September 18, 2024 • 54 mins
We sit down with Dr. Will Reedy, who leads health for Accenture NZ, to discuss the transformative potential of AI in healthcare.
Reedy explains how generative AI can enhance NZ’s health system, boosting both frontline and back-office operations and addressing workforce shortages.
He also raises the ever-looming spectre of the need to truly aggregate resident's health data, a must-have if we want to realise the full potential of generative AI.
Plus, an interview with Jack Pugh, the NZ winner of the James Dyson Award for design, about his winning medtech device.
The Business of Tech is sponsored by 2degrees for Business.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Our health sector is in trouble, short staffed, under resourced,
overwhelmed with patients who have complex and expensive needs.
Speaker 2 (00:08):
Our hospitals are among the least progressive in the Western
world when it comes to digital health. Just as artificial
intelligence shows potential to cut down ADMIN that sucks the
life out of our doctors and nurses.
Speaker 3 (00:23):
So what can AI do for our health system.
Speaker 2 (00:25):
And what are the barriers that are preventing its uptake?
Speaker 1 (00:29):
On the Business Attack sponsored by two degrees Business this week,
AI and how it could enhance public health if we
can get our digital health house in order.
Speaker 4 (00:38):
When you talk to healthcare organizations across New Zealand at
the moment and just ask the question of the board
or the teams or even their chief Data and Digital officer,
do you have an AI strategy? The answer is no.
Speaker 2 (00:50):
Doctor Will Reedy is on the EXTENTSI of New Zealand
leadership team and part of the consulting firm's global health team. Now, Ben,
your interview with will really get me one of the
best overviews of where we're at with AI in health
space in New Zealand. So everyone stick around for that.
Speaker 1 (01:07):
First though and keeping the health theme going. The Dyson
Awards for Design Excellence are underway now, annual global awards
where British inventor Sir James Dyson searches the world for
the best young designers. The New Zealand winner has just
been announced and Peter, I believe you were one of
the judges.
Speaker 2 (01:25):
Yeah, they got me on just to run the ruler
over it in terms of is this something that's going
to connect with a big audience. We had health experts
and other design experts. One of the top designers from
Dyson was a judge as well. This was the first
time I judged it, so it was really good. But
the one that won was one of the most simple
(01:47):
ones really, but was a design that we thought actually
had a really addressable market. No one else was doing
this and it's called the snapcap and a very simple
device that really helps frontline health professionals deal with the
containers that medicines come in, glass containers that need to
(02:11):
be dismantled quickly on the frontline inwards and put in
a sharpy bin to get rid of this stuff. You
need a device to pull these things apart and break
them open, and amazingly there wasn't one in existence.
Speaker 1 (02:27):
Yeah, it's a really cool looking device. It looks kind
of like our twenty first century bottle opener.
Speaker 3 (02:32):
Very simple in.
Speaker 1 (02:33):
Design, very esthetically pleasing in design, and obviously much needed
because I think in your interview you talk about the
fact that frontline health professionals actually do get injuries from
trying to open these little glass bottles.
Speaker 2 (02:49):
Yeah, they do, and it's just a reminder of how
much our frontline health workforce have to do all sorts
of little jobs you don't even think of that is
up to them. So if we can make their lives
easier and hopefully do it at a cheap price, that's great.
Speaker 4 (03:08):
You know.
Speaker 2 (03:08):
This was just one of about a dozen designs that
we looked at. Some of the other stuff that was
sort of like an exoskeleton that would replace a moon boot.
There was insuls that can be customizable and three D
printed for people who have problems with their feet if
they've got potentially diabetes or something like that. There was
(03:29):
a little sense so that you can put on the
back of your shoulder if you're out jogging on a street.
It will give you haptic feedback if it thinks that
a car is getting too close to you or is
potentially going to take you out, so great to see
that creativity.
Speaker 1 (03:42):
I love the simplicity of the winning design, and listening
to your conversation with the designer, Jack Pew gave me
some really good insight into kind of how he got
there and what he was thinking and the balance of
talent and pragmatism that led him there. So let's have
a listen to that interview.
Speaker 2 (03:59):
Now, Jack Pugh, Welcome to the Business of Tech, and
congratulations on winning the New Zealand Dyson Awards. You were
crowned the best design off about a dozen. I was
under the actual judging panel, so I saw all of
these dozen or so designs. Tell us a little bit
(04:22):
about yourself, Jack, Where are you from? How did you
get into design?
Speaker 5 (04:26):
Thanks Peter. So, I'm from christ Church, originally going to
sort of come up to the Capital to study at Messy.
I've always been quite interested in design when I was little.
Any when you ask would tell you that I always
wanted to be an inventor. So it's really cool to
be able to sort of do this and have some
recognition for it.
Speaker 2 (04:44):
You've won this award. Take us through your winning design.
It's called the cap Snap what did you set out
to achieve with the cap snap and what is it?
Speaker 5 (04:53):
Well, it's a simple tool for a simple problem in
the theory, a medical ball opener for health professionals working
with medications where they'd look to recycle them by taking
off the aluminium crimpsy your caps, or to open these
little glass vials which old medication called ampules. There's some
safety risks associated with both, and so this tool kind
(05:15):
of lets them do it in a real quick and
easy way.
Speaker 2 (05:18):
Yeah, it looks like a bottle opener, and essentially it
is a bottle opener. Yeah, functionally, because you know, just
to try and visualize it, and unfortunately, I've got a
lot of experience of this now visiting sick relatives in hospital.
You'll see a medicine bottle. It's got some in this
in the case, i know, immunotherapy liquid drug in it.
(05:41):
It's a glass bottle, it's got an aluminium cap which
is snug on it. You've got to separate those two
things to recycle these materials. You also have, as you say, amples,
these little sort of sealed glass containers that you have
to crack open literally to get the liquid out of them.
But they're really damn fiddly. And what I loved about
(06:03):
your design. You've got one little device that probably doesn't
cost that much to make, and it does both. It
removes that aluminum casing from the top of the glass
bottle and you can insert the ampule into the bottom
of it and crack it in half. So the big
benefit I saw from it is cutting down on the
potential for nurses and doctors to hurt themselves going through
(06:26):
that process cutting themselves on glass or on aluminium or
a tool trying to take that off. It does both
of those purposes.
Speaker 5 (06:34):
Yeah, that's right. The ways that people are currently getting
around both those issues is they're using kind of makeshift
solutions alcohol pads to step ampules, so at least if
it does shedd in an unpredictable way, the glass will
go into the air instead. And with those aluminium's caps,
they're real fiddly. If you ever see someone tri tech
one off, they don't have good tools suited for it,
(06:57):
so they're using kind of four steps and kind of
straining their hands to be able to pull them off.
Speaker 2 (07:01):
Yeah, it's dangerous. And you know what I love is
the simplicity of this. Take us through your your design
approach to this. This is a very specialized piece of equipment,
So what work did you go through sort of talking
to frontline health practitioners to inform the design of cap snap.
Speaker 5 (07:23):
Kind of things started off just kind of as a
conversation on what are the issues, kind of what are
they doing at the moment, and then from there tried
to think of kind of a range of solutions. What
would something handheld and quick and easy look like. First,
maybe something mounted to the wall, or like a mix
of both, something that's kind of a bit of a hybrid.
And I kind of drew up some ideas and sat
(07:44):
down with a member of the team and we kind
of talked through some of the ideas and we ended
up deciding that something that was real quick and small
and portable to be able to be moved around if necessary,
would be the kind of the best way forward. And
then from there I kind of springboard into a bit
of antive process, found some studies overseas looking at similar things.
(08:08):
We kind of developed into well, it's a bottle with
a cap, how about a botopner. We'll keep it super
super easy. And then some of the issues that came
up with one size kind of we couldn't have a
one size fits all approach to have it be real functional,
So we looked at what if we had a bollopner
that could sort of adjust and through that kind of
(08:28):
linear motion, we could incorporate that really seamlessly into the
ampule snapping function.
Speaker 2 (08:34):
It's interesting this was not, by any means the most
elaborate or sophisticated design that we came across the judging
panel for the dice and awards. There were all sorts
of quite elaborate designs that were probably required a lot
more design forinesse. But what I loved about this one
(08:56):
is that when I did research about it, no one
was really doing this. Why do you think no one
in the health profession has come up with something like this?
Speaker 5 (09:03):
Yet you get these things like the ampules or the
crimseal caps, and then the job has to be done,
so you come up with a workaround and you kind
of that's just what you do. And then in a
good and bad way, people don't make a lot of
noise about it. And so we've got all this potential
for interesting design solutions or just simple little fixes for
(09:25):
these health staff we're having to do all these workarounds
on a kind of a daily basis that just sort
of someone needs to have a look at and try
to come up with something.
Speaker 2 (09:33):
And that's exactly what you've done. You know, Sir James
Dyson is famous for being a real iterative designer. Did
you go through numerous iterations and either future obvious ones
you see for this device to make it even better?
Speaker 5 (09:48):
Yeah, So a real fun part of the process is
going through the iterations. So we went through a few
different iterations looking at how I can keep it as
simple as possible and reduce kind of them the number
of mechanical parts in it. There are kind of two approaches.
One kind of looked like a claw, if that makes sense,
and that would allow it to be able to open
a bunch of different sizes without having to use that
(10:11):
slighter function. But the issue we kind of ran into
with some of those prototypes is that it wasn't sort
of snappy and intuitive. If you looked at it, you
wouldn't know what you're looking at, and that was a
real big part of the design is making it just
super straightforward that you can look at it and pretty
much figure out how to use it without any instruction.
There's so little time to teach people about new things
(10:34):
when a new tool comes in kind of in these busy,
bustling environments that that's kind of what you want from
a tool that's going to do just this quick and
simple job.
Speaker 2 (10:41):
Yeah, Because it literally looks like a bottle opener, so
instantly you have that recognition. You go, oh, I've got
a bottle here, This must go around the neck. And
then you've got the sort of the hidden compartment on
the bottom of it, which is for breaking the ample
as well. So I guess it's a bit of education
required so that people know to use that. You're sitting
in your lab as we talk here, you've got a
three D printer behind you. Was that useful? Do you
(11:04):
do a lot of iterations and design work and prototyping
using three D printing?
Speaker 5 (11:09):
Yeah? So three D printing was a huge enabler for
this project. That and water jet cutting was another super
useful part in sort of smashing out these prototypes. I
tried to kind of reduce the amount of plastic that
I used where I could, so I tried to make
my prototypes I can kind of hot swap between the
different inserts to try the different geometries to see what
(11:32):
was kind of the best and most effective fit when
kind of figuring out some of the measurements. So yeah,
definitely super pivotal and being able to quickly run through ideas.
Speaker 2 (11:42):
So design was only sort of part of the criteria
for winning this award. A big part of it was
the real world application and the potential for this to
actually go on and be used. It's all well and
good to design something that just never leaves the labor
with the prototyping stage, and we did see some sort
of designs like that. What are the next steps for you?
(12:05):
Is this something that you'd like to pursue potentially as
a business that the Snapcap try and get it out
there into hospitals and clinics around the world.
Speaker 5 (12:12):
I would like to see it in people's hands and
just making their lives a little bit easier for those tasks.
So I'm able to put it a little bit of
time in to be able to work up the design
a little bit further. I've got some help from some
awesome people at a couple of the other hospitals around
the country been able to kind of cast the net
a little bit wider see how the same issues are
(12:36):
sort of getting received across those different sites. We're hoping
to be able to work it into something that we
can get into people's hands.
Speaker 2 (12:43):
The judges we're talking about this, we sort of thought
this is great, and you submitted a video literally off
a nurse who was I think snapping an ampool who
actually cut herself.
Speaker 5 (12:55):
It wasn't scripted, but we had to use that tape.
Speaker 2 (12:58):
Well, that illustrates it very well. But we were thinking,
why don't you just I mean, you need to break
the ampull, but in terms of taking the aluminum cap
off the bottle, why don't you just chuck it all
in a bin and automate that later.
