Latest / Becker’s Healthcare Podcast / Colin LeClair, CEO of Connections Health Solutions

Transcript

Colin LeClair, CEO of Connections Health Solutions · Becker’s Healthcare Podcast

• 0:02
  - Pharmacy has long been viewed as a cost
• 0:03
  - center, but that framing is starting to shift,
• 0:06
  - and chief pharmacy officers are driving the change.
• 0:09
  - Don Wright, pharmacy is one of the most
• 0:11
  - powerful financial and clinical levers the health system
• 0:14
  - has. The question is whether its leaders can
• 0:17
  - prove it, fund it, and scale it before
• 0:19
  - that window closes.
• 0:21
  - Becker's third annual fall chief pharmacy officer summit
• 0:24
  - is where that case gets built with 500
• 0:26
  - plus attendees,
• 0:28
  - 90 plus speakers, and three days of of
• 0:30
  - candid executive level sessions on pharmacy's financial power,
• 0:34
  - three forty b sustainability,
• 0:35
  - specialty growth, automation, and earning real influence at
• 0:39
  - the top. You can earn up to 6.67
• 0:42
  - CE credits all while being in a room
• 0:44
  - full of peers solving the same problems you
• 0:46
  - are. Join us November fourth through the sixth
• 0:49
  - in Chicago.
• 0:50
  - Register on our events page at eckershospitalreview.com
• 0:53
  - by clicking on the events tab in the
• 0:55
  - upper right.
• 1:01
  - And I'll just give us a quick countdown
• 1:03
  - here.
• 1:05
  - Okay.
• 1:07
  - Three,
• 1:08
  - two, one.
• 1:09
  - Welcome to the Becker's Behavioral Health Podcast. I'm
• 1:12
  - Ella Reuter, your host. Today, I'm joined by
• 1:15
  - Colin LeClaire, CEO of Connections Health Solutions.
• 1:19
  - Today, we're discussing how Connections is using AI
• 1:22
  - to improve care delivery, reduce administrative burden, and
• 1:26
  - better support clinicians.
• 1:29
  - Colin, thank you so much for joining us
• 1:30
  - today.
• 1:32
  - Thanks, Sal. Thanks for having me.
• 1:34
  - Could you start off by telling us a
• 1:36
  - little bit more about Connections Health Solutions and
• 1:38
  - your role leading that organization?
• 1:42
  - Yeah. So, Connections Health Solutions builds freestanding psychiatric
• 1:45
  - emergency units. We specialize
• 1:48
  - in,
• 1:50
  - diverting patients away from justice system and inpatient
• 1:53
  - stays,
• 1:55
  - to provide immediate access to multidisciplinary
• 1:57
  - psychiatric care.
• 1:59
  - The goal being,
• 2:01
  - provide
• 2:02
  - faster access to higher quality care at much
• 2:04
  - lower cost.
• 2:05
  - And so, you know, we're focused on that
• 2:07
  - 5% of the acuity curve,
• 2:10
  - across the behavioral health population
• 2:12
  - that accounts for about 50% of total cost.
• 2:15
  - And of that 50% of cost, 25%
• 2:17
  - is,
• 2:18
  - is an inpatient.
• 2:19
  - And what we found empirically is that about
• 2:22
  - sixty percent of those inpatient stays are completely
• 2:24
  - avoidable
• 2:25
  - if patients have access to really high quality,
• 2:29
  - crisis care.
• 2:31
  - So we we'd launched our model for the
• 2:33
  - first time in Arizona
• 2:34
  - about eighteen years ago,
• 2:36
  - and, and that was sort of our test
• 2:38
  - case and
• 2:39
  - and
• 2:40
  - had really remarkable
• 2:42
  - outcomes. And I think we're now the most
• 2:44
  - studied and,
• 2:46
  - and
• 2:48
  - analyzed
• 2:49
  - the crisis provider in the country.
• 2:51
  - And we used that success to now expand
• 2:53
  - into,
• 2:54
  - by the end of this year, nine new
• 2:56
  - cities across the country,
• 2:59
  - across six states.
• 3:00
  - So we're we're we're creating the market for
• 3:04
  - this solution.
• 3:05
  - Demand is obviously,
• 3:07
  - unrepentant.
• 3:08
  - And,
• 3:10
  - and so what we're working through is creating
• 3:12
  - the the regulatory framework and the,
• 3:15
  - licensure
• 3:16
  - and the,
• 3:18
  - payment model
• 3:19
  - to allow this this model to thrive.
