Latest / Becker’s Healthcare Podcast / Sudipto Srivastava, Chief Data and Analytics Officer at Montefiore Health System

Transcript

Sudipto Srivastava, Chief Data and Analytics Officer at Montefiore Health System · Becker’s Healthcare Podcast

• 0:00
  Welcome, everyone, to the Becker's Healthcare Podcast. I'm
• 0:03
  Laura Dirda, vice president, editor in chief here
• 0:06
  at Becker's Healthcare, and I'm thrilled today to
• 0:07
  be joined by Sudipto Srivastava,
• 0:10
  the chief data and analytics officer at Montefiore
• 0:13
  Health System.
• 0:14
  Sudipto, it's a pleasure to have you on
• 0:15
  the podcast today.
• 0:18
  Thanks. It's great to be here. Big fan
• 0:19
  of Becker. So, so excited, for this opportunity,
• 0:22
  Laura.
• 0:23
  Absolutely. And we're big fans of yours and
• 0:25
  everything that you're doing at Montefiore Health System
• 0:28
  as well. I think it's such an interesting
• 0:29
  time in health care, and we're certainly excited
• 0:32
  for our conversation.
• 0:34
  Indeed. Yeah. It is very exciting.
• 0:36
  Absolutely. Well, I know we'll talk about a
• 0:38
  good many things in thinking about technology and
• 0:41
  data and AI and more. But before we
• 0:43
  do, can you introduce yourself and tell us
• 0:45
  a little bit more about the health system?
• 0:48
  Absolutely.
• 0:49
  And, actually, even before I do that,
• 0:51
  you know, because this is an audience that's
• 0:53
  near and dear to me, my, other colleagues
• 0:56
  in health care and and tech who are
• 0:58
  trying to solve big problems. I wanna maybe
• 1:00
  start with, a couple of hypothesis that as
• 1:03
  we're going through our discussion today, Laura, I
• 1:05
  just want, the audience to keep in the
• 1:07
  back of their minds.
• 1:09
  Number one,
• 1:11
  do you think that we have solved all
• 1:13
  of our problems in health care?
• 1:16
  And number two,
• 1:19
  given all those sort of focus on AI
• 1:21
  and things like that, but how would you
• 1:23
  change work
• 1:25
  if you had access to several really smart
• 1:27
  PhDs who are working for you? Of course,
• 1:30
  I'll be with a little contact with little
• 1:33
  context for operational and cultural context that you
• 1:35
  work in, like these, you know, AI tools
• 1:38
  can do that. But just how would you
• 1:39
  change?
• 1:40
  And I want the the the audience to
• 1:42
  keep that as a frame in the back
• 1:43
  of their minds because, you know, we'll be
• 1:44
  talking a lot about it. And I think
• 1:47
  as we have the conversation, it might,
• 1:49
  you know, prompt some thoughts and ideas
• 1:53
  that in their in their own heads, which
• 1:54
  is, of course, the goal of this discussion.
• 1:58
  Anyway, so with that said, you know, just
• 1:59
  to let me tell a little about, you
• 2:01
  know, our organization and myself, you know, Montefiore
• 2:04
  Health System, Montefiore Einstein,
• 2:06
  is a 13 hospital health system, covers the
• 2:08
  New York City, Bronx, Westchester, Hudson Valley area.
• 2:11
  About 7,500,000
• 2:13
  encounters in a year, about 36, 38, 7,000
• 2:17
  employees.
• 2:18
  We also are an academic medical center, in
• 2:20
  the Albert Einstein School of Medicine.
• 2:22
  And the, the interesting fact that some of
• 2:25
  your audience may or may not know is
• 2:27
  that we have a $0
• 2:29
  MD student tuition,
• 2:31
  because this was made possible by,
• 2:35
  doctor Ruth Gottesman who announced in February 2024
• 2:39
  that all current and future medical students will
• 2:41
  have their tuition waived in perpetuity
• 2:44
  through a billion dollar donation. So that's something
• 2:47
  that is fascinating,
• 2:48
  and we love, sort of sharing that back.
