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

Newsday: Application Rationalization, Securing Big Wins, and Data Pitfalls with Dave Dyell

to serve as a true partner to clients by providing healthcare's most agile IT solutions. Visit ThisWeekHealth. com slash innovative consulting group. That's innovative dash consulting dash group to learn how innovative can help you stay ahead

Bill Russell: Today on Newsday.

Dave Dyell: when the question goes back, what AI do you want to do? And what problems are you trying to solve? They don't necessarily know how to articulate it, right? They just know that, they're seeing the hospital down the street, advertise how they're using AI.

They want to remain competitive in their market without necessarily having an individual problem they want to solve. 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. Newstay discusses the breaking news in healthcare with industry experts

Bill Russell: Now, let's jump right in.

All right. It's Newsday. And today we're joined by Dave Dyell with Innovative Consulting Group Managing Partner. Dave, welcome to the show.

Dave Dyell: Thank you, Bill.

Bill Russell: First time on the show. And the first time anyone from Innovative Consulting Group has been on the show, great name, by the way, tell us a little bit about Innovative Consulting Group.

Dave Dyell: Sure. So we're a little over 20 years old now. So the firm's been around for a long time. All of it dedicated to healthcare IT. We like to think we offer, the industry's most flexible approach to healthcare IT consulting. We specialize in everything from, strategic solutions that you would find like disaster recovery planning, infrastructure reviews.

Cloud migration, application rationalization to managed services, as well as, into traditional professional services to help, CIOs get the projects across the finish line. We have a special, program around project management. It's one of our, best managed services that we offer.

So if they want to outsource their entire PMO or, have us come in and, augment it, we can do that as well. But we like to think that our agility. is one of the things that really differentiates us and allows us to really truly partner with our clients to help really solve true business problems.

Bill Russell: we're going to dive into that. I'm curious the PMO stuff's interesting. I'm curious on the AppRat stuff. Because I've talked to a couple of people over the last two weeks that have been talking to me about. And they're like, we're not getting enough juice out of this program.

There was this belief that we were going to get more out of it than we're actually getting out of it. Are you seeing that with clients or is there something? that people are doing that they're, they are able to get more out of those application rationalization projects.

Dave Dyell: So when we looked at building a program around this, one of the things that was key for us is understanding that, clients can't wait to see value, right?

If you look at some of the old traditional applications You'd have firms that would come in and they do a 12, 18, sometimes 24 month study to go through and identify all these areas where they felt like there could be consolidation or reduction, right? They produce this massive, 200 page report, and they dump it on a CIO's desk, and what are they supposed to do with that, right?

Now they have to try to figure out how to take that information. Large group of paper and sort through it and figure out what is valuable, where the returns could be, to then try to go execute. So I think when you hear what you're hearing, what you're getting a reaction to is they're just dumped this report that they don't know what to do with.

Our methodology is a little bit different. We come in a four to six week type engagement, and we actually turn around a very rapid, deliverable that actually includes a truly prioritized list of applications. And part of the ways we can do that is that we have an algorithm. that we've designed to take these inventories and help identify that.

So it's a little bit more of a tech enabled version of application rationalization than I think what you're seeing out there in a traditional consulting approach. And then, as I mentioned in my overview of the company, we don't just leave clients by themselves then. We try to do something with that.

We can then go into our phase two where we help them grab, that really low hanging fruit right at the top that provides that biggest bang for their buck, so they can immediately see an improvement to their bottom line. It's sad, unfortunately, in a lot of, especially our community hospitals that are struggling to stay out of the red and get back up into the black.

And, if you look at something like application rationalization, there is a ton of waste across application portfolios that can provide value. But when you just dump it on the CIO's desk as one big report after an 18 month waiting period, by then the organization's moved on, priorities have shifted and trying to get those projects into the limelight to take on just doesn't happen.

Bill Russell: Yeah it's a really interesting problem because there's a technology aspect of it, so it gets dumped on the IT, it gets dumped on the CIO's desk, like you said, and that's absolutely true, but there's an organizational change management effort to this.

There's so much of it that's not technology per se, but it's leadership and it's culture and it's a bunch of things. I remember with radiology, I was able to go. the radiologists in a room, say, we're going to consolidate. They all looked at me and said, yep, two thumbs up.

