Latest / Becker’s Healthcare Podcast / How Ambient AI is Helping Health Systems Address Financial, Operational, and Clinician Pressure

Transcript

How Ambient AI is Helping Health Systems Address Financial, Operational, and Clinician Pressure · Becker’s Healthcare 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 how Ambient
• 0:08
  AI is helping health systems address financial,
• 0:11
  operational,
• 0:12
  and clinician pressure.
• 0:14
  To cover this topic, we're joined by Nikhil
• 0:16
  Baduma,
• 0:17
  the cofounder and CEO of Ambiance Health Care,
• 0:20
  where he leads a team dedicated to empowering
• 0:22
  clinicians and health systems with AI technology,
• 0:25
  drawing on his deep background in machine learning,
• 0:28
  and as author of O'Reilly's Fundamentals of Deep
• 0:30
  Learning, Nikhil brings a unique perspective on how
• 0:34
  technology can drive both clinical and operational outcomes
• 0:37
  across health care organizations.
• 0:39
  Alfred Lumsdane is the chief financial officer of
• 0:42
  Arden Health. He brings more than thirty years
• 0:44
  of experience in health care finance and enterprise
• 0:46
  operational leadership.
• 0:48
  He oversees financial strategy, capital planning, and organizational
• 0:51
  performance with a career spanning executive roles, including
• 0:55
  CFO of Quorum Health and senior leadership positions
• 0:58
  at Sharecare and Tivity Health. Across these roles,
• 1:01
  he has guided organizations through major financial transformations,
• 1:04
  strengthen revenue integrity,
• 1:06
  and improve long term financial resilience.
• 1:09
  Nikhil, Alfred, thank you for being here today.
• 1:11
  Thank you. Thank you. Alfred, let let's begin
• 1:14
  with you. Can you really sort of set
• 1:16
  the stage here for our CFO listeners out
• 1:18
  there?
• 1:19
  What what challenges are health systems facing right
• 1:21
  now, and how is AI positioned to help?
• 1:23
  Can you give us a lay of the
• 1:24
  land here?
• 1:25
  Yeah. At at its most basic,
• 1:27
  you know, health systems are facing, you know,
• 1:29
  a a, I'd say, a confluence of challenges,
• 1:32
  both operational
• 1:33
  and financial at the operational side. You know,
• 1:36
  we have a workforce that is stretched to
• 1:39
  its capacity.
• 1:40
  You know, I'd say stressed out is not
• 1:42
  too strong of a word. Then financially,
• 1:45
  we've got headwinds from challenges from, the payer
• 1:48
  environment.
• 1:49
  Payers have struggled on their finances over the
• 1:52
  last several years, and that has shown up
• 1:53
  in a number of ways as that challenge
• 1:56
  moves downstream
• 1:57
  and gets expressed in things such as,
• 2:00
  significantly
• 2:01
  more denials than we've ever faced as an
• 2:03
  industry.
• 2:05
  That that's a significant overview, the for a
• 2:07
  significant challenge for many health systems when you
• 2:09
  talk about denials and sort of that that
• 2:12
  challenging payer environment.
• 2:14
  From sitting in your seat as CFO
• 2:16
  with with Arden,
• 2:18
  how are you sort of leveraging Ambient,
• 2:21
  right, to to help address some of these
• 2:23
  pressure points and get some relief?
• 2:26
  Yeah. Absolutely.
• 2:27
  You know, historically,
• 2:29
  whenever we would add technology
• 2:31
  to our clinicians,
• 2:33
  it became more of a burden. It didn't
• 2:35
  make their jobs easier.
• 2:37
  It didn't actually make delivering care any easier.
• 2:40
  Technology actually increased the both the amount and
• 2:44
  complexity
• 2:44
  of their jobs. And as a consequence, of
• 2:46
  course, you've seen,
• 2:48
  much higher levels of burnout,
• 2:50
  much higher levels of turnover.