• 3:22
  - By way of background,
• 3:23
  - my career started in, globally capitated medicine working
• 3:26
  - first for health plans
• 3:28
  - and then moving into
• 3:29
  - globally capitated IPAs,
• 3:32
  - learning about the development of,
• 3:34
  - you know, high functioning, high quality
• 3:36
  - health care delivery systems for vulnerable populations primarily
• 3:39
  - focused on
• 3:40
  - Medicaid and de eligible Medicare populations.
• 3:44
  - And through the course of that early career,
• 3:47
  - you know, what I found was that behavioral
• 3:49
  - health was so often carved out
• 3:52
  - and delegated to, you know, a separate entity.
• 3:55
  - And
• 3:56
  - and and there weren't great solutions, particularly for
• 3:59
  - the really acute. And so,
• 4:02
  - about seven years ago, I came across this
• 4:03
  - organization,
• 4:05
  - and they were
• 4:06
  - delivering
• 4:07
  - high acuity
• 4:09
  - psychiatric care to the most vulnerable population, the
• 4:12
  - patients who needed it most. And they were
• 4:14
  - doing it in a value based fashion, so
• 4:15
  - I immediately fell in love with the model.
• 4:18
  - And,
• 4:19
  - and that's how I arrived in behavioral health.
• 4:23
  - Wonderful. Well, it's so exciting to see connections
• 4:26
  - really growing, especially as you mentioned with acuity
• 4:28
  - and demand continue to rise. And it's so
• 4:31
  - interesting to hear about that background in health
• 4:33
  - plans as well.
• 4:36
  - So thank you so much for that background.
• 4:38
  - I know we talked about earlier,
• 4:42
  - AI in behavioral health care, and it's been
• 4:45
  - a really big topic of conversation across the
• 4:47
  - board in health care. But I'm curious to
• 4:49
  - hear why has connections
• 4:51
  - made it such a priority right now.
• 4:55
  - Well, there's such an enormous opportunity to,
• 4:58
  - improve interoperability,
• 5:01
  - in behavioral health. You know, we were
• 5:03
  - we were not the beneficiary of of the
• 5:06
  - meaningful use,
• 5:08
  - policies that,
• 5:10
  - drove
• 5:11
  - physical health or accelerated physical health to move
• 5:14
  - into,
• 5:16
  - you know,
• 5:17
  - the the adoption of medical records and the
• 5:19
  - sharing of data.
• 5:20
  - And,
• 5:21
  - as a result of that and a number
• 5:23
  - of other factors, behavioral health was sort of
• 5:25
  - left behind.
• 5:26
  - And it's unfortunate because these patients are complex,
• 5:30
  - and
• 5:31
  - you you need to be able to coordinate
• 5:34
  - their social supports, their mental health needs, and
• 5:37
  - their physical health needs. And,
• 5:39
  - and behavioral health is sort of, you know,
• 5:41
  - in a siloed
• 5:42
  - space within health care.
• 5:44
  - So, you know, our our interest
• 5:48
  - in in AI solutions was born out of
• 5:50
  - necessity. We we had to find solutions that
• 5:53
  - could
• 5:55
  - aggregate all of these disparate data points from
• 5:57
  - across the patient's, you know,
• 6:00
  - lived experience
• 6:01
  - from all these sources, you know, prescribing and
• 6:04
  - lab and and all the different EHRs,
• 6:07
  - where their data lives and,
• 6:09
  - and try to create, you know, a a
• 6:11
  - more complete story so we understand our patient's
• 6:13
  - history, and we have
• 6:14
  - better tools to to to use to to
• 6:17
  - develop care plans.
• 6:19
  - And in a value based model, that's table
• 6:21
  - stakes.
• 6:22
  - In physical health, it's, you know, quote, unquote,
• 6:24
  - easier to do, and being brought up as
• 6:26
  - much harder. I think, you know, 42 CFR
• 6:29
  - has put a lot of fear in providers
• 6:30
  - that they can't share data,
• 6:32
  - and you do have to be very careful
• 6:34
  - sharing sensitive data.
