Latest / Becker’s Healthcare Podcast / Unlocking Value Inside and Outside the EHR With AI

Transcript

Unlocking Value Inside and Outside the EHR With AI · Becker’s Healthcare Podcast

• 0:00
  - Well, hello, everyone. This is Erica Spicer Mason
• 0:02
  - with Becker's Healthcare. Thank you so much for
• 0:05
  - tuning in to the Becker's Healthcare podcast series.
• 0:08
  - Today, we're going to walk through the value
• 0:10
  - that health systems are seeing both inside and
• 0:13
  - outside of their EHR environments,
• 0:15
  - including where AI tools are factoring in.
• 0:18
  - And joining me for this discussion,
• 0:20
  - we have Ed Ricks, the managing director for
• 0:22
  - health care IT practice at Baker Tilly,
• 0:25
  - and doctor Jeet Varmicha,
• 0:27
  - the chief information and digital strategy officer at
• 0:30
  - SBH Health System.
• 0:32
  - Ed, Jeet, it's great to have you both
• 0:34
  - with us today. Thank you so much for
• 0:35
  - joining. Thanks, Erica.
• 0:37
  - Well, it's so great to have you both
• 0:39
  - on the line. And before we get into
• 0:41
  - the the heart of the conversation,
• 0:43
  - I wanted to see if you could both
• 0:44
  - just tell me a little bit more about
• 0:45
  - your roles and the scope of the technology
• 0:48
  - work that you both oversee.
• 0:50
  - And go ahead and get us started with
• 0:52
  - this one.
• 0:53
  - Yeah. So at Big Cattell, I'm the managing
• 0:55
  - director of our health care IT practice, and
• 0:58
  - we help our clients every day with EHR
• 1:00
  - implementations,
• 1:01
  - optimization,
• 1:03
  - any kind of assessment work that they need
• 1:05
  - done.
• 1:06
  - But the best part is getting to talk
• 1:09
  - to different hospital systems every day, I think,
• 1:11
  - in this job. My career before joining Baker
• 1:13
  - Tilly for over twenty years, I was a
• 1:15
  - CIO in hospital systems
• 1:17
  - and then chief operating officer for about four
• 1:19
  - years. So I still remember all the things
• 1:21
  - they're going through. It's been a couple years
• 1:23
  - now, but it's a lot of fun that
• 1:25
  - I get to spend time talking to people
• 1:26
  - like Gee all the time, hearing what's going
• 1:29
  - well, hearing what's not going well, take that
• 1:31
  - to our next client and help them for
• 1:32
  - their planning purposes.
• 1:35
  - Fantastic. Great to have your expertise with us,
• 1:37
  - Ed. Thank you.
• 1:38
  - And, Gee, tell us about yourself.
• 1:41
  - Hi. Good afternoon, everyone.
• 1:43
  - Thank you, Erica, for having me.
• 1:45
  - I'm Jeet Barmecha, and I'm PGY 31. What
• 1:49
  - that means
• 1:50
  - is I'm an internal medicine practitioner,
• 1:52
  - almost thirty years
• 1:54
  - at my health system, which is SBHL system.
• 1:57
  - Apart from practicing
• 1:59
  - internal medicine at bedside,
• 2:01
  - I do have a full time role as
• 2:03
  - being a senior vice president and chief information
• 2:05
  - and digital strategy officer.
• 2:07
  - What that means is I have a wonderful
• 2:10
  - team,
• 2:11
  - who manages,
• 2:12
  - technology infrastructure,
• 2:14
  - technology applications,
• 2:16
  - telecommunications,
• 2:17
  - digital strategy, and, data and analytics.
• 2:21
  - We are very fortunate enough that,
• 2:24
  - about two and a half, three years ago,
• 2:26
  - we were able to implement,
• 2:28
  - Epic and Oracle together
• 2:30
  - as clinical EHR revenue cycle and ERP, with
• 2:33
  - financials.
• 2:35
  - Okay. Thanks for rounding us out with the
• 2:37
  - intros, and it's great to learn
• 2:39
  - more about you both and this technology lens
• 2:42
  - that you're bringing to our conversation today.
• 2:45
  - And, you know, I think our conversation really
• 2:47
  - is going to cover,
• 2:48
  - essentially the current state of of EHRs and
• 2:51
  - how they're lending themselves to innovation
• 2:53
  - and what you both see ahead.
• 2:55
  - So let's go ahead and start there.