We're going to be able to share our images across. This is great. Let's do this. And we did it. And I thought, man, that was easy. So I went to the radiologists and I said, hey, we just did this with with cardiology. It's pretty simple. It's pretty straightforward. You just, whatever.

And they just every wall possible went up and we never got that project done. It's definitely not a technology problem. It's a leadership challenge.

Dave Dyell: It can be. The organizational change management is absolutely an issue.

We run into this as well. I think helping those different departments understand that while yes, it is nice that you have your little tiny point solution, this other solution that we have at the enterprise level provides, 95 percent of what you need. So with some small workflow adjustment, the organization can save this much cost, right?

And you still are going to get again, 95, 96 percent of what you need out of this other application. The 4 percent is not worth the extra, costs, right? To the organization to allow you to have your own individual, application, especially when you consider the integration costs and the maintenance cost of that integration and everything else across the board, right?

So you're absolutely right. There's definitely a change management component to this for sure.

Bill Russell: All right. So we'll get to the news. We're taking a look at a Forbes article. What to consider before implementing AI into your healthcare organization. I just finished the 229 project meeting in DC.

With 13 CIOs, and we talked a lot about ai. You're working with a lot of clients around this. So this is a good article. It sets it up. It has , 10 items to consider. And some of these things are things I heard in the meeting. One is to identify the need for ai, right?

So right now there's a lot of talk about ai. AI it's showing up in existing applications. It's like you just do an update and all of a sudden it's there and hasn't gone through governance and anything. It's just showing up. Other organizations are coming in and saying, Hey, we got to get our AI message in front of people.

But what I'm hearing from CIOs is. Wait a minute. We don't approach technology this way anymore. We're trying to, we're trying to solve problems. So stop talking to me about the technology. It's let's talk about the problem set and see if AI can be applied to that. finding people are starting to coalesce away from the shiny object and more towards.

Okay let's start looking at the problem sets.

Dave Dyell: So it's interesting, right? We are hearing from our clients some of the same things you heard, right? Which is that pressure from the C suite, and even sometimes from the board level, that we gotta be doing AI. We have to be doing AI, right?

Everybody wants to put that pressure. on the IT folks to do AI. But again, when the question goes back, what AI do you want to do? And what problems are you trying to solve? They don't necessarily know how to articulate it, right? They just know that, they're seeing the hospital down the street, advertise how they're using AI.

They want to remain competitive in their market.

So they feel like they need to have AI as an answer to that competitive nature. Within the market, without necessarily having an individual problem they want to solve. And you're correct, IT's job is to find solutions to problems, right? And just like when I've worked with developers over the years, there's nothing a developer hates worse than when you walk in and tell them how to code something, right?

They want you to come to them with the business problem that you're trying to solve and let them figure out how to make the code that they want to solve. Solve the problem, right? It's the same thing that CIOs want to do, right? They want to hear what the organization's trying to accomplish, what the strategic initiative is, what the business problem is we're trying to solve, and then allow us to source and find the technology that can solve it.

'cause it may be back to our app, right? Conversation in our existing portfolio and we might be able to do it what we have already today without adding anything else to the organization. Or again, it could be a workflow issue there, there are other ways to solve the problem without necessarily running out and grabbing another, AI vendor.

Bill Russell: the second one is quality and source of the data. And it's this was a fairly robust conversation that we had with the CIOs. It just in terms of being concerned about bias, not because you're feeding it bad information, although there is that concern as well, but, one of the CIOs was making the case that, my population is the Southeast or my population is, West Texas.

And if you're going to feed it with information from people who are from Southern California, And then make determinations based on that for this population, that becomes a problem. there's this race towards aggregating. good quality data across the board. And so you have EPICS COSMOS, you have oh gosh, the one that all the health systems came together.

I'm blanking on the name right now, but there's this EPICS, COSMOS, and then there's the one that Providence and a whole bunch of the healthcare systems came together and formed. Anyway. I'll remember it sometime. I'm going to leave this in here so people can recognize that I forget things from time to time and it just happens.

but even that, as we were talking about that, it's important to understand the regional differences between the data sets and whatnot. There's a concern about this whole idea of what are we training these systems on and where are they getting the data and what's the quality of that data?

