Parliamentary Science and Technology Select Committee on homeopathy today

November 25th, 2009 by Ben Goldacre in homeopathy | 66 Comments »

I gave evidence at the Parliamentary SciTech committtee today for their enquiry into whether the government had used scientific evidence properly in making their decisions about MHRA licenses for homeopathic pills, and homeopathy treatment on the NHS. This was a mini-enquiry as a result of interest expressed by the public, which is excellently democratic, you can see the whole thing online here, and some of it is quite good fun.

www.parliamentlive.tv/Main/Player.aspx?meetingId=5221

Personal highlights, from memory, include:

  • Paul Bennett from Boots saying that there is no evidence showing homeopathy pills are effective at treating any conditions, but Boots are happy to sell them anyway, since the MHRA have given them a license. Wahey!
  • Robert Wilson, the head of the homeopathic pill manufacturers’ association, giving us a lecture on trial methodology, explaining that: 65 people in any trial cannot be statistically relevant (which is rubbish, if you have a pill that cures everyone from an incurable condition then 40 people is fine, hell, a dozen would do); that if you talk to statistically literate people they will tell you that size is everything (it’s one of many things); and anything less than 500 people is not going to be statistically significant (which is utter, utter nonsense).
  • Edzard Ernst pwning Peter Fisher with relaxed applomb in the second half, especially when Fisher wheels out some claims on homeopathic aggravations.
  • Some angry looking people in the room staring at me very intently which, from background research by Gimpy, may have been part of a magickal ritual.
  • Evan Harris MP doing his “that’s rubbish” face as it emerged that the Royal Pharmaceutical Society are still slowly investigating a complaint about dangerous homeopathic pharmacists from 2006.
  • Ian Stewart MP talking.

One thing that will never get old for the homeopaths, it seems, is the old practise of pulling out a single trial and saying “ah, but look, pish to your meta-analyses, here is a trial where homeopathy works”. No matter how many times you point out why this is foolish and wrong, they will always think you’re just being picky, and that is why they will always give us joy.

Anyway, it’s worth digging around on the site and watching a few other sessions, if you’re that way inclined. As I’ve said before, I think select committees are really interesting and informative, the one place where politicians do what you’d want them to do all the time, which is to say, sit down and have a good think about policy.

www.parliamentlive.tv/Main/Player.aspx?meetingId=5221


++++++++++++++++++++++++++++++++++++++++++
If you like what I do, and you want me to do more, you can: buy my books Bad Science and Bad Pharma, give them to your friends, put them on your reading list, employ me to do a talk, or tweet this article to your friends. Thanks! ++++++++++++++++++++++++++++++++++++++++++

66 Responses



  1. Ben Goldacre said,

    November 25, 2009 at 7:08 pm

    incidentally, at one point when ian stewart is banging on about nils bohr and homeopathy i say “physics isnt something that interests me”, i meant “the physics of homeopathy isnt something that interests me”. physics rules, and you shld totally buy yourself this for xmas bit.ly/4XosZs

  2. danielrendall said,

    November 25, 2009 at 7:59 pm

    When will homeopaths learn that the less they say, the more powerful their arguments will be? To achieve total NHS domination, they should say nothing…

  3. Ennui said,

    November 25, 2009 at 8:10 pm

    I live on the same road as the “Homeopathic Hospital” that the NHS run. I worry that it’s driving down the house prices in the area. After the recent reports of slashed MRC funding in the papers, I hope the “Homeopathic Hospital” is shut down and its NHS funds are allocated elsewhere.

    (P.S. Ben, boo for cancelling the UCL Division of Infection & Immunity talk. I was looking forward to that for ages. 🙁 )

  4. stever said,

    November 25, 2009 at 8:18 pm

    Good effort Ben.

    I gave evidence to the Home Affairs select committee on cocaine (as it were) a couple of weeks ago. Mentioned Amy Winehouse only to discover her Dad, somewhat bizarrely was on immediately after me and not best pleased.

    transform-drugs.blogspot.com/2009/10/transform-give-oral-evidence-to-home.html

  5. Dead Badger said,

    November 25, 2009 at 8:19 pm

    I liked the bit about 500 being a minimum requirement for statistical significance, too. I noticed that Wilson later mentioned a paper by Möllinger et al as proving homeopathy worked, so I looked it up. Would you care to guess how many subjects it involved?