• 2:57
  - Ajit, I know that the health care industry
• 2:59
  - has spent roughly two decades and an enormous
• 3:02
  - amount of money on EHR implementation.
• 3:05
  - So from where you sit at SBH Health
• 3:07
  - System, how do you assess what that investment
• 3:10
  - has returned so far? You know, whether we're
• 3:13
  - talking about clinical outcomes, operational, financial outcomes,
• 3:16
  - where has it clearly delivered, and where is
• 3:18
  - the value harder to point to?
• 3:21
  - So let... Erica, let me start with our
• 3:23
  - mission. I mean, our mission is fairly simple,
• 3:26
  - which is providing safer health care.
• 3:29
  - And
• 3:30
  - if
• 3:31
  - you understand the
• 3:33
  - journey of EHR in The US,
• 3:36
  - it used to be revenue cycle systems or
• 3:39
  - maybe some ancillary system.
• 3:41
  - The patient movements
• 3:43
  - in the country actually
• 3:45
  - give a very good springboard,
• 3:47
  - of digitizing
• 3:48
  - the clinical or operational or financial workflows.
• 3:52
  - And you're right. It's been only only two
• 3:54
  - decades.
• 3:55
  - And we have come so far,
• 3:58
  - either with the use of,
• 4:00
  - meaningful use
• 4:02
  - or
• 4:03
  - better interoperability
• 4:04
  - standards.
• 4:06
  - But the most important... What I realize is,
• 4:08
  - apart from digitization
• 4:09
  - of clinical
• 4:11
  - or operational or financial records is the patient
• 4:14
  - empowerment
• 4:15
  - or patient engagement.
• 4:17
  - Now we have the tools where patients
• 4:20
  - or the caregivers are empowered,
• 4:22
  - in their care decisions. So for me, that
• 4:25
  - is very
• 4:26
  - gratifying and very, informative
• 4:29
  - and very helpful
• 4:30
  - so that we all can participate in the
• 4:32
  - decision process.
• 4:34
  - Yes. There,
• 4:36
  - the speed of care
• 4:38
  - amongst the caregivers,
• 4:40
  - patients, and clinicians
• 4:42
  - have increased so rapidly because of, digitization.
• 4:47
  - We used to have paper records. I used
• 4:49
  - to be a resident on paper records, and
• 4:51
  - there was only one record on the
• 4:53
  - unit. And now it is almost anywhere in
• 4:56
  - the world.
• 4:57
  - You can have access to the patient records,
• 5:00
  - appropriately,
• 5:01
  - if you have access to.
• 5:03
  - But, nevertheless,
• 5:04
  - there has been some unintended
• 5:07
  - consequences,
• 5:08
  - especially on the clinical side
• 5:10
  - or clinician side, which is the cognitive overload,
• 5:14
  - because of better interoperability
• 5:16
  - standards,
• 5:18
  - what I call it, digital twins. It's not
• 5:21
  - only the
• 5:22
  - medical record or clinical record in front of
• 5:25
  - you, but you have the clinical record from
• 5:27
  - various
• 5:28
  - encounters,
• 5:29
  - whether through HIE exchange
• 5:31
  - or through patient portal,
• 5:33
  - so forth, and also post acute care.
• 5:36
  - So
• 5:38
  - it has become universal record for the continuity
• 5:41
  - or the longitudinal
• 5:42
  - care.
• 5:44
  - Yes.
• 5:45
  - In terms of outcomes,
• 5:48
  - definitely, there has been better clinical outcomes because
• 5:50
  - we are using more and more
• 5:53
  - evidence based, clinical guidelines at the point of
• 5:55
  - care. We have access to
• 5:57
  - real time data on diagnostics,
• 6:00
  - and,
• 6:01
  - as I mentioned regarding the longitudinal
• 6:03
  - care from other places.
• 6:06
  - In terms of financial outcomes, yes, definitely, we
• 6:09
  - have,
• 6:10
  - seen some significant improve,
• 6:13
  - improvement on the,
• 6:15
  - coding, billing, and dropping the claim side. So,
• 6:18
  - overall, it it has it has,
• 6:20
  - definitely
• 6:21
  - lot of positive outcomes.
• 6:23
  - And as I said, it is a journey.
• 6:25
  - We are not there yet. And,
• 6:28
  - the unintended consequences of clinicians becoming
• 6:32
  - a glorified typewriter in certain areas needs to
• 6:35
  - be tackled, how we can make it easier
• 6:38
  - so that we have more bedside time, more
• 6:41
  - time with
• 6:42
  - patients and caregivers, and less time on screens.
