Wales is building a dragon of an NHS app

Different country, different requirements

The plan for the app came to fruition after it became clear that the NHS App in England would not fit the Welsh requirements, particularly as it needed to be bilingual, but also due to different functionality requirements, including giving wider access to health and care services across Wales.

The DSPP programme had an ambitious goal: not to only to create an app, but to foster better relationships with suppliers and opening the NHS to innovation, as well as delivering fully joined up care.

The contract for the app development was awarded to Kainos, which worked with the DSPP programme to build the app. Towards the end of 2022, the app went into private beta with a thousand patients from 10 different GP surgeries trialling the app.

DSPP programme director Stephen Frith explained that he originally assumed that “we’d switch everything on and roll out the big bang. But that proved to not be a sensible suggestion.” 

The beta testing was originally due to run until Christmas 2022 but has been extended until the end of March 2023.

“Fundamentally, nothing has gone wrong,” he said, adding however that the support model has been “quite complicated” due to the ecosystem of different systems and different suppliers in use in the NHS and getting those to speak to each other.

“When you get a simple call on a help desk because something’s not working exactly as you would expect, it’s amazing how difficult it is to get service management back to that organisation to run it smoothly,” Frith said.

He added that just getting the help desks, which should be “fairly standard old-fashioned ITIL support arrangements with IT systems”, defined and bedded down is taking a lot of effort.

One issue the team ran into was the problem of the apostrophe. In one GP practice, they were keen to use the correct grammar, and put an apostrophe in the clinic name, but the apostrophe did not come across through the interface correctly, and the clinic failed to show up on the list of bookable appointments.

Apostrophe problems sorted, the app will soon appear in both Apple and Google’s app stores.

A different kind of help

However, the DSPP programme has decided against a publicity campaign for the app.

“While it will be in the app store and available for everybody, we’re not advertising that fact,” Frith said, explaining that instead the programme is going to focus its efforts on supporting the GP practices, slowly and continuously rolling it out.

“We’re going to build up around the GPs in Wales, making sure they understand what the app means for their patients and get them to own it because we’ve got a slight difference in the way we’ve set up the support model in that we’re not setting up any kind of central communication hub for people using the app directly, other than technical issues.”

This is because the experience in England has often been that when people phone up requesting help with the app, the issue is that there are no appointments available, which needs to be dealt with by the GP practice itself.

“At best, we delay them speaking to the person they need to speak to; at worst, if they’ve got some sort of medical condition or emergency, we delay them [in getting help],” Frith said.

He added that everyone expected the app to have a big call centre and a central phone number, but that would essentially result in setting up a call centre for the NHS. “Before you know it, you’re fielding all the calls – [but] that’s what NHS 111 is for,” he said.

The supplier, Kainos, is running the technology, and will run the app and all its support for the programme for the first 12 months.

“It is under contract to make sure the lights are on and everything’s working properly, so if the technology fails, we don’t expect it to be the public telling us that it’s failed, we expect [Kainos] to notice and fix it,” Frith said.

Once those 12 months are up, the operational system support will likely migrate to Digital Health and Care Wales (DHCW).

While the programme doesn’t “want to be beholden to an external company forever”, Frith said he expects there will still be an external contract for some development, new features and skills transfer, adding: “I suspect we’ll operate a hybrid model for some time. Kainos is very supportive of that and is ready to support that transition.”

Rather unusually, the software for the app sits in a Microsoft Azure DevOps environment, with the subscription held by DHCW rather than by Kainos. This means that DHCW owns the code and the licence for the software.

Assurance and testing

Another element key to the app development is bringing on board people themselves. DSPP co-programme director Matt Cornish, who runs the project together with Frith, said there has been a lot of work going on with assurance groups in guiding the path the app has taken, from assurance on technical expertise, service transformation, ethics and information governance. But what has been most important, he added, was people and making sure the app works for them and their needs and getting the NHS organisations on board.

As Cornish points out, there is only a finite size to the country. “So, despite previous history of some of the health boards operating in different ways, and ultimately having sovereign authority about what they procure or don’t procure, there are opportunities and lots of similarities,” he said.

“We’ve been able to connect with them on a regular basis, and as we’re about to go across the lines of public beta, there is opportunity for us to continue to collaborate in a number of areas.  Opportunities around understanding how we can use communication preferences of citizens to better communicate, reduce costs and improve personal experiences with hybrid mail, for example, are huge.

“These are things to go on our roadmap, and then there’s more to come after that. It’s not overpromising the jam tomorrow, but showing that if we work effectively together, we’re more likely to be able to deliver “once for Wales” solutions and, in time, be able to provide more equitable access to services.”

The road to private beta testing was also complicated. The team wanted to make sure there was a good representation of people both with a preference of English and Welsh language, people who were au fait with digital and those who weren’t, a mix of gender, ethnicity, age, health conditions and so on. Some 25,000 people were reached out to, and of the 4,000 people interested in taking part, 1,000 were selected.

Cornish said there is also ongoing work to better understand how people use the app. “We want to know how they land on particular pages, how long they use particular pages, so that we can continually adapt and improve – having kind of the data tool behind it is going to be important going forward,” he said.

