Simon Singh and Edzard Ernst

TRICK OR TREATMENT?

Alternative medicine on trial

This book provides a critical survey of the main forms of alternative medicine. Significantly, the authors prefer to use this term in preference to the currently fashionable CAM (complementary–alternative medicine), because they wish to distinguish sharply between orthodox and unconventional treatments. (The term 'complementary' does not appear in the index.) Their thesis, in a nutshell, is that hardly any of the unconventional treatments work except as a placebo.

The book is aimed at a medically naïve readership, so it begins by explaining the principle of the randomized controlled trial in considerable detail. Although studies of this kind were done as long ago as the eighteenth century, when James Lind demonstrated that fruit could prevent scurvy in sailors, they came into widespread use only after the second world war. Still more recently the term 'evidence based medicine' was introduced to describe the scientific testing of treatments.

Singh and Ernst adopt the fixed position that only treatments that have been evaluated by such criteria should be used but they don't make it clear that the results of clinical trials are often not as unequivocal as one would wish. Not all currently used conventional treatments are as soundly evidenced as they seem to imply.

The book considers four therapies in detail (acupuncture, homeopathy, chiropractic, and herbal medicine) and has an appendix in which the authors' views of a number of others are presented in summary form.

For acupuncture, the conclusion is that probably almost all the improvements reported can be ascribed to placebo, although there is a little evidence for some kinds of pain and for nausea. Not only is there little good evidence for the efficacy of acupuncture, the authors say, but the underlying theory is nonsensical.

A major omission in this discussion is that nothing is said about modern medical ('Western') acupuncture. This version partially or wholly ignores the traditional ideas of yin and yang, 'meridians', and so forth and instead is based on the modern understanding of anatomy and physiology. Practised by health professionals, it includes plausible explanations for how acupuncture effects may be produced.

Singh and Ernst make much of the recent large-scale German trials which showed consistent effects of needles in several different kinds of pain but found no difference between 'real' acupuncture and 'sham' acupuncture done superficially at non-traditional sites. They take this to be proof that acupuncture relies on the placebo effect. However, as the authors of those studies pointed out, one could also conclude from their results that needling does work but it often, though not always, makes little difference how or where it is done. This is actually my own view. Even sham needles that do not penetrate the skin, used in some other trials, may have a 'real' physiological effect, brought about by a particular light-touch system only discovered in the last ten years. This makes it very difficult to do plausible control treatments in acupuncture.

The section on homeopathy provides a fair-minded account of the origins and present state of the treatment. On the historical side it includes some interesting material I hadn't come across previously. I didn't know that the Nazis were interested in homeopathy and conducted some high-quality research, which was never published owing to the outbreak of war in 1939. There is also an account of how homeopathy reached India and became so succesful there, something I have always wondered about.

Ernst has himself practised homeopathy in the past and has received it as a patient, but the conclusion reached here is entirely dismissive. Research purporting to show how highly dilute preparations could have an effect in the body are flawed and the clinical evidence does not support the claims made for homeopathy as a treatment. In other words, it is simply a powerful placebo. I would broadly agree with this although I think it would be more accurate to say that the homeopathic consultation often acts as a form of psychotherapy.

Chiropractic is discussed more briefly. The authors find it to be based on an implausible theory and to be ineffective; moreover they warn warn prospective patients that chiropractic manipulation of the neck is potentially dangerous. They are slightly more positive about osteopathy as a treatment for back pain, though they think that cheaper alternatives are equally effective.

In the case of herbal medicine, also discussed fairly briefly, they concede that some of the treatments, such as St John's Wort for depression and Devil's Claw for musculoskeletal pain, do work, although probably no better than conventional treatment. The safety of some widely used herbal medicines is questionable and many are ineffective. A useful table summarises the evidence for a number of popular herbal medicines, and another table lists the common adverse effects. This is important because many patients assume that because a substance is natural it is therefore safe, which is far from the case. There are also important interactions between herbal products and conventional medicines.

Even if many alternative treatments are placebos, does it matter? The placebo effect can be surprisingly powerful so why not take advantage of it? This question comes up several times in connection with various treatments and it is discussed at length in a chapter towards the end of the book. The authors' conclusion is uncompromisingly hostile to the deliberate use of placebos. Mainly this is because it entails dishonesty, and they think that the systematic use of such trickery would have various undesirable consequences, such as medicalizing minor illnesses that would be better left untreated, endorsing bogus remedies, and encouraging patients to seek help from practitioners of alternative treatments.

Singh and Ernst score a lot of hits in this book. I agree with much of what they say; in fact, I have made the same general criticisms of alternative medicine myself in the past. But the boundaries between orthodox and unconventional treatments are less precise than they suggest here. Hypnotism, which they mention briefly, has already partly crossed the barrier, and modern acupuncture is also making the same transition. And there seems to be an unconscious tendency towards dualism in their treatment of the placebo—they almost imply that the mind and the body are separate entities. We need to remember that the placebo effect must, in the end, be produced by the brain. I think there is likely to be a 'final common path' by which many of these therapies produce their effects.

7 June 2008