Latest / Becker’s Healthcare Podcast / Scaling AI and Transforming Care Delivery at Cleveland Clinic with Dr. Rohit Chandra

Transcript

Scaling AI and Transforming Care Delivery at Cleveland Clinic with Dr. Rohit Chandra · Becker’s Healthcare Podcast

0:00
This is Laura Dierdahl with the Becker's Healthcare
0:02
Podcast. I'm thrilled today to be joined by
0:04
doctor Rohit Chandra,
0:05
chief digital officer at Cleveland Clinic. Rohit, it's
0:08
a pleasure to have you on the podcast
0:10
today.
0:10
Pleasure to be here.
0:12
Fantastic. And, you know, I'm excited to dig
0:14
into our conversation because I'm looking forward to
0:16
learning a little bit more about some of
0:18
the cool things you are doing at Cleveland
0:20
Clinic right now. I know you've got a
0:22
lot, with AI happening, and it's just really,
0:25
you know, important to have that perspective and
0:27
where you how you're thinking about the future
0:29
too. But before we dive in, can you
0:30
introduce yourself and just tell us a little
0:32
bit more about what you're doing at Cleveland
0:34
Clinic?
0:35
Sure.
0:36
I'm Rohit Chandra. I'm the chief associate officer
0:39
at the clinic. I joined the clinic about
0:41
four years ago actually from outside of health
0:44
care, so I'm a engineer by training.
0:48
And in my role here at the clinic,
0:50
I have responsibility
0:52
for
0:53
our technology footprint at the clinic.
0:56
And the way to think a little bit
0:57
about it is that technology is an critical
1:02
capability
1:03
for any industry, but for health care in
1:05
particular.
1:06
And so from that lens, I have responsibility
1:09
for traditional IT,
1:11
artificial health records,
1:13
data,
1:14
analytics,
1:15
and increasingly,
1:17
AI.
1:18
That's fascinating to hear. And, you know, a
1:20
a huge job in front of you. And
1:22
so I think if we narrow it down
1:23
to just the last year or so, can
1:25
you tell me a little bit about the
1:26
projects or initiatives that had the most impactful
1:29
results? What did you do, and how did
1:31
you measure those outcomes?
1:33
So,
1:34
for 2025,
1:36
I think,
1:38
for us and perhaps for many organizations
1:41
like us,
1:42
the single most impactful
1:44
AI initiative has been the AI Scribe.
1:47
And AI Scribe for the audience today, just
1:51
to simplify it, is
1:52
all of us will often, as patients, go
1:55
to a physician for an outpatient appointment. And
1:58
let's say it's a twenty or thirty minute
1:59
appointment
2:00
that we have, let's say, with our primary
2:02
care physician or some other physician.
2:05
What we don't realize is that after that
2:08
encounter,
2:10
physicians typically need to spend anywhere from ten
2:13
to fifteen minutes writing up their clinical notes
2:16
from the encounter.
2:18
So primary care doctor today can spend anywhere
2:22
from
2:23
one and a half to three hours
2:24
every day
2:26
after their clinic time
2:28
writing up their notes from each patient visit.
2:33
Now this is essential work. It's really important
2:36
for continuity of care, for safety, quality, and
2:40
everything else. So this documentation is a good
2:43
thing from every aspect. It's just that it's
2:45
very time consuming.
2:48
So what we,
2:49
what technology has increasingly emerged over the last
2:53
year or
2:54
two is something that I refer to as
2:56
an AI scribe.
2:57
So this is essentially
2:59
you can just on your iPhone, you can
3:01
have something that will listen to the provider
3:05
patient
3:06
conversation.
3:08
It will
3:09
record it,
3:10
transcribe it, and then automatically
3:13
generate a draft of the physician notes and
3:16
the patient notes.
3:18
And what is amazing is the technology has
3:21
become so good that it can generate really
3:23
high quality notes.
3:25
We still have a physician in the loop
3:26
to ensure that they're complete and accurate,
3:29
but,
3:31
it can
3:32
substantially
3:34
reduce the documentation burden on physicians.
3:37
And equally, if not more importantly,
3:40
it can
3:41
significantly
3:42
enhance the
3:44
patient provider
3:46
interaction
3:47
during the encounter.
