Aug 30, 2024
Generative AI is a known disruptor in healthcare today. Will this transformational technology be embraced by clinicians, patients, and healthcare organizations? Listen in as two experts from Accenture Healthcare and Oracle discuss the difference between traditional AI and GenAI, the opportunities that GenAI is presenting to the industry, and the need to lean in to utilize technology as an enabler and a change agent. Hear how care delivery can be reimagined with GenAI and how this technology has the potential to be applied to help reduce clinician burnout, augment the clinician-patient relationship, bridge workforce shortage gaps, reduce margin pressure, and more.
--------------------------------------------------------
Episode Transcript
00:00:00:17 - 00:00:22:10
Michelle
You're listening to Perspectives on Health and Tech, a podcast by
Oracle with conversations about connecting people, data and
technology to improve health for everyone. Hi, I'm Michelle
Flemings. I'm the industry executive director for Health Care for
North America Cloud Infrastructure at Oracle. Glad that you're
here. We're in the age of artificial intelligence.
00:00:22:12 - 00:00:41:29
Michelle
The opportunities that we're presented with using A.I. as an
industry are truly groundbreaking. And to be honest, I think we're
just getting started. I want to focus today on a subset of
traditional AI and want to talk a little bit about generative AI,
some of the ramifications, some of the risks, benefits and what
does it mean to health care.
00:00:42:01 - 00:00:47:29
Michelle
I am delighted to sit across today from Tej Shah and I'll have him
introduce himself.
00:00:48:01 - 00:01:08:00
Tej
Thanks, Michelle Tej Shah I'm an emergency medicine doctor, a
managing director, and Accenture's global health care practice.
I've got a ton of experience working across different parts of
health care as a consultant, obviously, but also as a venture
capitalist, investing in companies and most recently starting a
company in the health tech space.
00:01:08:00 - 00:01:16:07
Michelle
We're talking about AI and we always have to start with what is the
comparison between traditional versus generative AI?
00:01:16:09 - 00:01:17:21
Michelle
How would you put that?
00:01:17:23 - 00:01:41:11
Tej
We've been talking about AI for 50 plus years. This is not a new
thing. We've been talking about how we can leverage data to get
insights going from analytics to AI to GenAI. Really the
difference is, you know, before with AI we were able to extract
data and insights from data so we could figure out what the next
obvious data point was going to be.
00:01:41:13 - 00:02:19:03
Tej
And in 2017 there was an article that was published by Googlers
around attention is all that matters, right? And what happened with
that papers is we came up with a completely new framework where
we're now able through GenAI to not just predict what's the
next data point, but understand what's the next word that we can
generate. And it's taking into consideration the context of the
sentence to be able to make that prediction so that it's actually
appropriate and it's opened up a whole bunch of opportunities that
we're going to talk about today that is really transformative.
00:02:19:05 - 00:02:30:03
Michelle
Let's get into some of those opportunities. I think it's
fascinating that it's been as long ago that A.I. came about
and the general public really doesn't know.
00:02:30:03 - 00:02:35:14
Michelle
let's talk about some of the opportunities in patient care that
exists because of where we are now.
00:02:35:14 - 00:02:40:28
Michelle
With A.I.. We do a lot around documentation. Where else should we
be going?
00:02:41:00 - 00:03:08:03
Tej
Back in 1996, you know, we started using search engines and it
wasn't really until early 2000 when Google came around and the
search bot search box became our librarian and around that time
there was an article that was published by Will Carr in The
Atlantic called Is Google Making Us Stupid and what he really
meant.
00:03:08:03 - 00:03:33:29
Tej
when you read the article was, are we thinking differently? Are we
using our brains in the same way or reading as deeply as we did
before? And obviously the answer is no. I think we continued to use
our brains and we continue to, but we were using this tool as a
librarian. We were using it to identify information quickly and be
able to access it, you know, more readily.
00:03:34:01 - 00:04:04:19
Tej
And what GenAI has done for us is start to be an advisor. It's
enabling this transformation from technology, being a librarian to
an advisor and that's what we're using AI for now. And GenAI today,
right? So this idea of documentation and this is all relevant
because the next sort of wave of what we're going to do with GenAI
is it's going to start to act as an agent.
