Latest / Becker’s Healthcare - Clinical Leadership Podcast / Scaling Generative AI for Safer, Smarter Care with Dr. Mark Mabus

Transcript

Scaling Generative AI for Safer, Smarter Care with Dr. Mark Mabus · Becker’s Healthcare - Clinical Leadership Podcast

• 0:02
  The most important health care decisions don't happen
• 0:05
  in isolation.
• 0:06
  They happen when leaders come together.
• 0:08
  Becker's sixteenth annual meeting brings together more than
• 0:11
  3,500
• 0:12
  hospital and health system executives this April in
• 0:15
  Chicago.
• 0:16
  With 800 speakers from Ascension, Cleveland Clinic, CommonSpirit
• 0:20
  and more, the conversations get real. Leaders will
• 0:23
  share how their scenario planning for policy shifts,
• 0:26
  breaking through value based care barriers, and building
• 0:28
  clinical teams that translate new ideas into real
• 0:31
  world care.
• 0:32
  Join top decision makers in the room April
• 0:35
  13 through the sixteenth.
• 0:37
  For the agenda and event details, visit beckershospitalreview.com
• 0:41
  and click on the events tab in the
• 0:43
  upper right.
• 0:45
  This is Laura Dirda with the Becker's Healthcare
• 0:47
  podcast.
• 0:48
  I'm thrilled today to be joined by doctor
• 0:50
  Mark Mabus, chief medical informatics officer at Parkview
• 0:53
  Health. Doctor Mabus, it's a pleasure to have
• 0:54
  you on the podcast today. Thank you so
• 0:56
  much, Laura. It's a a great opportunity to
• 0:59
  be here, and I'm excited to share with
• 1:02
  the community.
• 1:03
  Oh, absolutely. And and this is gonna be
• 1:05
  a fun conversation because I know there's so
• 1:07
  much you're you're doing right now at Parkview.
• 1:09
  Some really cool things in generative AI and
• 1:12
  more. And so I know we'll dive into
• 1:14
  some of that. But before we do, can
• 1:15
  you introduce yourself? Just tell us a little
• 1:17
  bit more about yourself and Parkview Health.
• 1:20
  Sure. Sure.
• 1:21
  Like you said, I'm doctor Mark Mabus. I'm
• 1:23
  a practicing
• 1:24
  family physician. Still see patients, a day a
• 1:27
  week.
• 1:28
  CMIO at Parkview. Parkview is a large,
• 1:31
  not for profit, health system in Northeast Indiana,
• 1:34
  Northwest Ohio with multiple hospitals,
• 1:36
  a large employed medical group,
• 1:39
  a strong focus on
• 1:40
  community health.
• 1:42
  We have we have urban centers, suburban centers,
• 1:45
  and rural,
• 1:46
  all within
• 1:47
  our, area there too.
• 1:50
  My role,
• 1:51
  you know, I I sit at the intersection
• 1:53
  of, you know, clinicians and technology
• 1:56
  and then governance too. You know, the,
• 1:59
  application analyst for the EHR report to me,
• 2:01
  but I'm also in charge of, you know,
• 2:03
  provider efficiencies there too. So making sure that
• 2:06
  EHR,
• 2:07
  AI, which we'll get into, and all these
• 2:10
  digital tools, you know, create efficient workflows,
• 2:13
  safer care for patients, and quality pair care
• 2:17
  for patients.
• 2:18
  You know, it it's a it's a fun
• 2:20
  job to be in between all of that
• 2:22
  and be able to see all the different
• 2:24
  areas of care throughout our health system.
• 2:28
  That that's, you know, so cool and definitely
• 2:31
  an interesting place to be having that clinical
• 2:33
  background, the technology, and and bringing that all
• 2:35
  together.
• 2:36
  You know, it is truly a a view
• 2:38
  of health care that not many others have.
• 2:40
  Now I I'm curious. When you think about
• 2:42
  the last year or so, could you tell
• 2:43
  us about the most important initiative that you
• 2:45
  led? What did you do, and what were
• 2:47
  the result?