• 6:37
  - But that kind of culture and the lack
• 6:39
  - of, you know, kind of EHR adoption has
• 6:41
  - has really set us back. So we've looked
• 6:43
  - to AI solutions to to
• 6:46
  - expedite
• 6:47
  - the collection of that data and make it
• 6:49
  - simpler and put it in the hands of
• 6:51
  - our clinicians to make better decisions at the
• 6:52
  - point of care.
• 6:53
  - And these are tools that weren't available to
• 6:55
  - us, you know, just a few years ago.
• 6:58
  - Now they are. And because of competition
• 7:01
  - in in the AI space,
• 7:03
  - that technology has never been more affordable, and
• 7:06
  - the data has never been more available.
• 7:08
  - And so,
• 7:09
  - you know, it's a value based provider that's
• 7:11
  - focused on outcomes.
• 7:13
  - It's it's
• 7:15
  - it's time to make this to make this
• 7:17
  - leap.
• 7:19
  - Mhmm.
• 7:21
  - And I love that focus especially on the
• 7:23
  - continuity of care
• 7:25
  - across the behavioral health spectrum, especially because we've
• 7:28
  - seen that be a huge hurdle and challenge
• 7:31
  - across the industry,
• 7:33
  - overall, especially,
• 7:35
  - as you mentioned too, with data operations
• 7:38
  - as well.
• 7:42
  - It sounds like that's really the main problem
• 7:44
  - that connections is really trying to solve, that
• 7:46
  - continuity of care.
• 7:48
  - And,
• 7:49
  - I know we've had conversations,
• 7:51
  - and I've had these conversations with other behavioral
• 7:54
  - leaders as well,
• 7:56
  - that really say AI should support clinicians, not
• 7:59
  - replace them. And I know you mentioned,
• 8:02
  - really trying to connect clinicians to AI. What
• 8:04
  - does that look like at Connections?
• 8:08
  - Yeah. So
• 8:09
  - there are
• 8:10
  - kind of a few,
• 8:13
  - you know,
• 8:15
  - major areas of focus
• 8:17
  - in health care AI. One is administrative simplification,
• 8:21
  - making the administration of health care much simpler,
• 8:23
  - more efficient, or eliminating it altogether.
• 8:26
  - There's care augmentation,
• 8:28
  - which is supporting clinicians to do the work
• 8:30
  - that they do, practice at the top of
• 8:32
  - their license, and not spend, you know, more
• 8:34
  - time than is necessary,
• 8:36
  - doing that administrative work and helping with clinical
• 8:39
  - decision making.
• 8:40
  - There's synthetic and computational biology, which is not
• 8:43
  - something that we're, you
• 8:45
  - know, in need of yet. But I think
• 8:47
  - as we move more toward
• 8:48
  - precision psychiatry,
• 8:50
  - there there may be opportunity for us there,
• 8:53
  - as that as those
• 8:54
  - solutions and tools reach, you know, a a
• 8:57
  - more of a Medicaid population.
• 8:59
  - Then finally, there's, you know, consumer empowerment.
• 9:02
  - And these are, by the way, you know,
• 9:03
  - categories that were defined really elegantly by Eric
• 9:06
  - Larson.
• 9:08
  - We we subscribe to this this kind of,
• 9:11
  - you know, framework.
• 9:12
  - We are focused almost exclusively on administrative simplification.
• 9:16
  - There's an enormous amount of administrivia
• 9:19
  - in behavioral health
• 9:21
  - that that wastes up to about 40% on
• 9:23
  - average
• 9:24
  - of our clinicians' time. Each discipline has a
• 9:26
  - different constraint, a different burden of adminware in
• 9:29
  - terms of administration.
• 9:31
  - But on average, on a weighted average basis,
• 9:33
  - our clinicians
• 9:35
  - spend about 40% of their time just moving
• 9:38
  - paper, chasing paper, chasing data.
• 9:41
  - And this is an in an industry that
• 9:43
  - are already, you know, has a pencil labor
• 9:45
  - shortage.
• 9:47
  - And and then that problem is exacerbated in
• 9:49
  - an emergency medicine environment when, you know, a
• 9:51
  - patient arrives at your door. They may not
• 9:53
  - be a great historian and then be able
• 9:55
  - to tell you succinctly or accurately what their,
• 9:57
  - you know, medical history looks like. And you
• 9:59
  - need to have really good clean data at
• 10:01
  - your fingertips the moment that patient arrives.
• 10:04
  - That's just not available. And so,
• 10:09
  - you're you're you're starting treatment blind.
