Latest / Becker’s Payer Issues Podcast / Connected Intelligence: How AI Is Reshaping Group Health Insurance

Transcript

Connected Intelligence: How AI Is Reshaping Group Health Insurance · Becker’s Payer Issues Podcast

• 0:00
  Hi, everyone. This is Brian Zimmerman with Becker's
• 0:02
  Healthcare. Thank you so much for tuning into
• 0:04
  the Becker's Healthcare podcast.
• 0:06
  Today, we're going to talk about connected intelligence,
• 0:08
  how AI is reshaping group health insurance.
• 0:11
  Joining me for today's discussion is Mark Jeffries,
• 0:14
  general manager health with Gradient AI. Mark, thank
• 0:17
  you so much for coming on the podcast
• 0:19
  today. Hey. Good morning, Brian. Really appreciate the
• 0:21
  opportunity.
• 0:22
  Yeah. So so let's get into it. Maybe
• 0:24
  first with a with maybe you can introduce
• 0:26
  yourself a bit more, share a bit about
• 0:28
  your professional background,
• 0:30
  and the and the current work you're doing
• 0:31
  at Gradient AI in the in the group
• 0:33
  health insurance space. Yeah. Happy to. So, Mark
• 0:37
  Jeffries. I am the general manager for the
• 0:39
  health practice here at Gradient.
• 0:41
  I joined Gradient a little over a year
• 0:43
  ago. Prior to that though, I was,
• 0:45
  spent most of my career in
• 0:47
  the insurance carrier space, most recently ten years
• 0:50
  with Cigna, where I helped build and ultimately
• 0:53
  led the supplemental health business within The US
• 0:55
  Healthcare Organization.
• 0:57
  But also in my career, I've spent time
• 0:58
  at Aetna, in the Medicare Advantage business, at
• 1:01
  MetLife, in the Hartford, and group benefits and
• 1:03
  financial services.
• 1:05
  So I have a, kind of a broad
• 1:06
  and well rounded background in, in the insurance
• 1:09
  space.
• 1:11
  Specifically to Gradient, I'm accountable for strategy, execution,
• 1:14
  and growth of the health business.
• 1:16
  But primarily,
• 1:18
  the focus is on how do we develop
• 1:20
  our portfolio of solutions to ultimately meet our
• 1:23
  customer needs, right, to create value and drive
• 1:26
  solutions that impact, our customers' businesses.
• 1:29
  Excellent. Well, I appreciate that background in bringing
• 1:32
  that perspective to today's conversation, your work with
• 1:35
  Gradient, but, of course, tons of experience in
• 1:37
  in in the health plan space. So So
• 1:39
  I wanna ask
• 1:40
  about AI adoption in in group health has
• 1:43
  accelerated quickly, of course.
• 1:45
  Kinda feels like AI just got here, but
• 1:47
  it really has not.
• 1:49
  Yeah. But a lot has happened fast, I
• 1:51
  guess, I would say. But how how how
• 1:53
  would you what would you say is the
• 1:54
  biggest shift in how insurers are actually using
• 1:56
  AI today compared to, let's say, just just
• 1:58
  a couple few years ago?
• 2:00
  Yeah. Yeah. That's a it's a it's a
• 2:02
  great question. I mean, it's hard to believe
• 2:04
  that ChatGPT
• 2:05
  was, you know, just launched in in January
• 2:08
  2022. It seems like it's been around forever,
• 2:10
  but I do look at that, that data
• 2:12
  as kind of a a point where there
• 2:13
  was kind of a radical shift in, you
• 2:15
  know, understanding of what, AI and tools like
• 2:18
  this can do to help businesses, you know,
• 2:20
  drive, you know, value, deep value.
• 2:23
  So I I I agree, right? AI adoption
• 2:26
  is accelerating quickly. But I would argue that
• 2:28
  in healthcare, it's still a laggard compared to
• 2:30
  other industries.
• 2:32
  That really shouldn't surprise us though, because healthcare
• 2:35
  has historically been the late adopter of emerging
• 2:37
  technologies.