    25. And that’s three study groups.

    It also did not study efficacy, but instead counted “remedy-specific symptoms”, whatever they are. Apparently since arsenicum and natrum cause fewer “remedy-specific symptoms” than placebo (although how there can be a remedy-specific symptom for placebo is beyond me), they must therefore be different to placebo. And therefore, of course, homeopathy works. Right?

    Either way, cracking entertainment. I’m going to move to Oxford so I can vote for Evan Harris.

  6. scotslawstudent said,

    November 25, 2009 at 8:20 pm

    danielrendall:

    Actually what they need to do is dilute their substantive arguments in massive quantities of waffle. To their credit they’re trying their hardest in this.

  7. xtaldave said,

    November 25, 2009 at 8:50 pm

    “The best possible outcome of having an argument with a Homoeopath is that you won an argument with a Homoeopath.”

    How Homoeopathy EVER got into the NHS is beyond me.
    There is no way public funds should be used to pay for unproven quackery.

  8. allopathicdoc said,

    November 25, 2009 at 9:10 pm

    We had a homeopathic GP in our practice. He wrought havoc by medicalising the non-medical and created enormous dependency in patients who would come along for a remedy for every minor symptom. He clogged up the system and confused the patients as well. He wasn’t a clear thinker and he transmitted that to the patients. It was enormously disempowering for them. Since he has left all those people have stopped consulting for every itchy little finger and hot left eyebrow and they seem none the worse for it. In fact, they don’t even seem to miss it.

    I am all in favour of people having the liberty to consult (mostly) harmless quacks, but for God’s sake make them pay for it. The idea that this is subsidised is absurd. I’d like to think this is the beginning of the end for homeopathy on the NHS, a long overdue triumph for rationalism, but with Charlie on board I somehow doubt it.

  9. jdc said,

    November 25, 2009 at 9:14 pm

    There were a couple of moments where other people giving evidence seemed to repeat (verbatim) comments that you had made – perhaps the intention experiment went a bit wrong?

    On the whole, I enjoyed watching/listening to the committee – but felt that a few knob gags would have lightened the mood. At least Evan managed a gag about homeopathy remembering people’s poo.

    One minor point: at least two people (I think it may have been three) referred to the powerful placebo effect. I’ve written about this recently and as far as I can tell, apart from in terms of pain, there seems to be no good evidence of a powerful placebo effect. Is there any reason to think that placebos are really useful outside clinical trials (with the possible exception of pain)?

  10. Mikee23 said,

    November 25, 2009 at 9:30 pm

    I managed to watch some of it this morning – it was pretty entertaining stuff. I particularly enjoyed the homeopath boasting about his baby teething ointment. The placebo effect can’t work on babies, you know! What a dimwit. Who was the guy who kept asking you if you could “categorically” prove it didn’t work?

    I was most amused by your physics comment – shame you didn’t mean it. Physicists are often pretty flaky. I reckon too much cosmology drives you a bit batty.

    Changing the subject Ben, you have to look at the insane “23 Year Coma” story. Facilitated Communication is alive and well, and getting pages of uncritical endorsement in our beloved Grauniad! Check out these pics. Madness!

  11. smithers said,

    November 25, 2009 at 9:43 pm

    A magickal ritual? Is Ben about to be turned into a frog?!

    Got to love the fact that their response to questioning of their pseudo-scientific remedies is to er invoke arcane magicks 🙂

  12. Spencer said,

    November 25, 2009 at 9:52 pm

    oh my gosh. I did so laugh when having watched the first hour, decided to check the comments here and read Dead Badger’s research on the number of participants of the trial that was quoted by Wilson “… by Möllinger et al as proving homeopathy worked”

    haha I cackled loudly. If only this could have been pointed out at the time.

    Only 25? naaaa balderdash. Needs to be more than 500 nudge nudge wink wink ahahah

  13. matthewc said,

    November 25, 2009 at 10:10 pm

    Hummm Yep brought a smile to my face and laughed at stever’s post expect the evils where dished out.

    I have a thought though, is homeopathy the last vestiges of the decline of magic in medicine, too be honest I would much rather they did a witch dance with some crazy out fit on it would defiantly make me feel a whole lot better.

