Latest / Becker’s Healthcare Digital Health + Health IT / Unlocking Data and Efficiency: How a Major Health System Transformed Workflows with Data Fabric and AI.

Transcript

Unlocking Data and Efficiency: How a Major Health System Transformed Workflows with Data Fabric and AI. · Becker’s Healthcare Digital Health + Health IT

• 0:00
  Hi, everyone. This is Lucas Voss with Becker's
• 0:02
  Healthcare. Thanks so much for tuning in to
• 0:04
  the Becker's Healthcare podcast series. Fantastic to have
• 0:07
  you. Today, we're talking about unlocking
• 0:10
  data and efficiency, how a major health system
• 0:12
  transfer workflows
• 0:14
  with data fabric and AI. And I'm so
• 0:17
  excited
• 0:18
  to be joined by Sarah Galyon, senior director
• 0:20
  of health care solutions at Intellistax. Sarah, thanks
• 0:23
  so much for being here today. It's great
• 0:25
  to have you. Yes. Thanks for having me.
• 0:27
  I'm excited.
• 0:28
  Absolutely. For those that might not know you
• 0:30
  yet, at least, could you just share a
• 0:32
  little bit about yourself and and your work
• 0:33
  in health care?
• 0:35
  Yeah. Sure. So I've been with Intellistac for
• 0:38
  about a year now. Prior to that, I
• 0:40
  worked for the University of Tennessee Medical Center
• 0:43
  for seventeen
• 0:44
  years. I did a lot of things when
• 0:47
  I was there. Most of my adult life
• 0:49
  spent there, but, most of my tenure was
• 0:52
  focused on digital transformation,
• 0:55
  process improvement initiatives, and, like, clinical informatics.
• 0:59
  I had the privilege of working alongside
• 1:01
  and learning from some of the most,
• 1:04
  brilliant minds I would consider in health care
• 1:06
  process improvement.
• 1:07
  So that was a really unique opportunity.
• 1:10
  And then through that work is how I
• 1:12
  got to know the folks at Intellistac,
• 1:14
  because we use their products
• 1:16
  in the hospital for a lot of our
• 1:18
  digital solutions.
• 1:20
  So and now I am here,
• 1:23
  trying to help kind of spread the message,
• 1:26
  so to speak, to a wider audience.
• 1:28
  Right. And we're certainly spreading the message today.
• 1:30
  Excited to hear about all of your expertise
• 1:31
  and then some of the things that we're
• 1:33
  going to discuss today. The name of the
• 1:35
  game for health care for health systems right
• 1:37
  now seems to be to do more with
• 1:39
  less
• 1:40
  from workforce shortages, finance always. Right? Finances, patient
• 1:44
  expectations, etcetera. It it's really tough.
• 1:47
  In this environment, how do you see technology,
• 1:50
  and that includes AI,
• 1:52
  helping teams work more efficiently and focus on
• 1:56
  what matters most?
• 1:57
  So the first thing that comes top of
• 2:00
  mind for me is is a quote by
• 2:02
  Hal Wolf, who's president and CEO of HIMSS.
• 2:05
  He said in a keynote speech a few
• 2:07
  years back, technology alone
• 2:10
  doesn't do anything.
• 2:12
  It's people, process,
• 2:14
  and technology. All three have to come together.
• 2:17
  And I thought that was just brilliant
• 2:20
  at the time because it's so true. Like,
• 2:22
  you you've got all these tools and things
• 2:24
  that you you wanna use,
• 2:26
  but you can't just throw tech at the
• 2:28
  problem and think it's gonna fix it. Right?
• 2:30
  In this particular scenario, I don't really see
• 2:33
  this as doing more with less as much
• 2:35
  as I see it being smart about providing
• 2:38
  your very highly valuable staff and team members
• 2:41
  with the right tools to be as efficient
• 2:44
  as possible.
• 2:46
  There's a man called Oliver King Smith, who's
• 2:49
  the founder of SmartR AI,
• 2:51
  and he refers to AI as assistive intelligence.
• 2:56
  And I think that's a much better way
• 2:57
  of thinking about this kind of technology because
• 3:00
  when we think about AI, it can be
• 3:01
  very intimidating.
• 3:03
  But, really, it's there to enhance your overall
• 3:06
  productivity and allow you to shift your focus
• 3:09
  away from ancillary tasks
• 3:11
  and work more towards that kind of higher
• 3:13
  caliber of work.
• 3:15
  There's a, I think, a major opportunity
• 3:18
  to streamline workflow processes and efficiencies
• 3:21
  using tools like that,
• 3:23
  specifically in the administrative
• 3:25
  space of health care.
