Latest / Becker’s Healthcare Digital Health + Health IT / Using AI to Turn Patient Data Into Actionable Insights and Proactive Care with Dr. Russell Horton

Transcript

Using AI to Turn Patient Data Into Actionable Insights and Proactive Care with Dr. Russell Horton · 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,
• 0:25
  - it can benefit health systems, and most importantly,
• 0:28
  - it can benefit patients and their families.
• 0:33
  - Phillips.
• 0:34
  - Hi, everyone. This is Gracelyn Keller with Becker's
• 0:36
  - Healthcare, and we are recording live at the
• 0:38
  - eleventh annual IT and revenue cycle conference, the
• 0:41
  - future of AI and digital health. Thank you
• 0:43
  - so much for tuning in to the Becker's
• 0:45
  - Healthcare podcast series. Joining me for today's discussion
• 0:48
  - is doctor Russell Horton, who is the division
• 0:50
  - medical director at BMG Primary Care. So doctor
• 0:53
  - Horton, thanks for being here today, and let's
• 0:55
  - jump right into our conversation. Yeah. Thanks for
• 0:57
  - having me. So health systems are receiving a
• 1:00
  - growing stream of patient data from bedside monitoring,
• 1:03
  - ambulatory devices, and care delivered at home. Where
• 1:06
  - is that data creating actionable insights actionable insights
• 1:08
  - for clinicians, and where does it still create
• 1:10
  - more work than insight?
• 1:12
  - Yeah. And that's a a great point about
• 1:14
  - all this data. It's really good to have,
• 1:16
  - but I think the important thing that we
• 1:18
  - need to remember is that
• 1:19
  - the data needs to have an action action
• 1:21
  - item. Like any good meeting, it needs an
• 1:23
  - action item to take away from it.
• 1:26
  - You know, where do they need to follow-up?
• 1:27
  - What do they need to do? Who do
• 1:29
  - they need to see? What refills do they
• 1:30
  - need? Where it's creating more work, I think,
• 1:33
  - is when the data just comes in in
• 1:34
  - a big lump.
• 1:35
  - And so we're seeing, you know, a lot
• 1:37
  - of things that we still have to parse
• 1:39
  - through, we have to make decisions about as
• 1:41
  - opposed to, here's all the action items. You
• 1:43
  - can jump on those. When you need more
• 1:45
  - data, here it is.
• 1:47
  - And so we're still yet to get there,
• 1:49
  - but, hopefully, that's where we're heading to.
• 1:51
  - And when clinicians are trying to reach a
• 1:53
  - diagnosis, they often must piece together information from
• 1:57
  - radiology, cardiology, pathology, all these different systems.
• 2:01
  - So where does that fragmentation
• 2:03
  - create the most friction in your organization, and
• 2:05
  - how do you bring together data sources and
• 2:07
  - address the fragmentation or close the gaps?
• 2:10
  - Yeah. I I think there's a couple issues
• 2:12
  - there. One is kinda like we just said,
• 2:14
  - you know, we get a lot of data
• 2:15
  - all at once, and there's still a lot
• 2:17
  - of parsing that needs to be done. But
• 2:19
  - I think it's also context. Right? So cardiology
• 2:23
  - and pathology
• 2:24
  - and radiology may have all been looking at
• 2:26
  - the same problem,
• 2:28
  - but weren't talking to each other. And so
• 2:30
  - our data systems really need to be able
• 2:31
  - to put that into context of, hey. This
• 2:33
  - is all about the same thing, the same
• 2:35
  - patient, and here's what we're doing about it.
• 2:38
  - Unfortunately, a lot of the times we get,
• 2:40
  - you know, something from urology, but then something
• 2:42
  - from pathology,
• 2:43
  - and it's not clear if it's the same
• 2:45
  - time or when or where or who. So
• 2:47
  - that's really, I think, where the next step
• 2:49
  - has to be is that AI and data
• 2:51
  - needs to give context to things so that
• 2:53
  - we can look at it quickly, make decisions,
• 2:55
  - and move forward ultimately so the patient gets
• 2:57
  - the best care.
• 2:59
  - And patients are increasingly moving among the hospital,
• 3:02
  - outpatient settings, and the home. Where does clinical
• 3:05
  - contacts most often get lost during these transactions
• 3:08
  - and what would it take to give care
• 3:09
  - teams a more continuous view of the patient?
• 3:13
  - Well, I think first of all, when you're
• 3:15
  - coming from the hospital to outpatient to other
• 3:17
  - services, like, the first thing we look at
• 3:19
  - at Banner, of course, is is all this
• 3:21
  - information getting where it needs to go safely.
• 3:23
  - Right? We need a lot of checks and
• 3:25
  - balances right now with AI. We probably always
• 3:28
  - will, but,
• 3:29
  - a lot of kind of important stuff can
• 3:31
  - get lost in that mix. And so what
• 3:32
  - we want is for,
• 3:34
  - a lot of safety first. And then,
• 3:38
  - the context
• 3:39
  - usually tends to get lost when
• 3:42
  - the different data sources, so different providers, different
• 3:45
  - nursing units
• 3:46
  - are not aligning with what the real purpose
• 3:49
  - of that data is.
• 3:50
  - And so,
• 3:52
  - I think what it's gonna take to get
• 3:53
  - a continuous view of the patient is that
• 3:56
  - probably AI or some very smart people are
• 3:58
  - gonna have to have that immediately go into
• 4:01
  - one clinical picture.
