On July 24th and 25th the NHS Citizen project ran what was advertised as a design workshop in Bristol. I attended the first day in person, and I followed some of the second day online. What the project will deliver is already clear in outline, but there is much detail yet to fill in. Behind the missing detail I suspect some important unanswered questions lurk.
The workshop took the form of presentations, discussions and drop-in areas, with the discussions and presentations repeated throughout the event. The intention was that to take all of it in would require four sessions or two full days, and that was true. I took part in lots of discussions, and I dropped in to a drop-in, but I mostly ignored the presentations. My perceptions of the whole thing are undoubtedly half-baked.
Of course, if you attended all the discussions, or even all the presentations, on a single topic you would experience a very different discussion or a slightly different presentation each time, so it’s unlikely that any two people’s experience of the event would be the same.
I mostly participated fully rather than taking notes, so this is in no sense a comprehensive report. It’s more a collection of rambling thoughts thrown up by the workshop. Parts of the event were webcast live and you can replay the webcasts to see what took place for yourself.
Rather than really being a design workshop, the event seemed to be an opportunity for the project to showcase its current thinking and prototypes, and at the same time to get feedback on its current concerns.
Welcome and sign-in
One current concern is around what impression citizens will get on first contact with NHS citizen online. Apparently a decision has been made to use a commercial product called MyDex to verify identity and store personal data, but this will mean going off to a MyDex site to log in to the NHS citizen site. It won’t be clear to everyone what they’re logging in to and why.
I suppose the process will be a bit like using your Facebook, Twitter, WordPress or Google account to log in to another web site, like some people often do now, except that MyDex isn’t trusted the way those other identity providers are. On the other hand, MyDex is technically more trustworthy for technical reasons, even though the reasons for this are technical, technically speaking. This means MyDex is all good. Honest. Yep. For sure. Really. What it all boils down to is that you trust that nice guy, don’t you? Even though he’s a sales rep, huh? Of course you do…technically.
MyDex has ambitions to be a kind of basket into which the Government will give you little choice but to put all your personal data eggs. At some stage during the journey from here to there, of course, it will become worth someone’s while to figure out how to suck your data out or tamper with it in more subtle ways, and as just soon as you can say “Jack Robinson” after it becomes worth someone’s while it will happen, but that will be then and this is now. For now it’s all good on that front.
Anyway, little of that is NHS citizen’s problem. The problem for NHS citizen is that there will be citizens who have no idea what MyDex is and who won’t trust it, and who therefore won’t trust NHS citizen. And that’s right from the start, before they’ve had any meaningful interaction with NHS citizen whatsoever.
Actually, as someone else at the workshop pointed out, there are citizens who won’t trust NHS citizen just because it says “NHS” in the name, and even less so if it requires them to identify who they really are in any way at all. If you think that’s far-fetched, have a look on this blog for the posts that describe my experience of hospital treatment last year. You won’t find any, because I never published any. Why not? Because my life might soon depend on that hospital again.
The part of NHS citizen that connects it to other conversations about the NHS is called Discover. For example, it is likely to connect with Patient Opinion. In the case of patients of the 2gether trust this can only be a good thing, because 2gether doesn’t appear to respond to any patient stories on Patient Opinion, and at least by sharing stories more widely there is more of a chance that 2gether’s patients won’t completely have wasted their time typing them. But there’s no guarantee 2gether will respond to patient stories on NHS citizen either.
Part of Discover’s role will be to connect with citizens offline, through people who will do the connecting for them. For example, a local charity that runs a discussion group for sufferers from some long-term condition could designate a member of the group to type information about what was said into Discover, or that’s my impression of the intention.
A current concern about Discover is that it won’t know who isn’t there. The plan is to use things like demographic analysis to try and ensure good coverage, but I’m not happy that such techniques will be adequate. For example, that local charity might only advertise its discussion group through the local hospital outpatients department, and it might have good representative coverage of the people who attend for appointments there. But that good representative coverage might mask a hidden group of sufferers who are so ill they have to be treated at home, and another hidden group of former sufferers who gave up on the NHS, paid for private treatment, were cured, and don’t have the condition any more. Ideally, Discover has to be able to discover such hidden groups, or it will merely reinforce the status quo, but that might be an impossible ideal.
That’s an important generalization about public engagement in the NHS, of course. The angry people, the committed volunteers, the expert patients and so forth, are generally the people with long-term conditions, or conditions that have been made long-term by NHS treatment. The ones who were cured, like the ones who died, don’t turn up for the meeting.
A completely different and remarkable aspect of Discover is that it will embody data analysis, so that it’s not just a discussion forum full of opinions, but a place where people can examine the facts and turn opinion into informed opinion. This seems to me to be very ambitious. For many of the healthcare issues that I take an interest in there is either no useful data at all or only secret data. Indeed I was hearing about plans for yet more secret healthcare data just a few days ago.
It’s very easy to go in the other direction, to ask, “What data is there?” and then to connect up some of the data you find, and finally ask a question that the data happens to be able to answer. But if you’re a citizen you don’t start with a list of known data sources. You start with the question you have, and in most cases there is no public data to answer it. Anyway, in a sense that’s the point of the system, to reveal what’s broken about the NHS to the NHS England Board.
