For screening services, a result leaves the lab as structured data. Then it duplicates and splits. One copy goes into the screening programme’s own system – structured, readable, never intended for anything directly user‑facing. The other copy goes to the GP record – also never designed for users – but eventually reaches them in the NHS App. By the time it arrives it has been translated, truncated and reformatted by each of the many systems it passed through. What the user sees is a bundle of strings, and the strings are unpredictable. Sometimes they’re readable. Sometimes a result arrives titled “Hum papill vir DNA dtctn assay”.
For our services, we want to help users get test results in a way that gives them the best chance of being confident they know what it means – and if they need to do something, how to do that thing. If we can consistently and safely match the two copies, the readable version tells us what the messy version means. Then we can enhance what reaches the user – clear, well‑structured information and what to do next. But it might mean hooking into reporting systems that were never built for this, to decode strings that were never meant to be read.
Anything we build works across other people’s software. We won’t know what this changes for other teams until we’re deeper into the detail.
For information governance reasons, we can’t look at the data reaching users to test any of this at scale. We can test if the matching works with a handful of results. But a handful will only be a proof of concept. To know if it works safely at scale, we’d need to test against real data. That might look like a ‘dark’ version of the service. Running against real results under full clinical and information governance, producing evidence for us, while users see nothing until we know it works.
The NHS App alone cannot change this system – it doesn’t own any of this. Designing services within the NHS is like trying to build a rocket to land on the moon – but the component parts were built 30 years ago, by organisations that never met and didn’t know they were building a rocket. There are teams trying to fix this at the core, but that work is big and really, really hard. It takes political leadership and consistency.
This is us making things better in the short term. Trying to untangle knots, or at least understand what thing goes where. It’s workarounds wrapped around workarounds bound up in workarounds. Simplicity for the user is paid for in complexity for services. That cost is real. We’re doing the hard work to make it simple, but we’re also adding to the spaghetti of complexity.
Jeannie had the idea to collect real examples and Alistair wrote up how prevention results appear in the NHS App.