Dave Dyell: Yeah, so this is the one that's near and dear to my heart and is the one I was hoping we were going to spend, some good quality time digging into today. I'm an old integration guy. I started my healthcare career in the hospital side. I was the Cloverleaf interface guy eventually went to work in consulting, then to the vendor, actually worked for HIE, back in the late nineties who manufactured that interface engine.

And so I've lived through, the best of breed ages into now the big, vendor ages and, integration is not solved yet, right? I mentioned Tefka a minute ago that a lot of the organizations are joining now because the government's finally regulating. That interoperability has to occur within our industry.

It was done out of the University of Pennsylvania. I think it was called the Medical Diagnostic Accuracy Study, and they essentially took a bunch of de identified patient charts from very, chronically ill patients with multiple comorbidities. They sent it out to a whole group of physicians trying to determine whether or not, physicians that have been trained 40 years ago, 30 years ago, 20 years and fresh out of school, how would they do from a diagnostic perspective on those charts?

And surprisingly, they got pretty much Accurate diagnosis is across every generation of physician, no matter when they've been trained across the different medical schools here in the U. S. Great for our medical schools, right? But then they wanted to test the study. So they took the exact, different set of de identified charts, but the same conditions, and they actually started taking out pieces of the data.

And removing just arbitrarily different information about patient history, medications, things like that, and send it back out to the same group of physicians and got back completely different results. Inaccurate diagnoses, sometimes treatment plans that would have resulted in harm to the patient, and so it shows you that if the data is not accurate, The model's not going to be built right.

If our human beings that are training the models who are going to then, make mistakes because there's missing pieces of data, then that's exactly what's going to happen. And you see it in education, right? My wife's a first grade teacher. If you're teaching children, new vocabulary, right?

We all remember the vocabulary tests we used to get every week. If you stop adding new things and new information, to that child's learning, they're going to stop learning. And so these models need to have new and updated and accurate information. I myself, I'm in my 50s, I have three physicians.

They all use three completely different EMRs. None of them share data. None of them are connected. So any AI model, right? plus, I don't, thankfully, go to the hospital very often, right? I don't think I've been to a hospital in over 15 years here locally. So if I did walk into that hospital, they don't have any of my data.

Anything they have is 15 years old. So if they have an AI model that's trying to do anything clinical on me, I'm not trusting that model, right? Not today. And yes, I understand that they can look at historical data. They can use that historical data to help predict the future. I do think there's some great use cases out there that I think CIOs can help their organizations improve care.

I think some of the things the vendors are doing today are phenomenal around AI, especially you mentioned, imaging. Imaging is a great place, right? I'll tell a personal story. I've got a daughter who, you know, when she was an infant kept having urinary tract infections. We couldn't figure out why, right?

After three hospitalizations, we end up, in a specialist and they're looking through, old imaging records and sure enough, this specialist sees on the image a little pinch in her urinary tract that was creating reverse vacuum and causing the UTIs. One little simple procedure done, on an outpatient basis and she's never had a problem since.

AI would have probably caught that, but the radiologist who read it the first time didn't see it. So I do think there are particular use cases where you absolutely can have AI help, but until we solve this issue of interoperability and truly get patient charts in one place where we have a complete view of that patient, would have a hard time.

And I think that's where, from a CEO perspective, they should be insisting that there's good governance policies put in place in their organizations for how AI is going to be used for the ethical use of that AI. And then how that AI is going to be leveraged to actually impact patient care.

Hey everyone, Drex DeFord here, and we have an exciting webinar on October 22nd at 1 p. m. Eastern. It's sponsored by CrowdStrike and AWS. We're diving into building a resilient healthcare system, cloud security strategies for today. With cloud breaches up 75 percent over the last year, healthcare systems can't afford to rely on outdated defenses.

So join us as industry experts share practical strategies to strengthen your cloud security posture and adopt zero trust and boost operational resilience. Don't miss it. Register now at thisweekhelp. com slash cloud security. That's thisweekhelp. com slash cloud dash security.

Bill Russell: You have a strong background in integration. And my time in healthcare overlaps with yours. we on the right path? It's slow. I love the stuff that Mickey Trapathy's doing and before him. It's been pretty consistent that what they're doing.

