NHS patients could soon be making appointments, renewing prescriptions and even declaring what sort of end-of-life care they want to receive by using a new app, the details of which Gizmodo UK can exclusively reveal today.
Thanks to freedom of information laws, we’ve got our hands on a cache of documents that describe exactly what the app will do, what it will look like and the hugely complex technical challenges that lay behind connecting Britain’s health services to our phones.
In fact, we can today reveal that thanks to the byzantine structure of the health service, developers of the app nearly made an utterly insane decision about how it should work - before user-research intervened and stopped the final app doing something that would have been utterly nightmarish.
Read on to find out more - and don’t worry, we’ll start at the beginning.
What will the new app do?
The intentions behind the app are exactly what you might expect: Make it easier for patients to access healthcare services, cut down on administrative bullshit, and eventually provide a means to connect with other health apps, like Fitbit and Apple’s Health app. That’s right, this is the app that could mean that you no longer have to spend ages on the phone trying to book a GP appointment ever again.
This new app is the first stage. And it was on the 12th September last year that erstwhile Health Secretary Jeremy Hunt set out “8 challenges” that provided the skeleton specification for the app. They were:
- Symptom checking and triage (i.e. figuring out if you’re just being a hypochondriac or whether you should be in an ambulance right now)
- Access to your medical records
- GP appointment booking
- Repeat prescription ordering
- Changing data sharing preferences
- Changing organ donation preferences
- Changing end of life care choices
- Promoting “approved apps” to patients
There are already some apps available that can carry out some of these functions (such as GP booking) - but this app will be the first time that all are unified into one simple app, and promoted nationally, using the actual NHS brand. So expect big things.
Here’s some screenshots of what to expect. Of course, things may change during development and before release - but it’ll definitely look something like this. Here’s appointment booking - you’ll be able to make a booking and view upcoming appointments:
And here’s what it will look like ordering a repeat prescription - note that this doesn’t change the existing system of requiring GP approval, it’ll just take place inside the app instead of in person:
Here’s what the home screen and medical records section might look like:
One of the most interesting things about the documents we’ve obtained is that it shows how user research informed the design of the app - and consequently, the other objectives that Hunt set out are hidden in the “more” tab.
For example, below are the sections to change your data sharing preferences and organ donation preferences. That’s right - at the press of a button on an app, you’ll soon be able to dictate which parts of you can be sliced out after your death, which while weird to think about, though is absolutely a good idea.
These two settings are also interesting for how they could be used in the future. This is just us wildly speculating now, but both data sharing and organ donation could be ripe for some “nudge”-style experimentation: It’d be interesting to see what would happen to the rates of organ donation if the “yes” button was much larger than the “no” button, for example. And it’d be cool to test whether more people would sign up if prompted by a notification on their phone reminding them to do so, just like how Domino’s Pizza always manages to time its push notifications suggesting you order a pizza for whenever it knows you’re vulnerable.
It’s also interesting to consider why data sharing has been included: One of the big opportunities for healthcare in the future is going to be in terms of big data, and using algorithms to comb through medical records to look for unexpected correlations and the like, as a key to unlocking new cures and therapies. If the app can get more people to consent to sharing their data, that could be good for humanity too. Presumably putting control over this into the app is a good, pre-emptive way of demonstrating to people who might complain just how easy it is to opt out.
And yes, the app really will also be used to let you specify what sort of “end of life” you want to receive.
With a few drop-down menus, you’ll be able to specify where you’d like to be cared for, as well as your religion so they know how to deal with you after your death. This is what the NHS refers to as an “advance statement” - and which features in the first version of the app. Probably not best to take the piss and put “Jedi” on this one.
In the future, the goal appears to be to offer a number of other options here, but there are legal and clinical hurdles that make it difficult to include stuff like lasting power of attorney and your advance decision on whether to refuse treatment.
There’s also a simple technology challenge. Though you might assume when you fill this in that it goes off to some big NHS database, the documents reveal that the reality is that what this does is basically fire off an email to your GP, whose poor work experience kid will then have to manually re-enter the information into the GP’s database. But hey, this is a good start.
