Latest / Becker’s Healthcare Podcast / Dr. Yasir Tarabichi, CHAIO, CMIO of Ovatient, MetroHealth

Transcript

Dr. Yasir Tarabichi, CHAIO, CMIO of Ovatient, MetroHealth · Becker’s Healthcare Podcast

• 0:00
  Philips is a health tech leader focused on
• 0:03
  innovation that improves the health and well-being of
• 0:06
  people. Our health care technology and informatics solutions
• 0:09
  help care teams diagnose, treat, and manage more
• 0:12
  patients with greater precision, speed, and confidence across
• 0:16
  the care journey. With Philips, clinicians are empowered
• 0:19
  with streamlined insights in the moments that matter
• 0:21
  for every patient. Better care for more people.
• 0:25
  Philips.
• 0:28
  This is Gracelyn Keller with the Becker's Healthcare
• 0:30
  Podcast, and we are recording live at the
• 0:32
  tenth annual health IT, digital health, and RCM
• 0:35
  meeting. I'm currently joined by Yasser Tarabichi, who
• 0:38
  is the CAIO
• 0:39
  and CMIO of Ovation's with MetroHealth.
• 0:43
  So, Yasir, thanks for being here. We'd love
• 0:45
  to have you start by telling us a
• 0:47
  little bit more about yourself and your work
• 0:48
  in health care.
• 0:50
  Sure. So I would identify myself as a
• 0:52
  physician informaticist
• 0:54
  first and foremost. So I'm a pulmonary and
• 0:56
  critical care doc who still practices.
• 0:58
  I've had progressive,
• 1:00
  roles in the informatics space over the last
• 1:03
  decade or so. And at this point in
• 1:05
  time in my life, I am chief health
• 1:07
  AI officer at MetroHealth. So thinking about how
• 1:10
  to strategically
• 1:11
  steward all of these new health innovations into
• 1:13
  the health care space, both responsibly
• 1:16
  and with, good ROI and and outcome monitoring.
• 1:19
  And I'm also CMIO
• 1:21
  of a virtual care startup called Ovation, which
• 1:23
  is actually a shared platform
• 1:25
  that we're developing with leaders at MUSC Health.
• 1:29
  And so it's a virtual care platform that
• 1:31
  serves patients in different states in an integrated
• 1:33
  fashion where you're seen online,
• 1:36
  and you can get follow-up care and downstream
• 1:39
  opportunities
• 1:40
  at the brick and mortar facilities that we
• 1:41
  integrate with. So it's integrated virtual first care
• 1:44
  as opposed to point solutions, keeping patients in
• 1:47
  your system rather than
• 1:49
  losing them.
• 1:51
  And let's start our conversation today talking about
• 1:54
  AI
• 1:55
  because nearly half of medical practices reported using
• 1:58
  AI in some capacity in the last year,
• 2:00
  and it remains a key topic for health
• 2:02
  IT leaders.
• 2:03
  So from your perspective, what are the use
• 2:05
  cases that are making a difference right now,
• 2:07
  and how are you leveraging them in your
• 2:08
  organization?
• 2:10
  Yeah.
• 2:11
  So we've been doing predictive
• 2:13
  AI for several years now when we've had
• 2:16
  a lot of use cases that we've published,
• 2:19
  we've spoken about,
• 2:20
  a lot of energy around sepsis being a
• 2:23
  a good one as a as a frame
• 2:24
  of reference for the work that we've done.
• 2:26
  So from the predictive perspective, we obsess about
• 2:29
  are the models accurate, fair, and more importantly,
• 2:32
  as a safety net system,
• 2:34
  MetroHealth is worried about
• 2:36
  equitable implementations and outcomes. So can you use
• 2:39
  a reasonable AI model
• 2:41
  in a way that actually enhances care for
• 2:43
  all
• 2:44
  and then in a way that does not
• 2:46
  exacerbate
• 2:47
  disparities that we see and deal with, on
• 2:49
  a regular basis amongst our patient population or
• 2:52
  within our patient population?
• 2:53
  And so we've had a decent track record
• 2:55
  of getting those innovations and solutions out, implementing
• 2:58
  them in a safe and responsible and reliable
• 3:00
  way, and they do drive several aspects of
• 3:03
  our business.
• 3:04
  Obviously, with generative AI, there's been a a
• 3:06
  big,
• 3:07
  there's been an increase, I should say, in
• 3:09
  opportunity. And also
• 3:11
  along with that risk,
• 3:13
  risk for safety, reliability, and risk as as
• 3:16
  it pertains to,
• 3:18
  spending money in the wrong wrong place. So,
• 3:21
  lot of potential sunken cost there as well.
• 3:24
  The way we've approached the this the second
• 3:26
  part of the, AI,
• 3:27
  coin here, the generative AI solutions,
• 3:29
  we are focused on enterprise wide solutions,
• 3:33
  a rising tide that can lift many boats
• 3:36
  rather than point solutions.
• 3:37
  We focus on integrated solutions with a clear
• 3:40
  value proposition that touches multiple aspects of our
• 3:43
  business,
• 3:44
  clinical care
• 3:45
  efficiency,
• 3:46
  clinical outcomes, and even, yes, ROI in terms
• 3:50
  of revenue, either generation or recovery.