00:04:04:21 - 00:04:30:28
Tej
It's going to start to enable us to take these tasks away from the
from our day-to-day workflows and enable us to operate more
efficiently. There is no reason, for instance, that a doctor like
you or I should have to, you know, once we've submitted that a
patient needs to do a specialist appointment, continue to follow
up, make sure that that appointment was scheduled and have a bunch
of people that are in that workflow follow up on it.
00:04:31:01 - 00:04:49:27
Tej
It should just happen. And I think that generative AI is going to
create these agents and these agents are going to be linked to one
another. We're going to chain them together and it's going to do
that follow up. It's going to make that experience more seamless,
really enabling clinicians to do the work that gives us joy.
00:04:49:29 - 00:05:07:07
Michelle
Let's drill down on that now, because you said something
fascinating there being an agent. How I see it as being is now, I
don't have that extra chaos and clutter to remember. As an ER doc,
we're in the midst of the chaos we are in and we're trying our best
to multitask. And there are fewer of us now.
00:05:07:10 - 00:05:25:16
Michelle
There are others that are doing an exit now and then. We also don't
have as much of a pipeline because, as you know, some of our
residency programs didn't fill out. So three years from now, we're
talking a massive deficit. And across the board in health care,
there has been an exodus and a shift out of health care to other
careers.
00:05:25:19 - 00:05:45:23
Michelle
How do you see then us utilizing Gen AI to augment and maybe bridge
some of those gaps with our shortages? The agent is great, but can
we maybe think about a couple of other things that maybe might be
even more magnificent? Like it's maybe front office, back office?
Is there opportunity there, you think? Because we don't think about
them a lot.
00:05:45:23 - 00:05:47:04
Michelle
I do believe, absolutely.
00:05:47:04 - 00:06:10:27
Tej
So let me just start. The clinician shortage is durable and it's
secular. This is something that I say over and over again. You
know, we have projected forward what we anticipate the workforce in
health care is going to look like. We know there's going to be a
shortage. But if you look at some of the research that's being
published now, just at the end of last year, there was a study that
published that looked at nurses and doctors that are in school
today.
00:06:10:27 - 00:06:43:15
Tej
And, you know, you might know that of those surveyed students,
between 20 and 30% of them said they're going to drop out. They
don't see the future of health care as a promising career that they
want to pursue. That's going to be fulfilling in the way that maybe
when we were going to school, we saw. And I think that's really
disheartening because truly I think that, you know, when the way I
thought about health care and the way the reason I pursued a degree
is because I wanted to take care of people.
00:06:43:18 - 00:07:05:18
Tej
And what's happened along the way is so many things get in the way
of that, Right. That joy that we were talking about that I
mentioned, that's gone because we've got so much administrative
burden, so many distractions that sort of pull us away from that
day to day care. It's actually started to deteriorate or continue
to deteriorate.
00:07:05:18 - 00:07:28:18
Tej
That relationship between the doctor and the patient that I think
is so sacrosanct. It's so important. And I think patients want it.
Physicians want it that nurses want it back. And so when I think of
front office and back office, it's not just agents that are going
to be doing this work. What it's going to do is it's going to take
those tasks away, but it's also going to start to transform.
00:07:28:25 - 00:07:55:11
Tej
What role, as a clinician, I play in the delivery of care, right,
where maybe 30, 40% of my time I was it was taken up by doing that
administrative work. Now I'm going to have it back to refocus on
the patient. And it doesn't just drive productivity improvement, it
drives greater engagement, it drives better experience for us as
clinicians and for patients.
00:07:55:13 - 00:08:19:10
Michelle
So back on Joy, I would love to be back there as well, cause I
remember that feeling my first, but the patients and I could do
anything and everything and this was going to be magnificent. And
then you're right. Little by little, the administrative stuff
started to become a burden. With GenAI,, how do you see us being
able then, in this world of technology, having unfortunately had
that adverse effect?
00:08:19:10 - 00:08:45:15
Michelle
And I think it was an unexpected consequence of all of the
information coming into the EHR and all of the information coming
at us, period, whether it's on our phones, on our on our laptops
and still faxes and message centers and inboxes. How do you see Gen
AI as being able to be the thing that we can get providers to say,
I trust this, I'm going to buy in, I will do this.
