The NHS Citizen project is nearing the end of its design phase, and on February 24th it published a discussion paper on technology platforms and standards, inviting comments. In this post I’ve mostly worked through the discussion paper randomly nit-picking as I go, so unfortunately the result probably isn’t very coherent or readable on its own.
Anyway, here are my comments on NHS Citizen Technology Infrastructure, the discussion paper as described in the project blog.
Overall, I found this discussion paper too long. Much of it strays far from the subject of technology platforms and standards, while in other areas there appear to be gaps.
For example, there’s an unnecessarily lengthy description of Mydex CIC (the provider chosen to secure personal data) but no description of its backer, The Young Foundation. Lumen, the chosen online community platform, is mentioned without any description whatsoever of its backer, the New Economics Foundation.
For other components of the system no technology platform or standards are mentioned at all. For example, the project website is currently built on the WordPress platform and I subscribe to it using the RSS standard, but there’s no mention of either.
Advertisements for the “participation passport” and the associated “personal data store” clutter much of the paper unnecessarily. It would have been better to deal with these components in their own sections, and more succinctly.
Too much of the paper describes aspects of the design itself instead of the platform and standards requirements. For example, there is a description of a “balanced scorecard” for evaluating evidence, but no mention of technology platforms or standards to support it. If the scorecard has no technology implications then it doesn’t need to be mentioned in this paper at all.
Part of the problem with the paper is that the design is undefined, or there’s a reluctance to define it. At this stage of the project I feel that’s becoming a hindrance. Without a clear design there’s no good way to address platform and standards requirements.
2.1 What is NHS Citizen?
The “two simple questions” at the start are a bit complex, because it’s not clear who “we” are:
- “How can we better take into account the views of patients, service users and the general public when taking decisions about the NHS
- “How can we ensure that the board of NHS England is held to account by the public which it serves?”
In the first one it looks like “we” are maybe NHS England, but in the second one it looks like “we” and NHS England aren’t the same and “we” are maybe NHS Citizen. I think this is important to get right. Early on it was OK to keep the purpose of the project a bit vague so as to see what emerged, but now at this stage I think enough is enough.
The two simple questions make it seem that the target audience is the public at large, but then the description of the model talks about:
“a culture of participation within NHS Citizen”
“pre-existing patient participation”
So it seems as if there are three different target audiences. The widest one is the public at large. Then there is the subset of the public who already participate in NHS decision-making as patients. And finally there are people who take part in the culturally separate NHS Citizen (which I guess is probably a subset of people who already participate, but maybe includes people like me who participate but not as patients).
I was astonished to see that the remit of the project now includes this:
“[NHS Citizen] is designed to support the way in which citizens and staff use social media today and to encourage and spread best practice”
Was there a need for “support”? Had there been bad practice? Was this ever NHS England’s job?
“It has been developed using Agile methods”
That’s like saying beef is vegan because cows eat grass. Developing a design that’s so complex no two people can describe it in quite the same way, and then discussing the platform and standards requirements, and then laboriously implementing the thing isn’t really what Agile is about, you know.
So, although this introductory section is not key to the purpose of the discussion paper, its lack of clarity is striking. It probably would have been better to omit this section entirely, to avoid conflict with other descriptions of NHS Citizen in other places.
2.2 What do we mean by Citizen centred architecture?
The following view is patently untrue, or else worded so badly that whatever it’s trying to say doesn’t come across:
“the view that people are citizens and owners of the NHS, rather than passive recipients of services”
Last time I had NHS treatment (on Wednesday morning last week, as it happens) I was a passive recipient of its services, and I certainly don’t have any ownership of it despite for several years having been just about as involved as a citizen can be. So what’s being described here is perhaps a delusion, or a utopian ideal, but it can’t seriously be a view.
“the architecture must protect [citizens’] identity”
Why? The only possible reason is if the architecture forbids citizens from protecting their own identities. (And of course we shall see later that this is the case.) It would be less Orwellian to describe this feature as forbidding instead of protecting.
“to collect and protect the widest range of participants in an environment where they feel in control”
I like the sound of that, although I doubt the practicality of an environment where a wide range of participants all feel in control at once. I think it would be more sensible to acknowledge that however you design the thing it will exclude some people, and it will include some people who don’t feel in control, and some other people will feel very much in control, and then to understand who all those people are.
“Bringing the widest possible range of voices in, and maximising their impact, by using methods to which they are already adapted;
not creating a culture of debate that privileges some voices and excludes others;”
I fear these may not be compatible. Some voices might be excluded because the methods to which they are adapted just aren’t effective.
