Speaking at the CogX conference in London this week, health secretary Matt Hancock projected “huge opportunities” for NHS data use.
He said: “You are responsible for your health data. The citizen owns their data, and should be able to control it. There has been an attitude in the past within healthcare data, that it somehow belongs to the GP. That is fundamentally untrue. And it is anti-democratic.”
NHS Digital announced recently that it would defer the collection of GP data for planning and research until 1 September 2021.
The decision to delay on the General Practice Data for Planning and Research (GPDPR) service formed part of the background to a fireside chat at CogX with Maxine Mackintosh, research associate and fellow at the Alan Turing Institute, the Health Foundation and the University of Oxford.
Hancock said the Covid-19 pandemic health crisis “has engaged citizens with their health data like never before”, adding: “We’ve now had 25 million downloads of the Covid-19 app. That’s a new figure out today. We’ve had over six million downloads of the main NHS app, which you can now use to demonstrate that you’ve either had a vaccine or you’ve had a negative test result in the last 48 hours.”
Hancock said he believed this engagement with data would create “huge opportunities to improve and save people’s lives”.
He said the pandemic had forced “some big changes that we must not go back on and we must continue to push forward and the COPI [Control of Patient Information] Notice is one of them”. This was a notice that Hancock issued to NHS Digital on 17 March 2020 to require it to share confidential patient information with organisations entitled to process it for Covid-19 purposes.
But there had not been enough progress in sharing data more openly, said Hancock. “Separating the application from the data layer in any large organisation is complicated, and many large organisations have done it over the last decade or so. But in the NHS, in particular, this will be a mammoth task that will take a long time, because procurement has been done in so many silos, and so many so many parts of the system think they are responsible for the data in a particular area,” he said.
“And because there has been a culture of data sharing being done, especially in the research world, on a pseudonymised basis, by individual sharing of raw data that everybody then separately has to curate, we are going to change that. We’ll change it to a ‘TRE [trusted research environment]-first’ approach.
“This would mean that the data, which is the responsibility of the citizen whose consent is central to this, is held by the NHS on their behalf, can then be curated in a secure environment. And in such a way that each project that gains insight from it can build on it within an open source manner with an attitude to research which is essentially collaborative.”
Hancock added: “A technical debate about whether pseudonymisation [of data] is good enough to stop re-identification is ‘old school’. There is a far better way of approaching it, which is absolute privacy requirements for the citizens’ individual data, within a research environment where people are accredited to be able to enter. And openness about who’s looking at what.
“And so, if you can use the open data approach, where the data itself is more protected, and yet at the same time we are more open about who is researching on it, and for what purpose, what mission, then I think we’ll be in a far stronger place.”
Hancock referred to Ben Goldacre, director of the DataLab at Nuffield Department of Primary Care Health Sciences at the University of Oxford, as his “lodestar”. Goldacre is undertaking a review into the use of health data for research and analysis purposes.
Hancock added: “I’ve asked Ben Goldacre to answer the critical question: ‘Where does the commons essentially having the data end … and where can you then build on and gain value from the insights that are generated on top of the curated data?’ And the boundary there is incredibly important and highly technical. And I’ve got an expert in.”