3:49
So today, very often, if you were to
3:50
go to your see your doctor, the doctor
3:52
is often sitting at the computer sometimes looking
3:55
at you, but most of the time looking
3:56
at the screen typing up their notes. Well,
3:59
guess what? With this technology, they don't have
4:01
to do that. They can have their full
4:03
attention on the patient,
4:04
and, the documentation
4:07
happens almost magically.
4:09
So I would say looking back to 2025,
4:12
this has been
4:13
a game changing experience in reducing the documentation
4:18
burden,
4:19
enhancing the provider experience, enhancing the quality of
4:22
the provider patient interaction.
4:24
We've been very successful at it. We've had
4:27
more than 4,000 physicians adopted, and we've had
4:30
millions of encounters that have now been
4:34
documented
4:35
leveraging AI as opposed
4:37
to manually.
4:38
So I would say looking back in 2025,
4:40
that has been a big success story for
4:42
us.
4:43
That's great to hear. And I know that,
4:46
you know,
4:47
documentation technology and
4:49
AI scribes,
4:50
have a lot of value when it you
4:52
it comes to supporting the clinicians and making
4:55
sure that they're able to have more of
4:57
that human to human interaction.
5:00
As you look into the next year, where
5:02
do you see some of the big,
5:04
priorities as well as headwinds that you're focused
5:06
on for 2026?
5:09
So we're looking at bringing technology
5:12
in both
5:13
clinical and nonclinical
5:15
domains or operational areas.
5:18
I think just building on the Scribe conversation
5:21
we just had,
5:23
we think that the potential for AI
5:26
to build significant that the Scribe is just
5:29
the beginning.
5:30
We think that there are so many more
5:33
powerful capabilities
5:34
that can help a physician
5:36
in delivery of care. It can be
5:40
summarization
5:41
summarization of a patient chart before an encounter.
5:44
It can be engaging with the patient chart
5:47
in a much more conversational and intelligent way.
5:50
All of these are capabilities that will significantly
5:53
provide assistance to the provider at the level
5:57
during a patient encounter. So those are areas
6:00
where we are trying to enhance and build
6:02
upon
6:03
the AI scribe, which we think is just
6:05
a starting point.
6:07
Other areas are, I'll say, more in the
6:09
back office space. We've done good work over
6:12
the last year in bringing AI into coding
6:15
automation.
6:16
That has been pretty helpful for us and
6:18
has potential to significantly
6:20
accelerate that work for us going forward.
6:23
We've also
6:25
started to use AI and machine learning in
6:29
clinical risk prediction. So we launched something to
6:32
do sepsis prediction last year
6:34
that was very helpful.
6:37
Sepsis, just for our listeners,
6:39
is a
6:41
a bloodstream infection that is very complex to
6:44
detect
6:45
but and has very high mortality rates.
6:48
And so with AI, we can we built
6:50
a system that does early detection of sepsis
6:54
and that can significantly
6:55
improve our ability to intervene early
6:59
in patients that show the onset of sepsis.
7:02
When I'm going with this is all of
7:03
these are areas
7:05
beyond just the AI scribe
7:08
where we think we have significant potential to
7:10
apply AI powered capabilities
7:12
in other aspects of health care developing.
7:15
That's helpful to understand and really cool to
7:17
think about the technology and and implications on
7:21
how it can augment it and boost the
7:23
quality of care you're able to deliver and
7:25
especially with early detection on something like sepsis,
7:28
I know is, life saving.
7:30
I'm curious. When you look at bringing in,
7:33
more of the technology into the clinical setting,
7:36
how do you think about,
7:38
governance and monitoring and just making sure that
7:41
AI stays on track and is giving you
7:43
the results that you need? I know it's
7:44
not always just, kind of implementing the technology
7:47
and and letting it go, but truly making
7:49
sure that over time, it's still performing at
7:52
the rate you want it to be performing.
7:54
Oh, I think, Laura, you're exactly right, which
7:57
is
7:58
health care
8:00
is complex.
8:01
It's
8:03
tech and technology is only an
8:06
enabler.
8:07
It's very important for us to work
8:10
super closely and in deep partnership
8:13
with the clinical leaders at the clinic to
8:17
ensure that this is helping with health care
8:19
development.