• 2:48
  I think our most
• 2:50
  important initiative
• 2:52
  last year was getting
• 2:54
  assistive AI, you know, generative AI,
• 2:57
  deployed throughout our health system. We did that
• 3:00
  in a couple different ways. You know, the
• 3:03
  ambient documentation
• 3:05
  was a big one for our providers.
• 3:07
  You know, we, partnered with a vendor to
• 3:09
  provide that service,
• 3:11
  through Epic,
• 3:12
  and we use a lot of Epic's generative
• 3:14
  AI tools that they have available as well.
• 3:18
  And in fact, we just got our recent,
• 3:20
  class research arts collaborative survey that that said
• 3:23
  that 88%
• 3:24
  of our providers
• 3:26
  also said
• 3:27
  that the best improvement
• 3:29
  in 2025
• 3:31
  to our EHR was the deployment of generative
• 3:34
  AI in some form of an or another,
• 3:36
  like that assistive AI. We moved from limited
• 3:40
  pilots
• 3:41
  to broad access, you know, and made sure
• 3:43
  that we were able to deploy these tools
• 3:46
  across all areas of care delivery, you know,
• 3:48
  ambulatory, emergency department, and inpatient. And at the
• 3:52
  same time, we were very
• 3:54
  intentional about how we message this to providers.
• 3:57
  You know, this was
• 3:59
  the advent of assistive
• 4:01
  AI. We're not we're not getting into automation
• 4:03
  and, you know, computers making diagnoses for you.
• 4:06
  Nothing like that. Nothing going to replace clinical
• 4:08
  judgment.
• 4:09
  It's kind of, you know, building that trust
• 4:12
  with with your providers. Hey. We've got these
• 4:14
  new tools that can help you be more
• 4:15
  efficient.
• 4:16
  And then, you know, deploying that out, you
• 4:19
  know, with the with piloting first, we're able
• 4:21
  to make sure it works for our workflows,
• 4:24
  then we have the the mass adoption. So,
• 4:27
  very early in 2025,
• 4:30
  we had,
• 4:31
  let our, ambient
• 4:34
  documentation
• 4:35
  be available to all of our providers.
• 4:37
  And then in October
• 4:39
  2025,
• 4:40
  we made over 18 generative AI features available
• 4:44
  in Epic to all of our providers
• 4:47
  as well.
• 4:48
  So looking at, you know, different areas where
• 4:50
  we can deploy AI and do so
• 4:54
  at mass scale. You know, we prepped it
• 4:56
  with all these little pilots,
• 4:58
  especially using some of our super users,
• 5:02
  our, what Epic calls physician builders, you know,
• 5:04
  to help us with testing and feedback,
• 5:07
  and then getting it,
• 5:09
  you know, out to the masses
• 5:11
  that I
• 5:12
  you know, specific points in time and bang,
• 5:15
  big bang. I'll I'll give you another statistic
• 5:18
  with the generative AI features like I mentioned
• 5:20
  that we, deployed in October. We were running
• 5:23
  at about, you know, 60,000 tokens per month
• 5:26
  or something along those lines. In October, since
• 5:29
  we turned everything on, that jumped to 2,000,000
• 5:32
  tokens in that month. And just this past,
• 5:35
  January 2026,
• 5:37
  we hit 3,000,000
• 5:39
  tokens. So it was like, yeah, just a
• 5:41
  little bit here and there, you know, getting
• 5:43
  our pilots underground and then, you know, throughout
• 5:45
  the whole health system, these things are available.
• 5:48
  People are using them, you know, not just
• 5:50
  our providers. You know, there's there's obviously things
• 5:53
  that help our providers, but, you know, we
• 5:54
  also deployed in coding and revenue cycle,
• 5:58
  side of our nonclinical folks can benefit from
• 6:01
  AI. But, yeah, planning those things out took
• 6:04
  a lot of work, and then you can
• 6:06
  go
• 6:07
  at scale and get these in the hands
• 6:09
  of everybody.
• 6:11
  Yeah. Generative AI is the thing. It's it's
• 6:13
  what everybody needs to be doing across their
• 6:16
  health care system to, to enable these tools
• 6:19
  for for efficiency,
• 6:21
  ROI, all those things. You know? It it's
• 6:24
  it's here. Use it.