  14. smithers said,

    November 25, 2009 at 10:59 pm

    I think Edzard Ernst could be my new hero. Each time someone to his left says something ridiculous he just smiles and ever so politely pwns the living crap out of them!

  15. Richard Palmer said,

    November 25, 2009 at 11:00 pm

    @allopathicdoc

    Out of interest, how did you end up with a homeopathic GP in your practice?

  16. timbod said,

    November 25, 2009 at 11:08 pm

    Robert Wilson made the basic statistical mistake of confusing the two statements
    A) I conclude that a drug has a beneficial effect and
    B) I conclude that a drug has no beneficial effect.

    As Ben points out, if a drug has a very beneficial effect which “works” for most patients then only small numbers (eg 10) are needed in the trial to demonstrate that (A) is statistically significant. If you wish to show (B) then you do need a large number and 500 would be typical. (It’s was easy to find an undiscovered Pacific island by stumbling across one, it’s was a much harder task to say there are no more Pacific islands to be discovered as you have to chart the whole of the ocean).

    The problem for Wilson is that for a drug to be approved by NICE, the homeopaths need to show A) (more: they have to show that that (A) is the conclusion in the majority of properly conducted trials), the argument that no-one has yet conducted a trial to falsify (B) is just nonsense.

    He might as well be saying “take this pill to cure your discombibliatitus, it must work because no-one has yet shown that it doesn’t cure discombibliatitus”

  17. chris lawson said,

    November 26, 2009 at 2:32 am

    Even a binomial trial with a sample size of 1 can be statistically significant depending on the hypothesis being tested.

    For instance if someone hands you a coin that looks fair but your friend tells you that the coin will land on tails 9 times out of 10, and you toss the coin and get heads, then the short confidence interval on that is 0-0.776 which means that you already have >95% confidence that your friend is wrong.

    If you toss the coin again and get another heads, then your confidence has gone up to >99% that your friend is wrong. A third heads will get >99.9% confidence. Not bad for 3 samples of a binomial outcome.

    The calculator used is at www.causascientia.org/math_stat/ProportionCI.html?

  18. BlinkingRegistration said,

    November 26, 2009 at 2:41 am

    Any chance of someone prividing the timestamps of the highlights? It’s a bit long to scan through.

  19. Jbags said,

    November 26, 2009 at 4:10 am

    #17 BlinkingRegistration

    It is genuinely worth watching the first hour, it is all interesting stuff. The first five minutes gives a good taste of what you find later on. At about 10 minutes (I think) there’s an attempt to pin Robert Wilson down on the subject of meta-analyses, it amused me watching him wriggle. At about 54:00 the point about Nils Bohr comes up, well worth catching.

    It was this section that actually outraged me the most, the analogy of homeopaths to particle physicists, in their struggle for evidence and proof of theories. I couldn’t believe they let him get away with that! The fact that there is an absolutely huge (and importantly, sufficient) mountain of evidence studying the real world effects of homeopathic treatment (which we can measure) is the reason it hasn’t been widely accepted and not because the general scientific community “haven’t recognised the genius of homeopathy”.

    His point about anaesthesia etc (treatments which work where we are unable to identify the mechanism) is entirely moot. We don’t know the mechanism, that is a gap in the scientific knowledge, alright, but you can’t point to one hole and say that therefore justifies your entire theory. OK the mechanism of anaesthesia is unknown, _the effects_ however are very much entirely known, they have been studied and well documented (if he doubts this he’s free to refuse a local next time he has serious dental work done). The _effects_ of homeopathy are also well known i.e. no better than placebo.

    It infuriates me no end when homeopaths bring up this distraction of unknown mechanisms such as anaesthesia, it is irrelevant and a smokescreen, pointing at the incompleteness of scientific knowledge to bolster their claim. Its shameful, and they shouldn’t get away with it.

    Robert Wilson coming out with “bespoke remedies” (quote at around 57:00) was also one of my favourites and made me laugh.

    Well done Ben, irrespective of the decent scientific arguments you presented, you were also the most eloquent member of the panel, and that I thought was a serious help to your cause; serious scientific argument, clearly and concisely presented. Bravo.

  20. BlinkingRegistration said,

    November 26, 2009 at 4:29 am

    Many thanks, Jbags. I don’t have a spare hour this week, so it’s handy to be able to check out the best 5/10 minutes.

