2 min read

Guest editorial Monday 17 September 2012: “You are not a beautiful and unique snowflake” - melting healthcare data

This guest editorial by PCT Cassander, one of the wittiest and best-informed commentators on health policy and its discontents, discusses the potential abuses and uses of data to improve health services.

I had to read George Orwell's 1984 at secondary school. I remember laughing archly with the other quiet, bookish types in my year at the irony of being forced to read a book about totalitarianism.

I also remember it being the first book that really frightened me – not because it depicted a state which continuously monitored its subjects, but because it postulated that from this constant surveillance, it would be possible to accurately deduce a person's deepest secrets.

......................................................................

Click here for details of 'The red ties that bind Comrade Sir David: postmodern NHSCB to commission itself (oh yeah, and what cowboy drafted this mandate?)', the new issue of subscription-based Health Policy Intelligence.

......................................................................

I thought of 1984 again recently when I read an article in the New York Times article with the bland yet unsettling title of “How Companies Learn Your Secrets”. It tells the story of what happened when the marketing department at a large American retailer asked a company statistician the question “If we wanted to figure out if a customer is pregnant, even if she didn’t want us to know, can you do that?”

The answer may be less surprising to you than it was to the angry Midwestern father - who wanted to know why his teenage daughter had been sent discount coupons for baby accessories.

The underlying reason for wanting to find out a customer's gestational status is that becoming a parent is one of the few events in life when a person is particularly receptive to an attempt to change their shopping habits. At other times, people tend to follow an established routine, which is very difficult for even the most tempting of promotions to substantially alter.

Which brings me to the NHS part.

Collecting data has long been an NHS pursuit. Linking these data together and making sense of them is a much rarer activity, as highlighted by David Colin Thomé back in 2009.

The value of these data is obvious, as evidenced by a queue of pharmaceutical companies ready to pay six figures for anonymised access to patient records – an information trove which will doubtless lead to the discovery of any number of exciting new medical treatments.

Or exciting (and profitable) new medical conditions.

Ideally, though, the information flow should be going both ways. There is a wealth of data collected by the private sector which could help the NHS understand its patients and design new services. For instance, Tesco preside over a gargantuan database named Crucible – perhaps as a sly nod to Arthur Miller's classic depiction of a community riven by fear of the dark arts.

If pregnant women can be spotted through their shopping habits, what about people who are trying to stop smoking?

Can future diabetics or hypertensives be predicted through their purchase history?

Do people who recover well from a PROMs procedure have different shopping habits to those who don't?

Linking a database like Crucible to existing NHS data could answer these questions.

Of course, there would be privacy implications; although these would likely be no worse than the current move to presumed consent for patient records being available to paying customers.

Ultimately, the amount of information which is held about each of us, and the number of ways in which it's linked together are only going to increase. As Chuck Palahniuk puts it in Fight Club, “you are not a beautiful and unique snowflake.”

The big retailer of the world are well on their way to proving this – it would be nice if better health services, as well as laser-targeted marketing, could be part of the end result.