Speaker 5 (13:12):
That was an idea that we kind of tossed around
at the start. But the kind of reality of kind
of this product is that it'd be all well and
good to make something real, awesome and automated, and I
actually going into this project that's kind of what I
was going to look to do. But then talking with
the people and like hearing about the issues and then
(13:33):
learning about the ampuel side of things, we've sort of
determined that while technically it could be considered a band
aid solution, it would be on the fastest track to
be able to have this issue kind of be sold.
Speaker 2 (13:45):
That's what we loved about it, the simplicity of it.
I mean, presumably this, if you get this into production,
it wouldn't be a super expensive device to create either.
Speaker 5 (13:54):
Yeah, we're hoping to well, we're working to get the
part count down as much as possible. Engineers in christ
Church I've had some great input on on clever ways
to really simplify that mechanism to even just three or
four parts.
Speaker 3 (14:09):
Tops.
Speaker 2 (14:09):
Where are you working at the moment?
Speaker 5 (14:11):
So I'm working out of Wellington Regional Hospital, part of
the Futtow Water Improvement team based out of here, so
day to day I'm sort of looking at other issues,
but it's still we're able to work on some innovative
solutions sort of within the hospital, which is a really
exciting sort of opportunity.
Speaker 2 (14:28):
Well, brilliant, you're in exactly the right place and they're
lucky to have this design brain on the team there.
So good luck for the next phase. And he's hoping
we see snapcap in hospital wards before too long.
Speaker 1 (14:41):
Well hat fingers crossed, great young talent, Jack Pugh. Great
to have those kinds of people coming through New Zealand,
so big well done to him and look forward to
seeing what he does next. So that's innovation and health hardware,
and our topic of focus this week is AI in healthcare.
(15:02):
Doctor Will Reedy is one of the country's leading experts
in digital health. He joined Accentua around a year ago
and is still a practicing doctor one day a week
in the county's Monaco area.
Speaker 2 (15:13):
He's a lot of experience helping with the rollout of
digital health systems all over the world, so he's pretty
well placed to compare and contrast our preparedness and progress
in the digital health space compared to the likes of Europe,
the US and Australia, where he's also worked.
Speaker 3 (15:30):
Will's also really interested in the potential to reduce health
inequities using digital health tech, that is, if we can
trust machine learning and large language models to get it right.
Speaker 2 (15:40):
So here's Ben's interview with Accentures doctor Will Reedy.
Speaker 3 (15:50):
Good, Hey, Will, how are you great? Thanks?
Speaker 1 (15:52):
Ben good, Welcome to the Business of Tech podcast. Thanks
for joining us. As so, why don't we start with
if you could just give us a very quick summary of.
Speaker 3 (16:00):
Who you are, what you do, and a little bit
about your background.
Speaker 4 (16:03):
So yeah, look, it's pleasure to be chatting with you
this morning. So I guess my full time job is
working for Accenture, is their how order health and wellness
lead here in New Zealand with the goal to help
the health system transform given some of the real challenges
that many of us aware of at the moment in
the health system. And then one day a week I
do a shift in the surgical services at Middlemore so
(16:26):
that can be in the emergency department and the clinics,
in the wards and sometimes in theater. So I guess
that keeps it real in terms of understanding the pulse
of the health system at the clinical front line. But
I guess ultimately the passion is to try and transform
health systems with technology and things like AI.
Speaker 1 (16:42):
It must be quite start going from in Accenture talking
about the latest and greatest worldwide AI this, and then
you go into a hospital and very different story.
Speaker 3 (16:53):
I'd imagine, Yeah, it is.
Speaker 4 (16:54):
It's really interesting and it's generally accepted that in our
hospital system, not so much in our GP systems or
primary care systems. We are the least digitized hospitals now
in the developed world. So it is a little bit
of a change in terms of the global work that
I do and seeing what's going on in terms of
transformation and how far some other countries are ahead of us.
(17:16):
Also being quite digital my day job at Accenture and
then you know, having to use a pen and paper
on Fridays pretty much.
Speaker 3 (17:23):
Yeah.
Speaker 1 (17:24):
But also I guess conversely, that also just helps you
to understand the potential for change, right, and that actually
how far we can go when we start to implement
modern tech.
Speaker 4 (17:34):
Yeah. I think it's easy to kind of look at
the I guess, the lower level of digitization and tech
enable transformation in the New Zealand health system today. But
it's also an opportunity, and my words are an opportunity
to kind of leap frog some of the approaches and
the thinking. So I do I'm quite optimistic about some
of those opportunities for New Zealand just to go, hey,
(17:54):
where are we trying to get to? What would we
do differently? Could we leap frog ahead of some of
the other jurisdictions.
Speaker 1 (18:00):
Yeah, it's interesting you use that term leap frog. That
seems to be kind of a relatively common New Zealand
experience where we kind of fall behind a little bit
and then we go, oh, let's catch up, and in
doing so we kind of go ahead and really hit
that cutting edge again. Is that kind of what you're
seeing happening at the moment?
Speaker 4 (18:17):
I guess, I see the opportunity and it's really interesting
to share with you. It's funny how some of the
AI technologies are becoming quite pervasive in healthcare. And a
colleague of mine has been leading the way in gp
Land all it or primary care in driving the adoption
of a product called Nabler, which allows you, with the
(18:37):
permission of the patient, to record the voice around the
interaction with the patient and then convert that to text
and then put it into the GP system. And our
GPS I think about ranked about third in the world
in terms of their levels of digitization. And that's been
kind of the same kind of I guess measurement that's
been in place for about twenty years. So primary care
in New Zealand's actually been I guess at the forefront
(18:59):
of tech for some time. And then you know, what
he's seeing is first and foremost the patients and they're
farno who are in the room with him, are seeing
some benefits around more eye contact not turning around and
typing stuff in the computer and in terms of the
fifteen minute consult that is common in primary care. The
other side that he's finding really interesting is efficiencies within
(19:20):
the consult around typing everything down because we often paraphrase
what the family or the patients say to us given
the time limit. But what he's also finding, as our
colleagues across about one dred and one hundred and fifty
practices in New Zealand, is that the cognitive load, that
the mental load is a lot less because you're having
to recite and think things through. So piece of technology
(19:41):
came along relatively recently and it's been adopted in about
three hundred practices across New Zealand, which is about fifteen
percent of the GP practices. So it's really interesting around
that leap frog concept. Grab an idea and go with it. Yep.
Speaker 1 (19:56):
It is an interesting idea and it is one that
has you know, it could have been earlier probably, but
the reality is is that the technology that we have
now around AI has just made it super accessible.
Speaker 3 (20:08):
Are there other.
Speaker 1 (20:08):
Areas that you're seeing that trend that the modern AI
tools that we have now in the last couple of
years have made things possible suddenly that would have seemed
really onerous in the past.
Speaker 4 (20:19):
Yeah, look, it's interesting. I'll probably give some broader contexts.
So one of the challenges in the health system at
the moment is how do you adopt AI? And I
guess people translate AI to generative AI at the moment,
so it's just careful to be specific about that. And
so the adoption piece is and I'll get down to
some of the use cases shortly but effectively, when you
(20:40):
talk to healthcare organizations across New Zealand at the moment,
and just ask the question of the board or their
exec teams or even their chief Data and Digital officer,
do you have an AI strategy? And the answer is no,
which is really interesting. And then the second question you
ask in the context of adoption of AI, if you
went down that path, is do you have a policy
(21:01):
around how you'll adopt AI as an organization? And generally
speaking not many have that either. But then you go, okay,
put that to one size. You don't have a strategy,
you don't have a policy. What are the core use
cases that you've been thinking about that you'd like to
get into your organization? In the next twelve months and
again that's where it opens up interesting conversations. And so
(21:21):
when we ran a leadership summit earlier in the year
with the Chief Medical Officer for Health New Zealand and
the head of the AI Advisory Group, the most common
use case was surfacing genitive AI experiences to patients or
family or FANO, so things like education CHAP as an example,
I've just been diagnosed with diabetes. What can I expect
and how do you get repeatable advice to patients where
(21:43):
doctor or nurse isn't available as an example.
Speaker 1 (21:45):
That's kind of very similar to the recent announcement around
GOVGBT right where it's the ingesting a bunch of government documents,
government pages and then being able to chatbot style ask
questions and get information about this governments.
Speaker 4 (22:00):
And I think I think in New Zealand, you know,
the health systems around the world, particularly in socialized health
systems like New Zealand, Australia in the UK, is everything's
quite fragmented. So if you've got a consistent education piece
experience for patients where no matter what question they ask,
they are a consistent answer, it's actually a big benefit
in terms of patients being empowered to manage a chronic
(22:24):
I guess condition like diabetes. So yeah, we are quite
surprised that everybody is going, how do we surface stuff
to patients? Which is really interesting, So that's open up
a new world. The second kind of area, broadly was
what we call the clinical or the front line or
the front of office, and so lots of use cases
around voice detext and reducing the burden of me using
(22:45):
pen and paper as an example, some benefits around managing inboxes,
around lab results coming in because a lot of lab
tests that we order for patients, we're kind of trying
to rule something out and we want it to be
if it comes back normal, then we don't really need
to process that forget where I'm coming from. And then
the back of office piece around workforce management and finance,
(23:06):
procurement and supply chain. So those are the broad areas.
So the reason I explain those broad areas around use
cases is people are understanding there is potential to apply
in particular generative AI to those use cases. It's just
where are they going to get the biggest impact around
healthcare in New Zealand. So I guess your question was, Hey,
are people's eyes being opened up. Yes, they are because
(23:28):
they're learning around use cases against offshore at the moment
and going, hey, that could easily be applied here in
the New Zealand context.
Speaker 1 (23:35):
Yeah, sticking with the generative AI theme, you know, I've
often looked at the health of Fire website as it's
now called, and just that that's such a huge corpus
of information and data, like it seems like a great
opportunity for something like an educational chatbot, But there is
a lot of risk with that. Where chatbots are known
to want to please people, they're known to kind of
(23:57):
sometimes make things up if they're relying heavily on those
LLAM models in the background. How are people in the
health sector thinking about those risks considering the sensitivity.
Speaker 4 (24:08):
Yeah, it's a really good question. I think the interesting
context is do healthcare organizations need to adopt genai or not?
And the general trend overseas is they're all adopting it
to see what the potential is, but not necessarily to
roll it out at scale. And the reason they're doing
(24:29):
that is it a competitive advantage? Does it deliver a
better service? Those types of things. So one of the
things I've seen In New Zealand. We have six and
a half thousand healthcare organizations, of which one is Health
New Zealand. It is the biggest, but that's the wider context.
So organizations need to think about AI from a responsible perspective.
And the biggest concern slash barrier is exactly what you
(24:51):
just articulated, which is hallucinations. I think the soft words
are unreliable outputs, and so I think a lot of
that has to be un understood. Choosing our use case,
learning about the use case, having the governance and leadership
in place to go. Actually, we've tried this in a
small use case. We have actually looked at what's happening
(25:12):
overseas and these use cases do actually add value, but
you'd got to go on the journey around I guess
the hallucination side of things. The other thing that's interesting
in New Zealand, in Health New Zealand as an example
of starting to do it, is they need to get
all their data in one place to run the GENAI
lms across the top of it. And so there's definitely
one that's the strategy for Health New Zealand, and they've
(25:34):
started investing in something called National data platform, so they
can put their data all in one place, whether it's
clinical data, workforce data, finance data, and then they can
start training the GENAI tools on top of their own
data and then managing that piece. You still have hallucinations,
don't get me wrong around data quality, but at least
you're doing it on the data that you have governance
(25:55):
over a couple of things to share with you. In
terms of New Zealand's text for GENAI adoption, mass of
concerns around liability. So if I have run a GENAI
tool and I've surfaced it to a patient and it's
given them some advice and something doesn't go as well
as it would have liked, and there's more risk in
terms of patient care, who's liable Is it the clinician
(26:18):
or is it the GENAI And how does it actually work?