• 3:53
  And I can talk about a few of
• 3:54
  those. So, obviously,
• 3:56
  we,
• 3:56
  with several other health care systems, are moving
• 3:58
  forward in a big way with Ambien AI.
• 4:01
  We are optimistic
• 4:02
  about where this is going in terms of
• 4:04
  relieving the burden,
• 4:06
  on our clinicians and providers at the point
• 4:08
  of care,
• 4:09
  improving documentation and even potentially revenue
• 4:12
  as a result of that.
• 4:14
  And the other aspect is we're optimistic that,
• 4:17
  you know, with computer vision and other ambient
• 4:19
  technologies that that input aspect of care is
• 4:23
  going to become easier. The bar becomes lower.
• 4:25
  We have richer, more accurate information entering our
• 4:28
  electronic health records. And as a result,
• 4:31
  better detail on our patients, how we're doing,
• 4:34
  ideally, also in the future,
• 4:36
  some guardrails around clinical
• 4:38
  decision support that, you know, in an in
• 4:40
  an ideal universe happens in near real time.
• 4:42
  So optimistic about that strategy, optimistic about that
• 4:45
  technology.
• 4:47
  We're also using generative AI
• 4:49
  to summarize and retrieve information from our clinical
• 4:52
  documentation.
• 4:54
  So getting clinicians up to speed in a
• 4:55
  quick fashion with medicine today, there's a lot
• 4:58
  of shift work. So providers,
• 5:01
  nurses are coming in and out of care,
• 5:03
  cycling through on a regular basis,
• 5:05
  having to to understand their patients quickly. So
• 5:08
  we're using this technology to surface
• 5:10
  pertinent information
• 5:12
  at the point of care in real time.
• 5:14
  We're using the technology to help us draft
• 5:16
  progress notes and discharge summaries to, again, reduce
• 5:19
  the burden on our our providers,
• 5:21
  also with the potential prospect of improving documentation
• 5:25
  to the point of, you know, accurately reflecting
• 5:27
  our case mix index and and billing more
• 5:29
  appropriately, concordant with the actual complexity
• 5:32
  that we see within our health care system.
• 5:34
  And that that is a solution that's, again,
• 5:36
  enterprise wide, starting inpatient, moving to ambulatory
• 5:40
  ultimately with a very solid vendor,
• 5:43
  partner who's been really focused on the seamlessness
• 5:46
  of the integration.
• 5:47
  And so those are, I would say, the
• 5:49
  two big spaces that we've moved into.
• 5:52
  We work with other vendors with several other
• 5:53
  AI solutions, including things like computer vision. And
• 5:57
  so there's other solutions that we're evaluating in
• 5:59
  in real time. And what we're trying to
• 6:01
  build are really responsible,
• 6:04
  reliable partnerships
• 6:06
  with companies that are forward thinking, that are
• 6:08
  open to feedback and codevelopment and opportunity, and
• 6:11
  even potentially partnering.
• 6:13
  And as virtual care expands from AI enabled
• 6:16
  tools and remote monitoring to broader digital health
• 6:19
  platforms, introducing these new technologies can bring challenges.
• 6:23
  So what advice do you have for leaders
• 6:25
  navigating everything from governance to patient engagement? And
• 6:28
  can you share an example of how your
• 6:30
  organization has balanced innovation with operational constraint?
• 6:34
  So one of my biggest pet peeves in
• 6:37
  the AI space is companies that lead with
• 6:40
  AI.
• 6:41
  So the value proposition becomes AI for the
• 6:44
  sake of AI.
• 6:46
  We are problem oriented.
• 6:49
  Solutions
• 6:50
  based on the problem, cart not before the
• 6:53
  horse. Right? So don't tell me why I
• 6:55
  need an AI solution. Tell me about the
• 6:57
  problem you're trying to solve
• 6:59
  and why AI is the answer to that
• 7:01
  problem.
• 7:02
  And so if you think about how that
• 7:03
  percolates up to the system, then you have
• 7:05
  to make sure that you know what the
• 7:07
  problems you're trying to solve are. I do
• 7:09
  not need to be sold on solutions
• 7:11
  for problems that I, you know, quote, unquote,
• 7:13
  don't know that I had. Right? Or people
• 7:15
  telling me that there's magically some revenue generation
• 7:18
  or money under the couch cushions that I
• 7:20
  didn't quite know about. I mean, we're obviously
• 7:21
  always looking to
• 7:23
  optimize our, you know, our systems and our
• 7:26
  operations, but we know
• 7:28
  ourselves enough to know what are we trying
• 7:30
  to do
• 7:31
  strategically. So
• 7:33
  as that percolates up to the system,
• 7:35
  strategically aligned solutioning is exactly what we're working
• 7:39
  on, and AI is not always the answer.
• 7:41
  And actually, as chief health AI officer, more
• 7:43
  often than not, I'm telling people why AI
• 7:46
  is not where you should be looking.