• 2:53
  And then when you look at markets like
• 2:55
  the ones that,
• 2:56
  Arden is in, we're in generally second tier
• 2:59
  urban markets,
• 3:01
  markets like Albuquerque, New Mexico, Tyler, Texas, Tulsa,
• 3:04
  Oklahoma.
• 3:05
  You know, there are, at times, more limited
• 3:08
  work pools from a workforce standpoint. And so,
• 3:11
  you know, managing,
• 3:13
  that burnout is,
• 3:15
  is extremely important.
• 3:16
  And so
• 3:18
  I would say for the first time
• 3:20
  with ambient listening, with implementing the Ambiance product,
• 3:24
  we've seen a tool
• 3:26
  that was specifically
• 3:27
  designed to make the clinical
• 3:30
  jobs easier.
• 3:32
  And we've seen tremendous adoption that, you know,
• 3:35
  I could talk at length about, but,
• 3:37
  you know, I'd say it's the first tool
• 3:39
  in my career
• 3:41
  from a technology
• 3:43
  where the,
• 3:44
  the clinicians have embraced it almost in a,
• 3:47
  say, a manic way. It's been so well
• 3:50
  received.
• 3:51
  And that's and to to your point, Alfred,
• 3:53
  that is very unique when you think about
• 3:55
  perhaps the history of clinician adoption of technology
• 3:58
  solutions that are being brought to them. Was
• 4:01
  there any, I guess, hesitancy in the beginning
• 4:03
  when you think about sort of the history
• 4:06
  of technology integration in in health care?
• 4:09
  I I I imagine there could have been
• 4:11
  some skepticism just based on the past experience
• 4:13
  to your point where technology has been brought
• 4:16
  in to make their jobs ultimately, maybe not
• 4:18
  intentionally,
• 4:19
  but more complicated.
• 4:20
  Absolutely. There always is skepticism and is well
• 4:23
  earned, if you will, because, again, historically, we've
• 4:26
  taken, you know, these highly trained,
• 4:29
  highly credentialed,
• 4:31
  workforce that is, geared towards patient care and
• 4:34
  improving lives and achieving optimal outcomes.
• 4:38
  And, being a little bit facetious, you know,
• 4:40
  ask them to be data entry clerks and
• 4:42
  not practice at the top of license doing
• 4:45
  the things they do,
• 4:47
  which is delivering quality outcomes.
• 4:49
  So, yes, I would say, anytime we've got
• 4:52
  a new tool, it's,
• 4:53
  from a technology, there is a normal skepticism.
• 4:56
  And, again,
• 4:58
  the adoption curve that we've seen with Ambiance
• 5:01
  has been,
• 5:02
  exceptionally high. For us, it's been, nearly 90%,
• 5:07
  across the entire enterprise and across all specialties.
• 5:10
  Right. That those are impressive numbers.
• 5:12
  Nikhil, I wanna come to you now and
• 5:14
  get your perspective.
• 5:15
  Really share how Ambiance approaches this this space.
• 5:19
  And and, of course,
• 5:21
  we're talking more here than simple scribing. Correct?
• 5:24
  The I mean, not not to say that
• 5:26
  that scribing tools can't be
• 5:28
  incredibly beneficial for for clinicians, but this is
• 5:31
  this is something different. Can can you expand
• 5:33
  on that for our listeners?
• 5:34
  Yeah.
• 5:35
  You know, even taking a step back, what's
• 5:37
  what's probably kind of wild for a lot
• 5:39
  of folks
• 5:41
  to to to witness for the last couple
• 5:42
  of years is just how rapidly
• 5:45
  we've become accustomed
• 5:47
  to AI in our daily lives.
• 5:50
  My mom is not a particularly tech savvy
• 5:53
  user, and yet she uses ChatTBT and Cloud
• 5:56
  almost every single day now. And it's one
• 5:58
  of those things where
• 6:00
  even though we've seen AI sort of take
• 6:02
  off across
• 6:04
  our personal
• 6:05
  sort of use of technology,
• 6:08
  more often than not, you step in inside
• 6:09
  of a health system today and it almost
• 6:11
  feels like the organizations
• 6:13
  are largely operating in a pre AI world.