• 6:27
  I love that. You know, such a a
• 6:28
  clear call for having this technology and doing
• 6:31
  what you can do in order to make
• 6:33
  the most of it. And it seems like
• 6:35
  you've got a really
• 6:36
  strong process,
• 6:37
  not only for, you know, having some of
• 6:39
  those tests and those pilots, but then bringing
• 6:41
  things to scale. And I think that's a
• 6:42
  place where a lot of health systems and
• 6:44
  organizations really struggle to, you know, do that
• 6:47
  effectively.
• 6:49
  So
• 6:50
  what have you done or what have you
• 6:51
  found that, you know, worked well when you
• 6:52
  are taking some things from, you know, those
• 6:54
  smaller pilots
• 6:55
  into a a broader and scalable space and
• 6:58
  seeing this kind of, like, adoption, which is
• 7:01
  wild.
• 7:02
  A lot of it is figuring out workflows
• 7:05
  and where this fits properly. You know, you
• 7:07
  can have your vendors give you,
• 7:10
  the instruction manuals or or, you know, how
• 7:13
  this actually works,
• 7:15
  in technically,
• 7:17
  you know. But when you dive into the
• 7:19
  workflows of an actual user and start using
• 7:23
  these things, you get to figure out that,
• 7:26
  maybe it's not so cut and dry out
• 7:28
  of the instruction manual. You've got to create
• 7:31
  those best practices.
• 7:32
  That's what pilots were able to do for
• 7:35
  us on these in in these scales. And,
• 7:38
  you know, it does take you
• 7:42
  adopting the right people for your pilots. I
• 7:45
  mentioned that we used physician builders, but, you
• 7:48
  know, those physician builders are also in tune
• 7:50
  with,
• 7:52
  non savvy users' workflows too.
• 7:55
  There's kind of that spectrum. You got your
• 7:57
  your, early adopters and people who do great
• 8:00
  with anything that you throw at them technology
• 8:02
  or change,
• 8:04
  management wise. Got your middle of the road
• 8:06
  folks, and then you got your your,
• 8:09
  we'll just call them slower folks, you know,
• 8:10
  that that, a little more resistant to change
• 8:13
  or,
• 8:14
  you you know, you know, the the ones
• 8:16
  that take some extra education. So you've gotta
• 8:18
  keep all those different areas in mind. And
• 8:20
  so when we are developing best practices,
• 8:34
  practice.
• 8:35
  We use our informatics teams to write those
• 8:38
  best practices up, create videos,
• 8:42
  you you know, document in in,
• 8:44
  educational emails
• 8:46
  or in, you know, in workflow
• 8:49
  education as well. We utilize in basket messaging
• 8:53
  for a lot of our informatics type and
• 8:55
  workflow education,
• 8:56
  because people actually look at their in baskets,
• 8:59
  in in the EHR rather than having to
• 9:02
  go out to, to, you know, their email
• 9:04
  program to look at it that way.
• 9:06
  So connecting with providers, getting boots on the
• 9:09
  ground, our informatics team knew that these were
• 9:11
  high initiatives for the year. So they actually
• 9:14
  scaled back some of their just kind of
• 9:17
  regular
• 9:18
  scheduled clinic supports and instead focused on these
• 9:21
  high level deployments, whether it was ambient,
• 9:24
  documentation
• 9:26
  or, you know, October when all 18 of
• 9:28
  these features went live. They made sure they
• 9:30
  scheduled time in each of the departments
• 9:32
  to go over
• 9:34
  the features that applied to the users in
• 9:37
  those departments.
• 9:38
  So so it's possible to get this out
• 9:41
  to all those users, but it is a
• 9:44
  lot of coordination,
• 9:45
  a lot of effort,
• 9:46
  and, you know, kinda all hands on deck
• 9:49
  when these things do go live. Think of
• 9:52
  it like a mini EHR go live. You
• 9:55
  wanna be as prepared as possible because these
• 9:58
  are high impact tools
• 10:00
  and kind of a different type of tool
• 10:03
  that they've had available to them than in
• 10:05
  the past.