• 7:49
  But unfortunately,
• 7:50
  you know, when when vendors lead with
• 7:53
  AI
• 7:54
  less informed
• 7:56
  operational leaders or clinicians will think that there
• 7:59
  is some kind of magic instilled in that
• 8:01
  it's going to be an ideal or perfect
• 8:02
  solution, or that it's an independent autonomous system,
• 8:06
  or that I'm going to be able to,
• 8:08
  you know, reduce my need on certain FTEs
• 8:10
  in the organization and unlock a whole bunch
• 8:13
  of revenue, and none of that has panned
• 8:15
  out. Right? And so at the end of
• 8:17
  the day, I think the answer to the
• 8:19
  question is
• 8:20
  start with your strategic strategies, your operational priorities.
• 8:24
  Always start with a, you know, vendor partner,
• 8:27
  I should say core vendor partner approach first.
• 8:29
  More often than not, they may already be
• 8:31
  working on something that you're you're you're trying
• 8:33
  to solve,
• 8:34
  and then make sure that you have a
• 8:36
  clear eyed view of the actual ROI. So
• 8:38
  give you an example, the Ambient Solutions,
• 8:41
  you know, I don't know that they're going
• 8:43
  to give us more revenue, direct hard dollars,
• 8:45
  but I do know that there's a chance
• 8:47
  that our providers are gonna be able to
• 8:48
  get home on time. They're gonna actually enjoy
• 8:51
  working for our organization even more than they
• 8:52
  did before, and we're more likely to less
• 8:55
  likely, I should say, to lose them to
• 8:56
  another organization that has an Ambien AI solution.
• 8:59
  So the ROI is not always measurable and
• 9:02
  hard dollars.
• 9:03
  I like to know that the outcomes are
• 9:07
  not just related to finances. There's operational efficiency.
• 9:10
  There's clinical safety,
• 9:12
  reliability,
• 9:13
  and then, obviously,
• 9:15
  hard hard dollars. And you don't have to
• 9:17
  feed all of those streams to make a
• 9:19
  case for your solution,
• 9:21
  but you have to hit as many as
• 9:22
  many of those boxes as you can for
• 9:24
  it to be viable and more importantly for
• 9:26
  it to be sustainable.
• 9:28
  And how are you seeing recent legislation, both
• 9:31
  state and federal, affect health care organizations
• 9:33
  and health care IT specifically, and have you
• 9:35
  adjusted any strategy in response? Yeah. So we're
• 9:39
  obviously really tightly integrated in terms of strategy
• 9:42
  with our compliance and risk groups. And so
• 9:45
  we're monitoring that landscape actively.
• 9:47
  We you know, there will be an ebb
• 9:49
  and flow of regulation and deregulation
• 9:52
  depending on the political lens and the climate.
• 9:54
  But as a as a safety net system,
• 9:57
  we hold our we are responsible for our
• 9:59
  own fate. We are responsible for how we
• 10:01
  are perceived by our patients and our community,
• 10:03
  and we take that responsibility
• 10:05
  seriously. And so what I mean to say
• 10:07
  there is we're not waiting to be told
• 10:10
  how to be safe and responsible when it
• 10:12
  comes to implementing these new technologies.
• 10:14
  We we prefer to think of ourselves as
• 10:16
  leaders
• 10:17
  in that space.
• 10:19
  And so really watching it, really trying to
• 10:21
  make sure we're ahead of the game there.
• 10:23
  And we're not we're not waiting for AI
• 10:25
  assurance labs to come and go. We believe
• 10:27
  that
• 10:28
  the implementation
• 10:30
  of reliable AI solutions within our system requires
• 10:34
  our own due diligence and requires internal validation
• 10:38
  to make sure that it works for us
• 10:39
  in our setting and our patients. And we're
• 10:41
  not going to be told externally that, you
• 10:43
  know, solution a versus solution b may be
• 10:45
  better because
• 10:46
  of some analysis run at a completely different
• 10:49
  health care system or by some AI assurance
• 10:51
  group that may have different
• 10:54
  interests,
• 10:55
  in mind.
• 10:57
  And final question as we wrap our conversation
• 10:59
  up, what is your top piece of advice
• 11:01
  for health care leaders as they prepare for
• 11:03
  further advancements in technology and rising demands for
• 11:06
  care?
• 11:08
  Go back to your strategy.
• 11:09
  Go back to your operational needs.
• 11:12
  Focus on the problems you're trying to solve.
• 11:14
  AI is not always the answer nor should
• 11:16
  it be. And I think if you if
• 11:18
  you if you really think about a solution,
• 11:21
  a problem
• 11:22
  solutioning approach or problems based approach,
• 11:26
  you'll you'll more likely come out with a
• 11:28
  successful outcome. And so I think
• 11:31
  go back to your basics.
• 11:33
  Awesome. Well, Yasir, thanks so much for joining
• 11:35
  me today on the Becker's Healthcare podcast and
• 11:37
  sharing your thoughts and insights on these topics.
• 11:40
  Again, we are recording live at the tenth
• 11:41
  annual health IT, digital health, and RCM meeting.
• 11:44
  Thank you so much for having me.