Latest / Becker’s Healthcare Digital Health + Health IT / Transforming Healthcare with AI: Insights from Rohit Chandra, PhD, Chief Digital Officer at Cleveland Clinic

Transcript

Transforming Healthcare with AI: Insights from Rohit Chandra, PhD, Chief Digital Officer at Cleveland Clinic · Becker’s Healthcare Digital Health + Health IT

• 0:00
  Imagine this. You're in the heart of Chicago
• 0:02
  mingling with the brightest minds in health IT.
• 0:05
  You've arrived at the 9th annual Health IT
• 0:07
  plus Digital Health plus RCM conference
• 0:10
  taking place October 1st through 4th at the
• 0:13
  luxurious Hyatt Regency Chicago.
• 0:15
  Picture the excitement as you collect countless business
• 0:18
  cards, forging invaluable connections with over 25 100
• 0:22
  executive level attendees.
• 0:24
  Feel the buzz of ideas flowing as you
• 0:26
  engage in meaningful conversations about the future of
• 0:29
  healthcare technology.
• 0:30
  Envision yourself attending sessions led by over 415
• 0:35
  elite speakers
• 0:36
  gaining insights that could transform your organization.
• 0:39
  From digital transformation and telehealth to clinician burnout
• 0:43
  and cybersecurity,
• 0:44
  each topic is designed to spark new ideas
• 0:47
  and provide actionable takeaways, but it doesn't stop
• 0:49
  there. Imagine sitting in a packed auditorium listening
• 0:53
  to motivating keynotes from some of the biggest
• 0:55
  names in sports.
• 0:56
  Four time Super Bowl champion Rob Gronkowski,
• 0:59
  WNBA champion and author Lisa Leslie, and NFL
• 1:03
  legend and ESPN analyst
• 1:06
  Eli Manning will be there sharing their stories
• 1:08
  and inspiring you to reach new heights.
• 1:11
  All around you, the future of Health IT
• 1:13
  unfolds, and you're not just a spectator,
• 1:16
  you're an active participant.
• 1:18
  Don't wait. You can find the event website
• 1:20
  and get
• 1:21
  by visiting beckershospitalreview.com
• 1:25
  and clicking on the events page. That's the
• 1:27
  beckershospitalreview.com
• 1:30
  events page.
• 1:31
  Hello, everyone. This is Jacob Emerson with Becker's
• 1:34
  Digital Health and Health IT podcast. Thanks so
• 1:37
  much for tuning in to all of the
• 1:39
  media opportunities with Becker's Healthcare.
• 1:42
  Today, we're thrilled to be joined by a
• 1:44
  special guest. Doctor. Rohit Chandra is executive vice
• 1:47
  president and chief digital officer at Cleveland Clinic.
• 1:51
  Doctor Chandra, thank you so much for taking
• 1:53
  the time to be with us on the
• 1:54
  podcast today.
• 1:56
  Oh, thank you.
• 1:58
  So, Rohit, there's a lot we wanna talk
• 2:00
  with you about today. But to get us
• 2:02
  started more broadly, there's obviously a lot going
• 2:04
  on right now all across the healthcare tech
• 2:08
  world. So so talk to us a little
• 2:09
  bit about from your perspective, the the trends
• 2:11
  that you're watching most closely right now that
• 2:14
  you think isn't receiving enough attention either in
• 2:17
  the media or among your industry colleagues.
• 2:20
  What what's top of mind for you right
• 2:21
  now and within your role?
• 2:25
  I think there's a few different areas that
• 2:27
  we're trying hard to pay attention to.
• 2:30
  Stepping back. I think the
• 2:32
  developments
• 2:33
  in AI,
• 2:35
  especially over the last couple of years have
• 2:37
  been
• 2:38
  mind blowing. And I think the potential is
• 2:40
  remarkable.
• 2:42
  I think a few trends that I think
• 2:44
  we're trying hard to keep an eye on
• 2:45
  is
• 2:47
  this traditional machine learning, which is well understood
• 2:50
  and or at least much better understood in
• 2:52
  the industry and making sure that we're leveraging
• 2:54
  that where it makes sense is important.
