particularly during the COVID-19 pandemic. The conversation explores the critical role of data analytics in improving patient outcomes and operational efficiencies. How does UPMC's approach to AI and data analytics serve as a model for other healthcare institutions? What implications does this partnership have for the future of healthcare technology, and how might it influence patient care and operational strategies? This episode not only sheds light on UPMC's journey with AI but also prompts us to consider the broader impacts of technology on healthcare systems.
Categories: AI / Machine Learning, EHR
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. My name is Bill Russell. I'm a former CIO for a 16 hospital system and creator of This Week Health, where we are dedicated to transforming healthcare, one connection at a time. Today , we have an interview in action from the 2024 conferences, the spring conferences, VIVE in LA, HIMSS in Orlando.
Special thanks to our sponsors, Quantum Health, Gordian, Dr. First, CDW, Gozeo Health, Artisite, and Zscaler. You can check them out on our website, thisweekhealth. com. Now, onto our interview
(Interview 1) 📍 HIMSS, 2024. And I am here in the Microsoft Booth.
Tyler Bryson Head of Healthcare in the US for Microsoft. I'm shorting that long title. Chris Carmody with UPMC who has the privilege of heading up your EPIC implementation, which everyone has read about. Looking forward to having interviews with you along that journey. That should be fun.
Absolutely.
To no one's surprise, we're going to talk a little bit about AI, but we're going to talk specifically about UPMC employing Microsoft tools. Give me an idea of just a couple of the things that, we could jump off from there.
Sure. Probably the biggest thing we've leveraged Microsoft's AI platforms for is to help enrich our clinical analytics program to start. That's probably the number one thing. Dating back to the onset of COVID, we actually used Text Analytics for Health from Microsoft to do what's called entity extraction to pull out key pieces of information related to the different phases of COVID.
During the Delta variant, Our clinicians and researchers were able to quickly identify different patterns on what was working, what wasn't working with different patients that you couldn't readily get in the core EHR. That informed our models, it reinforced a lot of the decisions on how we were going to treat our patients, and it translated into better outcomes for our COVID patients at that time., everyone talks about telemedicine during COVID and the impact of it, but that work around analytics really helped transform us and evolve us as an organization to become more reliant upon our data.
Because we have data that digs back to 1992 that's been digitized. So almost 4 billion records. Moving forward from COVID, we continue to be the largest user of text analytics for health in the world as a Microsoft customer. We've gone back 12 years now worth of data and have used that to not only inform those clinical analytic models that do predictive risk scores across our continuum of care, but now we're using it for AI purposes and to leverage AI for the efficiencies that we can gain operationally.
Tyler, I want to come to you and talk a little bit about the partnership with UP MC. It's interesting because Microsoft's tools are being used in so many different ways in healthcare. These days, talk about the different partnerships you have with health systems and how they're utilizing these tools so differently because you're giving them a, like a toolbox of things they can do.
Yeah, many people don't know about depth that we've been engineering healthcare into our cloud. At almost every layer of our services, there are healthcare specific capabilities. Chris just mentioned a couple of them. These capabilities are becoming more and more AI ready and more and more data driven.
So that a health provider or network can just jump into the Azure cloud and use these services in a studio. And have this access to capabilities that five years ago would have taken a lot of engineering to do,
it's really interesting to me that when we talk about bringing AI into healthcare, having these kinds of tool sets in a sandbox and all this data together really creates a lot of opportunities.
Talk about your EPIC implementation a little bit and how you guys are stepping into that with regard to Microsoft.
Yeah, I think it's important to give the context of the relationship with Microsoft. We made a conscious decision to invest in their platforms and their technologies from an infrastructure perspective.
It really helped us scale with our growth. Over the last seven or eight years, UPMC has basically doubled. Number of users, number of physical locations. Through acquisitions and organic growth so we needed a partner with someone who could help enable that for us. We jumped on the cloud. We were one of the fastest users in deploying Office 365.
That became really useful for myself and my job. We were told to send, from 2, 000 or 3, 000 people per day working remotely 📍 to 📍 70 2, over 📍 a weekend , right? So that scale was important to us. And then as you saw that the strategy and the roadmap for Microsoft to embed AI in the products and services and allow us to control things.
