Latest / Becker’s Healthcare Podcast / Moving Healthcare AI from Experimentation to Execution with James Murray

Transcript

Moving Healthcare AI from Experimentation to Execution with James Murray · Becker’s Healthcare Podcast

0:00
This is Scott Becker with the Beckers Health
0:02
Care Podcast. I am thrilled today to be
0:04
joined by somebody who's had a brilliant career
0:06
at the intersection of health care and and
0:09
digital technology and a lot more. We're joined
0:12
today by Jim Murray. Jim's the managing director
0:15
of the BDO Center for Health Care Excellence
0:18
Innovation
0:19
and a brilliant thinker in person.
0:21
Jim, can I ask you to take a
0:22
moment and introduce yourself?
0:25
Thank you, Scott. I am absolutely thrilled to
0:27
be here.
0:29
I come to this discussion with a little
0:32
over twenty five years of health care
0:34
technology experience and
0:37
literally my entire career
0:39
I've worked in academic medical centers in retail
0:42
health pharmacy, payer environments,
0:45
all focused around technology.
0:48
And currently as as managing director,
0:51
with BDO,
0:53
I work with health care systems across the
0:54
country on digital strategy,
1:00
optimization
1:01
and my work is kind of at the
1:03
intersection of Technology and operations
1:06
and and leadership,
1:08
you know helping organizations
1:10
go after these unbelievably complex environments,
1:14
modernize their their clinical systems, or really translate
1:17
their technology investments into
1:20
value or measurable
1:22
outcomes. And,
1:23
big part of what I'm what I'm working
1:25
on, today,
1:27
advising,
1:28
executive teams on on how to take these
1:30
these fragmented
1:32
digital initiatives
1:34
towards a more, like, disciplined
1:36
enterprise level strategy
1:38
that that can really grow their business. And
1:40
I'm I'm just thrilled to be here with
1:41
you.
1:42
Well well well, thank you so much for
1:44
joining us.
1:45
Talk about currently the the trends that you're
1:48
watching in health care. What are you watching?
1:50
What do you have your eye on? What
1:51
are you most what what are you watching?
1:53
The the biggest thing that I'm that I'm
1:55
watching or seeing today is kind of a
1:57
shift from
1:59
experimentation
2:00
to execution,
2:02
and that's that's especially with artificial intelligence. And,
2:05
you know, we we have to say artificial
2:07
intelligence in every every paragraph we we talk
2:10
about now or every topic we talk
2:12
about. But, health systems aren't asking,
2:16
can we do this anymore? Instead, they're asking,
2:19
you know, what value does this create?
2:23
And so we're
2:24
we're seeing that,
2:26
happen across the industry.
2:28
We're seeing a very strong emphasis. I'm seeing
2:31
it on governance and kind of simplification
2:34
across the organizations. You know, these health systems
2:36
have have too many tools. They have overlapping
2:39
platforms. They have
2:41
disconnected
2:42
and kind of siloed
2:44
initiatives.
2:45
And so leaders are are
2:48
starting to do a a good job with
2:50
prioritizing,
2:51
you know, getting getting fewer but more integrated
2:54
solutions
2:55
and solutions that are are starting to move
2:58
metrics. So in, like, you know, the big
3:00
challenges around access or challenges around cost.
3:04
And then I think the the last thing
3:05
I'll mention in this space is,
3:08
a real push to to modernize
3:10
legacy environments. And and it's not not just
3:14
to not just for innovation, but really to
3:17
to bring costs down, bring friction down.
3:20
It isn't
3:22
always about more technology.
3:23
It's it's more discipline with technology
3:27
and ensuring that everything they're using kind of
3:30
ties to value and outcomes.
3:32
Thank you. And you you see systems become
3:35
more intentional.
3:36
How hard is it to do that with
3:38
so many different
3:39
applications and ideas
3:41
being pitched to them? How do they sort
3:43
out
3:44
where they should be intentional about use cases,
3:46
about AI, about technology,
3:49
and where they should sort of say no
3:51
or or say we just can't take that
3:53
on right now? How how do they sort
3:54
of
3:55
sort that out?