• 2:51
  As for me, you know, my title is
• 2:53
  VP data analytics,
• 2:54
  and my role includes,
• 2:56
  oversight for
• 2:58
  data analytics, organizational
• 3:00
  reporting,
• 3:01
  AI governance, and also research IT, partnering with
• 3:04
  our sort of school
• 3:06
  of medicine.
• 3:07
  And then as we sort of think of
• 3:10
  roles, titles, and what we,
• 3:12
  a lot what we'd like to
• 3:15
  pivot on within our organization is
• 3:18
  that we have made a conscious decision not
• 3:20
  to create a separate, like, say, czar for
• 3:23
  AI or anything else. And, again, there's nothing
• 3:25
  wrong with organizations that are doing it, but
• 3:27
  our philosophy is that
• 3:29
  AI and tech is
• 3:31
  everyone's responsibility.
• 3:33
  And we find that this frees up our
• 3:34
  staff and docs to explore things
• 3:36
  from their unique vantage unique vantage point.
• 3:40
  Now, of course, we have guardrails,
• 3:42
  and but we wanna be deliberate in not
• 3:44
  taking and making AI
• 3:46
  an individual's response.
• 3:49
  Well, that makes a ton of sense. And,
• 3:51
  you know, thank you so much for that
• 3:52
  explanation. It's, so fascinating to see the different
• 3:55
  perspectives that health systems are taking as technology
• 3:58
  and AI move so quickly. And
• 4:00
  I I love the idea that it's everyone's
• 4:02
  responsibility that you want everyone from their seat,
• 4:05
  to to take that unique vantage point and
• 4:08
  really figure out how they can apply it
• 4:10
  within their own roles and workflows and,
• 4:13
  get the best results possible. So that that's,
• 4:16
  really cool to hear about. Exactly.
• 4:18
  When you think about the last year or
• 4:20
  so, what was the most most important initiative
• 4:22
  that you led? What did you do, and
• 4:23
  what were the results?
• 4:25
  Yeah. Well,
• 4:27
  you know, plenty with, like, 37,000
• 4:29
  employees,
• 4:31
  thinking and working every day, smart bunch of
• 4:33
  clinicians and staff and nurses that we have.
• 4:35
  So lots, I guess, you know, in terms
• 4:38
  of this conversation, I'll pick up a few
• 4:40
  to talk about.
• 4:42
  And, of course, as with every self respecting
• 4:44
  health system, you know, we launched, Ambient Voice
• 4:47
  AI solutions.
• 4:48
  You know, so we have over
• 4:51
  550,
• 4:52
  clinicians enrolled in that, you know, covering areas
• 4:55
  such as primary care, medical and surgical specialties,
• 4:57
  and we have an ongoing demand
• 5:01
  of additional departments that want to use, ambient
• 5:04
  scribes.
• 5:05
  We have a very successful implementation there. We
• 5:08
  have over 48% high u utilizers. You know,
• 5:11
  these are folks
• 5:12
  that who are using,
• 5:14
  scribes more than 50% of the times in
• 5:16
  their sort of visit and over 50% of
• 5:18
  their visits.
• 5:19
  So we've had about,
• 5:21
  like, over 300,000
• 5:23
  encounters,
• 5:24
  and, you know, these things have continues to
• 5:27
  grow. And, you know, back to the question
• 5:29
  that we sort of prompted the audience to
• 5:31
  think of is, you know, think of, you
• 5:33
  know, have you solved all problems in health?
• 5:35
  And I think Ambient Scribe becomes a very
• 5:37
  interesting one because
• 5:39
  this was a problem that
• 5:41
  we like, over the last
• 5:44
  few decades, we kinda created for our doctors
• 5:46
  and nurses where documentation burden sort of, kept
• 5:49
  growing. And here we have a technology solution
• 5:51
  that allows them to, you know, solve that
• 5:54
  problem, that that classical
• 5:56
  pajama time that they were sort of all
• 5:58
  giving. And you've done many of podcasts with
• 6:00
  Becker's, and articles around that. So I want
• 6:02
  to report the audience with it. But, you
• 6:04
  know, here we have is a huge problem
• 6:07
  that now we that we can try to
• 6:09
  solve with, with tech.