8:20
The second aspect that I think is really
8:22
important is,
8:24
you know, AI is in its early stages,
8:26
so it's important for us as we navigate
8:28
these waters to do it as safely and
8:31
as responsibly as possible.
8:33
And so we always have a human in
8:35
the loop so that, you know, the technology
8:38
is helping assist.
8:39
But, you know, the technology is never perfect.
8:42
And so,
8:44
we just need to be very thoughtful about
8:47
what are the areas that we bring AI
8:48
to bear and how do we apply,
8:51
apply it in a way that has a
8:53
human in the loop so that we can
8:54
be safe and responsible.
8:57
So that's something that's core to us when
8:59
we
9:00
look ahead,
9:02
in terms of bringing AI and applying it
9:04
to other areas that I just touched upon.
9:07
That makes a lot of sense, and it's
9:09
helpful to understand how you're thinking about those
9:11
things as I know it's a such a
9:13
fast evolving field.
9:15
What do you think the hardest thing you'll
9:17
have to do in the coming year will
9:18
be?
9:20
I think the
9:22
main
9:23
opportunity
9:24
ahead of us
9:26
is really
9:27
how do we scale up our
9:30
ability
9:32
to
9:34
not just the two or three things that
9:35
I mentioned,
9:37
but 10 or more different initiatives across the
9:41
organization. And expands,
9:44
we need to expand our ability to mobilize
9:47
ourselves as an organization
9:49
to
9:50
pick and choose the right problems,
9:52
mobilize the technology teams, move the clinical and
9:56
operational leaders
9:57
to say, what are the problems that we
9:59
want to go after? How do we
10:01
work with
10:03
our technology partners to properly build the right
10:06
solutions?
10:07
And as I mentioned,
10:09
bringing an air in a clinical setting is
10:12
hard work.
10:13
We need to make sure that we are
10:16
not bringing in just technology, but we're doing
10:19
this in a way where we cannot just
10:20
apply it safely, but, you know, apply it
10:23
at scale and stay with it till it
10:25
is deployed enterprise wide. So, actually, the change
10:29
management is
10:31
essential
10:32
to do carefully and responsibly. That is actually
10:35
as big a lift as the technology.
10:38
And looking ahead,
10:40
I think it's really twofold.
10:43
Let's make sure that we're picking the right
10:45
areas to apply it to. And then number
10:47
two, to be able to do these efforts
10:49
at scale with
10:51
that carries the organization and carries health care
10:53
forward.
10:55
That makes a lot of sense. It's helpful
10:57
to think about it in that way. I
10:58
mean, I know, you know, building a team,
11:01
that can be nimble and,
11:03
ready to,
11:05
you
11:06
know, incorporate technology is
11:09
easier said than done in many ways. And
11:11
so when you look at the culture at
11:13
Cleveland Clinic, I know it's a special place
11:15
and certainly has been an innovation leader for
11:17
a long time.
11:19
What do you do to make sure that
11:20
your teams are,
11:22
you know, ready to
11:24
to go with the AI technology and are,
11:27
staying innovative while also staying safe?
11:33
So it's a work in progress. I think
11:35
that
11:37
all of us, as well as folks,
11:40
in other industries, are
11:42
grappling with how to bring the right technology
11:44
and tools into the organization,
11:47
how to give ourselves a safe space to
11:49
explore, to experiment, to learn.
11:52
So I think that,
11:55
it's a journey.
11:57
A few things that we have done
11:59
are we hired a chief AI officer,
12:04
eighteen months back from outside the industry,
12:07
and we are increasingly building up a team
12:10
of folks who have background in technology.
12:14
And
12:15
the approach that we're taking is
12:19
how do we
12:21
build
12:22
deep relationships and partnerships with clinical and operational
12:26
leaders at the clinic
12:27
so that we can,
12:30
bring a technology first mindset to many of
12:32
the important areas within health care. That's the
12:35
opportunity that I think is ahead of us.
12:37
AI is a very powerful tool,
12:40
but in our business,
12:41
how do we apply it in a health
12:44
care setting is the challenge and the opportunity.
12:47
So bringing in fresh talent is a thing.
12:49
Picking and choosing problems is a thing. Giving
12:52
ourselves a safe space to experiment, to try,
12:55
to learn as we go along.
12:57
These are all the
12:59
ways in which we are
13:01
increasingly activating our organization
13:04
to be,
13:06
to apply AI to health care.
