Jan 31 2011

Homeopathy Pseudoscience at the HuffPo

Dana Ullman, a notorious homeopathy apologist, actually has a regular blog over at HuffPo. For those of use who follow such things, the start of his blog there marked the point of no return for the Huffington Post – clearly the editors had decided to go the path of Saruman and “abandon reason for madness.” They gave up any pretense of caring about scientific integrity and became a rag of pseudoscience.

Ullman’s recent blog post is typical of his style – it is the braggadocio of homeopathy. I am sure others will skeptically dissect his piece so I won’t go into every point here. I want to focus on Ullman’s claim that the clinical and basic science research supports homeopathy. Here is the paragraph on which I want to focus:

Most clinical research conducted on homeopathic medicines that has been published in peer-review journals have shown positive clinical results,(3, 4) especially in the treatment of respiratory allergies (5, 6), influenza, (7) fibromyalgia, (8, 9) rheumatoid arthritis, (10) childhood diarrhea, (11) post-surgical abdominal surgery recovery, (12) attention deficit disorder, (13) and reduction in the side effects of conventional cancer treatments. (14) In addition to clinical trials, several hundred basic science studies have confirmed the biological activity of homeopathic medicines. One type of basic science trials, called in vitro studies, found 67 experiments (1/3 of them replications) and nearly 3/4 of all replications were positive. (15, 16)

Those numbers are references that allegedly support his claims – 14 papers (they are not all studies, some are reviews) that allegedly make the case that homeopathy works. Most reader do not independently check references to see if they say what the author claims. Some may foolishly assume that the editors at the HuffPo have done that already.

First, Ullman is a notorious cherry picker. Any large and complex body of research will have enough noise that you could support just about any claim you wish regarding the research in you cherry pick only results that support your conclusions. The way to get to the essence of a body of research is through a systematic review – a review that looks at all the research and examines each piece for quality. You need to examine the relationship between quality of research and outcome. Ullman, rather, prefers to simply count studies.

I don’t know if his claim that most homeopathy studies show positive results is true, but I am willing to concede that this is probably true – because this is true of most research areas, even into therapies that we now know do not work. We know from the work of John Ioannidis that most published studies, in retrospect, are wrong. This is because there is a large amount of preliminary and poorly controlled research leading up to the large definitive trials that finally answer questions. Preliminary research is unreliable and biased – most of it is wrong. But we can still get to reliable answers in the end. Meanwhile, there is also researcher bias, publication bias, and the various placebo effects that conspire to make medical research look positive, even when there is no effect.

Ullman’s first reference to support his claim, however, is this meta-analysis: Are the clinical effects of homeopathy placebo effects? A meta-analysis of placebo-controlled trials. This study does not show that most published homeopathy studies are positive – that’s not what a meta-analysis is for. Here is what they concluded:

The results of our meta-analysis are not compatible with the hypothesis that the clinical effects of homeopathy are completely due to placebo. However, we found insufficient evidence from these studies that homeopathy is clearly efficacious for any single clinical condition. Further research on homeopathy is warranted provided it is rigorous and systematic.

This is hardly a ringing endorsement of homeopathy – and that’s the best reference Ullman could come up with. I disagree with the authors that the evidence is not compatible with placebo – I think it is, even by their own data. The authors should read John Ioaniddis. These results are perfectly compatible with just the expected noise of clinical research. But they were on the money with their second sentence – you cannot conclude from the evidence that homeopathy actually works for any specific indication. This is a good clue in itself that we are dealing with noise – when you focus on any one indication, the evidence is not there. Yet Ullman includes this as a reference in a paragraph in which he claims the opposite.

His second reference to back this point does not even address his point. It was a re-analysis of the Shang study that showed that homeopathic treatments are placebos, and the analysis concluded that “The conclusions on the effectiveness of homeopathy highly depend on the set of analyzed trials.” Well, of course they do. That’s why systematic reviews are better than meta-analysis. What do the systematic reviews of homeopathy show? Edzard Ernst did a systematic review of systematic reviews of homeopathy – that’s about as thorough as you can get. He found:

The findings of currently available Cochrane reviews of studies of homeopathy do not show that homeopathic medicines have effects beyond placebo.