• 6:12
  Continuing with, you know, our other initiatives,
• 6:15
  you know, we also launched our AI governance
• 6:17
  framework and our AI policy for the entire
• 6:20
  health system to guide how things go.
• 6:22
  You know? And we've had our evolution in
• 6:24
  that. Last year, when we started, we launched
• 6:27
  what we call governance one point o. We
• 6:29
  learned from their lessons. The starting of this
• 6:31
  year, we launched governance two point o. And
• 6:33
  we're already thinking about governance three point o.
• 6:36
  And if your audience is saying, well, what's
• 6:38
  the difference between each of these? I think,
• 6:41
  you know, as we went through our phases,
• 6:43
  we started thinking of AI governance
• 6:46
  in terms
• 6:47
  of quick decision making because we were facing
• 6:49
  a lot of volume of requests that are
• 6:51
  coming in.
• 6:52
  And then as we were able to stabilize
• 6:54
  that, we realized that we needed, you know,
• 6:56
  significant input in governance two point o from
• 6:59
  all walks of our
• 7:01
  ecosystem. You know, of course, the clinical teams,
• 7:04
  the nursing teams, as well as people in
• 7:06
  finance and
• 7:08
  legal and privacy and cyber and so on.
• 7:11
  So that is the combination of our governance
• 7:14
  two point o. And then, you know, we're
• 7:16
  seeing
• 7:17
  how even this process can be improved given
• 7:20
  the new volume of request that we're getting
• 7:22
  on AI, and we're starting to think of
• 7:24
  governance sort of three point o. So it's
• 7:26
  a very, you know, deliberate iterative process, and,
• 7:28
  you know, we're really sort of proud of
• 7:30
  what we've done. You know, we've,
• 7:33
  looked at
• 7:34
  over, like, 80 like, we've launched over 80
• 7:37
  some solutions. We have,
• 7:39
  several,
• 7:40
  close to a 100 more in the pipeline
• 7:42
  and covers the entire gamut of, like, radiology
• 7:45
  and acute care, cardiology, revenue cycle, patient access,
• 7:49
  GI, IT security, nursing, just to sort of,
• 7:52
  you know, name a few.
• 7:54
  Now with that sort of said, we also
• 7:56
  feel that
• 7:58
  the most important initiatives are also the ones
• 8:00
  where we have deep partnerships with our business.
• 8:03
  So in this area, we've done a lot
• 8:04
  of work in our with our population health
• 8:06
  teams and within our care management organization,
• 8:09
  you know, when when it comes to having
• 8:11
  data on
• 8:13
  care gap closures, having data on where they
• 8:15
  how are they're they're doing with their heat
• 8:17
  is measures. On the clinical side, you know,
• 8:19
  we've, started
• 8:21
  auto tagging of sort of cancer types using
• 8:23
  GenAI tools, like the bio clinical BERTs sort
• 8:26
  of,
• 8:27
  capabilities.
• 8:28
  On the research side, we actually entered last
• 8:30
  year into a partnership with Dandelion Health.
• 8:33
  You know? And there's details written up about
• 8:36
  it, but, essentially, you know, we will
• 8:40
  partner we are part of a consortium of,
• 8:42
  four other health systems around the country
• 8:45
  which are offering de identified data
• 8:48
  to, you know, test out,
• 8:51
  AI tools, AI algorithms,
• 8:54
  advanced research, all in a safety and invite
• 8:56
  sort of framework. So we're very proud of
• 8:58
  that sort of partnership.
• 9:00
  And that's just to name a few. You
• 9:01
  know? I'm,
• 9:02
  of course, not talking about, you know, things
• 9:04
  that we did with our pharmacy. We have
• 9:05
  a strong one Amazon One Medical partnership and
• 9:08
  so on. But I'll I'll I'll shut up
• 9:09
  for now, and hopefully that gives you a
• 9:12
  little bit of a perspective of,
• 9:14
  what we've been doing.