13:09
That's helpful to understand and, you know, really
13:11
a great approach,
13:13
in strategy, it seems like, in order to
13:15
be ready for whatever comes next. And, you
13:18
know, speaking of that, how do you see
13:20
health care changing and shifting and evolving because
13:22
of the technology and because of AI as
13:24
it's coming more rapidly into the space? Where
13:27
are those some of those opportunities that you
13:29
see and are excited about for 2026 and
13:32
beyond?
13:33
So let me actually look at it,
13:37
from a, perhaps, a multi year perspective.
13:41
I think
13:43
that
13:44
I approach it both from
13:49
a operational as well as a clinical setting
13:51
where I think over time,
13:54
each of these areas have the potential to
13:56
be substantially
13:58
transformed using AI.
14:00
If you just look at the from an
14:03
operational lens,
14:04
we are a complex organization
14:06
that does more than 30,000
14:09
patient encounters a day.
14:12
Every patient encounter
14:14
requires the coordination
14:16
of
14:17
an exam room, an operating room, providers,
14:20
nurses, physician assistants.
14:22
So we have to coordinate the activity
14:25
of tens of thousands of
14:27
caregivers
14:29
and orchestrate
14:30
30,000
14:31
encounters.
14:32
But guess what? For every patient, it's perhaps
14:35
the most important
14:37
event of the day, and we need it
14:39
to go perfectly.
14:41
So this is an area where we're using
14:43
technology to help do the coordination of all
14:46
of these activities
14:47
so that they work like clockwork.
14:50
We want technology and algorithms to automate
14:53
to and to optimize
14:55
these encounters.
14:56
The benefit over time for the industry is
14:59
we should be able to do our activities
15:01
more efficiently
15:03
at greater scale and give us the ability
15:05
and maximize our ability to see as many
15:08
patients as we can. So this is an
15:10
example of an opportunity over time in an
15:13
operational setting.
15:16
In
15:17
a couple of examples I'll give from a
15:19
clinical setting perspective,
15:21
we know that,
15:22
health care is a very
15:24
labor intensive industry,
15:26
and our ability to deliver care is actually
15:29
limited by our capacity
15:31
by providers and nurses.
15:33
At the same time, providers and nurses are
15:36
overworked,
15:37
and there's significant burnout.
15:40
So there's an opportunity to say, how can
15:42
we bring technology
15:43
to substantially
15:45
transform and ease
15:47
the ability, the efficiency, and the experience of
15:50
our caregivers.
15:51
And ASRIVE is an example that I just
15:53
gave, but,
15:55
over time, I think all of these technologies
15:58
have significant potential to reduce the
16:02
busy work that caregivers go through. This is
16:05
good for doctors. It's gonna be good for
16:07
nurses, and it is going to be good
16:09
for patients because it will expand our ability
16:12
to deliver care.
16:14
The third area I'll touch on over time
16:16
is just improving
16:18
patient clinical care and safety.
16:21
Having technology
16:24
that can, let's just say, do risk prediction
16:27
and risk stratification of patients
16:30
helps us focus our energy on the patients
16:32
who are at highest risk for adverse
16:36
symptoms and consequences,
16:38
and that helps us focus our clinical energy
16:41
on the patients that need it most. So
16:44
this will help us actually take better care
16:46
of patients. So those are examples
16:48
where we think
16:50
we can substantially
16:52
improve
16:53
the quality and the safety of the care
16:55
that we deliver to our patients.
16:57
So if I step back over time,
16:59
if I look at
17:01
the macro landscape,
17:03
I am a deep believer that AI can
17:06
help us make health care safer,
17:09
cheaper,
17:10
and more scalable.
17:13
I love that. Roa, thank you so much
17:14
for joining us on the podcast today. This
17:16
has been a a really fun and fascinating
17:18
conversation, and I look forward to seeing you
17:20
as well in April. I know you'll be
17:22
speaking at our annual meeting, and doing a
17:24
fireside chat, which will be amazing to dig
17:27
deeper into many of these themes and and
17:29
just the things that you're doing at Cleveland
17:30
Clinic. So I appreciate your time today and
17:32
look forward to seeing you in a couple
17:34
months.
17:35
Thank you so much, Lauren.