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

[00:00:00] This episode is brought to you by Dr. First. Their AI powered solutions improve the accuracy of medication history, enhancing patient safety and reducing clinician administrative burdens.

From e prescribing to medication adherence, Doctor First Solutions integrates seamlessly with your EHR, informing clinical decisions and improving care. Visit thisweekhealth. com slash doctorfirst today to transform your medication management with Doctor First's innovative tools.

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.

Now, onto our interview

(Interview 1) All right, here we are from HIMSS 2025 out here in Vegas. And today I'm joined by Dr. Colin Banas with Dr. First and Dr. Tom Wells with Piedmont and a welcome gentlemen. First day of the conference. Here we are out in [00:01:00] Las Vegas. It's cold here. Just like it was cold at Vive and it's cold and it's windy. It is very windy. Let's get right into it. Tom, I'd love for you to tell us a little bit about

your role.

I'm a general internist by training. I went to the Medical College of Georgia. Did my residency there as well and have been in Athens practicing medicine for 28 years now. And we were a privately owned group until about 16, until 2013 and we sold our practice.

Cause we saw the foreshadowing of value based contracting and more of the government programs and realized even as a 10 person group we didn't have the resources to. figure out how we were going to manage that and joined the local health system. At that same time, they were looking for a medical director for their insurance plan.

They had a small health plan that did third party administration of their, of them, as well as a couple of other hospitals right around and had some commercial and I got into that and got captured by the world of care management and population health and have tried to figure out how to leverage that medical knowledge with [00:02:00] that.

Trying to figure out how to take care of people on a larger scale using that knowledge and since I've been at Piedmont We joined Piedmont in I've been fortunate enough to help oversee their employee health plan from a medical supervision standpoint well as looking at their population health and recently have been the chair of their clinical governance council, which is made up of primary care from across the system And we help try to drive quality and safety for our primary care network for the Clinically Integrated Network.

So just to give people an idea of the scale, so

Piedmont number of associates in that plan?

Well over 20 hospitals including acute care, rehab, and in the employed network there's well over Specialties across the state, the clinically integrated network is may have both employed and independent and is close to 3700 providers.

So we're at HIMSS. So we're going to talk about some of the challenges facing healthcare. I want to bring you into this What do you expect to hear while you're here this week? What are some of the challenges that health [00:03:00] systems are facing?

Okay.

I think it's a lot of the same year over year with just incremental progress. It's easy to get disheartened when you I just left an interoperability talk. And I was like, yeah, we're still talking about interoperability.

Interoperability, really?

Yeah, it was actually a really good talk.

But there's going to be a lot of AI. And I think the trick is going to be figuring out like what is actual real AI and what is just like the buzzword AI, just like. buzzword interoperability has become. In fact, , one of the quotes upstairs was when a word becomes the solution to everything, it means it's lost all of its meaning, altogether.

I do expect to see incremental progress. I do expect to see real world use cases of some of these technologies in use with real impact on patient outcomes. And that's actually what I get excited for is when somebody can show me the patient outcome or the provider outcome. So it's it's travel season that we're finally winding down.

But this one I'm excited for. So I got my fingers crossed,

Interoperability is an interesting topic, especially you mentioned clinically integrated. Yeah.

Interoperability for us is still something we're working on and I [00:04:00] agree with you. It means a lot of things. For me, it means trying to drive all the results of our entire network together into one.

Manageable, platform.

You really can't do a clinically integrated network without

So we're working on a better solution. We have, we get the data, but really to drive it in a meaningful way. One of the ways we started working was four years ago the primary care network portion of it is we mandated that you had to be on one of three EHRs. To stay in the network.

I hate to ask the question, how many was there before that? Twenty something. Oh my goodness. Yeah, so that's challenging. interesting because I've heard people say, our strategy is we're going to get everybody on Epic, but that's not, when you're talking about the whole state of Georgia that's not realistic, is it?

hard, because some of these people are entrenched in their EHR, so we looked at ones that we thought we could better pull the data out of that. Also, we looked at which ones most of the network was on and then tried to narrow it down [00:05:00] to that. What are some of the other thing? For me, I think AI, how do we leverage it?

