Healthy Skepticism Library item: 12524
Warning: This library includes all items relevant to health product marketing that we are aware of regardless of quality. Often we do not agree with all or part of the contents.
Publication type: news
Goldacre B.
The data belongs to the patients who gave it to you.
The Guardian 2008 Jan 5
http://www.badscience.net/?p=603
Full text:
It is rare that a bad pharma story is left untouched by the British media, but this one unfolded while everyone was drunk in December, and perhaps it was just too geeky. Luckily, you share my taste for details. Ezetimibe is a best-selling cholesterol drug with sales of more than £2bn last year. It can modify cholesterol levels but no one knows whether it cuts the incidence of real outcomes such as heart attacks, or, you know, death. Is that the bad thing?
No. There was a trial, called ENHANCE, which looked at this more closely, running for years and ending two years ago. Doctors still give out the drug and they could do with this medical evidence to help with prescribing: but it’s nowhere to be found.
Is that the bad thing? Kind of. Let’s imagine we’re playing snakes and ladders. I roll the die three times in a row then pick the best score as my actual roll, as if the other two were just practice. I invite you, winningly, to ignore those other rolls. You would rightly kick the board over, declaring I was a cheat. I waited until after I knew the results and then I chose the score that suited me best.
Similarly, in a trial, you might measure many things but you have to say which is the “primary outcome” before you start: you can’t change your mind about what you’re counting as your main outcome after you’ve finished and the results are in. It’s not just dodgy, it also messes with the statistics.
You cannot find your starting hypothesis in your final results, unless you are a time lord. What if the results were completely random, like our die, and you just selected the one six, making it look as if you were a winner?
But the people running the ENHANCE trial altered their chosen endpoint when the trial was over. They say they did so before they knew the results. That may be so, but it doesn’t look good, and they’ve now had a very serious letter from a US congressional committee demanding to know why it was done. (I seriously recommend you read it, the letter is a hoot, and it pulls no punches). Then they announced they were changing the endpoints back, in a press release FAQ dated the same day this committee’s letter was sent.
You cannot change the rules after the game has started. You cannot even be seen to do that.
To stop firms doing this kind of thing we have registers of clinical trials. Before your trial starts you put it on the register and publish detailed information about what you’re going to do – so you can’t change your mind once the game has begun or bury unfavourable data. That’s the point of a clinical trials register.
Was the Enhance trial registered? Yes. Oh no, hang on, yes, but registered on October 31 2007, more than 18 months after it finished. And looking at some original papers, this register entry in fact had the cheeky altered outcomes. So is that the really bad thing?
No. Here is the very bad thing. It also transpires there are more unpublished studies on Ezetimibe, from five years ago, and some regard safety. A Schering executive, Robert J Spiegel, says the firms had not considered those studies scientifically important enough to publish. Some, he suggests, could be eventually placed in the public domain.
And that is the final straw: because all of this breaks the most fundamental and sacred contract in clinical research, the moral contract between patient and researcher. Patients are in the hands of researchers on the understanding they are taking modest risks to benefit mankind as a whole, to improve prescribing. Ezetimibe may well be safe. But nonetheless the data belongs to patients – and to the people whose bodies are used.
