Latest / Becker’s Healthcare Digital Health + Health IT / Scaling AI Across Healthcare With Better Data, Governance and Clinical Context with Nari Gopala

Transcript

Scaling AI Across Healthcare With Better Data, Governance and Clinical Context with Nari Gopala · Becker’s Healthcare Digital Health + Health IT

• 0:00
  - What we wanna do with AI is help
• 0:01
  - clinicians go from being reactive
• 0:04
  - to proactive.
• 0:05
  - At Philips, we can recognize patterns in monitoring
• 0:08
  - data across populations
• 0:10
  - and apply those insights to specific patients.
• 0:14
  - Safe and responsible AI can support sharper insights,
• 0:18
  - pave the way for faster diagnosis,
• 0:21
  - and help streamline care.
• 0:23
  - That can benefit clinicians, it can benefit health
• 0:26
  - systems, and most importantly, it can benefit patients
• 0:29
  - and their families.
• 0:33
  - Phillips.
• 0:34
  - Hi, everyone. This is Laura Dirdo with Becker's
• 0:36
  - Healthcare. Thank you so much for tuning in
• 0:38
  - to the Becker's Healthcare podcast series. Joining me
• 0:40
  - for today's discussion is Nari Gopala, chief digital
• 0:43
  - officer at Mayo Clinic. Nari, it's a pleasure
• 0:45
  - to have you on the podcast today.
• 0:47
  - Thank you. Thank you for the, opportunity to
• 0:49
  - be here. Absolutely. Now I'm looking forward to
• 0:52
  - our conversation. I know we're gonna dig into
• 0:55
  - thinking about patient data, access, and more. And
• 0:58
  - so,
• 0:59
  - we'll just dive right in. Health systems are
• 1:01
  - receiving a growing stream of patient data from
• 1:03
  - bedside monitoring, ambulatory devices, and and care delivered
• 1:06
  - at home. Where is the data creating actionable
• 1:09
  - insights for clinicians, and where does it still
• 1:11
  - create more work than insight?
• 1:14
  - That's a great question. I think, especially
• 1:17
  - as we think about large health systems like
• 1:20
  - ours where
• 1:21
  - as you transition from what used to be,
• 1:24
  - like, collect all the data that you can
• 1:26
  - and and and get as much signal as
• 1:28
  - you can from the patient even if they
• 1:30
  - are not in the, in the facility or
• 1:32
  - in the hospital, I think that thinking has
• 1:35
  - changed in the world quite a bit because
• 1:36
  - I think, ultimately,
• 1:38
  - you want to
• 1:40
  - separate the the the right signals
• 1:42
  - from the noise. And and oftentimes,
• 1:46
  - the wearable devices and the tech that's out
• 1:48
  - there,
• 1:49
  - they can collect all sorts of data. And
• 1:51
  - and the key question for for a system
• 1:54
  - like ours means then, okay. When that data
• 1:57
  - comes in, what do you do with it?
• 1:59
  - Who's who's actually gonna look at it?
• 2:02
  - And and if the right signal doesn't get
• 2:05
  - separated out, extracted out,
• 2:08
  - then how are we gonna build that trust
• 2:10
  - within
• 2:11
  - who is gonna look at it on the
• 2:12
  - clinical side to really take actions based on
• 2:15
  - that?
• 2:16
  - So that went back... I mean, we kind
• 2:18
  - of went back to our... Some of our
• 2:20
  - highly successful remote patient monitoring programs and looked
• 2:23
  - at how do we how do we make
• 2:25
  - them successful.
• 2:27
  - Like, for instance,
• 2:28
  - we deal with mostly complex and serious conditions
• 2:31
  - at Mayo Clinic.
• 2:33
  - And,
• 2:34
  - let's say somebody just went through a,
• 2:36
  - heart surgery or or some sort of a
• 2:39
  - transplant...
• 2:40
  - Organ transplant journey.
