Unleashing the Potential of Technology in Value-Based Care with Jean-Claude Saghbini, President at Lumeris and a Leading Expert in Technology and Engineering for Population Health Management

[00:00:00] (upbeat music)

[00:00:02] - Hey everybody, welcome back to the Health Matters Podcast.

[00:00:12] I'm Saul Marquez, hosting the podcast today

[00:00:15] straight from Las Vegas on the show floor.

[00:00:18] Today I have the privilege of hosting Jean-Claude,

[00:00:21] Sag Beni, who leads technology and engineering

[00:00:24] for population health management at Lumeris.

[00:00:27] To improve outcomes and reduce care variability

[00:00:30] and costs in value-based care.

[00:00:32] With more than 15 years of experience in healthcare,

[00:00:35] Jean-Claude is passionate about forging

[00:00:38] the seamless integration of technology

[00:00:40] into clinical workflows.

[00:00:42] Jean-Claude, thanks for joining me today.

[00:00:43] - Thank you very much, thanks for having me

[00:00:44] and for the great introduction.

[00:00:46] - Of course, and so look, we're really excited

[00:00:48] to dive into Lumeris and the work that you guys do.

[00:00:51] Before we do, I'd love if you could just share

[00:00:53] a little bit about what inspires your work in healthcare.

[00:00:56] - Yeah, so I've been in technology since I can remember

[00:00:59] sometimes 20 years ago, I decided to get into

[00:01:03] more of a vertical space that is mission-driven.

[00:01:06] Healthcare was it, been in healthcare since 2005,

[00:01:10] done a variety of things, primarily technology-based,

[00:01:14] but with a focus on driving outcomes on quality and cost.

[00:01:18] The drive of what, behind what I'm doing at Lumeris

[00:01:22] is that we're at the front line of working

[00:01:25] with health systems to help them succeed in their mission

[00:01:29] and succeed in value-based care.

[00:01:31] And the technology that we bring to bear,

[00:01:34] what's exciting about it, and I look,

[00:01:36] I've talked about it myself, but also my entire team

[00:01:39] is driven by the same mission, which is putting technology

[00:01:42] that they can see it produce the outcomes, right?

[00:01:45] They can see it produce higher quality of care.

[00:01:49] They can see it make meaningful dense in cost.

[00:01:52] And that's what drives me, what drives my team.

[00:01:54] And many people here in health.

[00:01:57] - That's a great mission, Jean-Claude,

[00:01:58] and really let's dive into some of the work that you guys do.

[00:02:01] What role does technology play in value-based care enablement?

[00:02:06] - Yeah, it's a central role.

[00:02:08] It's actually a foundational role.

[00:02:10] It's not the entire role.

[00:02:12] You can do it with technology alone,

[00:02:13] but you cannot do it without technology.

[00:02:15] In order to succeed in value-based care,

[00:02:17] There are some imperatives that you have to do. One is you have to be able to understand the populations that you are serving and you need to also be able to understand the end of one patient that you are able to, that you need to serve. To do that, you need to get access to vast amounts of data that unfortunately in healthcare is not data that today is cleansed or normalized or follows any particular logical standards.

[00:02:47] You have to cross the board. So this is where technology comes in, right? Getting the data, aggregating it, normalizing it, deriving insights that are beyond the descriptive nature of that data, primarily machine-learned insights, sometimes rules-based insights based on experience and whatnot, and then taking these insights and putting them in workflows to drive action. All of that entire stack is for the purpose of improving quality and reducing cost.

[00:03:17] But like I said, it's not enough to only have the technology, so you need all of this technology I just described, which is difficult to put together, it's putable together, but it's quite an endeavor given the vast amounts of data that's needed.

[00:03:29] But then you have to make sure that you have change management and aligned incentives to act on the data to drive for the outcomes that you're trying to drive towards.

[00:03:39] And that's also part of the data that's needed, right, to get the right incentives to whether the part of the same challenge.

[00:03:46] Exactly. And I was just on a talk earlier, and we were describing how we're not only getting typical clinical data, but we're finding sort of major improvements when we are taking that data, claims data, clinical data, etc.

[00:04:00] And then combining it with social determinants of health data, such as transportation challenges, language barrier data, affordability data, whether the presence or non-presence of caregivers.

[00:04:13] And how do you take all of these sort of these data elements that are typically not healthcare specific, but then when you combine them with the typical health care data, you get much more powerful insights that are way more predictive and with highly measurable outcomes.

[00:04:29] Thank you for that, Jean-Claude. How are you currently using generative AI? That's a big thing to do.

[00:04:35] today both within your organization and to transform care delivery. Yeah, let me just

[00:04:40] go back a tad bit to before Generative AI. We have a great question. It's off of my

[00:04:46] trust in these days, Generative AI. Before AI had been a core foundational element to

[00:04:53] our technology, machine learning, building machine learning models for predictive models,

[00:04:59] detection models for various conditions, things like I talked about earlier, which is combining

[00:05:05] social literacy as data with clinical data. Those were great capabilities that AI brought

[00:05:11] to be able to achieve the results we were able to achieve. Generative AI starting last

[00:05:16] November with a big announcement has been even outside of healthcare is a revolutionary

[00:05:22] technology, right? It's off the scale of bringing the internet to the home, right, with web

[00:05:28] browsers, that the internet existed since the 60s. But with the advent of the web browser,

[00:05:35] it became accessible to everybody, right? So, Generative AI is of that scale, has that

[00:05:40] capability. At the same time, we also hear about the limitations, right, and the issue

[00:05:44] that can happen. So, the excitement is paired with how do we put the guardrails? So, which

[00:05:50] describes in our approach? So, our approach to Generative AI is one, making sure that

[00:05:55] wherever we are exploring Deploying Generative AI, we're still going with results of quality

[00:06:02] and costs, right? So, versus technology for the sake of technology. Number two is keeping

[00:06:08] physicians in the driver's seat in the middle, keeping clinicians as the decision makers,

[00:06:13] and then putting guardrails around it. Beyond that, it's using the technology to automate

[00:06:19] as much as possible of the tasks that are automatable, and letting the experts deal with the exceptions.