• 9:16
  Absolutely. That's incredible. I mean, you know, to
• 9:19
  have so many different AI solutions in the
• 9:21
  pipeline close to a 100 is,
• 9:24
  really, really a huge accomplishment right now. And
• 9:26
  I, appreciate you talking through, you know, whether,
• 9:30
  the governance process, that structure. I know a
• 9:32
  lot of health systems and leaders that I
• 9:35
  talk to, that's a sticking point for them,
• 9:37
  trying to figure out how they're incorporating
• 9:39
  AI smartly into that governance process to make
• 9:42
  sure they're not missing anything, but also not
• 9:44
  reinventing the wheel and keeping the aspects of
• 9:46
  governance
• 9:47
  in place that have worked across the board.
• 9:50
  So it seems like you know, being able
• 9:52
  to move quickly is important, but obviously you
• 9:54
  want to mitigate the risks and make sure
• 9:55
  you're not moving too quickly and finding that
• 9:57
  right balance,
• 9:59
  is critical, it seems like.
• 10:01
  Indeed. Ashley Inkumsah: Absolutely. Well, I'm curious, you
• 10:04
  know, kind of building upon all of these
• 10:06
  things that you've been talking about, these different
• 10:08
  AI applications and really incorporating that into,
• 10:12
  the the daily workflows for your team members,
• 10:14
  what do you see as being some of
• 10:16
  the big priorities and headwinds that you're focused
• 10:18
  on for the rest of this year?
• 10:21
  Oh, yeah.
• 10:23
  Well, I think,
• 10:24
  from a when it comes to priorities,
• 10:27
  you know, we to the point that I
• 10:28
  made earlier, we really wanna align with our
• 10:30
  business and clinical teams in, you know, exploring
• 10:33
  the problems that
• 10:35
  they face day to day and how it
• 10:37
  can be all through tech, through AI, and
• 10:39
  and other tools. And in that, you know,
• 10:41
  there are certain areas that have popped up,
• 10:44
  big in that space.
• 10:46
  You know, many of your our our colleagues
• 10:48
  here who are listening in will have been
• 10:51
  thinking about rev cycle.
• 10:53
  And, you know, we see a huge opportunity
• 10:56
  in revenue cycle, whether it's in terms of,
• 10:59
  you know, denials management, prior authorization,
• 11:02
  you know, other sort of functional areas where,
• 11:05
  you know, know, we have a huge opportunity
• 11:08
  to work better, to have better throughput, get
• 11:12
  better outputs, you know, as we partner with
• 11:14
  our, health,
• 11:15
  payers as well. Supply chain is another huge
• 11:18
  area.
• 11:20
  You know, if you think about,
• 11:21
  you know and most health systems, you know,
• 11:24
  supply chain is a significant significant probably number
• 11:27
  two or number three when it comes
• 11:29
  to the amount of cost and revenue that
• 11:31
  they sort of have to handle.
• 11:34
  And, you know, they look at contracts.
• 11:37
  There's a lot of information that they have
• 11:39
  to extract. You know, they have so many
• 11:41
  different vendors looking across the different
• 11:44
  vendor landscape. You know, I kinda always say
• 11:46
  it's, you you know, they're buying pencils to
• 11:48
  MRI machines.
• 11:50
  So, you know, have using AI in a
• 11:53
  thoughtful manner there
• 11:55
  is huge.
• 11:56
  HR is another area, you know, from a
• 11:58
  business perspective, you know, because we get a
• 12:00
  lot of internal questions around basic things like,
• 12:02
  you know, what's the vacation policy, how it's
• 12:04
  happening,
• 12:05
  what are the rules around x, y, and
• 12:07
  z
• 12:08
  that we get a lot of sort of
• 12:09
  staffing separate question.
• 12:11
  On the clinical side, you know, there's a
• 12:14
  a ton that we're thinking of in diabetes
• 12:16
  management outreach
• 12:18
  and population health care gap closures
• 12:20
  in our sort of surgical areas. You know,
• 12:22
  we do so much in our sort of
• 12:24
  peri ops
• 12:25
  space, with that. So, you know, all those
• 12:29
  are active exploration and discovery
• 12:32
  and sometimes even some deeper that's happening. And
• 12:35
  finally, from a tech perspective, from a priorities
• 12:38
  point of view, you know, we're looking at
• 12:41
  conversational AI
• 12:42
  because, you know, that allows our nurses
• 12:45
  and sort of front desk staff to,
• 12:48
  you know, focus on the higher acuity sort
• 12:50
  of, you know, patients. And so we're looking
• 12:52
  into sort of implementing tools in that space.