• 6:45
  - And that there are some tools available which
• 6:47
  - are emerging.
• 6:50
  - Again, we have lot to,
• 6:52
  - make these,
• 6:54
  - physician and patient encounters
• 6:56
  - much more meaningful.
• 6:58
  - Mhmm. Yeah. Ajit, thank you so much for
• 7:00
  - the overview. It sounds like to your point,
• 7:03
  - there's there's some way to go in terms
• 7:05
  - of improvements, you know, getting clinicians off of
• 7:08
  - screens more. That's definitely a through line and
• 7:11
  - a theme that we've heard among leadership conversations
• 7:14
  - at Becker's.
• 7:15
  - But then also, you know, you note on
• 7:16
  - some really important improvements here, patients being more
• 7:19
  - engaged and empowered in their care through the
• 7:21
  - EHR.
• 7:22
  - So, you know, a a mix of both
• 7:24
  - worlds, I guess you could say.
• 7:26
  - But overall,
• 7:27
  - driving improvements certainly.
• 7:30
  - Ed, when you look at the organizations that
• 7:32
  - you work with, how consistent is that picture
• 7:34
  - and and how are you seeing leaders measure
• 7:36
  - these clinical, operational, and financial outcomes?
• 7:40
  - You know, the,
• 7:41
  - as Jeet mentioned, the intended and unintended
• 7:44
  - circumstances or, you know, consequences of an implementation.
• 7:47
  - I... I'd say throughout my career as a
• 7:50
  - CIO,
• 7:51
  - I always felt bad because we felt like
• 7:53
  - we were doing the right thing for the
• 7:55
  - right reasons,
• 7:56
  - And and Jeep was clear on that. Like,
• 7:58
  - he got this foundational value with more accessible
• 8:00
  - longitudinal record, legible orders,
• 8:04
  - medication, safeguards, all the things, standardized documentation for
• 8:08
  - quality reporting, all those things that come from
• 8:10
  - that. But you've sort of impeded the clinician's
• 8:13
  - ability to deliver care, I think. And, you
• 8:15
  - know, when when I talk to places, probably
• 8:17
  - the best part of this job is I
• 8:19
  - get to talk to folks like g every
• 8:20
  - day,
• 8:21
  - who've done greater things than I ever did.
• 8:23
  - And you learn so much from all of
• 8:24
  - them, I think, and can sort of translate
• 8:26
  - on. And I would say that, like, all
• 8:28
  - organizations have done a pretty good job
• 8:31
  - of trying to measure
• 8:33
  - value somehow. But for some reason, that degree
• 8:35
  - of value realization is not equivalent across there.
• 8:38
  - And and I think that I sort of
• 8:41
  - mean that, like,
• 8:43
  - I wouldn't ask a health system to produce,
• 8:45
  - like, one enterprise wide
• 8:47
  - EHR ROI number. I I don't think that's
• 8:49
  - even possible. I'd say let's look at a
• 8:51
  - portfolio
• 8:52
  - of kind of clinical and operational and workforce
• 8:55
  - and financial outcomes that are tied to the
• 8:57
  - workflows we've improved. And that's how you can
• 9:00
  - measure everything. You have to do it that
• 9:01
  - way. So
• 9:03
  - financial is big. Adding capacity is big. Risk
• 9:06
  - avoided is big. You know, it's hard to
• 9:08
  - put numbers on some of those, but I
• 9:09
  - think measuring all those things makes it a
• 9:11
  - little bit more interchangeable
• 9:13
  - between organizations. And it's, it's tough. I see
• 9:17
  - a lot of, I will say,
• 9:19
  - waste in what people leave behind from their
• 9:23
  - EHR and don't don't get... I love what
• 9:25
  - Jeet said in his introduction. Like, he's... It's
• 9:27
  - a journey. A lot of people look at
• 9:29
  - it as the go live, and then they're
• 9:31
  - live and you go forward, and that's just,
• 9:33
  - like, step one. So, you know, I I
• 9:35
  - feel like the EHR created the capabilities,
• 9:38
  - workflow redesign, and I guess governance determine whether
• 9:40
  - that capability becomes value over time.
• 9:44
  - Mhmm. And if I can add point, Sam.
• 9:46
  - To... If I can add to Ed's point,
• 9:49
  - we have not even touched the post acute
• 9:51
  - care or the population health, social drivers of
• 9:54
  - health.