00:08:45:17 - 00:08:51:16
Michelle
Because the last time we petted the dog, that was technology, it
bit us and we have a long memory.
00:08:51:19 - 00:09:23:16
Tej
Yeah, you know, the EMR is a beautiful thing, right? It actually
transformed and improved the quality of care. It helped us, you
know, document how we were taking care of our patients and enabled
us to move patients seamlessly from one provider to another. But
you're absolutely right. When you look at the data around the
impact to productivity of the EMR, there's a 13% reduction in
clinician productivity because of it.
00:09:23:18 - 00:09:51:05
Tej
And I think, as you said, clinicians have been burned by it. I
think we feel as though we were betrayed. We feel as though, you
know, it didn't help us do our jobs better. It didn't help us
improve the relationships that we have with our patients. And I
think as we think about deploying new technology into care
delivery, we're going to have to do a much better job of
articulating the value proposition of what the technology is.
00:09:51:07 - 00:10:07:24
Tej
And we're going to need to pull in clinicians to think about where
it gets deployed, what part of the workflow should be addressed by
technology and what should it take over. I think there's a lot of
work for the industry to do. We're not good at it. We haven't
been.
00:10:07:27 - 00:10:11:29
Michelle
So how does Accenture then take away that hesitation?
00:10:12:02 - 00:10:35:28
Tej
You know, so we're in the early innings of what GenAI is going to
be able to do of where technology is going to be deployed. Truly, I
think the organizations that are taking the leap forward and
adopting and figuring out, you know, how technology is going to be
integrated into care delivery, are taking some really basic steps.
They need to have a policy on generative AI.
00:10:36:01 - 00:10:59:06
Tej
They need to have governance, they need to figure out what is their
data architecture and what's that foundation going to look like so
that they can access that data to enable their clinicians to
deliver care. They're going to need to think about on the back end
how they train their clinicians and, you know, when they actually
eventually do deploy technology, what is it going to look like?
00:10:59:07 - 00:11:03:26
Tej
How is it going to transform the way that they work, the way that
they deliver care.
00:11:03:29 - 00:11:25:28
Michelle
Wow, transforming the way that we do work? When you say that, I
think to myself, gee whiz, it would be marvelous. And in my mind, I
would love to be able to be the ER doc seeing the cardiac chest
pain patient and the intuitiveness of the system being that it
understands who I am, what I do, what's going on with my
patient.
00:11:26:00 - 00:11:42:25
Michelle
It then brings in from the EHR what I need to know then, but then
also helps me to go along with the best practice guidelines and
recommendations while not forgetting that other stuff that you were
speaking about before. With regard to referrals. Do you think that
that's possible?
00:11:42:27 - 00:12:14:07
Tej
So when I built my company, the fundamental premise that I was
going for is technology should always be in the background. It
should not be the reason that we get up in the morning and do
anything right. It should be the enabler, it should be the
assistant, it should be that third year medical student or third
year resident that just enables you to do your job and perform at
your peak all the time and think somewhere along the way, you know,
we didn't sort of catch on to that.
00:12:14:09 - 00:12:29:12
Tej
And I do think that with generative AI that we have the potential
to drive that type of experience for clinicians who really are
customers and clients of health systems. And we need to start
thinking about them that way.
00:12:29:15 - 00:12:52:13
Michelle
So I have two questions on that. We also spoke about the P word
productivity. And as physicians, especially ER, we know that that
can be something of a double-edged sword. So a lot of us want to
frame it, couch it, present it more so as efficiency and
effectiveness. How do we bring that message across with trust?
00:12:52:13 - 00:12:54:22
Michelle
Because there is trust to be rebuilt.
00:12:54:24 - 00:13:20:08
Tej
Yeah, you know, it's a great question. I agree that it's a
double-edged sword. I think, you know, we productivity is what the
organization cares about, but it's not necessarily what a clinician
cares about. You know, their schedules are full, they're overloaded
already. They're seeing more patients than they want to, and
they're stretched way beyond, you know, the schedules that they're
supposed to be working.