I think that's key because, you're going to see lots of different solutions that are built for someone else's problems, quite honestly. And I think Microsoft enables us to. Focus our AI efforts to do things like we're focused on, which is building efficiencies into the health care system.
So some of the traditional approaches to things change. Yeah. So you guys have the big announcement, you're moving to Epic and people are like, all right, how has that changed since we moved to Epic all these years ago? And you guys are stepping into it in a very different way than most have.
Yeah we're taking a little creative approach especially again, back to the infrastructure of NIS.
We're going to have a pretty large instance of Epic once we're wholly deployed in a couple of years. But in order to help us get there and evolve and leveraging new technologies and platforms that allow for that scale and flexibility, we consciously made the decision to put all of our pre production environments of Epic, all the configuration build test environments, into Azure, into our tenant of Azure.
So we can control, we can scale it, we can spin it down at night when no one's on the system and really be efficient in our use when we're dealing with, constraints from a financial perspective. So it's all positive things to position us moving forward into the future.
Tyler, what's it like to, adapt the cloud to healthcare? Because we're talking about technology that was built a long time. I mean, No offense to Epic, the foundation, though, is fairly old. It's reliable, but it's fairly old. Does that just plug into the cloud, or does Microsoft have to adapt for that?
Yeah, one of 📍 the things we've been working on is, with Epic's help, really designing our cloud from the ground up. To me, that makes requirements. Right. While at the same time, leveraging the capabilities that the cloud has, that environments 10 to 15 years ago didn't. Working closely across our engineering teams, we're building solutions together, building and testing and certifying environments.
Epic is also using Fabric as their data platform for many of their new services. So you can see there's integration and we're heading stronger 📍 and stronger.
So,
Chris, we're going to close with you. What's next? I know you're focused on the EPIC implementation, but you guys can walk and chew gum at the same time.
What are you thinking about with regard to AI and what's going to be possible?
I think, again, leveraging the platforms from Microsoft, we're focused on those healthcare One of the biggest, wins we're going to see very shortly using that Text Analytics for Health solution to do the entity extraction out of voice recording.
We have about 800, 000. Patients and health plan members are calling for their prescription rebill. A human being has to listen to it, write it down, check the different systems to verify that it's accurate, and then send it to the pharmacy to be fulfilled. We're working right now.
800 thousand,
800 thousand. Wow. Is
that a year?
That's a year. Wow. Think about the efficiencies that'll create in our system by automating that entire process. That's the journey we're on from an AI perspective, among many others, but I think, when you look at AI concern with the hallucinations and things like that and training models around your data, I think being narrowly focused, what we're trying to do at UPMC is going to help push us forward and actually be successful with it.
So I'm that pragmatic guy, so phone call comes in, transcription happens, that same technology we were talking about that went into the node and pulled out the key data is going to pull out the prescription data. Yes.
then, a service like Azure OpenAI can 📍 help with natural language processing. What we're seeing is multiple AI services working together to improve.
it's the AI studio then that takes that extraction, now puts it into a workflow, that avoids a human being to check the system, that the AI platform will check the system, validate it, and then send the prescription off, and notify the patient, Hey, your prescription has been sent to the pharmacy, it's ready for pickup.
As you're talking, I'm thinking refills, but now I'm thinking of A host of other things. Absolutely.
Let's walk before we run. I
know. And that's part of the problem with AI. We have so many people talking about five years down the road, but there's a lot of practical things that we can bring today.
By doing, right? Absolutely. We can harness the power of these programs 📍 safely .
And that's part of our governance model. Again, there's many AI solutions that are out there in the industry right now. Not fully baked, but As an organization, you have to prioritize what makes sense. Where you get the biggest bang for your buck in investment.
And that's, our approach. I think that's why we'll be successful.
Well, .Thank you for your time.
Thank you, Bill. 📍 📍 Thanks
for listening to this Interview in Action episode. If you found value in this, share it with a peer. It's a great chance to discuss and in some cases start a mentoring relationship. One way you can support the show is to subscribe and leave us a rating. If you could do that, that would be great, and we want to give a big thanks to our partners who make this possible.
📍 Quantum Health, Gordian, Dr. First, CDW, Gozio Health, Artisite, and Zscaler. You can learn more about them by visiting thisweekhealth. com slash partners. Thanks for listening. That's all for now.