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

The 229 Podcast: Balancing Innovation vs. Demand and Top CxO Focuses with Chris Paravate

Bill Russell: [00:00:00] Today on the 229 podcast.

Chris Paravate: you know what the physicians tell me.

I am not worried about how many new patients you give me. I'm not worried about how many complex patients you give me. And I'm actually having some fun 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 229 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 229 podcast and today we are joined by Chris Paravate, CDIO of Northeast Georgia Health System. Chris welcome to the show.

Chris Paravate: Thanks, bill. Super excited to talk with you.

Bill Russell: Well, when we started doing this podcast, one of the things we wanted to do is continue the conversations we're having at the [00:01:00] 229 project. But yesterday you were on stage at Georgia HIMMS. Yeah. And I wanted to, I mean, let's start with, you know, what are you hearing?

What are you seeing out there?

Chris Paravate: Yeah, I mean, I'm sure I'm not the first one to say that AI is not a fad. It's interesting for me to hear from other leaders about how that's taking shape in their organization. Governance is a big thing. I hear different strategies around speed and adoption.

I hear. Different risk tolerances. So it's kind of, and then, you know, I, a good colleague of mine, Jeff Brown, said, there, there will be some aggregation, right? And I think that's a trend we've seen in other technologies. Well, I suspect we'll see it in this space as well. But I hadn't even thought about that, right?

I've been so focused on, you know, find a solution, make some transformation, demonstrate value. Accelerate. Right? And so I thought that was [00:02:00] some sage wisdom.

Bill Russell: Well, lifetime Achievement award. Love that for him. I'm gonna dive into the topics you just mentioned. So we'll start with ai. Okay. Most people apologize for talking about ai. I'm gonna, I'm gonna stop apologizing for talking about it.

I think it's a real thing You say it's, you know, it's not a fad. And I agree with you. The tension that we're hearing at the city tour dinners and the summits is , between the ethereal and the real practical applications of ai. And you know, and a lot of organizations are saying, look we're trying to figure out how to sort of weed these out.

They are worried about consolidation. Like, you know, you go with some small player and all of a sudden they get acquired and now you're looking at something completely different. But , how are we gonna find those practical applications? Are we just gonna wait for enough people to try 'em and be fast followers?

Chris Paravate: I'm such a no frills person, bill, and I think, you've kind of got to [00:03:00] know me a little bit. So I, you know, I don't grandstand on that stuff. I really do look for the very practical and then I look for architecture and design, right? So, you know, we're a big snowflake shop. We do a lot of stuff with Snowflake tools.

We look at architecture and we're saying, okay, well if it's a small startup. What's their architecture, what's their design, and what's our ability to maybe even bring that in-house? Right? But you know, speed is a big deal, we weren't the first to adopt ambient voice recognition. But we've gotten in two months, we went to 80% adoption.

You know, eight in 10 providers in our organization use ambient voice recognition now. Just hot off the presses. because you know, we're chasing KPIs in a big way. Those physicians are seeing 2.75 more patients per week.

Bill Russell: And we're not allowed to talk about that. Why not?

Chris Paravate: What's wrong with that?

And you know what? The [00:04:00] physicians like it and the patients like it. I know.

Bill Russell: I agree with you. And it's, but it's an interesting thing. You can't bring it into the system and say you're gonna see more patients.

Chris Paravate: I didn't ask a single person to ask to see more, but you know what the physicians tell me.

I am not worried about how many new patients you give me. I'm not worried about how many complex patients you give me and I'm actually having some fun again.

Bill Russell: Yeah.

Chris Paravate: And doing what I'm doing. And I'm doing it without the worrying about what is piling up. Right? I mean, think about. You know what our lives might be if we didn't worry about how many emails were gonna be in our inbox.

Right. I mean, that's the parallel,

Bill Russell: right.

Chris Paravate: You know? Right. You have no, I'm in meetings all day. Guess what I'm gonna do tonight?

Bill Russell: We haven't really figured out email yet. No,

Chris Paravate: No, we haven't.

Bill Russell: We come outta UGM and they're launching, you know, another.

Whatever number of AI features on top of the a hundred and something they launched last year, and I don't see that slowing down. I think every year we're gonna see a [00:05:00] whole bunch of things added to the let's just call it the Chester drawers over here that we can start opening up and implementing what does it take to, you know, go to UGM, sort of digest that and then.

And then operationalize the ones that make the most sense for Northeast Georgia.

