Latest / Becker’s Healthcare Podcast / Turning Healthcare’s Toughest Operational Challenges into Breakthrough Solutions: Part 1

Transcript

Turning Healthcare’s Toughest Operational Challenges into Breakthrough Solutions: Part 1 · Becker’s Healthcare Podcast

• 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, and welcome to
• 0:07
  part one
• 0:09
  of our series on turning health care's toughest
• 0:11
  operational challenges into breakthrough through solutions.
• 0:15
  And I'm so excited to be joined by
• 0:17
  Ben Fichtner. He's the America's president at UiPath
• 0:19
  and Jason Buie, who leads the health care
• 0:22
  industry practice also at UiPath.
• 0:25
  Both have spent years working closely with provider
• 0:28
  and payer organizations tackling
• 0:30
  the operational challenge that sit behind how health
• 0:33
  care actually works. And as most of you
• 0:36
  know, obviously,
• 0:38
  AI is a big topic right now
• 0:40
  for health systems across the country, really, across
• 0:44
  the world. But what health care leaders really
• 0:46
  want to know is where it's actually
• 0:48
  improving
• 0:49
  how work gets done, and that's exactly what
• 0:53
  we'll be talking about today, what's happening on
• 0:56
  the ground right now, and where organizations are
• 0:59
  already seeing
• 1:00
  real
• 1:01
  impact. Ben and Jason, thanks so much for
• 1:04
  being here today. It's so great to have
• 1:06
  you. Appreciate you having us. This is gonna
• 1:08
  be fun.
• 1:09
  Lots to talk about. I do wanna kick
• 1:11
  us off just with a brief introduction
• 1:14
  so everybody knows where we are, level setting
• 1:16
  us a little bit. Ben, I start off
• 1:18
  with you. For those that might not be
• 1:20
  familiar,
• 1:21
  can you just share a little bit about
• 1:22
  what UiPath does and how
• 1:25
  it fits into the health care ecosystem that
• 1:27
  we're seeing today?
• 1:29
  We've been around for a decade plus helping
• 1:31
  health care systems specifically.
• 1:33
  But we started in general as a task
• 1:35
  based automation tool.
• 1:37
  So and we've grown up from there. Right?
• 1:39
  So I think a lot of times, Jason
• 1:40
  will go deeper on some of these use
• 1:41
  cases. But when you think about it, there's
• 1:43
  a ton of tasks that happen across health
• 1:45
  care systems, but more importantly, across a ton
• 1:47
  of IT and back end systems. And we're
• 1:50
  the the technology, hopefully, that orchestrates,
• 1:53
  and doesn't replace what's already there as it
• 1:55
  relates to EMRs or other things, but working
• 1:57
  across those things. So health care processes can
• 1:59
  run more end to end.
• 2:01
  And what we've seen, and Jason will talk
• 2:02
  about a lot of great customer use cases
• 2:05
  that we've had, of course, is where the
• 2:07
  biggest issues are is not just in one
• 2:09
  specific task or one specific process, but more
• 2:12
  on these highly complex end to end workflows
• 2:15
  that we've been able to help deliver significant
• 2:17
  ROI,
• 2:18
  again, to both payers and providers. But we're
• 2:20
  really happy to be here and really happy
• 2:21
  to dig in. Yeah. I'm excited to dig
• 2:23
  in as well. There's so much to talk
• 2:24
  about here as well. And I wanna touch
• 2:26
  a little bit on your personal journey,
• 2:29
  on this as well
• 2:31
  and how it's shaped you, right, your journey
• 2:34
  in this space.
• 2:36
  You spend a lot of time working with
• 2:38
  provider and payer organizations, but, again, you also
• 2:40
  have that personal connection, which is so crucial.
• 2:42
  How has that shaped
• 2:44
  the way you think about the operational challenges
• 2:46
  health care leaders are dealing with today?
• 2:49
  Great question. And, yeah, I grew up in
• 2:51
  a family full of physicians,
• 2:53
  and I was kind of the black sheep
• 2:54
  of the family who started off actually,
• 2:56
  selling capital medical equipment for GE Healthcare.