As you know that, we're in a provider shortage and we're a booming population, especially in the state of Georgia. And so how do we make the physicians more efficient? And , how do you make the staff more efficient? And what is the right AI and what's not the right AI? Because it has to fit, especially in the state of Georgia, from downtown Atlanta to the executives living in Atlanta to the patients, some of the patients I service who are farmers out in Madison County who, may not have high speed internet.

You really do cover everything. When I think of Northeast, Southeast,

We go all the way. We have a network in Columbus. We have Macon. We have Augusta. Pretty much from Macon to Columbus to Augusta, that line and up. We have a network in a few clinics, ambulatory clinics below that.

Dr. First, you guys doing anything with AI?

Yeah,

There's a couple things. Now,

everybody's [00:06:00] talking about it. I'm going to hold you to, How is AI being applied to med rec and the things that you guys do?

Yeah one of the things that we've been doing for a decade it's actually, it has four different patents on it, is around medication data and the normalization of that data.

And I think you've heard me talk about this a little bit before, but So this is actually not LLM agentic AI. This is actually something that was homegrown a decade ago, and I like to call it a very narrowly focused clinical grade AI, meaning it was created by clinicians for a very specific clinical purpose, and that is to achieve that semantic interoperability with medication data.

Tom and I both know we, we do a good job usually writing our scripts perfectly structured and you send them out into the ether. And if you're

talking about doctors right now, they usually do a good job. It's usually readable. It's usually,

yeah. He and I come from when it wasn't, definitely did paper.

But one of the dirty secrets around medication data and the medication management networks that we have is that even though we've sent the data out, decently structured. When we bring it back in from a [00:07:00] variety of different places, it has lost its structure. It reverts to free text. And then you lose all of the positive things that you can do with it.

The drug interaction checking, the a seamless med rec where the stuff can flow. And so this particular AI, which again we've been doing for 10 years, is around renormalizing anything that has lost its structure along the way. even on very basic use cases, we found that we can make a pharmacy technician, 25 to 30 percent more efficient, they can see more patients, or you can free up we were talking about Epic a second ago, you can free up a willow analyst from curating a very specific manual translation table to go do something clinically meaningful.

And so that's just a really small use case, not trying to boil the ocean, but, solve a very specific problem that, that doctor first saw over a decade ago. And then as a company, of course, we're trying to embrace AI in productivity, the agents, the transcriptions, the things like that.

want to come back to, you're talking about physician shortages, nurse. Technicians [00:08:00] just in general, front staff to the back to the guy right in the prescription, everybody. We're looking at shortages. I'm hearing more and more people not looking for the LLM that knows everything, but more AI that is very narrowly trained on a specific task.

Therefore, you don't have as many errors or those kinds of things going on. I'm curious, is there a specific area that you think AI will advance more quickly than others to add value to the clinician?

I think we are starting to use some AI scribes, right? And one of the things that we continue to add value as I'm learning how to use my AI scribe is in , suggesting specific diagnosis to help with.

Making sure that we're correctly, accurately coding the patients as well as the formulation of how do you order the appropriate tests. If it hears X, it suggests that these tests be ordered. And if you're talking through it with a patient, then it tees it up for you to order. So more and more the [00:09:00] AI is going to be an assistant.

Yes. It's going to be that person that sits there with you. And then learns from you because it learns your preferences and learns your patterns.

And that's interesting. That's going to be an exciting future as we look at it. And one of the things about ambient listening more and more it's becoming common.

Yeah. The table stakes, right? Yeah. If you were a physician going to a new organization and they said yeah, we don't, we haven't invested in ambient listening yet. You'd

sort of

look at them and go,

well.

so I'll tell you, as a regional medical director for primary care, I have the responsibility of interviewing new docs for the Athens, Gwinnett, that region.

That's a question that comes up. What are you doing? leveraging the technology? What are you doing with AI scribes? Those are questions that these folks coming out of residency programs are asking. Out of

residency?

I believe it.

It's like the ATM machines of old, right? The first bank that had an ATM was probably a great differentiator.

Past four or five years if you didn't have an ATM, you were not a bank anymore. It was an expectation.

Yeah.

What are some key initiatives

you're looking at in the year coming up, or?

We're certainly looking at AI and [00:10:00] improving that. I think one of the things is leveraging telemedicine.

and telehealth to yeah,

You cover the entire state.