• 9:55
  - When I used to be a resident and
• 9:57
  - new attending, we never
• 9:59
  - documented in so much detail because we felt
• 10:02
  - only illness was the
• 10:04
  - care we were supposed to take, disease management
• 10:07
  - more specifically. But now with the help of
• 10:09
  - these tools, it goes way beyond
• 10:12
  - the illness.
• 10:14
  - We look into the environment and the factors
• 10:16
  - which contributes to those factors.
• 10:18
  - So we could not have done this without
• 10:20
  - the, emerging technologies or the enhancements in the
• 10:24
  - electronic health records.
• 10:26
  - Mhmm. Yeah. Great points from you both. And
• 10:30
  - and I wanna go back to something, Ed,
• 10:32
  - that you had touched on. I I think
• 10:34
  - you had mentioned something around
• 10:36
  - the fact that you see some organizations kind
• 10:38
  - of kind of fall victim to this concept
• 10:41
  - of waste within the EHR, you know, implementing
• 10:44
  - it and seeing it as a as a
• 10:46
  - one time implementation
• 10:47
  - and and things are set from there.
• 10:50
  - While much of the conversation jumps straight to
• 10:52
  - the next AI platform or point solution. So,
• 10:56
  - Ed, I'll have you get us started on
• 10:57
  - this one and then, Keith, we'll go back
• 10:58
  - to you for some other insights.
• 11:01
  - Ed, when you go into an organization, how
• 11:03
  - much unused capability do you typically find sitting
• 11:06
  - in the EHR that was already implemented? And
• 11:08
  - and what does a rigorous look at that
• 11:10
  - involve before anyone signs a new contract for
• 11:13
  - that point solution or AI tool?
• 11:16
  - Yeah. I I think if most people are
• 11:18
  - honest about this in their own organizations, they'd
• 11:20
  - probably agree. There's probably more,
• 11:23
  - unused capability than people realize, you know, for
• 11:26
  - sure. It's... I love this, like, EHR first,
• 11:29
  - whoever our vendor is. In in Gene's case,
• 11:31
  - they implemented Epic, you know, on the clinical
• 11:34
  - side for everything. And so try to implement
• 11:37
  - as much Epic as you can and go
• 11:38
  - forward. If it's another vendor, same thing. Makes
• 11:41
  - a ton of sense without point solutions, but
• 11:44
  - it seems like things kind of fall into
• 11:46
  - four categories, I guess, maybe of unused or
• 11:49
  - untapped capability. And so
• 11:51
  - it's not always, like, functionality that might have
• 11:53
  - been licensed but not deployed,
• 11:55
  - maybe functionality that was deployed but poorly configured,
• 11:59
  - which can cause problems and functionality that works
• 12:01
  - technically but was never really adopted because of
• 12:03
  - the work flow changes
• 12:05
  - and the... You change management and driving that
• 12:07
  - user adoption. And then maybe the last piece
• 12:09
  - to me is really around data and analytics
• 12:11
  - and how how do you get your arms
• 12:13
  - around standardizing that data so it makes sense
• 12:15
  - to your organization and you can drive it
• 12:17
  - to be, you know, I guess, provide meaningful
• 12:19
  - insights at... When it's necessary.
• 12:22
  - I... People always look at or ask me
• 12:24
  - about adding more stuff, and I think we'll
• 12:26
  - get to that in the AI realm here
• 12:27
  - in just a moment. But, you know, I
• 12:29
  - think that can make sense, but I... I've
• 12:31
  - heard somebody say this once, and I like
• 12:33
  - this. Like, optimize before you accessorize.
• 12:36
  - So take full advantage of what you have
• 12:37
  - first. If there's first... If there's some reason
• 12:40
  - it doesn't fit the niche, which is very
• 12:42
  - possible,
• 12:42
  - then look at alternatives. But don't just,
• 12:45
  - you know, more technology doesn't necessarily mean you're
• 12:48
  - delivering better care. It's just more technology and
• 12:51
  - that can get in the way sometimes.
• 12:53
  - Mhmm. Yeah. Optimize before you accessorize.
• 12:56
  - Love that.
• 12:58
  - So, Jeet, I'm curious how that matches your
• 13:00
  - experience at SBH.
• 13:02
  - Where have you found room inside your existing
• 13:04
  - EHR environment?
• 13:06
  - I definitely echo every statement,
• 13:09
  - every adjectives,
• 13:11
  - what I'd mentioned.
• 13:13
  - Technology is just an enabler.
• 13:15
  - For me, people
• 13:17
  - is important.
• 13:18
  - So our important is the processes.