• 10:07
  So we made these, you know, kind of,
• 10:09
  our two big key points, you know, ambient
• 10:12
  documentation and then the big bang of generative
• 10:15
  AI.
• 10:16
  High points for the year, big focuses for
• 10:19
  those months that though that happened, and I
• 10:21
  think that targeted approach, targeted education, and targeted
• 10:25
  feedback, you know, being able to take that
• 10:27
  back to our vendors,
• 10:28
  and have improvements made from there, that all
• 10:31
  led to
• 10:33
  very successful deployment
• 10:35
  of of these tools.
• 10:37
  Got it. And that makes a lot of
• 10:39
  sense and is, you know, really helpful to
• 10:40
  understand that mindset and how you go into,
• 10:43
  understanding where different people are at, along the
• 10:46
  adoption scale and then, you know, looking at
• 10:48
  those workflows and figuring out how to communicate
• 10:50
  with them and and having AI really be
• 10:53
  supportive,
• 10:54
  to the things that they're doing on a
• 10:55
  daily basis. Mhmm. Now something else you mentioned
• 10:58
  there too is thinking about the value to
• 11:00
  the health system and the ROI. How do
• 11:03
  you think about that? How you're measuring that?
• 11:05
  I I think, you know, it's, one thing
• 11:07
  to look at some of the soft ROI,
• 11:09
  time saved, and everything else. But, you know,
• 11:11
  when you're thinking about the big investments that
• 11:13
  you have to make in order to get
• 11:14
  this right,
• 11:15
  what makes the most sense for you when
• 11:17
  you're trying to figure out, hey. Is this
• 11:18
  actually,
• 11:19
  doing what it needs to do? And and,
• 11:21
  you know, what is that return on investment?
• 11:24
  For sure with today's, financial climate and uncertainty
• 11:28
  in health care, you do want to be
• 11:30
  a good steward of the resources that you
• 11:32
  are provided.
• 11:33
  The kind of initial approach that we took,
• 11:35
  you know, with with the ambient documentation,
• 11:37
  we knew there could potentially be a financial
• 11:40
  ROI with that. Yes. There are soft ROIs
• 11:43
  with, like you mentioned,
• 11:45
  decreased time in notes, decreased time in the
• 11:47
  EHR,
• 11:48
  more time with patients, decreased burnout. Those are
• 11:51
  all kinda soft and hard to quantify,
• 11:55
  financially
• 11:56
  there too. But if you have, you know,
• 11:58
  say, an extra fifteen minutes,
• 12:01
  were you actually, you know, using that fifteen
• 12:03
  minutes a day to throw another patient on?
• 12:06
  You know, that those questions were, you know,
• 12:09
  of course, big on our our finance side.
• 12:12
  And, actually, with some of our pilots, we
• 12:14
  did require,
• 12:15
  that,
• 12:16
  our, pilot users, if they were gonna use
• 12:18
  the technology to help cover the cost, you
• 12:20
  know, that they would have an additional patient
• 12:22
  per per,
• 12:24
  clinic day there.
• 12:26
  But as we went on and saw some
• 12:28
  of the other benefits of this, a little
• 12:30
  bit lower provider turnout rate and definitely decreased
• 12:34
  burnout metrics that we were seeing, you know,
• 12:36
  we did become a little more
• 12:38
  lax on on that requirement, and I know
• 12:41
  a lot of other health systems have done,
• 12:43
  the the same thing there. With some of
• 12:45
  the generative AI functionalities,
• 12:47
  I mentioned the nonclinical
• 12:48
  side of things, encoding revenue cycle.
• 12:51
  Those are ones that are,
• 12:54
  kind of, I would say, no brainers to
• 12:56
  implement. You know, that is something that can
• 12:59
  absolutely,
• 13:00
  potentially get you a financial ROI. We even
• 13:03
  have one for our, providers that helps them
• 13:06
  calculate their,
• 13:07
  standard level of service e and m codes
• 13:09
  based on medical decision making. Were they coding
• 13:12
  things properly with, mild, moderate, or high,
• 13:16
  levels of complexity there? They have a little
• 13:19
  calculator that AI can read their note and
• 13:21
  tell them, you know, which
• 13:23
  which level of service to code. So that
• 13:25
  brought a little bit of, financial
• 13:27
  ROI
• 13:28
  there as well.
