Dr Des Spence is a Glasgow GP who writes a regular column in the British Medical Journal. I frequently find myself agreeing with what he writes and this week is no exception. He looks at the way doctors (and the public) are presented with “the facts”. It’s all in the way the statistics are chosen and published. He takes the West of Scotland Coronary Prevention Study as his working example. This large study showed a reduction of 32% in the deaths of men in the study who took statins to lower their cholesterol. I won’t re-iterate the detailed statistical analyses here, but that’s the “relative risk”, and another way to present exactly the same findings is to show that the “absolute risk” shows reduction in mortality from cardiovascular disease was 0.7%. That’s startling enough, but what it means is that you have to treat 715 men with the statins to save 1 life. In other words, one person benefits and 714 take the pills but don’t benefit.
I wrote about this in an earlier post where I reviewed “Reckoning with Risk” by Gigerenzer. I urge you to read either that book, or his “Gut Feelings“. You’ll never swallow “the facts” so easily again.
Heroes Not Zombies 
Waht was it Mark Twain said? Something like “there are lies, damn lies, and then there are statistics.” Something like that, I think.
Dr. B
Everything we perceive is through our own filters! It is so interesting to find out what different people “hear” in the same message.
Yep, that’s what he said, Dr.B! It’s a quote the more extreme EBM fanatics forget! Actually, a serious point I think in that context is that there is a world of difference between “evidence” and statistics!
So true, quotesqueen, and if we can park our judgements and prejudices we can actually hear what other people are hearing!