For anyone in health analytics from the private sector, readers' online comments on news articles about Care.data can make pretty depressing reading. The loudest voices seem to be those whose main worry is that private companies will profit from personal data that does not belong to them. Most of them will be among the 1.2 billion Facebook users worldwide.
A fascinating October 2013 article entitled Romantic Partnerships and the Dispersion of Social Ties, describes a way of predicting those Facebook relationships that are most at risk of failure. The researchers analysed approximately 1.3 million Facebook users aged at least 20, with between 50 and 2000 friends, who list a spouse or relationship partner in their profile.
Then I read this week an announcement that the European Union has given a grant to a team led by Sheffield University to develop a real time lie detector for social media. The EU-funded project aims "to classify online rumours into four types: speculation, controversy, misinformation and disinformation". In the words of lead researcher Dr Kalina Bontcheva, the system aims to help "the emergency services to quash a lie in order to keep a situation calm".
This morning before I set off to the office I checked the traffic on Google to decide whether to go via the motorway or through town. The information I used to help me has come from millions of motorists over time and a few others who were ahead of me in the jams.
Facebook, Twitter and Google collectively are worth over US$600bn, making money from content that we freely provided- and which we assume is routinely mined by the National Security Agency's or the Government Communications Headquarters.
My social relationships, my political views, the places I visit, how fast I was driving – all analysed for patterns in order to make a prediction. Add on to that my emails, my documents, my stocks and shares, my internet searches (including those for the health conditions I am worried I might have).
So why is there so much fuss about anonymous health records? What's left for anyone to know about me? When data anonymity is protected in an act of parliament why should I put more trust in the privacy settings of Facebook, Twitter or Google?
Maybe part of the problem is that our medical data doesn't seem to belong to us. Request your GP record and if you are as old as me you will need to part with £50 to cover the cost of photocopying the pre-computer age, handwritten part. Then before handing the record over the GP must review it in case anything in it might cause distress. When he does, you will realise that the GP record is just that – a view of your medical history through the eyes of your GP.
Information about your hospital stays is contained to discharge letters – written to your GP. There is no such a thing as a personal record that joins together all of your interactions with the NHS over your lifetime. This is precisely the problem that the Care.data project seeks to address. So here are some suggestions for our next steps:
• Press ahead with Care.Data but focus on the primary purpose of the improving care.
• Make a commitment not to provide any data for secondary research purposes until the whole record can be made available for the public to see their own NHS record online. When HMRC get 10 million UK taxpayers to file assessments online, can it really be that difficult?
• Let people control their privacy settings in the same way that they do for social networks . Those of us that want to actively support research could go further and add in other information to provide a richer background context.
• Provide a dashboard showing which organisations and projects have used my data so I can see how I have contributed to research, and acknowledge my contribution as a legacy to science.
However it is done and however long it takes, the Care.data project is too valuable to lose. Let's start by letting people feel that they really do own their own data even if it takes a little longer.
Jonathan Burr is chief executive of Intelesant
Content on this page is produced and controlled by Intelesant