But when you have to do it at the federal level, It moves really slow because you have to, get everyone to buy in, you have to give them a period of time to comment and those kinds of things, you generally go like one data set at a time. The US CDI keeps growing, which is great, but it's, it has taken a fairly long time.

What's going to accelerate us getting good quality data, a longitudinal patient record at the point of care whenever somebody presents?

Dave Dyell: Regulation. I've been doing this, like I said, for 30 years now. Nothing in this industry happens unless the government regulates them to do it, right?

If you think back to, any aspect, right? Think of EMRs, right? Look at the EMRs. For a long time, we were still on paper. A large percentage of this country was still on paper until, somebody passed a piece of regulation that said you have to now, get rid of paper. We used to file claims by paper until the Administrative Simplification Act was passed, and then all of a sudden everybody had to endorse X12 and had to do it.

I would love to say that similar to how, in the late 80s, just the whole, just groundswell, of desire from the hospitals was enough to force the market to do something. And that's how HL7 happened, right? HL7 was a truly, down to earth program that was started by, a group of just really, smart forward thinking individuals who decided to create a standard, right?

And I had the great pleasure of getting to know Dr. Clem McDonald, who's one of the founders of HL7, when I did some work with his research institute up in, Indiana. And it's like when you start to, see that groundswell, it can happen. But I haven't seen that in years. The only way I've seen us move forward has been through government regulation.

And I know nobody wants to hear that. I know everybody hates that. But unfortunately in this industry, things don't move until the government forces this or, this industry to do

Bill Russell: let's close out on this topic. So ROI. of AI projects. Where are we going to, in your estimation, where are we going to find ROI on these AI projects?

So right now

Dave Dyell: what we're seeing is candidly around patient access and revenue cycle. There are a lot of manual tasks that are completed by human beings in both of those processes today that AI can absolutely replace. It is a generational thing. My 87 year old mother is not going to use a chatbot to register, right?

So that's definitely where, AI can, add a lot of value. But if you look at my kids, anybody in their generation, they're absolutely happy to use a chatbot and prefer that, right? As their primary way to interact with the organization. If they have questions about their bill, they want to pay their bill, they're more than happy to interact with a piece of AI, right?

To turn around and do those things. So I think if you're looking for low hanging fruit with solid ROI right now, I would look in patient access and I would look in the revenue cycle side of healthcare because I think that's where you can get some immediate return. on your investments and you stay away from this data problem that I was just talking about, right?

Because you're leveraging the AI to replace those manual, human tasks. And if you look outside of healthcare, the industries that are leveraging AI to the most effectiveness, especially in terms of getting back ROI, are in those situations, right? Where they're just removing those manual tasks that human beings are doing today and replacing them with, something that can be done very easily and much more efficiently.

Probably more accurately, with AI,

Bill Russell: I agree. The, revenue cycle, coding, We have ambient listening. I'd like to see ambient listening move all the way through coding because AI could do that if it can generate the note, it can do better coding. like the whole idea of access.

Anytime we can move the patients towards healthier behavior. We always say that, healthcare is only accounts for 20 percent of a person's health. Let's reach outside of that and help them to make healthier choices and those kinds of things. And I think there's messaging platforms.

There's, follow up things that can be done through AI that quite frankly are really expensive to do with teams of people. And that's why it's broken in the past, big call centers, lots of investment, that kind of stuff. So I agree with you. I think there's an awful lot of things we could do.

Patient access and around the patient as well. And then, Obviously imaging is one of those things that is right on the cusp of really everybody doing a read in the world being able to do, 20 percent more reads because they're AI assisted reads because the data quality around images, it's the image.

you're not relying on somebody else doing something. It's the image is the image. And we have that same thing with telemetry data as well. We could do some AI stuff around the telemetry data because it's just collecting my heart rate or whatever those vitals are on an ongoing basis and maybe, predict code blues and that kind of stuff ahead of time.

Some exciting stuff going on. It's a great time to be in healthcare, I believe. Dave, want to thank you for your time. I want to thank you for coming on the show. It's been great. Appreciate it. Thanks for having me, Bill.

Thanks for listening to Newstay. There's a lot happening in our industry and while Newstay covers interesting stuff, another way to stay informed is by subscribing to our daily insights email, which delivers Expertly curated health IT news straight to your inbox. Sign up at thisweekealth. com slash news.

Thanks for listening. That's all for now