Wow that was a bit of an introduction
We we've had
quite a bit of what and how and
I thought I'd give the CIO's perspective
bit of a confession first
Before I started doing this crazy day job. I'm actually I suppose always will be a software engineer, so I'm actually obsessively
Interested in the what and the how
But you know as a CIO, it's not about the what and the how it's all about the why initially
So I thought I'd just spend a little bit of time explaining the thought process we had to go through to set out our strategy
Because if you can't make the case for this type of approach
To integrate a digital care record you're never going to get into the what and the how
And you know it'll be hopefully useful to other CIOs so Derriford hospital. I'm the CIO
director IM&T here
still I
Think one of the largest single-site hospitals in Europe one massive block of concrete doesn't win any architectural awards at all
But it is in a very very nice place and very functional because it's all in one place
We're pretty much typical of an acute trust I think
talking to colleagues, so we're about 900 to 1200 beds depending on what you can as a bed
We've about 6,000 staff. We're a major trauma center, and we're a tertiary center, so we're pretty normal acute trust
We've got pretty standard IT I believe
We've had a Best of Breed Interface strategy
we'd buy better preclinical products since 99 when I joined the trust and we interface them together with HR 7
and standards
We've replaced up housing 99
We've got an open standard interface engine we use into systems ensemble for those indices with people in the crowd
I've done your plug. I want to reduce license fee next year, please.
We've got fully rolled out order comms.
Pack summaries, I'll talk about a bit later on we've done three times
And we've got something unique called Salus, which is home developed. It's a patient flow solution
that becomes a bit more interesting later on where I'll talk about what we're using that for and how it will benefit others.
And
Thomas said we've got a 170
departmental solutions, you've underplayed the problem. We've actually got a 190
departmental clinical solutions, so these are wholly independent
proprietary clinical applications
by specialty, and they say I don't think we're unique. I think most people are nodding that's pretty sort of normal, I think.
And this is the diagram that
freaks me every time I see it. This is actually our interface architecture, and you can see at the top
July 2017 so that's pretty current so the level of complexity
to do just what we've done and actually we've still got 190 per spoke solutions, it's incredibly complicated
So that's where we started from
We've got a few issues other than complexity
We're wholly locked in
We can't change many of those solutions
because we're wholly locked in and we find ourselves in a really strange place.
I'm sure this will
some of you will actually pick up on this one, I'm sure
We buy licenses in perpetuity
To make software read-only because we can't migrate their data when we buy a new one
and we do that a lot if I'm honest
That's a big big problem for us
proprietary vendor locking because of the proprietary nature of the data
Note to migration nightmare and it's there's no strategic fit at all if we're really really honest
so
we're starting from a place where we know we've got to move
and where are we trying to get to?
There's quite a few people here of my age, so you probably remember in old money we had
EPR one to six, everybody's trying to get to EPR six
Nowadays, it's HIMSS level 7
In EMR adoption model we're all trying to get to 7. That's our aspiration
You know we don't do this sequentially. We're probably around 2 at the moment
We've pretty much done PACS three times, we've got e-notes. We've done observations,
order coms, so we're starting to do this in pieces
OPENeP I'm going to talk about in a minute
So what we're starting to get towards that level 6, level 7 a bit at a time
But there's one real kicker. There's one thing that we can never nail on our current strategy, and that's level 4
Clinical decision support, proper, true clinical decision support, connecting those 190 solutions
in a real way that makes absolute sense
and enterprise-wide scheduling. So that's the type of thing that if a doctor on the ward
requests a specific test and it comes back and says the patient's allergic to an antibiotic
it goes off and tells pharmacy and pharmacy prescribes another one
That's the level we're talking about real-time decision support.
I don't know anybody worldwide. That's managed to achieve that with the architecture we've got
and there's one fundamental reason that would be
Unbelievably complicated if we try to do that in that architecture.