• 2:39
  And it's, you know, it sounds bad, but
• 2:41
  really it it's not due to a lack
• 2:42
  of desire to innovate. Right? I I think
• 2:44
  there's some really innovative thinkers within, you know,
• 2:47
  the healthcare ecosystem.
• 2:48
  It's just the ecosystem's really complex. Right? And
• 2:51
  it has, you know, significant regulatory,
• 2:54
  requirements. There's massive data security requirements.
• 2:57
  All of those things, you know, act as
• 2:59
  really strong headwinds against this type of innovation.
• 3:03
  That said, there's a significant focus within healthcare
• 3:07
  on leveraging the power of AI to deliver
• 3:09
  meaningful value.
• 3:11
  And early on, right, that focus was on
• 3:13
  using AI to drive operating efficiencies,
• 3:17
  which to me isn't surprising given, you know,
• 3:19
  the margins are know, razor thin often on
• 3:22
  health care solutions.
• 3:24
  And there was a lot of low hanging
• 3:25
  fruit. Right? And so saw just a ton
• 3:27
  of, you know, how how can we, you
• 3:29
  know, empower AI to improve just the, you
• 3:32
  know, the operating models, the underpinnings of these
• 3:34
  organizations.
• 3:36
  However, you know, the value of AI isn't
• 3:38
  just improved speed and cost reduction.
• 3:40
  It's also
• 3:42
  clarity.
• 3:43
  You know, AI tools enable organizations to unlock
• 3:46
  actual insights from vast assets that just isn't
• 3:49
  physically possible for humans to do on our
• 3:51
  own.
• 3:52
  So,
• 3:54
  you know, when you think about,
• 3:56
  you know, these insights are being used to
• 3:58
  bring
• 4:00
  clarity to complex decisions,
• 4:02
  which ultimately enable better underwriting decisions on new
• 4:05
  business, more impactful engagement with high cost members,
• 4:08
  and more informed conversations and decisions at the
• 4:10
  time of renewal.
• 4:12
  But, frankly, the biggest shift we're seeing is
• 4:14
  really this move from experimental
• 4:16
  operationally focused AI to practical actionable AI that
• 4:20
  supports kind of real risk decisions within the
• 4:23
  ecosystem.
• 4:25
  Yeah. And, Mark, maybe you can tell me
• 4:26
  if I'm getting this right too. When when
• 4:28
  it comes to AI and health care, there's
• 4:30
  complexity, there's regulation.
• 4:32
  Also, processes, these these weren't exactly built for,
• 4:36
  AI to just come in and use. Right?
• 4:37
  Like, I've heard the I've heard the phrase,
• 4:39
  like, some of the companies that are or
• 4:41
  or that are doing the best with AI
• 4:43
  right now were kind of, like, are being
• 4:45
  built now or being built for AI. So
• 4:47
  can you say a bit more about, like,
• 4:48
  some of the processes you here are just,
• 4:51
  you know, not necessarily like, you can't just
• 4:53
  plug and play with AI with those. You
• 4:55
  know?
• 4:55
  Yeah. I think that's a that's a great
• 4:57
  observation, and it is a challenge especially for
• 5:00
  the, the organizations who have been been around
• 5:02
  for a long time. And that's, you know,
• 5:04
  when I think about my statement about, you
• 5:06
  know, health health insurers being laggards and adoption
• 5:08
  of new technology,
• 5:09
  it's it's also driven by, you know, just,
• 5:12
  you know, legacy systems,
• 5:14
  fractured, you know, data ecosystems,
• 5:16
  you know, different
• 5:17
  types of technologies
• 5:19
  that, you know, took years and years to
• 5:21
  implement and,
• 5:22
  you know, and it they're they're kind of
• 5:24
  ingrained. And so the cost of change,
• 5:27
  is it can be really daunting. And so,
• 5:29
  you know, I think the the opportunities that
• 5:32
  we're seeing is, you know, how do you
• 5:34
  focus your attention on on AI that can
• 5:37
  be,
• 5:37
  empowered within, you know, current workflows,
• 5:41
  can be integrated in ways that bring information
• 5:44
  and insights,
• 5:45
  to the to the top of the
• 5:47
  the pile. Right? So that, your risk managers
• 5:50
  and underwriters have, you know, a new point
• 5:52
  of view on,
• 5:54
  the risk of, you know, groups that they're
• 5:56
  looking at or, you know, populations
• 5:59
  or renewals,
• 6:00
  but to to find ways to to kind
• 6:02
  of leverage and integrate,
• 6:04
  you know, AI into the current processes and
• 6:06
  current systems.
