Hello, and welcome to Talking digital health, a podcast by law firm Mills & Reeve, hosted by health and tech lawyers, Sophie Burton-Jones and Charlotte Lewis.
Our aim is to discuss key topics of interest to those procuring tech in the NHS and those selling tech to the NHS.
I'm Charlotte Lewis. And I'm Sophie Burton-Jones, and we're your hosts for this episode.
We're delighted to be joined by David Hancock, an independent consultant working with the NHS on the definition and implementation of interoperability standards.
Hi, David.
Hi there.
Great to have you.
In this episode of Talking digital health, we'll be talking about digital transformation in the NHS and specifically the drive to improve interoperability and sharing of data between systems and organisations within the NHS and the wider health and social care setting. So first of all, could you explain what interoperability is for the layman, please?
Yes, certainly. I think the best thing to do is to relate it to a story from a completely different industry. And in fact, time. In 1871 in Chicago, and in 1872 in Boston, there were two great fires and they needed help from other fire crews from outside of the city. And when those fire crews came to help, they couldn't because of the size of their fire hose couplings were different to the size of the fire hydrant. And that is an interoperability problem.
Yeah, I mean, when you say it like that, it makes perfect sense.
It's that sort of mic drop moment. So then thinking about the NHS. The NHS and social care have been attempting to improve interoperability for over 25 years.
Why do you think progress as being so limited?
After all, for clinicians, it obviously remains one of the top three issues that they say contributes to burnout and threatens patient safety.
Yeah, well, I'd like to continue the analogy of fire hoses, because there was another fire in Baltimore in 1905. And at that time, there were 600 different sizes and variations of fire hose couplings and hydrants. And so in 1905, as a result of that fire, there was a definition of a standard for hoses and couplings known as the Baltimore standard. And the hydrant and hose couplings were set at two and a half inches or 64 millimetres in new money. But a hundred years later, after the definition of that Baltimore Standard, only 18 of the 48 most populous American cities had the national standard for fire hoses. And indeed, the Oakland firestorm of 1991, its severity is put down to the fact that in Oakland, they had three-inch hose couplings, which meant that other crews, again, couldn't go and help. So you can see that there was no technical issue stopping people standardising on those couplings. And so clearly, there was another reason, and that reason is actually because it's a social problem.
Because in order to be able to solve that, cities needed to make a decision to change their couplings and their hydrants. But they chose not to. And who knows why they chose not to?
Perhaps they couldn't. Perhaps it was too expensive. But if it actually mattered, really mattered, they could have found the money. You know, so if it matters, you can find the money. And clearly, it didn't matter enough to them. And so that's why it's a social problem. And if you look at it across the NHS, it is actually the same social problem.
Clinicians say that interoperability is one of the most important things that they need to improve because if there's the not enough, a good enough flow of information between systems, there are technical standards already in place. So why aren't they used? And the reason is multifold. So, for example, if there's a standard to be able to send information like a transfer of care, which has essentially been mandated since 2019 for trusts to be able to support in order for them to have their license to practice as a healthcare provider, how many have actually done that.
When you actually consider that there technically are ways to be able to do that today. For example, I know that Nerve Centre, the UK electronic patient record company, satisfies all four transfers of care, yet nobody has asked them to implement it, and nobody's got that on their roadmap or to be implemented.
Similarly, InterSystems implemented the transfer of care into their integration engine. And what that meant is that it didn't matter whether trusts had all the information in one system, they could be spread across multiple systems. And all they had to do was collect that information from multiple systems and put it into Ensemble or Health Connect, the InterSystems product. And once they got that in there, the InterSystems solution could then package it up, put it in the right format, in the right standard, send it out across mesh into the NHS infrastructure.
It would work out from the NHS infrastructure, would tell which GP practice it had to go to, and it could send it and make sure it ended up in the GPP system. Why didn't anybody use that? Well, it's just it was too difficult.
All the data that was had to go into that, many trusts didn't actually collect all that data.
The data that they needed to be able to collect and to move was quite often in very old legacy systems, which themselves had very, very poor interoperability, which meant you couldn't get the data out of it.
The trusts, just in with all their other priorities, could not prioritise to be able to do that because we know trusts have huge numbers of other priorities around focusing on elective, there's cybersecurity, there's sorting out what's going on at the front door with ambulances and there's loads and loads of different things which change over time. So it's just not prioritised enough, which again is a social issue.
In order to be able to do this, it will require change in clinical processes to be able to collect that information. And clinicians are already massively overstretched.
All trusts are running hot. So have they got the time to be able to do this transformation and do this clinical change, to be able to collect that data and to then be able to send it across and collect it accurately? For transfer of care, this is very often ends up being done by a house officer or a senior house officer, not the consultant. And so have they got all the information and the skill to be able to put that information in? Can they get the information correct? And then when they send it, it becomes very difficult because if they might send it and they're so busy, they might go, oh, did I send that? I can't remember. Oh, I'm going to send it again. And then GP practices end up with three or four versions of the transfer of care, then they themselves don't know which one's correct.