• 2:57
  And separate that a little bit from generative
• 2:59
  AI, which has shown up in the flavor
• 3:01
  of chat GPT
• 3:03
  that is has
• 3:05
  remarkably different potential to help streamline healthcare.
• 3:08
  So I think keeping an eye on both
• 3:10
  capabilities that are emerging from technology, I think
• 3:12
  is important.
• 3:14
  The second part I would say is that
• 3:16
  within healthcare,
• 3:18
  we all get excited by the,
• 3:21
  I will say the shiny objects that have
• 3:24
  to do with frontline clinical care.
• 3:27
  But given that AI
• 3:30
  is an imperfect technology that we're still trying
• 3:32
  to learn how to deploy and use in
• 3:34
  a responsible and safe way,
• 3:36
  it's important to not just
• 3:39
  look at clinical use cases, but to also
• 3:41
  look at
• 3:42
  nonclinical use cases where we can learn how
• 3:44
  to use it, apply
• 3:46
  it in a much safer setting where there
• 3:48
  is minimal risk to patient harm.
• 3:52
  The
• 3:53
  third area that we're trying to keep an
• 3:54
  eye on is
• 3:56
  just technology is an enabler. It's important to
• 3:58
  think in terms of the healthcare problem so
• 4:00
  that it is our responsibility to connect the
• 4:03
  dots from technology to the change that we
• 4:05
  wish to drive within healthcare.
• 4:09
  Absolutely. And I wanna follow-up with a really
• 4:11
  important point you just made about AI's use
• 4:13
  cases beyond just clinical care. As I understand
• 4:17
  it at Cleveland Clinic, you're using AI in
• 4:19
  a lot of different ways right now, including
• 4:21
  accelerating the pace of medical research and
• 4:24
  addressing staffing and organizational
• 4:26
  challenges.
• 4:27
  So Rohit, if you were going to look
• 4:30
  back over the last, let's say 6 to
• 4:31
  18 months, is there any initiative or accomplishment
• 4:34
  that that you're most proud of within this
• 4:36
  space? And and can you tell us a
• 4:37
  little bit about that?
• 4:40
  So we focus on a few different areas
• 4:42
  that actually span both clinical and nonclinical.
• 4:48
  I give a few a few examples from
• 4:50
  perhaps each of those. In clinical care, we've
• 4:53
  been working on an effort to do
• 4:56
  prediction of the onset
• 4:57
  of sepsis.
• 4:59
  Now, sepsis is a
• 5:02
  bloodstream infection that has a very high mortality
• 5:04
  rate,
• 5:05
  and and it kills nearly like a 1000
• 5:07
  people in the country almost every day.
• 5:11
  And with those numbers, it becomes
• 5:15
  quite imperative for us to bring everything we
• 5:17
  can to bear on treating sepsis.
• 5:20
  So that's an area where we're using machine
• 5:22
  learning and AI to predict the onset of
• 5:24
  sepsis,
• 5:25
  to
• 5:26
  integrate that with our clinical care protocols
• 5:30
  so that we can alert our caregivers
• 5:32
  to the onset of sepsis and drive early
• 5:35
  intervention.
• 5:36
  And that is the best known methodology
• 5:38
  to help take care of our patients better.
• 5:40
  So that's a clinical use case and I'm
• 5:42
  quite encouraged by the data that we have
• 5:44
  so far where if we can get roughly
• 5:46
  6
• 5:47
  times better positive predictive value of patients with
• 5:51
  the onset of sepsis
• 5:52
  with very few false alarms. And this is
• 5:54
  critical
• 5:55
  to our caregivers so that their resources can
• 5:58
  be deployed on the patients that need it
• 6:00
  the most.
• 6:02
  That's an example in clinical care.
• 6:05
  For caregiver,
• 6:07
  there's a second area which really has to
• 6:09
  do with work,
• 6:11
  staffing
• 6:12
  and
• 6:15
  caregiver burnout
• 6:16
  where we are leveraging
• 6:18
  AI powered scribes.