00:13:20:10 - 00:13:40:09
Tej
And it's a lot a large part of why we're seeing the burnout that
we're seeing amongst clinicians today. And so I think as we again,
as part of how we think about technology and where it gets deployed
and how it gets deployed, we have to bring the clinician along. And
I think we have to think about what are the messages that are going
to resonate.
00:13:40:09 - 00:14:04:12
Tej
Just like when you do any sort of change program at any
organization, they care about taking care of that patient and how
do we articulate it to them that this technology is going to enable
that, that this technology is going to power the experience that
they want? Again, going back to this idea of the clinician as a
customer or a client of a hospital system.
00:14:04:19 - 00:14:27:22
Tej
Right. Because 75% of clinicians are employed now. Right. And if we
think about them as customers that are, you know, they're producing
the outputs that we want, which is units of care, we've got to
figure out what resonates with them, what makes them tick, what's
going to get them up in the morning and help them use the
technologies that we're going to make available.
00:14:27:24 - 00:14:33:28
Michelle
What's going to get them back at the table to actually have that
conversation right, Because they need to be there. We need to be
there.
00:14:33:28 - 00:14:36:07
Michelle
So going back to burnout, burnout is not new.
00:14:36:07 - 00:14:52:20
Michelle
And the question I have for you there is with regard to burnout and
AI, do you think that there is meaningful use in there that we can
honestly really tackle that meaningfully and impactfully?
00:14:52:23 - 00:15:00:25
Michelle
Or do you believe it's going to take a long time, like decades
worth of time to get there? Big questions.
00:15:00:25 - 00:15:24:22
Tej
Yeah. And you know what, Michel? I'm not sure I have the perfect
answer to it. I think what I'd say is we have to try. I think this
technology has the promise to change how we practice, to create the
space for us to do the thing that we had intended to do and take
away all of that distraction.
00:15:24:25 - 00:15:58:27
Tej
But I think we also need to think about ways that the technology
can change the roles that we play as clinicians right. And you
know, how it can augment us in different ways or automate some of
the work that we're doing. And that's going to require real
analysis of workflows. It's going to require real, you know, work
and effort to rethink and rejigger the jobs for clinicians as a
whole.
00:15:59:00 - 00:16:05:22
Michelle
So there's a receptiveness there to an acceptance there that we are
going to have to change. And we're not known for being change
agents now,
00:16:05:22 - 00:16:31:23
Tej
There's a lot of fear. I think there's,I want to keep doing things
the way that I've done them, but I know that what I've been doing
is not working and I think it's going to be a very tricky
sort of journey for health systems and for us as we work with them
to help them navigate it, because every system is going to be
different and you know, it's going to depend on the culture.
00:16:31:23 - 00:16:53:09
Tej
But I think we need to get to and this will take a little bit of
time is for clinicians to start having the culture of being
accepting of change because you know it, I know it. I think
everyone that's going to be listening to this knows that this is
not the end. This is really the beginning of real
transformation.
00:16:53:11 - 00:17:10:01
Tej
And the change is going to come fast and furious. And what we need
to do is prepare our clinicians and organizations to know and
expect that and help them understand how to navigate that. And I
think that's going to be a long and durable journey.
00:17:10:04 - 00:17:12:10
Michelle
It will be. And you say fear.
00:17:12:10 - 00:17:30:08
Michelle
how do we start working to allay some of those fears and trying to
get to the point of balancing out even just being straight up
facing Gen AI as a functionality that has so much opportunity and
then proceeding to that lean in, how do we help people to do
that?
00:17:30:10 - 00:17:34:02
Tej
I think we're never going to take humanity out of health care.
00:17:34:02 - 00:17:43:12
Tej
Health care is about, you know, the clinician and the patient in
that relationship and navigating the health and staying and
remaining healthy.
00:17:43:12 - 00:17:52:10
Te
In the short term and in the medium term, we're in this transition
phase from technology being librarian to an advisor.
00:17:52:14 - 00:18:20:17
Tej
And it's going to be an advisor for a very long time. That
transition to agent is going to happen, you know, very quickly on
administrative work, but it won't become an agent on the clinical
side because I think we still need to use our judgment. That's what
we were trained for. We know what's right for that patient and
getting that advice, getting that support being augmented by
technology is not outside the realm of what should be possible
today.