Ullman then goes on to claim that there is evidence for homeopathy for specific conditions – despite the conclusions of his own references that he neglected to mention. I have to note at this point that Ullman takes on the skeptics in his article, writing:

It is remarkable enough that many skeptics of homeopathy actually say that there is “no research” that has shows that homeopathic medicines work. Such statements are clearly false, and yet, such assertions are common on the Internet and even in some peer-review articles.

This is a typical Ullman strawman. Skeptics don’t say there is “no research” – what we say is that there is “no good research” – meaning large, blinded, placebo controlled trials that show a replicably positive effect. What we do see is the positively-biased noise of placebo vs placebo research. The better controlled the study, the smaller the effect and greater the chance of no effect. Systematic reviews reveal this pattern – Ullman’s cherry picking does not.

Ullman references one study and his own review for the next claim dealing with rhinitis. But an independent review, which Ullman did not reference, found:

Some positive results were described with homeopathy in good-quality trials in rhinitis, but a number of negative studies were also found. Therefore it is not possible to provide evidence-based recommendations for homeopathy in the treatment of allergic rhinitis, and further trials are needed.

Next up is homeopathy for influenza. He chose a Cochrane review of Oscillococcinum for influenza. There are two big problems with this reference. The first is the conclusion:

Though promising, the data were not strong enough to make a general recommendation to use Oscillococcinum for first-line treatment of influenza and influenza-like syndromes. Further research is warranted but the required sample sizes are large. Current evidence does not support a preventative effect of Oscillococcinum-like homeopathic medicines in influenza and influenza-like syndromes.

It seems Ullman saw the word “promising” and his eyes glazed over with such joy that he could not read the rest of the conclusion. He also missed the other major problem with this reference – it has been WITHDRAWN (in big capital letters at the beginning of the title). So even the wishy-washy support for this treatment was thought to be not up to the Cochrane’s usual standards (which, in my opinion, have slipped recently). If you want to see how silly Oscillococcinum is (beyond the generic homeopathic silliness) read this article by Mark Crislip.

I could go on in detail, but it will get tedious. Suffice it to say that the rest of Ullman’s references show the same pattern – they are to small or unblinded studies with weak evidence, or reviews of the same. Most are unblinded, which in this context means they are worthless. The one blinded study he directly references is for homeopathy in ADHD – which was a small study with barely significant results. Again – I prefer systematic reviews, like this one, which concludes:

There is currently little evidence for the efficacy of homeopathy for the treatment of ADHD. Development of optimal treatment protocols is recommended prior to further randomised controlled trials being undertaken.

Conclusion

Again – there is no good evidence for homeopathy, but there is low-grade evidence that apologists like Ullman can cherry pick and misrepresent. Ullman’s references do not support the claims he is making – sometimes directly contradicting them. But most readers will just see lots of reference numbers after Ullman’s claims and be impressed – and that’s probably what he is counting on.

The goal of the apologist is to provide cover, not to make a fair and scholarly assessment of the evidence. I think we can see what Ullman is doing here.

Share

22 responses so far

22 Responses to “Homeopathy Pseudoscience at the HuffPo”

  1. Karl Withakayon 31 Jan 2011 at 11:52 am

    FYI:

    RE: “If you want to see how silly Oscillococcinum is (beyond the generic homeopathic silliness) read this article by Mark Crislip.”

    That post lists Peter Lipson as the author

  2. Enzoon 31 Jan 2011 at 12:09 pm

    I think you caught Ullman doing something I like to call review pulling. It’s when you read a single review, take it on faith everything in it is right, and reference a separate article by using the references from the original review without reading them. Intellectual laziness at its best.

    Regardless of what the real science has to say about homeopathy, if you are going to claim you are an expert (or go against the experts publicly) then you could at least read the articles you cite. And not just the abstracts!