• 3:27
  There's so much that goes on in patient
• 3:29
  care coordination
• 3:31
  that most clinicians are just kind of blissfully
• 3:33
  unaware of, just to be frank. And that's
• 3:36
  just kind of the way it is. There's
• 3:38
  this unfortunate kind of division between the clinical
• 3:41
  side and the administrative side of the house.
• 3:44
  So kind of just an example of this,
• 3:46
  I was working on a project,
• 3:48
  helping
• 3:49
  a physician practice transition to CPOE,
• 3:52
  And
• 3:53
  I got a lot of resistance,
• 3:55
  from the physicians. They did not wanna have
• 3:57
  to use a computer.
• 3:58
  And
• 4:00
  one said to me point blank, I
• 4:02
  I don't wanna have to do this. All
• 4:04
  I have to do right now is tell
• 4:06
  my nurse what I want and it magically
• 4:07
  happens. Like, those words came out his mouth.
• 4:09
  Magically
• 4:11
  happens. And so I had to explain to
• 4:12
  him it doesn't magically happen. Like, your nurse
• 4:14
  is now gonna spend two hours of her
• 4:16
  day coordinating
• 4:18
  with the Surgical Suite,
• 4:19
  with,
• 4:20
  your administrative staff, with your schedulers, with the
• 4:23
  patient trying to get this all set up
• 4:25
  so your patient can come in for their
• 4:27
  procedure.
• 4:28
  Whereas if you would just take five minutes
• 4:30
  and put the order in, now we can
• 4:32
  automate a lot of that on the backside
• 4:34
  using that as the trigger.
• 4:36
  Right?
• 4:37
  And he was completely unaware that that was
• 4:39
  a possibility.
• 4:40
  So
• 4:41
  there's so much that goes into preparing a
• 4:43
  patient for arrival to a facility and also
• 4:45
  a lot of follow-up work
• 4:47
  after their visit's over. And there's a lot
• 4:50
  of opportunity now to kind of let the
• 4:52
  computer take care of a lot of that
• 4:54
  manual stuff
• 4:55
  that allows you to then transcend to a
• 4:57
  higher level of work. Right? So now the
• 4:59
  nurse doesn't have to spend two hours of
• 5:00
  a day chasing people around and trying to
• 5:02
  coordinate and orchestrate things. The computer can kind
• 5:05
  of kick off those
• 5:07
  those workflow patterns and and, you know, trigger
• 5:12
  certain notifications and and kind of keep people
• 5:14
  in sync with what's happening
• 5:16
  while she can go now put her focus
• 5:18
  back on other patients. Right?
• 5:21
  So I know people are kind of afraid
• 5:23
  that they're gonna be replaced by AI, but
• 5:26
  I would argue we're not even close
• 5:28
  to being there yet.
• 5:29
  But your job will evolve into something else
• 5:32
  through the incorporation of it.
• 5:34
  And I think that
• 5:36
  we should be using it to offset
• 5:39
  that work burden, not thinking it's just gonna
• 5:41
  replace entire jobs.
• 5:43
  Well, it comes back to the trifecta that
• 5:45
  you mentioned. Right? People process technology, and I
• 5:48
  feel like this it's doing more with more
• 5:51
  instead of the do more with less. Right?
• 5:52
  I think that's very, very key.
• 5:55
  From your work, as you look at some
• 5:56
  of the partners that you're working with and
• 5:58
  certainly your own work, you made this experience
• 6:00
  firsthand. You have this experience firsthand.
• 6:03
  What are some of the results that these
• 6:05
  organizations are achieving when they're changing the approach
• 6:08
  and they're not just saying, hey, nurse, can
• 6:10
  you do this for me? They're actually handling
• 6:13
  handling on their own. And what are some
• 6:15
  of the success factors that stand out for
• 6:17
  you
• 6:19
  that have really been able to make these
• 6:21
  changes stick?
• 6:23
  So making the changes stick,
• 6:26
  there's
• 6:27
  a few things that there's a lot that
• 6:29
  goes into play.
• 6:31
  Let me,
• 6:33
  I'm gonna tell a story real quick and
• 6:35
  Yeah. Go ahead. Help illustrate this. Back in
• 6:37
  the height of COVID,
• 6:40
  we had,
• 6:41
  instruction at UT Medical Center that we needed
• 6:44
  to stand up a monoclonal antibody unit.
• 6:46
  This was an experimental IV infusion therapy that
• 6:50
  preceded the vaccine, and it was intended for
• 6:52
  folks that were already sick with COVID.