And as you might expect, end of life care is clearly something the designers have been worrying about from a pure user experience perspective too, and the documents make repeated reference to how “end of life care content” presents “design challenges”. (Our top tip: Probably not the best time for emojis.)
The good news though is that according to the docs, “once people saw the End of Life Care prototype they were more supportive of it within the app”.
Dodging A Design Bullet
This looks like a relatively simple app, right? But what might look to end uses like four main functions in some relatively friendly packaging actually hides some rather breathtaking complexity.
Why? Think for a moment about what the NHS is. The NHS isn’t one single system that operates as a single whole. It is instead the name we give to hundreds, if not thousands of different services, organisations and systems.
Every GP’s surgery, every hospital and every pharmacy will have their own computer system for handling patient records and other healthcare information. And because of the way the NHS is structured (here you can insert your own little rant about the Tories marketising healthcare here if you like), every one of these organisations has a lot of autonomy in how it designs these systems.
So the challenge for NHS Digital, the group making the app, is how to get these systems to talk to each other.
And as the design documents reveal, the group initially proposed something that would have been horrendous. This is our editorialising, but you’ll definitely agree with us when you read about what the original plan was.
Here’s what the documents call the “original solution hypothesis” - the original plan for the app. This shows how the various systems would have interacted with each other. The crucial part are the boxes on the left. It appears that the original idea was to have the user use the NHS app to simply select their local GP, and then insist that the user go back to the app store and download yet another app, presumably built by a private provider, that works for their local GP - or “clinical commissioning group” (CCG), to use NHS Jargon.
We let out a little scream of horror at the thought of this, as you can just imagine what a terrible user experience it would be: Your NHS app immediately telling you that you downloaded the wrong app, and then you’d have to download an app for your CCG. It’d be like having to download London Facebook when in London, and then have Manchester Facebook for when you visit friends up north. Hideous.
Mercifully then, common sense won in the end, as user research prior to development revealed that “all users” thought the two app model was insane, as this slide shows:
In fact, this is so crazy that we wonder if the developers knew this all along, but needed to pitch the idea and ask the question to normal users, in order to persuade their UX-design-illiterate managers what a deeply terrible idea it would have been.
The developers avoided one UX nightmare, but this still doesn’t solve the overall problem: Just how do you get all of those different systems to talk to each other? To make it all accessible through an app, it would have to give you a NHS account and verify your identity so that it can be sure that is you accessing what is effectively the number one example of sensitive personal information. And then it would have to connect up your NHS account to your records on your GP’s system, and the records at your pharmacy - not to mention stuff like the organ donation database.
The new app manages to solve this, and images from the development process show how NHS Digital has managed to make this inherently complicated process slightly more bearable between iterations of the app.
It isn’t entirely clear whether this screen would have ever made it to the final public version, but check out this if you like design nightmares:
In order to hook up the app to your GP, it would have required users to enter an “account linkage key” and an “ODS Code” supplied by their GP, as well as their own user ID. Yes, there is no way in hell that your Gran would ever use this app if this were ever the case.
The good news then is that the developers later built a much friendlier sequence - albeit one that still shows the necessary hoops you’ll still have to jump through to use the app.
Here’s a complicated diagram showing the login/signup process:
What this effectively shows is the sequence of events when you first sign up to the app after downloading. What’s most interesting is the section in the top right:
It’s hard to see but this shows how you’ll have to prove to the app who are you: Not only will you have to supply a photo of your passport or driving license (or other identity document), but you’ll then have to record a “video selfie”. The app doesn’t say whether you’ll also have to hold up today’s newspaper in it to provide proof of life. After you’ve submitted your details, you’ll then have to wait up to 24 hours for the NHS to approve your identity and let you into the app.
Once you’ve proven your identity, the app will then automatically connect up to GP systems, without any need for the end user to worry about “ODS” codes, whatever they are. And this step in the process illustrates another complication.