It's not only that, it's allowing us to have dedicated virtualists and hiring dedicated virtualists in our primary, in our employee primary care network and allowing that coverage transition of care from hospital if they're not able again to see their primary care doc right away.

Yeah, I think one of the things otherwise is behavioral health. We have a true mental health crisis, as we all know, and being able to leverage telemedicine. That I'm not stuck with trying just to leverage the behavioral health in my area. I can pull in from other areas so that where there's maybe some excess care, caregivers here.

That we can use it in different parts of the state. In developing a network that's willing to have that give and take with us.

Heard that coming up at five for the first time of people starting to use the televisions in the room as a true [00:11:00] device to they put a camera there and whatever. And then you have The specialists are able to just get around a lot quicker.

Oh, as far as that, we do have tele neurology for some of our, for our smaller hospitals.

That they're able to come in. And we started a tele nursing program where we're using centrally based RNs to watch over and help with some of the management in some of our hospitals where there may be some staffing.

want to close with this question. I'd love for the two of you to weigh in on it, which is you talked about incremental change.

We're seeing incremental change over time. What would need to change to have a significant move forward in terms of the quality of health we can provide or access to care or just any of the major problems

I'll go back to interoperability and actually Lacey Knight and I used to talk about this a lot.

In order, I think interoperability is largely solvable. It's that the incentives aren't quite aligned where they need to be. If, and so it's been this carrot and stick for the past 12 years, meaningful use and promoting [00:12:00] interoperability, whatever you want to call the programs, and now TEFCA, and then the administration's changed, and we can go forward and come backward.

If the incentives were aligned, and if there was, appropriate financial and clinical and you would think that those things are all there. They're just not. Cause if they were, this stuff would be happening already. Just like it happens in banking and aviation and retail. And so, I do think, to get that big leap instead of these incremental steps and there's nothing wrong with, crawl, walk, run, but I do think incentives is probably the key.

I

think part of it is getting

past

I just want to start you off at the conference with a really difficult question.

No, I've got, I think, a decent answer for this. I think it's getting past the bias and misunderstanding that AI technology makes medicine impersonal. I think it makes it more personable.

I think my AI scribe allows me to sit in my room seeing the patient. And talk to them and explain what we're going to do and not be worried about typing on the keyboard and not be worried about looking at the computer. We're able to have a discussion and get back to practicing medicine the way I wanted to practice [00:13:00] medicine when I came out 28 years ago.

And having the ability in the office to show the images and talk about radiology and look at lab trends with the patients. Is all making using technology make it more personable. And some of it is a story we're here to tell this week of using technology. predictive analytics and looking at the different technology overlays into your system to be able to find those most at risk and who are the ones that really need a more personable touch and who doesn't and allowing us to be able to give that human touch to those people.

So I lied. I said that was the last question. I'm just curious, and this is just a goofy question to end with.

Growing up, who was like the doctor on TV or in a movie that you related to? You're like, yeah that's what

practicing medicine is going to be like. Oh so I was a huge Doogie Howser fan as a kid. And I think he was my age, or is my age, he's still my age, I'm not sure that's the right answer. I will say that I found that the most clinically accurate television show, I believe, is Scrubs. [00:14:00] On NBC. I swear to God. Really? I do. It's comedic, but And yeah I fancied myself a Dr. Cox. Do you remember Dr. Cox? Yes,

the relationship between the intern and the

Intern resident, and then also as we were talking before we started Medicine versus surgery, plays out very much so on Scrubs, that's interesting. How about you? Hawkeye

Pierce from M. A. S. H. I thought he made medicine personable to people. Yeah. And carried it out. I would have to say, wasn't TV or movie, but the, probably one of the characters in fiction that most influenced me with medicine was in the Rules of the House of God, the resident, the fat man.

The one who was Oh, it's a,

it's a famous book that we all must read as internists. Creighton, right? Yeah.

I can't remember exactly, but it was, and the fat man was the resident that everybody went to ask the questions and kept his calm and never got flustered. That was my goal was to

Yeah, I love that question.

I just learned it. Yeah, I might steal that from next

time I'm doing [00:15:00] something.

Gentlemen, I want to thank you for your time here. Really appreciate it. And thanks for listening. That's all for now.

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 would be great, thanks for listening. That's all for now.