On Aon Podcast: Technology Impacting the Future of Health and Benefits

Intro:
Welcome to “On Aon,” an award-winning podcast featuring conversations between colleagues on, well, Aon. This week, we hear from Doug Melton for a discussion on the impact of technology on the future of health and benefits. And now, this week’s host, Sheena Singh.

Sheena Singh:
Hi, my name is Sheena Singh and I'm senior Vice President in Aon's healthcare industry vertical. Today we're talking about two critical elements for all companies in the future: technology and health benefits. Advances in technology will not only transform healthcare and treatment outcomes, it will radically change benefits, personalization, access to care, diagnosis, treatment, and affordability. HR leaders are thinking about how these new and emerging capabilities, like AI, will change health and benefits, and our healthcare clients and carrier partners are thinking about how it will impact how they do business.

I'm going to do that one again. HR leaders are thinking about how these new and emerging capabilities, like AI, will change health and benefits, and our healthcare clients and carrier partners are thinking about how it will impact how they do business. What will the future of health and wellbeing hold, and how can employers take advantage of new developments to mitigate healthcare costs, improve workforce resilience, and create a differentiated employee experience? With me today to discuss is Doug Melton, Global Analytics and Actuarial Leader for Health Solutions. Thanks for being here today, Doug.

Doug Melton:
I'm happy to be here. Thanks for inviting me.

Sheena Singh:
Good. Well, before we get started, I'd love to ask you a quick warmup question, if you don't mind.

Doug Melton:
Sure. Nothing like warming up.

Sheena Singh:
All right. So, tell me, Doug, what is it about your profession that most inspires you?

Doug Melton:
Oh, that's a good one. I'd say what the basis of our entire profession is driven around, to the best of your ability, facts, and specifically being able to use data and translate that data into actionable insights. Now, what makes it a passion of mine is the use cases that we particularly focus on aren't necessarily automating cars, or it's not necessarily advancing agriculture. Instead, we really focus on three things within the institutional space. Number one is employee health outcomes. Number two is if you impact those employee health outcomes, they impact businesses, so it impacts business performance. And then third, if you can impact people's health so that they come up to work and be more productive, other aspects of their life, particularly home life and community life, should as default improve as well.

So, the unique part about specifically looking at different types of employer-sponsored benefits, different types of employer-sponsored incentives, and not look at them just narrowly through the lens of what makes a total reward strategy effective. Instead, we say “what specific attributes are associated with higher level success based on what we know about these individuals from our data?” And therefore, we're able to put together a sort of a fact-based plan for employers on an annual basis for them to be able to maximize their employees' health outcomes, their business performance, and ultimately their success at home life as well as work life. That's what inspires me about this role.

Sheena Singh:
I love that. I love adding the structure and processes to make sure that you're able to really create the right approach in order to help people at the end of the day.

Doug Melton:
Absolutely. Absolutely.

Sheena Singh:
All right, so let's just jump right in then. So, Doug, what are some of the new technologies impacting healthcare?

Doug Melton:
Sure, some of the new technologies, well, let's start off with by definition, let's just define, we say new technologies, what space are we talking about? So, they could be remote devices or smart sensors. Those have been a while for quite some time. We're going to spend a lot of time talking about artificial intelligence and really unpacking AI and kind of trying to make it simple. And then sort of traditional analytics and machine learning is something that we're going to cover as well. But let's just, when we say what new types of technology, let's start there. I think with the wearables, what that's been able to do is it's been able to allow us to actually see individual level data in terms of the home. So before we started wearing wearables, the ability to really understand someone's metabolic health was primarily limited to lab-based studies.

But now we're able to look at things like recovery rates from sleep. We're able to look at heartbeats that might be irregular, and these are allowing us to be able to do some of the testing that we would normally do on an annual basis during a physical, and now it's something that we're doing on a daily basis. So that's just one example where wearables and telehealth that has allowed us to do what I would call daily care versus the annual physical care. If we move kind of beyond primary care and then maybe move into diagnosing before we go to treatment, you want to prevent, you want diagnose, you want to treat. If you move into column two around diagnosing specifically around NLP, natural language processing, and what I'm going to talk about in a few minutes, image recognition, those have particularly been used a lot in the MRI and the CT scans.

And that's been helpful specifically to be able to reduce the amount of human errors. But now we're not saying we're replacing radiologists, it's not what's happening there. But across the majority of major medical systems, particularly those that are tier one, they're using this type of technology to enhance the diagnosing, so we have fewer errors that are happening there. And then if you go on the treatment side of this, you certainly are thinking about things like gene and cell therapy that allows us to be able to be in a position to have life-changing events for rare therapies. Now you say to yourself, well that's not necessarily a technology, Doug, right? That's a transfusion of gene and cell.