  21. ellieban said,

    November 26, 2009 at 5:41 am

    Absolutely agree with Jbags. I loved watching the whole thing, especially the first 10 mins and I was very impressed by Mr Phil Willis, the chairman.

    But how on earth did you let them get away with all that physics nonsense? Even allowing for the need to succuss a remedy to activate it, following homeopathic theory to its logical conclusion means that all water must contain the memory of every chemical and at exceedingly high potencies. It isn’t the memory of water that homeopaths need to explain (although that would be good too), it’s the ability of the remedy to pick the correct memory for each disease. I suppose “but maybe we just don’t understand the mechanism” applies to that too though, so the real answer had to be “prove you have an effect and THEN we’ll talk about how it works”.

    One of my favourite moments was one of the homeopaths (I’m not sure who, it may have been Robert Wilson or it may have been someone at the start of the second session)saying “we don’t cherry pick data because… um… (whining) they do it too!”

  22. ellieban said,

    November 26, 2009 at 6:03 am

    BTW, there is a blow by blow text account here:

    www.guardian.co.uk/science/blog/2009/nov/24/homeopathy-science-technology-committee

    Useful for distilling the most relevant parts from the less interesting (read “homeopath dominated”) second session.

  23. Grumpy Bob said,

    November 26, 2009 at 6:27 am

    It may be democracy, but the website excludes me as a Linux user. I installed the recommended plugin, but it won’t work with that version of Silverlight.

    Is there another format available?

    R

  24. Jbags said,

    November 26, 2009 at 6:55 am

    Bob,

    I found it hosted by the beeb here:

    news.bbc.co.uk/democracylive/hi/house_of_commons/newsid_8371000/8371398.stm

    any luck?

  25. kinginsan said,

    November 26, 2009 at 7:01 am

    @Grumpy Bob: it worked fine for me in Linux (Ubuntu 9.10) with Moonlight:

    www.mono-project.com/Moonlight

    Agree with everyone else: hilarious stuff and the first 10/15 minutes were sheer class. I particularly enjoyed the comparison of homeopathy with prostitution!

  26. Gooster74 said,

    November 26, 2009 at 7:51 am

    Daily Telegraph in on the act focusing on the Boots clanger – although journalist still not quite grasping the dilution theory…

    www.telegraph.co.uk/finance/newsbysector/retailandconsumer/6658864/Boots-we-sell-homeopathic-remedies-because-they-sell-not-because-they-work.html

  27. CoralBloom said,

    November 26, 2009 at 8:43 am

    Robert Wilson thought the value of the homeopathic industry should be considered as relevant and important. Chicken soup is a big industry too, but that doesn’t mean we should support that industry in helping with flu and colds via the NIH! I’m glad that didn’t work with the committee.

    I liked the bit about the vigorous shaking to remove memories from water. Joy! Still trying to work out why, if this so-called memory is a property of water molecules, why de-ionising and double distilling would alter the memory state.

    Is there machine used for this – obviously with a timer to ensure the shaking happens for long enough. Maybe my uncle could make them and sell them to homeopaths at great expense, naturally.

    Well done Ben, Tracey and Prof Edzard. Succinct.

  28. SimonW said,

    November 26, 2009 at 9:25 am

    Surely the best statistical argument refutation about the 500 people is to point to those rational skeptics who’d be convinced of the power of prayer if it healed even one amputee.

    They aren’t even asking for a double blind controlled trial, although the idea of double blinding could be a source of humour: “Please deity 1 heal patient 2134’s missing limb (arm or leg as appropriate)”

  29. MarkEO said,

    November 26, 2009 at 9:31 am

    I saw the homeopaths asking for more trials as basically saying “best of 3″… “okay, best of 5″… “damn, best of 7” etc…

    I found the bit about maybe there will be explainations in the future frustrating, it missed the point of evidence completely. There is a chance of anything being explained in the future – that shouldn’t mean everyhting (including pumkins on Mars) should be taken seriously without evidence.