Supplier relationships

While the DSPP programme had a slightly different vision for the app than in England, the big thing was that the app had to be in both Welsh and English.

“It has to be dual language to comply with Welsh statutory requirements, and we wanted to design that in to be completely multilingual, should we ever need that,” Stephen Frith said. “We’ve localised the deployment of the software. That took quite a bit of work.”

The NHS app in England is also very primary care-focused, while in Wales, the team wanted to make it “very clear that we’re doing more than just primary care”, Frith said.

He added that even in primary care, things can get complicated. Wales GP practices either use the Vision system from Cegedim, or Emis Web. Both systems are widely in use in England as well, but Wales runs separate instances of the systems.

When Kainos was awarded the contract, the supplier suggested lifting the code base from the NHS App in England and use that as the basis of the Welsh NHS App. While this has been done, simply replicating the code did not work in all instances. Because they are different products than the English versions, the interface built between the GP systems in England and the NHS App there would not work in Wales.

“We’ve had to pay the companies to re-deploy those interfaces in Wale and make them work. That has not been straightforward,” Frith said. The Emis interface has just been released into the supplier’s testing environment and should be live soon.

This huge undertaking of a project requires competent and skilled people, which can sometimes be a challenge to find, and the DSPP programme is currently recruiting for several roles in the core team.

Building blocks

While the app in the first instance will give people the ability to perform actions such as book GP appointments, check test results, order repeat prescriptions and use NHS 111 online, what’s going on in the background is much bigger.

One of the key elements of the DSPP programme is to give patients much more control. The team has built several proof of concepts in the background that, while not yet accessible to the public, are built into the codebase of the app.

One of those is a piece of work currently dubbed “patient captured data”, or “about me”, where the patient’s information and things that are important about them are listed.

A typical use case, Frith explained, is when a patient is in long-term care and then gets transferred to hospital. “There are a lot of paper forms that get completed by either the patient or the nurse on their behalf, that may or may not find its way to the hospital and may or may not get read by the nurse there,” he said.

This includes information such as what’s important to them, what language they speak, whether they have bed sores, a pet at home, and so on.

This has all been digitised and can now be captured in the app. The DSPP programme is currently working on building an API (application programming interface) to interface with the Wales Nursing Care Record, used by health boards across Wales, so that nurses on the ward can see the patient-entered data captured in the app when using the system. Patients have control over access themselves, and are able to switch consent for access on and off.

This is only the first step towards giving patients more ownership of both their own data and the access to it. Wales also plans to expose waiting list information to the public on the app, so that if a GP has made a referral, the patient will be able to see that they are now on the waiting list.

While patients will get access to the summary care record through the app, Wales has no plans to open the full, written and uncoded record to patients. Frith explained that the topic is not only highly contentious in the medical profession, in terms of whether it’s beneficial, but it’s also not on the top of the list of what patients want.

“We’re sort of avoiding that one,” Frith said. “I’m sure it’ll happen at some point, but in a much more controlled and collaborativefashion.”

Patient-held records

The programme also wants to give patients access to all their medical information in one place, whether it’s from primary care, secondary care, private care, or information collected at home.

“We’ve deployed technology that allows us to do that, that’s just invisible in the background at the moment,” Frith said.

The technology is the EventStoreDB, an open-source database, which stores state-transition events. In layman’s terms, this means that any information that is date stamped, gets thrown into a big pile in the database which you can then filter through to find what you’re looking for and pull it through to the app.

As an example, if a patient has a blood test done by a laboratory, they send across the information on the date the test was done, the patient’s NHS number and details of where those test results are stored.

On the app, the patient will have a health timeline with their health interactions, and they will be able to filter it by test result, appointments and several other things, then click on the button that says “show more”, at which point the app goes to the database through an API and pulls that data back to display it to the patient.

“That concept is phenomenally powerful, and it’s a technical deployment, not just a conceptual one,” Frith said.

This can be used as a basis for a range of events and information and is already being used in the app for journal entries where patients can capture how they’re feeling on a particular day. That then goes into their event timeline and can in the future be shared with NHS staff, subject to access controls, and allowing them to navigate to different care events.

“That will be the backbone of a patient-held record in our system going forwards,” he said.

In the future, the plan is to set up a full API management service that allows multiple suppliers to register, access and post data through structured APIs. “That is part of the long-term vision in Wales, and this will just facilitate that,” Frith added.

The team has also been working on integrating third-party solution. One of the Welsh health boards, Swansea Bay University Health Board, already has a patient portal, provided by Patients Know Best (PKB) which can be accessed through the app using API-based integration technology. This is acting as a demonstrator to show how third-party solutions can co-exist alongside the NHS Wales App.

The next 12 months are going to be pivotal with the programme pushing ahead on many fronts:  Improving the underlying NHS Wales App experience is only part of the plan. Increasing the health and care data provided to the patients and public of Wales through the App as well as offering more patient interactions for things like appointment bookings and patient recorded outcome measures are being progressed in the background.

Continuing to support the integration of innovative third-party services to enrich the overall experience is also on the roadmap for the next year. For now though, the focus is on getting the App out of the door and operating in a fully live environment, quietly and without fuss.