[00:06:26] So, we started, even internally, right, inside the company, by deploying co-pilot and Generative

[00:06:33] AI capabilities to our software development teams, to enable them to get much more efficient

[00:06:39] at software development. We deployed Generative AI capabilities to our call centers, to allow

[00:06:45] our call center team to be able to access information in a much faster way, and to the

[00:06:50] extent that a patient is on the phone asking questions.

[00:06:52] the answer to that patient can come in a much faster way and increase patient

[00:06:57] satisfaction. Examples are using generative AI in we talked about data in health care

[00:07:04] and the messiness of their data. Turns out that Gen AI capabilities to understand

[00:07:10] messy data and extract normalize it and extract insights from it is quite

[00:07:15] impressive and while we still have humans in the loop to validate that we call it

[00:07:20] the scaled human in the loop and we're using it for data understanding data

[00:07:25] manipulation and scaling our abilities to deal with vast amounts of data.

[00:07:29] It's certainly promising Jean Claude and it sounds like you guys are embracing it

[00:07:33] and it's those organizations that embrace the technology with the right guardrails

[00:07:39] that are going to lead the way and a big thing primary care physician shortages

[00:07:45] clinician shortages overall what does next generation AI powered primary care

[00:07:50] look like yeah a good question next generation AI powered primary care is

[00:07:58] primary care that heads further and further towards automation of primary

[00:08:03] care services again I go back to not taking the physicians out of the

[00:08:08] driver's seat but much of what happens in health care today requires lots of

[00:08:14] automatable blocking and tackling and stretches you mentioned a second ago

[00:08:20] this physician shortage stretches physicians into areas that they could

[00:08:26] not be dealing with and takes away their attention from critical areas where

[00:08:30] nobody else can deal with except them. So next gen is automation of primary care

[00:08:35] services that goes from automating who needs what care and getting them to the

[00:08:42] right venue of care outreaches patients through seamless experiences

[00:08:47] of scheduling based on needs that they have using AI to risk stratify patients

[00:08:56] and guide clinicians to the highest risk categories but not only that with

[00:09:02] recommendations on what are the optimum modalities of care to apply to these

[00:09:06] patients again ultimate decision of the modality stays in

[00:09:10] in the hands of the physician. But to the extent that we can surface the inside that says this is the person, this is the situation they're in. And these are recommended venues of care or approaches to go about them and because of these reasons. And this is where AI can bring a lot of value. That cuts a lot of the data, many people insights double clicking that we're asking physicians to do in a very short amount of time to deal with patients.

[00:09:39] But certainly a promising care model with these new technologies, really not knowing where to start. What tip or advice would you leave them with as it relates to how to approach it?

[00:09:52] If I maybe scope the organization word and call that health systems or provider networks, I would say embarking on the journey of value-based care because this is the future.

[00:10:06] And we're starting to see very positive results in how these models are starting to bend the cost curve.

[00:10:13] Number two is it requires a change, a DNA change in operational practice at a health system or a provider network and aligning with the right partners and aligning with partners who can bring both technology as well as operational expertise

[00:10:34] and making sure that they develop the understanding that these are not point solutions, right? They require alignment across the board from incentives to practice transformation with technology to proper governance models.

[00:10:54] So these are comprehensive changes that need to be done. The good news is that they have been done and there are many success stories out there. But then going back to why we're having this conversation around technology, where we started on technology and AI, etc.

[00:11:09] That the foundation to it all is powerful technology stacks that are future-proof. So gen AI is only the beginning, but we're going to see more.

[00:11:20] So picking technology stacks that are future-proof that would allow them to evolve into the future.

[00:11:27] That's really helpful. And look, folks, the opportunity to succeed with value-based

[00:11:32] cares here. And thanks to ideas like Jean-Claude is sharing and the work that Lumeris does, I think

[00:11:38] we all have a shot at this now. Jean-Claude, this has been really interesting and I want to thank

[00:11:43] you for your time today. Can you share a closing thought and the best place the listeners could

[00:11:48] learn more about you, Lumeris, and the work that you guys are up to? Yeah, absolutely. Look, very

[00:11:53] excited about AI and generative AI. The way I look at it is it is not a silver bullet.

[00:12:03] It's more how to deploy it across the multitude of workflows that are in front of us. I look at

[00:12:11] that technology as technology that has to be deployed across the entire stack of capabilities, again,

[00:12:17] from data to action with everything in between. One AI use case is not going to do it, but think of

[00:12:25] it as the early days of the internet where now the internet is literally in every workflow we do,

[00:12:31] from email to web browsing to banking, etc. This is how we should be thinking about these new

[00:12:35] technologies versus find one point use case and apply to. That doesn't mean that the starting

[00:12:42] point is not one use case, but thinking about these technologies more broadly as being pervasive

[00:12:48] and exploring where they can happen across the continuum of capabilities. Lumeris.com.

[00:12:53] Outstanding. Jean-Claude, can't thank you enough, folks. Again, remember all of the things that

[00:12:58] we discussed are in the show notes, so make sure you take a look at those ways to get in touch with

[00:13:03] Jean-Claude, ways to learn more about Lumeris. It's all there, so take action on what you learned

[00:13:08] today. Jean-Claude, I can't thank you enough for being with us. Absolutely, thank you very much.

[00:13:12] Thanks for having me.

[00:13:21] [Music]