• 12:54
  And we're also, you know, starting to look
• 12:57
  deeper into the partnerships that we already have
• 12:59
  with our sort of vendors. I mean, you
• 13:00
  know,
• 13:02
  we of course, it's the we use Epic
• 13:04
  as our EMR, and Epic has been, you
• 13:06
  know, pretty amazing in terms of, exposing tools
• 13:11
  in the space of AI and productivity,
• 13:13
  and we're actively,
• 13:15
  looking at those. We have about 10 or
• 13:17
  12 act in in our pipeline that are
• 13:19
  going some many of them are live in
• 13:21
  use right now. Many are about to sort
• 13:23
  of go live, and we have a longer
• 13:24
  sort of road map of how we switch
• 13:26
  those capabilities on. And it doesn't just stop
• 13:29
  with our EMR systems. You know, we're looking
• 13:31
  at our operational systems. You know, we use
• 13:33
  ServiceNow.
• 13:34
  What can we use there to increase efficiency
• 13:36
  and respond better to our customer requests? You
• 13:39
  know, Workday.
• 13:40
  We went through that implementation,
• 13:42
  and Workday itself has tools. And as you
• 13:44
  as you look at our at our application
• 13:47
  landscape, every
• 13:48
  vendor is trying to sort of build build
• 13:51
  tools in for added efficiency, cost savings, and
• 13:53
  so on. So, you know, those would be
• 13:55
  sort of the a little bit of a
• 13:56
  smattering of the priorities that we're looking at.
• 13:59
  I think you asked about headwinds as well.
• 14:02
  So, well, there there are plenty, and your
• 14:05
  audience is,
• 14:06
  is no stranger to those.
• 14:08
  You know, I like to say that
• 14:11
  health care likes to go through,
• 14:13
  not likes to. It just goes through feast
• 14:15
  and famine cycles, you know, when it comes
• 14:17
  to funding and so on. So, you know,
• 14:19
  as we look at 2026 and 2027,
• 14:22
  the the macro environments will kind of force
• 14:24
  us to,
• 14:26
  focus a lot on the cost savings.
• 14:29
  You know, the other part will be filtering,
• 14:32
  you know, signal from the noise. You know?
• 14:35
  Health care is embracing,
• 14:37
  unlike
• 14:38
  like many or similar to other industry,
• 14:42
  but I think you need to health care
• 14:43
  now is we are embracing,
• 14:47
  AI at a faster pace. You know, the
• 14:48
  interest is there both the leadership level, the
• 14:50
  clinical level, the nursing level to do that.
• 14:53
  And then how do you filter out the
• 14:55
  signal from the noise? What is worth investing
• 14:57
  in? What is not worth investing in?
• 14:59
  The other headwinds are around protecting our data
• 15:02
  and our information, you know, because,
• 15:04
  I mean, of all the big bold ideas
• 15:06
  that we have, we have to be mindful
• 15:08
  that, you know, we are we are guard
• 15:10
  you know, we have to protect our patients'
• 15:11
  set of data.
• 15:14
  You know, we there there's always a potential
• 15:17
  for breaches, so we partner very closely with
• 15:19
  our, you know, cybersecurity
• 15:20
  teams.
• 15:22
  Especially with AI. There's new aspects of data
• 15:25
  come that come in, which is,
• 15:26
  you know, what is the vendor doing with
• 15:28
  your data? How are they training this?
• 15:31
  Partner with,
• 15:33
  you know, the large LLMs.
• 15:35
  You know, what is going on in terms
• 15:37
  of the safety of our data? What kind
• 15:39
  of agreements we need to have there? So
• 15:41
  those are, like, going to be I mean,
• 15:43
  they've always been challenges for us, but I
• 15:44
  think they will continue to be challenges.