• 6:15
  And yet the the opportunity is so big.
• 6:17
  You know, Alfred talks about this sort of
• 6:19
  workforce crisis.
• 6:21
  Most organizations are also facing a profitability crisis.
• 6:25
  And I I think for us, what we've
• 6:27
  realized is that
• 6:29
  even though we are on this exponential trend
• 6:31
  or is that models are getting smarter and
• 6:33
  smarter and smarter every quarter,
• 6:35
  there's still so much work to do
• 6:38
  to be able to translate that increase in
• 6:40
  intelligence
• 6:42
  to AI technology that's actually usable instead of
• 6:45
  the context of the health system
• 6:47
  and then the tie that usage of technology
• 6:50
  to real value that makes a dent in
• 6:52
  that operating margin crisis that most of these
• 6:55
  organizations are are are navigating right now.
• 6:58
  The first
• 6:59
  part of of that, Brian, I I think
• 7:02
  we think about how do we earn the
• 7:03
  right
• 7:04
  to do the window in front of the
• 7:06
  clinician. And you're right. This is much more
• 7:08
  complicated than just described, partially because doctors are
• 7:12
  doing all sorts of work before they even
• 7:14
  walk into a room preparing for a visit.
• 7:16
  They're navigating all this complexity in the billing
• 7:19
  system of the EHR, thinking about what codes
• 7:21
  to pick.
• 7:22
  They're being chased down by CDI folks to
• 7:25
  make sure that they change their documentation to
• 7:27
  support the codes.
• 7:29
  And all of this work,
• 7:31
  today sort of really eats at the psyche
• 7:34
  of the clinician who went to school to
• 7:36
  take care of people, not to not to
• 7:37
  wrangle with technology and not to become an
• 7:40
  expert in coding and billing. And so in
• 7:42
  many ways, what we think about is how
• 7:44
  do we take AI technology that deeply understands
• 7:47
  the medicine being practiced, understands the workflow of
• 7:50
  the clinician in that specialty area,
• 7:52
  has the full context of the patient, and
• 7:54
  also understands what the payers are looking for
• 7:56
  and what compliance is looking for, what auditors
• 7:58
  are looking for in coding and documentation,
• 8:01
  and how do we make it easy to
• 8:02
  do the right thing in workflow,
• 8:04
  so that doctors can focus on what they
• 8:06
  do best and not the stuff that really
• 8:08
  drains them up their energy.
• 8:09
  Yeah. And and, Alfred, I wanna come back
• 8:11
  to you here in a moment and get
• 8:12
  to some results. But, Nikhil,
• 8:14
  follow-up for you there. Just thinking about how
• 8:16
  to to your point, how fast this technology
• 8:19
  is moving, how advanced it is. It is
• 8:21
  just anecdotally from my personal experience too in
• 8:23
  the work that I do.
• 8:25
  Completely different than than what I was doing
• 8:27
  just before before it takes off. That said,
• 8:30
  a lot in the health care system has,
• 8:32
  you know, not necessarily been designed to move
• 8:34
  fast, I guess, I will say.
• 8:36
  The way I've heard it put before in
• 8:38
  terms of,
• 8:39
  the the the disparity in terms of the
• 8:42
  technology landscape is you've got robots in the
• 8:44
  building and you've got fax machines in the
• 8:46
  building.
• 8:47
  And so, Nikhil, wanna get your take in
• 8:49
  terms of how,
• 8:51
  I guess, you think about Ambiance meeting folks
• 8:53
  where your partners where they are. You know,
• 8:55
  where they're able to move, how fast they're
• 8:57
  able to move no matter how awesome the
• 8:59
  technology is. You kinda gotta meet folks where
• 9:01
  they are. Right?
• 9:02
  Look. I think one of the things what
• 9:03
  we found is that part of the reason
• 9:05
  why it's been hard to diffuse technology across
• 9:08
  the health system in the past is that,
• 9:10
  it came with so much friction.
