Kathleen Wessel (Host): Achieving finance sustainability and managing financial performance is top of mind for every AHA member.

Hello and welcome to AHA Associates Bringing Value, a podcast from the American Hospital Association. In the series of podcasts, we speak with the AHA Associate Program business partners, check in on their healthcare initiatives and learn how they support AHA Hospital and Health system members. Join us as we explore how AI is fundamentally transforming the revenue cycle management by addressing one of its most persistent challenges, claims denial and revenue recovery. I am Kathleen Wessel, Vice President of Business Management and Operations at the AHA. And today, I am joined by Glen Pendley, President with Ailevate. Glen, welcome to the podcast.

Glen Pendley: Thank you for having me. Super excited to be here.

Host: I think we are going to talk about a pretty pressing issue for member organizations right now, and that is how do we speed up this process? How do we make revenue recognition just a super streamlined process? So, I want to dive right into this discussion. What are the most significant challenges healthcare organizations face with claims denials and revenue recovery? And how do they impact financial performance?

Glen Pendley: That's a great question. The thing that I'm not going to do is be one of those like vendors that sit there and try to explain the problem to people that have been living this nightmare forever. What's interesting is that I actually don't have a background in the healthcare industry. I have a background in software technology. And we have identified, you know, a problem that we felt that could be solved with AI.

And as I've done a ton of research in this space over the last nearly a year now of just really trying to get familiar with it, it's kind of mind-blowing to me the challenges that every single hospital, big or small, are facing when it comes to the entire revenue cycle management and, you know, in particular denials and all these other things. And based on the research and conversations that I've had, I really believe a big part of it is just a capacity problem, you know, when it comes to people, right? We have more humans being born every day, which means there's more need of going to, you know, see doctors and nurses and everything. And the challenges that come from the infrastructure of getting people seen through the process healed, and then the hospitals to get paid for the work that they're doing.

Seeing that, like, I mean, some statistics say like 40-50% of hospital staffs are just administration. It's a manual job to do this. And the more capacity goes through, more water through the dam, only so much space that can go through. So, you're just compounding a problem that's existed for a while, and it's a challenge. You know, there's definitely a better way to do it and that's why, you know, when we identified the problem, it's just how do you streamline the process itself? Like, I want to say, remove the people, allow the staff that hospitals have to do higher order work, and it really comes down to it. Because the more of a backlog of, you know, claims that are going through, the more denials that are happening, there's more issues on the front end. You know, it's just like set a compounding effect. The longer a lot of these claims are going to sit in accounts receivable, the more revenue they have to write off, the less, you know, money they have-- these hospital systems have-- to reinvest in patient care and to make the investments in the facility. So, we're just trying to streamline that process as much as possible to allow these hospitals to reinvest the money that they're owed.

Host: Yeah. I think, you hit on a couple things we'll touch on later, but this is a pacing issue where--

Glen Pendley: It is.

Host: You know, we've got imbalance of, you know, the resources the hospitals have and then what the payers, you know, are kind of leveraging. And we need to really kind of match those capabilities. And I'm really excited about what you guys are doing here. So, could you maybe tell us a little more about Ailevate?

Glen Pendley: Yeah. And just one other point, what's fascinating is the resources that, you know, these providers are hiring to do that it's a high turnover job. You know, so even if you have the ability to hire the right amount of people, turnover in any role, in any industry just slows down production. So, I do not envy people that are trying to go through this.

But to your question around Ailevate and what we're trying to do, like I mentioned, my background's more in technology and software, and I was fortunate enough to be senior executive at one of the largest cybersecurity companies in the world prior to starting this. AI is such a huge-- like it's a blanket term. There's so many different things around AI, but nowadays when people think of AI, they think of generative AI. And what's interesting is when ChatGPT came out, generative AI really started becoming a thing. I spoke to, honestly, over a hundred of our customers at my previous employer, some like Fortune 10 companies to some like small mid-market companies. And I really wanted to get a sense of what they thought, you know, generative AI was going to do to their business and concerns and fears and all this other stuff, just to get a sense of where the mindset of folks were in general, and then tried to account for that.

Universal-- it didn't matter to the industry. My last company, we had healthcare, we had, you know, banks like everything across the board. And a universal sort of concern people had was fear of the magic black box in the sky, right? It's nice you can prompt something and upload a document and have it summarized and those sort of things were nice. But when it came to like real confidential information or like proprietary stuff, people were like, "Yeah, there's no way, you know, we're going to pass our data off to that magic black box in the sky."