• 13:30
  The inpatient folks,
• 13:32
  you know, being able to get their discharge
• 13:34
  summary as done
• 13:36
  faster too, you know, so they could, move
• 13:38
  on to to see their their patients. So
• 13:41
  finding,
• 13:43
  those
• 13:44
  key
• 13:46
  AI functionalities
• 13:47
  where you absolutely
• 13:49
  can see a financial
• 13:51
  ROI is kinda key
• 13:53
  to getting things started.
• 13:55
  And then those help to pay for, you
• 13:57
  know, some of the other,
• 13:59
  functionalities
• 14:00
  that may have
• 14:02
  more of that soft ROI
• 14:05
  based more on efficiency and time savings.
• 14:09
  I can use some of these to pay
• 14:10
  for those. That was kind of the mentality
• 14:13
  that that we used. It's also nice to
• 14:15
  have, you know, for example,
• 14:18
  with,
• 14:19
  Epic Systems, you can have a,
• 14:21
  monthly fee for
• 14:23
  many, many different generative I, generative AI functionalities.
• 14:27
  So you know that you are,
• 14:30
  you know, getting your money's worth by having
• 14:33
  the revenue cycle side to help pay for
• 14:35
  the clinician,
• 14:36
  side of the,
• 14:39
  AI functionalities there too. So,
• 14:42
  focusing on I guess I could mention that
• 14:44
  too. You know, focusing on your key strategic
• 14:46
  vendors that you already have, there are some
• 14:49
  vendors that health systems, you know, pay 7
• 14:51
  figures or more to that offer embedded
• 14:55
  generative AI now. Why not
• 14:57
  use those as opposed to
• 15:01
  searching out a third party? You know, that
• 15:03
  is a,
• 15:04
  that's a niche area. You know?
• 15:07
  Get your money's worth for what you're for
• 15:09
  what you're doing. I think with financial headwinds
• 15:12
  this year, there's gonna be a lot of
• 15:13
  health systems that have to think that way.
• 15:16
  You know, you know, if they do have
• 15:18
  to prioritize things,
• 15:19
  prioritize what is already
• 15:22
  included or included with a minimal or a
• 15:26
  bump in, you know,
• 15:28
  licensing fees
• 15:29
  as opposed to, you know, a huge investment
• 15:32
  in something else. You know, we we, we
• 15:35
  have to
• 15:36
  weigh the options
• 15:38
  of each of those and what works best
• 15:40
  for our system at that time, what works
• 15:42
  best for
• 15:44
  that service line or service area at that
• 15:47
  time.
• 15:48
  There's there's a lot of different ways to
• 15:50
  go about it while still being, you know,
• 15:52
  financially responsible.
• 15:54
  That's helpful to understand. Thank you so much
• 15:56
  for providing a bit of extra context there.
• 15:58
  And, you know, it seems like,
• 16:00
  a very,
• 16:01
  important thing to be thinking about as you're
• 16:04
  looking into the future. And speaking of that,
• 16:05
  what are some of your big priorities and
• 16:07
  headwinds that you're focused on for 2026?
• 16:10
  Oh, for sure, it's it's the scale
• 16:13
  of AI, you know, not just,
• 16:16
  getting,
• 16:17
  you know, things available to all our providers
• 16:19
  because it's really you know, even though it's
• 16:21
  available to all of our providers, not all
• 16:24
  of them are are using it. I'll be
• 16:25
  I'll be honest.
