By Michael Walsh, BBSRC Policy Fellow at the Society of Biology
In the digital age, so many of us feel that it is our right to have access to as much information as possible. However, certain exceptions apply: information relating to ourselves shouldn’t be freely available, and huge storms gather whenever leaks of personal data from medical or financial institutions occur. So it is perhaps surprising that the NHS in England has announced plans to place huge amounts of (anonymised) patient data in the public domain, a programme called care.data.
The plan is to collect and store healthcare data alongside some personal identifiers in a central database. This information would then be ‘pseudonymised’, and made available to healthcare professionals, academics, and (here’s where a lot of the controversy is) private companies. On the face of it this may seem like a worrying development, as statistics on our health are some of the most important and private information about us. However in a drive to reassure us, the Government is sending out a leaflet to households across the country outlining the purpose and benefits of the exercise, and also how to opt out. They argue that the information will be secure, and will lead to better patient care through improved planning and identification of risk factors.
Critics have been quick to condemn what they see as interference from ‘Big Brother’, and to point out the potential for shady uses of this information: if drug companies or insurers are allowed access (and it seems like they will be) then there are worries that they might be able to identify specific patients. The right for patients to know who has accessed their data and the ease with which they are able to opt out are also points which have been raised.
Other groups such as Nature have welcomed the potential benefits of the programme, while warning against abuse of the system and lack of clear information to patients on some topics. A survey last year by the Wellcome Trust showed that people generally agree with this point of view, and are positive on the use of their anonymised medical data within the NHS but less comfortable when it comes to it being shared with private companies and employers. It is clear that the positives from the scheme are widely recognised, but that proper safeguards and perhaps limits on who can access data are desired.
It is clear that there is great potential for both benefit and misuse, and it will take time before we know the net impact of care.data. Though one thing is certain, knowledge is power!
I heard a debate on this on Radio 4’s ‘Today’ programme a couple of days ago. A supporter of the new scheme and an opponent were being interviewed, and the opponent was challenged to cite a single example in which NHS medical data had been obtained without consent. Although he could not (or did not), thereby considerably weakening his own case, I have personal experience of just how seriously compromised the existing system appears to be. I am of the opinion that the existing security arrangements for the protection of patient data are by no means as secure and anonymised as we are led to presume.
The inquest on my wife in 2012 was a very high profile event, with full media coverage. She died from a highly atypical form of early-onset dementia, which appears to have been linked to her exposure to the 1988 Camelford aluminium sulphate water poisoning incident. For over 20 years the medical sector, including the NHS, had gone to great lengths to obstruct proper clinical investigation of this incident. (I have direct evidence of this, through my 12 year stint as a member of the CoT Sub-Grouip investigating it’s medical effects). Highly sensitive and, for the NHS and Dept of Health, damaging information was emerging during the inquest.
The barrister for another witness, a major player in the original incident, revealed that she had obtained a copy of my wife’s medical records, without my being aware of this. How this was achieved was not explained, but is came as a severe shock, since I could see no reason for this action. But she then attempted to establish to the Coroner that my wife had been a life-long drug addict, a disgraceful and totally untrue accusation. I had to bring my daughter into Court the next day to disprove this deliberate attempt to damage my wife’s reputation – remember that she was the victim, not someone under suspicion of wrong-doing.
My impression, and that of others present, was that this could have been an attempt to imply to the Coroner that my wife was therefore in some way responsibloe for her own death, despite the clear clinical evidence that it was due to aluminium contamination of her brain. If NHS records can be obtained without permission, in highly traumatic circumstances such as this, by the legal representatives of potentially hostile and prejudiced witnesses, then the security of this proposed system must itself be seriously in question.
Assurances that patients’ confidential medical data will be impossible to trace to the original sources cannot be accepted as adequate. There is certainly a case to support the use of suitably anonymised data for large-scale tests of association (not causation, of course) in medical epidemiological analyses. But where there is even a remote risk that they could, with improper human intervention, be put to less ethical use demands the utmost caution before any such scheme can be given approval.