So at Ailevate, we really wanted to start off with, you know, a foundation of like security and compliance and really elevate the fears and concerns that I heard from so many people, you know, over the last few years. So, what we've done is we've built an agentic sort of AI platform that runs completely in a customer's environment. Like, we don't touch any health records, you know, with our revenue recovery application, we touch none of it, right? It all stays within their enclave to alleviate a lot of those sort of concerns.

And the one thing that we wanted to focus on, whenever you're building a product, you know, the success is always focusing on one problem and doing it really, really well. So when we looked at the entire revenue cycle management, I mean, it's huge, right? There's so many different pieces to it. We wanted to really identify one area that we felt could genuinely be addressed through automation using AI the way that it could be used to really expedite this sort of process. And we landed on denials management for, you know, like handling denials and all that, specifically because it's almost a straight pitch up the middle sort of thing. Like, if you do the right thing, you understand the contracts that they have with the payers. You get the denial, you have the claim, you have all of those things. And we're essentially replicating the work that these administrative people are trying to do manually, you know, which takes a lot of time. And we've identified and figured out a way to essentially do that all in a very secure and compliant fashion where we don't touch their data, we don't do any of that.

So, our goal is ultimately, at the end of the day, you know, people didn't want to learn something new. They don't want to, you know, change their workflow. Everything that we've done is to kind of run seamlessly behind the scenes and just provide results so that the staff that people have can streamline that process.

Host: Change is hard.

Glen Pendley: It is hard.

Host: Even when it's obvious that there is a better way. We battle with that on the number of fronts. You've heard on kind of the process. So, the traditional approach to claims denial is manually reviewing, you know, as we've discussed, correcting, resubmitting. It is a very kind of person labor-intensive prospect, you know, throwing more people at it isn't really going to solve this. But this whole process creates bottlenecks, inaccuracy, delaying reimbursements and strains our really already stretched staff. How do we elevate what they're doing and allow them to kind of work at the top of their game? So, how can Ailevate help members transform the revenue cycle management process by addressing very specifically claims denials and revenue recovery?

Glen Pendley: I like how you said, how can we elevate the process? I like what you did there. I don't know if you did that on purpose, but that was good. So, the interesting thing when it comes to just AI solutions in general, especially over like the last few years, people have tried to jump like on like the AI bandwagon and implement gen AI in ways to help summarize. There's some value in doing that. Like, I don't want to dismiss anything anybody else has done.

But really, like our whole mission and what we're trying to accomplish is to really provide functional AI, like identify a problem. It's really process automation at the end of the day, right? The interesting thing here, you said it's very manually intensive, the current workflow. We're not asking them to learn something else. If you're not trying to boil the ocean, you can have AI produce extremely valuable results, putting it on rails and say, "All right, you know, I want these agents to use these different models to do this." Like, "These are your tasks. Talk to this, figure this out. Get a response based on the response that you get. If it's this, then do that." If you are very focused in what you're trying to accomplish and automate, it's a perfect sort of solution.

And when it comes to denials in particular, organizations have, hospitals have all of this information. People are doing it today. So, it's not like an impossible problem. It was just more around how do we use technology to replicate the work that people are doing manually, produce a result for them in seconds versus hours, days, whatever it might be. And then, give the stretch staff the opportunity to validate the results, tweak them if they have to, because no solution is a hundred percent perfect. If anybody tells you that, they're lying. And then, make a decision. Like, you can reallocate staff that you have to do higher order problems. There's lots of work to do. So, this isn't an attempt to like cut staff and do that sort of stuff. It's just the tedious sort of things of handling denials. Let the machines do that. Have the people validate and also do, you know, higher order sort of work. And that's really what our mission is.

Host: No, perfect. But it really addresses a pretty significant problem that's existed. And, again, you know, how do you match the pacing that is happening with claims reviews and denials. We may have touch on this a little bit, but how is AI changing the approach to revenue cycle management overall, particularly around the denied claims? And what makes this different? What makes what you do different from previous technological approaches?

Glen Pendley: That's a good question, because you always have to be prepared, and you don't want to try to build a better mouse trap and just hope people like your mouse trap better. I like to build things that are unique and solve real problem and help organizations of people. And it's interesting, it's not like this is a new problem, like people have been dealing with this for so long. And you've seen evolutions of either hospitals trying to do it organically or vendors trying to do it. You know, you see the concept of automation has been in place. You know, like people have tried to automate this process in the past, but it was kind of an RPA approach like robotics, process automation.