Or is it the same one as I got one before? So that's complicated for them as well. So you can see all these things come together to actually make interoperability really difficult. And it's not really a technical problem. This is why adoption is really, really hard.
Yeah, and why people have been talking about it for a long time.
Yeah, correct. And there are other things which contribute to this.
I mean, there is a construct in sociology called the collective action dilemma, and the collective action dilemma is where organisations realise that they need to be able to collaborate for the greater good. But there are, inside each individual organisation, there are other things they have to do which take higher priority, which mean that they cannot contribute to this collective action dilemma and do this thing for the good of all the organisations. And again, the transfer of care is a great example of the collective action dilemma, because if you talk to consultants in a hospital, they say, this is really, really important I get the transfer, if I could get the transfer of care out, because if I want to make sure that the patient gets the ongoing care and the GP knows what they need to do to be able to look after this patient. But it is so difficult for them to perhaps, you say, get that information together, do that clinical change, and with all the other priorities they've got, that they can't, they don't actually do it and they can't do it. So there's many things to consider and the clinicians want interoperability, but they actually can't do it because of the problems going on inside their organisation.
Yeah. I guess we've started to talk about it anyway in terms of what could interoperability do for people across health and social care, but do you have any kind of specific thoughts on what the role could be of interoperability?
Absolutely, because interoperability is about getting information to flow between organisations.
Now, if you want to be able to do digital transformation and change processes, you're going to need to change the information that is going to flow, and you're going to need to be able to change that between different systems. when you want to define your interoperability requirements, you say, what's the business process or clinical process that I'm going to follow?
What are the systems that are going to support that? And then for each of the systems, what is the information they require? And if you're going to do digital transformation, you're going to change clinical workflows, clinical pathways, and what have you.
You're going to change the required information flow and information that's required to go between different systems. So in fact, it is very difficult to do the transformation.
You want it, for example, in an integrated care system without solving interoperability because you're changing the information flow.
That is why it's so important.
That is at the core of why interoperability is key to the transformation.
Thank you, David.
We're going to pick up now on interoperability standards, data and the impact of AI. So, David, are there national interoperability standards across the NHS? And if so, are they mandated and enforced?
There are interoperability standards that have been defined.
You know, so we have the transfer of care, for example.
There are also other standards which have got their DAPBs. So there's one around interoperable medicines - to be able to exchange medicines information. And that is becoming more sophisticated when you want to add things like dosage instructions, which is quite complex. And that, again, is struggling to be adopted when you've got lots of different EPRs. needing to be able to support that.
There are specialist, you know, EPMA solutions which have been able to support that, but that's because if it's the only thing they do is the electronic prescription and medication administration.
When you're looking at other solutions, you know, EPRs, there's so many other priorities that they've got to do and that their customers want.
It's sometimes just become the most important thing that they want to do.
Now, there are other things that are being developed.
Now, we all might have heard of this interoperability standard called HL7 FHIR, which is Fast Healthcare Interoperability Resource. And this is a exciting development of standards because of it tries to solve the problems that we've had with standards before. And the other thing is it brings healthcare into the age of the internet because it actually works in the way the internet works, HTTP, using REST APIs and what have you, which means that you can bring healthcare, say, into the internet age, so you can share information technically more easily.
Most graduates coming out of university, having done computer science, actually understand the base technology, which is in HL7 FHIR, unlike having to learn quite obscure standards when you have things like HL7 version 2 and version 3 and what have you. So that's also an advantage. So there are standards being developed using HL7 FHIR, but they're the problem that we've got is that these are being defined by a central NHS England Interoperability Group. Now these people are doing a great job.
They're talented people working very, very hard, but there's not enough of them.
You can't have a central team defining all the interoperability needs.
You need across the NHS because of different parts of the NHS and social care have different priorities depending on what they want to do.
What we need to do is actually find a way of democratising the development of standards. So we do it bottom up rather than top down. And that is quite a cultural change that needs to happen. And that is something that we're trying to work towards. But within these standards, it becomes vital that we're going to cover the ground fast enough. So we need to be able to do these bottom up, but with an overall control to make sure that they are consistent. And when they're defined in one place, they can be adapted and they can be used in other parts of the service.
On that point, in terms of democratising from the bottom up, do you think it's actually going potentially the opposite way at the moment, just with the changes in NHS leadership and news about NHS England size being reduced or being brought into DHSC, there's a risk that it's actually going to go the opposite way to that?
No, this is the biggest opportunity we've got to change things, is that they absolutely are not going to have the capacity and the resource to do this. And we've got, obviously, this legislation coming down the line, which is going to mandate suppliers to meet certain interoperability standards. And that is in the belief in NHS England that the problem why they haven't got interoperability is because of suppliers not being prepared to do it.
Now, I hope in what I talked about earlier is that it is not a supplier problem per se.