Even as a software engineer, I don't believe that's possible and I've never ever said anything is impossible as a software engineer everything's possible
But I don't believe we would ever get there
And the fundamental reason, I believe, this is this is my take on this is that to achieve that you've got to have a truly
integrated single digital care record. You've got to have that data in a single integrated form
to achieve that level of interoperability for enterprise-wide scheduling and real-time clinical decision support
Couple of ways you can achieve it
You can buy your way into it
So you can go down the "Big Box" solutions. You can do the EPIC's, the Cerner's, the Lorenzo's, the TrakCare's
There's you know, as well as I do, the Allscripts has lots of them
very, very proven, it's a very proven model
you deploy it for the
level of functionality it provides you, you pretty much get benefit out of the box when you start, so it works
and and you've got a consistent user interface. It's one system for the functionality it provides, and you heard earlier
it doesn't provide everything obviously. A couple of cons however,
boy, are you locked in!
You know EPIC for one, and I'm sure the others will tell you the same. They make a big play about the fact
they've never lost a customer
It is a good system
but you know if you actually take all the effort to put in something like EPIC or Cerner or Lorenzo or whatever
would you really want to migrate away?
So true, but you know you are wholly locked in and you've got that same data migration nightmare
and as we know it's incredibly expensive
Mostly probably followed Gartner reports. I read a couple of weeks ago about the US actually that they've now cited
implementations of Allscripts, Cerner and EPIC, I'm not singling any out here,
as "eye-watering" the costs, the level of costs. Now ten years ago, they were citing these as a model of how to do it
They've suddenly realized now billions and billions of dollars are spent on regional implementations
That's probably not surprising to anybody, but here's the thing
they're no closer now to having a truly integrated digital care record for the patient than they were twenty years ago
Because none of them interoperate in the states
21st century Cures Acts which you're probably familiar with it's probably going to
make that change over the next few years at an interoperability level
But if you're a patient, and you're at the center of a region where you've got an Allscripts
and you've got an EPIC and you've got a Cerner you will have three patient held records
You won't have one.
And to me that's an absolute frightmare, if I'm really honest, and that's not where we want to go
I know we're a little bit behind in terms of implementing big boxes, but I don't want to be there at all.
So we wanted another approach
So we thought okay if we can't buy our way in, can we build our way in?
So if there was an open standard framework where we could
set out an aspiration of true
vendor neutrality at every level of the stack
wouldn't that be something sensible to do?
So that was kind of our vision. It fits with our strategic thinking
we've always had that best-of-breed interface on standards approach, so it kind of fit.
We could do it a bit at a time. We wouldn't have to try to make this case this huge thing and do it all at once and
a key one for me is, we would start to reinvigorate
that small to medium enterprises marketplace that the national program killed stone dead
You know those of you that used to go to Harrogate like me saw the marketplace pre-national program
And if you go to the NEC and see it now
it's wholly different, and you know the national program did that so we want to really get that back.
We want to reinvigorate the SME marketplace and really start to get this ecosystem built
Once we've done it, we've got no vendor lock-in
Because our data is completely abstracted
We'll have no data migration issues. We can swap pieces of functionality in and out and
everything will just carry on working as it did before and
Incremental investment is both less, and you don't have to make these massive investment cases
to an NHSI
You know try taking an 80 million pound investment case to the NHSI... good luck on that one
It's not something I particularly would like to do
So it does give us the ability to incrementally invest year-on-year business cases year-by-year to do pieces of the puzzle
So we've ruled that one out and we set our strategy on an open standard approach.
This is interesting. How do you do one bleep on two strings?
So bottom right in that diagram
The bottom layer is the data layer
The bottom right actually says PACs, hopefully
We've changed PACs three times in the time I've been at the hospital
we had Agfa for pre-national program
the national program came along and asked us politely to change it, so we changed it to GE
and then the national program kind of came to an abrupt stop
as did the contract, so we had to do something else, and we've now got Insignia
So we've had three completely separate pieces of software for clinicians to use and you know what?