• 16:27
  We have about a third of our providers
• 16:29
  who are in that high user category that
• 16:32
  we will,
• 16:33
  you know, see broad adoption of. We have
• 16:35
  another third that are middle of the road
• 16:37
  where they'll use some but not the others,
• 16:39
  and another, third that seem to be a
• 16:41
  little bit more disengaged. So our focus,
• 16:44
  is going to be this year on the
• 16:46
  education of that middle and lower,
• 16:49
  thirds. And maybe we can hit, you know,
• 16:51
  fifty fifty or sixty forty,
• 16:53
  you know, this year. I think,
• 16:55
  education
• 16:56
  and, you know, perhaps even one on one
• 16:58
  training on some of these things may help
• 17:01
  us
• 17:02
  move that needle, for clinicians
• 17:05
  and how they are,
• 17:06
  utilizing this AI. Another big area is to
• 17:10
  be able to, you know,
• 17:13
  properly inventory
• 17:14
  and document
• 17:16
  now that all these
• 17:17
  AI functionalities
• 17:19
  are live,
• 17:20
  you know, you you gotta make sure there's
• 17:22
  ongoing review, make sure there's not bias or
• 17:25
  drift that happens with with some of the
• 17:28
  models, you know, making sure that these tools
• 17:31
  are still functioning
• 17:32
  in a safe and high quality
• 17:35
  manner.
• 17:36
  We're doing some of these things through our,
• 17:38
  what we call AI steering committee. You know,
• 17:40
  every health system should have some sort of
• 17:42
  AI governance. We call it a steering committee,
• 17:44
  and that is overseeing and directing some of
• 17:46
  these, tools. As we're waiting on, different, you
• 17:50
  know, regulatory bodies or legislation
• 17:52
  to, you know, specifically
• 17:54
  say things, be proactive and, you know, start
• 17:57
  audits yourself and keep track of all the
• 17:59
  functionalities that you already have
• 18:01
  live. So it's it this year is the
• 18:04
  age of continuing education for users. It's the
• 18:07
  age of
• 18:08
  inventories, libraries, and, feedback mechanisms and and upkeep,
• 18:13
  you know, all while,
• 18:15
  weighing additional options with not just the generative
• 18:19
  AI solutions, but
• 18:20
  agentic and,
• 18:22
  even perhaps autonomous in in the coding arena.
• 18:26
  Nonclinical,
• 18:27
  definitely best to go nonclinical for automation there
• 18:30
  first as we're waiting on, some of the
• 18:33
  other things to happen with, regulatory
• 18:35
  there too.
• 18:37
  But,
• 18:38
  getting getting that groundwork so you're prepared to,
• 18:42
  you know, take on
• 18:44
  additional functionalities
• 18:45
  or more advanced functionalities
• 18:48
  while keeping in the back of your head,
• 18:50
  you know,
• 18:51
  how do I educate my users? How do
• 18:53
  I ensure safety, quality,
• 18:55
  and,
• 18:56
  regulatory compliance
• 18:58
  as all these tools are are there?
• 19:01
  Yeah. That makes a lot of sense. And
• 19:02
  I know it's always that fine line between
• 19:04
  wanting to be innovative, continuing to push the
• 19:07
  boundaries, but also not taking on too much
• 19:10
  risk within the system in the health care
• 19:12
  space too. Yes. Yes. Absolutely.
• 19:15
  Well, what do you think the hardest thing
• 19:16
  you're gonna have to do in the coming
• 19:17
  year will be?
• 19:20
  I I think it's saying no.
• 19:23
  Actually, you know, because of that pacing, you
• 19:25
  know, are there things that
• 19:28
  need to
• 19:29
  need to wait? In, you know, my health
• 19:31
  system in particular,
• 19:32
  you know, we're one of the one of
• 19:35
  the leading medium sized health care organizations
• 19:38
  in the country. You know, we don't have
• 19:42
  unlimited resources.
• 19:43
  So what makes sense for us, you know,
• 19:46
  may be
• 19:48
  something that is provided through an existing vendor,
• 19:50
  like I was saying, rather than,
• 19:52
  some new latest and greatest things. What can
• 19:55
  we take on,
• 19:57
  throughout all of this?
• 19:59
  It may we may have to say no
• 20:00
  to focus on instead
• 20:03
  upscaling what we already have, educating
• 20:06
  who we already support, and things like that.