• 2:42
  - When they go back home,
• 2:44
  - we make sure... First of all, we we
• 2:46
  - make sure
• 2:47
  - the remote patient monitoring program for that is
• 2:50
  - really fine tuned to that
• 2:52
  - specific set of patient groups, and it's it's
• 2:54
  - a fairly small group. I mean, roughly, you
• 2:57
  - can say
• 2:58
  - most remote patient monitoring programs that are successful
• 3:01
  - usually targets
• 3:03
  - x percent of of of the patients. Right?
• 3:05
  - I mean, it could be 10, could be
• 3:06
  - 20 depending on the system and the program.
• 3:09
  - But that population set is important because then
• 3:12
  - you're only tracking what's important for those patients,
• 3:16
  - and you're able
• 3:18
  - to easily,
• 3:20
  - separate that signal from the noise. So then
• 3:23
  - you then, factor in your your thresholds for
• 3:26
  - the others.
• 3:26
  - So typically, again, in in in a in
• 3:29
  - a primary care, preventative care space, there are
• 3:31
  - lots of devices out there that measure the
• 3:33
  - same thing but slightly differently, like blood pressure
• 3:36
  - monitoring
• 3:36
  - or your steps or your food and calorie
• 3:39
  - intake.
• 3:40
  - It becomes harder because the baseline is set
• 3:43
  - based on certain characteristics that you have, age,
• 3:46
  - all of that.
• 3:47
  - Whereas for
• 3:49
  - successful remote patient monitoring programs that are specialty
• 3:52
  - based,
• 3:53
  - you gotta fine tune that buffer. You gotta
• 3:55
  - fine tune that
• 3:56
  - baseline for each and every patient.
• 3:59
  - And then when the signal comes through,
• 4:01
  - it becomes really actionable for the for the
• 4:03
  - clinical
• 4:04
  - staff.
• 4:05
  - Then on the workflow side, we had to
• 4:07
  - make changes to then say, okay. If these
• 4:09
  - signals come through, who's gonna actually take action
• 4:12
  - on that? And then how do they then
• 4:14
  - trigger
• 4:15
  - subsequent care plan for for those patients? So
• 4:18
  - so I think if if you think about
• 4:19
  - it as
• 4:21
  - what is the patient
• 4:22
  - group that you really wanna monitor,
• 4:25
  - what are you actually monitoring,
• 4:27
  - how can you specify
• 4:29
  - the the the base levels of what you're
• 4:31
  - monitoring and what the trigger should be,
• 4:34
  - And then
• 4:35
  - on the practice side of it, how do
• 4:37
  - you get the accountable person
• 4:39
  - that's gonna be actually looking at it and
• 4:41
  - taking some action on it, whether that's a
• 4:43
  - care coordinator, whether that's a nurse, who it
• 4:45
  - might be depending on on the program.
• 4:48
  - And then how do you build that trust
• 4:50
  - in that system, especially now with AI? How
• 4:53
  - do you build that trust?
• 4:54
  - So
• 4:55
  - when those alerts come through, when those signals
• 4:58
  - actually become insights
• 4:59
  - that has to be acted on,
• 5:01
  - the clinicians
• 5:03
  - truly trusted as opposed to ignoring that as,
• 5:05
  - oh, that's set another alarm that went off.
• 5:08
  - That's a false alarm. Right? So that's important
• 5:10
  - from from how you make that,
• 5:12
  - make those remote program successful.
• 5:15
  - Absolutely. I think that makes a lot of
• 5:16
  - sense. And having that trust from the clinicians
• 5:19
  - is no easy task. As you mentioned, they're
• 5:21
  - being,
• 5:22
  - notified constantly
• 5:24
  - about different things that need to happen, and
• 5:25
  - they have so many responsibilities on a daily
• 5:27
  - basis with their patients. So that's really incredible
• 5:30
  - to hear how you're thinking about that and
• 5:32
  - doing that at Mayo Clinic.
• 5:34
  - Now I know when clinicians are trying to
• 5:36
  - reach a diagnosis, they're often then piecing together
• 5:38
  - information from radiology, cardiology,
• 5:41
  - urology, pathology, so many other systems where patients
• 5:44
  - have done tests and more.
• 5:46
  - Where does that fragmentation create more friction in
• 5:49
  - your organization, and,
• 5:51
  - how do you bring together the data sources
• 5:53
  - then to address the fragmentation or close those
• 5:56
  - gaps?