So he's a little bit of maturity around what I
would call regulation and compliance to think through, because at
the end of the day, most healthyic organizations have a
clinical obligation around the safety of the care that they provide.
And if you introduce AI alongside the clinician, how does
(26:38):
it actually work and what are the implications? And the
last year that's really interesting is privacy and security of
the data. So you might bring it all in, but
you know, how do you actually control it in the
world of more cyber attacks, particularly in health systems around
the world. So those are kind of the core barriers.
But the number one is obviously the hallucinations piece.
Speaker 1 (26:55):
Yeah, so you know, I guess part of that is
having the confidence to experiment and trial and go at it,
but also having the confidence to say, actually, in this case,
the risks are too high, the technology is not there
yet or may not be, and so we're going to
choose to not make this a customer facing or patient
facing thing. Keep it for doctors or healthcare providers and
(27:17):
they can be the intermediary between the chatbot that's getting
a lot of information and the patient at the other end.
Is that kind of how thinking is going, Yeah.
Speaker 5 (27:27):
It is.
Speaker 4 (27:27):
It's an interesting piece because if you're wanting to experience
an experiment and dip your toes and genitive AI, I
think that the two pieces of conversation we're having in
New Zealand at the moment is so, what are the
use cases at a gathering momentum offshore and generally speaking
in the healthcare context using GENAI to re platform, recode
(27:48):
old applications, and New Zealand has a problem at the
moment around that we've got a lot of legacy applications
being used, particularly in the public health system. The second
area is around contact center experience, and the third one
is that voice to text that I talk about. So
those are the three broad areas. The other thing that
we've packed up offshore, which is interesting to share with you,
is generally speaking, most of the GENAI use cases have
(28:10):
been done on top of platforms. They're leveraging Salesforce, Microsoft,
our electronic medical record platforms where genitive AI is being
built in as a feature if you like, in these platforms.
So the hallucinations piece isn't as big a risk because
it's built into a platform. It's well tested, So I
think it's a fine balance. But it's just interesting to
(28:31):
see what the trends are in terms of the practical
elements of gen AI, the hype versus in reality what's
going on.
Speaker 1 (28:38):
Yeah, absolutely, that's kind of a lot about primary care,
and we talked about how that's really advanced. What about
in the hospital world.
Speaker 4 (28:47):
Yeah, Again, it's an interesting discussion around looking what happens
off sure, and I think there's a couple of contextual things.
So as a practicing clinician, when I think about technology
like genitive AI, I think of it as another tool
in my clinical practice, like over stethoscope around my neck.
Most of the younger generation clinicians, I'm I'm a I
guess I called a veteran these days because they've been
(29:09):
around the health system for twenty five years. But if
I look at some of the newly trained doctors and nurses,
they all expect if you like, generative AI to be
available for some of the use cases around again that
voice detext piece, and we haven't done anything in New
Zealand at the moment, but in Australia, the first use
(29:29):
case is using voice to text and busy theaters to
drive through throughput so you can hit health targets around,
you know, like the waiting lists for hypophens and things
like that. So there is a willingness for in that
surgical operating theater for surgeons, anetheists, theater nurses to use
those technologies to be more efficient in terms of through
(29:50):
put through the theater because they're not writing notes after
an operation. It's been done real time, so yes, there
is potential, just not quite happening at the moment, and
I guess that's the clinical frontline piece. Really interesting report
that we did with Microsoft recently for New Zealand looking
at how genitive AI could be applied to the nursing
workforce and based on using some of the Microsoft technologies
(30:14):
voice to texts predictive analytics around patients becoming unwell because
they have temperatures gone up or their blood pressures dropped.
They said, if you put on these common tools, you'd
get a productivity increase of around nine to ten percent
for each nurse across the country. So that's proven offshore,
done a little bit of analysis around how I guess
nursing workforce works the New Zealand today and our public hospitals.
(30:36):
You apply these technologies and process improvement augmentation around genitive
AI would actually lift productivity ten percent, which I guess
you know eight to twelve hour shifts is you know
one to one and a half hours, which does make
a big difference.
Speaker 1 (30:50):
It does, especially if you're looking at you know, a
shortage and if ten percent is saying you have the
equivalent to ten nurses that have nine nurses on shift.
Speaker 3 (30:58):
That actually does make a difference end of the day.
Speaker 4 (31:00):
It does, and you know, we've got to you know,
as you've heard in the media, we've got to adopt
a shortage of nursing shortage and in some areas what
I would call an allied health professional shortage BUZZIO is
occupational therapists. The biggest year actually at the moment is
anesthetic technicians. So that's actually preventing some of the through
put in private and public hospitals where there's just enough
(31:20):
of anesthetic technicians while you're sleep to look after you.
It's really interesting at the moment.
Speaker 1 (31:25):
What about the casting forward into the future. Do you
see it, as you know, potentially being a kind of
first point of contact for patients to be able to
say instead of just pushing a button waiting for a
nurse pushing a button and saying, oh, I think my
leg's really hurting and I'm not sure, and so then
that can automatically go into like a database and be
(31:46):
triaged and get the nurses to who needs to. Like,
it strikes me there is a fair bit of potential
in that first line of support for generative AI in
the future.
Speaker 4 (31:56):
Yeah, look, I totally agree with you. So I think
technologies like unit of AI we often talk about in
the current health system of New Zealand, we've got to
transform it generally. That means you spend in the next
five years trying to change the reality is we need
to ask New Zealanders, you know, what's your experience like
in terms of their most recent interaction with a healthcare
professional or a period of care and hospital. So my
(32:19):
view is first and foremost is that's where things are going.
I guess it's probably called consumerism and what do people
expect in terms of their health experience and how do
they compare it to other experiences like banking just to
use it? And I guess at parallel so I do
think one our customers, patients like you and I, will
expect a better level of experience and driven by digital
(32:42):
touch points, of which GENAI will power some of them.
Speaker 3 (32:44):
It's number one.
Speaker 4 (32:45):
The next question is do New Zealanders want it? And
the other that is yes. So about seven or eight
years ago a survey was done that I was involved
and where we went around the country asking patients what
their expectations were around digital tools to manage their disease
and illness if they had something like diabetes, or their
health and wellness where they're trying to prevent themselves from
(33:06):
getting diabetes, and the overwhelming response was yes, we want
a digital experience. I think the last piece to share
with you, and it's a really interesting thing to think through,
is we've got these workforce shortage challenges, which is a
big problem to solve. But one of the things in
other health systems have reimagined what healthcare could look like
in terms of a journey for patients and their families
(33:27):
is it's an interesting concept where there's kind of they
have the capacity if you empower the right cohorts to
manage their health and wellness to take some of the
burden off doctors and nurses if they've got the right
tools in front of them, or they can manage their
diabetes themselves. So I think the other thing that's coming
is we enable more digital journeys, patients and their families
(33:50):
and farnes will take more control over their health and
wellness and that'll take some of the burden off the
health system and it waits up all the workforce shortage problems.
Will help in terms of at the moment, you probably
got demand, but they can meet that with their own capacity.
It's a funny way of thinking about things, but definitely
that's the general trained off shore at the moment.
Speaker 1 (34:17):
The other thing that strikes me is as not generative
AI specifically, but machine learning based tools, the kind of
classic AI as I kind of refer to it in
my head. The application of that is becoming so much
more capable and so much more broadly applied that actually
some of these widgets may not necessarily be needed where
(34:41):
they definitely were before.
Speaker 3 (34:42):
And I'm thinking of things.
Speaker 1 (34:43):
Like taku eyes take a photo of your eye to
get certain diagnoses. That strikes me as a way that
AI is helping us directly move towards some level of
equity because you don't need to have as much equipment
in order to be able to actually address or monitor
in some ways.
Speaker 3 (35:03):
Are you seeing that as well?
Speaker 4 (35:04):
You see the channel you know, I probably call that
democratization of health and wellness, right, It's a really interesting story.
So yeah, I've been involved with Tokui since the start
Spark Health in my previous role provide U some innovation
funding to get them using aws on the cloud to
run their algorithms around the email that you just talked about.
So and again, that's a little bit of that of
(35:24):
a consumer customer experience piece where you gather a photo
of the retina. How you gather it can be multiple
different ways, and then you teaching this algorithm to kind
of go, have you got diabetic retinopathy or hypertensive retinopathy?
Speaker 2 (35:38):
Yes?
Speaker 4 (35:38):
Or no? No, don't worry about it. Come back in
a year and get screened. Oh yes, you do set
up the referral path. So and I guess that's that's
one people like tokuais going, hey, we need to unlock, democratize,
provide better X this provide better experience. I definitely see
that coming. It's an interesting area because, as you have alluded, so,
(36:00):
it's diagnostics, right, and again my views medicine traditionally is
you come and see me, I take a history from you,
I run a whole lot of tests, and we work
out what your diagnosis is. With the sophistication of diagnostics
these days, you almost do the diagnostic first to help
you get to the diagnosis, because sometimes it's the gold
(36:21):
standard and you don't necessarily need to take the history
of how you've been feeling. Those types of things so
from my perspective, I think diagnostics are going to get better.
One of the real challenges that's happened off shore is
when patients engage in those diagnostic type tools. If it's right, great,
if it's kind of a little bit on the fence
(36:42):
around what your diagnosis is and how do you kind
of enter into the system to get it clarified. Because
there's the science of medicine that I talked about before,
and there's the art of medicine. And sometimes what you
find and a diagnostic is that it can be one
diagnosis or sometimes it can mean three or four other
diagnoses and you need to do further diagnostic tests to
kind of rule bring out till you get to the one.
(37:04):
So it's a little bit a little bit to think
through there. But I agree with you in terms of
companies like toku Wai's kind of setting the standard around
machine learning and changing the access to those types of services.
Speaker 1 (37:17):
I imagine, you know, maybe ten twenty years, You've got
somebody at home and it's like our time for my
medical checkup, and they get out their smartphone and they
take some photos of various things part of their body,
and they say some words and they you know, maybe
get a smart cheap smart watch and it takes some
stuff like that, and then that can feed into an
algorithm which can be sent to a GP to go,
(37:39):
oh yep, it all looks okay, Like we can give
you the medical tick really, like you said, democratizing it
but also taking it out of urban centers as well
and allowing people who may not have as much access
to primary healthcare to really receive the early intervention care
that can actually make a big difference.
Speaker 4 (37:58):
Yeah, yeah, look, I totally agree with you. One of
the things I was going to share with you today
is being some more and I'm involved in Martin and
PACIFICA getting into digital health tech. And the reason I'm
bringing this up is there's an element of driving kind
of a national way of working rural or urban around healthcare.
But the other thing that's really interesting is sometimes communities
(38:20):
need to solve for themselves. So they understand the business
problem or the health problem. They've got some tech that
they could use, but they actually come together as a
community around how they solve some of these problems. So
I think what we'll see in the future in terms
of what I'm trying to cover here is you'll have
national ways of working around these health checks that might
work for eighty percent of New Zealanders, but it doesn't
(38:41):
work potentially behaviorally for twenty percent. And I'll follow this
through So in the PACIFICA communities, it's really really hard
to get Pacifica to engage with the health system, and
so how do you engage with them? And if you're
offering this tooling, how digitally enabled are my mum, she's
saw them on Chinese she's seventy five, not particularly digitally enabled.
(39:04):
But if you had the tool to do a health
check remotely, who would go and do that with her?