• 20:10
  We don't want AI to overwhelm
• 20:13
  all of our,
• 20:14
  users,
• 20:16
  but we do wanna be transparent with what's
• 20:18
  available,
• 20:19
  how we can support things now,
• 20:21
  things like that. It's not
• 20:23
  it's yeah. I guess it's I guess you
• 20:25
  could call it pacing yourself. You know?
• 20:28
  Yes. We we deployed all this stuff. We
• 20:30
  we got things going.
• 20:32
  Can we maintain that pace? We'll we'll see.
• 20:34
  We have to balance the yeses with the
• 20:36
  noes.
• 20:37
  Absolutely. And I I know that can be
• 20:39
  certainly no small task, especially with the volume
• 20:42
  of things that are coming at you and
• 20:44
  your colleagues on a daily consistent basis. And
• 20:47
  so,
• 20:47
  yeah, I I know you've mentioned being able
• 20:49
  to, you know, kind of, like, make sure
• 20:51
  you've got, things covered with your existing vendors
• 20:54
  and partners along those lines. And so, you
• 20:56
  know, when you do have, folks that are
• 20:58
  excited about something new or or coming through,
• 21:01
  how do you,
• 21:02
  communicate, you know,
• 21:04
  and identify the things that are gonna actually
• 21:05
  make a difference,
• 21:07
  and and decide on taking a new partner
• 21:09
  when that, situation arises?
• 21:11
  That's one of the ways that our AI,
• 21:14
  governance helps out as well. We do have
• 21:17
  a,
• 21:18
  an update to our our vendor intake forms.
• 21:21
  You know? Everybody's got their own way of
• 21:24
  of getting,
• 21:25
  new technology or new services into a health
• 21:28
  system.
• 21:29
  So embedded within that, we now created our
• 21:32
  own, AI review. The AI,
• 21:35
  does focus on generative agentic or autonomous.
• 21:39
  We have other mechanisms in place for, you
• 21:41
  know, machine learning and some of those, quote,
• 21:43
  older forms of, AI.
• 21:47
  And one of those,
• 21:49
  questions, you know, that's on there, you know,
• 21:51
  first off, what's what's the need? You know,
• 21:53
  why are you looking at AI for for
• 21:55
  this?
• 21:55
  And then another question is,
• 21:58
  does
• 21:59
  our what we call trusted vendors, you know,
• 22:02
  the those vendors that you spend a lot
• 22:04
  of money on and you better be getting
• 22:05
  your money's worth from, do they offer a,
• 22:08
  similar technology and have you explored that? And
• 22:11
  there's always that. If not, please discuss with
• 22:14
  information services or informatics,
• 22:16
  because they're, you know, up on this technology.
• 22:19
  They're in the know of what our,
• 22:22
  vendors are bringing. They know the road maps
• 22:24
  of where our vendors are going.
• 22:27
  And so that is that point of entry,
• 22:30
  you know, for for our vendor,
• 22:32
  I think, if it hadn't happened before then,
• 22:36
  it gets caught in that process, and it
• 22:38
  allows our, informatics and IS teams to take
• 22:41
  that opportunity
• 22:42
  to work with the requesters
• 22:45
  and, you know, show show features that are
• 22:47
  available. And our,
• 22:49
  our our vendors have plenty of documentation on
• 22:52
  that or, you know, happy to jump on
• 22:54
  calls to explain some of those things. So,
• 22:57
  embedding that in a
• 23:00
  review process or that formal intake to get
• 23:03
  your budget dollars approved,
• 23:06
  that is
• 23:07
  our way of,
• 23:09
  making sure everybody knows what's going on.
• 23:13
  Got it. That's helpful to understand. Well, before
• 23:15
  we wrap up here, I wanted to ask
• 23:16
  you about growth. Where do you see some
• 23:17
  of the best opportunities for organizational growth in
• 23:19
  the future, especially considering
• 23:21
  all the the different use cases for AI
• 23:23
  and how quickly the technology is evolving?
• 23:26
  AI itself can bring,
• 23:29
  so many efficiencies there, you know, to give
• 23:31
  time back to our providers,
• 23:34
  our nurses, our care teams. You know, one
• 23:37
  of the one of the top reported
• 23:41
  benefits of using
• 23:43
  AI charting
• 23:44
  from our providers has been I get to
• 23:46
  spend more time with the patients.