• 5:57
  - So,
• 5:58
  - from a Mayo Clinic perspective, I mean, this
• 6:00
  - is a this is a fairly unique situation
• 6:03
  - for us. So we are
• 6:05
  - known from the get go about a hundred
• 6:07
  - and sixty plus years ago as a truly
• 6:10
  - integrated multidisciplinary
• 6:11
  - practice.
• 6:13
  - So it's it's in our DNA
• 6:15
  - when when a patient shows up with a
• 6:17
  - with a complex set of conditions that require
• 6:21
  - multiple people,
• 6:22
  - multiple specialties, subspecialties
• 6:24
  - to work together,
• 6:26
  - we kind of already have an operating model
• 6:28
  - to do that. Now I understand for many
• 6:30
  - other systems, it may not be easy. And
• 6:33
  - even for us,
• 6:34
  - building a a system around it,
• 6:37
  - the technology and the support for it
• 6:39
  - took us multiple phases and iterations to get
• 6:42
  - it get it to a point. I mean,
• 6:44
  - even now it's not perfect, but we'll we'll
• 6:46
  - take it and we'll continue to iterate on
• 6:47
  - it. Right? So so now back to the
• 6:49
  - question of
• 6:51
  - how do you do that? Who who this
• 6:52
  - Who, how do you how do you avoid
• 6:54
  - that sort of a fragmentation?
• 6:56
  - So for us, the most important concept is
• 6:59
  - we wanna look at the the needs of
• 7:01
  - the patient always comes first. And what we
• 7:04
  - mean by that more so from a care
• 7:05
  - perspective is
• 7:07
  - we look at the patient as a whole.
• 7:09
  - Mhmm. Yeah. You may have come to me
• 7:11
  - as the cardiologist or gone to somewhere else
• 7:14
  - or something else,
• 7:15
  - but I wanna look at the whole of
• 7:17
  - you and see what are all the things
• 7:19
  - that we need to do to make sure
• 7:20
  - you return back to health or as you
• 7:22
  - leave the clinic,
• 7:24
  - you're you're a healthier version of what you
• 7:26
  - came into the to the clinic with flow
• 7:28
  - for the original problem.
• 7:29
  - So in that sense, the approach we typically
• 7:32
  - take is
• 7:33
  - on the initial
• 7:34
  - triage and diagnosis, we we often say
• 7:38
  - you'd have to start somewhere. So
• 7:40
  - we'll identify
• 7:42
  - what we think at that point reasonably is
• 7:44
  - the major problem,
• 7:45
  - and then that specialty
• 7:48
  - takes ownership
• 7:49
  - for that patient and then coordinates the rest
• 7:52
  - of the the actions from that point on.
• 7:54
  - So,
• 7:55
  - there are there are care coordinators and others
• 7:57
  - that are assigned where
• 7:59
  - as the various tests are done, as things
• 8:01
  - are done, you kinda bring it together,
• 8:04
  - in a way
• 8:05
  - that still makes sense for the individual specialties.
• 8:08
  - Now with with AI,
• 8:10
  - this is becoming a little bit easier because,
• 8:14
  - you can now
• 8:15
  - build some reviews or or things that are
• 8:19
  - distilling what,
• 8:21
  - what's relevant for a cardiologist versus nephrologist versus
• 8:24
  - neurology, whatever might be your specialty.
• 8:26
  - And and it's easier to do that now
• 8:28
  - than in the
• 8:30
  - prior,
• 8:30
  - iterations of the tech where
• 8:32
  - we had to literally build that wheel for
• 8:35
  - each specialty kind of. So technology gives us
• 8:38
  - a little bit of a help there. But,
• 8:40
  - ultimately,
• 8:41
  - it's...