And it's probably been me or one of my three
younger brothers actually facilitate the processes. They've been independent enough
to do it. So there's a few but the technology
is enabled, but how do you practically get people to
use the tech to do the remote health and wellness
checks on an annual basis with the GP So there's
(39:26):
some of that. And the reason I'm showing that with
you is is once you understand the opportunity the technology,
it's really interesting to see how communities solve for how
they practically get the adoption of the tech to improve
health and wellness outcomes, and I think that's awesome. And
the other side of that is I think in my
(39:46):
experience at Accenture that the work we're doing in our
words around mary Indigenous people of New Zealand, some of
the problems you're trying to solve in New Zealand at
the moment, we're ahead of other parts of the world.
So I think about the Native Indian in America, I
think about Aboriginal and Tory Straight Islanders in Australia. Some
of the things we're doing in New Zealand already around
(40:07):
machine learning junior of AI sometimes through that EWI actually
is ahead of everybody else. So there's a little bit
around what we do in New Zealand. If we get
it right and get the adoption, we can kind of
show the world how you can improve equity for Formardi
pacifica as some examples.
Speaker 1 (40:23):
Can you share a kind of an example of what
you mean by that? What are some of the interventions? Y?
Speaker 4 (40:28):
Yeah, so no, it's really interesting to share with you.
So probably two places to start. One of the things
that I think, you know, what does the future of
health and wellness look like in New Zealand and it's
a really interesting to think thing to share with you
that there's this this concept called social determinants of health.
And effectively, if you digitize you every experience in the
health system, from a GP through to physio through to
(40:50):
beteen a hospital doctor, you'd only get one fifth of
the data that determines yours and my health and wellness outcomes.
Then the number one data items your post code and
where you live. But there's a whole lot of behavioral
stuff around do you exercise, do you smoke, do you drink?
What are your family what's your family history. I've got
a strong family history of a schemic heart disease in
my family. Where do you live, how educated are you?
(41:12):
As your house warm or cold? Those types of things.
So one of the opportunities in New Zealand is to
have that holistic approach to social atterminans of health. It's
called now in Martyrdom, there's another concept called tafade Tapafa
and that looks at your emotional health and wellness, your
physical you're spiritual and your mental health and wellness and
(41:33):
so some of the things that you're doing around how
they're applying the tech to that whole person. In the
community is actually having better health and wellness outcomes and
just doing a health system response. So coming back to
some of the use cases that happening at the moment,
So there's a lot of the communities in a number
of EEHE where they're collecting data around all those things
(41:54):
mental health, I guess, general physical health and well being,
emotional and spiritual health and wellbeing, and is starting to
algorithms across the top of that and so and that
helps them do care plans as an example for that
are not just about you know, take your high blood
pressure medication, will walk for thirty minutes today, it's around
thinking about the holistic person. So, and they're using genai
(42:16):
to link that outcomes and drive insights. And they're also
using a bit of email around predictive analytics. And it
comes back to innovation in New Zealand thinking about health
and wellness differently, and then those models of delivering health
and wellness I think where things are going to go globally.
Speaker 1 (42:35):
That's really interesting because it's moving away from this concept
that population aggregated data is the best way of assessing
a population's health and saying well, if you can actually
take data from specific populations and you can really dynamically
split it out by post code, by you know, ethnic background,
(42:57):
and run analysis really quickly and easily over different areas.
And that is directly a result of modern digital data tools.
Speaker 2 (43:07):
Right.
Speaker 1 (43:08):
It wouldn't have been possible even twenty years ago because
everything was so slow and difficult to actually to do.
But the result of being able to do it dynamically
and quickly means that you can really look at areas
where it is most needed and make targeted interventions that
are taking consideration more than just you know, high blood pressure.
Speaker 4 (43:30):
Here's your medicine, Yeah, exactly, And so you know, to
share with you, we've started a conversation with the Ministry
of Social Development and with and also with bai Kaha,
the ministry of disabled people that got broken off from
the health system as part of the reforms, and with
ACC and the context for that conversation is in communities
(43:50):
where health and wellness is a real challenge, how will
you bring insights health social disability in some case is
injury prevention perspective, focusing on an individual or the healthhold
they live in or the community they live in. And
so there's an enabling policy. I'll call it from the
(44:11):
DIA at the moment in Wellington around sharing identity across
those government agencies. So NHI in the hospital context or
health context and acc have a unique number for us
as well, so you know, so they can link my
record from the health system to I guess a shoulder
dislocation I had a few years ago and I was
playing rugby from an accident perspective. So there's enabling policy
(44:35):
to share identity so you can have a single view
of somebody. Then the second bit of that is what
data can be shared, and the third part of that
is what are some of the GENAI use cases you
can run across the top of health data, social services data,
disability data, and so there's real buy in across these
four agencies to kind of look at what the art
of the possible is in bringing together these agencies and
(44:58):
sharing data and a entity. Then once you've got that data,
how can you improve the experience for these some of
these communities that are affected in those contexts. So early
doors at the moment for us, but something that we
do offshore a lot, which is, hey, how do you
get these government agencies to think about the total picture
of social permanence of health and how do you bring
(45:19):
them together to offer new experiences and what opportunities does
genetve Ai bring. So it's a fascinating project that we've
started about three months ago, and then we're about to
get those agencies to look at the ard of the
possible of some of the things we're doing off sure again,
coming back to part of the conversations day, the Genai
conversation opens up these conversations around thinking about reimagining health
(45:41):
and wellness.
Speaker 1 (45:42):
Is there anything that you would like to add, anything
that important you think we haven't covered.
Speaker 4 (45:46):
So something I've been doing for some time is wearing
auring and it's really interesting. It started off as just
I Guess a tool to collect health and fitness data,
and with Genai coming along and now takes that data,
runs algorithms and goes measures mental resilience and it measures
(46:09):
whether I should not go and exercise today because I'm
not ready to do that. So I use that every
day and it also has a I Guess a health
coach based on the data. So just to share with
you right now, my level of daytime stress based on
the ring and the algorithms to me and what I've
been like for the last five years, I'm relaxed, So
that Ben, you must be running a good into you today.
Speaker 3 (46:29):
Yeah.
Speaker 4 (46:30):
Then it measures my heart health and tells me how
old my heart is relative to my age. So according
according to the app, I'm fifty one, and according to
the app, my heart's forty seven. And so I guess
what I'm sharing with you is coming back to some
of these experiences over time, and that consumerism trend I
see generally speaking Kiwi's you know, we'll get some of
(46:52):
these experiences where they manage their health and wellness leveraging
machine learning, generative AI, and then and then when they
think they need the intervention of the health system or
to navigate the health system, then they'll touch the health system.
So right now I wouldn't get where I'm coming from.
But if my heart health went the other way, or
my pulse is getting higher, or I was getting my
mental resilience wasn't so good, I'd consider potentially, you know,
(47:15):
going and seeing my family GP.
Speaker 1 (47:16):
And I guess even if you couldn't put an awring
on every finger in New Zealand, you could have them
in digital check up spots and churches and might I
and you know, a GP clinic where you can just
pop in and get that done really easily and dynamically
and without and linking that back to your health data
so that it can directly feed and maybe flag something
at your GP if there's something wrong.
Speaker 4 (47:38):
Yeah, and look being a really important thing you just
packed up upon that I should have talked about before,
which is the data in the future is going to
be a mixture of what I put in as a
provider about you, but it also should include the information
you put in about you. I see that trend coming,
but pragmatically, do people want it? And I think the
answers over time yes, if they start learning about some
of the capabilities, but they need to be empowered to
(48:00):
manage the health.
Speaker 1 (48:05):
I have to say I thoroughly enjoyed that interview. I
thought Will was really great in the content and the
specificity that he went into around the use of AI
and healthcare, right from generative AI into kind of the
Toku Eyes style machine learning and everywhere in between.
Speaker 3 (48:21):
I thought it was really interesting.
Speaker 2 (48:23):
He was great, you know, and obviously Accentua is heavily
involved in what's going on in the health sector. They're
helping out the government and probably earning a lot of
good consulting fees. But it was a real sort of,
I think, frank and upfront sort of assessment of where
(48:45):
we're at, and a real takeaway for me is what
we're coming from behind. Not on the primary health I
was surprised, you know, Will suggested that we're maybe the
third best in the world when it comes to GP
clinics and that offering digital health service. So my experience
off it hasn't been particularly great with patient portals and
(49:06):
literally watching doctors punching stuff into their computer while I'm
paying for them to do that. But I think we
all know that in hospitals, our systems have not been
great at evolving to meet the modern needs of the population,
and he's definitely put his finger on that. So I
(49:30):
was surprised and pleasantly surprised to hear a lot of
the experimentation and innovation that's already going on with population
health data and that and applying machine learning to it.
So that's great. There's some really cool pilots and stuff
going on with medical devices and people's homes, and then
all of that data eventually could be used in conjunction
(49:52):
with AI for predictive health. So there's a lot of
good stuff going on, but some big barriers. There's well,
we're way behind, We're struggling with a funding crisis in
health so and the data is not yet in a
state where frankly it's going to be reliable enough to
(50:13):
feed into AI system. So a heck of a lot
of work to do.
Speaker 1 (50:16):
Yeah, project here wants an embattled, difficult project is now dead.
Project another failed attempt to conglomerate this healthcare data into
something national and consistent. But we have to get there
at some point. I have to believe that at some
point we're going to need to figure it out. And
(50:36):
maybe trying to create something from the ground up is
just hubris. Maybe we need to figure out a better
kind of glue wear approach or something. But whatever it is,
we need to unlock this data and start enabling it
the sharing between ACC and hospitals and GPS and like
will said in the interview, figuring out how people can
(50:57):
actually input their own data through their wearable where that's
Aora ring and Apple watch or a glucose monitoring system
or whatever else, it will really take us a step
forward and ahead of a lot of other countries if
we can manage to get there.
Speaker 2 (51:13):
Yeah, one of the barriers he mentioned there, which I
was a little bit surprised about, is massive concerns about liability.
So we might need some regulation and compliance tweaks to
give health providers the confidence to use AI and not
be worried about getting sued. I thought that would have
been a bigger deal in a place like the US.
(51:36):
And actually last week was a visiting expert in AI
came through New Zealand and she told me exactly that
she's working with doctors and radiographers and that using AI
and machine learning to try and improve testing and analyzing
test results. And she said the senior doctors in particular
(51:56):
are really pushing back because they're worried about getting sued
and getting their clinic or their hospital suit as well
by making a wrong diagnosis. So that's clearly an issue
here as well. So having that confidence to be able
to put some trust in these systems is going to
(52:16):
be key, and maybe we don't have quite the regulatory
environment to allow that.
Speaker 1 (52:22):
Yeah, just putting those policies in place so that you
know it's not going to be ramp and AI and
making misdiagnoses all over the place. They are going to
be in conjunction with good diagnosticians as well, so they're
being checked and human in the chair and all that
good stuff.
Speaker 2 (52:37):
Yeah. And the other interesting thing that will point it
out when he goes out and talks to hospitals and
clinicians and that sort of thing, a lot of them say, no,
we don't actually have a policy or even a strategy
around using AI. And this is I think we see
just about in any area of industry and business at
(52:58):
the moment, people immediately jumped to the use case. So
I haven't got a strategy for it, but I know
my customers want this. And when it comes to healthcare,
it's using generative AI to surface experiences for patients. So
giving them a chatbot that they can query about sensitive
(53:18):
health related issues and they'll get reliable information. It's that
sort of stuff that actually the health sector thinks is
the frontline of the generative AI revolution.
Speaker 4 (53:27):
Yep.
Speaker 3 (53:28):
It's a big revolution and it's coming. We've just got
to make sure we're ready for it.
Speaker 2 (53:32):
Absolutely. So thanks to doctor Will Reedy for coming on
the Business of Tech, and we'll have more episodes coming
in this ongoing series about how AI is impacting various
sectors and industries across a tea.
Speaker 1 (53:46):
Show notes in the Tech section of the Business Desk
website and the podcast is available on iHeartRadio, well your
favorite podcast platform.
Speaker 2 (53:55):
Let us know what you think of the show and
drop us a line with suggestions for future gare email
Ben on benat Businessdesk dot co, dot MZ, or you
can find both of us on LinkedIn and x.