• 2:59
  So I've been in the health care game
• 3:00
  for, you know, twenty to twenty five plus
• 3:02
  years, and it's been an interesting journey. I
• 3:04
  think
• 3:05
  thinking back though around, like, my family's
• 3:07
  dining room table, if you will, over over
• 3:09
  dinner,
• 3:10
  my parents oftentimes,
• 3:12
  you know, would would compliment the certain parts
• 3:14
  of their day, and they were always very,
• 3:15
  very excited the patient and the clinical experiences.
• 3:18
  But the things that always bug them the
• 3:19
  most, of course, was the administrative and operational
• 3:21
  strain that they had every day. And it
• 3:23
  was how do they get time back to
• 3:25
  spend more time with patients.
• 3:27
  And I think that's still ever true. Right?
• 3:28
  And I think when we think about a
• 3:30
  lot of customers' examples, and I know Jason
• 3:32
  will will share a few, but one that
• 3:34
  we just touch base with,
• 3:36
  a massive,
• 3:37
  provider here was talking to us about reviewing
• 3:39
  their medical records, right, as part of their
• 3:41
  authorization decisions.
• 3:42
  And their team was spending seventy plus minutes
• 3:45
  just going through a chart to find the
• 3:46
  evidence they needed to make a decision. So
• 3:48
  that's seventy minutes they weren't spending with patients,
• 3:50
  right, or they weren't spending on on the
• 3:52
  work that really matters. That's seventy minutes of
• 3:54
  searching instead. So using u I UiPath tools,
• 3:57
  they were able to cut that down to
• 3:58
  a handful of minutes,
• 4:00
  which is super, super important. Right? We wanna
• 4:02
  make sure that we're doing everything we can
• 4:04
  to help folks spend as much time in
• 4:06
  the places that matter and not doing back
• 4:09
  end administrative work.
• 4:10
  And it's kinda cool to see my story
• 4:12
  come for full circle right from sitting at
• 4:14
  the dinner table there to now being able
• 4:16
  to do this with with clients on a
• 4:17
  day to day basis.
• 4:18
  Yeah. Giving providers a little bit more time
• 4:20
  back to care for their patients, which is
• 4:22
  so crucial as you've mentioned.
• 4:24
  Jason, I wanna touch a little bit on
• 4:26
  what we've mentioned in the intro. Right? AI
• 4:28
  is inevitable right now. The conversation is there.
• 4:31
  Everybody is talking about it. From your perspective,
• 4:34
  when you're working with health care organizations today,
• 4:37
  what are they tackling with AI specifically in
• 4:39
  advanced automation? What are you seeing most?
• 4:43
  Thanks for the question, Lucas, and thanks again
• 4:45
  for having me on here. You know, I've
• 4:46
  been a longtime listener and consumer of Becker's
• 4:49
  content and just really appreciate what you guys
• 4:50
  are doing for the health care space. You're
• 4:52
  keeping everyone informed, and I just love everything
• 4:54
  you're doing. So keep up the good work.
• 4:57
  But I'll start off by saying that I
• 4:59
  think it's really great everyone is talking and
• 5:01
  thinking about AI right now in health care
• 5:03
  because it's already unlocking new efficiencies.
• 5:05
  And, you know, these efficiencies are are something
• 5:08
  that we could only dream of not too
• 5:09
  long ago.
• 5:11
  And it's what we've ultimately been waiting for
• 5:13
  in health care for the longest time. Right?
• 5:14
  We've been wanting more connectivity.
• 5:16
  We've been wanting better coordination
• 5:19
  and elimination
• 5:20
  of administrative functions that aren't value add and
• 5:23
  that aren't helping us focus on the patient
• 5:25
  or member. So I think that AI is
• 5:27
  gonna be a really big part of this
• 5:28
  transformation as a whole and something to look
• 5:30
  forward to.
• 5:32
  But, anyway, back to your original question on
• 5:34
  where I'm seeing AI being applied today.
• 5:36
  I'll have to split my response,
• 5:38
  between providers and payers because I think that
• 5:40
  would be the easiest way to lay this
• 5:42
  out and just explain to folks what's happening
• 5:44
  on the ground level from what we're seeing
• 5:46
  at UiPath.
• 5:48
  Yeah, Jason. Appreciate you breaking it down like
• 5:50
  that. That's helpful framing.
