The Business of Healthcare Podcast, Episode 137: A Push Toward Smarter, Simpler Care with AI

We're not just developing AI and technology for technology's sake. For us, this is really about a commitment to simplifying healthcare. Welcome to the Business of Healthcare podcast from the Center of Healthcare Leadership and Management at the University of Texas at Dallas' Naveen Jindal School of Management. Here at UT Dallas, we bring together business executives and other thought leaders to help navigate the challenges of a rapidly changing, increasingly complex healthcare ecosystem.

I'm your host, Professor Daniel Carnuda, Director of the UT Dallas Professional Program in Healthcare Management. Be sure to subscribe on Apple Podcasts, Spotify, or your favorite podcasting app to ensure you don't miss any future episodes. You can also join us online at businessofhealthcarepodcast.com.

As the artificial intelligence phase of technology's evolution in our personal and business lives continues at a rapid pace, several of our recent episodes have focused on AI. In this episode, we're going to further that theme by taking a look at AI from the healthcare payer's perspective. Our guest is Anita Allemand, who is the chief growth officer for one of the biggest health insurance companies in the country, Elevans Health. Anita has a diverse background in healthcare.

She holds a doctor in pharmacy degree and previously was senior executive at Walgreens Boots Alliance, leading healthcare services. Anita, welcome to the Business of Healthcare podcast. Thanks, Dan. So will you start by giving us and our listeners just a background of your current focus with Elevance?

Yeah, as you mentioned, I'm the Chief Growth Officer at Elevance Health. Here at Elevance Health, we serve about 109 million consumers and members and have a diverse portfolio of insurance, pharmacy, behavioral, clinical, and other care solutions. As the chief growth officer, one of my main focus areas is to deliver and scale consistent, you know, human centered whole health experiences at every life stage. So what this means is solutions that truly elevate the experience and drive better health care outcomes for all the consumers and the members that we serve.

I've spent my entire career and I feel like I've grown up in healthcare coming from a family of clinicians. I, in fact, started as a practicing pharmacist and a pharmaceutical researcher before that. So my entry point into healthcare was very much at the patient level over the past 25 years. My roles have changed over time, but the lens has always been the same.

It's about how we make the system, which tends to be very complicated, work for people and improve care in ways that truly matter in everyday life. So how are you using AI to help transform your area of the business? Do you have certain, I don't know, do you break it up into main areas that you focus on? Yeah, we're not just developing AI and technology for technology's sake.

For us, this is really about a commitment to simplifying healthcare. The four main areas that we are leveraging AI is in care navigation, how we support providers and practitioners, personalizing experiences, and then clinically, how we close healthcare gaps. So let me get into maybe a bit of specifics on each one of those. Yes, please.

So when we say care navigation, we know that healthcare can be really daunting and confusing, especially when it comes to navigating some of the more common questions. We've developed an AI-enabled virtual assistant to help our members and our consumers understand their benefits. What is covered? Why is my claim the way it is?

What coverage limitations do I have? And really guiding them with personalized support. And this is available 24-7 to more than 10 million of our members. Early results in our pilot programs found that almost 9 out of 10 users were able to leverage this virtual assistant and help them find answers.

And the beauty of this is they can do this at their convenience with their digital infrastructure. And all this information is available at their fingertips. So that's really around understanding and helping someone navigate their benefits and care. Can they do that by phone and by app?

Yes, it's a digital, it's called our Sydney app. So it's an app that is available to all of our members. But we have rolled it out to 10 million of our consumers and our members right now with a plan to broaden that and scale that across our entire book. But if they're not good with technology, they can call a number?

Yes, this is really about choice. If they want to use the virtual assistant on their phone, they can use that. If they prefer the traditional methods of telephonic and calling someone, they can do that as well. And the same information is available in both sources.

So it's really about what works best in personalizing the experience to preferences. Are they getting an AI generated voice talking to them or is it a human? It is a human. They have prompts, but ultimately, if they want to have a conversation with a human, they can have a conversation with a human.

Some things can be automated using IVR technology, but in almost every instance, if they want to speak with a human, they have the ability to do that. So that's the care navigation piece on how we're leveraging AI. The second I would say is how we support providers and practitioners. And frankly, as a prior practitioner, this is one that I really feel very strongly about.

The ability to reduce administrative burden associated with providing care is huge. So we have a proprietary technology called HealthOS. Think of it as a bi-directional pipe that shares relevant data back and forth between providers and back to us. And so what this does is it helps the providers.

So we get data from multiple sources and we provide that information to providers. So it gives providers a really complete view of their patient. And so it can help reduce administrative burden. So by surfacing these insights, sharing the data early with the providers, the providers can make decisions in real time.