    @Mikee23 I pointed out the Ideomotor effect coma bloke to Ben on Twitter yesterday, I don’t know if he saw it or not – v. busy man. I,however, blogged about it yesterday – William Satire Jr. – in a jokey fashion

  30. mus said,

    November 26, 2009 at 10:13 am

    @ MarkEO: Martin tackled this as well over at layscience.net: layscience.net/node/814

  31. MrWeeble said,

    November 26, 2009 at 11:05 am

    > Out of interest, how did you end up with a homeopathic GP

    You take a competent GP with a sensible level of scientific awareness and remove one drop of blood, which you then inject into a random man off the street. Then you take one drop of *their* blood and inject it into another random bloke off the street, one mumbling to himself and shuffling aimlessly is best. Lo and behold he now has teh Awesome Power of Homeopathy and there you have your Homeopathic GP

  32. PeteX said,

    November 26, 2009 at 11:38 am

    I’m sure the witchcraft will work, so there will be no change to the regulations on labelling homeopathic remedies (cause and effect). 🙂

    I was wondering, though: is using witchcraft on MPs punishable as contempt of the House? Surely people must have bewitched MPs before, back in the dark ages, when we thought this stuff actually worked. I mean, back in the dark ages, when most of us thought this stuff actually worked… I mean…

  33. merrick said,

    November 26, 2009 at 12:13 pm

    Ennui (cooment 3):
    “I hope the “Homeopathic Hospital” is shut down and its NHS funds are allocated elsewhere.”

    Perhaps we should permit it to continue, but with homeopathic funding. Take 1 trillionth part of the active funding – I’m imagining shaving the edge off a penny or somesuch – and see if that buys cures.

  34. sciencerocks said,

    November 26, 2009 at 1:48 pm

    Thank Elvis for Ben and co.

    Working in the area of hearing and dizziness, I was interested in the reference by supporters of homeopathy to a systematic review of a ‘homeopathic’ remedy for vertigo (dizziness) called Vertigoheel. While I recognise Prof Ernst’s point that the remedy is not strictly speaking, I thought I’d look it up anyway. I suspect it is this article:

    “Treatment of vertigo with a homeopathic complex remedy compared with usual treatments: a meta-analysis of clinical trials” Arzneimittel-Forschung. 2005;55(1):23-9. Abstract is on PubMed.

    While I am no expert on meta-analysis and have not seen the original article, the abstract seems to suggest that VertigoHeel is no better and no worse than “usual therapies” to which VertigoHeel was compared, namely betahistine, Ginkgo biloba extract, dimenhydrinate.

    It is my understanding that (1) it is not clear whether betahistine is helpful for the main cause of vertigo it is used for (Meniere’s Disease) (see for example: www.cochrane.org/reviews/en/ab001873.html) let alone any others; (2) I can’t find a systematic review from an obviously trustworthy source of Ginkgo biloba for vertigo but the conclusions of all SRs (e.g. Cochrane) I have seen for other conditions (e.g. tinnitus) seem to include the phrase “there is no evidence that Ginkgo biloba is effective for…” – of course, we cannot generalise to vertigo from these; (3) I couldn’t find (in 15 min) any systematic reviews on dimenhydrinate for vertigo or dizziness – it seems to be used to reduce nausea and vomitting, which often occur with vertigo.

    Putting this very brief and flawed survey of the background literature together, it seems to me that there is a good chance that VertigoHeel is no more or less effective in the treatment of vertigo than three other ineffective treatments.

  35. skyesteve said,

    November 26, 2009 at 2:38 pm

    @sciencerocks – spot on! Or as the Cochrane reviewer put it for beta histine:

    “Main results
    Seven trials involving 243 patients were included. No trial met the highest quality standard set by the review because of inadequate diagnostic criteria or methods, and none assessed the effect of betahistine on vertigo adequately. Most trials suggested a reduction of vertigo with betahistine and some suggested a reduction in tinnitus but all these effects may have been caused by bias in the methods. One trial with good methods showed no effect of betahistine on tinnitus compared with placebo in 35 patients. None of the trials showed any effect of betahistine on hearing loss. No serious adverse effects were found with betahistine.

    Authors’ conclusions
    There is insufficient evidence to say whether betahistine has any effect on Ménière’s disease”

    If you compare one thing that doesn’t work with another thing that doesn’t work all you prove is that neither of them work!