• 5:52
  Let's start with
• 5:54
  providers.
• 5:55
  What are you seeing there first?
• 5:58
  In general at this time, you know, they're
• 6:00
  going after large and complex processes within their
• 6:02
  organizations, and that's because it's finally possible to
• 6:05
  tackle these challenges in a meaningful way with
• 6:07
  AI.
• 6:08
  They're honing in where there's a lot of
• 6:10
  human intervention and areas where humans just were
• 6:13
  never really meant to be facilitating the process
• 6:16
  in the first place.
• 6:17
  On the provider side of things, you know,
• 6:19
  currently,
• 6:21
  we're working with a large academic medical center
• 6:23
  to streamline the patient triage process. You know,
• 6:25
  they came to us with this unique use
• 6:27
  case and wanted us to transform it with
• 6:29
  them. And we got,
• 6:31
  on the process level to understand what's happening,
• 6:33
  and we found that there's just a lot
• 6:34
  of referral documents to come through just to
• 6:36
  make
• 6:37
  a decision
• 6:38
  on where to put the patient next and
• 6:40
  where to route them and the types of
• 6:42
  programs
• 6:42
  that the the patient needs to have wrapped
• 6:44
  around their care journey.
• 6:46
  Now with agentic automation, you know, the triage
• 6:49
  and decisioning can happen so much faster,
• 6:51
  and the patients
• 6:53
  reach the right level of care a lot
• 6:54
  sooner
• 6:56
  because the details
• 6:58
  are finally being summarized for the human reviewer
• 7:00
  along with recommended next best steps.
• 7:03
  And so this has ultimately changed the game
• 7:06
  for how we triage the patients once they're
• 7:08
  in our facilities.
• 7:10
  At another academic health system, we've helped this
• 7:12
  customer with monitoring,
• 7:15
  the medication reconciliation process and then ultimately alerting
• 7:18
  clinicians when something needs attention.
• 7:20
  We saw compliance improve pretty significantly. Right? Going
• 7:24
  from about eighty five to 95% in just
• 7:26
  a few months, which is pretty quick turnaround
• 7:28
  time for an AI use case.
• 7:30
  And what's been pretty revealing to us is
• 7:32
  that agentic automation is allowing for health care
• 7:34
  organizations to just be a whole lot more
• 7:36
  proactive.
• 7:37
  Catching stuff before it happens,
• 7:40
  preempting ourselves and preparing ourselves for different scenarios,
• 7:44
  and then engaging with the member in a
• 7:46
  big way, at scale now that we can
• 7:48
  use AI to tap into the various channels
• 7:50
  that patients leverage.
• 7:52
  And we spend a whole lot of time
• 7:53
  in the rep cycle space at UiPath and
• 7:55
  working with our provider customers.
• 7:58
  And this is where we found most organizations
• 8:00
  are trying to apply AI first, actually,
• 8:03
  because it's a high volume area, complex area,
• 8:06
  and always changing of of the the process
• 8:09
  and rules that that govern it.
• 8:11
  Now we have agents that are picking up
• 8:13
  denials,
• 8:14
  analyzing it, and understanding its intent. And then
• 8:17
  going back into the EMR and collecting the
• 8:19
  supplemental information to resubmit this with high confidence
• 8:22
  that this will ultimately be the last transaction
• 8:25
  on this case, and we don't have to
• 8:26
  worry about it any longer.
• 8:29
  Some unique observations
• 8:30
  there from a provider point of view.
• 8:33
  What are you seeing on the payer side?
• 8:37
  On the payer side of things, we're seeing
• 8:39
  them apply AI in the back office functions.
• 8:41
  You know? This is where we see the
• 8:44
  most opportunity.
• 8:45
  But so think of
• 8:47
  some of the areas like claims processing,
• 8:49
  enrollment,
• 8:50
  payment disputes, prior auth intake and review,
• 8:53
  you know, as well as just auditing important
• 8:55
  aspects of our overall operations
• 8:58
  and doing so in a much more robust
• 9:00
  fashion.
• 9:01
  They're going after, like, the last mile of
• 9:03
  the process where humans were always required,
• 9:06
  but they're not removing the human entirely from
• 9:08
  the equation. Right? They're designing their AI solutions
• 9:11
  with us to maintain full control and then
• 9:13
  allow for human operators to intervene at the,
• 9:16
  like, most critical parts of the process.