Speaker 1 (54:06):
Next Thursday, we'll be talking mergers and acquisitions and the
complex technical decisions relating to digital infrastructure that need to
be made when businesses come together.
Speaker 2 (54:16):
Until then, have a great week.
Our health sector is in trouble, short staffed, under resourced,
overwhelmed with patients who have complex and expensive needs.
Speaker 2 (00:08):
Our hospitals are among the least progressive in the Western
world when it comes to digital health. Just as artificial
intelligence shows potential to cut down ADMIN that sucks the
life out of our doctors and nurses.
Speaker 3 (00:23):
So what can AI do for our health system.
Speaker 2 (00:25):
And what are the barriers that are preventing its uptake?
Speaker 1 (00:29):
On the Business Attack sponsored by two degrees Business this week,
AI and how it could enhance public health if we
can get our digital health house in order.
Speaker 4 (00:38):
When you talk to healthcare organizations across New Zealand at
the moment and just ask the question of the board
or the teams or even their chief Data and Digital officer,
do you have an AI strategy? The answer is no.
Speaker 2 (00:50):
Doctor Will Reedy is on the EXTENTSI of New Zealand
leadership team and part of the consulting firm's global health team. Now, Ben,
your interview with will really get me one of the
best overviews of where we're at with AI in health
space in New Zealand. So everyone stick around for that.
Speaker 1 (01:07):
First though and keeping the health theme going. The Dyson
Awards for Design Excellence are underway now, annual global awards
where British inventor Sir James Dyson searches the world for
the best young designers. The New Zealand winner has just
been announced and Peter, I believe you were one of
the judges.
Speaker 2 (01:25):
Yeah, they got me on just to run the ruler
over it in terms of is this something that's going
to connect with a big audience. We had health experts
and other design experts. One of the top designers from
Dyson was a judge as well. This was the first
time I judged it, so it was really good. But
the one that won was one of the most simple
(01:47):
ones really, but was a design that we thought actually
had a really addressable market. No one else was doing
this and it's called the snapcap and a very simple
device that really helps frontline health professionals deal with the
containers that medicines come in, glass containers that need to
(02:11):
be dismantled quickly on the frontline inwards and put in
a sharpy bin to get rid of this stuff. You
need a device to pull these things apart and break
them open, and amazingly there wasn't one in existence.
Speaker 1 (02:27):
Yeah, it's a really cool looking device. It looks kind
of like our twenty first century bottle opener.
Speaker 3 (02:32):
Very simple in.
Speaker 1 (02:33):
Design, very esthetically pleasing in design, and obviously much needed
because I think in your interview you talk about the
fact that frontline health professionals actually do get injuries from
trying to open these little glass bottles.
Speaker 2 (02:49):
Yeah, they do, and it's just a reminder of how
much our frontline health workforce have to do all sorts
of little jobs you don't even think of that is
up to them. So if we can make their lives
easier and hopefully do it at a cheap price, that's great.
Speaker 4 (03:08):
You know.
Speaker 2 (03:08):
This was just one of about a dozen designs that
we looked at. Some of the other stuff that was
sort of like an exoskeleton that would replace a moon boot.
There was insuls that can be customizable and three D
printed for people who have problems with their feet if
they've got potentially diabetes or something like that. There was
(03:29):
a little sense so that you can put on the
back of your shoulder if you're out jogging on a street.
It will give you haptic feedback if it thinks that
a car is getting too close to you or is
potentially going to take you out, so great to see
that creativity.
Speaker 1 (03:42):
I love the simplicity of the winning design, and listening
to your conversation with the designer, Jack Pew gave me
some really good insight into kind of how he got
there and what he was thinking and the balance of
talent and pragmatism that led him there. So let's have
a listen to that interview.
Speaker 2 (03:59):
Now, Jack Pugh, Welcome to the Business of Tech, and
congratulations on winning the New Zealand Dyson Awards. You were
crowned the best design off about a dozen. I was
under the actual judging panel, so I saw all of
these dozen or so designs. Tell us a little bit
(04:22):
about yourself, Jack, Where are you from? How did you
get into design?
Speaker 5 (04:26):
Thanks Peter. So, I'm from christ Church, originally going to
sort of come up to the Capital to study at Messy.
I've always been quite interested in design when I was little.
Any when you ask would tell you that I always
wanted to be an inventor. So it's really cool to
be able to sort of do this and have some
recognition for it.
Speaker 2 (04:44):
You've won this award. Take us through your winning design.
It's called the cap Snap what did you set out
to achieve with the cap snap and what is it?
Speaker 5 (04:53):
Well, it's a simple tool for a simple problem in
the theory, a medical ball opener for health professionals working
with medications where they'd look to recycle them by taking
off the aluminium crimpsy your caps, or to open these
little glass vials which old medication called ampules. There's some
safety risks associated with both, and so this tool kind
(05:15):
of lets them do it in a real quick and
easy way.
Speaker 2 (05:18):
Yeah, it looks like a bottle opener, and essentially it
is a bottle opener. Yeah, functionally, because you know, just
to try and visualize it, and unfortunately, I've got a
lot of experience of this now visiting sick relatives in hospital.
You'll see a medicine bottle. It's got some in this
in the case, i know, immunotherapy liquid drug in it.
(05:41):
It's a glass bottle, it's got an aluminium cap which
is snug on it. You've got to separate those two
things to recycle these materials. You also have, as you say, amples,
these little sort of sealed glass containers that you have
to crack open literally to get the liquid out of them.
But they're really damn fiddly. And what I loved about
(06:03):
your design. You've got one little device that probably doesn't
cost that much to make, and it does both. It
removes that aluminum casing from the top of the glass
bottle and you can insert the ampule into the bottom
of it and crack it in half. So the big
benefit I saw from it is cutting down on the
potential for nurses and doctors to hurt themselves going through
(06:26):
that process cutting themselves on glass or on aluminium or
a tool trying to take that off. It does both
of those purposes.
Speaker 5 (06:34):
Yeah, that's right. The ways that people are currently getting
around both those issues is they're using kind of makeshift
solutions alcohol pads to step ampules, so at least if
it does shedd in an unpredictable way, the glass will
go into the air instead. And with those aluminium's caps,
they're real fiddly. If you ever see someone tri tech
one off, they don't have good tools suited for it,
(06:57):
so they're using kind of four steps and kind of
straining their hands to be able to pull them off.
Speaker 2 (07:01):
Yeah, it's dangerous. And you know what I love is
the simplicity of this. Take us through your your design
approach to this. This is a very specialized piece of equipment,
So what work did you go through sort of talking
to frontline health practitioners to inform the design of cap snap.
Speaker 5 (07:23):
Kind of things started off just kind of as a
conversation on what are the issues, kind of what are
they doing at the moment, and then from there tried
to think of kind of a range of solutions. What
would something handheld and quick and easy look like. First,
maybe something mounted to the wall, or like a mix
of both, something that's kind of a bit of a hybrid.
And I kind of drew up some ideas and sat
(07:44):
down with a member of the team and we kind
of talked through some of the ideas and we ended
up deciding that something that was real quick and small
and portable to be able to be moved around if necessary,
would be the kind of the best way forward. And
then from there I kind of springboard into a bit
of antive process, found some studies overseas looking at similar things.
(08:08):
We kind of developed into well, it's a bottle with
a cap, how about a botopner. We'll keep it super
super easy. And then some of the issues that came
up with one size kind of we couldn't have a
one size fits all approach to have it be real functional,
So we looked at what if we had a bollopner
that could sort of adjust and through that kind of
(08:28):
linear motion, we could incorporate that really seamlessly into the
ampule snapping function.
Speaker 2 (08:34):
It's interesting this was not, by any means the most
elaborate or sophisticated design that we came across the judging
panel for the dice and awards. There were all sorts
of quite elaborate designs that were probably required a lot
more design forinesse. But what I loved about this one
(08:56):
is that when I did research about it, no one
was really doing this. Why do you think no one
in the health profession has come up with something like this?
Speaker 5 (09:03):
Yet you get these things like the ampules or the
crimseal caps, and then the job has to be done,
so you come up with a workaround and you kind
of that's just what you do. And then in a
good and bad way, people don't make a lot of
noise about it. And so we've got all this potential
for interesting design solutions or just simple little fixes for
(09:25):
these health staff we're having to do all these workarounds
on a kind of a daily basis that just sort
of someone needs to have a look at and try
to come up with something.
Speaker 2 (09:33):
And that's exactly what you've done. You know, Sir James
Dyson is famous for being a real iterative designer. Did
you go through numerous iterations and either future obvious ones
you see for this device to make it even better?
Speaker 5 (09:48):
Yeah, So a real fun part of the process is
going through the iterations. So we went through a few
different iterations looking at how I can keep it as
simple as possible and reduce kind of them the number
of mechanical parts in it. There are kind of two approaches.
One kind of looked like a claw, if that makes sense,
and that would allow it to be able to open
a bunch of different sizes without having to use that
(10:11):
slighter function. But the issue we kind of ran into
with some of those prototypes is that it wasn't sort
of snappy and intuitive. If you looked at it, you
wouldn't know what you're looking at, and that was a
real big part of the design is making it just
super straightforward that you can look at it and pretty
much figure out how to use it without any instruction.
There's so little time to teach people about new things
(10:34):
when a new tool comes in kind of in these busy,
bustling environments that that's kind of what you want from
a tool that's going to do just this quick and
simple job.
Speaker 2 (10:41):
Yeah, Because it literally looks like a bottle opener, so
instantly you have that recognition. You go, oh, I've got
a bottle here, This must go around the neck. And
then you've got the sort of the hidden compartment on
the bottom of it, which is for breaking the ample
as well. So I guess it's a bit of education
required so that people know to use that. You're sitting
in your lab as we talk here, you've got a
three D printer behind you. Was that useful? Do you
(11:04):
do a lot of iterations and design work and prototyping
using three D printing?
Speaker 5 (11:09):
Yeah? So three D printing was a huge enabler for
this project. That and water jet cutting was another super
useful part in sort of smashing out these prototypes. I
tried to kind of reduce the amount of plastic that
I used where I could, so I tried to make
my prototypes I can kind of hot swap between the
different inserts to try the different geometries to see what
(11:32):
was kind of the best and most effective fit when
kind of figuring out some of the measurements. So yeah,
definitely super pivotal and being able to quickly run through ideas.
Speaker 2 (11:42):
So design was only sort of part of the criteria
for winning this award. A big part of it was
the real world application and the potential for this to
actually go on and be used. It's all well and
good to design something that just never leaves the labor
with the prototyping stage, and we did see some sort
of designs like that. What are the next steps for you?
(12:05):
Is this something that you'd like to pursue potentially as
a business that the Snapcap try and get it out
there into hospitals and clinics around the world.
Speaker 5 (12:12):
I would like to see it in people's hands and
just making their lives a little bit easier for those tasks.
So I'm able to put it a little bit of
time in to be able to work up the design
a little bit further. I've got some help from some
awesome people at a couple of the other hospitals around
the country been able to kind of cast the net
a little bit wider see how the same issues are
(12:36):
sort of getting received across those different sites. We're hoping
to be able to work it into something that we
can get into people's hands.
Speaker 2 (12:43):
The judges we're talking about this, we sort of thought
this is great, and you submitted a video literally off
a nurse who was I think snapping an ampool who
actually cut herself.
Speaker 5 (12:55):
It wasn't scripted, but we had to use that tape.
Speaker 2 (12:58):
Well, that illustrates it very well. But we were thinking,
why don't you just I mean, you need to break
the ampull, but in terms of taking the aluminum cap
off the bottle, why don't you just chuck it all
in a bin and automate that later.