And the problem with that, and this is another like kind of mind-blowing thing that I've learned over the last nine months, is that when you have an RPA-based sort of process, everything has to be perfect, right? Like, the data has to be perfect, the system, it's all tied together. And as long as everything is perfect, you know, the end of the road looks great. And the mind-blowing sort of thing is that every single hospital that we've, you know, interacted with and spoke to, they are all special snowflakes. Everybody does something a little bit different. They could be using their entire, like technology stack may be exactly the same, but how they use it, what they do, their processes are completely different. And it's very difficult when you're building software because you want to build something that scales to as many people as possible. It doesn't scale. If you try to build like a bespoke solution for every single potential customer you have, that just doesn't work, right? So, the solutions that have come over time, the RPA approach doesn't really scale because, again, special snowflakes, and then things always change within these organizations. So even if you get it working, it breaks pretty quickly. And when you're building more modern software, it's very difficult to build something that scales to everybody you know. And one of the benefits of the approach that we're taking and how we're doing what we're doing is we're leveraging AI in multiple different fashions.

When you see some of the vendors and approaches people have taken, and to be honest, in almost every industry, every product, like the majority of people are taking the approach of we have an existing product, we're going to throw some AI on there and allow people to make better natural language searches or like the results of some sort of query, you know, we can summarize for it. And again, there's value in that, but it's not fundamentally changing the workflow, right? It's just augmenting existing stuff that they do. We're trying to do the inverse, like we're trying to leverage AI, not to make existing workflows better, but we're trying to do the work people are expecting or I'm trying to execute.

We have accomplished our mission. If nobody ever logs into our solution, like it produces valuable output and then goes and resides in their existing tool, right? We're a black box in their organization just doing the work for them. That's a huge departure from how a lot of folks have been trying to solve the problem to date.

Host: I know you, you've referred to, you know, not having come from healthcare originally coming from outside of healthcare. There are definite benefits to that, but it seems you've learned the one golden rule here is that a hospital is a hospital is a hospital, is a unique setup each time. A lot of reasons for that.

But I'm glad Ailevate really looks at it from that perspective and really understands that. Kind of going into some of the very specific kind of use cases, you know, you've had a front row seat to how members are using AI for denials management. You know, can you share what that's looked like in practice? Like, how are they actually kind of taking the tools and resources and deploying them? And then, what should others expect when they're getting started with something like this? What kind of disruptions or surprises come up along the way?

Glen Pendley: What success looks like to us is that, you know, the EHR that they work in, you know, like the tools that already exist, we want to leverage a lot of the investments that people have made. We're not asking people again, to replace stuff, to change stuff. We're trying to work seamlessly behind the scenes, pull the information that we need to pull from the right resources dynamically, and then produce the output that we want. And the successful sort of outcomes that we're seeing is we wanted to give guidance and input recommendations on like, "Here's what you should do based on this contract, this, this, and this. You should do X, Y, and Z, right, in certain circumstances."

In other cases, we actually will recreate the document to get resubmitted for them. Like, we'll provide it for them and say, "Hey, we feel very confident and we have like a confident score." Again, we try to be as transparent as what we're doing. It's still a black box of people. Even though we don't see the data, you know, AI isn't perfect, so we want to be as transparent with that. And we'll actually provide the output for the staff to look and be like, "Oh, yep, this looks exactly right. It goes in line with what, you know, what they're saying." I'll just resubmit this to the clearinghouse of payer, whoever that they're interfacing with the time from denial to resubmittal has just dropped drastically.

And our goal for every organization, and we'll continue to get better over time, is we'll continue to iterate. Denials are still going to happen. That's at the beginning of the process. And someday, we might move further to the left. But right now, we're going to focus entirely on just the denials. If we can get to a point where, like, you're able to recognize the revenue that you're owed, like a substantial increase in that, that's money back in the pocket for these health systems that can reinvest.

The transparency thing, some really good feedback that we've gotten from folks is that like one of the things that we've really tried to do is not just, "Here's the output," you know, "And good luck, God speed. That's just the results." We try to actually explain every step, all of our different agents that are doing all of these different works, we explain exactly how we came to that answer. So, as you know, the existing staff that folks have, if they're looking through and they're like, "Oh, that's not exactly right," or like, "Oh, that name is off again," whatever it might be. They could sit there, and instead of trying to, you know, debug and figure out, like we explain exactly, "Oh, we pulled this information from this," what we've been able to find is that like, in certain circumstances, we recognize that, you know, some hospitals aren't getting paid exactly what they should based on their contract. You know what I mean? And like, there's positive like output, because we're being so transparent and declarative around the actions being taken behind the scenes in the system and the output that we're providing. So, those are some examples of benefits and positive sort of outcomes that we've been able to provide.

Host: Yeah, no, I think those are all helpful. And just the ability to have the recheck, you know, seeing exactly what's being pulled and exactly how it's being evaluated from rules applied is critical to building trust with the system and all of that. So, what key metrics should healthcare leaders be tracking to evaluate the success of their denials management strategy? And how does AI provide better visibility into some of these metrics?