• 9:19
  And then the claim space, we're seeing payers
• 9:22
  leverage, you know, that are already mature automations
• 9:25
  as foundations to incorporate AI on top of.
• 9:28
  They're having a a you know, these AI
• 9:29
  agents man these work queues to drive resolutions
• 9:32
  for the less complex scenarios.
• 9:34
  And they're having these agents compare the pricing,
• 9:37
  the contract terms, and identifying duplicates that are
• 9:39
  coming in and putting the the duplicate aside
• 9:43
  so they can truly work on the actual
• 9:44
  claim that needs attention.
• 9:47
  And then I'll just take a look at
• 9:48
  the care management space with you all. You
• 9:50
  know,
• 9:51
  everyone's been talking about applying AI in this
• 9:53
  space for the last couple of years.
• 9:55
  But now, Genentech AI is able to automate
• 9:58
  every step of the process. So think of
• 10:00
  the intake all the way through
• 10:02
  reviewing medical records, through, you know, the decision
• 10:05
  support handling,
• 10:06
  and then triaging into a clinician to make
• 10:09
  a final determination.
• 10:12
  AI is playing a role in virtually every
• 10:14
  step, and we're seeing that this is accelerating
• 10:16
  turnaround times pretty significantly
• 10:18
  and turning a process that is typically,
• 10:21
  you know, multi days down to minutes.
• 10:24
  So a lot of innovation happening there.
• 10:27
  And if you talk about medical records, you
• 10:29
  know, you have to really talk about the
• 10:31
  medical record summarization capabilities that we've put out
• 10:34
  there. It's been a game changer for the
• 10:35
  payer community that we work with and partner
• 10:37
  with.
• 10:38
  And this is pretty much generally where all
• 10:40
  our payer customers are applying AI first in
• 10:42
  the space
• 10:43
  because it just provides immediate relief to their
• 10:45
  clinicians.
• 10:47
  You know, we're able to ingest a stack
• 10:49
  of medical records and summarize it for the
• 10:51
  review nurse so they can
• 10:53
  be more in a validation role, if you
• 10:55
  will. Right? We provide the citations for them
• 10:57
  to double check
• 10:59
  the content.
• 11:00
  And the agent can also just read handwriting
• 11:03
  now, structured and unstructured documents and images.
• 11:06
  And we've seen review times be reduced by
• 11:08
  nearly 90%.
• 11:10
  You know? So this type of innovation is
• 11:12
  what we're excited for and is driving meaningful
• 11:16
  impact.
• 11:18
  And they touch so many aspects of the
• 11:20
  patient journey now too, which is so crucial.
• 11:23
  And and thanks for sharing all of these
• 11:24
  great examples. I think it really encompasses how
• 11:27
  holistically
• 11:28
  AI has been applied in the health care
• 11:30
  space right now. Now when organizations are start
• 11:33
  working on these kinds of workflows that you've
• 11:35
  mentioned and you've touched on a little bit,
• 11:37
  how do you make those
• 11:39
  operational
• 11:40
  so that they actually become part
• 11:43
  of that day to day process?
• 11:46
  Yeah. Good question. It's
• 11:48
  not a one size fits all. There's not
• 11:50
  one master playbook that works. Every organization is
• 11:53
  layered and designed so differently.
• 11:56
  But I I think the the way in
• 11:57
  which to operationalize it that I've seen success
• 12:00
  in is starting
• 12:01
  relatively small
• 12:04
  and proving it out and getting buy in
• 12:06
  from the organization top down.
• 12:09
  We're seeing customers being most effective when the
• 12:12
  employees, the clinicians are part of the journey
• 12:15
  to roll out the AI capability.
• 12:18
  We also
• 12:19
  see the need to constantly educate
• 12:22
  the workforce. One, because the technology is changing
• 12:25
  so quickly, but two, it's new to all
• 12:28
  of us and we all have different interpretations
• 12:29
  of how AI
• 12:36
  repetitive and
• 12:37
  having frequent check ins is what moves the
• 12:40
  needle.