As described above, GPs and CCGs will all have their own systems for managing patient records and appointments, but in most cases this doesn’t mean building their own systems from scratch. Instead they will have bought an off-the-shelf system from a healthcare software provider. Which is good news for the NHS app, because it means that they only have to build in support for a handful of major suppliers and they cover all GPs and CCGs. And the documents we’ve obtained also reveal the progress they’ve made integrating with each one:
What this boring chart reveals is that if your GP is using EMIS Health, there’s a good chance you’ll be able to use the NHS app from day one - but if you’re GP uses TPP or MicroTest, you might have to wait a bit longer for the companies to get their act together.
The documents also offer other glimpses of the design process - and what we can expect the app to do in the future. And you know they’re serious about testing - here’s an image showing how they literally strapped a camera to the heads of people to observe how they make use of the app:
There has also been plenty of testing of really subtle design changes - as demonstrated by this slide showing slightly different login page designs:
Perhaps our favourite example of this user-centred design approach is in the booking of appointments. You might think that this would be easy given we’ve been booking tables at restaurants online for years. But showing off the app to GPs, the NHS Digital team quickly revealed a problem: The need to “triage” requests for appointments.
One of the most underrated aspects of being a GP receptionist, it turns out, is the ability to figure out if someone really needs to see a doctor, or if they are just wasting the doctor’s time. In person or on the phone, it would be relatively easy to ask questions and divine from the patient’s actions whether or not they truly need an appointment, but online this isn’t so easy.
So the developers added in a little nudge. When you go to book an appointment, rather than simply taking you to the booking form, the app will throw up a prompt, suggesting that instead of diving straight in and bothering your GP that instead you should take a look at the online symptom checker. A brilliant example of how good design could conceivably save thousands of wasted hours.
The documents also point towards enhancements planned for further down the pipeline too. On a slide outlining “priority” products and enhancements for 2019, it reveals that they want to add the following to the app:
First, there’s “Market Interoperability” - turning the NHS app into a platform that can work with third party providers, and a “Personalised app library”. Both basically describe enabling the app to hook up with the likes of Fitbit or your Apple Watch, to share data. Then there’s fingerprint login, which seems like a no-brainer for health data.
They also want to include built in support for GP video consultations, so that instead of going to the doctor, you can Skype with them instead. This is already being tested in some parts of the country in a separate app, so bringing into the app seems like an inevitability.
And finally, eventually the intention is to hook the app up to another big NHS service: The E-Referral Service, which is used to book hospital or clinic appointments.
This is pretty cool, right? It seems like the NHS might be… slightly modernising, at long last. So when and how will you get a chance to use it?
The good news is that your phone will almost certainly be able to run the app: The developers are supporting mobile operating systems going back to iOS9 and Android 5 (Lollipop) - which will apparently cover around 99% of users. If you have another device, or a computer, there will also be a mobile web interface… and even “partial support” for Internet Explorer 8, though we secretly suspect that is more for the benefit of NHS computer systems in GP surgeries and hospitals that have long needed upgrading.
The documents also reveal the rollout plan: At the end of September, the NHS will launch a private beta with around 2000 patients in ten GP practices, apparently picking practices which are in areas with “high wifi uptake” and “high 111 online uptake”, that use EMIS’s system - that was the one with all of the ticks on the boring graphic above, which the developers have most thoroughly managed to integrate so far.
Then, more excitingly, at the end of December there will then be a full public beta, albeit with controls on exactly what features you can access based on what the NHS wants to test. The documents don’t reveal a target date for the full, proper release, but it is estimated that a year after full roll-out, the app will have 16m users registered for online GP services, 3 million GP appointments being booked or cancelled online, and with 7 million prescriptions being ordered.
Like the plans for the app overall, this is definitely ambitious. But if the app launches successfully, the prognosis for the future of the NHS could be a much better health service.
James O’Malley is Interim Editor of Gizmodo UK and tweets as @Psythor.