Chris Paravate: You know, there is 160 use cases at UGM published from Epic on AI capabilities. And when you look at the rate of adoption, it's really low. It's really low. In fact, Northeast Georgia's pretty progressive in that adoption.

And not as deliberately pursuing that as I would like to. That was like an accidental, so,

Bill Russell: so do you take those use cases, bring them back, and there's a group that you sort of sit down and go, okay let's go through these,

Chris Paravate: working

towards that. I mean, I'd love to tell you there's two, there's three things that are on my mind.

One is how am I [00:06:00] differentiating? Those AI use cases from the rest of the work. Right. So I just finished a call and we're looking at our HIS system and we have 320 enhancements in that list. 320, and that's. You know some people who listen to this and say, ah, that's nothing. Some people would be like, that's crazy.

It doesn't really matter. Here's the thing is I can't throw those AI solutions in that pool. They just get lost. Right? I need to differentiate, and frankly, I'm working with Dr. Hanley, our CEO to say, Hey, how do we pull these things up? And I can say to you, Hey, I can get you a $3 million. Reduction in costs and denial appeals letters in six months.

Would you like that or would you like me to redesign the registration screen? It is just, they're not in the same hunt. Right. And these solutions can be [00:07:00] implemented frankly in weeks. Right. So, and, you know, for a decade we've thought about implementations in terms of. Years or months, or, you know, six months, we're talking about, with some focused energy, right?

Think about like COVID focused energy. You could turn up these things in a couple weeks and say, are you realizing value or not? I don't have that figured out, but I'm literally, and when I'm talking to whether it's Epic or Workday, I'm asking the question, can we do this in two weeks?

Bill Russell: That would be really nice. At a recent city tour dinner, 229 City tour dinner, we we were having a conversation and people are like, Hey, nothing's changed. We have a set of problems that we need to fix. And if AI helps to solve those problems we're going to do it.

And you know what I do? I mean, I generally agree with that statement. But my job in those meetings is to push a little bit and say, okay, is there a different way to think about this? Because essentially we could just keep, okay, we have this set of [00:08:00] problems and we have these tools and we're gonna solve these problems with these tools.

It depends how strategic those tools are. It's

Chris Paravate: magnitude though, right? I mean, if the AI solution among, compared to an enhancement request, right? Enhancement requests make five or six people happy. An AI solution allows those five or six people go do something different. , The scale is far more transformational.

And that's what I'm trying to bring that attention and I'm trying to figure out how to harness that to go. Enhancements are good and projects are important, but I got this category over here that's like a really hot catalyst and how do I maybe think about how we prioritize that work differently?

Bill Russell: I mean, one of the things I tried to do as a CIO and I had a great conversation this week with a Chief financial informatics officer, and it was a really good discussion. I know that individual Yeah, you do know that individual. And she was she was talking about all the things that they target.

And [00:09:00] what I like to do as a CIO is to say, okay, what we need to do is. I it's sad to say this, but but money is the fuel that makes IT go. Money is the fuel that makes, you know, gives us the ability to invest into right. New access methods and all or sorts of other things. And so I would sort of align our resources around things that were going to get us a couple million dollars here and a couple million dollars there.

And, you know, if you can take denials from. 5% to 2%, you just freed up a couple million dollars for the health system. Yeah. So I mean that work to me seemed I mean this is the application of AI to just solve the problems we have today is one way to go. The other way to go is to say, alright, these things could be transformational.

I mean, you talk snowflake, you talk, I mean, you have a lot of really good tools. If we elevate it even one step higher, is there a chance? To fundamentally change the economics and the quality and the outcomes that are happening [00:10:00] at the health system.

Chris Paravate: I think the answer is both. And you know, I talked a little bit about this yesterday at the center, we're still a people business.

Bill Russell: Yeah.

Chris Paravate: Right. And as a community health system, we take care of everyone regardless of their economic status, their whatever. Caring for patients. Just the medical care, the clinical care is complex, right? And it's a dangerous business, but that's not all that comes in the door, right?

What comes in the door is every family, you know, disagreement, every economic, social situation that you could imagine. And so for those clinicians to be. A little bit more in the moment and a little bit less distracted by the noise to care for that. That's at the core of our business. That is our tangible value.