Speaker 5 (13:12):
That was an idea that we kind of tossed around
at the start. But the kind of reality of kind
of this product is that it'd be all well and
good to make something real, awesome and automated, and I
actually going into this project that's kind of what I
was going to look to do. But then talking with
the people and like hearing about the issues and then
(13:33):
learning about the ampuel side of things, we've sort of
determined that while technically it could be considered a band
aid solution, it would be on the fastest track to
be able to have this issue kind of be sold.
Speaker 2 (13:45):
That's what we loved about it, the simplicity of it.
I mean, presumably this, if you get this into production,
it wouldn't be a super expensive device to create either.
Speaker 5 (13:54):
Yeah, we're hoping to well, we're working to get the
part count down as much as possible. Engineers in christ
Church I've had some great input on on clever ways
to really simplify that mechanism to even just three or
four parts.
Speaker 3 (14:09):
Tops.
Speaker 2 (14:09):
Where are you working at the moment?
Speaker 5 (14:11):
So I'm working out of Wellington Regional Hospital, part of
the Futtow Water Improvement team based out of here, so
day to day I'm sort of looking at other issues,
but it's still we're able to work on some innovative
solutions sort of within the hospital, which is a really
exciting sort of opportunity.
Speaker 2 (14:28):
Well, brilliant, you're in exactly the right place and they're
lucky to have this design brain on the team there.
So good luck for the next phase. And he's hoping
we see snapcap in hospital wards before too long.
Speaker 1 (14:41):
Well hat fingers crossed, great young talent, Jack Pugh. Great
to have those kinds of people coming through New Zealand,
so big well done to him and look forward to
seeing what he does next. So that's innovation and health hardware,
and our topic of focus this week is AI in healthcare.
(15:02):
Doctor Will Reedy is one of the country's leading experts
in digital health. He joined Accentua around a year ago
and is still a practicing doctor one day a week
in the county's Monaco area.
Speaker 2 (15:13):
He's a lot of experience helping with the rollout of
digital health systems all over the world, so he's pretty
well placed to compare and contrast our preparedness and progress
in the digital health space compared to the likes of Europe,
the US and Australia, where he's also worked.
Speaker 3 (15:30):
Will's also really interested in the potential to reduce health
inequities using digital health tech, that is, if we can
trust machine learning and large language models to get it right.
Speaker 2 (15:40):
So here's Ben's interview with Accentures doctor Will Reedy.
Speaker 3 (15:50):
Good, Hey, Will, how are you great? Thanks?
Speaker 1 (15:52):
Ben good, Welcome to the Business of Tech podcast. Thanks
for joining us. As so, why don't we start with
if you could just give us a very quick summary of.
Speaker 3 (16:00):
Who you are, what you do, and a little bit
about your background.
Speaker 4 (16:03):
So yeah, look, it's pleasure to be chatting with you
this morning. So I guess my full time job is
working for Accenture, is their how order health and wellness
lead here in New Zealand with the goal to help
the health system transform given some of the real challenges
that many of us aware of at the moment in
the health system. And then one day a week I
do a shift in the surgical services at Middlemore so
(16:26):
that can be in the emergency department and the clinics,
in the wards and sometimes in theater. So I guess
that keeps it real in terms of understanding the pulse
of the health system at the clinical front line. But
I guess ultimately the passion is to try and transform
health systems with technology and things like AI.
Speaker 1 (16:42):
It must be quite start going from in Accenture talking
about the latest and greatest worldwide AI this, and then
you go into a hospital and very different story.
Speaker 3 (16:53):
I'd imagine, Yeah, it is.
Speaker 4 (16:54):
It's really interesting and it's generally accepted that in our
hospital system, not so much in our GP systems or
primary care systems. We are the least digitized hospitals now
in the developed world. So it is a little bit
of a change in terms of the global work that
I do and seeing what's going on in terms of
transformation and how far some other countries are ahead of us.
(17:16):
Also being quite digital my day job at Accenture and
then you know, having to use a pen and paper
on Fridays pretty much.
Speaker 3 (17:23):
Yeah.
Speaker 1 (17:24):
But also I guess conversely, that also just helps you
to understand the potential for change, right, and that actually
how far we can go when we start to implement
modern tech.
Speaker 4 (17:34):
Yeah. I think it's easy to kind of look at
the I guess, the lower level of digitization and tech
enable transformation in the New Zealand health system today. But
it's also an opportunity, and my words are an opportunity
to kind of leap frog some of the approaches and
the thinking. So I do I'm quite optimistic about some
of those opportunities for New Zealand just to go, hey,
(17:54):
where are we trying to get to? What would we
do differently? Could we leap frog ahead of some of
the other jurisdictions.
Speaker 1 (18:00):
Yeah, it's interesting you use that term leap frog. That
seems to be kind of a relatively common New Zealand
experience where we kind of fall behind a little bit
and then we go, oh, let's catch up, and in
doing so we kind of go ahead and really hit
that cutting edge again. Is that kind of what you're
seeing happening at the moment?
Speaker 4 (18:17):
I guess, I see the opportunity and it's really interesting
to share with you. It's funny how some of the
AI technologies are becoming quite pervasive in healthcare. And a
colleague of mine has been leading the way in gp
Land all it or primary care in driving the adoption
of a product called Nabler, which allows you, with the
(18:37):
permission of the patient, to record the voice around the
interaction with the patient and then convert that to text
and then put it into the GP system. And our
GPS I think about ranked about third in the world
in terms of their levels of digitization. And that's been
kind of the same kind of I guess measurement that's
been in place for about twenty years. So primary care
in New Zealand's actually been I guess at the forefront
(18:59):
of tech for some time. And then you know, what
he's seeing is first and foremost the patients and they're
farno who are in the room with him, are seeing
some benefits around more eye contact not turning around and
typing stuff in the computer and in terms of the
fifteen minute consult that is common in primary care. The
other side that he's finding really interesting is efficiencies within
(19:20):
the consult around typing everything down because we often paraphrase
what the family or the patients say to us given
the time limit. But what he's also finding, as our
colleagues across about one dred and one hundred and fifty
practices in New Zealand, is that the cognitive load, that
the mental load is a lot less because you're having
to recite and think things through. So piece of technology
(19:41):
came along relatively recently and it's been adopted in about
three hundred practices across New Zealand, which is about fifteen
percent of the GP practices. So it's really interesting around
that leap frog concept. Grab an idea and go with it. Yep.
Speaker 1 (19:56):
It is an interesting idea and it is one that
has you know, it could have been earlier probably, but
the reality is is that the technology that we have
now around AI has just made it super accessible.
Speaker 3 (20:08):
Are there other.
Speaker 1 (20:08):
Areas that you're seeing that trend that the modern AI
tools that we have now in the last couple of
years have made things possible suddenly that would have seemed
really onerous in the past.
Speaker 4 (20:19):
Yeah, look, it's interesting. I'll probably give some broader contexts.
So one of the challenges in the health system at
the moment is how do you adopt AI? And I
guess people translate AI to generative AI at the moment,
so it's just careful to be specific about that. And
so the adoption piece is and I'll get down to
some of the use cases shortly but effectively, when you
(20:40):
talk to healthcare organizations across New Zealand at the moment,
and just ask the question of the board or their
exec teams or even their chief Data and Digital officer,
do you have an AI strategy? And the answer is no,
which is really interesting. And then the second question you
ask in the context of adoption of AI, if you
went down that path, is do you have a policy
(21:01):
around how you'll adopt AI as an organization? And generally
speaking not many have that either. But then you go, okay,
put that to one size. You don't have a strategy,
you don't have a policy. What are the core use
cases that you've been thinking about that you'd like to
get into your organization? In the next twelve months and
again that's where it opens up interesting conversations. And so
(21:21):
when we ran a leadership summit earlier in the year
with the Chief Medical Officer for Health New Zealand and
the head of the AI Advisory Group, the most common
use case was surfacing genitive AI experiences to patients or
family or FANO, so things like education CHAP as an example,
I've just been diagnosed with diabetes. What can I expect
and how do you get repeatable advice to patients where
(21:43):
doctor or nurse isn't available as an example.
Speaker 1 (21:45):
That's kind of very similar to the recent announcement around
GOVGBT right where it's the ingesting a bunch of government documents,
government pages and then being able to chatbot style ask
questions and get information about this governments.
Speaker 4 (22:00):
And I think I think in New Zealand, you know,
the health systems around the world, particularly in socialized health
systems like New Zealand, Australia in the UK, is everything's
quite fragmented. So if you've got a consistent education piece
experience for patients where no matter what question they ask,
they are a consistent answer, it's actually a big benefit
in terms of patients being empowered to manage a chronic
(22:24):
I guess condition like diabetes. So yeah, we are quite
surprised that everybody is going, how do we surface stuff
to patients? Which is really interesting, So that's open up
a new world. The second kind of area, broadly was
what we call the clinical or the front line or
the front of office, and so lots of use cases
around voice detext and reducing the burden of me using
(22:45):
pen and paper as an example, some benefits around managing inboxes,
around lab results coming in because a lot of lab
tests that we order for patients, we're kind of trying
to rule something out and we want it to be
if it comes back normal, then we don't really need
to process that forget where I'm coming from. And then
the back of office piece around workforce management and finance,
(23:06):
procurement and supply chain. So those are the broad areas.
So the reason I explain those broad areas around use
cases is people are understanding there is potential to apply
in particular generative AI to those use cases. It's just
where are they going to get the biggest impact around
healthcare in New Zealand. So I guess your question was, Hey,
are people's eyes being opened up. Yes, they are because
(23:28):
they're learning around use cases against offshore at the moment
and going, hey, that could easily be applied here in
the New Zealand context.
Speaker 1 (23:35):
Yeah, sticking with the generative AI theme, you know, I've
often looked at the health of Fire website as it's
now called, and just that that's such a huge corpus
of information and data, like it seems like a great
opportunity for something like an educational chatbot, But there is
a lot of risk with that. Where chatbots are known
to want to please people, they're known to kind of
(23:57):
sometimes make things up if they're relying heavily on those
LLAM models in the background. How are people in the
health sector thinking about those risks considering the sensitivity.
Speaker 4 (24:08):
Yeah, it's a really good question. I think the interesting
context is do healthcare organizations need to adopt genai or not?
And the general trend overseas is they're all adopting it
to see what the potential is, but not necessarily to
roll it out at scale. And the reason they're doing
(24:29):
that is it a competitive advantage? Does it deliver a
better service? Those types of things. So one of the
things I've seen In New Zealand. We have six and
a half thousand healthcare organizations, of which one is Health
New Zealand. It is the biggest, but that's the wider context.
So organizations need to think about AI from a responsible perspective.
And the biggest concern slash barrier is exactly what you
(24:51):
just articulated, which is hallucinations. I think the soft words
are unreliable outputs, and so I think a lot of
that has to be un understood. Choosing our use case,
learning about the use case, having the governance and leadership
in place to go. Actually, we've tried this in a
small use case. We have actually looked at what's happening
(25:12):
overseas and these use cases do actually add value, but
you'd got to go on the journey around I guess
the hallucination side of things. The other thing that's interesting
in New Zealand, in Health New Zealand as an example
of starting to do it, is they need to get
all their data in one place to run the GENAI
lms across the top of it. And so there's definitely
one that's the strategy for Health New Zealand, and they've
(25:34):
started investing in something called National data platform, so they
can put their data all in one place, whether it's
clinical data, workforce data, finance data, and then they can
start training the GENAI tools on top of their own
data and then managing that piece. You still have hallucinations,
don't get me wrong around data quality, but at least
you're doing it on the data that you have governance
(25:55):
over a couple of things to share with you. In
terms of New Zealand's text for GENAI adoption, mass of
concerns around liability. So if I have run a GENAI
tool and I've surfaced it to a patient and it's
given them some advice and something doesn't go as well
as it would have liked, and there's more risk in
terms of patient care, who's liable Is it the clinician
(26:18):
or is it the GENAI And how does it actually work?
So he's a little bit of maturity around what I
would call regulation and compliance to think through, because at
the end of the day, most healthyic organizations have a
clinical obligation around the safety of the care that they provide.
