A properly designed clinical trial? I have my doubts.
Your reviewer of this series, John Garrow, says that the presenter
shows "it is possible to apply properly designed clinical trials to
alternative therapies, for which she deserves congratulation." Like your
correspondent Les Rose, I was intrigued to witness "live" the planning of
what was meant to be a ground-breaking research study into the effects of
acupuncture using fMRI scanning. All I can say is thank heavens this is
not how research studies are devised and conducted in practice.
Fortunately most meaningful studies have a proper protocol with clear aims
and objectives, with input from those who know about the design of such
studies and who have statistical expertise, and the study findings and
conclusions are not presented to the lay public via the medium of national
television by one of the designers enthusing about its amazing results.
The study premise was that placing needles into the skin using a
particular technique at an acupuncture meridian point might reveal changes
detectable by fMRI. We are told the results of the study suggest that the
deep needling elicited a different response to that of shallow needling,
with deactivation rather than stimulation of part of the brain involved in
the perception of pain, a result prompted the programme's presenter to
declare this was proof that acupuncture worked. I am quite prepared to
believe that sticking needles into people elicit neurophysiological
responses, but to adequately investigate whether acupuncture is a valid
phenomenon using this type of study there needed to be a far clearer
protocol and adequate study groups and controls. It was the acupuncturist
on the team who decided that needling should be confined to a single
acupuncture point on the wrist, and that that there should only be 2
groups of patients – "controls" who should receive shallow needling, and a
study group who had deep needling with rotation to elicit a de-chi
sensation. A wonderful opportunity to look at ways in which the brain
deals with different sensory stimuli seems to have been lost – there was
no control stimulus other than needling (how do we know that, say, a deep
skin pinch as opposed to a superficial one would not provoke the same
reaction?), there was no “sham” acupuncture needling (despite the
demonstration earlier in the programme of this technique), and no needling
of points placed away from the traditional acupuncture meridian points
(this would have clearly shown whether the site of acupuncture is
relevant, as acupuncturists would have us believe). Were the acupuncture
needles entirely non-ferromagnetic? Even if they were, how can we be sure
that they do not locally augment the strong 3T magnetic field generated by
the scanner, and that this induces a physiological effect dependent upon
the depth of needle insertion?
Perhaps we can now look forward to more traditional scientific
research methods coming into play – this study first needs to be published
following peer review (and which journal will do this now that the
"results" have been plastered all over the media?), and second it needs to
be redesigned and repeated with proper controls. There are already studies
demonstrating that fMRI can detect deactivation within pain processing
areas of the brain during painful stimulation with attention-diverting
counter-stimulation (1). We need to be sure that such studies are designed
to test exactly what they purports to test, and nothing else.
(1) Longe SE et al.
Counter-stimulatory effects on pain perception and processing are
significantly altered by attention: an fMRI study.
Neuroreport. 2001 Jul 3;12(9):2021-5.
Competing interests: No competing interests