The durable value that Northeast Georgia brings to our community is the care that we deliver, right? So if I can [00:11:00] reduce all that noise, whether it's denials or a keyboard. I could be in that moment with that patient, I could have the energy to be with that patient. I don't think that those things have to be there's nothing wrong with pressing that expense out of there, reducing that cost, because the truth of the matter is, you know, you drive an hour north of here and it's very rural and it's very poor, and those patients need care.

If we can care for 'em, you know, in that community then, we're one, we're providing a better service, but two, we're not taxing a resource in other communities like here in Gainesville, so this is it, you know, the equation is you can satisfy both, we can implement an AI tool that provides better care and reduce costs, that's not a sin.

Bill Russell: I think the biggest buzzword since cloud computing right now is upskilling. I mean, the number of times I hear the word upskilling . What does that [00:12:00] mean? I'm gonna give you a, an interesting story to, to bounce off of.

At one of our dinners, one of the I think it was A-C-M-I-O said they asked their IT organization, so they, the entire IT organization on a Zoom call or teams call, whatever happened to be. And they said, you know, how many of you have used a generative AI tool? Have used chat, GPT, Claude Grok Gemini.

Whatever it happens to be you know, for work or for personal use. And the person said, I got just shy of 60% of the people in the room. He said that the only problem with that is this is it. And this represents less than the general public Yeah. In terms of use. And he goes I was a little concerned like, you know, how are we gonna upskill this workforce and what does it mean to upskill this workforce?

Chris Paravate: Two comments I have about that. One is that there is a younger generation that have just grown up with these tools, right? So, know a young individual [00:13:00] who is working to sell ERP solutions.

He's never implemented an ERP system. In fact, I don't even think he's even logged into an ERP system. But he can use one of those AI solutions to say, where is this customer? What are they implementing? What are the key considerations? What would be a strategy to address backfill or conversions? And he can start to assemble a roadmap to really help.

That customer think through that planning for you and I, that was 10 years of experience, right? So there's this knowledge that this younger group is just going to leverage and use because it exists, right? And yeah, you see that in the consumer space, right? My daughter and her husband were sitting at the kitchen table the other day picking colors for their house.

Using chat, GPT. So I think that's one aspect. The other aspect is this, you know, you recently published an article about, AI adoption [00:14:00] and how do you get to transformational? And you spoke a little bit about education, and I've been struggling with this thing called education. And you use the skill, you know, upskill I'll talk about upskill in a minute.

So yesterday I was driving to that event. And I'm like, dang it, I still don't have a good answer. I still don't have, oh, here's how I'm gonna upskill my team on ai. So I'm gonna try this experiment. This is what I came up with at six 30 in the morning, driving to Duluth in traffic. Jeff Morrison is my chief analytics AI development guy.

He's really pretty brilliant but he is going to. Provide us a 15 minute education every week on a piece of capability. And I don't care where he sources it, he can source it from TikTok for all I care. But everybody's gonna watch it every week. And I'm just gonna start infiltrating this message of knowledge because you gotta, and you know, I mean, literally we go to sleep [00:15:00] tonight, we'll wake up tomorrow and there's something new.

And so, one that's a really low tech to really start to push that on the more formal side of what does upskilling mean. I think that's really a fancy word for something US technologists have always thought about, which is how do I remain relevant and highly employable in a market, right? I mean.

You know, I, hell, I've been doing this for 32 years. I got some great V skills bill and you wanna hook me up? I can, you know, c plus and I can do anything on a vax, but you know, for some reason I decided that maybe I needed to learn something a little bit more contemporary and I needed to broaden my skills.

I think this is just another chapter of that, maybe on steroids, but we certainly have more resources to learn than. Than we did 20 years ago.

Bill Russell: it's gonna be interesting. I'm gonna keep delving into that topic as we move forward and what does it look like to introduce the team to the tools that

Chris Paravate: keep chewing [00:16:00] at that because I'm stru, you know, I'm not gonna send somebody to get a master's in, AI development.

I don't need 'em to know development. I really need to know how do I put these tools to work?

Bill Russell: I was looking through the reason I'm, these are so relevant, so new, this feels like I'm hitting these things is I went through our notes for the 229 city tours today. Like, we just, Drex and I were sitting down, we were like what were the top topics that were talked about?

AI obviously was number one. Number two was money, as you would imagine. Number three is the one that surprises me. it's Governance and it's this whole idea of prioritization, governance and prioritization and demand, outstripping capacity. We talked about this a little bit earlier.

I mean, how do you stay in front of this?