• 13:21
  - And, he's absolutely correct. We are still in
• 13:24
  - the optimization phase. We are three years.
• 13:28
  - Maybe we are in little bit of more
• 13:30
  - of acceleration
• 13:31
  - phase, but clearly not transformational yet.
• 13:34
  - The reason is, as, Ed mentioned,
• 13:38
  - we do have
• 13:39
  - so many,
• 13:40
  - I would say,
• 13:42
  - configured tools within our EHR,
• 13:45
  - which we have not utilized,
• 13:47
  - to the fullest potential.
• 13:50
  - And as I'd mentioned regarding the data, so
• 13:53
  - the data gives a good feedback that how
• 13:56
  - we can reengineer
• 13:58
  - our clinical workflows
• 14:00
  - and make it more
• 14:02
  - enhanced. So it's like a continuous,
• 14:04
  - loop
• 14:05
  - or what in quality management is called PDSA
• 14:09
  - cycle. It's a continuous,
• 14:11
  - learning and improvement cycle. So we get the
• 14:14
  - data.
• 14:14
  - We identify some flaws in our workflows. We
• 14:18
  - reengineer workflows, or we advise,
• 14:21
  - the EHR
• 14:23
  - vendors,
• 14:24
  - or partners.
• 14:25
  - And,
• 14:26
  - in their next upgrades, they do implement those,
• 14:29
  - fixes.
• 14:30
  - So, again, it's a journey, and it's in
• 14:32
  - the exploration
• 14:34
  - or optimization
• 14:35
  - phase for us. And I'm I'm
• 14:37
  - I'm pretty much sure it's for everyone.
• 14:40
  - And then, hopefully, the health outcomes in terms
• 14:43
  - of longevity,
• 14:44
  - less morbidity,
• 14:46
  - decrease infant mortality
• 14:48
  - rates across the globe and across The US,
• 14:51
  - those are the good indicators.
• 14:53
  - And in terms of value based care, yes.
• 14:55
  - Definitely, we are seeing better outcomes,
• 14:58
  - based on the preventative,
• 15:00
  - or wellness visits and,
• 15:02
  - appropriate timeliness
• 15:04
  - of the interventions,
• 15:06
  - based on the chronic diseases.
• 15:08
  - So,
• 15:09
  - yes, it's...
• 15:11
  - As I mentioned, again and again, it is
• 15:13
  - still a journey.
• 15:15
  - We are continuously
• 15:16
  - learning from feedback from the clinicians, from the
• 15:20
  - caregivers, and for... From the patients,
• 15:23
  - and we enhance the capabilities of our tools.
• 15:26
  - So, again, processes also have to be looked
• 15:28
  - at. We cannot be using the same process
• 15:31
  - what we were using
• 15:32
  - in the past where we were on the
• 15:35
  - paper.
• 15:36
  - And, obviously, we'll have to, continuously improve on
• 15:40
  - the regulatory changes,
• 15:42
  - or clinical guidelines,
• 15:43
  - based on the evidence,
• 15:45
  - which are coming.
• 15:47
  - So, again, it's a continuous,
• 15:49
  - improvement for us. And I would definitely add
• 15:52
  - that, you know, they did a great job
• 15:54
  - at SPH of not the old school lift
• 15:56
  - and shift of just replacing one vendor with
• 15:59
  - another, but really looking at process redesign. That's
• 16:02
  - that's sort of how you become more transformational
• 16:04
  - and take advantage of the new tools that
• 16:06
  - you have.
• 16:07
  - And, Erica, if if I may want to
• 16:10
  - add, continuous
• 16:11
  - training and education
• 16:13
  - and upskilling of the workforce.
• 16:16
  - Because we have a very heterogeneous
• 16:19
  - workforce,
• 16:20
  - from
• 16:21
  - the CNAs
• 16:23
  - to fellows to attending physicians or consultants,
• 16:27
  - and not everybody has the same
• 16:30
  - level of training or understanding
• 16:32
  - of the use of the tool.
• 16:34
  - So it is very, very important that once
• 16:36
  - you implement, you have to continuously
• 16:38
  - train, retrain, personalize,
• 16:41
  - or thrive training,
• 16:43
  - the... What they call it. So,
• 16:45
  - the more you are trained, the better you
• 16:48
  - can utilize these tools.
• 16:50
  - Mhmm.
• 16:51
  - It's a great point, Jeet, and it really
• 16:53
  - underscores what you've shared throughout the conversation today,
• 16:56
  - which is that this is a continuous journey
• 16:58
  - re... Refining your team's use of the EHR
• 17:00
  - and and the tools and capabilities within it.
