Hello and

I'm Tim Harford.

few things in the news and in life.

And I'm Tim Harford.

Few things in Nurdland are as vexing as huge troves of medical data.

Big medical data sets pose an enormous risk to personal privacy.

They're so sensitive and it's surprisingly easy for bad actors to abuse them.

But big medical data sets also hold out an enormous opportunity.

There's so much we could learn from them about how to provide better treatments or simply

to organize health care more effectively. So how can we seize this opportunity without

unleashing a privacy apocalypse. This fellow has an idea.

My name's Ben Goldwaker. I'm Prophet Oxford. I run the Bennett Institute for Applied

Data Science. Ben made his name as a newspaper columnist and the author of Bad Science.

Now he runs a data lab at Oxford University.

Let's start by explaining the exceedingly tricky double-edged conundrum he's been grappling with.

So the very same records that generate important insights in medicine,

patients electronic health records at scale across the whole population also contain the most

confidential medical secrets of the whole population.

Medical records contain data about when you got ill and what you got ill with, what drugs you were prescribed and who you are, age, gender, ethnicity and where you live.

Across a wide and diverse enough population, England is excellent for this, you could in theory use

this data to identify all kinds of useful patterns, disease risk factors,

surprisingly powerful drug combinations, long-term trends in population health, and so on.

...