3:56
That's that's where the the governance
3:59
comes in, and that kind of
4:01
strategic mindset comes in and says, okay. We're
4:04
going to
4:05
approach this,
4:07
from
4:08
a a standpoint of we're we're not gonna
4:11
try
4:12
to keep everything in place. We're not gonna
4:14
try and, you know, cut things in half.
4:16
We're gonna actually rationalize
4:18
everything
4:19
that is there
4:21
and look at that. Like I mentioned, the
4:23
you know, moving those core metrics, figuring out
4:26
what is gonna impact the access, what is
4:29
gonna
4:30
impact the cost,
4:32
and, in a consistent way? So they're able
4:35
to to look at it holistically, and that
4:37
leads to the the less siloed initiatives too.
4:41
We're we're taking a a view of this
4:44
from,
4:45
an enterprise
4:46
perspective. And academic medical centers or or smaller
4:50
organizations that are doing this, they they look
4:52
at their whole ecosystem.
4:54
And we've seen that as a definite trend,
4:56
and and artificial intelligence is is driving a
4:59
little bit of that because,
5:02
people wanna jump into that space. But it's
5:05
definitely been a a theme that I've been
5:06
watching and and participating in.
5:09
Thank you. And talk about the work you've
5:12
been doing over the last couple years
5:14
and looking forward.
5:15
What are you most proud of from the
5:17
last year to two, and then what are
5:19
you most excited about and focused on going
5:21
forward?
5:22
What I've been proud about over the last
5:24
year or two has been the impact that
5:27
BDO is having and then other organizations as
5:30
well
5:31
are are bringing that focus and a little
5:33
more discipline and governance to
5:36
health care
5:37
IT. And we're able to
5:39
guide organizations
5:42
on on the value and and where to
5:44
find it. And we've got measurable results that
5:46
we can we can reference on that and
5:49
and really show that improvement. Because in this
5:51
crazy industry, you see things going
5:54
every which direction that you can possibly,
5:57
imagine and people, you know, falling in love
5:59
with new solutions every day.
6:02
And
6:03
what I like to say is is fall
6:05
in love with the problem.
6:06
You know, really get at a deep level
6:09
of understanding
6:10
why the organization,
6:12
why the system, why the process functions the
6:15
way it does,
6:16
And then you'll start to see
6:18
the the solutions or the logical next steps
6:21
kind of present themselves.
6:23
And and what excites me now is, like
6:26
I said, they're they're moving to execution,
6:28
and we're helping organizations move from vision to
6:31
execution.
6:32
We're we're turning,
6:34
you know, these these big digital ambitions
6:37
into real measurable
6:39
results,
6:40
whether we're helping organizations
6:43
understand the the value of improved data integrity
6:47
or taking,
6:48
you know, mundane
6:50
repeated tasks off the desks of of a
6:53
physician, of a nurse practitioner, nurse
6:57
down to the administrative staff, the support staff.
7:00
Anywhere we can do that, we're gonna
7:03
get everyone operating at the peak of their
7:06
license. The the physician will be practicing at
7:09
the peak of his or her license,
7:11
and that is really gonna
7:13
raise the whole system up.
7:15
And so, I think there's there's a lot
7:18
of value in in simplifying
7:20
those complex
7:21
environments, kind of figuring out what how your
7:24
technology decisions are are aligning with strategic goals.
7:29
And we're we're absolutely seeing,
7:32
you know, the improved access, the reduced burden,
7:35
the lower cost.
7:37
But
7:38
when you see an organization take advantage of
7:40
that without adding a bunch of complexity, that's
7:43
probably what I'm most proud of and most
7:45
excited about, that we're doing now.
7:48
When you see organizations
7:50
execute well on these strategies
7:53
versus not, what are you seeing them do
7:55
right when they're executing on digital strategies, integration
7:59
strategies,
8:00
and more? Where do you see it go
8:01
well?
8:03
Where it's where I see it goes well,
8:05
it's,
8:07
the organizations
8:08
that are
8:09
are looking at value
8:11
beyond just scale.
8:14
Value is gonna be created by by how
8:16
an integration
8:17
kinda reshapes the the economics, of course,
8:21
the the decision,
8:24
where where the decision sits, decision rights, and
8:27
really, ultimately, clinical
8:28
behavior.
