Weeknote, w/c 31 March 2025

It finally feels like spring. The weather is changing and the NHS is in a transitional phase (to put it mildly). Somewhere, the 10 Year Plan is being developed with the goal of setting the direction for the organisation for the next decade. There are a number of possible futures that could come out of this, each shaping what I will or will not be able to do. Ahead of any bold new plans being announced, it feels like a good time to re-establish the fundamentals of my job.

“The NHS App gives you a simple and secure way to access a range of NHS services.”

That’s the public description of the NHS App. It seems straightforward enough, but many of the words in that sentence are significantly more complicated than they appear at first glance. To wit:

You

Who is the person in question? It won’t always be the one who is using the software. “Proxy” access, where one person manages services on behalf of someone else, is an important part of healthcare (think about a parent getting a prescription for their child). This introduces questions of privacy and delegated authority.

Simple

How do we meet the public’s expectations about how digital services should work while also accounting for the inherent complexity of healthcare? The NHS App primarily fulfils administrative functions, but there are clinical risks associated with many of these, e.g. if you can’t get your prescription filled in time, it can have severe medical consequences. We need to balance designing services that are easy to use against providing care that is clinically robust, and many aspects of health are just not simple. Bodies aren’t digital.

Secure

How do we deliver a safe service that isn’t a burden to use? Health data is some of the most sensitive information that exists. It is our job to make sure it remains safe (data security). It is also our job to make sure the people it concerns remain safe too (clinical safeguarding). Consider the kinds of content that might appear in a push notification on a phone’s lock screen. Now consider all of the situations a person might be in while receiving that notification. We want to make things convenient, but we need to ensure we do not put people in danger.

Access

How do we design a digital channel so that it doesn’t disadvantage people who can’t or won’t use it? The first point of the NHS Constitution states, “The NHS provides a comprehensive service, available to all”, but everyone in the country can’t access digital tools and services. Unlike the private sector, we can’t choose to not serve a particular market because it would cost too much. We have an ethical and statutory responsibility to not cut people out.

Range

What data and services are available to any given person? The structure of the organisation and the choices local organisations have about how they set up digital services mean that any attempt to establish a unified national digital channel is going to involve a lot of work to harmonise parts that weren’t designed to fit together.

Services

What is the best approach for designing one leg of a journey when no one is responsible for the complete journey? Many patients think the NHS is a unified thing, yet many of the elements of each service seen in the NHS App sit outside of our control. Users expect services to work across artificial system boundaries and for data to be shared, which is entirely reasonable. In many part of the health system, there is an inherent tension between devolution (which includes market innovation) and a coherent service offer.

Over the last week I’ve seen a few comments and articles talking about what people believe to be the correct way forward for the NHS App (and possibly digital health services more generally). What role should the centre play? What role should the market play? I’ve written about some of the challenges inherent in trying to design in this space before. One possible future is Matthew Gould’s conception of the NHS App as a “platform for innovation”. That sounds exciting and, to a certain extent, is what we are doing. The difficulty I have with this phrasing is that “innovation” evokes a search for novel solutions (robots, lasers, AI), when what we really need is mostly just hard work focused on a stable set of targets and applied for a long enough period of time. Maybe I have a limited imagination, but we don’t lack for awareness of what doesn’t work. More than brand new ideas, we need help to fix all of the problems we’ve known about for years. “It isn’t complicated; it is just hard.”

---

Reading from the past week:

• Between the tepid bath and the cloud of vapour: a plea for pragmatic ambition. Matt Edgar on the sweet spot for teams to be working in. This is great and (unfortunately) incredibly familiar. I’d summarise this as “use design methods”. Nothing fancy, just the normal work-a-day stuff of design for technosocial systems.
• Potato, potato. Ralph on helping teams collaborate and cults. “I feel like I spend more of my time speaking to people about what they’re doing than I do working on things myself” – welcome to the manager club!
• One year in. Frankie with reflections on a year of working on digital prevention at NHS England. All true.
• Test, learn, and grow principles. A set of principles set out by a cross-Whitehall group for operating in complex domains. Via Frankie.
• The GIST checklist. A checklist by Peter Jenson for team’s to use as they work through relatively small problems. I really like the “look for trouble” part, which mirrors the idea of trying to invalidate hypotheses (rather than looking for evidence that you’re right).
• Incentive architecture. Not new, but I keep referring to this article by Abby Covert. “Priorities only change when the reason we can’t change becomes the reason we have to in order to reach our intention.”
• DOGE plans to rebuild Social Security code. At what point should I stop torturing myself by paying attention to this horror show?
• The shape of network society. Gordon Brander, who is now “a McLuhan absolutist”, on the ways in which there is no culture without media.
• What makes an app feel “right” on the Mac? Watts Martin on design conventions and being a good platform citizen. At some point there is an entire post to be written about how this relates to public sector product design.
• Who has permission to know things? Iris Meredith on how hierarchy distorts understanding, what counts as knowledge in stratified social groups, and how this affects those without power.