• 17:03
  - And I'm curious to
• 17:05
  - kind of... Along those lines of seeing this
• 17:07
  - as a journey, I... I've... I'm curious about
• 17:09
  - what you both see
• 17:11
  - three to five years out from now. You
• 17:13
  - know, much of our conversation has focused on
• 17:15
  - this balance of optimizing
• 17:18
  - what you have already in your EHR and
• 17:20
  - then layering on tools where those capabilities aren't
• 17:22
  - there. But, you know, in three to five
• 17:24
  - years, where do you expect
• 17:26
  - much of the AI tools that clinicians are
• 17:28
  - using like documentation,
• 17:30
  - decision support, workflow automation?
• 17:32
  - Where do you think they'll live? Will it
• 17:34
  - will it be inside the EHR or maybe
• 17:36
  - another layer that sits alongside it?
• 17:39
  - I mean, I I kinda feel like the
• 17:41
  - EHRs are gonna continue to remain like the
• 17:44
  - transactional backbone.
• 17:46
  - But the AI, it's got to become,
• 17:49
  - like, the intelligence and orchestration layer around the
• 17:52
  - EHR
• 17:53
  - and... Because you're gonna have multiple, multiple solutions
• 17:56
  - popping in. And clinicians
• 17:58
  - don't care where it lives, but they sure
• 18:00
  - don't wanna have to access 12 or 15
• 18:02
  - different things. You know, somehow, it's gotta be
• 18:05
  - native in our workflow. And I think that's...
• 18:07
  - You know, integration is gonna be key of
• 18:09
  - all those things. But as the HR vendors
• 18:12
  - develop more AI tools within their tool set
• 18:15
  - and and native to their environment, I think
• 18:17
  - that's helpful.
• 18:18
  - But, you know, sometimes I feel like they
• 18:20
  - should stick to what they do right and
• 18:22
  - let the innovation people really focus on innovation.
• 18:24
  - And and you'll see a lot of these
• 18:26
  - other AI tools that in... Kind of aid
• 18:29
  - the clinicians,
• 18:30
  - aid everyone
• 18:31
  - going forward, but it's got us feel native
• 18:33
  - somehow. So I see the HR being transactional.
• 18:36
  - I see the AI engines
• 18:38
  - and models all sort of sitting outside of
• 18:40
  - that, and it shouldn't matter to the clinicians
• 18:43
  - providing care. It should seem as one tool.
• 18:45
  - Mhmm. Yeah, Ajit. How about you? Yeah. I'll
• 18:48
  - take a little bit of a different approach.
• 18:50
  - Again, EHRs are foundational.
• 18:52
  - They they will be there for a very
• 18:54
  - long time.
• 18:55
  - But the
• 18:57
  - other solutions
• 18:58
  - which will be decreasing cognitive overload or improving
• 19:01
  - efficiencies,
• 19:03
  - whether they are, operational or financials or even
• 19:06
  - the clinical,
• 19:07
  - it may come from outside the e h...
• 19:09
  - Traditional EHR vendors,
• 19:11
  - but they all have to come together as
• 19:14
  - one platform.
• 19:15
  - We don't want to have several,
• 19:18
  - credentials to log into different systems
• 19:21
  - to provide the same care.
• 19:23
  - Now in terms of AI, not all AIs
• 19:26
  - are the same.
• 19:27
  - So what I mean is the generative AI
• 19:29
  - solutions,
• 19:31
  - which uses your own data
• 19:33
  - to start with. I think
• 19:36
  - having a good data foundation
• 19:38
  - is
• 19:39
  - very important.
• 19:40
  - But when it comes to agents or agentic
• 19:43
  - AI,
• 19:44
  - it may come from outside the HR vendor
• 19:47
  - industry
• 19:48
  - or it may come from the HR. It's
• 19:50
  - very difficult to predict
• 19:52
  - what is gonna happen from three years of...
• 19:55
  - Three or five years is a very long
• 19:57
  - time. The velocity of the information which is
• 20:00
  - coming in the pipeline,
• 20:01
  - perhaps maybe two to three years.