• 6:55
  And the goal was to keep them from
• 6:58
  getting bad sick, technical term,
• 7:01
  and ending up in the hospital.
• 7:03
  It wasn't intended to be a cure. It's
• 7:04
  just to help bolster their immune system so
• 7:07
  that they could fight it off. But
• 7:09
  our instructions were if you can service eight
• 7:11
  to 10 patients a day, that would be
• 7:13
  good enough. And we've looked at that and
• 7:15
  went,
• 7:16
  eight to 10 patients. They're like, why even
• 7:17
  bother? Eight to 10 patients isn't gonna put
• 7:20
  a dent in this problem. Right?
• 7:22
  And we knew that the real driver
• 7:24
  of
• 7:25
  volume was gonna be how quickly can we
• 7:29
  triage
• 7:30
  the referrals
• 7:31
  to get the patients through the system and
• 7:33
  into the schedule. Right?
• 7:35
  So
• 7:36
  we brought
• 7:37
  everybody together and
• 7:39
  talked about it with the people that we're
• 7:41
  gonna have to be administering the medications, the
• 7:44
  pharmacists, the nurse,
• 7:46
  the schedulers, you know, everybody that was gonna
• 7:48
  be involved in this, and we designed
• 7:50
  a system. We used a combination of digital
• 7:53
  tools in the EHR and also products from
• 7:56
  Intellistack as well,
• 7:58
  to design and implement a referral and triaging
• 8:00
  system that would allow us to quickly capture
• 8:03
  the referrals
• 8:04
  and streamline scheduling and eliminate those bottlenecks so
• 8:08
  we could get people through faster and then
• 8:10
  scale as needed.
• 8:12
  At our peak,
• 8:13
  we serviced around a 120 patients a day.
• 8:17
  Our clinic was open two full years, and
• 8:20
  we triaged over 10,000
• 8:22
  referrals during that time.
• 8:24
  Other health care facilities
• 8:26
  in that area
• 8:27
  who also were offering this therapy closed in
• 8:30
  six months,
• 8:31
  and as far as I know, never saw
• 8:32
  more than eight to 10 patients a day.
• 8:35
  So
• 8:36
  the reason that worked was a lot of
• 8:39
  contributing factors. Right? But one of the biggest
• 8:42
  ones I would attribute to, not just the
• 8:44
  tech,
• 8:45
  but we brought in
• 8:47
  the people that were gonna do the work.
• 8:50
  And we talked to them, and we tried
• 8:52
  to understand,
• 8:54
  from their point of view what they were
• 8:56
  gonna need in order to be successful. Right?
• 8:59
  So
• 9:00
  to me, that's what really makes it stick
• 9:02
  because you can throw
• 9:04
  whatever tech you want, and it doesn't matter
• 9:06
  how cool it is, you could build a
• 9:07
  rocket to the moon. If nobody knows how
• 9:09
  to fly the rocket ship, you're not going
• 9:11
  anywhere. Right?
• 9:12
  So
• 9:13
  I I always tell people you've got to
• 9:15
  talk to the people doing the work
• 9:18
  and,
• 9:19
  give them a seat at the table and
• 9:20
  a voice to provide input on how to
• 9:23
  design and implement those processes. If they're involved,
• 9:27
  you will have a much easier time, getting
• 9:29
  buy in and adoption for the implementation.
• 9:33
  That's one key factor that's been emerging as
• 9:35
  well for us in a lot of conversations
• 9:37
  is the adoption piece. Adoption is crucial, especially
• 9:40
  for AI success
• 9:42
  in this space, which is so, so critical.
• 9:45
  Looking then to again, you've highlighted this process
• 9:47
  and and how it worked. When we look
• 9:49
  at leadership and and specifically health care leaders,
• 9:52
  obviously,
• 9:53
  what is one step that they can take
• 9:55
  now to better prepare for this future that's
• 9:58
  obviously driven by technology and specifically
• 10:01
  AI to ensure their organizations are set up
• 10:04
  for lasting efficiency gains?
• 10:07
  So
• 10:08
  I have a a saying that I like
• 10:10
  to throw around a lot, which is
• 10:12
  managers theorize about problems,
• 10:14
  and the frontline worker actually knows what the
• 10:17
  problems are.
• 10:18
  Not just knows what the problems are, but
• 10:20
  they know what the work grounds are. Right?
• 10:23
  So it kind of goes back to, one,
• 10:26
  you've got to involve the right people, and
• 10:28
  two, don't assume that AI is gonna be
• 10:30
  the magic potion that solves all of your
• 10:33
  problems.