And if you introduce AI alongside the clinician, how does
(26:38):
it actually work and what are the implications? And the
last year that's really interesting is privacy and security of
the data. So you might bring it all in, but
you know, how do you actually control it in the
world of more cyber attacks, particularly in health systems around
the world. So those are kind of the core barriers.
But the number one is obviously the hallucinations piece.
Speaker 1 (26:55):
Yeah, so you know, I guess part of that is
having the confidence to experiment and trial and go at it,
but also having the confidence to say, actually, in this case,
the risks are too high, the technology is not there
yet or may not be, and so we're going to
choose to not make this a customer facing or patient
facing thing. Keep it for doctors or healthcare providers and
(27:17):
they can be the intermediary between the chatbot that's getting
a lot of information and the patient at the other end.
Is that kind of how thinking is going, Yeah.
Speaker 5 (27:27):
It is.
Speaker 4 (27:27):
It's an interesting piece because if you're wanting to experience
an experiment and dip your toes and genitive AI, I
think that the two pieces of conversation we're having in
New Zealand at the moment is so, what are the
use cases at a gathering momentum offshore and generally speaking
in the healthcare context using GENAI to re platform, recode
(27:48):
old applications, and New Zealand has a problem at the
moment around that we've got a lot of legacy applications
being used, particularly in the public health system. The second
area is around contact center experience, and the third one
is that voice to text that I talk about. So
those are the three broad areas. The other thing that
we've packed up offshore, which is interesting to share with you,
is generally speaking, most of the GENAI use cases have
(28:10):
been done on top of platforms. They're leveraging Salesforce, Microsoft,
our electronic medical record platforms where genitive AI is being
built in as a feature if you like, in these platforms.
So the hallucinations piece isn't as big a risk because
it's built into a platform. It's well tested, So I
think it's a fine balance. But it's just interesting to
(28:31):
see what the trends are in terms of the practical
elements of gen AI, the hype versus in reality what's
going on.
Speaker 1 (28:38):
Yeah, absolutely, that's kind of a lot about primary care,
and we talked about how that's really advanced. What about
in the hospital world.
Speaker 4 (28:47):
Yeah, Again, it's an interesting discussion around looking what happens
off sure, and I think there's a couple of contextual things.
So as a practicing clinician, when I think about technology
like genitive AI, I think of it as another tool
in my clinical practice, like over stethoscope around my neck.
Most of the younger generation clinicians, I'm I'm a I
guess I called a veteran these days because they've been
(29:09):
around the health system for twenty five years. But if
I look at some of the newly trained doctors and nurses,
they all expect if you like, generative AI to be
available for some of the use cases around again that
voice detext piece, and we haven't done anything in New
Zealand at the moment, but in Australia, the first use
(29:29):
case is using voice to text and busy theaters to
drive through throughput so you can hit health targets around,
you know, like the waiting lists for hypophens and things
like that. So there is a willingness for in that
surgical operating theater for surgeons, anetheists, theater nurses to use
those technologies to be more efficient in terms of through
(29:50):
put through the theater because they're not writing notes after
an operation. It's been done real time, so yes, there
is potential, just not quite happening at the moment, and
I guess that's the clinical frontline piece. Really interesting report
that we did with Microsoft recently for New Zealand looking
at how genitive AI could be applied to the nursing
workforce and based on using some of the Microsoft technologies
(30:14):
voice to texts predictive analytics around patients becoming unwell because
they have temperatures gone up or their blood pressures dropped.
They said, if you put on these common tools, you'd
get a productivity increase of around nine to ten percent
for each nurse across the country. So that's proven offshore,
done a little bit of analysis around how I guess
nursing workforce works the New Zealand today and our public hospitals.
(30:36):
You apply these technologies and process improvement augmentation around genitive
AI would actually lift productivity ten percent, which I guess
you know eight to twelve hour shifts is you know
one to one and a half hours, which does make
a big difference.
Speaker 1 (30:50):
It does, especially if you're looking at you know, a
shortage and if ten percent is saying you have the
equivalent to ten nurses that have nine nurses on shift.
Speaker 3 (30:58):
That actually does make a difference end of the day.
Speaker 4 (31:00):
It does, and you know, we've got to you know,
as you've heard in the media, we've got to adopt
a shortage of nursing shortage and in some areas what
I would call an allied health professional shortage BUZZIO is
occupational therapists. The biggest year actually at the moment is
anesthetic technicians. So that's actually preventing some of the through
put in private and public hospitals where there's just enough
(31:20):
of anesthetic technicians while you're sleep to look after you.
It's really interesting at the moment.
Speaker 1 (31:25):
What about the casting forward into the future. Do you
see it, as you know, potentially being a kind of
first point of contact for patients to be able to
say instead of just pushing a button waiting for a
nurse pushing a button and saying, oh, I think my
leg's really hurting and I'm not sure, and so then
that can automatically go into like a database and be
(31:46):
triaged and get the nurses to who needs to. Like,
it strikes me there is a fair bit of potential
in that first line of support for generative AI in
the future.
Speaker 4 (31:56):
Yeah, look, I totally agree with you. So I think
technologies like unit of AI we often talk about in
the current health system of New Zealand, we've got to
transform it generally. That means you spend in the next
five years trying to change the reality is we need
to ask New Zealanders, you know, what's your experience like
in terms of their most recent interaction with a healthcare
professional or a period of care and hospital. So my
(32:19):
view is first and foremost is that's where things are going.
I guess it's probably called consumerism and what do people
expect in terms of their health experience and how do
they compare it to other experiences like banking just to
use it? And I guess at parallel so I do
think one our customers, patients like you and I, will
expect a better level of experience and driven by digital
(32:42):
touch points, of which GENAI will power some of them.
Speaker 3 (32:44):
It's number one.
Speaker 4 (32:45):
The next question is do New Zealanders want it? And
the other that is yes. So about seven or eight
years ago a survey was done that I was involved
and where we went around the country asking patients what
their expectations were around digital tools to manage their disease
and illness if they had something like diabetes, or their
health and wellness where they're trying to prevent themselves from
(33:06):
getting diabetes, and the overwhelming response was yes, we want
a digital experience. I think the last piece to share
with you, and it's a really interesting thing to think through,
is we've got these workforce shortage challenges, which is a
big problem to solve. But one of the things in
other health systems have reimagined what healthcare could look like
in terms of a journey for patients and their families
(33:27):
is it's an interesting concept where there's kind of they
have the capacity if you empower the right cohorts to
manage their health and wellness to take some of the
burden off doctors and nurses if they've got the right
tools in front of them, or they can manage their
diabetes themselves. So I think the other thing that's coming
is we enable more digital journeys, patients and their families
(33:50):
and farnes will take more control over their health and
wellness and that'll take some of the burden off the
health system and it waits up all the workforce shortage problems.
Will help in terms of at the moment, you probably
got demand, but they can meet that with their own capacity.
It's a funny way of thinking about things, but definitely
that's the general trained off shore at the moment.
Speaker 1 (34:17):
The other thing that strikes me is as not generative
AI specifically, but machine learning based tools, the kind of
classic AI as I kind of refer to it in
my head. The application of that is becoming so much
more capable and so much more broadly applied that actually
some of these widgets may not necessarily be needed where
(34:41):
they definitely were before.
Speaker 3 (34:42):
And I'm thinking of things.
Speaker 1 (34:43):
Like taku eyes take a photo of your eye to
get certain diagnoses. That strikes me as a way that
AI is helping us directly move towards some level of
equity because you don't need to have as much equipment
in order to be able to actually address or monitor
in some ways.
Speaker 3 (35:03):
Are you seeing that as well?
Speaker 4 (35:04):
You see the channel you know, I probably call that
democratization of health and wellness, right, It's a really interesting story.
So yeah, I've been involved with Tokui since the start
Spark Health in my previous role provide U some innovation
funding to get them using aws on the cloud to
run their algorithms around the email that you just talked about.
So and again, that's a little bit of that of
(35:24):
a consumer customer experience piece where you gather a photo
of the retina. How you gather it can be multiple
different ways, and then you teaching this algorithm to kind
of go, have you got diabetic retinopathy or hypertensive retinopathy?
Speaker 2 (35:38):
Yes?
Speaker 4 (35:38):
Or no? No, don't worry about it. Come back in
a year and get screened. Oh yes, you do set
up the referral path. So and I guess that's that's
one people like tokuais going, hey, we need to unlock, democratize,
provide better X this provide better experience. I definitely see
that coming. It's an interesting area because, as you have alluded, so,
(36:00):
it's diagnostics, right, and again my views medicine traditionally is
you come and see me, I take a history from you,
I run a whole lot of tests, and we work
out what your diagnosis is. With the sophistication of diagnostics
these days, you almost do the diagnostic first to help
you get to the diagnosis, because sometimes it's the gold
(36:21):
standard and you don't necessarily need to take the history
of how you've been feeling. Those types of things so
from my perspective, I think diagnostics are going to get better.
One of the real challenges that's happened off shore is
when patients engage in those diagnostic type tools. If it's right, great,
if it's kind of a little bit on the fence
(36:42):
around what your diagnosis is and how do you kind
of enter into the system to get it clarified. Because
there's the science of medicine that I talked about before,
and there's the art of medicine. And sometimes what you
find and a diagnostic is that it can be one
diagnosis or sometimes it can mean three or four other
diagnoses and you need to do further diagnostic tests to
kind of rule bring out till you get to the one.
(37:04):
So it's a little bit a little bit to think
through there. But I agree with you in terms of
companies like toku Wai's kind of setting the standard around
machine learning and changing the access to those types of services.
Speaker 1 (37:17):
I imagine, you know, maybe ten twenty years, You've got
somebody at home and it's like our time for my
medical checkup, and they get out their smartphone and they
take some photos of various things part of their body,
and they say some words and they you know, maybe
get a smart cheap smart watch and it takes some
stuff like that, and then that can feed into an
algorithm which can be sent to a GP to go,
(37:39):
oh yep, it all looks okay, Like we can give
you the medical tick really, like you said, democratizing it
but also taking it out of urban centers as well
and allowing people who may not have as much access
to primary healthcare to really receive the early intervention care
that can actually make a big difference.
Speaker 4 (37:58):
Yeah, yeah, look, I totally agree with you. One of
the things I was going to share with you today
is being some more and I'm involved in Martin and
PACIFICA getting into digital health tech. And the reason I'm
bringing this up is there's an element of driving kind
of a national way of working rural or urban around healthcare.
But the other thing that's really interesting is sometimes communities
(38:20):
need to solve for themselves. So they understand the business
problem or the health problem. They've got some tech that
they could use, but they actually come together as a
community around how they solve some of these problems. So
I think what we'll see in the future in terms
of what I'm trying to cover here is you'll have
national ways of working around these health checks that might
work for eighty percent of New Zealanders, but it doesn't
(38:41):
work potentially behaviorally for twenty percent. And I'll follow this
through So in the PACIFICA communities, it's really really hard
to get Pacifica to engage with the health system, and
so how do you engage with them? And if you're
offering this tooling, how digitally enabled are my mum, she's
saw them on Chinese she's seventy five, not particularly digitally enabled.
(39:04):
But if you had the tool to do a health
check remotely, who would go and do that with her?
And it's probably been me or one of my three
younger brothers actually facilitate the processes. They've been independent enough
to do it. So there's a few but the technology
is enabled, but how do you practically get people to
use the tech to do the remote health and wellness
checks on an annual basis with the GP So there's
(39:26):
some of that. And the reason I'm showing that with
you is is once you understand the opportunity the technology,
it's really interesting to see how communities solve for how
they practically get the adoption of the tech to improve
health and wellness outcomes, and I think that's awesome. And
the other side of that is I think in my
(39:46):
experience at Accenture that the work we're doing in our
words around mary Indigenous people of New Zealand, some of
the problems you're trying to solve in New Zealand at
the moment, we're ahead of other parts of the world.
