Find your NHS number service shows how to fix interoperability

This week NHS Digital launched the Find your NHS Number Service. It is fantastic and I am delighted to see it exist, but not for the reasons you might initially think - it showed the way to a better, more Interoperable NHS.

First lets look at what the service does from a user journey perspective.

It collects some basic information to assert a person exists on PDS (Personal Demographic Service). It doesn't assert that the user and the record are the same person at the point of submission, but that's the important bit, it doesn't have to, because it's only going to send the NHS Number to the patients telephone number that the GP system has recorded. It's requested every time you attend your GP so they can send you an appointment reminder too and in turn updated on PDS. It's not perfect but it will work for a lot of people. It needs something you know to be able to send a message to something you have.

The real penny drop moment is that up until this point we assumed the user and that patient are the same person, they are not, the service is designed with the system in mind, the patient is only going to be really involved in the step that is missing from the user journey above. Receiving the SMS. The rest will probably be done on their behalf.

I suspect one of the main drivers for this service is so various non NHS organisations can record COVID-19 test results/vaccinations and have a reliable identifier for when a mass of CSV files are pushed around and imported back into GP records to avoid the mountain of governance and technical complexity of integrating with the GP systems directly. In these times we cannot wait that long. So this service was built to fix system needs and they are pretty important ones too.

So why does this change things for Interoperability?

Let me go through a sequence of events:

  • I first noticed the service on Wednesday, I saw the tweets, clicked a link and saw the front page.
  • I came back to it on Thursday and went through the journey to completion.
  • 15 minutes later I had built a simple prototype API service, this service used a few common libraries and 47 lines of code.
  • The service allows me to pass the demographics to a simple API that will automagically fill in each page on the users behalf and click submit.
  • I can now use this in all my back office systems, it's quicker to develop than it is to read a twitter timeline.
  • The patient is sent an SMS or an email but it's easy to select SMS if present.
  • A lightweight proof of concept app I have (and therefore any app with the correct permissions) was able to read and obtain the NHS number for my patient's SMS that was received.

So within days of this service going live, I was able to build an prototype of a new service that would allow my users to obtain their NHS number and pass it to me with minimal intervention. I could then use it to reduce retrieval time of records from SCRa if I had a clinical service, improve the processing of Discharge Summaries I may generate back to General Practice, heck even if I was sftp-ing CSV files around they would be easier to handle and match.

Not only this, I didn't have to speak to a person, be reviewed at a board, fill in mountains of documentation, submit evidence, buy an HSCN connection, PDS compliance and spend £10,000's on resources to get through these things, the result however is the same, I can obtain the users NHS number.

Why is this better?

The reason is simple, the API prototype I built changes who the user is, it becomes the patient not the system and they are put at the centre, we trust the user. It is after all (or should be) their data, we sent it to them, they can chose to share with who they want to. It would be good to remove the need for a hack, expose the API, enable user centric applications to work better for the users by making it properly open.

It is not lost on me that this is a simple dataset and it gets more complex with sensitive data, but the pattern is now there, NHS Login can assert identities and the NHS App can be the vault of data or a conduit to it that the patient controls. Two brilliant services that will only get better and make it easier to focus on the user and could be the key to a more interoperable NHS.

The purpose here isn't to cover all bases or make the prototype live, but to share that there is another way other than system to system integration for system needs, they will always have a part to play for all the big scary stuff. That's policy though, not delivery. Bravo NHS Digital, you have delivered another fabulous service that is challenging old assumptions.