• 10:13
  - So
• 10:14
  - eliminating all the administration required to collect that
• 10:17
  - collateral, collect that data, develop a more fulsome
• 10:19
  - picture of the patient and start a really
• 10:21
  - good care plan
• 10:22
  - is is is absolutely critical. So that's our
• 10:25
  - first priority.
• 10:26
  - Eliminating the administration,
• 10:28
  - giving our our team back more time to
• 10:30
  - work with the patient directly, spend less time
• 10:32
  - chasing paper,
• 10:34
  - and and looking through their EHR.
• 10:38
  - And, and more time with the patient is
• 10:40
  - gonna lead to better outcomes. It's also gonna
• 10:42
  - lead to faster treatment and and faster stabilization.
• 10:45
  - We are also looking at care augmentation
• 10:48
  - opportunities,
• 10:49
  - but we're being very, very cautious.
• 10:51
  - We have a very sensitive population,
• 10:54
  - very high risk.
• 10:55
  - The data on, you know, the efficacy of
• 10:58
  - these tools, the accuracy of these tools is
• 11:00
  - getting better and better and and and demonstrating,
• 11:03
  - you know,
• 11:06
  - pretty, you know, pretty high
• 11:09
  - accuracy rates, if you will,
• 11:12
  - in terms of supporting,
• 11:14
  - you know, diagnosis and supporting, you know, development
• 11:16
  - of treatment plans. But,
• 11:19
  - more than just human in the loop, we
• 11:21
  - want human
• 11:23
  - at the helm
• 11:24
  - and supported by,
• 11:27
  - AI,
• 11:28
  - you know, research and insight
• 11:32
  - so that we get it right. Over time,
• 11:34
  - we may
• 11:35
  - steer more toward, you know, evolving
• 11:39
  - AI,
• 11:41
  - supported diagnosis
• 11:42
  - and AI supported treatment plan development
• 11:45
  - more than we are today.
• 11:47
  - But for now,
• 11:48
  - we wanna give our our clinical team better
• 11:50
  - tools to use and not certainly not replace
• 11:53
  - them.
• 11:56
  - And then finally,
• 11:57
  - the I mentioned that we're gonna, you know,
• 11:59
  - we're probably going to pause on synthetic and
• 12:01
  - computational biology,
• 12:02
  - use cases for now. But consumer empowerment is
• 12:05
  - pretty you know, patients who are often experiencing
• 12:08
  - homelessness,
• 12:09
  - who are, you know, incapable of navigating the
• 12:12
  - health system on their own need to have
• 12:14
  - really good tools to keep the patient tethered
• 12:16
  - with their care plan, to keep them engaged.
• 12:18
  - And AI is an excellent solution
• 12:19
  - to follow-up with patients post discharge, to follow-up,
• 12:23
  - between visits, to check-in on them, and and
• 12:26
  - continue that line of communication
• 12:28
  - with an efficiency
• 12:30
  - for low touch, low risk,
• 12:32
  - engagement
• 12:33
  - opportunities,
• 12:36
  - and, and make sure we we understand where
• 12:38
  - they are and how they're doing
• 12:40
  - at all times.
• 12:41
  - Pretty powerful.
• 12:43
  - Yeah. It's a pretty powerful tool to to
• 12:45
  - facilitate that. So that's where we're focusing our
• 12:48
  - time.
• 12:51
  - And have you run into any challenges so
• 12:53
  - far,
• 12:54
  - or where are you,
• 12:56
  - really seeing success otherwise?
• 13:00
  - Yeah. One of the
• 13:01
  - the challenges that we're we're hyper aware of
• 13:04
  - and and we're considering with every AI investment
• 13:07
  - we make and every project that we explore
• 13:10
  - is
• 13:10
  - the risk of obsolescence.
• 13:12
  - There are so many AI solutions that are
• 13:14
  - popping up that are competing with one another.
• 13:17
  - It often feels as though somebody can
• 13:20
  - vibe code a solution that replaces the one
• 13:22
  - that you've just implemented.
• 13:24
  - And,
• 13:25
  - and all of a sudden, your solution is
• 13:27
  - one of dozens and,
• 13:29
  - and may very soon be be gone.
• 13:32
  - And so we're being very careful about who
• 13:34
  - we select, making sure that they have, you
• 13:35
  - know,
• 13:36
  - proven the track record and customer base and
• 13:39
  - and, you know, and they're viable.