Chris Paravate: It's a constant battle, right? So first of all and I didn't appreciate this till I had some other conversations and we were going through our digital most wired level 10 certification. And really reflecting on our [00:17:00] governance. Our governance right now is our C-suite.

They are prioritizing our IT demand. And you know, once upon a time we used to get project requests and we'd kind of look at 'em and go, yeah, whatever. Right? It's not a good request or it's not a priority. We get really good requests now. There's no bad requests, right? It becomes what is the next most important thing for the organization.

And, the whole, you said the second thing it comes up is budget. It's like, well, what's really important for the organization and the budget don't always, they're not always congruent. Right. And and that's where the rub really comes down to is I could add 50 60 FTEs to my IT team tomorrow.

And We could maybe, you know, increase our output by 20%. And I'm not sure that we would satisfy that demand. People would feel like it was 20% better. There would be more demand. And I [00:18:00] also, I wonder at some point, I've always worried about outstripping operations ability to absorb all that change.

Right. You know, that used to be a really big thing at the top of my mind. It doesn't, I mean, we just keep cranking out more and it doesn't seem that we've hit that, that you know, nursing, particularly when I worked at another hospital system, they would be like, stop. No more. Uh, I don't get that anymore.

So I, you know, I think demand will continue to be. a focus. We've been very fortunate. We don't have a huge technical debt. I've been really deliberate over my 11 years of really managing that, really progressively. We're moving our production all production systems to the cloud actually this weekend.

That will avoid a big capital investment. You know, but that's the thing I, you know, so here's the AI again, right? How do these three things play together? Well, if I can pull $4 million. Outta revenue cycle in waste, can I take [00:19:00] 2 million of that and put it towards additional innovation?

We haven't gotten that crisp yet. We got work to do.

Bill Russell: In this time where there are financial pressures and we're wondering about Medicaid cuts and all those kinds of things. what does the innovation budget look like? I mean, can you we used to talk in terms of placing bets.

It feels to me like we don't place bets anymore. We only invest in really sure things at this point.

Chris Paravate: Well, you know, I'm testing those edges, right? I'm looking at a, another piece of AI technology on our revenue cycle side. And it's not in budget and it's not planned. But I think it has a pretty strong RROI and I'm putting that together and frankly, I'm gonna sit down with our CEO and our CFO and say.

Hey, do you wanna place a bet? It's not in budget, but if we yield the outcome, it'll actually help the budget, right? , It will drive more profitability. [00:20:00] So, you know, are you interested that, and think about on the revenue cycle, and why is that so attractive?

Let's say I save a million dollars in revenue cycle. What does that represent in gross charges? I mean. It's gotta be five or $6 million in gross charges, right? How many encounters is that? How many visits is that? So you don't wanna talk about materially changing the bottom line. It's incredibly impactful.

So I'll report back to you in about three months of what's our desire to place bets. I will tell you that our leadership we're getting a lot of attention. I did a physician retreat just before UGM our medical staff leadership retreat. I did a two hour presentation on ai.

The physicians were on the edge of their seat. They wanted more. And I could have spent a whole day on it and, so a lot of people are really [00:21:00] seeing the way that we could potentially make some big changes. And we're doing stuff, a lot of stuff on the medical. You know, an example, I didn't mention earlier, but we have a high propensity for lung cancer in Northeast Georgia. And it's not because there's a huge smoking population. So we, we've implemented an incidental findings module that let's say, you know, someone like you or I get a go and get a calcium CT score well now I got a CT of your whole chest and I can move it through an AI algorithm and if there is a pulmonary finding, I can route you to the right person.

And that intervention could be addressed, probably and identified earlier and faster. And so we're seeing some big outcomes. I mean, that's. That's the stuff, the medical stack you're really excited about.

Bill Russell: Yeah, I can see that. This may not be the way we think anymore, but is there a technology out there that you are looking at and keeping an eye on 'cause you feel like it's gonna have a significant [00:22:00] impact?

Chris Paravate: know what I watch, I always watch the money. I mean, when you kind of pull back and you look at, you know, some of the technology companies, and where the dollars and the investments are being made that often is the leading indicator for me of a technology that I need to watch. And there are some big bets being put out there right now.

An early bet that we made over two years ago, it was on Snowflake and that has you know, really paid strong dividends. That's One that we watched and quickly jumped on early. And that's been a strong platform for us as well.

Bill Russell: We've had two mergers this week.

Two tech companies. So you had Press Ganey and Qualtrics come together. You had HealthLink advisors go with who'd they go with? Anyway, they went with another Chartis, I think.