8:30
There there are so many variables
8:33
when you look at
8:36
what's happening now, and and there's probably there's
8:39
probably way too many to discuss in in
8:41
the short time. But
8:42
I think organizations that focus on the data
8:46
and
8:47
data being used to drive behavior
8:50
rather than looking at data as a reporting
8:52
mechanism.
8:54
You know, the the real value
8:56
is driven by data that is that is
8:59
timely, clinically credible, and and tied directly to
9:03
consequences, ultimately.
9:05
And consequences can be both positive and negative.
9:08
But it's gonna have to
9:11
you know, if you have a dashboard or
9:12
you have a dataset of some sort, it's
9:14
gonna have to answer questions that physicians
9:16
really care about, that all providers really care
9:19
about.
9:20
And and I think that's that's what's going
9:22
right is when there's a focus on data
9:24
integrity and meaningful metrics coming out of these
9:27
organizations.
9:29
Right. That's when you can create real value
9:32
versus causing challenges. And and you talk about
9:35
the opposite when people are giving throwing data
9:37
at you,
9:38
but it's not actionable
9:40
or or it's not really done with the
9:41
user in mind.
9:43
Are there examples of spots where you see
9:46
integrations don't really help or can destroy value?
9:50
Any examples or thoughts there?
9:52
I I think, you know, it's,
9:55
maybe not destroy value, but eat away at
9:57
value. How about that? It's, it diminishes the
10:00
value when
10:02
when the data is,
10:04
not timely, for example.
10:06
You know, it's, like, okay. Great. You told
10:09
me something, but I needed to know that,
10:11
yesterday, a week ago, a month ago to
10:14
really make change
10:15
when it arrives late
10:17
or when the data
10:19
feels
10:20
punitive to the the providers or the clinical
10:24
staff that you're trying to move,
10:26
in a different direction.
10:28
They're they're not gonna be wanna be called
10:30
out for that. They're gonna wanna see
10:33
how they can change and and be better.
10:36
Metrics that are kind of disconnected from
10:40
from the levers that are in the practice
10:42
or in the organization become meaningless. And,
10:46
providers of all levels will will
10:49
begin to
10:51
show a distrust,
10:53
in the numbers or or even ignore them.
10:56
And and once that's once that's done, then
10:58
the value is gone.
11:00
So it's you know, you can throw a
11:01
lot of data out there. You can be
11:03
very transparent
11:04
in your data.
11:05
But without that actionability,
11:08
it's gonna just breed cynicism,
11:11
and and that's where it's gonna where it's
11:13
probably gonna
11:15
fall apart.
11:16
And I think the
11:17
you know, there's some areas around that and
11:19
and how you prevent that and and, you
11:22
know, make the metrics more meaningful, but
11:25
it's all gonna come down to standardization
11:28
and and treating standard data standardization as as
11:32
more of an ideology
11:33
than just a tool or process,
11:36
because then and you can then you can
11:37
impact clinical pathways. You can impact revenue cycle.
11:40
You can impact payer contracting
11:43
with by bringing standard data definitions. So it
11:46
kinda drive the the business intelligence.
11:50
Thank you. And are there places where you're
11:53
seeing very real measurable return on investment
11:57
from AI and automation in health care? Are
11:59
are there spots that you're seeing, you know,
12:01
real measurable
12:02
return on
12:03
investment? Absolutely.
12:05
The the
12:06
largest measurable returns that I'm seeing
12:09
are coming from,
12:12
automation
12:13
that that takes the friction
12:15
out of the work that that clinicians and
12:18
operators
12:18
do every day, you know, those repeatable tasks.
12:22
You know, it's it's in clinical documentation,
12:25
you know, is
12:27
is the ROI comes in when you're saving
12:30
time,
12:31
not just chasing, you know, a particular coding
12:34
dollar.
12:36
That's that's where the r, ROI is is.
12:39
And and we're seeing it now, or I'm
12:41
seeing it in, you know, ambient documentation for
12:44
outpatient visits
12:46
that that scribe in the background that's listening
12:48
in an automated note taking or note drafting,
12:51
I'm sorry, for,
12:53
clinician review,
12:54
inbox management, you know, just the the amount
12:57
of messages providers get on on a constant
13:00
basis,
13:01
chart summarization.