Well no, a lot of the analytics that are going around evaluating it and understanding efficacy and durability, and do these individuals regress to the mean, because the denominator's so low, it's so small for those kinds of randomized trials, we're able to use different types of actuarial analytics that we haven't been able to use before to kind of forecast what we think the future might be, even if we don't have mounds of data that you say you might have for a specific generic drug that's been on the market for 20 years managing a chronic condition. You won't have that length of observation periods or the number of people that are large enough to have that level of confidence. So again, being able to use some of our actuarial forecasting in a way that's different before would be an example of how we're using around gene and cell therapy, durability measurement. So those are three examples from the primary annual prevention to the diagnostic to the treatment of how it's actually being used today.

Sheena Singh:
That's great. And you started to touch on this, and I wanted to better understand what's the difference between this new technology, like generative AI, for instance, from the more traditional analytic approaches like predictive modeling that have been used?

Doug Melton:
Yeah, we knew this question was coming, and so my consistent answer has always been, the primary difference very simply is, AI can be autonomous and learn on its own versus the predictive analytics that we've done of the past. Mounds and mounds of data scientists have had to go in and kind of figure it out through their own sort of human-based iterative learning. That's the most simplest answer possible is, autonomous learning on its own, not having to have any sort of human intervention. Whereas our traditional predictive analytics and our classification models, you actually have to have some individuals in there. So, for example, in the ML world, generative model and image classification are very common in healthcare. I talked about them before, specifically image classification, is used in most major medical centers now uniquely for the MRI and CT scans.

And then in comparison, the traditional predictive analytics are primarily using forecasting for our actuaries. And then for a lot of the self-insured health plans and the self-insured hospitals, they're using some of the traditional predictive analytics to predict disease progression and to figure out which cohort of individuals do you want to engage today versus those, their disease is not as progressing as bad and you might either delay or give them a lessened base treatment. So very simply, autonomous versus human intervention. And again, two specific most common ones in ML would be AI degenerative model as you described, as well as image classification, healthcare. And then also, again, going back to more of the actuarial side separate from ML, but using predictive analytics for forecasting reasons to look at disease progression and future medical cost trends.

Sheena Singh:
Well, that is really helpful, and I think you started to touch on this. Why do these new technology advancements matter for employers at the end of the day?

Doug Melton:
Well, the new technologies matter for employers because it should make the employee experience better, unique when it comes to not just healthcare, but just health and benefits in general. So again, if we use the previous framework or prevention, diagnosis and treatment, let's take that and kind of put that on the H&B sort of point of view, and that would be sort of open enrollment to getting the benefits that you need using those benefits. And then at the end of the year, you actually see a return on your investments and those benefits that you had. And so, these new technologies, again, going across that value chain, will now help us do what's called smart enrollment.

Not to say that folks weren't making intelligent decisions before, but now you have this next best action approach that is in almost every part of the open enrollment season where you give the platform whenever platform that you're using a little bit of information about you information coupled with whatever historical data they have, and they will recommend what range of health benefits, what range of voluntary benefits, or even some non-insured perk benefits at a global level that you think would benefit you the most given your situation or your family's situation.

We had some of that sort of conjoint analysis before, but not at the way that we have it today. And that's probably less of a function of the technology, and that's more a function of big data. And then sort of post H&B enrollment, employers could be happy because now you're saying, "My NPS score should be happy, my retention rate should be happy because I know that my employees are enrolling and benefit that they are benefiting from at work and at home." And then on the back end of that, "You should be actually be able to say, am I getting an ROI on this?"

And so, a lot of what I would call multi regression analysis that's done today is able to tell an employer, "Hey, look, I think your disease management vendor and diabetes is doing a fantastic job, that behavioral health is really working for the adolescents, but it's not working for the sort of adults." And so how do we start to think about different strategies we want to take and how do you maybe bring in multiple behavioral health vendors versus a single one based on the results around efficacy that we've seen? Again, sort of multi-variable aggression analysis. So again, an employer benefits from this very simply, they keep their workforce happy through smart enrollment, they're able to then see which products are or are not working, and hopefully that impacts their P&L and their bottom line on the backend.

Sheena Singh:
Wow, that's great. I love the idea of making sure that employers can use this technology to focus on the employee experience and making sure that the experience is customized and personalized to them. But at the same time, we all know employers are struggling with cost and affordability and using this as an avenue to make sure people are in the right programs for them that are the best fit and also affordable for the employer and the employee. So that's great. So, let's talk a little bit more about how Aon specifically is approaching the use of such technology to help our clients.