  36. PreviousChemist said,

    November 26, 2009 at 3:05 pm

    Re: Homœopathy and anæsthesia. The difference is that anæsthetics can easily be shown to work, that is, have an effect demonstrably greater than placebo, whereas homœopathic preparations cannot. You do not need to know the mechanism of action to be able to show efficacy. Newton did not know how gravity worked, but gave a good description of its effects and shewed them to behave in a way that could be accurately modelled by the inverse square law. He did not need to know quantum mechanics to be able to do this. In the same way, even if we do not know how anæsthesia works, we can still use it. Thankfully.

  37. NeilHoskins said,

    November 26, 2009 at 3:58 pm

    Good for democracy? Maybe. When it comes down to it, though, we all know that the important decisions are made on a sofa at No 10.

    Incidentally, all the discussion in this column on the statistics of clinical trials is deeply flawed: any physicist with a knowledge of Schrödinger’s cat knows that a trial subject can get better and die both at the same time 😉

  38. Grumpy Bob said,

    November 26, 2009 at 6:48 pm

    @kinginsan – Ubuntu 9.10 here, just installed moonlight, but doesn’t appear to fit the bill. Perhaps the repos are a bit out of date…

    R

  39. Grumpy Bob said,

    November 26, 2009 at 7:09 pm

    @Jbags – BBC link works well, many thanks.

  40. adamk said,

    November 26, 2009 at 7:09 pm

    Perhaps homeopathy is one of the theorems suggested by godel’s incompleteness theorem –

    there are theorems in every logically consistent system which are true but cannot be proved

    or something like that…….

  41. quasilobachevski said,

    November 26, 2009 at 9:00 pm

    PreviousChemist,

    I’m no expert, but I’m pretty sure that we still don’t have an explanation for gravity in terms of quantum mechanics. That’s one of the major challenge in theoretical physics, these days.

  42. quasilobachevski said,

    November 26, 2009 at 9:01 pm

    I tried to enclose that in pedantry tags, but WordPress swallowed them. Oh Ben, when will this blog have a preview button?

  43. Psythe said,

    November 27, 2009 at 1:20 am

    Interesting session – I did rather like the gentle “don’t worry; its not too technical; this is the science committee” retort to Ernst.

    They should have got David Dimbleby to take all nine of you into a room afterwards and have you argue directly with one another (with him asking the questions – I always thought that Question time would be better without the audience).

    Hmm. Do Robert Wilson and Nick Griffin really look similar or am I just associating them after watching them both wriggle under public scrutiny?

  44. mrmuz said,

    November 27, 2009 at 5:03 am

    Surprisingly watchable and informative for some nitty gritty government biz. Good stuff.

    Anyone know about those recent tests featuring highly dilute drugs the guy at the end was talking about? They couldn’t possibly be the same sort of ‘dilution’ levels the homeopaths favour, surely.

  45. lorcancoyle said,

    November 27, 2009 at 7:59 am

    BBC Breakfast being their usual pentrative investigators – homeopathy advocating GP (claiming double blind trials etc.) vs. a tabloid journalist (I wouldn’t have the time for homeopathy), very balanced.

  46. Ian Harvey said,

    November 27, 2009 at 9:49 am

    Thius Morning on the BBC one of the Magic Water merchants was given a ludicrously easy ride. news.bbc.co.uk/1/hi/health/8382265.stm I think there may be a drinking game here.

  47. csrster said,

    November 27, 2009 at 11:07 am

    I though we had decided that a Homeopathic Hospital was just a huge car park with a tiny lego-brick somewhere in the middle.

  48. rob9son said,

    November 27, 2009 at 11:11 am

    Did anyone happen to catch BBC breakfast’s horrendous discussion about homeopathy in response to the select committee this morning? A proponent of homeopathy discussed the ‘numerous trials’ that had ‘proven’ it to be effective treatment, while the other side of the debate was provided by a daily telegraph columnist who’s opinion seemed to consist of ‘I wouldn’t try it, I’m too busy’. Pathetic.

  49. skyesteve said,

    November 27, 2009 at 11:39 am

    @csrster – it used to be but in order to make the hospital more effective they took the Lego brick away…

  50. emen said,

    November 27, 2009 at 12:02 pm

    csrester,

    “I though we had decided that a Homeopathic Hospital was just a huge car park with a tiny lego-brick somewhere in the middle.”

    hahahahahaaaaaa!
    brilliant!!