• 12:41
  And then during the change management experience,
• 12:44
  you know, we're finding that it's not as
• 12:47
  daunting or as rigorous as people initially thought.
• 12:52
  We thought we had to all become, you
• 12:53
  know,
• 12:54
  prompt
• 12:55
  engineers and learn technology and code it. But
• 12:59
  because it's so simple to engage with now
• 13:02
  and it's low code and nearly no code,
• 13:04
  the clinical workforce is finding that it's straightforward
• 13:07
  and they're embracing it in a in a
• 13:09
  much bigger way than we anticipated.
• 13:11
  And with them being advocates,
• 13:13
  we are seeing that people are rallying behind
• 13:16
  this rather quickly.
• 13:18
  And the other aspect to this is,
• 13:20
  you know, we want to
• 13:23
  ensure that people are building the right things
• 13:25
  and deploying the right things from the get
• 13:27
  go. And so it's crucial to have
• 13:30
  champions,
• 13:31
  partners in the ecosystem,
• 13:33
  and domain experts that can select the right
• 13:37
  use cases to operationalize.
• 13:40
  And then, you know, you don't have to
• 13:42
  do big bang. Do this in modular form,
• 13:45
  you know, improve
• 13:46
  or apply AI into specific aspects of the
• 13:49
  workflow that you know that humans are struggling
• 13:52
  with or that there's repetitive
• 13:54
  nature to it.
• 13:56
  And you'll you'll start to see that people
• 13:58
  will gravitate towards this and and be excited
• 14:01
  about it.
• 14:02
  I love that you mentioned the change management
• 14:04
  aspect of this because I think it's so
• 14:06
  crucial. The leadership piece is often underrated when
• 14:09
  it comes to AI deployments and pilots, etcetera,
• 14:12
  whatever happens
• 14:13
  within a given health system. And I I
• 14:14
  wanna stay with the leadership perspective here a
• 14:17
  little bit because, again, we hear a lot
• 14:18
  about bright shiny objects in AI. Right? There's
• 14:22
  a lot of promise
• 14:23
  around certain AI tools, especially over the last
• 14:26
  couple of years. And, again, we know some
• 14:28
  work really, really well. Jason, you mentioned Agentic
• 14:31
  AI. Right? That's applied virtually everywhere now, and
• 14:35
  and some just didn't work out. Ben, I
• 14:37
  I'd love to stay with the leadership theme
• 14:40
  here. When leaders start seeing examples like the
• 14:42
  ones Jason described, right, what actually convinces them
• 14:45
  that this is not a bright shiny object,
• 14:47
  but it's actually real and it's going to
• 14:50
  make a difference? Yeah. I think it's it's
• 14:53
  healthy skepticism. Right? And I think the the
• 14:55
  article we all were talking about, call it
• 14:56
  last October, was ninety five percent of AI
• 14:59
  pilots fail, right, the one from MIT,
• 15:01
  which I kinda got a chuckle on because
• 15:02
  there's the same thing in across all organizations
• 15:05
  and not just health care. We talk a
• 15:06
  lot about, like, seventy five percent of IT
• 15:08
  projects fail in general. Right?
• 15:10
  So there should be skepticism on some of
• 15:12
  these things. I think what we've seen and
• 15:14
  where you can know when it starts to
• 15:15
  be real is, a, the the massive amounts
• 15:17
  of savings that we've seen from both payers
• 15:20
  and providers and, you know, their numbers, not
• 15:22
  ours of tens to hundreds of millions of
• 15:23
  dollars a year, saving things just for,
• 15:26
  operational costs, but a lot of things that
• 15:28
  just went missed, right, in in processing
• 15:31
  paperwork or documentation or coordination across systems and
• 15:34
  all the things that they have to do,
• 15:36
  to keep the system running.
• 15:38
  That's the one thing that we typically see
• 15:39
  is it's it's not about to Jason's point,
• 15:42
  I think he gave practical advice about how
• 15:44
  to get started. But for us, we talk
• 15:46
  a lot about, like, the process being the
• 15:47
  hero more than the technology. And when you
• 15:50
  start to see the end to end process
• 15:51
  become more efficient or deliver better outcomes,
• 15:54
  you know you're on to something. Right? And
• 15:56
  I think we've seen that in a handful
• 15:57
  of cases that that Jason helped illustrate.
