This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.

Matt Sullivan: I would love it if our clinical teams went about the business of taking care of patients. Never had to touch a computer again.

Bill Russell: My name is Bill Russell. I'm a former health system, CIO, and creator of this Week Health, where our mission is to transform healthcare one connection at a time. Welcome to the 2 29 Podcast where we continue the conversations happening at our events with the leaders who are shaping healthcare.

Let's jump into today's conversation.

All right, it's the 2 29 podcast, and today I am joined by Matt Sullivan, CMIO, for advocate. And 2 29 Podcast is about continuing the conversations. We start at the 2 29 project meetings, and Matt, you guys have such a great group with the C MiOS. It's been fun to see the conversations that you guys have had over the years and the group of you know, of CMIOs.

Hey, welcome to the show, by the way.

Matt Sullivan: Yeah, thanks for having me. Good morning.

Bill Russell: What's new in your world, CMIO for advocate? I mean, at one point you were down at Atrium, then it's advocate Aurora and Atrium come together. How does the job change with scale?

I mean, it's huge.

Matt Sullivan: Boy. That's a really a great question. I was hoping to ask you the same and maybe get some insight to sort of. Yeah. But the reality is, is a, a ton of our time is spent trying to figure out how to scale. I mean, here we are at 11 o'clock this morning and already three times today I've had conversations, have meetings, and closed the meetings, and then gone back into the simple conversations around what we just talked about seem to be coming at it from two different vantage points.

It's an ongoing problem to figure out how to take different groups that maybe come at it culturally from the same, you know, good spot in your heart but tactically come at it from a different perspective. Which things do you keep from one versus the other? And then how do you bring them together so that the marriage of the two is actually effective and hopefully if we do it right, more effective than either of the two that you had.

So it's a challenge is the answer is the singular answer is it's a challenge, but it is, part of the work that's required in order to get us where we need to be.

Bill Russell: It's interesting. We had similar problem sets in that at St. Joe's. I'm thinking culturally I had West Texas and Southern California, if you can imagine two culturally different areas.

And I remember having some of these meetings where I mean essentially physicians were telling me we don't practice medicines the same way they do in Southern California. And I don't doubt that some of that is true. I mean that there are distinct regional kind of things and approaches.

But it didn't make those decisions any less difficult?

Matt Sullivan: Yeah, I think the decisions will always be difficult. And I actually completely agree with you there, there are always the physician groups or even nursing groups that say, Hey, what we've got's a little different over here than what you've got.

So I think the issue for us initially is to ground on what's important. And so I think that is. The beauty of our senior leadership has come together and said, look here's the strategic drivers for Advocate Health. It's not a holding company of a bunch of disparate little hospitals that do their own thing.

And we have to be mindful of our mission and at Advocate Health, the mission of sort of serving all is really huge. So we're a for all company where. We have to tackle things that work both in downtown Chicago and rural Georgia. And that, that is like your example although in Texas, I think they probably would consider themselves to be unique in all cases.

Bill Russell: They were. They made it clear. They made it clear. Every time I flew in there, it's like, you're from California. I'm like, I'm not even from California. I grew up in Pennsylvania. I lived 20 years in Missouri. I'm like Midwest, and that's why they kept sending me. They're like, yeah, you're, you know, you speak their language.

I'm like, St. Louis and Lubbock are very different places, different animals.

We gotta achieve this. How are we doing it now? What's really different? Like, because there are differences and then there's real differences. And when you get down to it. The way you practice medicine isn't substantially different from one place to the other. Each medical case, each patient is different, but the drive, the focus, the execution, and I learned that sort of in my own practice over a series of years, moving from an academic center to working in the community to going to freestanding eds, to driving pretty far out to understand that.

The patients that I was seeing when I was practicing. As it turns out, those communities aren't different. When you have chest pain, you have chest pain. When you have abdominal pain. Abdominal pain.

It doesn't matter where you show up, it's the same workflow. We're gonna document the same way. Because we have to document the same way. It's driven regulatory and those where do you like, draw the line and say, look, there's no discussion here. Up here. We can talk about the variability, but below here let's try to drive systemness if we can.