• 13:42
  - We don't wanna spend enormous amount of time
• 13:44
  - implementing a solution that that may be gone
• 13:45
  - tomorrow.
• 13:46
  - We're also building
• 13:48
  - safeguards into our contracts so that if the
• 13:50
  - company does full, that we have the ability
• 13:52
  - to absorb that technology and the team to
• 13:54
  - to operate,
• 13:56
  - the the, the solutions that we've built.
• 13:59
  - So keeping a very close eye on that.
• 14:03
  - The the other is is team adoption. Right?
• 14:06
  - We're asking
• 14:07
  - our clinical team to
• 14:09
  - to support
• 14:11
  - you know, these new technologies. And change management
• 14:13
  - is hard in healthcare. Generally, there's risk, and
• 14:15
  - so you have to be very thoughtful and
• 14:16
  - careful in in your approach. But,
• 14:19
  - just change in general. You're you're asking clinical
• 14:21
  - clinicians to take a leap of faith and
• 14:23
  - trust technology.
• 14:25
  - And, and we hold technology to
• 14:28
  - a generally very high standard, maybe a higher
• 14:30
  - standard than we hold our own clinicians to.
• 14:32
  - So technology has to be infallible,
• 14:35
  - in order to
• 14:37
  - to develop and and and and retain the
• 14:39
  - trust of the team that will be using
• 14:41
  - it. And so we spent a lot of
• 14:43
  - time ensuring that the team is very much
• 14:45
  - on board and,
• 14:47
  - and that we get the adoption that we're
• 14:48
  - looking for. And and then finally, we spent
• 14:51
  - a lot of time looking at the before
• 14:52
  - and after to prove the use case to
• 14:55
  - the team. Right? So how what is the
• 14:57
  - goal of this
• 14:58
  - implementation?
• 14:59
  - How are we measuring success? What what what
• 15:02
  - was our success rate
• 15:03
  - before?
• 15:04
  - And then after adopting and and implementing this
• 15:08
  - technology and and and after some runway,
• 15:11
  - how is how does what does performance look
• 15:13
  - now look like now? And and,
• 15:16
  - proving to them kind of objectively that we
• 15:19
  - were good before, and now we're great. And
• 15:21
  - it's a result of you guys engaging
• 15:24
  - and adopting.
• 15:26
  - That's really important. I think it's those proof
• 15:28
  - points may otherwise be invisible to them in
• 15:30
  - the, you know, in the chaos of a
• 15:32
  - very busy day.
• 15:34
  - The front line doesn't usually have the time
• 15:36
  - to stop and look at
• 15:37
  - those types of metrics. It's important to show
• 15:39
  - them the kind of the fruits of their
• 15:40
  - labor, that the change that they just went
• 15:42
  - through was worth it. So spending a lot
• 15:44
  - of time on that.
• 15:46
  - And as you know,
• 15:49
  - many leaders are looking
• 15:50
  - for a clear return on investment from AI.
• 15:53
  - Could you talk more about how you're measuring
• 15:56
  - success? And are there any surprising
• 15:58
  - outcomes or results that you've seen?
• 16:03
  - Yeah. So, you know, number one is, like
• 16:07
  - I said, in in improving
• 16:10
  - the
• 16:11
  - time
• 16:12
  - and spend
• 16:13
  - on the administrations
• 16:15
  - of medicine. The all of the the busy
• 16:18
  - work that is not what they signed up
• 16:20
  - for.
• 16:23
  - It's rendering,
• 16:24
  - redundantly,
• 16:25
  - patient information.
• 16:27
  - How many clicks does it take to complete
• 16:30
  - an assessment or a psychosocial,
• 16:32
  - or an evaluation?
• 16:34
  - So we're doing time studies
• 16:36
  - of of all these actions to turn them.
• 16:38
  - Like,
• 16:39
  - is it working? Are we actually delivering on
• 16:41
  - our promise to our to our team that
• 16:43
  - we're taking
• 16:44
  - a painful administration out of their work? That's
• 16:46
  - the stuff that's burns them out. It frustrates
• 16:48
  - them. It it annoys them. This is not
• 16:50
  - what they, you know, became clinicians to do.
