Care.data: What’s in a dot and what’s a dot not?

When is a period not a period?

The period, as my six-year-old son knows, terminates a sentence, brings a discrete thought to an end (like this). Nevertheless, thinking has momentum; it trips over the period into the next sentence (like this). He also knows that a hyphen elides two or more thoughts, like the thought that he is a six-year-old boy (the number of counts, the unit of measure and the directional accumulation of time as ageing). How, then, might I explain to him the meaning of 'care.data'? Are the ideas separate but contiguous, in tension with each other, or do they flow into one another, become con-fused? 'Care.data' or 'care-data'? One should not assume lightly the Adamic privilege of naming, for names assign natures. To the web-aware (that’s all of us), ‘care.data’ already sounds as though health information will be spread across the internet, like a miasma in the data cloud.

The tension between care and data has caused some well-publicised difficulties for care.data recently. Their effect has been to force a halt to the programme. They have centred on the means by which notice of the programme was served on every British household in a way that too few can recall (and that left those who could little the wiser). Since then, there has been a toxic drip-feed of negative stories in the media, hinting at poor data security, questionable competence in information governance, and a lack of clarity about how, with whom, for how much, and for what purposes, care data will be 'shared'. The fact that the answers to these questions appear to be not only occult, but actually undetermined, seems somehow more grievous.

In the face of all this, NHS England has announced a pause. The privacy interest groups now surmounting the barricades are no doubt congratulating themselves at having taken the streets. But it’s worth pausing to note that the risks are potentially symmetrical: there are risks to not using information as well as in misusing it. It’s just that they are differently framed and distributed. However, we have no common ground on which to evaluate these: hype on the one hand, scaremongering on the other. A plurality of priorities and interests. There is nowhere for sober public reasoning to take place (which is the same as to say: there is not enough public reasoning to create this ground).

We should also recall that the HSCIC did not appear from nowhere: its predecessor bodies have been marshalling national hospital admitted patient care data (HES data), for example, since 1989-90. At the same time, the Clinical Practice Research Datalink is able to deliver answers to research questions right into a clinical consultation. So why has care.data become such a monster? Why has the collection of HES data or the CPRD data attracted nothing like the attention that the proposed data scoop from the primary care sector has?

Doubtless there has been some very bad planning, bad communication and misunderstanding of the values and interests of communities that care.data connects together (particularly in primary care).
 There is the fear of a centralised government ‘data bank’ that is almost part of the national character. There is the taint and tarnish of commercial interest, not helped by the confusion of imperatives that lie behind the project. (Is it for improved health outcomes, biomedical research, resource allocation, or national industrial strategy?) Nor is it helped by myopia about where the drugs our doctors prescribe us really come from. All of these things are relevant, but the way in which they figure is unclear. Perhaps the failure of communication is so severe as to deprive patients of a meaningful opportunity to opt out, or to meaningfully inform any choice in the matter. Perhaps there are conditions in which this would not matter so much. What seems clear is that the government and NHS England have mistaken the moral basis for care.data, perhaps (who knows?) because they have not thought about it in those terms.

What, then, of our own work? We have been talking about care.data within the Council's current Working Party on biological and health data. I wrote, in an earlier post on this blog, about how I thought a major difficulty for initiatives of this type is finding a moral basis for the transition from a perspective of ‘data protection’ to ‘value extraction’. Care.data has provided us with one of the most fecund examples of this (not that there has been any shortage of examples). Now that our consultation is closed, the deliberative work is well underway, and we have an increasingly clear sense of our approach. Our report will be out towards the end of the year but we’ll continue to track the shifting sands of data governance and the building works placed on them in mean time.

Meanwhile, many questions remain unanswered. What, specifically, will fill care.data’s period of reflection and re-engagement? Is the Government going to take this, as a live issue, into the next general election? Or will it (like the National Programme for IT) continue piecemeal, albeit without the toxic ‘care.data’ banner?

What happens in the current period of reflection and debate will be crucial not only for care.data but also for the moral basis of future data initiatives in healthcare and biomedical science. As one of our Working Party members points out, mishandling care.data risks undermining confidence in the health service with the result that people seek treatment late or not at all. So getting this right is important. If care.data is to go ahead, there needs to be more than the hope that a period of positive publicity will drive out the scaremongers. There needs to be some genuine public reasoning.

So: when is a period not a period? When it’s a full stop.