Matt Sullivan: Well, we drive systemness by saying here are the lines. And I think the lines are very clear and they're actually pretty similar, regardless of your size. So the lines for cyber, right? Right. We're not gonna let you do this. These companies, these software programs, these tool sets are dangerous. So no.

Okay. And that's one of the very few. So,

Bill Russell: so they're not copying, pasting from Epic and putting it into chat GPT and doing that stuff.

Matt Sullivan: You can use the brand names of whatever, it's that principle that we're saying, Hey, time out. We can't do that. Right, the safeguards, although they say they're there, they may not actually be there.

Let's use our EMR vendor as an example. If the EMR vendor hands you a solution set that says, look, it's gonna work like foundational approach, and it's gonna get you all the things you need, the stars and the magic. And it turns out that workflow is really efficient and it drives the work in a better way for the downstream clinician, whether that's a doctor or a nurse, or a physical therapist.

Those folks see that pretty quickly. I mean, that. to be really frank, that is the beauty of working in medicine. These systematized issues happen probably all over the world in all sorts of industries, but in medicine we have people that are driven for patient care and generally are very smart.

That doesn't happen very often anymore. It used to be the case, right? Hey, look what we've got. People go, oh, that's horrible. You mean I have to type all this stuff in and then I have to go to this other place and then do, why are you having me do that? Can't we have somebody else do that? Yeah. We're starting to that inflection point change to where the vendors are making the work better and making the outcomes better, and some of that is fueled by AI through things like, you know, ambient dictation where.

Some people love it. They walk out and they go, oh my gosh, look at this. Did my whole note for me. And then there's always people that have a little bit more, you know, OCD tendency and they wanna review the whole thing, rearrange it. That's fine too. Both of those scenarios were far less than somebody typing stuff in

I mean, why are you. Confident in a future that has AI embedded in the care setting.

Matt Sullivan: So I'm gonna frame it using the computer vision a little bit. For a long time I would say to people like, really close friends, look, if you're really sick, you're gonna wake up one day in the middle of the night and I'm gonna be sitting in your hospital room.

Because I've worked at lots of hospitals at night and I know that like there's not everybody's watching all the time. And did that

Bill Russell: comfort them or were they were a little scared,

Matt Sullivan: frightened to death and it's probably fair. But the idea that someone else is watching over you all the time, that has some knowledge of what should be happening, what shouldn't be happening in a sense of trying to wrap your.

Bill Russell: let's, Let's hope.

Knowing

Matt Sullivan: her, I'm pretty certain that she's gonna be at the hospital and we see patients and families who are exhausted. They've been, in some cases, depending on where you are, sleeping in a recliner and eating food out of a cafeteria that's not sleeping in their own bed. It's not their own stuff and they're discombobulated, which doesn't help the patient when we've put ambient

Someone will come on that camera and someone will come on the TV screen and be right there for him because, you know, pushing the call bell button hasn't been working for us. But this is a much better situation and so that's why I'm bullish on it. That's why I think there's a value there. Now,

Bill Russell: The alternative would be if you could put a physician or a nurse in my room, 7 24 365. That would be my preference. We just can't do that.

Matt Sullivan: we've seen that in the medical literature that lots of specialty care is primarily driven off having that specialty, highly trained nurse at the bedside continuously in like a neuro ICU versus regular floor.

Now that has nothing to do with efficiency, that has nothing to do with sort of cute or high tech fun. Like we all love high tech fun, but that's really around making sure that we have somebody watching. And then the second component is there some AI that can not only watch but watch with them for specific things relative to the disease process?

And so I think about what the future might look like for respiratory disease. And we'll just pick an example. I don't, there's lots of examples, but someone's having trouble breathing and there's a certain, there's a certain rate at which you breathe and there's some difficulty and there's a inhale component and an exhale component, and some of that stuff we check off like once every eight hours.

Is somebody getting worse? And if we know somebody's getting worse, can we identify that slightly earlier than we previously been able to do? And does that actually change their outcome? And in order to do that, we've gotta intervene with something. And I'll say something like a breathing treatment just to sort of.

Be generic about it, but if your respirations are, and your respiratory function as viewed by an AI agent who can sit there 24 7, 365 and watch as it goes down slightly, if we intervene early and it flattens or improves and we show, then people get out of the hospital a little bit faster, don't have as many problems, don't need as many treatments, lots of outcomes.