  3. bluedevilRAon 31 Jan 2011 at 12:12 pm

    Karl, I think this is the article he was referring to by Dr. Crislip. I remember reading it not that long ago. It tackles osillococcinum in traditional Crislip fashion.

    http://www.sciencebasedmedicine.org/?p=9064

  4. Steven Novellaon 31 Jan 2011 at 12:12 pm

    Karl – That was not the SBM article I meant to link to. It’s fixed now, thanks.

  5. daijiyobuon 31 Jan 2011 at 2:42 pm

    When science is used as the sword of sectarian medicine…

    It leads to a certain kind of self-inflicted absurdity.

    Why don’t they just claim the whole homeopathy thing a religion,

    get some nice duds, some symbols, an odd name for their deity,

    like Dynamis or Lebenskraft,

    and carve out a niche on the whackaloon fringe.

    -r.c.

  6. thequiet1on 31 Jan 2011 at 6:51 pm

    daijiyobu,

    There’s no need to bother with the first few steps when they already have the last one taken care of :)

    To be honest, I think they could drop their whole sciencey charade and it wouldn’t make a difference to whatever level of acceptance homeopathy has. It’s not like people who use homeopathy do so because they sought out impartial evidence that it works. It’s underlying premise is already so absurd as to be immune to satire.

    Once a person establishes that something works through personal experience evidence becomes superfluous to their needs. They ‘know’ it works. Homeopaths could drop the science charade today and claim magic fairy’s make it work, people who already ‘know’ it works won’t be budged from their position.

    In fact, I’d wager their resolve to defend homeopathy would be strengthened.

  7. cwfongon 31 Jan 2011 at 7:30 pm

    If these charlatans sold only to the faithful, and were satisfied with that, they wouldn’t be responsible for what seems like half of the advertising on television – which used to be confined to cable channels and apparently was so successful that the broadcast channels are being overrun with it.
    Not to mention the google ad running on this page as I type.

  8. Khym Chanuron 31 Jan 2011 at 8:02 pm

    I wonder why Ullman would want to support Oscillococcinum, since it doesn’t follow the “Law of Similars”:

    1) It was invented by someone who thought that the flu was caused by a vibrating bacteria which was also found in duck livers, but the flu isn’t caused by bacteria.

    2) No such vibrating bacteria exists, so it can’t cause the same symptoms as anything.

  9. wertyson 31 Jan 2011 at 8:18 pm

    Braggadocio is exactly the word to describe Ullman’s characteristic style. Those with a penchant for personality psychology will, once they have follwed his tweets and drive-by comments for a while come to note the combination of grandiosity, entitlement, shallowness, lack of empathy and fits of rage that typify the high-functioning narcissist. Ullman has an extreme love-hate relationship with science. he seems to recognise that it is the true path to legitimacy, but he is so thin-skinned about any reasonable criticism of his positon that he can never get the science right.

  10. daijiyobuon 31 Jan 2011 at 8:36 pm

    Bob Park has something up about Montagnier

    http://bobpark.physics.umd.edu/WN11/wn012811.html

    -r.c.

  11. BillyJoe7on 31 Jan 2011 at 10:29 pm

    cwfong,

    “Not to mention the google ad running on this page as I type.”

    Funny, there are no google ads running on my page as I type.

  12. nybgruson 31 Jan 2011 at 11:02 pm

    Billyjoe: I have five. 4 of them are about how to learn to be a homeopath (including an online course!) and 1 is about asking an expert about your itchy scalp. Go figure

  13. BillyJoe7on 01 Feb 2011 at 4:05 am

    nybgrus,

    “I have five. 4 of them are about how to learn to be a homeopath ”

    Interesting.
    But puzzle solved: my son, who is typing away beside me here and who is a computer nerd, just reminded me that he installed an “ad blocker” on my computer some months ago. I had completely forgotten.