• 15:47
  And finally, just, you know,
• 15:49
  the other headwind is, like, we have to
• 15:50
  go be steady in in the storm. I
• 15:52
  think there's so much
• 15:54
  news, media, articles
• 15:57
  that's coming out. All the things about the
• 15:59
  glorious things that AI can do and all
• 16:01
  the mistakes that it can do and all
• 16:03
  the doom,
• 16:04
  that it can cause or all the amazing
• 16:07
  things it can do. So staying true to
• 16:09
  our mission
• 16:11
  and aligning with the needs of our staff,
• 16:13
  doctors, and nurses,
• 16:15
  that will be a big sort of headwind
• 16:17
  that we have to manage as well. Anyway,
• 16:19
  I said a lot,
• 16:20
  but that's, that's what we're sort of tackling
• 16:22
  right now.
• 16:24
  Yeah. Absolutely. I mean, you know, it's a
• 16:27
  huge space and and certainly a lot of
• 16:29
  challenges out there,
• 16:31
  across the board. And so, you know, when
• 16:32
  you're looking at,
• 16:34
  the opportunities
• 16:35
  out there, I I think, so many possible
• 16:39
  partnerships you could have, new companies coming up
• 16:41
  every day as well as, like you mentioned,
• 16:43
  you know, existing,
• 16:45
  partners coming out with new AI applications
• 16:48
  in ways to ideally increase efficiencies and and
• 16:51
  become, more critical within your daily workflows. So
• 16:54
  I'm curious when you're evaluating all the opportunities
• 16:58
  out there and and trying to bring, your
• 17:00
  your teams along, and especially in health care,
• 17:03
  you know, there can be some,
• 17:04
  resistance to change in in certain ways. How
• 17:07
  do you, kind of bridge that gap and
• 17:09
  really make sure you're, one, selecting the right
• 17:12
  things, and then, two,
• 17:13
  putting your team in the position to actually
• 17:15
  use it so, you know, these,
• 17:18
  new applications that you're bringing in or or
• 17:20
  working through
• 17:21
  aren't left dormant,
• 17:22
  you know, a year from now.
• 17:25
  Yeah. No. I think, you know
• 17:27
  and this is where all '37, thirty eight
• 17:30
  thousand employees, you know, have a role to
• 17:32
  play.
• 17:33
  And,
• 17:34
  you know, there are many aspects that go
• 17:36
  in.
• 17:37
  One, we don't wanna
• 17:39
  force
• 17:40
  AI down
• 17:41
  any sort of, particular area. You know, there
• 17:44
  has to be a conversation and understanding of
• 17:48
  what it does, what it doesn't do. There's
• 17:50
  a lot of,
• 17:51
  training and coaching that goes in there in
• 17:53
  terms of what is the value of that,
• 17:56
  you know, talking and being actually honest about
• 18:00
  some of the skepticism that people have. And
• 18:02
  that is a journey. And sometimes you take
• 18:03
  two steps forward and one step backward.
• 18:06
  You know, when it comes to, you know,
• 18:08
  people
• 18:09
  in teams who may be genuinely,
• 18:11
  you know, nervous and anxious about the quality
• 18:14
  of the tools, the,
• 18:15
  stability of these tools, the accuracy of these
• 18:17
  tools. So, you know, so we have to
• 18:19
  definitely
• 18:21
  sort of do that. We also,
• 18:23
  you know, have certain measurements. So we haven't
• 18:26
  talked a lot about, the control towers and
• 18:28
  the control centers that we have to overall
• 18:31
  monitor the use of AI. And not just
• 18:33
  monitoring from a performance drift other perspective, but
• 18:36
  also from a, hey. Did it meet the
• 18:39
  hypothesis that you had going into it?
• 18:41
  Because if you think about it, you know,
• 18:43
  many large organizations like ours and especially in
• 18:46
  health care,
• 18:47
  you know, I like to say that we're
• 18:48
  really good at addition, but not good at
• 18:50
  subtraction.
• 18:51
  And And let me tell you what I
• 18:52
  mean by that. What I mean is we're
• 18:54
  really good at adding more and more
• 18:56
  solutions and tech to it, but we're not
• 18:59
  as deliberate about
• 19:00
  subtracting it and taking them away. So we've
• 19:03
  been very conscious in terms of
• 19:06
  our approach to tech and AI
• 19:09
  is as we work through our governance process,
• 19:12
  we will go back to solutions and say,
• 19:15
  you have you know, when you launch a
• 19:17
  solution and when AI governance approved it,
• 19:20
  the hypothesis was it was going to do
• 19:22
  x, you know, x number of patient encounters
• 19:25
  and improve win productivity or maybe some dollar
• 19:28
  threshold or numbers or volume thresholds or something
• 19:31
  something something.
