Adam Smith Science and Communications Officer, ANH-Intl
KEY POINTS
- The UK's first Professor of Complementary Medicine, Edzard Ernst, has retired and is seeking a successor
- In several interviews with UK and international publications, he appears to be raising his public profile
- Despite his high visibility, his research methods and conclusions are extremely questionable
- ANH-Intl awaits his next move with interest
Edzard Ernst, Professor of Complementary Medicine at Peninsula Medical School in Devon, UK, became an ex-Prof when he retired in May. According to a recent interview in the UK's Telegraph newspaper, his post will be retained and Ernst is helping to find a successor. It will be interesting to see what Ernst does next, especially since he can't wait to be “Outspoken about quackery and charlatans. I look forward to that. Hopefully, UK libel law has [sic] changed by then".
After this interview and another in the UK's Guardian newspaper, Ernst has now taken his opinions on natural healthcare to an international audience, via an interview in New Scientist magazine (login required).
Another change of heart?
Interestingly, between the end of July, when the Guardian piece appeared, and the 23rd August that saw publication of the New Scientist interview, he has gone from proclaiming that homeopathic efficacy is "absolutely nil" to saying that "I still think that homeopathy works, the question is: why?...It's a powerful placebo effect". Even in the Telegraph piece, published a whole day earlier on 22nd August, he said no positive conclusions can be drawn about homeopathy from the published research. As we've said before: which is it, Edzard? Even if you do think that the placebo effect entirely accounts for homeopathy's positive results, that's a bit different to saying that there are no positive results. Neither does it begin to explain homeopathy's usefulness in animals.
Nice flip-flops!
But we digress. Ernst is no stranger to U-turns on the effectiveness of natural healthcare. Not only that, but, as he is at pains to mention in every interview, he once worked in a homeopathic hospital, and he is on record as saying that "As a kid I was treated homoeopathically...This made a big impression on me". This was before he became the fearless Inquisitor-General of natural healthcare for whom evidence-based medicine (EBM) is the one true guiding star.
A word on Ernst's methodology
As we have pointed out before, the reason why Ernst's assessments of natural healthcare are often negative is not simply that these healing techniques shrivel under the harsh light of scientific scrutiny. Ernst's methods, which we see applied in the same manner over and over to many different forms of natural healthcare, involve selecting what he deems to be high-quality studies and performing meta-analyses or systematic reviews on them. Studies of studies, in other words.
The problem comes when trying to separate out the 'specific therapeutic effect' from the plethora of related phenomena that must also be considered, and which together make up the 'total therapeutic effect' (see Figure 1).
Figure 1. Variables operative during clinical trials that contribute to the 'total therapeutic effect'
The RCT can only separate out the specific therapeutic effect from the total effect if all of the other phenomena in the Figure 1 are independent of each other, and operate the same under experimental and real-life conditions. But just think of the practitioner–patient relationship for a moment. As we said in our earlier piece, "Crucially, the bio-physical-energetic interactions between practitioner and patient, as well as internal neurophysiological and metabolic processes, are likely to be very different in these two contrasting circumstances...it is perhaps no wonder that Ernst finds that treatment effects are so often lost in the overall noise of the experiment – falling foul to the netherworld of statistical insignificance."
We submit that it is no coincidence that many of Ernst's studies have been negative.
Fantastic Mr Fix
In addition, never forget that in our opinion, Ernst is not above a bit of figure-fudging. At one point, he and his colleagues rejected 99.8% of the available evidence on herbal medicine and concluded that "there is no convincing evidence that it is effective in any indication" [1].
Can you imagine the response if anyone involved in natural healthcare examined any aspect of orthodox medicine with the same laughable methods? There would be an outcry, and rightly so. What's so dangerous about it is that Ernst's propaganda is taken seriously by commentators in the media and presented to the public as if it represents some kind of unquestionable, scientific truth, providing useful ammunition for the anti-nature skeptics to continue their campaign. Much like Ernst's entire career, in fact.
So what's next?
Having said that, Ernst's closing comments to the New Scientist may be encouraging. "If...the clinical evidence shows that my present conclusions are wrong then I will change my mind again...it is a sign of intelligence to change your mind when the evidence changes". It will be interesting to see what happens now that Ernst is a commentator on natural healthcare, rather than a researcher.
For the full story, read our previous expose, entitled Professor Edzard Ernst: Master Trickster Of Evidence-Based Medicine?
[1] Guo R et al. Postgrad Med J 2007;83:633–37.
ANH-Europe Homepage
Comments
your voice counts
27 August 2011 at 6:12 pm
A quick rebuff to this article:
Ernst's assertions that 1. "Efficacy is nil" and 2."it works by placebo effect" are not mutually exclusive. Here you are confusing efficacy with effectiveness. Efficacy is the specific effect that the treatment has - attributable to the treatment alone. A placebo effect may make a treatment effective for certain conditions but does not make it any more efficacious than a similar placebo (e.g. sugar pills).
When Ernst refers to positive findings this is a reference to a positive effect beyond that produced by a similar placebo. Therefore saying "Even if you do think that the placebo effect entirely accounts for homeopathy's positive results, that's a bit different to saying that there are no positive results" is wrong.
Homoeopathy's usefulness in animals is attributable to the fact that the placebo effect is recognised as having an effect on animals. This is possibly due to the attention and expectations of the carer of the animal.
Ernst's change in views is nothing if not commendable as he displays one of the most noble attributes of the true scientist, the ability to admit he is wrong when presented with evidence contrary to his beliefs.