• 23:48
  And that's what medicine
• 23:51
  is all about, and we've thrown so many
• 23:53
  other, you know,
• 23:55
  hoops to jump through just to be able
• 23:58
  to deliver care to patients. So I'm very
• 24:01
  excited and glad that,
• 24:03
  AI charting has been able to provide that
• 24:06
  connection
• 24:07
  back with our providers. And as we take
• 24:09
  things live with our nurses,
• 24:12
  this year as well, our nurses can spend
• 24:15
  more time with their patients too. So giving
• 24:18
  time back, that it that is precious and
• 24:22
  is so helpful for growth
• 24:25
  for our, you know, employees,
• 24:27
  there too, our users of of this technology
• 24:30
  of AI. You know, the same thing can
• 24:31
  happen on the on the nonclinical side of
• 24:34
  things, Getting administrative tasks done faster,
• 24:38
  getting, through that coding work queue a little
• 24:41
  faster,
• 24:42
  that will will help in their efficiencies
• 24:45
  too. As a as a health system,
• 24:48
  we are,
• 24:49
  blessed with organizational
• 24:51
  growth as well where we are expanding
• 24:53
  our,
• 24:54
  reach throughout the state, able to support,
• 24:58
  especially some of our rural hospitals
• 25:00
  in the state of Indiana,
• 25:02
  with our technology too offering
• 25:05
  through our,
• 25:07
  EPIC,
• 25:08
  Connect program
• 25:09
  as well,
• 25:10
  and also some managed services agreements.
• 25:13
  Rural health is very important and
• 25:17
  is very much struggling,
• 25:20
  and and so we're happy to be able
• 25:22
  to provide services there. You know, some of
• 25:24
  this this AI technology,
• 25:26
  you gotta think that there is some,
• 25:29
  equity concerns going on too. How can a
• 25:32
  small,
• 25:35
  hospital,
• 25:36
  or maybe FQHC
• 25:38
  clinic, you know, afford
• 25:40
  some of these,
• 25:43
  technology tools?
• 25:45
  And, you know, there are,
• 25:47
  larger health system partners that can help out
• 25:49
  in that area too.
• 25:52
  There is some growing change at least at,
• 25:55
  in Indiana's legislative level to also,
• 25:59
  offer assistance,
• 26:00
  to critical access hospitals and and others
• 26:03
  that may need access to these tools too.
• 26:06
  Because, you know, the
• 26:08
  the technology
• 26:09
  will help promote
• 26:10
  safer and higher quality care.
• 26:13
  All
• 26:14
  areas of care delivery
• 26:16
  should have equitable
• 26:18
  access to these tools. And I'm happy to
• 26:20
  be a part of a system that wants
• 26:23
  to share that, you know, with these, other
• 26:26
  areas that may, you know, have a bigger
• 26:29
  bit bit of a struggle to to get
• 26:30
  there.
• 26:32
  Absolutely. I think that makes a lot of
• 26:33
  sense. And, certainly, you know, having that ability
• 26:36
  to support other facilities,
• 26:37
  really keeps that care local and make sure
• 26:40
  everybody has access to what they need. And
• 26:42
  so, I think that's very much a point
• 26:44
  well taken in thinking through how quickly technology
• 26:47
  is changing. There's so much possibility and potential.
• 26:49
  It seems like a a really exciting time
• 26:51
  in health care right now.
• 26:53
  100%
• 26:53
  agree.
• 26:55
  Absolutely. Well, doctor Mavis, thank you so much
• 26:57
  for joining us on the podcast today. This
• 26:58
  has been a a really,
• 27:00
  fun conversation. I can tell you're very passionate
• 27:02
  about everything you do, which is so much
• 27:04
  fun. And, you know, look forward to seeing
• 27:06
  you as well at our annual meeting. I
• 27:07
  know you'll be speaking on a panel. We'll
• 27:09
  dig deeper into many of the things we
• 27:11
  talked about today, and so I'll look forward
• 27:12
  to seeing you there.
• 27:14
  Thank you so much. It's been a pleasure.