Huge amounts of data don’t in themselves guarantee a safer and better NHS - they need constant challenge and scrutiny of interpretation, writes Roger Taylor

The Brexit referendum exposed how far the ‘official’ narrative of what is happening in our country can lose sight of the views of much of the population, despite the availability of polls, economic data, expert analysis and a vigorous media debate.

Transparency, open government, freedom of information, a free media – these are the mechanisms by which we try to make sure this does not happen. These are the tools through which our democracy should remain coherent and avoid fracturing. The evidence of the last few weeks is that it is not working.

Transparency became a central part of health policy after scandals showed how easy it was for the insider view to lose sight of reality. The Bristol Royal Infirmary inquiry revealed an organisation giving assurances about safety that had no connection to what was happening in the operating theatres.

A decade later, the Mid-Staffs inquiry found that quality assurance systems created in response to this – performance score cards, public reporting and reviews by Monitor – were still able to miss the fact that a hospital with patients protesting outside it had experienced a collapse in standards of care.

Testing theories to destruction

So what is the problem with transparency? The analysis put forward in Transparency and the Open Society is that our narrative around transparency is out of date. We regard more information as more transparency, when what we need is more control over information. Most of the time, when we talk about transparency, we refer to powerful organisations – whether it be the NHS or the government – giving an account of itself and publishing information. This does not increase transparency if these organisations are able to present a narrative that cannot be challenged.

In a digital age, transparency must be understood as the ability to contest the interpretation of data in a scientific manner – being able to examine the way that one person has understood information and test it to destruction. Our current conception of transparency falls short of this. The creation of data portals with endless tabulations of ever increasing irrelevance does not put the ability to reinterpret data into the hands of others, as is so often claimed. Nor does giving patients the ability to view their records.

The heart surgeons are sometimes held up as a model of transparency. But they only published adjusted mortality figures after Dr Foster did so

True transparency means a genuine sharing of control over information to the extent that the narratives available to you and me are not controlled by those responsible for services – whether as professional, provider, regulator, commissioner or Whitehall department. It means a genuine democratisation of control of information.

That requires two conditions. The first is that I can take data used about me in standard machine readable formats and share them with anyone I chose. There must be an exception for confidential information that might harm me. But outside that, all information about the patient should be portable and standardised because I have little desire to look at it myself. I simply want the interpretation of it to be contestable.

Constant challenge

The same goes for information about the system as a whole. It is not sufficient that the NHS publishes score cards and performance statistics. What is required is that this interpretation can be challenged and alternatives presented. Again, I have no desire to compare different sets of statistics. I simply want to know that it is possible for a third party to be able to interpret that data in relation to my circumstances and advise me whether I am receiving the right service.

Transparency has already done much to improve the quality of healthcare by making variation in quality more visible. But this process relies on constant challenge. The heart surgeons are sometimes held up as a model of transparency. But they only published adjusted mortality figures after Dr Foster did so – and since Dr Foster ceased to publish, the information is no longer available.

This view of transparency is not just a challenge for public healthcare systems – it is an equal, if not bigger challenge to private organisations. We will see more companies enter the field of healthcare by collecting data from us and offering interpretations of that information – whether through a phone diagnosis service or watches that measures your heart rate. These services need to be subject to the same level of transparency – an issue that may cut across business models built upon plans to exploit proprietary control of patient data.

There are legal and institutional challenges to bringing a more authentic form of transparency to healthcare. But there is an urgent need to do it. Healthcare is at the forefront of the need to use computer driven intelligence to interpret data. As the volumes of genetic, epigenetic and biomedical data explode we have no option but rely on machines to makes sense of it all. I may be willing to put my trust in a doctor without transparency but the same does not apply to a computer. Such systems must be wholly transparent and open to scientific scrutiny and challenge if they are to be trusted.

Transparency and the Open Society by Roger Taylor and Tim Kelsey is published by Policy Press.