Thoughts on the Death of the NHS CUI Guidelines

I see that Matt Edgar, Head of Design at the NHS, has announced the withdrawal of the NHS Common User Interface (CUI). The reason given is that the CUI had not “changed with the times” following the closure of NHS Connecting for Health in 2013.

I think this is a pretty interesting episode in the social anthropology of digital systems (if that’s a thing). The CUI was the first, and let’s hope therefore also the worst, attempt at addressing some obvious problems in the design of digital systems in a huge organisation.

And as humans do when faced with problems too large for them to see, let alone understand, those tasked with its creation adopted a sort of religious veneration of their work, in the hope that perhaps heat would bring light. A sort of Calvinistic belief in the power of toil – like monks creating illuminated scrolls.

For example, a simple date & time input pattern consisted of two separate documents totalling 45 (PDF) pages. I’m sure I was not alone in concluding that the purpose of these documents was to exist rather than be read by an audience who also probably never really existed. Certainly, had I worked for  the NHS as a UX designer, I would have had neither the time nor the inclination to read them beyond tracking down a live example of the pattern in question too look at. And for all the impressive verbosity of the documents produced, they covered a rather forlornly small number of UI elements in any case.

In saying this, I do not wish to denigrate the efforts of those involved in the creation of the CUI. I personally know some of them. But I think we can learn something from this chapter in the history of how we make digital systems because failure (and if the CUI was anything, it was that ) is a good thing.

Will the second attempt at the problem of UI standards across the NHS fare any better? What, as Edgar asks, “if anything”, should replace it? Now the responsibility of NHSX, here’s the same date (but not time) input pattern in what is now the NHS Digital Service Manual.  Interestingly, while the old CUI guidelines addressed the issue of what to do if you needed an imprecise date, the new pattern just says not to use it. You instead need to read what is presumably the “official” guidelines on the (also new) gov.uk Design System. I can’t resist the thought that in the future this pattern will evolve into simply allowing the user to input anything. The system will just parse that input into the format required. Indeed, that’s probably possible today and seems the most user-centred approach to me. But I digress.

So paradoxically for the problem of standardisation, the answer seems to be at least partial devolution to different government departments. Here, for example, is what looks like potentially overlapping  discussions of patterns for the NHS. I wonder who wins in the event of conflict? And do they even know where, or even if, their patterns are being used?

I don’t work at the NHS or for gov.uk, so I’m only observing what are great efforts to keep things public and visible. Nor do I know what they should be doing. But I think the history of what’s been going on is interesting and useful and should be discussed more in the context of what happens over the long term in user interface design.

UPDATE March 2022

I see the successor (of the successor?) of the CUI Guidelines is now “live” in so far as it meets an internal standard  after having been launched in 2018.  It will be interesting to see how this develops, and whether it means the hitherto perpetual dawn has finally ended.

The GOV.UK Design System is now live