Characteristic and incidental (placebo) effects in complex interventions such as acupuncture

Paterson and Dieppe1 summarise the results of clinical trials of
acupuncture: acupuncture to genuine points on meridians has large effect
sizes compared with no treatment, but sham acupuncture to ‘inappropriate’
points off meridians is almost as good. The authors call this a ‘paradox’
and argue that it has arisen because the RCTs have not tested the
‘holistic’ effects of acupuncture. Their own qualitative research has
demonstrated an effect which could be described as psychotherapeutic.2
This effect seems to arise from two sources: firstly, from a sound
therapeutic relationship that is an essential feature of good medicine of
any sort, and of course warrants further research; and secondly, from the
characteristic (or specific) features of traditional Chinese medicine
particularly what they call ‘the emergent and contingent diagnosis’. This
needs more careful scrutiny.

Traditional Chinese medicine includes a number of concepts that are
contrary to the current prevailing understanding of anatomy and
physiology. One is the whole idea of ‘meridians’. Another is the very
basis of diagnosis and treatment: that a particular set of points can be
chosen to treat ‘Liver fire rising’, and a different set of points chosen
to treat ‘Dampness in the Spleen’. This belief system requires the
suspension of disbelief by many Western practitioners and patients, and
raises questions about the use of belief in medicine, credulity, and
patient autonomy. Traditional Chinese acupuncturists themselves show
‘reticence in challenging the status quo’ of their received health
beliefs. Indeed, the whole thrust of the RCT evidence already referred to
– that treatment effects do not depend on needling the precise appropriate
points – calls these beliefs into question.

There is an alternative theoretical model that provides a solution to
the ‘paradox’ of RCT results in acupuncture: that the needles stimulate
nerves, not meridians and points. Therefore, for many conditions, the
needle can be inserted almost anywhere within the relevant spinal segment:
it is the skin or muscle penetration that is sufficient. So needling
‘inappropriate’ points, the usual procedure adopted in the past as a
‘control’ in RCTs, is a physiologically active treatment. There is no such
thing as a nearby point that is ‘inappropriate’ and there is no paradox
after all: the results of previous RCTs are exactly what you would expect.
Recently, a credible sham procedure was developed that does not involve
skin penetration, and some early results have shown that it is less
effective than needling.3;4 The apparent paradox considered by Paterson
and Dieppe may not be a paradox and alternative models need considered.
There may be, after all, an important future for a biomedical approach
to the evaluation of acupuncture, investigating the effect of the very
thing that makes the therapy instantly recognisable – needle insertion.

Adrian White

John Campbell

Reference List

1. Paterson C, Dieppe P. Characteristic and incidental (placebo)
effects in complex interventions such as acupuncture. BMJ 2005;330:1202-5.

2. Paterson C, Britten N. Acupuncture as a complex intervention: a
holistic model. J Alt Complement Med 2004;10:791-801.

3. Kleinhenz J, Streitberger K, Windeler J, bacher A, Mavridis G,
Martin E. Randomised clinical trial comparing the effects of acupuncture
and a newly designed placebo needle in rotator cuff tendinitis. Pain
1999;83:235-41.

4. Guerra de Hoyos JA, Martin MC, Leon EB, Lopez MV, Lopez TM,
Morilla FA et al. Randomised trial of long term effect of acupuncture for
shoulder pain. Pain 2004;112:289-98.

Competing interests:
Adrian White is editor of Acupuncture in Medicine, the journal of the British Medical Acupuncture Society