They can see what their patients are eligible to receive. They can see what criteria or guidelines we might have and provide all that in real time. And so we found that this really removes a lot of the administrative burden and allows our providers to spend time with their patients really focused on patient care which is what all clinicians want to do is take away the burden so I can actually spend time improving care for my patients Is one of the goals with that to reduce claims denials It is to provide the right information necessary to providers so they can ultimately make the right decisions on behalf of their patients.

The technology is never used for denial. So when it comes to denial, it's always a human in the loop, a provider speaking with a provider, but the technology is really meant to surface all the data that's necessary to make that approval go smoothly. But if there's a denial, it's not technology that does that. Now, that's what I meant.

If you give the information to the providers on the front end, then they can get their claims right the first time and there'll be nothing to deny on the back end when you guys... Absolutely. And that's in real time, right? So we don't have to put the patient and the consumer in the middle of navigating that and getting more information from the provider and providing it to us.

You're right. It really helps reduce denials. And how does the interface work with the provider? Like, is it agnostic, regardless of the provider's patient system that they're using?

Your pipeline can hook into their system? We have the ability to. However, it does require integration on the part of the health system or the provider. And as you know, data interoperability is one of the major challenges in healthcare and with EMR systems.

So we are working with the most common providers to ensure that it is easy to integrate that directly into their workflows. And so that is the intent. But in some cases, it is health system by health system, depending on what systems they use and what EMR providers they use. Yeah.

But the intent is to work with all of the EMR providers to make it easy for the providers to actually connect into us. All right. And then I would say the other two areas that we're using, digital tools and AI, is on personalized experience. And so what this means is that, imagine, if you will, all the data that we have on our members and our consumers.

In our Sydney Health app, which is the app that I mentioned when we talked about care navigation, we use over 500 plus data points on very specific personalized data points to help our members find the right provider. And so that's based on their data, their preferences, their locations, you know, availability of appointments, et cetera, to really be able to, so if someone goes on our app and say, I need to look for a provider for a knee replacement, it really surfaces based on all their preferences and data and provides a really good match.

So that's one that, you know, we've learned not just finding care, but connecting care to the right providers is very important to our members and our consumers based on a lot of the focus groups we've done. So that is another way in which we're using data and AI to personalize finding care. And then the last I would say is proactively closing care gaps. And what does that mean in simple talk?

It is about identifying the needs and using the data to identify someone who might have a chronic condition, but for example, hasn't taken a preventative test or isn't on the write therapy or medication or has medications that might have interactions. So to be able to close those gaps in the care requires us to be able to see all the data across all the providers. So that is another way in which we are leveraging AI to help identify those care gaps and to push that information out to providers.

So the next time they see their patient, that comes up as an opportunity for an intervention for them. So that's another way we're thinking about how we leverage technology and AI. So care navigation, supporting providers, personalizing experiences, and finally, from a clinical perspective, making sure that we are seeing the whole health and providing data points into providers to close care gaps. So I can imagine this is changing rapidly on you, the technology that's available.

Is this an everyday thing where you come in and say, oh, now we can do this? I mean, or does your team get together and say, we'd like to do this for our beneficiaries, and then take that to the technology guys and say, is this type of technology available? How does that process work? I would say it's a fantastic partnership with our technology teams.

We always start, in my motto, especially as we think about product design, benefit design is human-centered principles and design. Start with how do you make experience better? How do you simplify? And when you start with the patient, the consumer in the middle and work backwards, you know, we start with what we want to achieve.

And our technology teams help us figure out the right solution and the right organizations to work with. So technology truly becomes an enabler. So we're not going out and saying this technology, how do we retrofit it into our products? It really is around starting with the experience and the outcomes in mind.

As you know, the technology is rapidly evolving in this space. And so we have partnerships. We build some of this ourselves. And we also have partnerships externally so that we can learn and grow at the same pace that the technology is rapidly evolving in.

So I know this really isn't your area per se, but are you seeing how they're using AI to streamline the adjudication, the claims adjudication process? Yeah, you know, I would say, yes, we I have seen that it isn't where I spend the most of my time. But but adjudication, real time adjudication is a key area because it is also one of the big friction points. You know, we start with what are the friction points in health care and how do we before you can transform, you have to foundationally be able to to reduce, as I've mentioned, burden and friction.

So it is definitely one that we are spending a lot of time is how do we leverage both the technology and AI to get ahead of any of the claims issues we might face. So it's not a retrospective review, but we're doing it prospectively at the time of adjudication. And I understand Elevance has a relationship, some kind of relationship with open AI. What does that involve?

You know, as I mentioned earlier, you know, we partner with organizations externally help us in our roadmap. We've partnered with OpenAI on multiple fronts, but I'll talk about one. It's to train our associates in responsible AI through a very formal certification program. program.