• 16:54
  - And,
• 16:55
  - and so saving
• 16:57
  - them time is priority number one. Priority number
• 16:59
  - two is ensure that that at the very
• 17:02
  - least,
• 17:03
  - the quality of care
• 17:04
  - and safety
• 17:06
  - of the the environment
• 17:08
  - has not declined and then certainly preferably that
• 17:11
  - we're actually improving,
• 17:12
  - quality. So we're measuring that very, very closely.
• 17:19
  - So far
• 17:20
  - and we're taking a very targeted approach to
• 17:22
  - the the AI initiatives that we're pursuing.
• 17:25
  - It's been very
• 17:26
  - successful, but we've also been very, very careful.
• 17:29
  - I think as we gain confidence in these
• 17:31
  - tools and we learn, you know, how to
• 17:33
  - implement
• 17:34
  - faster and getting the
• 17:37
  - faster, we will likely accelerate
• 17:39
  - our adoption
• 17:42
  - and and start working down our Pareto of
• 17:44
  - of opportunities.
• 17:45
  - I think there is risk in acceleration.
• 17:48
  - And so,
• 17:49
  - we haven't had any
• 17:51
  - major challenges yet, but,
• 17:53
  - as we accelerate,
• 17:54
  - I think we do have to keep an
• 17:56
  - eye
• 17:57
  - on change fatigue, and the team continue to
• 18:00
  - absorb new processes, new technologies,
• 18:02
  - because we may replace administrative fatigue with change
• 18:05
  - fatigue, and then we gotta be very careful
• 18:06
  - that we don't do that.
• 18:08
  - And, of course, again, we gotta keep a
• 18:10
  - very close eye on obsolescence
• 18:13
  - and make sure that the tools that we've
• 18:15
  - spent so much time, you know, earning their
• 18:17
  - trust around and implementing
• 18:19
  - don't go away tomorrow. If that only to
• 18:21
  - restart,
• 18:22
  - that will be endlessly frustrating for a team
• 18:25
  - that's already working really hard.
• 18:28
  - So no no none of those challenges yet,
• 18:31
  - but we are anticipating them and then planning
• 18:33
  - to
• 18:35
  - to, you know, to to
• 18:37
  - ensure we don't,
• 18:39
  - get out of our risk fees.
• 18:43
  - Mhmm. And I really love how you highlight
• 18:45
  - that importance on
• 18:50
  - implementing
• 18:51
  - AI and having those relationships with clinicians
• 18:55
  - that you're able to
• 18:58
  - comfortably adjust as things continue to develop and
• 19:01
  - as new pilots come out and as new
• 19:04
  - tools come out. I think that's super important,
• 19:06
  - and thank you so much for for highlighting
• 19:08
  - that.
• 19:09
  - In terms of the next
• 19:11
  - three to five years, where do you really
• 19:14
  - see AI changing behavioral health care?
• 19:18
  - And do you have any advice that you
• 19:19
  - would give to other behavioral health leaders
• 19:22
  - who are getting started in AI?
• 19:26
  - Yeah. I mean, I think the biggest change
• 19:27
  - that we'll see is
• 19:29
  - is the value that having a complete picture
• 19:32
  - of the patient will bring.
• 19:36
  - And what I mean by the complete picture
• 19:37
  - is all of their psychosocial needs, all of
• 19:40
  - their
• 19:40
  - psychiatric supports and needs, and all of their
• 19:42
  - medical supports and needs
• 19:44
  - all in one picture.
• 19:46
  - Too often, we we address these things in
• 19:48
  - a silo, and, you know, these conditions are
• 19:50
  - polygenic. There's they're the they're
• 19:53
  - multifactorial
• 19:56
  - additions and sorters. And so
• 19:58
  - to to look at just one area in
• 20:01
  - isolation
• 20:03
  - is is is not enough for a complex
• 20:05
  - patient. We do we need to consider the
• 20:07
  - whole patient?
• 20:08
  - We we don't have the ability to do
• 20:10
  - that
• 20:11
  - without these tools. I think that's gonna be
• 20:13
  - the biggest change is is practicing full person
• 20:16
  - medicine
• 20:17
  - in psychiatric care.
• 20:19
  - And that's gonna be a major shift. I
• 20:21
  - think we take for granted in physical health
• 20:23
  - that
• 20:25
  - more more often than not, we do have
• 20:26
  - a more complete picture. The patients in your
• 20:29
  - primary care practice, you generally are aware of
• 20:31
  - the
• 20:32
  - the
• 20:33
  - the
• 20:34
  - the the heart health care that they're receiving
• 20:35
  - if they're if they have a, you know,
• 20:37
  - CHF or
• 20:39
  - if they have COPD. You're you're you're aware
• 20:41
  - of, again, their their,
• 20:44
  - yes, all the medications that they're on, all
• 20:46
  - the services they're receiving.