• 8:42
  - If if you put the patient at the
• 8:43
  - center and if you think about everything the
• 8:45
  - patient goes through,
• 8:47
  - and even if you don't have the answer
• 8:49
  - as the specialist,
• 8:50
  - as long as you know that you can
• 8:52
  - reach out to your colleague down the hall
• 8:54
  - or wherever
• 8:56
  - to get that,
• 8:57
  - then we believe you you you do better
• 8:59
  - as a multidisciplinary
• 9:00
  - system. So that's kinda how the approach we
• 9:02
  - have taken to to,
• 9:04
  - get that. Doesn't mean it's perfect. Again, I
• 9:07
  - think, as you know, many of these specialties
• 9:10
  - have their own
• 9:11
  - viewers, so to speak. Right? They... They'll create
• 9:13
  - a a a a very specialized viewer for
• 9:17
  - radiology images.
• 9:19
  - They may create something else for pathology, and
• 9:22
  - then you... As a technology person, you gotta
• 9:24
  - then figure out how do I bring it
• 9:27
  - into all together. I mean, even even today
• 9:29
  - when I go into any of my specialty,
• 9:31
  - I'll see multiple tabs open. They'll say, oh,
• 9:34
  - this is where I go for this. This
• 9:35
  - is where I go for that. So that
• 9:37
  - part of it hopefully, again, in the AI
• 9:39
  - era gets a little bit even more, homogenized
• 9:42
  - to an extent while still retaining the the
• 9:44
  - personalization for that specialty.
• 9:46
  - But right now, yes, it is multiple tabs
• 9:49
  - and
• 9:50
  - some sort of a typical user interface where
• 9:52
  - they do have to tap through.
• 9:54
  - But when you go to each tab or
• 9:56
  - for their role, they still have summarized views
• 9:59
  - that are very specific to what's relevant for
• 10:02
  - them, what action they can take from that,
• 10:04
  - and how do they proceed from there. Yeah.
• 10:06
  - Absolutely. So it seems like you are with
• 10:08
  - the technology in many ways stitching together a
• 10:10
  - lot of the different,
• 10:12
  - information so you can pass it more easily
• 10:15
  - between folks, but I I know no easy
• 10:17
  - task. And so putting that all together, is
• 10:19
  - so helpful. Is that something you've done primarily,
• 10:22
  - in how's that Mayo or or been through
• 10:25
  - partnerships or a little bit of both? A
• 10:26
  - little bit of both. I think, some of
• 10:28
  - our tech is definitely homegrown.
• 10:30
  - We we often,
• 10:32
  - joke about our our clinical practitioners
• 10:35
  - having a a plus one always that they
• 10:37
  - all dabble in technology and and innovation.
• 10:40
  - So some of the tech is definitely homegrown
• 10:43
  - with the... Coming from the practice and and
• 10:45
  - the technologies working together. Some of it has
• 10:47
  - been partners,
• 10:49
  - like many that you see here today on
• 10:51
  - the on the floor... The conference.
• 10:53
  - Absolutely. Fantastic. And, you know, I know beyond
• 10:56
  - just what's happening within the system, patients are
• 10:58
  - increasingly moving among hospital, outpatient settings, getting care
• 11:02
  - at home.
• 11:03
  - So where does the clinical context most often
• 11:05
  - get lost during those transitions, and what would
• 11:07
  - it take to give care teams a more
• 11:09
  - continuous view of the patient?
• 11:12
  - Yeah. I mean, if you if you think
• 11:13
  - about that question,
• 11:15
  - in today's world, in the AI era, even
• 11:17
  - the word context takes a very different meaning
• 11:19
  - and and and importance. Right? I think, for
• 11:22
  - us,
• 11:25
  - if you if you take a a distinct
• 11:26
  - patient journey,
• 11:28
  - there's three parts where
• 11:30
  - we ought to be careful in terms of
• 11:32
  - the context getting passed in right. So as
• 11:34
  - I said, we are
• 11:36
  - we are into,
• 11:38
  - a multidisciplinary
• 11:39
  - practice for the complex and serious mostly.
• 11:42
  - So a lot of our patients come through
• 11:43
  - the referral route. So
• 11:46
  - your initial...
• 11:48
  - As the patient gets referred in Mhmm. What
• 11:51
  - is that data... What is that context getting
• 11:53
  - passed from the primary care to,
• 11:56
  - our specialties?