• 6:08
  And that's that is, you know, unfortunately, is
• 6:10
  a headwind for for larger,
• 6:12
  more tenured organizations.
• 6:13
  And you're right. I we're seeing you know,
• 6:15
  there's there's tons of health care startups
• 6:17
  that are coming to market with, you know,
• 6:20
  digital first
• 6:21
  builds bias. Right? And so their entire organization
• 6:25
  is built on, you know, modern, you know,
• 6:27
  cloud technology
• 6:29
  with kind of the latest and greatest. So
• 6:31
  there is a separation between old and new.
• 6:34
  Yeah. For sure. And let and let's get
• 6:35
  into sort of that,
• 6:37
  I guess,
• 6:38
  crowded,
• 6:39
  I guess, marketplace for for lack of a
• 6:41
  better word. When you when you look at
• 6:43
  all the tools out there, can you talk
• 6:44
  a little bit about the challenge that comes
• 6:47
  when maybe trying to implement different tools or
• 6:50
  point solutions for quoting renewals, population count?
• 6:54
  Is that a challenge? Is that is that
• 6:55
  problematic? And when you look at at at
• 6:58
  the market today in this competitive landscape?
• 7:01
  It is. It is. And
• 7:04
  and I'll kind of answer this from my
• 7:06
  own personal kind of perspective, you know, having
• 7:09
  spent thirty years sitting primarily on the insurance
• 7:12
  carrier side of the the house. Right? I
• 7:13
  know how difficult it is to
• 7:16
  source, onboard, and operationalize
• 7:18
  vendor solutions.
• 7:19
  But when when those solutions are are siloed
• 7:22
  point solution, you know, the value is limited
• 7:25
  often to just one targeted outcome.
• 7:28
  And that holds true for purpose built homegrown
• 7:31
  solutions too.
• 7:32
  Alright. So rarely, if ever, do these solutions
• 7:35
  interact in ways that provide compounding value for
• 7:38
  the organization.
• 7:39
  So each of these sit within their own
• 7:41
  silo. Do they do their own thing, but
• 7:43
  they don't talk to each other. Right? And
• 7:46
  it often you know, the the burden of
• 7:48
  trying to connect
• 7:50
  the pipes or connect the dots between these
• 7:52
  things often falls on the human capital, the
• 7:55
  re the teams, right, that sit within these
• 7:57
  organizations.
• 7:58
  And it becomes a a massive, massive work
• 8:01
  effort to do that. And often it means,
• 8:03
  you know, the the the the potential power
• 8:05
  of, you know, connecting these things is is
• 8:08
  never truly realized.
• 8:09
  And so that you know, and that's particularly
• 8:11
  true in the kind of risk management life
• 8:14
  cycle. And when I say life cycle, I'm
• 8:16
  thinking of quoting new business, managing renewals, and
• 8:20
  ultimately supporting the population
• 8:22
  of members that were in in those plans
• 8:24
  of population health.
• 8:26
  You know, each of these are, you know,
• 8:28
  frankly, incredibly important pieces of the puzzle, and
• 8:31
  they can be fine tuned on their own.
• 8:33
  But the real advantage comes when insights and
• 8:36
  intelligence from each of these compound across that
• 8:39
  life cycle. So I think of this as,
• 8:42
  a compounding engine.
• 8:43
  So where data from each part of the
• 8:45
  cycle is fed back into AI enabled models
• 8:49
  so that each future prediction becomes smarter.