• 19:32
  Is it doing that? And let's be very
• 19:34
  honest to measure if it's doing that. And
• 19:38
  if it isn't,
• 19:39
  let's take it out. Let's take it out
• 19:41
  of our ecosystem
• 19:42
  and do that. And trust me, these are
• 19:44
  not easy conversations to have.
• 19:47
  But that is a framework that is needed
• 19:49
  so that, you know, two years, three years
• 19:51
  down the line, we're not taking on so
• 19:53
  much tech debt
• 19:55
  that we have to, you know, have a
• 19:57
  separate exercise to unwind all all of that.
• 20:00
  Does that make sense?
• 20:02
  Yeah. That's so helpful to understand how you're
• 20:04
  thinking about these things and, you know, really
• 20:06
  truly trying to avoid that,
• 20:08
  technical debt for sure.
• 20:10
  I'm curious, you know, when looking ahead, what
• 20:12
  do you think the hardest thing you'll have
• 20:13
  to do in the coming year will be?
• 20:16
  Yeah. I think,
• 20:18
  and again, this is something maybe your audience
• 20:20
  can appreciate as well,
• 20:23
  especially if it's, in a large health system
• 20:26
  like ours with an academic medical center as
• 20:28
  well, is just picking the right problems to
• 20:31
  solve. You know, I like to joke with
• 20:33
  my not joke, but at least have a
• 20:34
  conversation with my stakeholder team and my leadership
• 20:37
  is
• 20:38
  we have a 100 problems to solve,
• 20:41
  but the funding and resources to do 17
• 20:43
  of them.
• 20:45
  And then how do we do that? So
• 20:47
  and it is very hard because,
• 20:50
  you know, when someone comes in and says,
• 20:52
  this is going to help me
• 20:54
  detect
• 20:55
  a particular disease
• 20:57
  three days earlier, five days earlier, or this
• 20:59
  will have a clinical impact on someone who
• 21:02
  is working with diabetes, or here's a cost
• 21:05
  saving or split. It is very hard to
• 21:07
  make those decisions to pick those 17 or
• 21:10
  20 out of the 100,
• 21:12
  But we have to because we don't have
• 21:13
  unlimited resources.
• 21:15
  So allocating our funds
• 21:17
  wisely in that area, investing in the right
• 21:20
  things where we see the right quote unquote
• 21:22
  me. I'm doing air quotes right now ROI
• 21:24
  because ROI is defined very differently for every
• 21:27
  organization. So I'm looking at it from a
• 21:29
  purely DARS perspective. Some,
• 21:31
  identify other factors. You know, we, of course,
• 21:33
  have our framework for that. But
• 21:35
  it really will be how do we pick
• 21:37
  the right problems to solve.
• 21:40
  And to the question that I asked, sort
• 21:41
  of your audience to think through is, you
• 21:43
  know, do we feel that we have solved
• 21:46
  all our problems? So as we look at
• 21:47
  our health systems, what is the biggest one?
• 21:49
  Is access big? Is reaching out to a
• 21:51
  certain patient population big?
• 21:53
  Is, looking at the high acuity patients, you
• 21:56
  know, who may have a lung nodule that
• 21:57
  goes
• 21:58
  undiagnosed for a period of time to get
• 22:00
  a referral appointment back? Like, so there's so
• 22:02
  many of them, but just picking the right
• 22:04
  problems, I think,
• 22:05
  that has been always a challenge. But I
• 22:07
  think especially with the supercharged environment of tech
• 22:10
  that we're in right now,
• 22:11
  that I feel will be a pretty hard
• 22:13
  thing,
• 22:15
  for this year and many years to come.
• 22:19
  That makes a lot of sense. And certainly,
• 22:21
  you know, having that strong process, as you
• 22:24
  mentioned earlier,
• 22:25
  can make a big difference in in keeping
• 22:27
  things,
• 22:28
  you know, relatively focused in on what's most
• 22:31
  important for the health system, tying back to
• 22:33
  the mission and vision of where you wanna
• 22:35
  go. Before we wrap up here, I wanna
• 22:37
  talk about growth too. Where do you see
• 22:39
  some of the best opportunities for organizational growth
• 22:42
  in the future?