• 11:57
  - Mhmm. What are all the outside medical records
• 12:00
  - that you may have had
• 12:02
  - across multiple health system? I mean, one one
• 12:04
  - challenge that in health care we we all
• 12:06
  - still own it is we have talked about
• 12:08
  - data interoperably for
• 12:10
  - a decade now at least, but we still
• 12:12
  - haven't truly achieved it to the scale of
• 12:15
  - most other industries do.
• 12:17
  - So so that challenge still exists in terms
• 12:19
  - of how do you pull all these outside
• 12:21
  - records
• 12:22
  - into your system and make sense of it
• 12:25
  - the way that you wanna make sense of
• 12:26
  - it. Right? So so the first point of
• 12:28
  - drop off is that. I think, we have
• 12:30
  - our own,
• 12:32
  - AI driven tools that can do some of
• 12:34
  - that,
• 12:35
  - full, really, multi formatted outside records
• 12:39
  - and and start to create a a previsit,
• 12:42
  - diagnostic summary for the clinicians to look at.
• 12:46
  - The second piece of handoff,
• 12:48
  - really happens
• 12:49
  - when... Okay. So you're in the hospital. We
• 12:51
  - take care of you. We're about to discharge
• 12:53
  - you.
• 12:55
  - Again, because we are a specialty practice, we
• 12:57
  - don't really assume everybody is coming back to
• 12:59
  - us for primary care. They'll continue there. And
• 13:02
  - we are a destination site practice.
• 13:04
  - We we assume they're gonna continue their primary
• 13:06
  - care where where they came from. Right? Mhmm.
• 13:08
  - So there is a handoff back to to
• 13:10
  - primary care on the discharge. And then even
• 13:12
  - from a patient perspective,
• 13:15
  - let's just say the discharge summaries and the
• 13:17
  - discharge guides and all that becomes really
• 13:20
  - like a history book, to to read through
• 13:23
  - for some people. And I think how do
• 13:24
  - we as...
• 13:27
  - How do we,
• 13:28
  - make it easier for patient to understand
• 13:32
  - what was they treated for, what are the
• 13:34
  - things that they need to take care of
• 13:35
  - as they go back into their life at
• 13:37
  - home, and then how do we provide the
• 13:39
  - right level of context and information back to
• 13:41
  - the the primary care that they have so
• 13:45
  - they can continue the care,
• 13:47
  - before any other intervention is needed. Right? So
• 13:49
  - both of those places, we believe
• 13:52
  - there's a lot of context lost today that
• 13:54
  - that can be
• 13:55
  - addressed, and and it becomes
• 13:58
  - increasingly important because
• 14:00
  - without that context,
• 14:01
  - what you get out of any of the
• 14:03
  - AI models today
• 14:05
  - will be very different. And I think that's
• 14:06
  - why getting that right context is even more
• 14:09
  - important in the AI era than in the
• 14:11
  - previous year of technology.
• 14:13
  - That's such a great point. And especially when
• 14:15
  - you think about how AI technology works and
• 14:17
  - what really you need in order to continue
• 14:19
  - to ensure its effectiveness over time,
• 14:22
  - is really,
• 14:23
  - really helpful. I I know there's so many
• 14:26
  - AI use cases out there right now competing
• 14:28
  - for attention. So how do you distinguish the
• 14:30
  - most compelling ones to go from pilot,
• 14:33
  - to something that's ready to be scaled as
• 14:35
  - part of everyday care, and what must be
• 14:36
  - in place in order to scale that responsibly?
• 14:40
  - I think I'm gonna start with an admission
• 14:41
  - here.
• 14:42
  - I mean, just like any other health care
• 14:45
  - organization out there,
• 14:47
  - we went through our phase of pilothosis, if
• 14:50
  - you call that.
• 14:51
  - We had thousands of pilots running all over
• 14:53
  - the
• 14:54
  - all over the organization.
• 14:56
  - Again,
• 14:57
  - Mayo has a very decentralized model of innovation,
• 15:00
  - so it's not like we're gonna be able
• 15:03
  - to stop people from thinking about a innovative
• 15:06
  - solution in in their own area for a
• 15:08
  - problem that they're facing.
• 15:10
  - So what we quickly realized was though that
• 15:13
  - that funnel has to be quickly
• 15:15
  - shaped
• 15:16
  - efficiently.