• 8:52
  So sharper, more informed risk decisions translate into,
• 8:56
  you know, accurate pricing, better high cost member
• 8:59
  identification,
• 9:00
  more informed renewal decisions,
• 9:02
  all of which ultimately impact medical loss ratio.
• 9:06
  Yeah. Let let let's talk more a bit
• 9:08
  about that sort of, like, the the benefits
• 9:10
  that you're laying out there in terms of
• 9:11
  using sort of consistent logic, I guess, I'll
• 9:14
  call it, via technology across quoting renewals and
• 9:17
  population health.
• 9:19
  It improves price pricing consistency, but there's also
• 9:21
  the the trust component with employer groups too.
• 9:23
  Right? Can you tell me more about that?
• 9:25
  Exactly. Yeah.
• 9:27
  So I'll I'll I'll start by, you know,
• 9:29
  here's my perspective on on employers. So at
• 9:31
  the end of the day, employers are generally
• 9:33
  looking for
• 9:34
  affordable pricing for a competitive benefit offering for
• 9:38
  their employees. It's pretty straightforward.
• 9:40
  And they'd like that price to remain stable
• 9:42
  and predictable
• 9:43
  so that they can actually effectively manage the
• 9:46
  finance of their business. Right? So on the
• 9:48
  other hand, the health insurance carrier is looking
• 9:51
  to understand everything they can about the risk
• 9:54
  of the group
• 9:55
  to offer that competitive price that, but it
• 9:58
  has to cover expected medical and administrative costs
• 10:01
  with a profit for their business. Right? So
• 10:03
  those are
• 10:04
  the two scenarios, perfect situation, right?
• 10:07
  Unfortunately,
• 10:08
  those underwriting decisions are often done with a
• 10:11
  lack of historical experience for the group
• 10:13
  with little or no insight into the expected
• 10:15
  future experience of that group and in a
• 10:18
  hyper competitive market where they're forced to make
• 10:21
  trade offs between growth and profitability expectation.
• 10:24
  So this typically leads to aggressive first year
• 10:27
  pricing,
• 10:29
  right, to win the business with an expectation
• 10:32
  that profit shortfalls be made up with rate
• 10:34
  action and or benefit design changes at renewal,
• 10:38
  which is fundamentally
• 10:39
  at odds with the employer's desire for stable,
• 10:42
  predictable pricing and benefits.
• 10:44
  So using, you know, an integrated
• 10:47
  risk solution provides,
• 10:48
  you know, the actionable insights for underwriters to
• 10:51
  empower better decision making, which builds and supports
• 10:54
  a trusted relationship with the employer
• 10:57
  while also supporting kind of the financial objects
• 11:00
  objectives of the health insurance organization at large.
• 11:04
  Yeah. The the benefits you're laying out are
• 11:06
  are really clear of what this technology can
• 11:08
  do. But I do imagine for some, a
• 11:11
  point solution can feel maybe less intimidating than
• 11:13
  going in enterprise level AI. But for organizations
• 11:17
  interested in wanting to to make that move,
• 11:19
  to get to enterprise level AI,
• 11:22
  what would you recommend, Mark? Where should they
• 11:24
  get started? But what should they think through
• 11:25
  before jumping in? Yeah. So I I there's
• 11:28
  two kind of three things I would highlight.
• 11:30
  So first,
• 11:31
  it really is important to align on the
• 11:33
  business challenge or challenges that you're trying to
• 11:36
  solve
• 11:37
  and and clean and clearly define the outcomes
• 11:40
  that you wanna see.
• 11:42
  This will help focus the conversation on things
• 11:44
  that matter and avoid trying to boil the
• 11:46
  ocean.
• 11:47
  So to your point, right, it
• 11:49
  it it can be
• 11:51
  daunting. Right? And so so think so spend
• 11:53
  the time on the front end to think
• 11:55
  about
• 11:56
  what is it that we're trying to solve.
• 11:58
  Then second,
• 11:59
  and this may,
• 12:01
  you know, be offending to some people listening
• 12:03
  here, but the, you know, you do need
• 12:06
  to have a candid discussion with the leadership
• 12:08
  team in your organization
• 12:09
  about the ability of your organization to, quote,
• 12:12
  unquote, build this ourselves.