Chris Paravate: Chartis, yeah. Right.

Bill Russell: Is that a something that you keep in the back of your mind, like, will this company get acquired?

And is that part of the buying calculus for [00:23:00] you?

Chris Paravate: The short answer is yes. 'cause we've seen so many acquisitions that, you know. Products have been, you know, several products have been aggregated and then all of a sudden there's a change in leadership, a change in direction, and that product went from a, you know, an up and coming, leading to a bust.

Right? And so, you know, I've been very intentional with the architecture here. We have 408 applications, so we're, $3 billion organization, six hospitals, 13,000 employees, 408 applications. So that's pretty dang good number. Pretty low, comparatively speaking. When I look at benchmarks, probably more like 600 to 800.

And so we do pick partners very deliberately and we look for platform. Strategies along the way. And then, you know, I think there is, there's transitional solutions out there that you may say, Hey, you know what, that will buy me [00:24:00] two years and we'll see where the market matures. And I'll either get off that system and move to something else.

And I think these are, you know, these are pretty durable strategies I think a lot of CIOs are thinking about that. We have the benefit. Of being very centralized. And, all technology, for example all software is budgeted in it for the entire system. It's all under our cost center.

All the contracting is under our cost center. We manage that very deliberately so that we can get those legacy systems out.

Bill Russell: I mean a lot of people talk about that. It's like all the budget money's in it. So when the organization wants to spend more money, there's really no transparency or accountability to them.

And so you end up like, Hey, we need new this. Let's purchase more licenses, is whatnot, 'cause it's not hitting their budget per se. Is that something you guys have gotten around?

Chris Paravate: Well, it's a balancing act, right? So you gotta have it with governance, right? Part of our governance discussion, the C-suite is, Hey, you [00:25:00] want that piece of software?

It's not in the budget, and here's what it's gonna take to support it. And here's what the operational implication is. So when we go to budget season next year, the first $352 million of any investment, any incremental investment we can make, you're taking that right off the top. And that, that gives some pause and there's a little bit more peer accountability there to say, well wait a second do we really wanna spend that here?

What's the value? Is this something that's just keep the lights on, right? I mean, you know, there's some things gotta have. But I'll give you an example. You know, we haven't deployed Microsoft copilot. You say, well, why haven't you? Well, it's an incremental cost to my licensing structure, and I can't quantify the ROI on rolling out copilot to my user community.

Now, lots of people want it, but [00:26:00] I can't produce an ROI on that. And so the executive team has said, no, we're not gonna roll out copilot. Broadly, we'll use it in a very specific way. Now that's a tough decision, right? But that's kind of where we're at Now, I'll challenge, you know, that I think we will over time will become com commoditized, right?

Bill Russell: That one comes up a fair amount when people are saying, who stands to gain the most from copilot being rolled out across the health system? I mean, is it the email tools? Is it the and what do they do without it? Like, I go to Claude and Chat GPT about five times an hour.

Do they have an alternative there?

Chris Paravate: I mean, yes. And that's the thing is you can use g you know, and chat GPT. You can use Claude, you can lose lots of others. You don't have the full integration, you don't have capturing meeting minutes. There's some things we're not gonna allow you to do. We're gonna but [00:27:00] we're pretty liberal in access.

You know, so if I need a, draft a job description, you know, I'm gonna go to one of those publicly available tools and source that, or a policy or, you know, maybe even a thought process. So, you know, and I think then you get into that same comment you made, which is, you know, six and 10 are using a tool and that's in it well.

I'm gonna pay full price for a copilot and I'm gonna get 40% utilization or 30% utilization and what planet, does that make sense?

Bill Russell: Right. Right. Man, it's right there. It's just a little checkbox on your Microsoft.

Chris Paravate: Oh, it's so easy to add that to your ea. It's, I'm sure I'll have lots of incentives.

Bill Russell: They'll figure it out over time.

It'll be interesting. Hey, Chris, it's always great to catch up with you. I'm looking forward to seeing you in a couple weeks.

Chris Paravate: Excellent. Look forward to seeing you, bill.

I.

Bill Russell: Thanks for listening to the 229 podcast. The best conversations don't end [00:28:00] when the event does. They continue here with our community of healthcare leaders. Join us by subscribing at this week health.com/subscribe.

If you have a conversation, that's too good not to share. Reach out. Also, check out our events on the 229 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.