• 22:44
  Yeah. Well, that's a fascinating question.
• 22:48
  You know, if if I have to pick
• 22:51
  maybe
• 22:52
  one,
• 22:55
  I would say it's
• 22:56
  I see a unique opportunity in redefining
• 23:00
  the way we work.
• 23:03
  And, you know, and to those in the
• 23:04
  audience, you know, who are either rolling their
• 23:07
  eyes because every consultant walking to the door
• 23:09
  right now is using these big words,
• 23:11
  but
• 23:12
  I think there is a huge opportunity there.
• 23:15
  So
• 23:16
  maybe just let me share a story, a
• 23:18
  small one.
• 23:20
  So I was speaking recently to someone who
• 23:22
  is in ER radiology and also teaches radiology
• 23:25
  residents.
• 23:26
  And what they've been observing, they said, was
• 23:28
  that first year residents in radiology are catching
• 23:31
  up quite fast with their third year or
• 23:33
  even fourth year colleagues.
• 23:35
  And their hypothesis, again, not proven,
• 23:37
  was that this is due to AI because
• 23:40
  a lot of first year residents are adept
• 23:42
  at AI and are using it different ways.
• 23:45
  And then the discussion evolved into
• 23:48
  what can be done to further expand the
• 23:50
  scope and scale of what is taught to
• 23:52
  residents
• 23:54
  using now, perhaps AI as a tool. So
• 23:56
  now I'm not clinical, so I'm paraphrasing
• 23:58
  some of what they said. But can AI
• 24:02
  be used to expose
• 24:04
  radiology residents to other business fields that they
• 24:07
  interact with? Many earlier so much earlier. So
• 24:10
  that instead of looking at certain things from
• 24:13
  a very silo lens of radiology,
• 24:15
  they can look at it from a lens
• 24:16
  that a cardiologist would look at or a
• 24:18
  neurologist would look at. So now doctors in
• 24:21
  the audience may say, well, that's a cockamamie
• 24:23
  idea, but which is why I'm, you know,
• 24:24
  paraphrasing
• 24:25
  someone else, you know, who is way more
• 24:27
  intelligent than than I am in the clinical
• 24:30
  sphere. But I give that as a as
• 24:31
  an example to say,
• 24:33
  now
• 24:34
  imagine
• 24:36
  your processes, your workflows, the things that you
• 24:40
  do, whether it's operations or clinical or nursing,
• 24:44
  or call centers,
• 24:46
  supply chain or population health. Pick an area.
• 24:49
  Look upstream and downstream.
• 24:51
  Are there ways that we can redefine things
• 24:53
  a little bit? I think we have a
• 24:55
  once in a
• 24:57
  long time opportunity to be able to do
• 24:59
  that because we now have the tech tools,
• 25:01
  the AI tools, the capability, the desire, the
• 25:04
  capital required for it, the funding needed for
• 25:07
  it, the leadership and the board level interest
• 25:09
  in that, that it will be a missed
• 25:11
  opportunity if we don't do that from an
• 25:13
  organizational growth perspective. And this won't be a
• 25:15
  one year effort. This might be multiyear efforts
• 25:18
  and so on,
• 25:19
  but
• 25:20
  it allows us to focus on that.
• 25:24
  Yeah. Yeah. Definitely. I I think that makes
• 25:25
  a lot of sense. And, you know, it's
• 25:27
  really helpful to have that type of,
• 25:29
  focus in and then understanding
• 25:31
  where, you know, you're going from here. So
• 25:33
  I appreciate it. Sudipto, thank you so much
• 25:36
  for joining us on the podcast today. This
• 25:37
  has been a fun conversation. I'm really kind
• 25:39
  of focused, and I appreciate your time and
• 25:41
  energy on it and look forward to continuing
• 25:43
  the conversation soon. This was fascinating. Laura, I
• 25:46
  love your podcast, and I'm glad that we
• 25:48
  had an opportunity to discuss.