• 12:15
  While it's not impossible,
• 12:17
  I've seen I personally seen many projects fail
• 12:20
  because there was an attempt to build it
• 12:22
  internally.
• 12:23
  So I do challenge and ask, right, organizations
• 12:26
  to to, you know, ask yourselves, right, what
• 12:29
  is it that your business is good at?
• 12:30
  You know, if you're an insurance carrier,
• 12:33
  is your strength building complex models enabled by
• 12:36
  AI that leverage vast data ecosystems
• 12:38
  that exist outside your organization?
• 12:41
  Or is your strength
• 12:42
  assessing and pricing risk?
• 12:44
  Right? Have having those early conversations and making
• 12:47
  sure that your leadership team is aligned on
• 12:49
  the the answers,
• 12:51
  will really
• 12:52
  help you avoid,
• 12:53
  you know, delayed decision making down the line.
• 12:57
  And so I I do encourage you to
• 12:59
  do that. And then third, right,
• 13:01
  assuming, right,
• 13:03
  that, you know, there's recognition that, you know,
• 13:05
  you you may not be the expert and
• 13:07
  and you should think about a partnering type
• 13:09
  relationship. You know, spend the time to find
• 13:12
  a trusted partner who has experience
• 13:15
  and credibility in the space that you're looking
• 13:17
  to impact
• 13:18
  and engage with them early, leverage their insights.
• 13:21
  Right? Use them as a tool to really
• 13:23
  help you think about how you can deliver
• 13:25
  your outcomes
• 13:26
  and then, you know, drive forward.
• 13:28
  Yeah. The the a very honest self assessment
• 13:31
  mark and then really finding that right fit
• 13:33
  out there. Right? The right fit in the
• 13:35
  market, the right partner. Correct? Exactly.
• 13:37
  And and I'm gonna let you go here
• 13:39
  in in just a bit. But one final
• 13:40
  question for you. If you could leave executives
• 13:43
  out there listening with one takeaway
• 13:45
  about AI and underwriting today,
• 13:48
  AI in general, what are your what are
• 13:49
  your thoughts? What would it be? What's the
• 13:51
  one thing you wanna leave our listeners with?
• 13:54
  One thing is so hard to limit myself
• 13:56
  to. But sure. I I boil It's a
• 13:58
  very disciplined question mark. Yeah. It is. It
• 14:00
  is. It it might be a little bit
• 14:02
  more than one thing, but, but sure, I
• 14:04
  I boil it down to this. So first
• 14:05
  of all, the value of AI isn't just
• 14:09
  improve speed and reduce costs.
• 14:11
  It's clarity, clarity of information, right, and how
• 14:14
  that information can be applied,
• 14:16
  right, to change decision making. So when information
• 14:19
  and insights are understandable,
• 14:21
  consistent, and provided earlier,
• 14:24
  your underwriters and risk managers can make better
• 14:26
  decisions
• 14:27
  that they trust.
• 14:29
  And then the goal is more informed and
• 14:32
  responsible decisions.
• 14:33
  Right? And with those more informed and responsible
• 14:36
  decisions,
• 14:37
  you know, you can manage costs
• 14:39
  and profitability
• 14:41
  while supporting the ultimate goal, which is better
• 14:44
  outcomes for your employer customers and their members.
• 14:47
  Right? So that's how I think about the
• 14:49
  power of AI and would encourage you to
• 14:52
  think about it that way as well.
• 14:54
  Well, Mark, we'll leave it there. Thank you
• 14:56
  so much for coming on the podcast and
• 14:57
  and sharing your your insights and your perspective
• 15:00
  with our audience.
• 15:01
  I appreciate it. Thank you so much for
• 15:03
  your time, Brian.
• 15:04
  We also wanna thank our podcast sponsor, Graden
• 15:07
  AI.
• 15:08
  You can tune in to more podcasts from
• 15:09
  Becker's Healthcare by visiting our podcast page at
• 15:12
  beckershospitalreview.com.