Today at 5pm I’ll be giving the Bellwether Lecture at the Oxford Internet Institute. My topic is Big Conflicts: the ethics and economics of privacy in a world of Big Data.
I’ll be discussing a recent Nuffield Bioethics Council report of which I was one of the authors. In it, we asked what medical ethics should look like in a world of ‘Big Data’ and pervasive genomics. It will take the law some time to catch up with what’s going on, so how should researchers behave meanwhile so that the people whose data we use don’t get annoyed or surprised, and so that we can defend our actions if challenged? We came up with four principles, which I’ll discuss. I’ll also talk about how they might apply more generally, for example to my own field of security research.
Without reading Ross’s paper, this is a risky thing for me to say, but on its face, the proposition that “the law some time to catch up with what’s going on” in genomics can and should be challenged.
Medical ethics is pretty invariant isn’t it? Patients have always had a reasonable expectation of privacy, they’ve expected that their data will remain confidential, that information disclosures will be kept to a minimum, consistent with health care objectives etc etc etc. Why does that change now? What’s changed with modern genomics and medical Big Data is that researchers can divine details about people like never before. Usually this is for the best, but as we’ve seen with “DNA Hacking”, there is enough genomic data now in the public domain and enough other linkable personal data for rich medical information to be predicted about people behind their backs.
I wrote about this in a letter to Science in 2013: http://lockstep.com.au/library/privacy/letter-to-science-re-identifi.
The fundamentals of privacy and medical ethics should put the brakes on the excesses we’re starting to see in genomic research. It has always been the case that patient consent is needed before a physician takes blood for example and runs tests on my state of health. If a health informatician now can diagnose me from Big Data, without taking blood, the ethics haven’t changed a jot. Let’s stop getting carried away with the powers of new data analytics, and instead take some time to be sure that in this new world, patients are given proper information about new diagnostics and treatments, and every opportunity to participate in the clinical decision process.
Do you know if the talk was recorded?
I am expecting a video of the lecture to appear here in due course ; http://webcast.oii.ox.ac.uk/index.cfm?q=bellwether&view=search&Search=Search
The previous presentations in the series are also here.
The text may turn up here if not with the video; http://www.cl.cam.ac.uk/~rja14/econsec.html
The talk is currently at this URL and the slides are here.
Here is a more recent version of the same talk, this one given at a big data conference in Cambridge.