• 4:03
  - Because while, yes, there's a lot of specialties
• 4:05
  - and there's a lot of home health services
• 4:07
  - and different things going on, We really need
• 4:09
  - that to come to us in the form
• 4:10
  - of a patient and a person that we're
• 4:12
  - caring for and not just, you know, a
• 4:15
  - lot of really detailed information.
• 4:17
  - And with so many AI use cases competing
• 4:20
  - for attention, how do you distinguish a compelling
• 4:22
  - pilot from something ready to be become part
• 4:25
  - of everyday care, and what must be in
• 4:27
  - place to scale that properly?
• 4:30
  - Yeah. You know, even being at a at
• 4:33
  - a great conference like this, there's just so
• 4:35
  - many solutions. Right? And so for us in
• 4:37
  - primary care or even as a health system
• 4:39
  - at large,
• 4:40
  - what we,
• 4:42
  - need is someone who has experience in the
• 4:45
  - EHRs,
• 4:46
  - someone who knows,
• 4:48
  - sort of
• 4:49
  - front facing, patient facing information.
• 4:52
  - And then we need to be able to
• 4:53
  - take that
• 4:54
  - and really pilot it to a specific use
• 4:57
  - case. So, you know, if a solution has
• 4:59
  - a lot of great things around
• 5:01
  - ER workflows, that's great. But in primary care,
• 5:04
  - that doesn't really help us a lot. So
• 5:06
  - the the solutions that are coming to us
• 5:08
  - sort of vetted a little bit, have had
• 5:10
  - had some experience, really help with that. And
• 5:12
  - then, you know, to scale it responsibly, again,
• 5:15
  - a lot of safety checks and balances. We
• 5:16
  - need clinic... Clinicians,
• 5:19
  - there from the forefront.
• 5:20
  - But then,
• 5:22
  - we also need it to be something that
• 5:24
  - is, again, not so narrow a use case
• 5:26
  - that we can't expand it.
• 5:28
  - It's a tough line to to walk, and
• 5:30
  - I don't envy the vendors for this. But,
• 5:33
  - if we can really have something that has
• 5:36
  - some of that experience
• 5:37
  - but isn't so narrow that we can't use
• 5:39
  - it, then, you know, that's gonna be a
• 5:41
  - great system.
• 5:43
  - And as technology relationships move beyond onetime purchases,
• 5:46
  - what should a partner remain accountable for after
• 5:48
  - go lives to make the relationship genuinely valuable
• 5:51
  - over time?
• 5:53
  - Yeah. Accountability
• 5:54
  - is huge on both sides. I mean, us
• 5:56
  - as a health system, you know, we're a
• 5:57
  - nonprofit health system. We're trying to reinvest things
• 6:00
  - at Banner back into our people and our
• 6:03
  - patients,
• 6:04
  - and we really need a vendor or a
• 6:06
  - partner that's gonna do the same thing.
• 6:09
  - So,
• 6:09
  - it's the follow ups. It's the, hey. You
• 6:12
  - know, we've got an update or we've got
• 6:14
  - a change, and we can help you through
• 6:15
  - that.
• 6:16
  - It's about lowering the IT lift, frankly, a
• 6:19
  - lot of the time is,
• 6:21
  - kind of coming to us ready to go,
• 6:23
  - not saying, hey. You need to make a
• 6:25
  - bunch of changes so that we can do
• 6:26
  - this.
• 6:28
  - And if we can create that sort of
• 6:30
  - partnership, I think we can do that long
• 6:31
  - term value,
• 6:33
  - but, it really does have to go both
• 6:35
  - ways.
• 6:36
  - And finally, as care increasingly moves beyond hospital
• 6:39
  - walls, how are you using connected technologies and
• 6:41
  - AI to identify patients who need earlier follow-up,
• 6:44
  - support care coordination teams, and close gaps in
• 6:47
  - post discharge care? Yeah. And and this is
• 6:49
  - where we really have to
• 6:51
  - invest in our people to do
• 6:54
  - the most that they possibly can do, but
• 6:56
  - we also really need
• 6:58
  - to have AI and these different tech solutions
• 7:00
  - lift up the rest of it. Right? So,
• 7:04
  - I think what we're trying to do is
• 7:05
  - is prediction based analysis. What we're... You know,
• 7:09
  - we wanna be able to see here's who's
• 7:11
  - gonna need help, here's what's gonna come next,
• 7:13
  - and then be able to take those next
• 7:15
  - steps because, ultimately, we wanna care for the
• 7:18
  - patient, but we need that help. And so
• 7:20
  - that's where we
• 7:21
  - we see kind of that next step is,
• 7:24
  - if we're catching people if we're catching people
• 7:27
  - ahead of time. Right? Preventative care,
• 7:30
  - caring for the whole patient, value based care
• 7:32
  - is huge for us at Banner right now.
• 7:34
  - And so if we have something that's allowing
• 7:36
  - us to do that, to get ahead of
• 7:38
  - disease, not follow disease, that's really, I think,
• 7:40
  - the best of both worlds.
• 7:42
  - Wonderful. Well, doctor Horton, thanks so much for
• 7:44
  - joining me today and sharing these thoughts and
• 7:46
  - insights. And I also wanna thank our podcast
• 7:48
  - sponsor, Philips. You can tune in to more
• 7:50
  - podcasts from Becker's Healthcare by visiting our page
• 7:53
  - at beccarshospitalreview.com.