• 16:00
  But that's usually better than it is of,
• 16:01
  like, hey. How many times are people logging
• 16:03
  into this tool? Or how many times are
• 16:04
  people getting better outcomes? I think what we're
• 16:07
  looking at at the majority of our clients
• 16:09
  that are wildly successful is those that have
• 16:11
  taken the administrative end to end process and
• 16:14
  saying this is a hell of a lot
• 16:15
  smoother than it was a year ago. Right?
• 16:18
  And then that's when we know the change
• 16:20
  management you talked about, the impact that we're
• 16:21
  having with the leadership across health care organizations
• 16:24
  is delivering. So I think it's a great
• 16:26
  question, and I I
• 16:28
  appreciate us digging in, but it's a healthy
• 16:30
  level of skepticism, I would say, for sure,
• 16:32
  and we gotta continue to work at it
• 16:33
  to make it successful for everybody that touches
• 16:35
  it.
• 16:37
  If you had to
• 16:39
  point to a factor that truly separates
• 16:42
  organizations and leaders that are making a difference
• 16:44
  here,
• 16:46
  what would those factors be when you look
• 16:48
  at at it from a progress perspective? What
• 16:51
  are those factors that are crucial?
• 16:54
  That's a great question.
• 16:56
  I'd say a really big factor in all
• 16:58
  of this is the mindset, and the mindset
• 17:00
  from the leadership level all the way down
• 17:02
  to the frontline operators.
• 17:04
  The ones that are charging ahead where I'm
• 17:06
  seeing
• 17:07
  are the ones who believe they have to
• 17:09
  disrupt themselves internally first and foremost. Right? And
• 17:11
  they're not sitting around waiting for the industry
• 17:13
  to be disrupted or hoping that it gets
• 17:15
  disrupted and then having to react to it
• 17:17
  and change,
• 17:19
  their organization to fit the mold.
• 17:22
  And when they all commit to that mindset,
• 17:24
  they move out of this proof of concept
• 17:26
  phase and become more focused on redesigning their
• 17:28
  workflows
• 17:29
  with AI as a native part of their
• 17:31
  processes.
• 17:32
  And real progress happens when organizations are intentional
• 17:35
  about making AI work and wanting it to
• 17:37
  work.
• 17:38
  As the old adage goes, you know, journey
• 17:40
  of a thousand miles begins with a single
• 17:42
  step. And we have to be open to
• 17:44
  trying and progressing forward even after something may
• 17:47
  not have worked out. We should then, of
• 17:49
  course, try again.
• 17:51
  And the number of tries will be fairly
• 17:53
  minimal from what I've seen in health care.
• 17:56
  Organizations will see that it'll connect and work
• 17:58
  rather quickly when they have the right domain
• 18:00
  experts in the room designing these AI initiatives.
• 18:05
  And the value is absolutely there. I think
• 18:07
  it's undeniable.
• 18:08
  We collectively as an industry need to allow
• 18:11
  our staff to explore this new technology and
• 18:13
  this new capability and incentivize them to do
• 18:16
  so.
• 18:17
  To close out, I'd also say that the
• 18:19
  organizations that roll out these AI capabilities into
• 18:22
  production and do so in a controlled manner
• 18:25
  will be the ones that will witness and
• 18:27
  understand the value more clearly.
• 18:29
  So, therefore, they're able to make meaningful progress
• 18:31
  and compound their successes and move a whole
• 18:33
  lot quicker
• 18:34
  in the space.
• 18:36
  Jason and Ben, thank you so much for
• 18:38
  your time and insights today. So great to
• 18:41
  have you. And thank you to UiPath for
• 18:42
  bringing us together and sponsoring today's episode. And
• 18:46
  for our listeners, UiPath will be attending Becker's
• 18:48
  annual meeting April 13 through the sixteenth. If
• 18:51
  you're there, they'd love to continue our conversation
• 18:53
  from today about how AI and automation can
• 18:56
  help automate and orchestrate these workflows more effectively
• 18:59
  across your organization.
• 19:02
  You can also tune in to more podcasts
• 19:03
  from Becker's Healthcare
• 19:05
  by visiting our podcast page at beckershospitalreview.com.