• 10:34
  You need to do your research. You need
• 10:36
  to talk to people and figure out start
• 10:38
  by identifying
• 10:39
  what problems, what bottlenecks, what inefficiencies
• 10:43
  are you
• 10:44
  trying to solve for. I would start there.
• 10:46
  I wouldn't I wouldn't dive off the deep
• 10:48
  end of the pool and let's find some
• 10:49
  cool AI and figure out how to how
• 10:51
  to, you know, pigeonhole it into our organization,
• 10:56
  because it may not be the thing you
• 10:57
  need, right? That's backwards.
• 11:00
  Figure out what you wanna make better. Figure
• 11:02
  out by identifying what problems you wanna solve,
• 11:05
  and then you can go and start researching
• 11:09
  what AI tools are available. It doesn't even
• 11:10
  have to be AI. It can be anything
• 11:12
  supportive that will be
• 11:15
  a value
• 11:16
  to, like I said, support the process. Because
• 11:19
  that's all
• 11:20
  AI does. That's all technology does. It supports
• 11:23
  the process that you already have in place
• 11:26
  and have decided on. So if you try
• 11:27
  to pigeonhole something in there without taking a
• 11:30
  look at the bigger picture, you're gonna get
• 11:32
  failed implementation every time. You know, I think
• 11:35
  the best AI tools are the ones you
• 11:37
  never see. Yeah. Absolutely.
• 11:39
  A quick follow-up to this. I think this
• 11:40
  is an important question that I wanted to
• 11:42
  ask because I think it's key to this
• 11:44
  conversation.
• 11:45
  What's at stake
• 11:47
  if this doesn't happen?
• 11:49
  If they don't do this, if they don't
• 11:52
  follow the red thread? What's at stake here?
• 11:56
  Candidly, I think
• 11:58
  many
• 11:59
  health care
• 12:00
  organizations
• 12:02
  operate under kind of this
• 12:04
  comforting notion that health care as an industry
• 12:08
  is recession proof,
• 12:10
  and
• 12:11
  to a certain degree it is because everyone
• 12:14
  will always need health care in some form
• 12:16
  or fashion,
• 12:17
  what they tend to forget is your individual
• 12:20
  organization
• 12:21
  is not recession proof.
• 12:24
  Just because the industry is not going anywhere
• 12:26
  doesn't mean that if you decide
• 12:29
  not to take steps to make improvements
• 12:33
  doesn't mean that your particular organization will be
• 12:35
  around in the future. Right?
• 12:38
  And to be fair to to health care
• 12:40
  facilities, I think a lot of people think
• 12:41
  they've got deeper pockets than they really do.
• 12:44
  Right?
• 12:45
  From what I've seen, most hospitals, even large
• 12:48
  systems, have very narrow operating margins. They don't
• 12:51
  have a lot of money to throw around.
• 12:53
  So
• 12:54
  you've got to be intentional and deliberate about
• 12:56
  where you're spending.
• 12:58
  But the wonderful thing about AI, even though
• 13:01
  it's been around a lot longer than I
• 13:03
  think people realize,
• 13:05
  it's at a maturity level now where
• 13:08
  it can be implemented and
• 13:11
  offset, you know,
• 13:13
  your office manager who's been asking for three
• 13:15
  extra people to support the volumes for the
• 13:17
  past four years and never gets approved.
• 13:20
  Well, is there a way that we can
• 13:21
  implement AI to offset the need
• 13:24
  to hire those people so that the people
• 13:26
  she's got
• 13:27
  can be more more efficient and improve the
• 13:30
  patient throughput.
• 13:32
  And
• 13:33
  then not just stop the hemorrhaging of losing
• 13:36
  money, but now we can maybe start
• 13:38
  making money. I mean, health care is a
• 13:40
  business. We don't have to like it, but
• 13:42
  it is.
• 13:43
  So that that would be my thoughts on
• 13:45
  that is is if if you
• 13:47
  if you wanna maintain the status quo, the
• 13:50
  world won't move on without you. That's just
• 13:52
  the way it is.
• 13:54
  Especially
• 13:55
  around technology and AI. I think that's especially
• 13:57
  true in that space. Sarah, thank you so
• 14:00
  much for your time and insights. What a
• 14:01
  fantastic conversation. Thanks for being here. Yes. Thanks
• 14:04
  for having me. Absolutely. And we also want
• 14:06
  to thank our podcast sponsor at Telestack. You
• 14:08
  can tune into more podcasts from Becker's Healthcare
• 14:10
  by visiting our podcast page at beckershospitalreview.com.