• 9:12
  And it it it you weren't solving a
• 9:14
  burning enough pain point, and you were not
• 9:16
  solving it well enough for an end user
• 9:17
  for them to want to pull the technology
• 9:19
  out of the institution's hands.
• 9:22
  And I think
• 9:23
  it one of the things that we found
• 9:25
  that's probably the one of the most exciting
• 9:28
  revelations is that if you actually create something
• 9:31
  that's iPhone easy to use in the workflow
• 9:33
  and it solves a a burning pain point,
• 9:36
  the rate at which this stuff gets adopted
• 9:38
  is is shocking.
• 9:40
  I think,
• 9:41
  Alfred talked about sort of the adoption utilization
• 9:43
  numbers at 90% across the enterprise.
• 9:46
  We've seen organization go from zero to that
• 9:49
  level of adoption within months of just saying
• 9:52
  that, hey. We're ready to release the technology
• 9:55
  to our clinicians and make it available.
• 9:56
  I know those are not the numbers that
• 9:58
  most institutions are seeing with AI technology that
• 9:59
  they're buying, and I think part of it
• 9:59
  is there's a massive gap
• 10:11
  very sort of limited ambient listening capability actually
• 10:14
  looks like in the workflow.
• 10:16
  But what we found is if you solve
• 10:17
  that technology gap and that user experience gap,
• 10:19
  it just really takes off like wildfire. That's
• 10:22
  one component of it. I think the other
• 10:23
  component of it is
• 10:25
  this
• 10:26
  like like like implementing an EHR, implementing AI
• 10:29
  does require change management.
• 10:31
  And so we've built this incredible sort of
• 10:33
  care transformation team inside of Ambiance that works
• 10:35
  hand in hand with our partners
• 10:37
  to walk through what are the best practices
• 10:39
  of how you actually roll this technology out,
• 10:41
  achieve adoption, and actually achieve value creation from
• 10:44
  that adoption,
• 10:46
  all the way down to when a clinician
• 10:48
  has an issue sometime in the workflow, our
• 10:50
  median response time is ninety seconds to being
• 10:52
  at the elbow and resolving that issue for
• 10:54
  that clinician. And so we get all those
• 10:56
  pieces right, which is a magical user experience,
• 10:59
  strong change management, and then extremely strong at
• 11:02
  the elbow support in workflow for that clinician
• 11:04
  in real time twenty four seven. And I
• 11:06
  think that's the foundation to get these kinds
• 11:08
  of results.
• 11:09
  Excellent. I appreciate you going going a layer
• 11:11
  deeper there, Nikhil.
• 11:12
  Alfred, coming back to you now, we've touched
• 11:14
  on sort of the the 90% number with
• 11:16
  with adoption, which is excellent.
• 11:18
  Can can you share any other sort of
• 11:20
  results you've seen at Ardent Health since deploying
• 11:22
  Ambiance?
• 11:23
  Yeah. No. Happy to, kinda rattle off some
• 11:25
  of the underlying statistics in terms of what
• 11:28
  we've seen from, you know, at our core
• 11:30
  KPIs. And then more importantly,
• 11:32
  what is that translating into from an impact
• 11:35
  to the organization,
• 11:36
  you know, given the pressures that we talked
• 11:38
  about from, both a financial
• 11:41
  and, a workforce perspective. So, you know, just
• 11:44
  from a KPI perspective,
• 11:46
  we've seen the clinicians who are utilizing Ambiance.
• 11:48
  We've seen an over 44%
• 11:50
  decrease in documentation time,
• 11:54
  56%
• 11:55
  decrease in after hours documentation,
• 11:58
  almost a 20%
• 11:59
  increase in patient base time, so more quality.
• 12:02
  You know, again, clinicians doing what they were
• 12:05
  trained to do.
• 12:06
  95%
• 12:07
  of charts being closed on the same day
• 12:10
  and over 18%
• 12:11
  faster chart closure. So what does that mean?
