Hurrah for Health Data Research
A victory for the cause of sharing medical records for the public good
It is not often that I sit at the breakfast table cheering at the Today Programme, especially when the news is as grim as it has been lately. But this morning when I heard Cathie Sudlow making the case for sharing our health data for the common good I found myself punching the air.
Professor Sudlow, the chief scientist at Health Data Research UK, was talking to Mishal Husain about a hugely important piece of work, a study showing 7,000 Covid-related hospital admissions could have been avoided in 2022 if people had received their full quota of jabs. The research found that 44% of the UK population had not had all of the recommended doses of Covid vaccines, with younger people particularly likely to have skipped jabs.
There was a link for people of all ages between missing jabs and hospitalisation and death rates, with under-vaccinated people over 75 more than twice as likely to fall seriously ill with Covid than the fully protected. At a time when the voices of those claiming vaccines don’t work or are even harmful seem to be becoming ever louder, this study is obviously of great importance.
But the really remarkable thing about it was the extent of the data available to the researchers. Anonymised health records from everyone in the UK over 5 were analysed, with data from GPs and hospitals and from the Covid vaccination programme. Readers of this newsletter will know how difficult it has often been for reputable research programmes like UK Biobank to get their hands on this kind of data, and I have often expressed frustration that exaggerated privacy concerns seem to trump the benefits of such research.
Professor Sudlow was asked by Mishal Husain about such privacy worries:
“I would urge people to think very hard about the benefits as well as the risks, “ she replied. Questioned about whether there was a risk of people being identified from the anonymised records, she stressed that the security applied to the use of such data had been repeatedly tightened with ever greater oversight of the researchers:
“We’re getting better and better at running analyses at this sort of scale for massive public benefit and to inform public health policy without putting anyone in any danger of being reidentified or having to be concerned about their privacy.”
It struck me again that the risks - that some researcher only allowed access to the data in a controlled environment would nevertheless go to the trouble of piecing together the identity behind an individual anonymised record - were very minor, especially when compared to the benefits of such research. Professor Cathie Sudlow says the UK is rare in having the ability to collect such a vast pool of health data and hopes it can now be used to study other conditions such as cancer, heart disease and diabetes and discover new treatments. I, for one, will be cheering her on.
The problem here is that we must NOT leave health data sharing "to trust and altruism" - we have seen with the issues around the UK Biobank that it is not only shared and used for good - but also for profit. The mega data handlers/users/sellers are ready to promise anything and pay for this gold dust. Patients are often more willing to share their health data in the hope of benefits for them and their community, we owe them relevant research and return on engagement. We need to make sure there is no TRUST involved, but only good solid guardrails, cross-stakeholder monitoring, respectful data/research return and explanations to citizens, and swift effective, retribution when things go wrong. Genomic data (which can identify those that come after us) is even more tricky. As a Parkinsons patient Sara Riggare's views on patient data use is well worth a read : https://www.nejm.org/doi/full/10.1056/nejmp1704485
I don't think privacy concerns are exaggerated. The NHS doesn't seek consent from people, misuse the word 'anonymised' to persuade people their data isn't being linked and shared.
Its Also about patient autonomy - its simply polite to ask people to share their data, especially when healthcare professionals go on about consent, trust, privacy, confidentiality and all the rest of it