• 15:17
  - So
• 15:18
  - most of those solutions
• 15:20
  - pass through a filter of,
• 15:23
  - does this work beyond a pilot?
• 15:25
  - Can this actually scale to the enterprise?
• 15:28
  - And is it safe for everyone to then
• 15:31
  - use it and then adopt it across the
• 15:33
  - multispecialty
• 15:33
  - practice?
• 15:35
  - So, and there is a regulatory and governance
• 15:37
  - aspect of it, which we won't go into,
• 15:40
  - but we had to quickly figure that part
• 15:42
  - out. So we actually had... At the at
• 15:44
  - the highest levels of leadership, we had a
• 15:46
  - a leader assigned to do just that Mhmm.
• 15:49
  - To say, how do I take
• 15:51
  - AI solutions
• 15:53
  - that have already kind of been developed
• 15:56
  - through the implementation rigor for getting it incorporated
• 15:59
  - into our clinical practice
• 16:01
  - while taking care of any sort of regulatory
• 16:03
  - burdens and stuff that we need to do,
• 16:05
  - especially with patient pacing.
• 16:07
  - So, so in the last year or two,
• 16:09
  - we have kind of done that as a...
• 16:11
  - We we have centralized the the governance and
• 16:13
  - and that part of it while the innovation
• 16:16
  - still remains decentralized. Yeah. So you just go
• 16:18
  - through this. When you're ready, you go through
• 16:20
  - this one process
• 16:21
  - and and you kind of then get ready.
• 16:23
  - So we have now had,
• 16:24
  - over 300 plus AI solutions in practice
• 16:28
  - that have already been vetted, passed through that,
• 16:30
  - and primed for scale across the enterprise,
• 16:34
  - while we still have a lot of other
• 16:35
  - decentralized,
• 16:36
  - innovations going on. So,
• 16:38
  - so that's the approach we have taken from
• 16:40
  - from that perspective.
• 16:41
  - But but, yes, I think it is... Especially
• 16:44
  - with the way things are
• 16:46
  - emerging now with with with the models and
• 16:48
  - and and and how intelligent they are getting
• 16:50
  - with every iteration,
• 16:52
  - you are gonna get lot more,
• 16:56
  - solutions looking for problems than,
• 17:00
  - solutions that specifically address a problem that can
• 17:02
  - then be scaled Mhmm. Across an enterprise and
• 17:05
  - then across a large body of patients.
• 17:07
  - Absolutely. I think that's such a great point.
• 17:09
  - And, you know, even having the
• 17:11
  - innovation, the AI, going from idea to, hey.
• 17:14
  - It works. But then, you know,
• 17:16
  - becoming more of an enterprise wide solution,
• 17:19
  - takes a good bit more resources,
• 17:21
  - time,
• 17:22
  - as well as effort to get it there.
• 17:24
  - And so identifying those right things, and being
• 17:27
  - able to do it smartly,
• 17:29
  - is critical. And to your point, you know,
• 17:30
  - having more of the decentralized innovation, I can
• 17:33
  - imagine a lot of benefits, but does that
• 17:35
  - make you nervous or give you anxiety as
• 17:37
  - you were talking about? I was thinking, you
• 17:38
  - know, that, it'd be a lot to kind
• 17:40
  - of handle, but at the same time, it
• 17:42
  - seems like Mayo has done it beautifully for
• 17:43
  - years.
• 17:45
  - Yeah. Yeah. Yeah. I think I think for
• 17:46
  - the leadership,
• 17:48
  - sometimes,
• 17:50
  - the the good good part of it is,
• 17:52
  - you know, you're not you're not slowing down
• 17:55
  - the innovation.