You wrap your head around the possibilities. But I think that's where it's helpful.

But culturally, I, you know, there was a, there was a strike, I think it was in California, where the nurses went on strike and said, you know, no AI in our environment. And, there was a fear that it would do something. It would, you know, take away their work, which we have such a nurse shortage.

I can't imagine why that would be a concern. But that was one of their concerns. And people don't know what to make of AI yet, including nurses and some physicians and others within the health system. How do you say, Hey, let's try this out. Let's see it, let's see what's going on.

Matt Sullivan: This was my eight o'clock call this morning. And it was with an IT person, right? And a longstanding IT person who I've known for many years. Great guy. And we were trying to work on something like we always have been, right? And. Now that we have AI in the mix I had created an agentic agent to help with some of that work.

Point the agent at it and it'll just tell me what happened last month. And it'll just. No, ignore that. Tell me what happened last month. Put it in a chart and then put all the dumb stuff at the bottom. Put all the really cool stuff at the top and he just couldn't wrap his mind around it. And it took me 10 minutes to say it actually will work that way.

And then the next question, well, don't we have to then change everything to be in Excel? And the answer is no. We could just throw it in as Word documents. You can throw it as whatever you throw in the SharePoint folder. Point that there, it'll look through it all and it'll help. You want to change it around, you wanna alphabetize it.

If I said to a nurse, AI is coming and we're gonna try to get it to do everything that you do, and we're gonna try to get it so that you can just walk around and do the things that make you as a nurse feel good in the old way of caring for patients. So that you don't have to be a slave to the computer.

And then all of a sudden people go, oh yeah I want to, I wanna work at that place. And so when we wired up our entire hospital, it was amazing to hear the nurse leadership say, this was a, this was new for us. This was a challenge for us, and we were a little skittish. because we always have a nursing shortage.

We've got all these travelers that we've got and then we've got all these people leaving and have a single nurse leave. In the first year, everybody stayed. And I just, and I was sitting in this, you know, you're rounding with a lot of, you're in a suit, you're rounding with a bunch of leaders and you sit there and you go, oh my, did she just say that?

I won't say quite in the middle yet because I'm really hoping to have more room there. But I would love it if our clinical teams went about the business of taking care of patients. Never had to touch a computer again. Will we ever get there? Probably not.

Bill Russell: Yeah. I mean, that's the challenge with all of these projects is the getting the buy-in from people.

It's like, Hey, we're gonna try this. Let's see where it goes. Advocates, what, how many hospitals now? Is it 40, 50?

Matt Sullivan: I think it's seven or 8,000. I can't keep track of it, bill. Honestly, I think it's,

Bill Russell: They tell you, Hey, can you go up to this hospital? And you're like I didn't know we had a hospital there.

That's gonna be fun.

And that's pretty fascinating I think Drew.

Bill Russell: I'm sitting here. I used one of the tools to look up Matt Sullivan, C-C-M-I-O, advocate Health. I didn't know you spent time in at Hahnemann in Philly. Yeah,

Matt Sullivan: that's where I trained.

Bill Russell: That's where you trained. Wow. And yeah, I didn't, I didn't realize you spent time in Philly.

Is that where you started your career?

Matt Sullivan: Yeah, I went to medical school in Philly. It was just four years. Wow. And then I came immediately down to Charlotte. So I'm a lifer here in Charlotte. I was a medical student. I did my residency here at Carolina's Medical Center, did a research fellowship.

So

Bill Russell: Your career's not the norm. As you know. I mean, CIO means careers over CMIO probably means something like that. And you've been in that role for quite some time, even through mergers and acquisitions and other things.

Matt Sullivan: Yeah. I mean, I was A-C-M-I-O or I was an associate CMIO with Brent Lambert for a long time and we've been banging around through a couple of EMRs and probably now on six acquisitions of mergers and so.

It's been an interesting survival story.

Bill Russell: The first time I met you, we I know we ended up sitting at a table together. It was you and your bride who is phenomenal. Also works at advocate, and, and my daughter and myself, oh, were there.

And there was somebody else there. I forget who it was, but we were at that table that laughed all night and everybody else kept looking over going, what? Because you guys just,

Matt Sullivan: there was like three boring tables. And then we were just, oh my gosh.