00:18:20:22 - 00:18:30:02
Tej
It's available right now, but that role is going to need to change.
Right. What we do every day will change and will shift.
00:18:30:04 - 00:18:55:15
Michelle
I love, though, the vision of returning to being that empathetic
partner who helps to guide the patient through this journey back to
health or to maintaining wellness, that would be phenomenal as
opposed to always having people fall into gaps and into holes. And
with that, let's talk a little bit about what GenAI may or may not
be able to do with regard to access and health equity.
00:18:55:17 - 00:18:57:07
Michelle
Let's talk a little bit about that.
00:18:57:07 - 00:19:25:08
Tej
Yeah, You know, the shortage of clinicians that we're seeing around
the world, it's not just the U.S., right. The projection is 10
million shortage over the next decade. And that disproportionately
affects patients who don't have the means to access care. As an
emergency medicine doctor, you appreciate the number of patients
that come in as a site of last resort because they have no place
else to go.
00:19:25:11 - 00:19:52:20
Tej
It actually is not very convenient because the wait times are just
getting longer and longer. And so what we're seeing is the
disproportionate impact of the clinician shortage on populations
that, you know, have not historically had. The means to access
care. And so what I think and what I believe is with technology, as
we scale capacity and we retrain clinicians on what their role is
going to be, we're going to create access.
00:19:52:20 - 00:20:03:28
Tej
We're going to start to address the health equity challenges and
barriers and inequities, really that that we see as a society, not
just in the U.S., but everywhere.
00:20:04:00 - 00:20:06:05
Michelle
Beautiful, beautiful.
00:20:06:05 - 00:20:25:16
Michelle
How do we introduce GenAI to patients and have them accept that
sometimes they're not going to be able to talk to the nurse and ask
the nurse when the pharmacy is open and understand why that's
important and how this technology is also intended to serve them
and to help them to do better.
00:20:25:16 - 00:20:43:19
Tej
there's really two things that are really important related to AI
and how we make it available to patients. One is around having, you
know, a framework around responsible AI. And when we make the
technology available to patients and for what use cases, because
you're not going to it's not general availability for everyone.
00:20:43:21 - 00:21:01:28
Tej
And then the second thing is, you know, we have to understand how
we segment the patients that we serve, because your 91-year-old
aunt may never want to interact with technology in the way that I
do or that a 25 year old does. And I think we have to respect that
and we have to appreciate it.
00:21:02:00 - 00:21:32:17
Michelle
So we've hit now patient, we've hit provider, let's hit the
hospital organization. They’re the decision makers ultimately as to
what comes in, what gets implemented, what gets discarded or what
what's put on by the backlog or the back burner or how do you see
the conversation going with hospital decision makers about going
forward with a roadmap that includes AI?
00:21:32:17 - 00:21:49:03
Michelle
Because a lot of the discussions that I'm having are we want to see
somebody else dip their toe in, we want to see somebody else dive
in and survive. How do you see that conversation going or how would
maybe you approach that conversation with someone who is a bit
hesitant and wants to stand back a little bit longer?
00:21:49:03 - 00:22:09:01
Tej
the truth is there is not a choice that health systems are going to
have to make the investment in AI and GenAI, because we're all
facing, you know, really critical workforce shortages. And we need
to create capacity right now and we're going to need to do it in a
durable, you know, long term way.
00:22:09:04 - 00:22:29:19
Tej
The truth is there's a billion different applications of GenAI and,
you know, we don't need to explore all of them all at once. Right
now. What we need to do is very simple things around the
investments in AI. You know, we have to have a responsible
framework and governance and a policy for how we're going to deploy
it.
00:22:29:21 - 00:22:57:07
Tej
What's the oversight going to look like? We need to have that
technology foundation and the data foundation to make sure that
we're really leveraging all the data that we have. And truly we
have to think about how we train clinicians and the rest of our
workforce to adopt the technology so that when we get to scale,
it's going to be used, it's going to create and have the type of
impact and the ROI that we want.