• 17:56
  - Mhmm. And and and because of the multispecialty
• 17:58
  - nature of our practice, we also
• 18:01
  - believe true innovation
• 18:03
  - often happens very close to that specialty or
• 18:06
  - subspecialty,
• 18:07
  - not at an enterprise level. So so we
• 18:09
  - wanna preserve that. I think,
• 18:12
  - to your point, yes,
• 18:13
  - as as the as the chief digital officer
• 18:16
  - managing all of the technology underneath and the
• 18:18
  - innovation,
• 18:20
  - sometimes
• 18:21
  - it is it is quite a bit of
• 18:23
  - the wild wild west there, and then you
• 18:24
  - need to know what's going on and how
• 18:26
  - things fit in. And I think
• 18:28
  - I'll say this, I think we have been
• 18:30
  - very conscious in taking a a platform approach
• 18:34
  - to this problem from the get go. I
• 18:35
  - think six, seven years ago, we introduced our
• 18:37
  - our Mayo Clinic platform.
• 18:39
  - So everything that we do internally is like,
• 18:43
  - okay, how do you how do you... How
• 18:44
  - does this solution plug into the platform?
• 18:47
  - We we give you the chassis. We give
• 18:49
  - you the
• 18:50
  - the
• 18:52
  - the the rule set in terms of how
• 18:53
  - you can plug in.
• 18:55
  - So so we kind of force people to
• 18:57
  - think about,
• 18:58
  - if I really wanna take this solution enterprise
• 19:00
  - Mhmm. Then I gotta plug this into the
• 19:02
  - platform. So it's not a... Yet another add
• 19:04
  - on
• 19:06
  - piecemeal
• 19:07
  - point solution. It's actually part of the the
• 19:09
  - platform as well. So that has helped us
• 19:12
  - corral a little bit of that,
• 19:14
  - tendency to do that. Yeah. Absolutely. That's helpful.
• 19:16
  - And I think to your point, I mean,
• 19:18
  - the technology
• 19:19
  - relationships even with outside companies are are moving
• 19:22
  - beyond some of these one time purchases and
• 19:24
  - and,
• 19:25
  - getting into more of a partnership,
• 19:27
  - situation. So when you are developing a relationship
• 19:30
  - with a a new,
• 19:31
  - company, how do you make sure they remain
• 19:33
  - accountable for these technologies as they go live
• 19:35
  - and that relationship is genuinely valuable over time?
• 19:39
  - It's a great question. And I think I
• 19:40
  - think, as much as we, pride ourselves in
• 19:43
  - in being experts in in in the clinical
• 19:46
  - research and and and the spaces that we
• 19:49
  - know best, we also know that
• 19:52
  - we are we are increasingly
• 19:55
  - have to become a technology company, at least
• 19:57
  - an health care technology company, but we are
• 19:59
  - not the technology company themselves. So so the
• 20:02
  - type of partnership we are actually
• 20:04
  - looking for and and continue to do is
• 20:06
  - that strong blend
• 20:08
  - of the partner bringing the the technology chops,
• 20:11
  - we bringing our our,
• 20:13
  - depth and strength in clinical practice and research
• 20:16
  - and data, all of that, and then creating
• 20:19
  - something that makes a difference for the patients
• 20:22
  - that we see and care for and and
• 20:24
  - perhaps even for the larger
• 20:25
  - public. So in that sense, I think, whether
• 20:28
  - that's,
• 20:30
  - device manufacturers,
• 20:31
  - whether that's,
• 20:32
  - platform vendors, whether that's other sorts of technologies,
• 20:36
  - we typically,
• 20:39
  - look for that technology partner coming in with
• 20:41
  - their viewpoint
• 20:42
  - that pushes our thinking. And then we we
• 20:45
  - kinda push the same from our side to
• 20:47
  - say,
• 20:48
  - really, for that technology to be proven, it
• 20:50
  - needs to work in our environment. It needs
• 20:52
  - to be
• 20:53
  - providing trustworthy,
• 20:54
  - safe
• 20:55
  - data for our
• 20:57
  - physicians to to really use at the point
• 20:59
  - of care and and kind of build build
• 21:01
  - that model over time. So it's been a
• 21:03
  - it's been a great strategy for us. It's
• 21:04
  - worked out really well so far. And and,
• 21:07
  - again, increasingly, we are all becoming more and
• 21:10
  - more tech companies even in health care, but
• 21:12
  - you still gotta understand where your strengths are
• 21:15
  - and where the partner can bring theirs. That
• 21:17
  - makes a lot of sense. And, you know,
• 21:18
  - it's really a helpful framework to think through
• 21:20
  - as you're looking at, you know, I I
• 21:22
  - can imagine many potential partners on a daily
• 21:24
  - basis. So before we wrap here, I have
• 21:27
  - one more question.
