A recent HTN Now panel discussion focused on the question of how health and care can tackle interoperability, with HTN joined by Kate Warriner (chief transformation and digital officer at Alder Hey Children’s NHS Trust); Chris Johnson (chief medical information officer at Royal Papworth Hospital NHS Foundation Trust); and John Kosobucki (CEO and founder of OX.DH).
Providing some more context to her role and organisation, Kate highlighted how Alder Hey is a “really digitally mature organisation” and added that they were the “first paediatric trust in Europe to achieve HIMSS level seven”. As a children’s organisation, Kate noted that there tends to be a high expectation from service users and their families around the levels of digital offered by the trust, and added that they have played a “really active role” from an interoperability point of view in their local system.
Alongside his role as chief medical officer, Chris is also a respiratory consultant at the Royal Papworth. “We’ve done quite well with our digitisation – we are HIMSS level five at the moment with the systems that we’ve got. To achieve and maintain that we have had to do quite a bit in the interoperability space, both internally with our own devices and EPR, and externally, notably with our shared care record, provider for our lab systems, and GP Connect.”
John explained that OX.DH is a health tech spinout from the University of Oxford which started out around three years ago with the premise of bringing the digital technology experience from other industries into healthcare. Starting out by exploring how tech can support in assisted reproduction. Working with the NHS OX.DH has since been onboarded onto a number of frameworks, including the tech innovation framework focusing on bringing new technologies into primary care.
Challenges around interoperability…
Kate referenced Cheshire and Mersey’s shared care programme, designed to share information across the local system. It has been deployed over a number of years and has a lot of organisations connected into it, with 380,000 accesses of the system every month. It includes connection with a neighbouring integrated care system, which Kate called “really important in terms of the flow of patients we might receive – working across organisations and across ICS boundaries.”
The cross-boundary aspect can pose a challenge, Kate acknowledged, as can the need to win hearts and minds, and the need to focus on usability for our clinicians.
“We’ve got quite a bit of local complexity – we neighbour Wales for example, but we can’t seem to connect with Wales,” she added. “I would say another challenge sit around standards; we’ve got multiple suppliers involved in our systems, and managing maintaining standards for them has been quite challenging in our local area.”
Chris shared his view that the interoperability projects he has undertaken have been “the hardest things I have ever done in my career, whether that’s medical or digital. They’re very complex programmes. They involve a lot of stakeholders and a lot of knowledge; they are difficult to cost at the outset; they are difficult to keep to time, and that can cause further problems not only with costs but with expectations.
“Also, the space is so technically challenging, and that means there are multiple different ways to connect. We are steered towards FHIR, but HL7 is the reality for a lot of systems and is probably the appropriate choice for many use cases when it comes to interoperability. There are questions around whether we are persisting data or not persisting data in the downstream or the receiving system, and they offer fundamental challenges to how you connect and also with regards to the information governance around it.”
Chris raised some other technical challenges such as database structures and how they are modelled, and the “fundamental concepts which can be different – that’s the sort of thing that really trips you up when you are doing LIMS connections, for example. Different concepts can be shared by different LIMS systems or different EPRs, often depending on their origin, which country they are written in in the first place, and so on.”
Testing is another challenge, he continued, with multiple different scenarios in need of checking. “It’s really important that you are not just testing A to B, you are testing all the implications of the metadata. There is a lot of information that can be transmitted in an HL7 message or in a FHIR API call, but it’s not true database-level integration, and sometimes there is metadata missing. Sometimes that metadata can be clinically very important.”
The final challenge Chris highlighted is that of maintenance. “That could be an upgrade in either system that gives you a testing burden; it could be day-to-day maintenance and troubleshooting and monitoring.”
John commented that OX.DH’s relative newness can be an advantage in this area, as it means that the company does not have to find ways to deal with legacy technology debts. “But you still have to deal with the reality of what’s out there. You need to look at the history, the reality and the desired future state that we want to move ourselves to; which is of course a completely interoperable environment with integrated APIs, which allow any service to publish and subscribe, and knowledge of where that master data resides.”
He said: “We certainly do encounter challenges when we are working with other vendors. Some have the mindset that we want to control our data and everything within our own world, for security or integrity or other motives. But everyone needs to be adopting a more open, data-sharing model.”
… and tackling them
“How are we trying to tackle these challenges is the key question here, and it comes down to meticulous planning,” said Chris. “As John said, understanding the desired future state is really important. It’s about having a target model.”
In his experience, when his system has tried to do this, “it feels a little like we have been tripped up. That’s where we have disparities between developmental roadmaps of products, disparities between vendors, and the like. It can lead to compromises in the technology that we use, and a difficult mixed economy in your systems. That can make the application of business analytics, for example, really challenging. You might want to use data for population health planning or research, but sometimes your data is there and sometimes it is not. But I think pushing for that future vision and really putting the time in with planning is the way forward.”
At Alder Hey, Kate said, there are a number of projects and programmes aiming to tackle interoperability. “Internally, there is an example of an electronic anaesthetic record which is a specialist system in place for patients undergoing theatre procedures. It has enabled us to bring our drugs and vital signs into our electronic anaesthetic record, reducing the need for what used to be a big piece of paper with lots of manual inputting. That then integrates into our EPR.
“Coming back to the point about winning hearts and minds, making people’s jobs easier for them is such an important thing to keep in mind. I think integrating with your day-to-day systems is the best approach, because if you can make your interoperability feel seamless, it just becomes part of business-as-usual for your team. The end user isn’t needing to log onto multiple systems. Our shared care system integrates with our EPR, for example; so with a single click of a button from within their usual system, our clinical colleagues can access information from other settings such as primary care.”