Doug Melton:
Exactly. So, we are advising our clients on how they can utilize their data and analytics, and our analytics, to better understand everything from the quality of services to program engagement to return on your investment, to your last point about the importance of affordability. Affordability is one side of the equation. The other side of the equation is the investment. And so, do we get those cost savings associated with those expenses? And so, we're able to specifically use it like that. We even have an offering called what's called a cost efficiency measurement tool. And it's something that's been in high demand this last year as clients have really been trying to understand that question around ROI, specifically around chronic condition engagement and around provider accuracy. And then the second thing is we're also helping our employees with some innovative solutions. And those solutions really depend on what their needs are.

A middle market employer may not necessarily have a multi-state contractual relationship with seven or eight different hospital systems. They may have two systems in the one state in which they operate. And so their questions are vastly different from a multinational that's got captives, that's got expats, that's got different regulations in terms of how care or care cannot be delivered for their workforce. And so therefore you have to get creative because just simply saying telemedicine or simply saying, "Steer everyone to this tier one hospital," is not going to be effective for them. But I would say that is a space we spend a lot of time in is how do we're using these technologies to innovate with our clients and a lot of times, at least the funnest part, so when we co-create with them. And we're learning something, they're learning something and then, more importantly, we can take the learnings together and actually apply them to other clients who have the same situation.

Sheena Singh:
I love that collaboration so we can continue to progress in this area. So, Doug, another question that I have as we think about technology, we also have to think about data security. How can organizations stay safe while getting the benefits of all of these technologies?

Doug Melton:
Yes. So, the first one is this is a top priority and a top area of focus for us at all times. And it's been this way for many years. It's not something new. The data privacy is something we continue to focus on and the rules of safe use, and we share those explicitly with our clients. We share those explicitly with our vendor partners, and of course we share those, all the coworkers within the colleagues within the walls of Aon that are using this data to help our clients benefit. Some real practical examples, so it's not being so vague here. We have secure private health data for no health breaching, and we ensure safeguards are in place when we have computational errors that specifically occur.

We also have different data stacks and data lakes that are set up for purposes for transactional business, versus data lakes and data stacks that are stood up specifically for that previous conversation we had around innovation. And so, I don't want the innovation experience and that data that we're using for that to be mixed up with the enrollment of data or the prior authorization data that we're collected on behalf of the client. Let a person be able to get the care that they need to get the procedure that they need, and let's make sure that procedure is paid for appropriately, and there isn't a high cost bill there that's been out overpriced. And that kind of information and that interaction, those data sets sit in completely separate data houses from some of the stuff that we do around, say, innovation or predictive modeling. Does that help, Sheena?

Sheena Singh:
Yeah, that's really helpful, and I think of course important for everybody to keep in mind. But we know this space is continuing to progress. This was just scratching the surface hearing from you, Doug, today on where we're at. But I'm really looking forward to seeing how Aon is going to be partnering with our clients in this technology and AI space as we continue to enhance the health and benefits that employees are offered. So, thank you so much. I appreciate it.

Doug Melton:
Absolutely. Absolutely.

Sheena Singh:
So, before we sign off, I'd love to ask you another quick question just so our listeners get to learn a little bit more about you personally. Doug, what was the last book that you read?

Doug Melton:
Oh, okay. The last book that I read... And my wife is probably cringing if she listened to this podcast, so I tend to kind of follow her books for a while. So, she reads a lot of books. So, the answer to your question is Good to Great, right, by Jim Collins. And what I typically do, because she reads about a book every two weeks, is I see how many books she orders that sits on the kitchen table and never get read. And then I see the books that she orders that she reads, reads twice and then maybe recommends to a friend. And those are the books that I picked. And so, I read Good to Great probably about two and a half years ago when I was actually transitioning over to Aon, to try to figure out what are the behaviors, what are the organizational attributes why some companies are better than others.

And I thought I particularly needed to do that since I was kind of sort of pivoting my focus from just North America to more of a global role. And that book really helped me think about that in that regards. And so that was a good book to have, and she's recommending some other ones right now that have me focused more on servant leadership, is sort of the new sort of set of books that she's put in front of me. And so, I'm reading chapters one and chapters two of a full book, but it's something that I've embraced both at work and in the community. And so, trying to read and just get smart about that. And so, I'll have a second or third book for you by the next time we talk, Sheena.

Sheena Singh:
That's great. Perfect. I love that you inadvertently get a book club out of your wife getting access to all of these books, so that's awesome. Well, thank you so much, Doug. I appreciate all of the insight you provided today. That's our show for today. Thank you all for listening and look for the next episode of On Aon, coming soon.

Outro:
This has been a conversation “On Aon” and the future of health and benefits. Thank you for listening. If you enjoyed this latest episode, tune in soon for our next edition. You can also check out past episodes on Simplecast. To learn more about Aon, its colleagues, solutions and news, check out our show notes, and visit our website at Aon dot com.