• 6:20
  And think of this as physicians and nurses
• 6:23
  spend
• 6:25
  anywhere from
• 6:28
  1 to 3 hours a day
• 6:30
  on clinical documentation.
• 6:32
  Now this documentation is essential for patient care
• 6:36
  and for regulatory reasons,
• 6:37
  but it is very time consuming.
• 6:40
  And one can imagine that technology that can
• 6:43
  automatically listen to a conversation,
• 6:45
  transcribe it, summarize it, has the potential to
• 6:49
  significantly streamline
• 6:51
  the efficiency of the caregivers.
• 6:53
  So that's an area where
• 6:55
  we're actually working with a few different companies
• 6:57
  to test drive their technologies.
• 6:59
  And
• 7:00
  if you're successful in this effort, we will
• 7:03
  have the potential to substantially
• 7:05
  streamline the work of our caregivers.
• 7:08
  A third area, which is still in the
• 7:10
  early stages is actually the other extreme of
• 7:12
  the spectrum, which is back office.
• 7:15
  And there we're focusing on
• 7:17
  the problem of coding.
• 7:19
  Now coding is essentially the task of taking
• 7:23
  a patient encounter
• 7:25
  and the documentation associated with it and generating
• 7:28
  billing codes for reimbursement purposes.
• 7:31
  It's a necessary part of our business,
• 7:33
  but it is also very manual, very cumbersome,
• 7:37
  and very costly. I
• 7:42
  think
• 7:46
  Now if I step back, I think,
• 7:49
  the question you asked a little bit was
• 7:51
  what are the things that we're most proud
• 7:52
  of? I like these initiatives
• 7:55
  in that they touch on different areas of
• 7:58
  our operations.
• 8:01
  The other part I think that's important is
• 8:04
  they're a good set of initial steps to
• 8:07
  take in bringing AI into the organization.
• 8:10
  And they're teaching us the playbook, the methodology
• 8:13
  on
• 8:14
  how to engage with different partners, how to
• 8:16
  bring those technologies in, how to streamline the
• 8:19
  process of leveraging those technologies
• 8:21
  in a clinical care context or a caregiver
• 8:24
  context. And this methodology and playbook is I
• 8:26
  think going to be the foundation for our
• 8:28
  success going forward.
• 8:31
  Wow. Well, those are some really fantastic examples.
• 8:34
  I I appreciate you sharing those with us.
• 8:35
  And that that last one you mentioned about
• 8:37
  back office coding,
• 8:39
  you mentioned that was new, Rohit. Where are
• 8:41
  you at with that? What's what's the timeline
• 8:43
  of that most recent initiative?
• 8:46
  So,
• 8:47
  we're engaging with a variety of partners.
• 8:50
  The timeline, I would say, is we're somewhat,
• 8:52
  I'd say, in the early stages of doing
• 8:55
  a pilot where we've identified
• 8:57
  areas within coding,
• 9:00
  just the whole revenue cycle management in the
• 9:02
  industry is a complicated space.
• 9:04
  There is prior authorization
• 9:06
  for a set of encounters. There is the
• 9:08
  actual coding and billing exercise, and then there's
• 9:11
  a subsequent denials and claims exercise. So revenue
• 9:14
  cycle management is a complicated area.
• 9:16
  Where we are is
• 9:18
  we're pushing on each one of those, but
• 9:20
  we're at different stages on each one. So
• 9:22
  right now, we're currently more focused on the
• 9:24
  middle area, which is converting
• 9:27
  the a patient chart to the corresponding billing
• 9:30
  codes doing and being able to do that
• 9:32
  more automatically and more efficiently
• 9:34
  than we do today. And that's an area
• 9:36
  where we hope to have results within the
• 9:38
  next, I would say, 6 to 12
• 9:40
  That's really interesting. And I imagine someone who
• 9:42
  will be helping you on a lot of
• 9:44
  these initiatives
• 9:45
  is your new chief
• 9:47
  AI officer,
• 9:48
  who you actually just announced today here at
• 9:50
  the end of July.