Suppliers will not develop functionality which their customers won't use.
They will do things that their customers actually want.
Suppliers have finite development capacity and they will not develop stuff that their customers aren't going to use. So that's why, again, it's this social problem. And we need to be able to work towards that. And this is why democratising and decentralising it is the only way to succeed. And I am probably for the first time in maybe nearly 20 years that we've got a chance to do it because of the way in which NHS England and the DHSC are being cut by 50% because they know they haven't got the capacity to do it.
Well, in light of that comment, then, given the increasing digitisation across the NHS and EPRs implemented across organisations, is this actually reducing the need for interoperability as EPR systems replace many separate systems?
So it is not going to reduce it at all.
I suspected you might say that.
Because even though you're going to have systems which are going to create and have a larger scope, providers are still going to have tens of other systems which are critical to the operational processes going on inside a provider. So there is still a need for interoperability between other systems and the EPRs. And if you've still got this data, particularly in old systems, there is still the same problem of how you're able to get that data out.
Are there the APIs to allow you to do it?
These old systems will not have interoperability delivered because of their legacy, and suppliers are investing in their new systems. So they're absolutely still interoperability. And as I said, when you're looking at re-engineering clinical workflows, you are looking at change to flow of information across providers, and they have all got separate systems.
Even if you've got to provide a collaborative, who are standardising on a single system.
Patient flows, you know, are still very porous across ICBs. So even if you've got something in a provider collaborative, there is still the same need for interoperability, both across providers, across acute and community and what have you, and also across the ICB, when you've got patients flowing into other ICBs. And just moving on to AI briefly, because you talked about new systems, new technologies, does increasing use of AI tools, do you think they're going to support or hinder the aims of easier access and data flow and high quality data being shared? And what have you been seeing on this? So AI, requires high-quality data.
If you're going to be able to develop AI solutions and make sure that you are not introducing bias, if you want to be able to keep them currency, you're not going to get a drift in your AI models and all these things, you need high-quality data and you need it, say, at the scale that you need to manage your population.
That can only be done to interoperability to be able to get that information together.
The other thing we have to bear in mind, though, is that a lot of the AI that is being implemented in trusts, this is no criticism of anybody, anything they're doing, but it's very difficult for them to scale because it is being done, you know, solving one problem in one particular department or what have you, and they're not considering or they can't say that how it would scale. And if you consider that when you've got trusts using things like large language models, large language models themselves are pretty temperamental. They hallucinate.
You put the same question into a large language model, like a ChatGPT or what have you, you could get different answers for the same thing. And then even if you do get an answer, there's little traceability because like a ChatGPT just takes you to the document and doesn't take you to exactly where in the document. So this has difficulty in scaling, but you still need all that data to be able to train it.
There are many obstacles to overcome.
They will be overcome in time as we learn, but that's going to take many years. But you're not going to do that without interoperability.
Absolutely.
Final question, just to wrap up what has been an absolutely fascinating conversation, actually, on obviously what could be a very, very technical point, but you've really brought it to life. So I'm really grateful for you for doing that, David.
Just thinking ahead then, so what can organisations, suppliers of digital technology solutions and regulators do collectively to improve the adoption of interoperability standards in health and social care?
So you've got to realise that this has to be a community initiative. Every single organisation that has got the problem has got to be part of the solution, which means that, you know, we have to have the providers and clinicians working with suppliers, working with standards organisations, and working with the regulators to be able to come up with that solution to essentially what is a wicked problem. And we also know we can't do it top down. So in order to make this work, we need to have all those parties who share the problem to work together. And, as I said before, each organisation, for example, each ICB, ICS, has a different priority, a different problem. So we need to be able to have the ICBs working with providers, working with suppliers together to solve that problem. But if we're able to solve that problem and document it in a way which means that not only is it implementable, but it is shareable so other parts of the service can then pick it up when they need to and use it themselves to save the hunt to reinvent the wheel.
That's how we can cover the ground quickly and to be able to deliver what we need for health and social care.
I'm going to just ask you one final question, but to touch on that, is that work being done, that collective community work? And if so, by who or what organisations?
So, we do have suppliers working with their customers to be able to solve problems. The fact is though it's not considered to be part of the community. And so they're solving their own problems and they're not able to share or they're in fact developing it and not providing the necessary or necessary documentation to make it shareable.
That's why we've got to be able to define this standard definition of done and make sure that that is shared.
That again is really important, but with a agreed way and a commitment to do that across all sides, we will be able to achieve that.
That's great and thanks very much for your time today, David.
It's much appreciated as Charlotte says.
If you'd like to hear more from David, then you can contact him directly via LinkedIn and at Mills & Reeve we're very happy to discuss any legal queries you might have around NHS digital transformation, including on covering interoperability and data requirements in your procurement and or in your contract. And our individual contact details are available on our website at mills-reeve.com.
Thanks for joining this episode of Talking digital health by law firm Mills & Reeve.