Gather is the part of the system where you can raise a flag to indicate you think something about the NHS is an issue, and you can participate in conversations about these flagged issues. It’ll contain a collection of tools, but no one yet knows how the tools in the collection will fit together. Sometimes a flagged issue will arise out of something in Discover, and sometimes it will arise directly in Gather (I think).
One of my concerns about Gather is that it sounds potentially very complicated, so that it may tend to exclude people who aren’t power users. Well-organized individuals or campaign groups will tend to take it over. If the intention is to get ordinary people to contribute at this level, then I think it has to be made to look very simple. It doesn’t actually have to be simple, just look simple.
Partly, the idea of making everyone identify their real-life selves when they log in is to establish who is saying what, but it’s unlikely that, say, people who are receiving a particular medical treatment for a particular condition will be identified. So, if there are two treatments for a condition, and unknown to the NHS one of them works well and the other doesn’t, the people who come to NHS citizen to discuss the condition will mostly be people on the treatment that doesn’t work (because the others won’t be so upset by their circumstances). But NHS citizen won’t know this.
There might be a huge outpouring of support for the idea that, say, people with this condition should have more access to physiotherapy. But the right answer might be that people with this condition shouldn’t be receiving the treatment that doesn’t work in the first place.
You might wonder how a treatment that doesn’t work well could co-exist with a treatment that does. This can happen because in the NHS the treatment provider is paid for the work done, not for the outcome. A treatment that doesn’t work well means the provider gets to do more work and so gets paid more. The system rewards the worse choice of treatment — a kind of moral hazard. Patients rarely see the big picture. They’ll choose the treatment that’s more common and that the provider has resourced better so that it has nicer information leaflets and shorter waiting times. CCGs should root this kind of thing out, but it takes time.
Assembly agenda setting
Another current concern is how to go from the large number of issues flagged in Gather to the very small number of issues, maybe half a dozen or less, that can be discussed by the twice-yearly Assembly that’s envisaged. It’s well appreciated that simple numerical support for an issue in Gather is not in itself a reliable guide.
It seems to me that at some stage there has to be filtration in place with the effect that the system doesn’t allow certain issues to go any further. This is because, ultimately, the purpose of the whole exercise is to influence the NHS England Board, and there are issues that lie outside the Board’s remit. Sometimes, for example, that’s because an issue is too local, and the Board has no authority to stamp on an individual CCG or provider, and sometimes it’s because an issue is too big and political, and the Board has no authority to dictate to Parliament. And these are just examples.
There seems to be a design principle that filtration should happen as late as possible. That may be a false impression, however. It may be that filtration is just difficult for the designers to face, so that the parts of the system being prototyped first have shied away from attempting it.
Perhaps it would help to clarify NHS citizen’s relationship with other vehicles for public engagement if filtration happened as early as possible. I’m reminded of Tim Kelsey’s remark that the intention is not to nationalize other groups doing similar things. And one of the stories I saw in Patient Opinion, a story that got no response from the trust — if Patient Opinion had an early filter that told people to go to PALS instead for certain things, that person might have got a response immediately.
Another issue that emerged during the workshop is that there are two slightly mismatched visions of what NHS citizen is for. One vision, the simpler one, is that it’s for presenting important issues to the Board for its consideration. To do that, you need information about the issue itself, why it’s so important, and perhaps optionally what the solution might be.
The more complicated one is that NHS citizen is for holding the NHS England Board to account. To do that, you need all of the above plus knowledge of what the Board has done so far in that field, and of what the outcomes are. Experience of trying to hold directors of a provider trust to account suggest to me that you need very detailed knowledge of what the Board has done and of the outcomes to have any chance of getting anywhere. When it’s the NHS England Board, you also have the problem that the Board controls the available outcomes data.
There was supposed to be some discussion about connecting the outcomes from the Assembly back to citizens, but I found that hard to envisage without a concrete example. The solution I came up with, to assist with my own thinking, was to take an existing NHS England decision and work backwards.
The example that came to mind was an announcement from a couple of weeks ago: NHS England takes action to improve access to specialised mental health services for children and young people. I asked the question:
“What if NHS citizen had already been in place — how would that have been different?”
It’s not easy to understand what would have been different. Perhaps public pressure might have been more effective through the Assembly and the decision to do something would have been reached sooner. Perhaps the Board would have been held to account for its previous resource estimates at a much earlier stage.
And then a follow-up question is to imagine that, a year from now, there’s still a problem with specialised mental health services for children and young people:
“How effectively will NHS citizen be able to hold the Board to account for the action it took in July 2014?”
All in all, I’m still excited and hopeful about the NHS citizen project, but I still find it hard to imagine it working well in practice. This workshop helped, especially as it forced me to think up these examples, but at the same time it gave me the feeling that the project team is still focussed on the things NHS citizen can do. I’m beginning to want to understand what it can’t do, its boundaries, deliberate restrictions and known weaknesses.