• 20:04
  - So I think what would be transformational
• 20:06
  - is the entire
• 20:07
  - business transformation. Not only
• 20:10
  - the patient to physicians or clinician
• 20:13
  - encounter, but I'm talking about entire health care
• 20:16
  - business transformation
• 20:18
  - through agentic AI workflows. So we have to
• 20:20
  - relook
• 20:21
  - how these digital assistants,
• 20:24
  - will orchestrate
• 20:25
  - the
• 20:26
  - workforce
• 20:28
  - as a digital assistant
• 20:29
  - or what I call it human in the
• 20:32
  - loop
• 20:33
  - and very little autonomous
• 20:35
  - work processes.
• 20:36
  - That is going to be very, very transformational.
• 20:39
  - So, yes, there are some very good examples
• 20:42
  - or use cases at the moment.
• 20:44
  - What you just mentioned regarding the ambient listening
• 20:47
  - or chart summarization
• 20:49
  - or,
• 20:50
  - to some extent,
• 20:52
  - semiautonomous
• 20:53
  - coding. Yes. Those are important, but those are
• 20:55
  - not the only clinical workflows we have. There
• 20:58
  - are several other clinical workflows,
• 21:00
  - which happens at the bedside or in the
• 21:02
  - ambulatory care or in the post acute care
• 21:05
  - or in the virtual care. So
• 21:07
  - having agents,
• 21:08
  - which apps like a shadow shadow or the
• 21:11
  - digital assistants,
• 21:13
  - will transform
• 21:14
  - the business of, health care in my opinion.
• 21:19
  - And, g, you read my mind because before
• 21:21
  - I let you both go, I wanted to
• 21:23
  - make sure we just briefly touched on AgenTic
• 21:25
  - since that's getting so much attention,
• 21:27
  - as we head into 2027.
• 21:29
  - I know as health systems consider implementing AgenTic,
• 21:33
  - a lot of
• 21:34
  - a lot of that possibility depends on trust.
• 21:37
  - So for you, Jeep, what would have to
• 21:39
  - be true before you let AI take action
• 21:41
  - across your systems rather than just assist?
• 21:44
  - I I think the use of agents will
• 21:46
  - be inevitable. We will be, using it,
• 21:50
  - in our day to day processes.
• 21:52
  - But first and foremost is the,
• 21:55
  - internal,
• 21:56
  - upskilling
• 21:57
  - or education to all our
• 21:59
  - leaders and all our providers and all our
• 22:02
  - clinicians,
• 22:03
  - all the way down to the,
• 22:05
  - everyone at our health system or anyone who
• 22:08
  - touches our patients and the community at large.
• 22:10
  - That is number one. Number two, a good
• 22:12
  - governance.
• 22:13
  - A very solid
• 22:15
  - foundational governance,
• 22:18
  - including the use of agents. Are we going
• 22:20
  - to use them as a human capital?
• 22:23
  - Are they trustworthy?
• 22:25
  - Obviously, we are not in the business of
• 22:27
  - developing agents at our health system,
• 22:30
  - But,
• 22:31
  - any partner who will be developing these agents,
• 22:33
  - we have to be
• 22:35
  - making sure that,
• 22:37
  - the risks are identified.
• 22:40
  - And,
• 22:41
  - for example, low risk,
• 22:43
  - depending on the workflows,
• 22:45
  - can be,
• 22:46
  - taken easily. But when it comes to clinical
• 22:48
  - interventions
• 22:49
  - or very complex,
• 22:51
  - orchestration of agents,
• 22:54
  - we need to have a guardian agents. So
• 22:57
  - we all, as a human, have to act
• 22:59
  - like a guardian agents,
• 23:01
  - making sure that,
• 23:02
  - these digital assistants or tools,
• 23:05
  - which are going to enable us to provide
• 23:08
  - safer care,
• 23:09
  - are vetted through very, very,
• 23:11
  - carefully,
• 23:12
  - and they are trustworthy.
• 23:14
  - Again, it is,
• 23:16
  - we have just started to begin the scratch,
• 23:19
  - the surface of agentic AI,
• 23:21
  - transformational work,
• 23:23
  - and we still have to learn,
• 23:26
  - how these agents work,
• 23:28
  - making sure that,
• 23:29
  - they perform the task
• 23:32
  - they are intended to,
• 23:34
  - and how
• 23:35
  - somebody can
• 23:37
  - abort the task,
• 23:39
  - if they go beyond
• 23:41
  - their specific
• 23:42
  - task.
• 23:43
  - So it's it's a little bit complicated,
• 23:46
  - but I think, as long as we have
• 23:48
  - good governance, as long
• 23:50
  - as as long as we act like a
• 23:51
  - guardian agents or for the guardian agents for
• 23:55
  - these tools,
• 23:56
  - I think, we will be able to provide
• 23:58
  - very efficient and, safer care to our community.
