Ben Goldacre is a self-proclaimed “nerd evangelist”. He’s probably best known for his Guardian column and first book, Bad Science, which managed to be belly-laugh funny as he skewered pseudoscience, quackery, and more, with compelling logic and calm, sardonic, reasonableness. He’s a doctor, academic, and journalist, and you only have to listen to him speak to be convinced he cares deeply about the things he writes about. He makes as much noise about them as he can because he passionately believes somebody has to.
He was at the Brighton Science Festival on 2nd March to talk about the topic of his new book Bad Pharma. Unlike Ben’s previous writing, it is not particularly funny. There’s an occasional smirk to be had, sure, but you get the feeling it’s not trying to be funny, simply because he doesn’t think it’s a laughing matter. It is one thing when some quack is depriving a handful of idiots of their pocket money, but quite another when it is literally a matter of life and death – on an industrial scale.
He lost no time driving this point home by leading with the story of lorcainide. This is one of a class of drugs widely prescribed in the 1980s to people who had suffered heart attacks, as it was thought they saved lives by supressing irregular heart rhythms. But when a large trial was conducted to establish whether they really did reduce deaths (the Cardiac Arrhythmia Suppression Trial (CAST), which published results in 1991), it turned out they actually increased the risk of death. It’s estimated the drug’s popularity may have led to what he called “a biblical death toll” – of between 100 and 150 thousand Americans. This was all down to people acting with the best intentions, according to plausible beliefs.
We were then told about a 1980 study on lorcainide, in which nine out of 49 participants on the drug had died compared to one out of 46 given a placebo. Nobody knew about this. Why? The drug’s development was “abandoned for commercial reasons” and the research hadn’t been published. In an article published after the CAST trial, the researchers described these results and suggested they “might have provided an early warning of trouble ahead.” That’s a euphemism for the deaths of tens of thousands of people. Regarding the fact it wasn’t published at the time, they wrote: “it now provides an interesting example of publication bias.”
This is the heart of the matter. Published research has shown that not all research gets published. The majority of clinical trials are conducted, or at least financed, by pharmaceutical companies. Results that show their drugs in a good light are far more likely to be published than those that don’t – and they are perfectly entitled to withhold information like this. Ben explained: “We know from the best currently available estimate […] that about half of the trials that have been completed don’t go on to be published, and that trials with positive results are about twice as likely to be published.”
Think about that for a second. We like to think doctors make decisions based on a complete picture of all the scientific evidence. Not only is that not true, but the evidence they’re using is systematically biased to overestimate benefits or underestimate harms. Because of this we can’t know what the true benefit/risk balance is for many drugs (and all drugs have side-effects) and we can’t compare the relative benefits of different drugs to decide which is best. Imagine there’s one drug that saves six lives out of 10 and another that saves eight – but we don’t know that because we don’t have all the evidence. That’s a potential net cost of 2 human lives out of every ten people needing that kind of treatment. Apparently some drug company employees have a hard time understanding that point.
Are all drug companies evil?
Of course, not all ‘pharma’ is ‘bad pharma’. Drug companies have developed some of the most beneficial science mankind has ever known, saving countless lives. Ben was quick to point this out, but he doesn’t believe this gives them free license to distort evidence in ways which demonstrably harm people. Nor is he saying that all people working for drug companies are “bad, evil, people”. It’s more a case of “misaligned incentives” and a broken system.
And although big pharmaceutical companies take the brunt of the attack, it’s really a critique of the whole edifice of medicine: researchers who agree to be gagged by contracts, journals who fail to check proper methodology, editors who fail to enforce trial registers, regulators who maintain the culture of secrecy (and suffer from the same “revolving-door politics” as any other industry-related government structures). Even the doctors simply too busy to read past industry marketing – all must shoulder part of the blame for letting them get away with it.
Although this is weighty, important stuff, he didn’t forget people were there, at least in part, to be entertained. One of the highlights was a great explanation of funnel plot graphs – just so he could show us you can look for “evidence of publication bias – in studies of publication bias!” Explaining statistics to a lay audience is an achievement in itself, but managing to make it funny puts Goldacre in a different league. Never talking down to the audience, he’s like a likeable, articulate nerd, explaining something to less informed but intelligent pals. Without dumbing down, he led us through understanding and realisation to shock and anger – which, of course, was the object of the exercise.
The second half of the show was a Q&A session hosted by Sile Lane from Sense About Science. Dr Goldacre teamed up with SAS (and a few other organisations) to launch AllTrials.net. This is a campaign calling for all trials to be registered and all trial data to be released, including full clinical study reports, for all currently used treatments. It’s supported by over 200 organisations, including over a hundred patient groups, NICE, IQWiG (the German equivalent of NICE), the Royal Society of Medicine, the Medical Research Council, the Cochrane Collaboration, and the Wellcome Trust. Most excitingly, one of the biggest drug companies in the world, GSK, who were fined $3bn last year for “acts of criminal and civil fraud, including hiding data”, has recently signed up – which will hopefully open the gate for more to follow.
You can read more about the campaign, and sign the petition, on the website, or keep up to date with progress on Ben’s blog. If you want to know all the grimy details of specific companies’ bad behaviour, I recommend picking up Ben’s book.