• 20:48
  - That's not the case in psych psychiatric care
• 20:50
  - generally and historically. And so that will be
• 20:53
  - a a massive shift in how the industry
• 20:55
  - treats patients,
• 20:57
  - more holistically.
• 20:59
  - I think what
• 21:01
  - the availability data will also do is
• 21:04
  - usher in a new era of value based
• 21:07
  - era and value based contracting via real health.
• 21:09
  - If we now have the tools and the
• 21:10
  - data
• 21:11
  - to treat the whole patient and to manage
• 21:13
  - them longitudinally and to track progress,
• 21:15
  - and I think that's really important. We need
• 21:17
  - to move via real health to a more
• 21:19
  - outcomes focused
• 21:20
  - segment and approach, and I think we'll we
• 21:23
  - will soon have the tools to do that,
• 21:25
  - in a cost effective way. So I'm I'm
• 21:28
  - I'm very optimistic.
• 21:33
  - In terms of
• 21:35
  - what
• 21:36
  - I might recommend to other leaders and be
• 21:39
  - able to help her thinking about making this
• 21:41
  - leap is,
• 21:43
  - you can't use AI to replace that process
• 21:45
  - and bad data. So get your house in
• 21:48
  - order first. We spent several months
• 21:52
  - pulling together
• 21:53
  - a single data lake for all of our
• 21:55
  - different instances of our EHR across
• 21:58
  - seven locations at the time
• 22:00
  - and developing one single data lake with discrete
• 22:03
  - clean data so that we can run reporting
• 22:07
  - analytics
• 22:08
  - on all of
• 22:09
  - our locations,
• 22:10
  - all of our facilities in a common common
• 22:12
  - fashion. It also gave us now the ability
• 22:15
  - to use that data to feed, you know,
• 22:17
  - AI agents
• 22:18
  - and and have them start to do that
• 22:19
  - work for us.
• 22:21
  - Getting our house in order first was a
• 22:23
  - pretty critical first step, but I don't think,
• 22:26
  - that without without that, we we wouldn't have
• 22:29
  - been able to to make the progress that
• 22:31
  - we're making.
• 22:32
  - I think taking
• 22:33
  - that approach
• 22:35
  - to cleaning up data hygiene
• 22:37
  - and then
• 22:39
  - being very intentional about the problem that you're
• 22:42
  - trying to solve. And the more specific you
• 22:44
  - can be, the more effective the solution will
• 22:46
  - be.
• 22:47
  - What we did is is did time studies
• 22:49
  - across
• 22:50
  - the the the the processes that affected and
• 22:53
  - touched the most patient
• 22:55
  - involved the most in most of our the
• 22:58
  - the majority of our staff.
• 23:01
  - And and then isolated those that we believe
• 23:04
  - were the greatest opportunities
• 23:06
  - before we started to apply AI solutions to
• 23:08
  - this process even though we went in and
• 23:10
  - worked on work workflow first,
• 23:12
  - to see where there are opportunities to consolidate
• 23:15
  - and to to streamline. Once you've done those
• 23:17
  - two first steps, the outcome of your AI
• 23:19
  - implementation is gonna be dramatically more successful.
• 23:23
  - That was a discipline we kind of learned
• 23:24
  - the hard way. So,
• 23:27
  - yeah, if you if you take that kind
• 23:29
  - of disciplined approach
• 23:31
  - and and get hyper prioritized
• 23:34
  - and focus on,
• 23:36
  - you know, the top or the left side
• 23:37
  - of your,
• 23:40
  - I think, you're setting yourself up for success.
• 23:44
  - Wonderful. Well, that is all the time that
• 23:45
  - we have for today. Thank you so much
• 23:47
  - for joining us.
• 23:49
  - Yeah. My pleasure, Ella. Thank you for having
• 23:50
  - me.
• 23:51
  - Of course. And thank you so much to
• 23:53
  - our listeners for tuning in to Becker's Behavioral
• 23:55
  - Health Podcast.
• 23:56
  - For more conversations with behavioral health leaders across
• 23:59
  - the country, you can find us wherever you
• 24:00
  - get your podcasts.