So I think about the Native Indian in America, I
think about Aboriginal and Tory Straight Islanders in Australia. Some
of the things we're doing in New Zealand already around
(40:07):
machine learning junior of AI sometimes through that EWI actually
is ahead of everybody else. So there's a little bit
around what we do in New Zealand. If we get
it right and get the adoption, we can kind of
show the world how you can improve equity for Formardi
pacifica as some examples.
Speaker 1 (40:23):
Can you share a kind of an example of what
you mean by that? What are some of the interventions? Y?
Speaker 4 (40:28):
Yeah, so no, it's really interesting to share with you.
So probably two places to start. One of the things
that I think, you know, what does the future of
health and wellness look like in New Zealand and it's
a really interesting to think thing to share with you
that there's this this concept called social determinants of health.
And effectively, if you digitize you every experience in the
health system, from a GP through to physio through to
(40:50):
beteen a hospital doctor, you'd only get one fifth of
the data that determines yours and my health and wellness outcomes.
Then the number one data items your post code and
where you live. But there's a whole lot of behavioral
stuff around do you exercise, do you smoke, do you drink?
What are your family what's your family history. I've got
a strong family history of a schemic heart disease in
my family. Where do you live, how educated are you?
(41:12):
As your house warm or cold? Those types of things.
So one of the opportunities in New Zealand is to
have that holistic approach to social atterminans of health. It's
called now in Martyrdom, there's another concept called tafade Tapafa
and that looks at your emotional health and wellness, your
physical you're spiritual and your mental health and wellness and
(41:33):
so some of the things that you're doing around how
they're applying the tech to that whole person. In the
community is actually having better health and wellness outcomes and
just doing a health system response. So coming back to
some of the use cases that happening at the moment,
So there's a lot of the communities in a number
of EEHE where they're collecting data around all those things
(41:54):
mental health, I guess, general physical health and well being,
emotional and spiritual health and wellbeing, and is starting to
algorithms across the top of that and so and that
helps them do care plans as an example for that
are not just about you know, take your high blood
pressure medication, will walk for thirty minutes today, it's around
thinking about the holistic person. So, and they're using genai
(42:16):
to link that outcomes and drive insights. And they're also
using a bit of email around predictive analytics. And it
comes back to innovation in New Zealand thinking about health
and wellness differently, and then those models of delivering health
and wellness I think where things are going to go globally.
Speaker 1 (42:35):
That's really interesting because it's moving away from this concept
that population aggregated data is the best way of assessing
a population's health and saying well, if you can actually
take data from specific populations and you can really dynamically
split it out by post code, by you know, ethnic background,
(42:57):
and run analysis really quickly and easily over different areas.
And that is directly a result of modern digital data tools.
Speaker 2 (43:07):
Right.
Speaker 1 (43:08):
It wouldn't have been possible even twenty years ago because
everything was so slow and difficult to actually to do.
But the result of being able to do it dynamically
and quickly means that you can really look at areas
where it is most needed and make targeted interventions that
are taking consideration more than just you know, high blood pressure.
Speaker 4 (43:30):
Here's your medicine, Yeah, exactly, And so you know, to
share with you, we've started a conversation with the Ministry
of Social Development and with and also with bai Kaha,
the ministry of disabled people that got broken off from
the health system as part of the reforms, and with
ACC and the context for that conversation is in communities
(43:50):
where health and wellness is a real challenge, how will
you bring insights health social disability in some case is
injury prevention perspective, focusing on an individual or the healthhold
they live in or the community they live in. And
so there's an enabling policy. I'll call it from the
(44:11):
DIA at the moment in Wellington around sharing identity across
those government agencies. So NHI in the hospital context or
health context and acc have a unique number for us
as well, so you know, so they can link my
record from the health system to I guess a shoulder
dislocation I had a few years ago and I was
playing rugby from an accident perspective. So there's enabling policy
(44:35):
to share identity so you can have a single view
of somebody. Then the second bit of that is what
data can be shared, and the third part of that
is what are some of the GENAI use cases you
can run across the top of health data, social services data,
disability data, and so there's real buy in across these
four agencies to kind of look at what the art
of the possible is in bringing together these agencies and
(44:58):
sharing data and a entity. Then once you've got that data,
how can you improve the experience for these some of
these communities that are affected in those contexts. So early
doors at the moment for us, but something that we
do offshore a lot, which is, hey, how do you
get these government agencies to think about the total picture
of social permanence of health and how do you bring
(45:19):
them together to offer new experiences and what opportunities does
genetve Ai bring. So it's a fascinating project that we've
started about three months ago, and then we're about to
get those agencies to look at the ard of the
possible of some of the things we're doing off sure again,
coming back to part of the conversations day, the Genai
conversation opens up these conversations around thinking about reimagining health
(45:41):
and wellness.
Speaker 1 (45:42):
Is there anything that you would like to add, anything
that important you think we haven't covered.
Speaker 4 (45:46):
So something I've been doing for some time is wearing
auring and it's really interesting. It started off as just
I Guess a tool to collect health and fitness data,
and with Genai coming along and now takes that data,
runs algorithms and goes measures mental resilience and it measures
(46:09):
whether I should not go and exercise today because I'm
not ready to do that. So I use that every
day and it also has a I Guess a health
coach based on the data. So just to share with
you right now, my level of daytime stress based on
the ring and the algorithms to me and what I've
been like for the last five years, I'm relaxed, So
that Ben, you must be running a good into you today.
Speaker 3 (46:29):
Yeah.
Speaker 4 (46:30):
Then it measures my heart health and tells me how
old my heart is relative to my age. So according
according to the app, I'm fifty one, and according to
the app, my heart's forty seven. And so I guess
what I'm sharing with you is coming back to some
of these experiences over time, and that consumerism trend I
see generally speaking Kiwi's you know, we'll get some of
(46:52):
these experiences where they manage their health and wellness leveraging
machine learning, generative AI, and then and then when they
think they need the intervention of the health system or
to navigate the health system, then they'll touch the health system.
So right now I wouldn't get where I'm coming from.
But if my heart health went the other way, or
my pulse is getting higher, or I was getting my
mental resilience wasn't so good, I'd consider potentially, you know,
(47:15):
going and seeing my family GP.
Speaker 1 (47:16):
And I guess even if you couldn't put an awring
on every finger in New Zealand, you could have them
in digital check up spots and churches and might I
and you know, a GP clinic where you can just
pop in and get that done really easily and dynamically
and without and linking that back to your health data
so that it can directly feed and maybe flag something
at your GP if there's something wrong.
Speaker 4 (47:38):
Yeah, and look being a really important thing you just
packed up upon that I should have talked about before,
which is the data in the future is going to
be a mixture of what I put in as a
provider about you, but it also should include the information
you put in about you. I see that trend coming,
but pragmatically, do people want it? And I think the
answers over time yes, if they start learning about some
of the capabilities, but they need to be empowered to
(48:00):
manage the health.
Speaker 1 (48:05):
I have to say I thoroughly enjoyed that interview. I
thought Will was really great in the content and the
specificity that he went into around the use of AI
and healthcare, right from generative AI into kind of the
Toku Eyes style machine learning and everywhere in between.
Speaker 3 (48:21):
I thought it was really interesting.
Speaker 2 (48:23):
He was great, you know, and obviously Accentua is heavily
involved in what's going on in the health sector. They're
helping out the government and probably earning a lot of
good consulting fees. But it was a real sort of,
I think, frank and upfront sort of assessment of where
(48:45):
we're at, and a real takeaway for me is what
we're coming from behind. Not on the primary health I
was surprised, you know, Will suggested that we're maybe the
third best in the world when it comes to GP
clinics and that offering digital health service. So my experience
off it hasn't been particularly great with patient portals and
(49:06):
literally watching doctors punching stuff into their computer while I'm
paying for them to do that. But I think we
all know that in hospitals, our systems have not been
great at evolving to meet the modern needs of the population,
and he's definitely put his finger on that. So I
(49:30):
was surprised and pleasantly surprised to hear a lot of
the experimentation and innovation that's already going on with population
health data and that and applying machine learning to it.
So that's great. There's some really cool pilots and stuff
going on with medical devices and people's homes, and then
all of that data eventually could be used in conjunction
(49:52):
with AI for predictive health. So there's a lot of
good stuff going on, but some big barriers. There's well,
we're way behind, We're struggling with a funding crisis in
health so and the data is not yet in a
state where frankly it's going to be reliable enough to
(50:13):
feed into AI system. So a heck of a lot
of work to do.
Speaker 1 (50:16):
Yeah, project here wants an embattled, difficult project is now dead.
Project another failed attempt to conglomerate this healthcare data into
something national and consistent. But we have to get there
at some point. I have to believe that at some
point we're going to need to figure it out. And
(50:36):
maybe trying to create something from the ground up is
just hubris. Maybe we need to figure out a better
kind of glue wear approach or something. But whatever it is,
we need to unlock this data and start enabling it
the sharing between ACC and hospitals and GPS and like
will said in the interview, figuring out how people can
(50:57):
actually input their own data through their wearable where that's
Aora ring and Apple watch or a glucose monitoring system
or whatever else, it will really take us a step
forward and ahead of a lot of other countries if
we can manage to get there.
Speaker 2 (51:13):
Yeah, one of the barriers he mentioned there, which I
was a little bit surprised about, is massive concerns about liability.
So we might need some regulation and compliance tweaks to
give health providers the confidence to use AI and not
be worried about getting sued. I thought that would have
been a bigger deal in a place like the US.
(51:36):
And actually last week was a visiting expert in AI
came through New Zealand and she told me exactly that
she's working with doctors and radiographers and that using AI
and machine learning to try and improve testing and analyzing
test results. And she said the senior doctors in particular
(51:56):
are really pushing back because they're worried about getting sued
and getting their clinic or their hospital suit as well
by making a wrong diagnosis. So that's clearly an issue
here as well. So having that confidence to be able
to put some trust in these systems is going to
(52:16):
be key, and maybe we don't have quite the regulatory
environment to allow that.
Speaker 1 (52:22):
Yeah, just putting those policies in place so that you
know it's not going to be ramp and AI and
making misdiagnoses all over the place. They are going to
be in conjunction with good diagnosticians as well, so they're
being checked and human in the chair and all that
good stuff.
Speaker 2 (52:37):
Yeah. And the other interesting thing that will point it
out when he goes out and talks to hospitals and
clinicians and that sort of thing, a lot of them say, no,
we don't actually have a policy or even a strategy
around using AI. And this is I think we see
just about in any area of industry and business at
(52:58):
the moment, people immediately jumped to the use case. So
I haven't got a strategy for it, but I know
my customers want this. And when it comes to healthcare,
it's using generative AI to surface experiences for patients. So
giving them a chatbot that they can query about sensitive
(53:18):
health related issues and they'll get reliable information. It's that
sort of stuff that actually the health sector thinks is
the frontline of the generative AI revolution.
Speaker 4 (53:27):
Yep.
Speaker 3 (53:28):
It's a big revolution and it's coming. We've just got
to make sure we're ready for it.
Speaker 2 (53:32):
Absolutely. So thanks to doctor Will Reedy for coming on
the Business of Tech, and we'll have more episodes coming
in this ongoing series about how AI is impacting various
sectors and industries across a tea.
Speaker 1 (53:46):
Show notes in the Tech section of the Business Desk
website and the podcast is available on iHeartRadio, well your
favorite podcast platform.
Speaker 2 (53:55):
Let us know what you think of the show and
drop us a line with suggestions for future gare email
Ben on benat Businessdesk dot co, dot MZ, or you
can find both of us on LinkedIn and x.
Speaker 1 (54:06):
Next Thursday, we'll be talking mergers and acquisitions and the
complex technical decisions relating to digital infrastructure that need to
be made when businesses come together.
Speaker 2 (54:16):
Until then, have a great week.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com