With regards to the challenge around working with different suppliers, Kate suggested that there is a responsibility on NHS organisations to be as clear as possible on their strategies and vision, so it is understood where expectations lie from both sides.
“When it comes to standards, I believe that simplifying things as much as possible is important,” she concluded. “Maybe we even need less choice; a smaller, simpler set of standards. There has been some work in this area already, but I think there is more we could do to improve.”
John highlighted how OX.DH partners very closely with Microsoft in order to take advantage of the their assets and the national tenant. “More and more people are starting to use Microsoft assets for purposes beyond simple collaboration and communication with colleagues,” he said. “The tools that are natively available in the Azure cloud that can be used for a lot of these integration projects are just starting to scratch the surface of what’s out there, and that’s very much an area that we are focusing on – we have standard adaptors and end-points that understand and speak HL7 and FHIR.”
Requirements from a central perspective
“Resources are needed – the NHS is in financially challenged times as we know, and this area can be something that gets left behind,” said Kate. “In years gone by there have been a lot of funds around interoperability, but when you get into business as usual mode, it can be difficult to keep it up. Operationally, you have a lot of competing demands. Some reflections centrally in that space would be good.”
Kate also mentioned the links with other nations in the UK, with reference to how her system borders Wales. “It’s an issue for us, and it’s an issue in other parts of the country, for example the northern systems bordering Scotland. I think when it comes to our national strategy, because we are NHS England, we are not thinking about the UK as a whole. But we do see patients from the other nations, so I think some central strategic support on that would be really helpful in terms of patient flow.”
With regards to the NHS App, Kate noted that there are questions around the long-term future there; about how the app will or should be used in the future and how systems and information could be better connected through it. “It feels a little congested at the moment,” she observed. “I wonder if there is work to be done centrally around the architecture, to help join things together a bit more.”
Chris came back to the point around funding and spend. “There are a couple of things around funding for IT that need to be thought about, these days. There’s the old-fashioned model of spending according to turnover or bed number. But it’s totally irrelevant in the digital space. Computers don’t care how many beds you’ve got; but they do care about how many systems you are trying to connect. The level of support that is needed to maintain all of these connections of systems is hugely time-consuming, and I think that needs to be reflected in the approach to funding.”
Additionally, Chris mentioned the debate around funding coming down from the national source and how much of it should sit with trusts. “The problem is that we are all reliant on capital funding programmes and there is very rarely revenue beyond that, which brings us back to the maintenance issue. It’s these fundamental funding challenges that mean that, even we were to write down a really good target architecture, we would probably veer away from that very quickly.”
Short-term wins
“The first thing to do is to really get people together,” said Kate. “Understand, together, what the local challenges are, and what your plans need to be. Then as a group you can identify things that can be done quickly. There are always quick wins to be found. A big transformation change is made up of small changes incrementally. Just start with something small – it might be a clinical area or a team, but you start with them and then you grow.”
Kate also recommended that organisations look at their data and consider how they can undertake data-driven change; and also learn from elsewhere, to expand understand of benefits realisation seen elsewhere.
“I think work around authentication and single sign-on can be a quick win for many organisations,” Chris reflected. “Also, anything that you can do for keeping people in patient context is really helpful.”
Work that has enabled connection to other systems has “really made a difference”, such as shared care records and GP Connect. “They really make a transformative change to clinical teams and enable new ways of working.”
John reflected: “It’s about people, ultimately. The technology is there, it works, it has been proven in many different areas. It’s now about engaging and educating and bringing people with you on the change journey.”
How can progress be measured?
“You always have to start from a reasonably sold baseline,” said John. “It’s tricky, because there’s not always a clear start point. It’s always useful to establish those parameters from the start. When we work with fertility clinics, for example, we look at the number of staff they have, the number of patients, the treatments that they are having. Then we look at the amount of lag time for getting people through their treatments. Those are metrics that we can easily measure. We often try to engage with our clients and ask: what is your current pain point? If we could remove friction from that aspect of the flow, would that make the most difference to you?”
From Alder Hey, Kate stressed the importance of building progress measurement and the benefits realisation approach into project and deployment processes. “We have what we call drive and watch measures for our key progresses – drive being the ‘big ticket’ items and the watch measures being the smaller things we keep an eye on as we go. For each programme, we have a really clear set of measures which are categorised in this way and tracked over time. That can then lead into looking at things like quality benefits, productivity benefits, or cash-releasing benefits.”
Chris agreed that having the benefits realisation approach is vital, especially for larger projects bringing transformational change to the organisation. “We usually consider the digital part of these programmes to be an enabler – often digital has played a role alongside other factors and teams, but it’s important to ensure that it gets included when reviewing a project. That helps to prove the merits and the investment into digital.”
He also shared the view that the NHS should look at how “overall experience” can be measured with regards to the complex set of systems in use within NHS organisations. “We have the staff satisfaction surveys and so on which often raise issues with ease of use around IT and modernity of systems. We have started to look into using a system usability scale, and that can help to surface some of the issues. For example, one of the things where interoperability projects don’t tend to do so well, with two connected systems, is the fact that it’s often not easily flagged to the user when something isn’t working. Data doesn’t update or doesn’t become available or there’s a difference between one system to the next. It means there’s a blind spot, and usability scales are useful to highlight those so you can look into dealing with them.”
Many thanks to Kate, Chris and John for joining us and for sharing their insights.