• 12:15
  You know, this has translated
• 12:17
  directly
• 12:18
  into increased coding accuracy,
• 12:20
  fewer queries,
• 12:22
  more
• 12:23
  defensible claims from an audit perspective, which ultimately
• 12:27
  leads to a stronger and more robust revenue
• 12:29
  cycle.
• 12:30
  So when we really think about this then,
• 12:33
  you know, because, you know, we don't
• 12:35
  at the end of the day, you know,
• 12:37
  we are a mission driven company. And in
• 12:39
  order to execute on our mission, we have
• 12:40
  to have a margin.
• 12:42
  And so we have to have tools that,
• 12:44
  you know, agree to that overall mission.
• 12:48
  And we we don't view ambiance as an
• 12:50
  incremental cost, but actually as, you know, part
• 12:53
  of our rev whole revenue cycle process.
• 12:57
  You know, it's early in our implementation,
• 12:59
  but as you've already heard, we we've got
• 13:01
  almost 90%,
• 13:04
  utilization today.
• 13:05
  Our early analyses would indicate
• 13:08
  over $5,400
• 13:11
  in incremental top line revenue
• 13:13
  from e E and M coding alone,
• 13:17
  Approximately
• 13:19
  point four to point five additional patient visits
• 13:21
  per per day, and we value that at
• 13:24
  approximately an additional
• 13:26
  $8,400
• 13:27
  per per clinician per year.
• 13:30
  We're still,
• 13:31
  of course, you know, attached to all of
• 13:33
  that. There's incremental cost of, you know, providing
• 13:37
  those services of seeing an additional,
• 13:39
  patient,
• 13:40
  or half a patient a day.
• 13:42
  So we're still working through
• 13:44
  the net benefit to the organization, but it's
• 13:46
  actually it's you know, I'm I'm a very
• 13:49
  bottom line driven person. It's very easy to
• 13:52
  see,
• 13:53
  how this is in creating value to the
• 13:55
  organization.
• 13:56
  And it's early. Right? We're talking about,
• 14:00
  outpatient only, you know, when we look at
• 14:02
  the bigger opportunities still, you know, yet to
• 14:04
  be harvested.
• 14:06
  You know, we're really only at the tip
• 14:08
  of the spear.
• 14:10
  Yeah. Fascinating to see those results so early,
• 14:12
  and it it really
• 14:14
  does you know, the what's to come should
• 14:17
  be very interesting, and we look forward to
• 14:19
  checking in on that as you continue to
• 14:21
  have this experience with with Ambiance, Alfred.
• 14:24
  Before we close here, I want you each
• 14:26
  to maybe
• 14:27
  look to the future. As as, Alfred, I
• 14:30
  think your your comment suggests there's a wealth
• 14:32
  of opportunity out there for for this technology
• 14:34
  to come in and make a difference for
• 14:36
  patients and and clinicians.
• 14:38
  So as you look ahead and maybe putting
• 14:40
  on your CFO hat, Alfred, you can you
• 14:42
  can answer this question first. What does the
• 14:44
  future of Ambient Intelligence look like? What's your
• 14:47
  take?
• 14:48
  Yeah. Of course, you know, my say my
• 14:50
  crystal ball is probably no better than anybody
• 14:52
  else's, but it it it's easy to say
• 14:55
  the, you know, again, the persistence of the
• 14:57
  headwinds that we've already talked about
• 14:59
  and how
• 15:00
  we, as CFOs, need to prioritize
• 15:04
  compliance on a proactive basis, creating defensible documentation
• 15:08
  and implementing technologies
• 15:10
  that stabilize
• 15:11
  both the workforce as well as the revenue
• 15:14
  cycle.