Glen Pendley: At the end of the day, you know, cash is king, right? Like, are they getting more of the money that they're owed? I'm not a hospital CFO, but I would guess that that's probably the most important thing at the end of the day. But just looking at things as far as time spent in accounts receivable, amount of denials by payer, you know, getting a good sense of one payer versus another. If you're getting denied at a higher rate from one than the other, is it because on the front end of the process, could you optimize kind of the way that you're submitting them to begin with, right? It's not always, you know, things are wrong, like mistakes happen throughout the process and things are denied for legitimate reasons.

So, the idea is, again, tracking basic metrics, time accounts receivable, overall revenue received, you know, tracking that quarter after quarter, year over year, that sort of stuff. And then, you know, really kind of identifying and tracking metrics on types of denials by which payer and things like that, so you could optimize further in the process. So, there's less denials to begin with. And then, you know, kind of overall, you can measure the financial impact. Like if you build a feature in product, whatever it might be, you want to know people are using it, that they're actually bringing value. So if you're implementing, whether it's Ailevate and/or some other system or whatever you might be doing, tracking the financial impact that that investment has made in that solution to see if it's worth keeping and, you know, change is hard, but those are some of the things that I would recommend.

Host: How do you see revenue cycle management evolving over the next three to five years? I mean, what's happened in the last, you know, two years has been evolutionary. So, what about the coming years?

Glen Pendley: I would like to say kind of what we're doing is very forward leading. I haven't seen anybody else out there kind of taking the approach that we're taking, that's not going to last forever. I think more organizations, more companies are going to identify that trying to bolt on AI to existing tools and thinking that makes a material difference overall is going to fix the problem, versus taking the AI approach first to then how to solve a very specific problem.

I think more people will do that. I think revenue cycle management, like, you know, we touched on a few times, it's such a broad problem, like denials is just one aspect of it. The hospital systems, the folks that we've talked to, you know, as we've been building this up. Everybody talks about pre-authorization and like you guys looking at what we're doing, you can make a huge difference in like pre-auth and this and that and we could, but we have to nail this first, right? Like if stretched too thin, then the quality is where it needs to be. And I'd rather do one thing really well over the next three to five years, you'll see more from us, of course, but also new players in the space, touching on different areas where like functional AI can be applied to help streamline more of the process further.

You know, in the beginning, there's a lot of interesting ideas out there of the challenges people have of just like onboarding new patients. You know, like, there's so many things that, today, if you look at what people are doing to do the job, it's just monotonous, time-consuming, you know? And there's so many other things they could be doing. There's a lot of opportunity just to be like, that could be truly automated. You know, I believe we'll see more of that over the next three to five years.

Host: Yeah, I think it's definitely the way the field is moving and really any opportunities to kind of make things more efficient. Stop asking the same questions 15 different times during the admittance process. To your point, there's a lot of areas for improvement. Just figuring out what's the best application here. So, one last question. What guidance do you have for AHA members as they look to implement AI solutions?

Glen Pendley: One of the things that have become abundantly clear to me, you know, as I've gotten much more involved in this is the existing revenue cycle management process is kind of brittle and that's their lifeblood. Any solution or any tool or anything that is asking to like rip and replace anything, be wary because it's such a huge change and you're introducing so much risk to collecting the money that is owed to begin with. So, there are ways to apply AI and newer technology out there without during the baby out with the bath water. There's ways to do it and I think we've proven that. So, that'd be the first.

And then, second is any sort of vendor, especially kind of newer vendors. That are trying to boil the ocean and do too much. And this isn't just for hospitals, like I would say this in any industry, any software. I've, you know, sat on different boards at software companies. I've invested in them. And like, the ones that do better try to, you know, nail a specific problem. And as they get that expanding over time, you know, bigger company has more resources to solve more problems. But newer vendors that are promising. And we could solve this, we could solve that, we could solve that. You're going in an inch deep, in a mile wide, and I just don't feel like the success that you want to see from that sort of investment would be there. So, that would be my two sort of big recommendations, I think.

Host: Sound advice. I want to thank you for joining today's podcast and sharing your insights with members. I think what you guys are doing is very impactful for members. So, yeah, I'd love to continue learning more about the journey for the organization and really how you're working with members across the country. If you would like to learn more about Ailevate, please visit www.ailevate.com, and that is A-I-L-E-V-A-T-E.com.

If you would like to learn more about AHA Associate Program, please visit us at sponsor.aha.org. This has been an AHA Associates Bringing Value podcast, brought to you by the American Hospital Association. Thanks for listening.