The trouble with everything
November 16th 2007, 13:24
Homeopathy comes in for yet another pasting this week. It's the turn of Ben Goldacre, that scourge of all things alternative, who writes in The Lancet and The Guardian about the un-science of homeopathy.
His attack is three-pronged, and while we've all heard the arguments before, they are worth pondering anew.
- The placebo effect: This, for Ben, is the main reason why homeopathy 'works'. It explains why we "just feel better" after being given a homeopathic pill. (For the record, I've never felt any better having been given a homeopathic remedy, so, according to Ben's measure, my negative experience proves it works!) Medicine understands the placebo effect, and discounts it when analysing the results of a double-blind placebo trial. Homeopaths also claim there are plenty of trials that demonstrate an effect far greater than placebo, but this brings Ben to his second challenge.
- Publication bias. Homeopaths just publish the good stuff, and they hide the bad results in their filing cabinets. If they published every trial, good and bad, the greater-than-placebo effect would disappear. This doesn't happen in medicine, says Ben, because of the 'clinical trials database'. Everyone can see all the trials and their results. Sadly, this is not true. In the 50 years since 1948, there were around 1 million medical trials undertaken, yet only half of these were ever published. The rest were put into, er, filing cabinets (I wonder if they are the same ones the homeopaths use?) The EU clinical trials database is not available to the public because drug companies "need to protect commercially sensitive information about drug developments", Thomas Porstner, spokesman for the German Pharmaceutical Industry Association, has said.
- Regression to the mean. This is a scientific way of stating that disease is cyclical. Some days you feel better, some days worse. It's quite possible, says Ben, that you take the homeopathic pill when you're at your worst. Then you feel better, but not because of the pill, but because of the natural cycle of the disease. This argument can just as easily be laid at the door of conventional medicine.
For Ben, all of this un-science is dangerous stuff. Homeopathy could be killing people, he argues, which is a bit strange because a minute ago he was telling us it was utterly ineffective. Anyway, that's the province of the pharmaceutical industry, and they won't give that one up without a fight.
It's true that medicine is more 'scientific' than homeopathy and other alternative therapies, and for a very good reason. With the rise of mass production, the nascent pharmaceutical industry needed to establish its products were solutions to mass health problems. It shoehorned the art of medicine into a science, but the shoe fits poorly.
In truth, even the best medical trials establish only that a drug works for some people some of the time (usually in the first 6 months of starting treatment). The real question is: will it work for you? Well, you won't know until you take it and, when you do, you may well "just feel better". But isn't that what the homeopathic patient said, too?