Homeopathic anecdotes aren’t data either

When I tweeted the negative sentiment below, about the author of a book on how to ‘treat’ your toddler with homeopathy, little did I know that the irate replies would still be coming in 6 hours later.

homeopathyYes, my tweet was hyperbolic, and the answer to my question – in a direct sense at least – is most probably zero. But what the question hopes to provoke is reflection on the indirect consequences of recommending that parents treat one-year-old babies with homeopathic ‘remedies’ instead of medicine. Even if, as one person admonished me, the book claims to “aid basic ailments like constipation and insomnia. Hardly life-threatening.”

First, because the reply presumes that parents can diagnose a basic ailment in the first place. If a parent is told that her child’s constipation can be treated with homeopathy, she might persist with that course of treatment for long enough that the problem becomes more than “basic”, requiring proper medical attention. And the time wasted in seeking that, or in not giving the child proper medicine, could indeed be life-threatening. Ask Gloria Sam, the 9-month-old who died when her (perfectly treatable, and not life-threatening) eczema was ‘treated’ with homeopathy instead of a visit to a GP.

Second, because the book claims more than that, and my critic was cherry-picking examples. Other things that homeopathy can treat, only according to the Amazon blurb, are “breathing difficulties” and “vomiting”, both of which seem to be things that you’d hope concern parents more than simply inspiring the application of some sugar pills or water (in other words, a homeopathic “medicine”).

Because that is of course the third, and most important reason. Trial after trial has shown that there’s nothing to it beyond the placebo effect, something that a group of friends and I satirically demonstrated by joining the 10:23 protests a couple of years ago, where we each downed a bottle of a homeopathic ‘remedy’ (I think mine was arsenic). Here’s James Randi doing the same, taking a bottle of ‘sleeping pills’ as he often does to make this point (the clip also includes him making other arguments worth hearing).

And no, there’s no good evidence to suggest it has to be more than placebo, “because it works on animals” – we’ve got no reason to believe it works on animals any better than it does on humans (in fact, the perceived effect on animals seems to simply be an effect on humans, in terms of how they perceive the treatment and health of their pets).

But evidence isn’t what defenders of homeopathy are interested in. For them, anecdotal evidence is ‘argument’ enough, even though they would never stop to think about how they would reject similarly weak claims if they came in a version they don’t like. Kitten blood! It works for me! Crystals! Prayers to the Pink Unicorn! (Or, prayers to a ‘real’ god, just not one of the ones you happen to believe in.)

Racists defend their views with anecdotal evidence, as do sexists – reality is ignored in favour of confirmation bias. And we don’t think that doing so is a good, or a reasonable thing to do. Because the evidence is meant to matter, and the evidence isn’t “up to me”, and the experiences I might have had or not had. Part of the point of science is to provide us with resources that offer objective guidance, because we go into decisions knowing that – by and large – we’re too prone to various cognitive errors to be trusted.

The point is that a double-standard applies in people who are willing to defend their consumption and prescription of homeopathic ‘remedies’, in that they are willing to accept a very low standard of evidence on the grounds that the risks are low – “any responsible homeopath”, I’m told, “will advise their patients to take antibiotics where necessary, or to seek conventional treatment”.

But some homeopaths are less responsible than others. The coroners report was pretty clear in highlighting how Penelope Dingle would have suffered far less harm if not for her homeopath’s advice, and more generally, as Ben Goldacre makes clear in the Lancet, homeopaths simply get in the way of effective treatment:

Homoeopaths can undermine public-health campaigns; leave their patients exposed to fatal diseases; and, in the extreme, miss or disregard fatal diagnoses. There have also been cases of patients who died after medically trained homoeopaths advised them to stop medical treatments for serious medical conditions.

More prosaically, you’re simply wasting money if you spend it on homeopathy. This is one of the most annoying #middleclassproblems for me – alongside things like anti-vaccination sentiments, or obsessions with angels, or The Secret – in that it’s only the middle and upper classes who have the luxury of glamorising their anecdotal evidence in such a fashion. If homeopathy worked so well – given that it’s possible to produce it so cheaply – why would Bill Gates (etc.) not simply distribute it to those dying of malaria instead?

Of course “Western” or allopathic, or “chemical” (pick your favourite pejorative term) can’t cure everything. Nothing can. But homeopathy doesn’t outperform a straight placebo, meaning that any good effect you observe after taking a homeopathic remedy can’t have anything to do with that remedy. And much of ‘regular’ (by which I mean, real) medicine does. Furthermore, the stuff that doesn’t work – and the doctors that are quacks – tend to get driven out of the market over time.

Except for homeopaths, partly because of this almost religious devotion to “alternative” medicine (and the associated conspiracies around mainstream medicine), and partly because what homeopaths say, and prescribe, involves completely unfalsifiable claims. And that’s a bad thing – not only in general, but particularly when lives are at stake.

Check out What’s the Harm for a partial catalogue of homeopathy’s victims.