Wholly seamless. We trained clinicians to own a new product, and they just carried on as they did the day before simply because
it's got a data standard DICOM been there like forever as far as I can remember
So because the data was holding it in a data vendor neutral format
we could simply swap out one solution with another solution, train our users
and they just carried on as they did before and that was kind of the lightbulb moment for me
that was the
wow, what if we could do that on the left-hand end of that bottom stack which is the 190 solutions
If we could get to a point that each one of those has
unique functionality, fine, but had a
non-proprietary
open data model
we could then compete supplier against supplier,
swap software out and not have any lock-in at all. So that was my kind of lightbulb moment, all sat under our nice
open architecture.
For us, it's Ensemble, but it could be anything you like really as long as it's open standard based
We've had standards there for years. We've had hl7, we've had ITK remember that?
We've had IHE, now we've got FHIR and FHIR isn't
a competitor for openEHR if you want my opinion, I know there's a lot going on about this
They're totally and utterly complimentary to each other
discuss
Then we've got an application stack across the top, all delivered through Salus, which I showed you earlier
That's our view. We want it all delivered through a common portal to any device
So that's kind of our aspiration
Then we had a bit of good luck
We actually went out to procure an electronic prescribing solution. We went out to OJEU
We got right through the end of it
And we actually didn't shortlist anybody because we didn't find anybody actually that was capable of meeting the full requirement
That's not a good luck. The good luck is
CGI who we were working with at the time actually introduced us to MARAND and they've shown us their
OPENeP product, and they've shown us their
view of the world I guess their vision for how this should all hang together and
they're really alike and honestly we did not
rob either
We certainly ever said to Thomas before we certainly didn't rob theirs, that's for sure. These were created totally in isolation
But you can see a complete similarity and when we when we actually met with these guys we looked at the product
we looked at their philosophy and the openEHR platform the Think!EHR
that they developed and their ethos around open standards. There was a there was a unique fit there for us
We then said okay, we'll take OPENeP as the first application on that application stack
Based around an interoperable framework with the openEHR platform as the vendor neutral data layer
That would be our first line strategy
Where we are now? We started that
January? April? It seems a long time ago now, so we've been doing this
Yeah, you nodded, right about April time
We're first of type for OPENeP in the UK
It's open source
held in the Aperta foundation
Not-for-profit organisation which is fundamentally us, so we own the source
All set up nicely for us by NHS Digital, thank you very much Peter.
and
it's based on
the Think!EHR Platform on an openEHR repository, and it's going to be delivered through Salus, absolutely happy days
We go live in June next year. We have the test environment in place at the moment, and it's looking really really good
I've got to say
So all things are starting to really start to pull together with it
As I say the Aperta foundation is the means by which we're going to hang all the source together certainly
for OPENeP to start with
But other things that come which I'll talk about in a second, and there's a key thing for me
I'm not standing here saying we're trying to do away with commerciality
I think a lot of people see this as what's in it for the software supplier,
what's in it for the commercial partner and I get asked it a lot I've got to be honest and
what I try to say is
We we need relationships with software innovators. We just need a different type of relationship
What I don't want is
Licensed software that's licensed by site, by named user or concurrent user
specific to me for X amount of money
What I want is something a little bit more fluid
Something that we can create an innovation partnership around. For argument sake and this is just throwing it out there,
why wouldn't we license on patient rather than named users, sites, services and concurrency?
So that for me instead of having my remit stopped at the end of that concrete block
My remit now will have to go into the whole of Devon because we've got to interoperate across the whole of Devon
I want to deploy software out of the hospital into the community, into primary care, into other acute trusts
It's all the same patients. Why should I why should we pay again?
We need a slightly different model because when we've looked at our existing framework,
even if that software was capable of doing the whole community piece
you know, what we couldn't afford to deploy it based on our current license model
And that's just madness
So there's a whole lot of work that needs to be done
to work with our commercial partners to actually set that up.