• 21:28
  - As care increasingly moves beyond the hospital walls,
• 21:31
  - how are you using connected technologies and AI
• 21:33
  - to identify patients
• 21:35
  - who need earlier follow-up support coordinated care teams
• 21:37
  - and close the gaps in post discharge care?
• 21:41
  - I think that's a that's a... Kinda put
• 21:42
  - a full circle to this because we started
• 21:44
  - with remote patient monitoring, and we're ending with
• 21:47
  - care that that happens beyond the walls of
• 21:50
  - a hospital, so to speak. Right? And I
• 21:52
  - think this is one area where
• 21:55
  - we believe,
• 21:56
  - as health care organizations,
• 21:58
  - we may not be able to solve the
• 22:00
  - entire part of it. Like, for instance, in
• 22:02
  - one of our programs
• 22:04
  - that we have tried, it's still very successful
• 22:06
  - at a at a fairly pilot level.
• 22:09
  - The challenges
• 22:10
  - we we face are
• 22:13
  - typically,
• 22:14
  - for any sort of serious care to happen
• 22:16
  - at home or even,
• 22:18
  - let's say,
• 22:19
  - chemo at home kind of a thing,
• 22:22
  - it's getting the
• 22:24
  - the right
• 22:25
  - materials, so to speak,
• 22:27
  - dispatched and and available
• 22:29
  - at the patient's house
• 22:31
  - or home or whatever place of care
• 22:33
  - and ensuring
• 22:35
  - there is a coordination
• 22:36
  - between the the logistics of getting the right
• 22:39
  - stuff there, getting the right person, whether that's
• 22:42
  - a nurse or somebody
• 22:43
  - there, and and doing it in a in
• 22:45
  - a way that the patient feels like they
• 22:48
  - don't have to come to the hospital every
• 22:50
  - time for for chemo or something like that.
• 22:52
  - So
• 22:53
  - we have proven that it can it can
• 22:54
  - work at a smaller scale, but for it
• 22:56
  - to be,
• 22:57
  - the the economies to to work on this
• 22:59
  - work kind of model, we need partners. We
• 23:02
  - need partners who have the logistical chops, who
• 23:04
  - have the fulfillment chops, who can
• 23:06
  - get even rural care sort of a thing
• 23:09
  - work for this. So that's kinda where
• 23:12
  - we are right now. Again, there's... The the
• 23:14
  - hopes of the future is that we will
• 23:15
  - find those right partners that can help scale
• 23:17
  - it.
• 23:18
  - But, certainly,
• 23:20
  - as an ongoing
• 23:22
  - thinking, we do believe,
• 23:24
  - as as technology,
• 23:26
  - evolves here, there'll be more care done away
• 23:30
  - from
• 23:31
  - a hospital setting in the future
• 23:33
  - and... Which means
• 23:35
  - we as as one of the the the
• 23:37
  - destination,
• 23:37
  - medical
• 23:38
  - centers for, serious and complex care, we have
• 23:41
  - to think about
• 23:43
  - offering such care in a remote setting as
• 23:45
  - well. So we do have a a program
• 23:47
  - that's kind of specifically addressed to address it,
• 23:49
  - but
• 23:50
  - this is one area where I believe the
• 23:52
  - industry needs other partners
• 23:54
  - to to work with us. The health care
• 23:55
  - organizations alone may not be able to scale
• 23:58
  - it to the economies that we need to.
• 24:00
  - That's fascinating to hear, you know, a lot
• 24:02
  - of possibility, potential, and opportunity for the future.
• 24:05
  - Thank you so much, Nari, for your time
• 24:07
  - today and your insights. And we also wanna
• 24:09
  - thank our podcast sponsor, Philips. You can tune
• 24:11
  - in to more podcasts from Becker's Healthcare by
• 24:13
  - visiting our podcast page at beckershospitalreview.com.
• 24:17
  - Thank you.