• 9:52
  You're one of only a handful of health
• 9:54
  systems nationwide that now has a position
• 9:56
  like this in place.
• 9:58
  So can you talk a little bit about
• 9:59
  that? What went into that decision? And,
• 10:02
  do do you think that all health systems
• 10:04
  should be moving in this direction?
• 10:06
  So I'll
• 10:08
  I think it's important to step back, and
• 10:10
  our thinking on this has evolved over time
• 10:12
  as well.
• 10:13
  Our beliefs have evolved over the last couple
• 10:15
  of years to think that both the developments
• 10:18
  in technology and AI
• 10:20
  are actually accelerating
• 10:22
  And the potential for these technologies to transform
• 10:25
  entire industries
• 10:27
  is remarkable.
• 10:30
  You combine that with the
• 10:33
  health care as an industry
• 10:36
  that is in substantial
• 10:38
  need for this transformation.
• 10:40
  So sitting in our shoes, I think there's
• 10:42
  a unique opportunity to say healthcare as an
• 10:44
  industry really needs technology power transformation and the
• 10:47
  emerging technologies have remarkable capabilities.
• 10:51
  So that's what prompted us to say, how
• 10:52
  do we elevate this to a first class
• 10:55
  effort for us?
• 10:56
  And that's what led us to say that
• 10:58
  if it is an important enough area for
• 11:00
  us that has the potential to transform healthcare,
• 11:03
  then it needs to be something where we
• 11:05
  need to make sure that we have
• 11:07
  internal expertise,
• 11:08
  not just working with external partners, but we
• 11:10
  need to have in house expertise that can
• 11:12
  help us navigate
• 11:14
  these waters as technology evolves and as healthcare
• 11:17
  needs evolve.
• 11:19
  I think the related question that you so
• 11:21
  that's what went into our decision to say,
• 11:23
  hey, let's bring on somebody with AI expertise
• 11:26
  into the organization.
• 11:28
  I think the second part of your question,
• 11:30
  should all systems do the same?
• 11:33
  The way I would approach it is I
• 11:35
  think it is important for every system to
• 11:38
  approach AI with the potential for it to
• 11:41
  over time
• 11:42
  transform
• 11:43
  many of their core operations
• 11:46
  ranging from clinical to non clinical. And I
• 11:49
  think that positioning themselves to navigate those waters
• 11:51
  is important.
• 11:53
  Do you need a chief AI officer necessarily?
• 11:55
  Those organizational constructs may vary from organization to
• 11:58
  organization,
• 11:59
  but I would say that there needs to
• 12:01
  be a process by which organizations are being
• 12:03
  intentional
• 12:04
  about the areas that are most important to
• 12:06
  them,
• 12:08
  intersecting them with the potential for technology and
• 12:10
  AI,
• 12:12
  and engaging with the health tech ecosystem
• 12:14
  with that mindset. I think that's an imperative,
• 12:17
  a chief AI officer is obviously interesting construct
• 12:19
  to go about it. But I think that
• 12:21
  every health system should be thinking in those
• 12:23
  terms.
• 12:25
  Absolutely. That makes complete sense. I I appreciate
• 12:28
  you sharing a little bit about your thoughts
• 12:30
  on on that topic,
• 12:32
  which I imagine will be a big point
• 12:34
  of conversation at our upcoming,
• 12:37
  health IT event in Chicago in October, which
• 12:39
  I know you'll be a speaker at. So,
• 12:42
  one thing I did wanna ask you is
• 12:44
  if, and I and I imagine this some
• 12:46
  some thoughts you'll be sharing similarly at that
• 12:48
  event. But is there any advice that you
• 12:50
  would give to evolving leaders serving in similar
• 12:53
  roles like yours?
• 12:55
  What would you share on this podcast while
• 12:57
  you have the ears of of a lot
• 12:59
  of health system leaders from across the country?
• 13:02
  I
• 13:03
  think my learning
• 13:04
  over the last couple of years is
• 13:07
  it's very important
• 13:09
  to be
• 13:11
  intentional
• 13:12
  about
• 13:13
  the problems that we are trying to tackle.