We truly believe in elevating the skills of our associates and leveraging OpenAI to do that. So more than 60,000 of our associates and counting, frankly, have been upskilled through this program. And it covers foundational concepts prompt engineering and practical applications of AI in their daily workflows This partnership quite frankly it strengthens our commitment to leveraging and deploying AI and to do it in an ethical way. So we make sure we're embedding privacy, security, and transparency in all of our solutions.

So the partnership has been both around how do we leverage the technology to advance our roadmap and our products, but also how do we train our associates and upskill them so they are able to leverage that in daily work? Are you getting or feeling any pushback from your associates, your employees regarding AI? It's a scary term for most that don't really understand how it's being used. You know, I think I would say initially it wasn't pushback as much as lack of knowledge.

And so that really ties to why we've partnered with AI to train our associates when you don't know what's expected of you. And frankly, when you don't know how it's going to help you, it can be a confusing proposition. So we've spent a lot of time really making it relevant for our employees and our associates. So, you know, moving from the macro of AI being daunting to this is going to help you get more information.

So, for example, when you're a frontline associate speaking with a member or a patient, you now have all that data in front of you. You have the preference. So it's going to make it easier for you to have that discussion versus them thinking this is something that's going to eliminate my job. So we really had to figure out how to make it relevant for each person and based on what they're doing.

So I think the initial pushback was curiosity and wanting to do more. And the more we have focused on education, training, and specific investments to really be able to develop use cases and to scale them, that hesitation has grown into excitement. This episode is brought to you by the Center for Healthcare Leadership and Management, the definitive resource for healthcare management education in North Texas. The center is based in the Naveen Jindal School of Management at the University of Texas at Dallas.

It plays a unique role in training the next generation of healthcare leaders to meet local, regional, and national demands. The Jindal School uses its strengths in accounting, administration, finance, marketing, and information systems to educate highly qualified personnel for healthcare administration and executive leadership positions. The center is home to seven healthcare leadership and management programs, including undergraduate and graduate programs, as well as executive programs for physicians and working professionals.

For more information, visit us online at jindal.utdallas.edu forward slash healthcare. So let's shift a bit, or maybe not so much, to value-based care payment models.

How prevalent are such models in Elevent's business today? Yeah, value-based care models are a core part of our business. In 2024, almost 66% of our total expenses, our total medical cost was in value-based care arrangements. Wow.

In our Medicare Advantage population, as you can imagine, it's a high rate there. And frankly, the correlation of being in those arrangements to demonstrating value and outcomes in quality metrics has been pretty significant. And so, as I mentioned, almost 66% are in VBC arrangements for us. Is AI being used in those type of situations?

So I would say AI and technology is used in those arrangements to do two things. One is enable providers to have early identification of data and needs. So they're able to prioritize where and how to deliver care, the real-time guidance that I mentioned, and coordinating all the interventions. So I would say it's a combination of AI plus data in a timely manner, helps our providers know exactly where to focus their time and to make sure that they get ahead of, you know, before someone moves into really high risk.

So technology has provided that continuous loop for them and also to understand, quite frankly, what works and doesn't work, tracking the data, tracking the outcomes and the metrics. So then care can be really personalized because interventions that work for you, Dan, are very different from what might work for me. So we are, in fact, leveraging technology to say on a personalized level, what's the right outcome and what are the interventions that actually deliver that outcome?

So then there is that continuous feedback loop back to the providers as well. Yeah, that's one of the things that's needed for any type of value-based care arrangement would be access to data quickly so that quality and cost can be monitored in real time rather than after the fact. Surprise, here's your scores. I'm frankly losing track of, you know, what is AI and what is based technology anymore?

It seems like it's the same thing. You know, and I use it sort of interchangeably as well, because at the end of the day, I think all of these are technology investments, whether it's AI, agentic AI, data. So they're sort of core enablers to delivering healthcare in a more efficient and sort of outcomes-driven way. Yeah.

And I mean, everybody seems to be using it in some form or another. So with that in mind, do you have any longer-term views on where you see whatever we want to call it, the next phase of the AI evolution and where that might manifest in your business? What are you guys seeing in the not-too-distant future? I would say two things.

One is we still have a lot of work to do around AI and its connectivity across the healthcare ecosystem, because without it, AI can only see part of a picture, right? And this goes back to data interoperability. When we truly can connect all aspects of healthcare data, which today it's fragmented, then AI can sit on top of that and provide greater transparency, greater interoperability, and a full picture. So I do think foundationally, that's one that needs to evolve.

