I just said this on Radio 4, Homeopath David Spence responds on the show afterwards.
You’ll be able to listen to it again over the internet from 3pm:
Homeopathy is certainly popular, and I have no problem with people wanting to use it, enjoying it, and feeling a benefit from it. What I do object to is people misrepresenting the evidence for how useful it is, or misleading the public about the science behind it. Recently, a customer satisfaction survey from Bristol Homeopathic Hospital was being touted in the media as scientific evidence that homeopathy works. We were told, by the BBC amongst others, that this customer satisfaction survey “contradicted” in quotes, the robust scientific evidence which shows that homeopathy is no better than placebo.
When I say “scientific evidence”, there’s nothing mysterious or complicated here: you simply take a group of people who’ve been to a homeopath, and for half of them, you switch their homeopathic pills for placebo pills. It turns out that the people who get the placebo pills do just as well as the people getting the real homeopathy pills.
Sometimes, a study crops up where homeopathy performs better than placebo. But if you collect the figures from all the little studies, add them all up, and do one big count, you find, overall, again, that homeopathy is no better than placebo. That’s called a “meta-analysis”, and there was a very large wellâ€“reported one recently in the Lancet which showed, again, that homeopathy was no better than placebo.
Homeopaths have reacted to this placebo finding in various amusing ways. Many have simply denied that this damning evidence even exists: The British Homeopathic Association provide a list of systematic studies called “meta-analyses” on their website. I say list, but in fact, they only mention a few small flawed ones that go in their favour, and simply ignore all the large robust studies showing homeopathy is no better than placebo. I think that’s very deceitful.
Some homeopaths say, well, homeopathy, its not about the pills, it’s about the consultation, the ceremony, the whole process, comparing it to placebo is missing the point. A view I heartily endorse: I believe placebos are very powerful, and homeopaths are good at maximising the placebo effect. But to do that, they have to maintain the illusion that the pills work, and so they must deny the placebo evidence.
So what is this new evidence from Bristol, that we’re invited to believe can contradict all these large careful trials?
They took a few thousand patients, in a homeopathy clinic, and asked them if they thought they’d got any better since having homeopathy. That’s a customer satisfaction survey. It’s an interesting finding. But it’s got some major flaws: and most of them were glossed over by the media, and by the homeopaths who did the survey.
Firstly, for all we know, the patients could have got better by themselves. As people often do. Lots of things, even chronic illnesses, wax and wane: the menopause, one of the illnesses they studied, clearly gets better with time. For all we know, people on placebo might have felt better after that much time. People on a waiting list, even, might have felt better. Their study assumes that homeopathy takes the credit for the perceived improvement.
And there’s so much information missing. Were the people in their clinic also receiving conventional treatments? Presumably many were. For all we know, that’s what got them better. And what did they measure? They didn’t bother to use any of the standard research questionnaires for how unwell people feel, nor did they compare how unwell they felt to start with, with how unwell they felt later. People had to remember, on the spot, how unwell they were in the past, and then tell the homeopath whether they felt better or not.
And who were the homeopaths asking? People who chose homeopathy. And even then, they only looked at the people who came back to their clinic. They tell us nothing about the people who never came back after the first appointment. Did they decide it was rubbish? Did they get worse? Did they die? They are conveniently ignored, wiped from history, just like the British Homeopathic Association ignores the trials it doesn’t like.
Simply ignoring inconvenient patients, and inconvenient data, isn’t science.