• 13:16
  While the evolutions
• 13:18
  and sort of the developments in technology and
• 13:20
  AI are amazing,
• 13:22
  eventually they need to be in service of
• 13:25
  a healthcare problem or need.
• 13:28
  This becomes all the more important
• 13:30
  because
• 13:31
  technology is just an enabler.
• 13:34
  The adoption of technology, the adoption of AI,
• 13:37
  the culture, the process, the people, the change
• 13:40
  management required to actually drive
• 13:44
  change and change in outcomes and the transformation
• 13:47
  in healthcare is not easy.
• 13:49
  So it's important not to underestimate that adoption
• 13:51
  of AI
• 13:53
  that it makes sense is not easy. So
• 13:54
  it's important to start with a very intentional
• 13:57
  mindset that says, hey, these are our key
• 13:59
  problems.
• 14:00
  This is where technology fits in. And if
• 14:02
  we have technology that works, this is how
• 14:05
  we're gonna drive the change management, the adoption,
• 14:08
  and the process changes that need to go
• 14:10
  in
• 14:11
  into,
• 14:12
  actually helping transform healthcare.
• 14:14
  Just as an example, if you look at
• 14:16
  the sepsis example that I use where I
• 14:18
  said, hey, we're using AI for the prediction
• 14:20
  of the onset of sepsis.
• 14:22
  That's only part of the story.
• 14:24
  Those
• 14:25
  technology powered alerts
• 14:28
  need to be baked into the clinical workflows
• 14:30
  that are used by nurses and are used
• 14:32
  by doctors so that they adopt them, they
• 14:35
  take action based on them and it drives
• 14:37
  the right interventions.
• 14:39
  So the process change cannot be underestimated.
• 14:42
  So I think those are the 2 aspects,
• 14:43
  which is be very intentional about the problem
• 14:45
  areas that you're trying to go after and
• 14:47
  be very mindful that it's not just about
• 14:49
  technology. The change management is a significant lift.
• 14:52
  Mhmm. Well, it's it's great and timely advice.
• 14:55
  So appreciate you sharing that. Anything else we're
• 14:58
  missing? Rohit, any any final thoughts you'd like
• 15:01
  to share?
• 15:02
  I think the main thing is that,
• 15:06
  like the developments in technology are fascinating.
• 15:09
  It becomes tricky on how to navigate them.
• 15:11
  All of the things that I alluded to,
• 15:13
  which is, are you intentional about the problems?
• 15:15
  Are you intentional about
• 15:17
  the change management? All of this is hard
• 15:20
  work. I think at the same time,
• 15:22
  the part that fascinates and excites me is
• 15:27
  the potential for these technologies
• 15:29
  is amazing. One can imagine a future where
• 15:34
  the
• 15:35
  amazing amount of inefficient back office paperwork that
• 15:39
  we do is mostly automated
• 15:41
  with
• 15:42
  technology, AI, and generative
• 15:45
  AI. One can imagine clinical care that is
• 15:47
  increasingly powered by a highly intelligent agent.
• 15:51
  I don't know the path to get from
• 15:53
  here to there and I don't know how
• 15:54
  long it will take, but I am convinced
• 15:57
  that if you navigate these waters well,
• 16:01
  this technology has amazing potential
• 16:04
  to transform healthcare and make it
• 16:07
  more accessible,
• 16:09
  safer,
• 16:10
  more scalable, and available to many more patients.
• 16:13
  So I'm very intrigued by the potential
• 16:15
  and that's a journey that I think they're
• 16:17
  all on.
• 16:20
  Fantastic. It's exciting stuff. So, doctor Chandra, thank
• 16:23
  you for taking the time to sit down
• 16:25
  with us and for sharing your insights with
• 16:27
  our listeners.
• 16:30
  You're welcome, Jacob.
• 16:31
  If you'd like to listen to more podcasts
• 16:33
  from Becker's Healthcare, you can visit beckershospitalreview.com.