• 15:15
  Yeah. I do feel confident that underlying audit
• 15:19
  intensity will rise
• 15:21
  and that platforms like Ambiance that support coding
• 15:24
  and compliance are, you know, critical if we're
• 15:27
  going to, again, deliver on our mission. You
• 15:30
  know, from a next generation, I do see
• 15:33
  Ambient Listening,
• 15:35
  extending and and and really ambiance intelligence,
• 15:38
  not just listening, of course,
• 15:40
  extending into the inpatient realm. And, of course,
• 15:44
  the complexity
• 15:45
  is incrementally or exponentially
• 15:47
  higher on the inpatient side, and that actually
• 15:50
  then will yield the greatest,
• 15:52
  potential opportunity to harvest.
• 15:55
  Excellent. And, Nikhil, what what what's your take?
• 15:58
  Oh, how what would you add to what
• 15:59
  Alfred said? What do you what do you
• 16:01
  think where where are you thinking, in terms
• 16:03
  of what's next for
• 16:04
  for for Ambient technology, but for Ambient and
• 16:07
  for for partners like Alfred?
• 16:09
  No. I think, Alfred hit the nail on
• 16:11
  the head. When you've built a machine that
• 16:14
  can deeply understand
• 16:15
  every single one of the patients in in
• 16:17
  your population because you can analyze all of
• 16:20
  that past context that's loaded up in the
• 16:22
  EHR and
• 16:24
  unstructured and structured data, when you can listen
• 16:27
  to a conversation and reason about the medicine
• 16:29
  that's happening in that room,
• 16:30
  when you can distill the expertise
• 16:33
  of your your back office experts, your CDI
• 16:36
  analysts, your prior auth folks, your advisers,
• 16:40
  and you can then take that intelligence and
• 16:42
  you can distribute it as far upstream as
• 16:45
  possible so you can make doing the right
• 16:47
  thing easy and make doing the right thing
• 16:49
  obvious,
• 16:50
  proactively at the point of care as opposed
• 16:52
  to retroactively.
• 16:54
  In many ways,
• 16:55
  that sort of creates a lot of the
• 16:57
  foundational stability and opportunity that Alfred's talking about.
• 17:00
  I think the other opportunity that we've started
• 17:03
  to discuss is
• 17:05
  if you think about your organization
• 17:06
  and you've got a really deep understanding of
• 17:09
  your population
• 17:10
  and
• 17:11
  what they need, and you have a really
• 17:13
  deep understanding of your physical infrastructure
• 17:16
  and understanding of your clinicians and what they
• 17:18
  practice and what they're good at, the schedules
• 17:20
  and the work queues, there's also an opportunity
• 17:23
  to extend sort of the second order and
• 17:26
  third order opportunities from this intelligence layer,
• 17:29
  including
• 17:30
  how do we orchestrate
• 17:31
  care effectively across the system and solve for
• 17:33
  throughput and access at a system level? Because
• 17:36
  you've got a self learning AI machine that's
• 17:38
  able to solve this logistical optimization across the
• 17:41
  entirety of the population, the entirety of the
• 17:43
  physical infrastructure of the institution.
• 17:45
  And so in in some ways, we think
• 17:47
  once you once you've built this infrastructure, you've
• 17:49
  got an endless set of opportunities
• 17:52
  to both deliver better care
• 17:54
  with a better experience, but also at a
• 17:56
  lower cost and greater efficiency than we've ever
• 17:58
  done before. And so I think, you know,
• 18:00
  the Ardent team and the Ambiance team feel
• 18:02
  like we're barely 5% of the way through
• 18:05
  everything that is possible to create from the
• 18:07
  foundation that we've built together.
• 18:10
  That's exciting to hear. And, again, all of
• 18:12
  it says, excited to hear more as you
• 18:14
  you continue to build this foundation
• 18:16
  and build upon it. Nikhil, Alfred, thank you
• 18:18
  so much for coming on the podcast today.
• 18:20
  Happy to. Thank you. Thank you. We also
• 18:23
  wanna thank our podcast sponsor, Ambiance Health Care.
• 18:26
  You can tune to more podcasts from Becker's
• 18:28
  Health Care by visiting our podcast page at
• 18:30
  beckershospitalreview.com.