Bringing down the knowledge deficit

In this week’s budget the Chancellor reaffirmed the Government’s commitment to reducing the deficit. Deficit reduction is the Coalition Government’s one paramount policy objective and the definitive register for their political rhetoric. I’m now beginning to wonder whether we have unwittingly found ourselves in a deficit reduction paradigm, where the response to all political questions is ‘reduce the deficit’. (Like the ‘law of the instrument’, also known as ‘Marslow’s hammer’: ‘if all you have is a hammer, every problem looks like a nail’.

The reason I wonder whether public policy as a whole has fallen into a deficit reduction paradigm is that this deficit reduction appears to have defined the official line in an area that I’m currently working in, namely, the use of biological and health data. I’ll explain what I mean presently. We all know that a lot of criticism of NHS England’s care.data programme has been reported in the media. (This is the proposal to centralise social care and health care data – from both primary and secondary care records – in the Health and Social Care Information Centre, and to make them available for certain uses, including commissioning, service development and biomedical research.) However widespread public suspicion of official motives may be, there certainly appears to be widespread uncertainty about the details of how the data will be used and why this is a good or bad thing. This may not be just the result of poor communication or paternalistic obfuscation, since the uncertainty appears to be on the NHS England side as well as among the wider public. In any case, a large number of people seem to have no recollection of the leaflet by which notice of the programme was served on them. This put a spanner in the works.

What’s the response? ‘Reduce the deficit!’ The Health Secretary has announced an ambitious deficit reduction programme that would reduce the deficit to zero – in just six months! (I imagine the Chancellor, at this point, fainting over the Cabinet table from amazement at the scale of this ambition.)

It’s not only the government and its official organs that are guilty of rushing back to the deficit model when the implicit public mandate for their policies appears to slip through their fingers. Some medical research charities have been at it too. (In fact, I might be able to offer the Chancellor some small assistance with his own problems if I had a pound for every time I’ve heard or seen the phrase ‘if only the public understood’ or its cognates in discussions of the benefits of care.data.)

As I implied in a recent post on this blog, my fear is that instead of a genuine public debate we simply have a PR offensive. PR is a game that all can play. But lots of media stories about the benefits of biomedical research will not do the job and nor will anecdotes about identity theft and discrimination. (And we know which kind of story plays better in the press.) Cheering for abstract benefits and booing imagined costs leave the question of uncertainty unmediated on both sides. That is not to say that we can reasonably quantify the likely benefits or harms of a programme like care.data (I don’t think we can) but that we need to think about what they mean and how they should be approached. And this thinking needs to be undertaken collectively in some way, since we are all implicated. We need to establish a common moral ground.

I have been reflecting recently on why some controversial issues of public policy (like assisted conception) proceed in a relatively orderly way and others (like genetic modification of plants) have repeatedly blown up in official faces. From where we are now, engaging meaningfully in public debate must be crucially important, not simply ‘educating’ the public, and especially not hectoring or cajoling them. Care.data might have something to learn from the experience of GM Nation and subsequent attempts to get the GM ball rolling again. Which is, perhaps, a question for another time...

Comments (3)

David Coles 23 March 2014
Your article is spot on. However, I have little hope of Ministers and senior officials learning the lessons - their understanding of public perceptions and responses is where the true deficit model needs to be applied. I was working at OST on Public Policy in Science at the time of the GM debacle and despite trying to provide sensible advice to Cabinet Office it was clear that the writing was on the wall for the demise of AEBC following GM Nation - it wasn't the message the Government wanted to hear.
David Booth 23 March 2014
The advocates of the research use of data on patients' personal lives have at last been inveigled into public debate. However it is simply not possible to specify (let alone put cash value on) benefits and harms of such fundamental investigation of mechanisms for improving medical and social care. If Care.data's storage and processing of personal information are made fully secure and transparent, there remains the moral deficit of unreviewed communication to each patient of the justification for the exploitation by the aims of each study. Opting in or out does not address the issue. The UK research community has ethical review to the highest international standards (e.g. the ESRC's Framework). The technology required already for safe operation of Care.data could be extended inexpensively to incorporate ethically approved informing of patients in related series of longitudinal studies and insightful qualitative-quantitative designs. What are the data-sharing fetishists so afraid of?
Adam 21 March 2014
Controversial public policy issues cannot be implemented without a degree of transparency. From an outsider's perspective, the failed initial implementation of care.data appeared data-customer focused, not patient or GP-led.

When one starts with the customers' ideal (i.e. the maximum amount of data possible) and then attempt to mould all aspects of the system to fit that ideal, it is inevitable that many patients and healthcare professionals will be suspicious of the underlying motives. This is particularly the case when a perfunctory public information campaign and opaque opt-out system seem intended to disguise the design priorities.

At source this isn't a Big Data initiative, or clinical laboratory – it's about personal, individualised healthcare. It may be extremely useful to have identifiable, or pseudo-anoymised data available for clinical research, drug development or even private insurance allocation – but should that be the focus? Shouldn't the relationship between the patient and healthcare professional be the core of any digital records effort?

Why not start with an anonymised system and build from that? Why not spend years, rather than months, debating with stakeholders? Why not focus on the minimum data needed to improve overall care and save lives rather than the maximum?

Even if care.data had been implemented in this fashion there would still have been dissent. It would, however, have been informed dissent to which plans could be adapted. The controversy would have been healthy and beneficial rather than destructive.

Care.data is, above all, not a novel concept. Other European countries have managed to develop and digital records systems; systems that are designed to integrate with medical insurance companies as a default. That should be pretty controversial shouldn't it? Yet France is managing to do just that in integrating the DMP electronic healthcare record system into the existing Carte Vitale structure.

The principles of their approach?
- Total patient control of medical records (even to the level of hiding sensitive data from selected health professionals)
- Electronic auditing of healthcare professionals' access to individual records, with the audit trail being visible to the patient.
- Data to be anonymised at source before any extraction for bulk analysis.

The DMP is proving to be a costly system, as many large, state-backed ICT programs tend to be, yet it has roughly 85% approval public approval ratings. I would argue that this is mostly because it has been designed to serve the healthcare recipient rather than the data-customer. That might be the most important lesson for NHS England to learn.