“Giving participants control over their data and their identity, rather than just treating them as an entry on a database.”
This seems nonsensical, but perhaps it’s explained later in the paper.
“Creating an equal and co-productive relationship between the citizen and the NHS…”
There isn’t really any such thing as “the NHS”. It’s just a brand. Perhaps this is meant to refer to “NHS England”.
“Allows self-defined geographies”
That’s a problem here in Gloucestershire where we have a border with Wales, because Wales has its own NHS with subtly different rules, and NHS Citizen only applies in England.
“Similarly, the topic should be defined by participants and not imposed by outside forces.”
What does “outside” mean? Specifically, is NHS England outside or inside? As NHS England is certainly subject to outside forces (e.g. other bodies in England, global forces) and NHS England is part of the co-production, then the whole thing is indirectly subject to outside forces. For example, some citizens will almost certainly want to discuss topics that are beyond NHS England’s influence.
“should support…with respect to…a commitment”
That wording manages to imply that it won’t actually be open, transparent or private. This whole subject is extremely difficult. I think it would be better not to make promises you can’t keep.
“personal data would always remain private to the individual unless they chose to share it”
A meaningless tautology.
“with the proviso that published analysis must never made [sic] personal data identifiable”
An impossible requirement that can be used fraudulently to block publication of almost any analysis. Again, I think it would be better not to make promises.
“a standard that makes their contribution available to the decision making process”
This seems so opaque as to be meaningless, although perhaps it’s expanded on later. Clearly, there are decision-making processes in the world that don’t rely on personal identity at all, and in which no such standard is required to make anyone’s contribution “available”.
2.3 Do we need a social currency?
As this is not being proposed just now, I’ll ignore it!
2.4 How does it relate to other digital programmes?
This section is too long, I think because a whole lot of stuff that’s not terribly relevant has been pasted in from somewhere else.
“so the certified company does not know what services the citizen accesses”
Is “certified” the right word here? Perhaps you mean “certifying” or “authenticating”.
“This prevents two service providers from comparing notes”
No, it really doesn’t. It only prevents one possible method by which they might compare notes. It doesn’t have any effect at all on other methods they might use. In this sense it’s a risk to privacy, because it tends to make people think service providers can’t compare notes when in fact they can, and as a result people might place more trust in some service providers than they really should. I’m disappointed that NHS Citizen has allowed itself to be dragged into this murky area.
“take advantage of the security and identity benefits this system provides”
There are benefits for NHS Citizen? Spell them out! I can see that there might be benefits for some government transactional services.
“For NHS Citizen this will bring alive background information in local or national contexts”
I think it’s the other way around. I hope NHS Citizen will help bring alive government data that is currently useless because it’s been made “open” without any context of decision-making. For example, I recently worked with a dataset that’s only published at local government district level when the relevant policy decisions are made by CCGs. The data is “open” alright but almost impossible to make any sense of (and deliberately so, I rather suspect).
2.5 Open questions
While section 2.4 seemed to ramble on about stuff that’s not obviously relevant, this section seems to skim over things that need detailed explanation.
“One of the areas where NHS Citizen will need to break new ground is with respect to the use of user generated content…”
It’s not obvious why.
“Revocation of consent is a generally desirable feature of control for citizens…”
It’s not obvious why it should be so in the context of NHS Citizen.
I won’t continue with quotes from this section. It all seems very abstract and theoretical. NHS Citizen doesn’t need to construct a universal theory of data, it just needs to get a job done.
I suspect there’s some confusion going on between personal and public data, and because I’ve come across similar confusion elsewhere I think it may be that NHS Citizen has absorbed the confusion rather than generating it itself. If NHS Citizen had clearer boundaries, then I believe it could produce (co-produce, ideally) solutions that work within those boundaries even if the same solutions don’t work elsewhere.
“to be better connected and as a result more visible”
Being better connected results in being less visible, on average. If I show you a photo of a million people, and in there somewhere is your uncle Fred, connected to them all by being at the same event, do you think you can spot him? What happens in large connected systems is that concepts like “trending” and “viral” become more significant than visibility. So I don’t find this architecture overview believable.
“several connected tools”
This list makes it look like it’s not just that there happen to be several tools, which happen to be connected, but instead there are specific tools connected in a specific way to support a specific process. I think it would be clearer to come clean and admit that there’s a process with tools to support it.
“a ‘Participation Passport’ to connect active citizens”
Clearly, citizens who are active in the NHS are connected in various ways already, so the need for a ‘Participation Passport’ is unclear.