Above there is a gross misunderstanding of RCTs. The reason why such huge samples are desirable in RCTs is to avoid mathematical inconveniences such as regression to the mean. The randomisation of an RCT is done to remove all patients centred variables for example the relationship the patient has with their therapist. For more on this see http://www.amazon.co.uk/Testing-Treatments-Better-Research-Healthcare/dp/1905177356/ref=sr_1_1?ie=UTF8&qid=1314461218&sr=8-1
I cannot comment on Ernst rejecting 99.8% of the research on natural healthcare, but I postulate that he did not. He is likely to have rejected the following types of research which are often misleading
1. Anecdote (not to be confused as the singular for data (datum))
2. Poorly controlled trials/retrospective studies with small sample sizes, no blinding and no randomisation
3. Blog posts featured on www.naturalnews.com
02 September 2011 at 6:10 pm
Dear David, thanks for your comment. Please allow us to quickly rebuff your rebuff, however! Our reply applies equally to Anonymous’ comment, incidentally.
As to your first point, if you read the Guardian interview to which we referred in our piece, Ernst at no point indicates that he is referring to efficacy or effectiveness – he simply states that results of his clinical trials of individualised homeopathy in asthmatic children were “absolutely nil”. Without any qualification from Ernst, this is certainly different to saying that homeopathy has a powerful placebo effect, which is his current stance as outlaid in the New Scientist piece.
The whole issue of the placebo effect in animals is highly controversial, and to state that the matter is settled as it relates to homeopathy, or anything else, is simply incorrect. You might be interested to hear that the European Union is in agreement with our assessment, since it has just approved a 2 million Euro project to investigate the usefulness of homeopathy in animals (http://www.bva.co.uk/2417.aspx). Even the link we provided in our article (http://www.fwi.co.uk/Articles/2010/08/16/122560/2010-FW-Awards-Livestock-Adviser-of-the-Year-finalist-Christine.htm) shows that homeopathic treatment of farm animals is taken seriously enough that its exponents become finalists in national competitions!
We believe that the evidence we have presented here and elsewhere comprehensively discounts your assertion of Ernst’s scientific saintliness.
Finally, please thoroughly read both this article and the longer feature on Ernst that we linked to before accusing us of a “gross misunderstanding of RCTs”. It’s true that large sample sizes go some way to correcting certain variables encountered in the context of RCTs, but our point was that RCTs are in themselves an inherently artificial, clinical, cold and inhuman environment. We say that not to disparage RCTs in medicine, but merely to point out that that is partly their point: to control for ‘human’ variables and shine the supposedly clinical light of science onto the otherwise messy business of being ill and getting better.
In our longer piece, we point out that Ernst and similar researchers look only for the ‘specific therapeutic effect’ of the treatment under consideration, whereas a true appraisal of the healing dynamic would concern itself with the total effect. The artificial, often skeptical, surroundings of an RCT can actually increase the variability of the other x-axis components and hide the true effect in experimental noise.
Incidentally, you might wish to look back through Ernst’s own trials and see if you can find any with a sample size large enough, in your own estimation, to account for the patient-centred variables. That’s his own clinical trials, not meta-analyses or systematic reviews.
Your final comment betrays your prejudices, we’re afraid: it’s undeniable that he rejected those trials as we discussed, as you can see from the paper we linked to. All of the papers Ernst dismissed were genuine scientific studies and not anecdotes or ‘blog posts’, being taken from the electronic databases Medline, Embase, Cochrane Library, CINAHL and AMED. Although many were observational studies, even the originators of evidence-based medicine stressed that such evidence must be considered – not rejected out of hand (http://www.bmj.com/content/312/7023/71.long). To quote directly from that editorial: “Evidence based medicine is not restricted to randomised trials and meta-analyses. It involves tracking down the best external evidence with which to answer our clinical questions.”
It seems like your response appears to be somewhat of a kneejerk reaction that is the result of not having fully evaluated all available information. Perhaps a more considered response might have been in order had you taken time to read more widely around the area?
29 August 2011 at 1:08 pm
This article shows remarkable ignorance about the Scientific method and complex phisiological reactions such as the placebo and nocebo effects, for example the article states that placebo does not explain animal CAM results,but the function of the placebo effect in animals is well documented.
Why is it that CAM is exempt from the same rigours as other medicine? Why does it get a free pass?
23 September 2011 at 5:13 pm
I think saying that Ernst was a researcher is a questionable title, he was largely involved in the production of meta-analysis of existing research, so perhaps is best labelled an analyser.
The problem with his analysis, in the case of nutrients, was he was looking at the wrong data, namely RCTs. Nutrients are not drugs, they have several important differences and cannot be studied as such.
For example, in the case of his groups recent meta-analysis on calcium and weight loss (Nutr Rev. 2011;69:335–343.)they found a small but significant effect but instead said it was of "uncertain clinical relevance." I am not sure what they are uncertain about. Maybe they were expecting a drug like effect? At a population level the effect they found for calcium would translate to something very important.
Lets hope in his retirement Ernst has time to catch up on reading in the filed he claims to be an expert.
24 September 2011 at 11:57 am
Self-promotion?
Because homeopaths are all shrinking violets who don't post articles all over the web, invade blogs and online newspaper articles to promote their nonsense and generally try to infiltrate real medicine by any means possiblke?
Your voice counts
We welcome your comments and are very interested in your point of view, but we ask that you keep them relevant to the article, that they be civil and without commercial links. All comments are moderated prior to being published. We reserve the right to edit or not publish comments that we consider abusive or offensive.
There is extra content here from a third party provider. You will be unable to see this content unless you agree to allow Content Cookies. Cookie Preferences