Yesterday, I went to a talk given by Edzard Ernst at Bradford University. The talk was based on the book he co-wrote with Simon Singh: Trick or Treatment. This is my (unsystematic) review.

Professer Ernst began by telling us that the book has received much criticism from advocates of alternative medicine - and was met with bewilderment from some, who seemed surprised that a professor of complementary medicine would criticise CAM. To that, Ernst said that:

An uncritical scientist is a contradiction in terms.

His introductory slide* included statements to the effect that:

• A large proportion of the general population uses CAM
• Patient population figures for use of CAM are much higher
• In the UK, we spend £1.6 billion on CAM each year
• Even the most fundamental questions remain unanswered

Slide number two highlighted four ways of attempting to find out if something works: plausibility; the test of time; ask the patient; do the research.

Bloodletting was used as an example for each of the first two options. Plausibility: Ernst pointed out that (at the time) bloodletting was seen as a plausible treatment because of the belief that health was affected by the status of the four humours. He then went on to point out that acupuncture (based on the principles of "yin and yang"), chiropractic (based on the idea of "subluxations"), homeopathy (based on the principle that "like cures like"), and reflexology (based on the idea that points on the foot correspond to the internal organs of the body) are utterly implausible. Although he then pointed out that implausibility doesn't necessarily rule out a treatment being effective.

As for the test of time, Ernst noted that bloodletting was used for hundreds, if not thousands of years despite being ineffective. He then quoted Oscar Wilde: "experience is the name we give to our mistakes", which I thought was a rather neat way to illustrate the problem with assuming that the longevity of a treatment was a reliable measure of its effectiveness.

In tackling the option ask the patient, Ernst showed a slide from Marja Verhoef that showed a hierarchy of the factors important to patients. "Gut feelings" were at the top of the pyramid, "trial-and-error" and media articles were also listed but I was unable to note all the factors or the order of importance of the factors. There is an abstract of a qualitative study conducted by Verhoef and others which lists types of evidence relied upon by cancer patients searching for information on complementary therapies.

Scientific evidence ranked very low for most patients. Anecdotal evidence was among the top types of evidence patients relied on.

Then we came to research. Ernst related the proposal made by Jan Baptist van Helmont in 1662. Thanks to the James Lind Library, I am able to reproduce the quote:

'Let us take out of the hospitals...200 or 500 poor people, that have fevers, pleurisies. Let us divide them into halves, let us cast lots, that one halfe of them may fall to my share, and the other to yours; I will cure them without bloodletting and sensible evacuation; but you do, as ye know...We shall see how many funerals both of us shall have.'

This was perhaps the first time that an RCT had been proposed. Unfortunately, it wasn't until 1809 and Hamilton that a trial of bloodletting is believed to have taken place. Ernst then moved on to James Lind himself and the famous trial of treatments for scurvy. The James Lind Library has a commentary of the trial here: link. Ernst informed us that one of the treatments tried was "hard labour" and explained that as one of the symptoms of scurvy was lethargy, confusion over cause and effect had led to the belief that this may be a useful treatment. His comment that hard work was believed to be a treatment for "lazy bastards" brought laughter from the audience. I had expected the talk to be rather dry, but was pleased to note several moments where humour was employed by the professor.

Acupuncture

Ernst referred to a positive trial by Berman et al that studied the use of acupuncture for osteoarthritis, then pointed out that we must beware the cherry-pickers. A single trial cannot be expected to prove the efficacy of a treatment, but systematic reviews are more reliable. While acupuncture has been touted as a panacea (cure-all), Ernst noted that it has only been shown to be effective for pain and nausea.

Chiropractic

This section of the talk began with discussion of the origins of chiropractic, which some may have been surprised to find was not invented as a treatment for back problems. A slide was shown that listed systematic reviews of chiropractic spinal manipulation for various conditions alongside percentages of chiropractors who (wrongly in most cases) believed that chiropractic was an effective treatment.

Ernst then moved on to the risks of chiropractic, noting that mild to moderate adverse effects of the treatment were not disputed - in comparison to severe adverse effects which were hotly disputed. Discussion of chiropractic ended with a slide showing a pair of scales labelled "harm" and "benefit", with the harms being shown as heavier than the benefits.