And then the next time I saw you. Was at the Scottsdale Institute and it was like a complete role reversal. because you were going through a, an I forget like either your clinicals or your rev cycle had to be switched. I forget which one it was, but you were right in the middle of it and you were not home, so you were trying to troubleshoot that stuff over the phone.

I'm

Matt Sullivan: desperately out of fish outta water and exhausted. So I think we lived through all that stuff. But keen on a very important point bill, which is around how do you live and work in the environment that you're in? You know, I'm clearly not the smartest guy, as you know, um, and like, so why is it that, that you can be successful in an organization?

And I think part of it is that you've gotta be willing to tell the stories good and bad. Like we've had some amazingly. Fascinating mistakes and you have to be willing to sort of talk about those and learn from those and have a little bit of fun. Like we, we had a major script yesterday and I called somebody, I was like, really?

Have a little bit of humility and just try to get the right work done at the right time.

Bill Russell: I'll tell you, you did dodge a bullet and we've talked about this as well. The bullet you dodged was, I did interview for the Atrium CIO position, and I don't, well, I don't know if you were in the room for my interview or not. There was five people in the room and I did not get that job.

So you ended up with Andy Crowder and I think that was the right decision. Like if they had asked me. Between me and Andy, who should be the CIO, I would've said Andy. He has a better temperament for being a CIO.

Matt Sullivan: We would've survived, but it would've been rough.

I think

And it's, I don't know if the pace is quickening, it feels like it's quickening. Is that one of the key qualities you're looking for now? Is somebody that can adapt to change fairly quickly?

Matt Sullivan: the IQ EQ balance is always in the mix and the sort of that adaptability and ability to move and roll quickly and again, I think the pace of change over the last three years has been really much faster than it was.

Some of that is AI driven, and some of that is sort of venture capital into healthcare with lots of opportunities for different vendors, different solutions, and everything from bolt-ons to completely freestanding solutions that rip and replace what you already exist, like your existing stack. And so somebody has to go through and evaluate those things.

I want you guys to look at this. Sometimes you have to look at those things because you, it's politically the right thing to do. But sometimes you look at those solutions and you look and you think, this is not gonna work in our current environment, and here are the five reasons why. The ability to take that, understand where that sits in the strategic and tactical stack, and then pull that apart.

Have a couple of quick conversations, make sure you're on the right footing. Go back to that senior medical leader and say, I don't wanna look at this. It's not because I don't want to, or I won't do it. I will if you want me to, but here are the things that I think are gonna make it difficult for this to be successful in our environment.

Do we build it internally? And that's the conversation that we're having with ai. Right. How much do you take off the shelf? How much do you partner with people to modify and take a startup and do something new and different and help them and help yourself at the same time? And then how much do you just say, forget it, we're gonna build it internally because we have the tools and capabilities and scientists to do that, then that's our struggle.

Bill Russell: And I've been surprised how many of these conversations where I go, you know, well, how does that align to your strategy? They'll go, well, our strategy's kind of nebulous and we kind of tack things on here and we tack things on here. And I just, I sort of look at that and I'm going, these are billion dollar organizations that don't have a clear roadmap.

And when the clear roadmap is not set, which is the number one thing that leadership does for the people downstream is set the direction. When they do that, it becomes clear what to say yes to and what to say no to. Because it is hard to say no to things in healthcare.

But if it's not within the strategy, that's like the backstop to say, look, we're heading in this direction. This is cool, but that doesn't get us in that direction, but so let's keep looking at things in this direction. because that's where they told us we should go.

Matt Sullivan: Yeah. And sometimes that's hard because the cool stuff is.

Bill Russell: It's cool.

It doesn't make breakfast for you. Okay. Well, can it make toast? No. Can I get me coffee? No. Like that was my funny little Quip Now like, the funny thing is. Yeah, that's cool. But we can't do it because we're doing this cool thing like that. It's the, yeah, that's cool. But this is more cool.

It's this very strange interplay between the operators and the technology and the strategy that, well, and

Bill Russell: the other thing that's hard is when you have scale people are you can't use the, well, you know, we don't have funds for that. But I mean, you will say that. because it's like, look we're, it's not unlimited funds, right?