13:03
So they're not they're not flipping through page
13:04
after page, screen after screen
13:07
of of information.
13:09
They've got a a a virtual assistant.
13:12
And then in just the and you may
13:14
have seen this yourself, but your own experience
13:17
and and your listeners' experience,
13:19
the visit preparation and post visit documentation
13:22
are incredibly accelerated
13:24
by some of these tools, and there we're
13:26
seeing,
13:27
true value.
13:29
And there's there's, you know, there's those are
13:31
the those are the cool things that I
13:32
look at. You know, there's some, you know,
13:35
less glamorous, like the workforce management and, you
13:38
know, looking at demand forecasting
13:40
and and smart scheduling.
13:43
You know, all those
13:45
lead to an environment that helps with retention
13:48
of of staff and and providers.
13:51
You know, what what I'm saying,
13:53
I guess, is that all
13:56
all high ROI cases have
14:00
a a few things in common.
14:02
And,
14:03
I've summarized these or tried to boil these
14:06
down in in their, you know, their narrow
14:08
scope
14:09
focused on, you know,
14:11
one decision and one outcome.
14:13
They're, you know, keeping the human in the
14:15
loop. You know, AI is there. Automation is
14:18
there to recommend, but that the human is
14:20
making the decision.
14:21
It's it's easy sometimes when you look at
14:24
these systems and the solutions to,
14:27
quickly put it on the side of the
14:29
operation and reference it rather than truly embedding
14:32
it.
14:33
The ones that are bringing value are embedding
14:35
in the operation. There's no no swiveling on
14:38
the chair from screen to screen or system
14:40
to system.
14:41
Another one that is a differentiator that I've
14:43
seen is, the clear economic owner
14:46
of any initiative.
14:48
Like, who
14:49
where which p and l is this going
14:51
to? Where is this going?
14:53
And and getting that fast feedback,
14:57
fast feedback cycles so you can see the
14:59
value, you know, really in weeks, not years.
15:02
That's when it's gonna make a difference for
15:05
an organization.
15:07
No. I love that. And seeing sooner returns
15:09
versus later returns
15:11
is very important for a million reasons, momentum,
15:14
confidence,
15:15
having people continue to work with whatever it
15:17
is that you're doing. Is is that a
15:19
fair assessment?
15:21
Absolutely.
15:22
Absolutely. That's why we would go back to
15:23
the the strategy side and we say, hey.
15:26
You've gotta you've gotta map out a strategy,
15:28
so you're looking at this in kind of
15:30
a unified way.
15:32
That doesn't mean you have to wait a
15:33
year to start getting returns. You just have
15:36
to make sure that your initiatives
15:38
are are
15:39
accountable to your strategy and your mission,
15:42
and then you can start to to with
15:44
those fast cycles, the discipline, the standardization,
15:48
you can bring value in
15:51
rapidly
15:52
and build from there. And what and once
15:54
you start adding these stacking these,
15:57
strong,
15:58
returns, one on the other, it really is
16:00
it's surprising to a lot of organizations,
16:03
and they kinda wonder why they didn't do
16:04
it sooner.
16:06
No. I I love that. And it can
16:08
be hard to implement and to integrate, and
16:10
that's a lot of where you folks come
16:11
into, isn't it, in guiding people on how
16:13
to do that?
16:14
Absolutely. We have a a tremendous
16:17
team at BDO that
16:19
have spent decades in the industry and understand
16:22
how the different systems work and how the
16:24
how the data flows and probably most importantly,
16:27
the pitfalls of some of these, projects because
16:30
it it doesn't take,
16:32
much research to find, you know, projects that
16:34
have gone over time, over budget,
16:37
and not delivered the results. So very proud
16:39
to say we have an incredible team,
16:41
at PDO that understands the industry, understands the
16:45
solutions,
16:46
and have really been in the in the
16:47
trenches working on this stuff for many, many
16:49
years.
16:50
Jim, I wanna thank you for joining us.
16:52
Again, for our audience, for their understanding, for
16:55
the background, Jim is the managing director of
16:57
the BDO Center for Health Care Excellence and
17:00
Innovation. Jim, what a great pleasure to visit
17:02
with you today. Thank you so much. You
17:04
are too kind, Scott. It has been my
17:06
pleasure, and I really appreciate it.