“allow for responsible sharing of information”
This implies that those of us who are sharing information already are doing so irresponsibly, which I would suggest is a false accusation.
3.3 Open standards
3.5 Developing an open API
3.6 Interaction with other systems
The proposal seems weak. Use of the passport thingy isn’t meaningful interaction with other systems.
“help manage patient involvement”
Do you really mean “patient” and no one else?
3.7 NHS Citizen: work to date
Looks sketchy. It says what some of the component platforms are, but not all of them.
Lumen will inevitably be seen as politically tinged, even though you and I know it’s just software.
“As introduced in section 2.4, a principle of ‘minimum viable data set’ is to be applied…”
It’s not mentioned at all in section 2.4, as far as I can see. In 2.5 it’s referred to as an open question, making it surprising to see here that the question is settled and it “is to be applied”. I see no obvious justification for minimal data, myself, but I haven’t thought about it much.
4.2 Active Listening
This whole section is alarmingly vague.
“a window onto health and care discussions happening across potentially hundreds of active forums and networks”
So it’s like…what…a web browser? Or is it more like a web browser that only lets you see discussions that someone else has decided you can be allowed to see?
“Features in the active listening space are limited to searching, and tagging…”
But a few lines above, and again below, it says you can “join in discussions on any of the networks”!
4.2.5 rambles off topic into a discussion of the PDS.
“In order to post a comment on a particular network, users will need to choose to login [sic] to that network using their personal data store”
I don’t find that believable without much more explanation. Anyway, this is meant to be a paper about technology platforms and standards. What real-world technology platform or standard supports this?
This section doesn’t seem well thought out, and it doesn’t tell me anything about technology platforms and standards.
“to highlight (or ‘flag’) an issue as a high priority”
That wording implies that the issue already exists as an object in the system before it is flagged — as if the system contains a collection of unflagged and flagged issues. But the as yet unflagged issues will have titles, descriptions, etc. already. The highlighting user won’t be the one who provides these.
“Issues that have progressed past the initial comment and into discussion”
This is a third class of issue? So there are unflagged, flagged and in-discussion issues?
“responses from the NHS system”
What’s the NHS system? Does this create a fourth class of issue, an in-discussion issue that has an “NHS system” response?
4.4 System data
I’m bored now. Why do I need to know this?
4.5 Individual data
This seems like it should be in the PDS section.
4.6 The balanced scorecard: Data as evidence
An interesting summary of the basics of professional lobbying. Unclear what it has to do with technology platforms and standards.
4.7 Data licensing
Interesting suggestion in there that citizens are really going to be protected by contracts, not by the PDS as such or by statute law.
“Contract law empowers this [sic] citizen.”
A contract empowers the parties to the contract, and no citizen is a party to any of these contracts, as far as I can see. Protecting the public is what statute law is normally for. So this section is unclear.
4.8 Questions to discuss for security, privacy and anonymity
Interesting questions, though not obviously related to technology platforms and standards.
5.1 Why do we need a participation passport?
This doesn’t actually explain why we need a participation passport. It just hypes up why it might be nice, using words like “key” and “critical” to make it sound more…key and critical. I get it that the decision to use a participation passport has already been taken. It’s really pointless trying to justify it at such length.
5.2 Importance of the Citizen being in control
“This is about sharing decision making – and power…”
Not back in section 2.1 it wasn’t.
“a group of people who often feel disenfranchised and disconnected”
Really? Who do you mean, exactly?
“the overall concept has been extremely well received when discussed in the abstract”
Only when discussed by well-connected types who particularly enjoy abstract discussion, I suspect.
5.3 Prototyping and development process
This section doesn’t say anything about the process.
“as a CIC there are fewer commercial sensitivities”
We probably don’t need to know quite so much about Mydex CIC, but it would be good to know just a little about The Young Foundation.
5.4 PDS approach
Much of this is just hype, too.
“They will do so with the knowledge that they are sharing their information securely and that it won’t be misused.”
It may be true that the information won’t be misused, but not simply because of the PDS. The PDS only prevents certain theoretically possible forms of misuse. It has no effect on other possible forms of misuse.
“those they choose to share it with, in this case NHS Citizen”
Does this mean NHS Citizen is a legal entity which enters into a contract with every user?
6 Design process
“This paper provides an overview of the expected requirements to support the NHS Citizen design.”
It’s not a clear overview at all, though. Too much is left out, and there’s too much other information that doesn’t help in understanding the requirements.
There is no Appendix One.