Very early days, we don't have all the answers
But the software
certainly SMEs and some of the bigger players are working with
very, very interested in having that discussion with us because they see the future and the future isn't "sell us a licence
based on X number", that's not the future
So this is Salus, just to give you a quick flavour of what we've what we've actually done
This this is the piece we've written and you can see,
we probably can't actually it's not particularly clear,
there's a little tab at the top that says "meds" which was unique to this project. When they click on meds,
in context, this piece down the bottom is actually OPENeP, the MARAND product, and
my developers and the MARAND guys have been working joined-at-the-hip
And they've actually seamlessly integrated these two; we're not talking interfacing now. We're talking absolutely bolted these things together
at the, I guess, the doing level rather than the interface level.
We're not passing messages, I think was one saying so there's no joins whatsoever with this
It all appears to the user like it's Salus which is great for me because I get all the credit
but behind the scenes
MARAND have put in the OPENeP clinicality, that really safe clinical product
So where from here?
If you remember the application stack, we've done one, we've done OPENeP
We want to do more now all based on that same stack
We just want to roll applications out as quick as we can get our hands on them. We want order communications, we want electronic observations
anything you like; whether it comes from MARAND,
whether it comes from an open source community, whether it comes from other trusts that are involved in this game now
We want as much as we can get and we want to deploy as soon as we can
To start to build this this whole ecosystem up
We're looking to make Salus which is the product we brought to the table
open source through the Aperta foundation and NHS Digital so that for argument's sake our implementer is CGI
it could be one of a number, but we're using CGI
We want them to be able to come to another trust and for a fixed price deploy OPENeP
And if the trust haven't got a portal they can deploy Salus as well
And they can have all the benefits of bed management, patient flow and all that kind of stuff
it should be as simple as that. There's a fixed price model for a
commercial support partner to just come along and say you want that one, you want this one
If they've got an existing portal deploy OPENeP and integrate it with that portal
That's where we're trying to get that that framework set up
And lastly, but for me probably the most important,
we're all pioneers at the moment, you saw the nice diagram earlier. I wish I had that.
We're very much at the forefront of this, a lot of people don't even understand this at the moment and one of the things
we've got to do, I think one of our biggest challenges, is how we
we get out there
and we both evangelize it and we build this ecosystem; this thing that's going to take it from quite niche,
which is where it is at the moment,
to actually being fully mainstream such that it can challenge the EPIC, the Lorenzo, TrakCare and everything else
And for me involves a whole blended approach really
We need to invest locally in our own development teams and make them work differently because they're not going to be writing everything
We need to certainly invigorate local SMEs
and we need to be
connecting with bigger partners, for us people like education
when I was a kid in Plymouth
the school gates used to open in June and the dockyard gates used to open and kids used to flock out from school straight into the dockyard.
Now what we want to do is do that with with education. We want people graduates coming out of
University in Plymouth and spending 18 months to two years working with us
developing apps, building this ecosystem
Giving themselves a CV.
In two years they're going to go off to the industry they're gonna come up and work in London and make a fortune,
but for that 18 months to two years we want to get that vibrancy locally
with those those
very skilled people that can write these apps for a pastime if you put a little bit of rigor around them,
a little bit of structure
and
national global developments. Things like this, working with other trusts; we've got Salus you can have it for free
You've got a business analyst tool, we can have it for free. Moscow have written this fantastic thing, they'll charge us I'm sure,
but we'll come up with some deal
That kind of arrangement and the reason we can do it is because it's all based on one
open data record
Thank you
And II just as good as last time, thank you
We have a tech. We are a little bit over. But not too bad
thanks to all our speakers for keeping well to time and
You know it was a pact a pact program any questions before we break to break for lunch
We're hungry people
Somebody ought to have a slot before lunch
Yeah, we'll all be available to chat. Please go and enjoy lunch, and we're due to be back here