RogerBW's Blog

Ben Goldacre explains at length how pretty much everything about drug research and selection is rotten.

This book is a polemic and a call to action, which is generally the source of its problems as well as its good points. The basic information is solid, but it's written in a highly emotional manner (Goldacre keeps pointing out how angry he is, which gets fairly tedious), it's often repetitive, especially later in the book, and inconvenient details sometimes get skated over in the effort to provide a single consistent message. There's often an artificial dichotomy: people are either Evil Drug Company Stooges or Honest Brave Researchers. I think you'll find it's a bit more complicated than that.

A very brief summary of the content: it transpires that trials are poorly designed, biased, and selectively and misleadingly reported, regulators are underinformed, and doctors are lied to and bribed to select preferred drugs. (One wonders slightly how it is that the various companies don't cancel each other out in their efforts to get their drug prescribed more.)

In a curious divergence from the main thrust of the book, Goldacre spends a chapter explaining how he's in in favour of NHS data sets being made much more easily available to researchers; a little later he points out how drug companies already take shameless advantage of the data they can get, for example identifying individual doctors or patients by merging study data with public information and advertising to them directly. Are "researchers" somehow to be separated from "drug companies"? Goldacre even believes that anonymisation of data can work; that's a charmingly naïf viewpoint, but was sadly out of date even in 2012, and his support for the care.data plan earlier this year suggests that he's still talking well outside his area of competence.

There are all sorts of small annoyances. It's all very well to say that medical advertising exists solely to distort any possible evidence-based practice, but how is this any more true here than of any other sort of advertising? Medical advertising is no more uniquely horrid than any other. Describing Elsevier as a "respected international academic publisher" may just about have been tenable in 2012, I suppose; anyone who'd call them that now does it because he wants something from them. There are elementary mathematical errors ("a quarter of [...] revenue is spent on marketing [...] We pay 25 per cent more than we need to"). No, that's 33 per cent.

And of course there's nothing from the industry perspective. I don't mean the usual self-serving excuses, which certainly are covered here, but the elephant in the big pharma room: the number of new drugs that are worth bringing even as far as a stage 3 trial just keeps going down each year, and so the development cost per new drug actually made available keeps increasing (along with the incentive for companies to milk as much money out of it as possible, and out of copycats to get a fake extension on patent life). Reading Derek Lowe is recommended.

This is fine as a bloody rag to be waved from the battlements, but it's certainly not the last word on the subject. I hope this will be the most depressing book I read this year (it's edging out Chickenhawk and Save the Cat so far).