The second phase, I would say, is about proactive personalization at scale. So using it to anticipate needs proactively. So almost to the point you made, some things are still retrospective. Some things are still a dashboard after the fact.

So we have to push forward You know I say move left and make AI and personalization about real time guidance closing that care gap when it happens optimizing for example medication management when someone goes to pick up their prescription and anticipating everything earlier. So again, it's about moving from this reactive model, which by the way is also helpful at this point, to being far more proactive, moving left early, so then you can synthesize all the data you have, whether it's clinical data, benefit and plan data, social data, all of that in an integrated way and focus on prevention and identify those needs far earlier than we're doing today.

So I think it really is about proactive and then this personalization element, frankly, not just in pockets, but at scale where it becomes a core part of how we deliver healthcare. One of the challenges I see there is not on the technology side from where you sit, developing those tools, but from our side, the patients, and getting them trained on actually using this, I guess, technology that's available to them. Some people care, some people don't. Some people will use what's available to them to help manage their own health.

Others will completely ignore it. What are your thoughts on that, of training the patients once you've actually got the foundation of technology available to them? You know, that's an excellent point. And I would say one is we have to meet consumers where they are.

And what that means to me is that we have a responsibility to make it really easy, to make it simple, to increase adoption. So that's one part of it. And the training, it has to be really intuitive. all of our product design elements start as I mentioned with human-centered design we go out we test it with real patients across different demographics across different levels of understanding and really design for sort of the the wide variances amongst those so it starts with design elements and in philosophy so that's one so make it simple make it intuitive that training is not the primary barrier.

Second is make sure we are in fact training and there's always a human in the loop. So if the technology isn't working for someone, they have the button to say, let me speak with someone. And that's another element that we are working on. As I said, we leverage AI and technology to make sure that our frontline employees and associates have the same information.

So now they in fact can spend time having that conversation And maybe even training someone who calls them and says, I don't understand. This is not working for me. So that's the second piece. And then I would say this is also predicated on the belief that technology isn't meant for everybody.

I thrive on it. I don't read print mail. I'm getting to emails that way as well. But for my 85-year-old father, it's always going to be that face-to-face interaction that he craves and he learns from.

So how do we enable those providers and where they get that information? For my dad, it's his pharmacist. He is at that pharmacy counter all the time. So we have to make sure that those clinicians also have the ability to use the technology and, in fact, can be that trusted trainee, if you will, for the information.

So I do think it's meeting people where they are, making it very simple and intuitive and continuing to have other avenues if technology isn't your go to solution for understanding and using health care. So I suppose what might happen is I kind of think through this in real time here talking to you. It's a matter of at some point getting to like 90%, 10%. 90% of the people use it for their advantage to help keep them healthy.

And there's always going to be a 10% that just don't. But right now we're not there. And because of the baby boom population, which are notoriously non-technology as compared with the younger generations behind them, we might be 50-50 right now. 50% of the people use it.

50% of the people either don't care to use it or don't want to use it or whatever the reason is. And as we move along and the baby boom population sort of phases out, which will happen, and the new generations come in, which are more technology enabled, we may get to a point where it gets to 60-40 and then 70-30 and then 80-20. So I think it's going to be a slow crawl before population health through AI is really a thing. And by then there'll be something else that's even better.

Yeah, that's where I think you're right. It's on a different maturity curve right now. And I think as the population ages and the comfort with technology increases in aspects of life that are not just healthcare, then healthcare catches up to that, right? And if you think about most people are very comfortable using technology.

Think about even simple things like food delivery and others. But we had a big learning curve and a big barrier when I would say mail order pharmacy was first introduced and you had to use technology to say you're getting something delivered at home. But now it's common practice. So I do think as technology outside of healthcare becomes the norm, healthcare technology also has greater adoption, if you will.

Yeah. And where the patient is involved in thinking of corporate projects, I think whoever's monitoring those projects needs to be reasonable in determining what the actual patient uptake is going to be in order to determine success for a particular project. Absolutely. Yeah.

No, we measure a lot of consumer experience metrics. you know, do very real time feedback loops. So it's not a develop and deploy the technology and that's it. You've got to continuously be getting feedback and metrics to support how you have to evolve and what you need to do to improve it.

Great. Well, Anita, thank you for being with us today. This was fascinating. And I hope that we can catch up with you at some time in the not too distant future to see just how right or wrong we were in some of the things we're talking about here.

Absolutely. Thank you for having me. This has been my pleasure. Thanks for listening to the Business of Healthcare podcast.

To learn more about the Center for Healthcare Leadership and Management and the Healthcare Management Business Degrees and Certificates available to seasoned clinicians, master's students, and undergrads through the University of Texas at Dallas, go to jindal.utdallas.edu slash healthcare. Thank you.