But we've chosen to invest in X or Y. Speaking of which, you just you were up at UGM. Yes. because you blew me off up at UGM. That's right. You were there you

Bill Russell: I was very tired. You know what's your takeaway from UGM?

I mean, we have Emmy, we have Penny, we have Comet, we have Cosmos you know, along with a whole host of other things. I mean, plus your team presented a ton up there. I mean, advocate was all over the place.

Matt Sullivan: Well, it's a one scale and two, a great team of people that does great work. And so I love the fact that they showcase that.

We learned a lot internally. Still e even though it's a, a marriage of years now. We've bringing people from multiple areas of the company together and they meet together in 3D. And you learn a little bit there. It's just part of the deal which is fun. I really enjoyed that part.

And so. The secret sauce is to try to figure out how does our strategy line up and pair with the things that we think they're not gonna do very well at, so that we can continue to accelerate, maybe learn the ropes, maybe help them understand better so that in the end, their final product is really good while we keep iterating on the front.

And that is a little costly. But if we can justify the ROI and the investment of time and energy into it, and really sort of be on the leading edge in just a couple of spaces that'll allow us to focus. And it'll allow Epic to do some of the other bigger tasks that we know they're gonna be good at.

Some of the backend stuff, that they're really gonna be good at that stuff. And we know that they've been solid there.

Bill Russell: Well, we're getting close to the end of our time. And since you are a physician and your wife said I talked to a physician immediately following the presentation and he asked me, you know, what'd you think?

And I would like to know that they have you know, somebody looking over their shoulder saying, Hey, make sure you order that test for Bill too as well. And I said to him, I'm like. I think in a year or two I'm gonna only see doctors that are using Cosmos. because that gives me an added feeling of.

I don't know, se security, but whatever the feeling is that makes me feel better about that I'm being cared for. And it's sort of the opposite of what we hear about ai. It's like, oh, we're worried about AI's gonna hallucinate, whatever. And I'm sitting there going, look, I trust Epic. I trust 300 million patients in there.

The amount of data they can throw at this. I'm like, I want that second opinion. I want that second set of eyes, if you will.

And so 300 million patients and 50 million of 'em have yellow hats, and that's what you want to have is your oversight. There's more to you than just the yellow hat

Bill Russell: there. Absolutely.

Matt Sullivan: You're gonna need that person tired or not. We've been tired for a long time, but tired. You're gonna need them to use the actual human brain to analyze.

Justify that bill has a pink shirt and a yellow hat. That's different. Submit. Like, we don't want that. And we've seen that through alerting. Oh yeah.

Bill Russell: Yeah, absolutely.

Matt Sullivan: I think we're gonna be in that same swinging of the pendulum and my hope is that it will land where you and I want it to land, which is a.

Non-intrusive safety net oversight that allows that depth of knowledge of 300, soon, 500 million patients to be right there at the fingertips of the clinician so that they can use it exactly when it's appropriate to be used

Bill Russell: I mean, well, and that is one of the things I spent a lot of time reading about was just this whole idea of context and drift.

It turns out that AI models perform better. If you don't dump 300 million patient records into it. It performs better. If I'm seeing Bill and I just dump in that subset of yellow hat wearing patients instead of 300 million because it just, it's all, the rest of it is noise

Matt Sullivan: essentially. So how do we split?

Yet mistakes are made. We know that. We've seen that in the literature. You know, you can talk about the quadruple aim and how we should be hitting it all the time, and we don't. We know that. So there's power here. How we apply that power is, I think, the secret sauce to making healthcare better, making our teams more efficient, and making people feel like when they go to work, whether they're on the.

Doctor side of the house or the patient side of the house, that it's gonna be the best that we can deliver. And that's, I mean, that's why I go to work every day.

Bill Russell: Yeah, absolutely. Matt, it's always great to catch up with you. Love to hear what's top of mind for you guys and where you're going.

Thanks for your time.

Matt Sullivan: Thanks Bill.

If you have a conversation, that's too good not to share. Reach out. Also, check out our events on the 2 29 project.com website. Share this episode with a peer. It's how we grow our network, increase our collective knowledge and transform healthcare together. Thanks for listening. That's all for now.