Homeopathy

Beginning with a slide showing a homeopathic pamphlet featuring Queen Elizabeth II, we were told that the royal family were "staunch supporters" of alternative treatments such as homeopathy. Ernst went on to say that the treatment was based on the principles that "like-cures-like" and "less-is-more" - principles that are totally implausible. Ernst told us of a study he had conducted into homeopathic arnica which made him "very unpopular with British homeopaths" (who were "quite outraged" - but which was later confirmed by two independent systematic reviews.

Noting that some homeopaths complained that trials of homeopathy that did not study individualised treatment of patients, Ernst referred to a trial of individualised homeopathy for childhood asthma which provided no evidence that individualised homeopathy was superior to placebo as an adjunctive treatment. Cautioning us once more against the reliance upon single trials, Ernst then told the audience of a systematic review of systematic reviews. I believe this is the paper in question: abstract on Pubmed. Ending the section on homeopathy with reference to harm and benefit, Ernst discussed the possibility that, while an inert treatment such as homeopathy may not cause harm, homeopaths themselves might - for example by offering homeopathy as an alternative treatment for swine flu. Again, the slide showing a pair of scales with the balance tipped towards "harm" was shown.

Ernst digressed from discussion of the four complementary / alternative therapies at this point to point out that he doesn't see himself as a "Quackbuster" - but as having to choose whether or not to tell the truth about his investigation of complementary and alternative therapies. Telling the truth seems to lead to him being labelled as a Quackbuster. Ernst then pointed to studies into St John's Wort for depression as an example of investigation into CAM that found an effective therapy, before pointing out the risks (particularly that St John's Wort can interact with prescription medication). Leading on from this discussion of an effective CAM treatment, Ernst moved to the last of the four therapies that were the focus of his talk.

Reflexology

Ernst points out that reflexology may be relaxing and thus beneficial for patients. While the theory behind reflexology is implausible, massage therapy can be beneficial. I have struggled to find the systematic review of massage therapy Ernst referred to, but believe that this may be it: abstract on Pubmed. On the whole, the evidence was positive (albeit there were concerns over methodology). From the abstract:

[the trials] suggest that massage can alleviate a wide range of symptoms: pain, nausea, anxiety, depression, anger, stress and fatigue.

Summing-up

We were told that CAM is popular, used by patients, and in fact is driven by patients. Because of this, Ernst believes it is important to address the general public. He pointed out that of the research into CAM:

...you find some gems in this mess but you find a lot of disappointing results.

He then reaffirmed his stance that he is not a Quackbuster, but is a scientific investigator.

Questions

Having invited questions from the audience and promised a copy of his book for the three best efforts, the first questioner was a gentleman who had got his copy of Trick or Treatment signed by Ernst before the lecture and hence did not need a copy as a prize. He asked about the gold-standard of trials, the RCT, and whether there were problems with this approach to discovering the truth about medical therapies. Ernst stated that while there were problems with RCTs, they were the best option we had for investigating treatments. Referring to the issue that trials of this nature tell us about the group rather than individuals, Ernst reminded us that in medical science we must deal in probabilities.

A question about placebo brought a response to the effect that the placebo response can be powerful. The professor then pointed out that a genuine treatment given by a practitioner who was able to empathise with the patient would have both specific and placebo effects. CAM treatments that are no better than placebo cheat us of the specific effects that genuine treatments can offer.

A lady sitting to my left then asked about clinicians' views (with reference to the hierarchy of patients' views Ernst had shown us earlier). Ernst replied that clinicians often rely on experience rather than the best available evidence - and that "that is wrong" (I paraphrase here, as I cannot remember the verbatim quote).

There was also a question about the regulation of CAM (with regard to the Prince of Wales's FIH and the CNHC, better known as OfQuack). Ernst pointed out that regulated nonsense remains nonsense and complained that CAM practitioners have no requirement to use the evidence-based medicine.

More

Unfortunately, there were many elements of the talk (and questions & answers following the talk) that I failed to note down. I hope that my scrawled notes and imperfect memory have conveyed an accurate impression of the views of Professor Ernst but I would not be surprised to find that my notes and memory of the event are both far from perfect. I have also failed to recall or to make note of the humourous moments of the talk which brought laughter from the audience.

My overall impression was that Ernst is keen to engage with the general public and to inform them about evidence-based medicine and also the pitfalls that they should avoid. For example, being misled by cherry-pickers giving single studies as proof of a treatment's efficacy, or reliance upon inadequate evidence of other kinds - such as anecdotal evidence.

*Edit 26/10/2009: I now have a 41-page PDF of the slides used by Professor Ernst during his talk. I have uploaded the file to my Stuff And Nonsense blog here: link.