• 24:03
  - Yeah.
• 24:04
  - Important considerations, Jeep. Thank you.
• 24:06
  - And before we wrap, and, Jeep, any final
• 24:09
  - thoughts from you both before we close?
• 24:12
  - Again, I will emphasize on the journey.
• 24:16
  - This journey includes,
• 24:19
  - upskilling of task force,
• 24:21
  - which is very, very important,
• 24:23
  - and also education,
• 24:25
  - not only to the task force or the
• 24:27
  - employees, but also to the patients and caregivers.
• 24:31
  - How powerful these tools are if they can
• 24:34
  - be used
• 24:35
  - in a very, very safer manner.
• 24:38
  - And,
• 24:39
  - yes, we all are here to,
• 24:42
  - improve the health outcomes at the end of
• 24:44
  - the day,
• 24:45
  - and we will
• 24:46
  - need all sorts of partners in the industry
• 24:49
  - and outside the industry
• 24:51
  - to make health care safer
• 24:53
  - and,
• 24:54
  - to leave us,
• 24:56
  - on on have a better healthier outcomes.
• 24:59
  - Alright. That's good. I was trying to take
• 25:01
  - some notes too through a few things you
• 25:03
  - said. Like,
• 25:04
  - I think this EHR era, I guess, if
• 25:07
  - if you call it that,
• 25:08
  - sort of taught us that just acquiring technology
• 25:10
  - isn't the same as transforming care, and she
• 25:12
  - mentioned that earlier. You know, AI is not
• 25:15
  - gonna correct that either. Poorly designed workflow, unreliable
• 25:18
  - data, you know, unclear accountability,
• 25:21
  - in many cases, just gonna scale those things
• 25:23
  - faster and make
• 25:25
  - it worse. So I think that the people
• 25:26
  - who are creating, like, long lasting durable value
• 25:29
  - with their EHR implementations
• 25:31
  - and the tool sets around that are looking
• 25:34
  - at that brain problem a little bit differently.
• 25:35
  - They use what you already own. Use it
• 25:38
  - well. That's the optimization piece. Get full value
• 25:41
  - for the tools that you invest in already.
• 25:42
  - I think add when it makes sense incrementally,
• 25:46
  - but it's, in Chi Chi's word governance line.
• 25:48
  - I think as long as you see that
• 25:50
  - incremental value and you govern the autonomy and
• 25:53
  - split personal to risk, like, all those things
• 25:55
  - make sense. And I think that,
• 25:58
  - probably,
• 25:59
  - I'd say in health care, the the final
• 26:00
  - measure really can't be how many applications or
• 26:04
  - models we deploy.
• 26:05
  - It's really has to be whether clinicians get
• 26:08
  - meaningful time back, whether the patients receive more
• 26:11
  - safe care, more equitable care,
• 26:13
  - access improves. I mean, all the things that
• 26:15
  - we've been talking about for twenty five years,
• 26:17
  - it's time for that to happen. And then
• 26:20
  - I think the organization becomes more reliable. And
• 26:22
  - so the goal is not more technology.
• 26:24
  - The goal is to have better care with
• 26:25
  - less friction around the care.
• 26:28
  - Mhmm. So well said, Ed.
• 26:30
  - I just wanna thank you both for for
• 26:32
  - this conversation today. We've covered a lot of
• 26:34
  - ground from kind of the status quo of
• 26:37
  - EHRs to
• 26:39
  - improvements
• 26:40
  - and areas where you're focused on seeing greater
• 26:43
  - outcomes from your current system,
• 26:46
  - and then all the way to Agentic,
• 26:48
  - what's ahead, but not over indexing on the
• 26:50
  - technology itself, rather
• 26:52
  - the processes behind them, governance, and and the
• 26:55
  - patient outcomes, which, of course, are are steering
• 26:57
  - you in that direction. So I just wanna
• 26:59
  - thank you both again for the time that
• 27:00
  - you spent with Becker's today. Thank you. Alright.
• 27:03
  - Thanks, Erica.
• 27:04
  - Thank you.
• 27:05
  - Yeah. Well, it was a pleasure having you
• 27:07
  - both. And, of course, we also wanna thank,
• 27:10
  - Baker Tilly for sponsoring today's podcast.
• 27:13
  - Listeners, be sure to tune in to more
• 27:15
  - podcasts from Becker's by visiting our podcast page
• 27:17
  - at beckers hospital review dot com.