  14. scribe999on 02 Feb 2011 at 2:49 pm

    I don’t even bother looking once at Ullman’s crap on Huffpo any more. The dishonest way he argues with his commenters, his reliance on arguments from authority and popularity by constantly citing celebrities and historical figures (oftentimes the lead for his “column”) and his repetitious and terrible use of language just tire me out before I even reach paragraph two.

  15. hippiehunteron 02 Feb 2011 at 7:16 pm

    Lets not ever forget the scam that is homeoprophylaxis. These bastards are, along with other woo salesmen, a driving force behind the antivaccination lies that injure and kill children around the world.

  16. hippiehunteron 02 Feb 2011 at 7:20 pm

    SorryI forgot to include a link to show how we deal with these swine in Australia ( when the system works that is)

    http://www.homeoprophylaxis.com.au/

    Note how they retract the false advertising and then go on to lie about the “striking success of Homeopathic Medicine”

    AND link to the antivaccination liars at the AVN.

    They truely are scum of the lowest order GGGGGRRRRRR

  17. BillyJoe7on 02 Feb 2011 at 10:39 pm

    In fact, hippiehunter, they continue to make the same claim that homeoprophylaxis is an effective alternative to vaccines.

    So where exactly have we won.

    Sympathisers will simply see the TGA warning as Big Pharma trying to block the competition.
    The site almost wears the warning with pride.

  18. tmac57on 03 Feb 2011 at 10:17 am

    BillyJoe7- Do you know what the trends are in vaccine uptake in Australia? That might be a good indicator of which side is having the greater influence.

  19. zen_arcadeon 07 Feb 2011 at 2:38 am

    Back when HuffPo wasn’t so aggressively censoring comments that question their tireless promotion of pseudoscience I had a few disturbing volleys with Ullman. He mostly came across as a toad, quick to question credentials and tout his own without responding in any meaningful way to my questions. In my local library (suburb of Boston, not exactly backwater) I was disturbed to find countless seemingly-identical books by Ullman shelved a few feet away from reliable medical volumes. No wonder scientifically-illiterate people perceive him as legitimate medical authority, at least at first glance.

    HuffPost’s purchase by AOL promises to bring more brainless pseudoscience garbed as “health news” to the masses. Shame they never asked you or some of the SBM folks to blog there instead.

  20. [...] here for a good example of obfuscation by the illusion of academia. Steve Novella expertly demolishes a [...]

  21. norrisLon 13 Feb 2011 at 1:41 am

    tmac57
    The trends on vaccination in Australia appear to indicate a reduced percentage of children properly vaccinated, at least based on the increase in cases of pertussis over the last several years. Our herd immunity has dropped to the point where we have had a number of very young children die from whooping cough, and still the anti-vaccination yobbos continue to tell us how dangerous vaccinations are (hmm, what’s an American equivalent of the Australian word, yobbo? Moron perhaps?) These yobbos seem happy to have the blood of tiny babies on their hands.
    Last week I had an English lady in my veterinary surgery who told me that the breeder from whom she bought her dog told her not to allow a veterinarian to give the dog the heartworm injection because it kills Border Collies. The fact that this lady’s dog is a Cavalier didn’t seem to make a difference. I assured her that I have given over 10 000 injections of this excellent heartworm prevention with not a single negative effect. She then replied as follows:
    “Well, MMR vaccine causes autism!” I pointed out that there is ZERO proof of this, to which her reply was, “Well, a lot of parents think it does.” “But they’re wrong,” I said and pointed out that Wakefield has been deregistered as a doctor in the UK and may face charges over his intentionally fraudulent “research”. And yet, I am 100% certain that she went home convinced that she was 100% right. You cannot deal with people like this.

    Stuart

  22. npoljakon 18 Feb 2011 at 11:55 am

    Steve, if systematic reviews are so superior to meta-analyses, what is stopping them from achieving further widespread use?
    Is it because of the complexity involved and the fact that qualitative comparisons aren’t always possible given the availiable data?
    I ask this as an interested layman. Thanks!

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.