00:22:57:09 - 00:23:17:26
Tej
I guess the last thing that I would say is right now, GenAI
touches so many different parts of a health care organization. It's
not just the chief medical officer, it's not the chief nursing
officer, it's not the Chief HR Officer It's not the Chief
Operating Officer. It's not the CFO, it's not the CIO, it's not the
CTO.
00:23:17:28 - 00:23:43:23
Tej
And really figuring out in the context of, you know, an industry
wide sort of, you know, crisis around margins, right? Where margins
are in the low single digits, you know, and negative for more than
half of the health care systems in the U.S.. How do you make this
investment in the context of where a lot of other things are
burning?
00:23:43:26 - 00:23:59:11
Tej
How do you prioritize it and who do you go to and who makes the
decision? Who who's going to be accountable? Who's going to drive
the transformation? Who's going to drive the training? Who's going
to drive the change? I think health systems are trying to figure
all of that out and they need help.
00:23:59:14 - 00:24:28:23
Michelle
They do. They do. They're going to need a partner who's going to be
able to sit with them and understand where they are. More so,
though, a lot of other discussions I'm having are people want to
just kind of throw air out there as the panacea and the magic wand
to fix everything. And there's a real struggle to contain it and
say, let's pick one goal, something that is small yet impactful,
that doesn't derail the ability to do care, doesn't disrupt your
whole organization.
00:24:28:25 - 00:25:00:21
Michelle
And those are some of the most impactful and beautiful
conversations. And then when you get it right and they have a win
under their belt, maybe tomorrow, I in the way of provider
satisfaction, retention, patient engagement is better. Those sorts
of discussions and those sorts of wins are huge. So I think that it
is incumbent upon us and our positions and others out listening to
this and similar positions to be that selfless advocate slash
partner to your customer and help them along this path because it
is complicated.
00:25:00:21 - 00:25:16:00
Michelle
Let's talk a little bit then about health care, utilizing
technology to catch up with other industries, because other
industries are fairly far ahead of us when it comes to tech and
efficiency and productivity, using technology.
00:25:16:00 - 00:25:41:12
Tej
A lot of us like to say that if you want to know what's coming next
in health care, look at what happened in other industries 20 years
ago. What's incredible to me is that generative AI is being
contemplated and probably more advanced in health care than it is
in several other industries today. And it's almost moving at
pace.
00:25:41:15 - 00:26:03:14
Tej
And I'm not really sure exactly why that's happening. I think that,
you know, maybe it's the shortage, maybe it's the margin pressure,
maybe it's the potential, maybe it's, you know, some of these
thought leaders that are really expounding on its use and its
application, its potential in health care. Maybe it's just the
burnout that's happening amongst clinicians.
00:26:03:14 - 00:26:17:14
Tej
None of my clients on the provider side can afford to wait for the
perfect sort of, you know, what are all the use cases and how do I
sequence them and everything that I'm going to do for as you said,
for the next five or ten years?
00:26:17:17 - 00:26:52:29
Tej
I think we have to start now and there's a place to start right
now. There are things to do to prepare yourself for that future and
I think we're going to be astounded by the creativity of the
clinicians that that you and I work with and that work for these
organizations on how the technology can be deployed. And I think we
have to just be open to it because I think we're going to see the
potential and, you know, sitting behind a desk or sitting in a lab,
you know, looking at the technology and thinking of use cases,
that's going to get us only so far. We have to put it in the hands
of
00:26:52:29 - 00:27:04:27
Tej
clinicians having the appropriate sort of governance and
guardrails. But that's how we're going to learn what it can do and
how it can help and how it's going to impact, you know, how we
deliver care.
00:27:04:29 - 00:27:09:21
Michelle
That sounds like a great call to action. ‘Come to the table, your
seats waiting’. Let's do this.
00:27:09:21 - 00:27:13:05
Tej
Yeah, let's do it. I mean, ER doc, right? Like, let's go.
00:27:13:12 - 00:27:30:15
Michelle
I love it. I love it. Well, thanks for your time today, Tej. Thanks
to our listeners. Be sure to subscribe to the Perspectives on
Health and Tech podcast for more insights from industry experts,
visit Oracle Dotcom Slash Health or follow Oracle Health on social
media.