As a doctor I was brought up to view the randomised double blind placebo controlled trial as the “Gold Standard" for medical research.

It works more or less as follows:-

You have two groups of patients, preferably of adequate size (usually in the order of a couple of hundred or so, although statisticians will have much to say about that) with every one matched in each group, so that, for instance a 40 year male with severe disease in the placebo group is matched by a severely diseased 40 year old male in the active drug group.

Easier said than done.

The numbers must be adequate to show some effect but not so many that the result is spurious when extrapolated down to the individual or small group. One group is given the active substance and the other a dummy substance ("placebo"). The look, taste and packaging are identical so neither may be distinguished from the other. The identity of each is known only to an independent person, so that neither the researcher nor the subjects know whether they are taking the active stuff or not. Each is then treated in exactly the same way according to a protocol, the results collated and any effect determined.

So far, so good.

But the results are then applied to the general population. But doctors' patients are not seen regularly and fussed over in real life like they are in trials. Their pills are not counted every week to ensure compliance. As far back as 1999 an editorial from the British Medical Journal showed little correlation between trial results and real life.

Things have not changed, and remember we are thinking about the "Gold Standard".

It is clearly impossible to have a blinded placebo intervention with many Complementary Treatments. Placebo massage is still an intervention in its own right and both practitioner and person treated will know that massage is taking place. Similarly there cannot be blinded placebo hypnosis or acupuncture.

The medical model simply cannot work in this context.

Some years ago I adopted a before and after protocol for some research I did into hypnosis tapes and various diseases. The patient (I'm a doctor so regard all my clients as "patients") filled in a questionnaire for three months then had the tape and repeated the questionnaire. The only variable there was the tape. I had no actual contact with them so my face to face style and attitudes could not have influenced them. They were clearly exactly matched as they acted as their own controls. In fact there was less artificiality in that than in the gold standard trials

Doctors were sniffy, but I got published in the European Journal of Hypnosis, showed a useful effect and helped a few people (CDs are available from No Panic)

The challenge is to find a practical acceptable research methodology for those treatments which are not amenable to the randomised double blind placebo controlled trials.

I would like to suggest that the before and after trial is an acceptable place to start. Furthermore I don't think that CAM will gain mainstream acceptance until it has some large trials under its belt and the only way to do that is to co-operate and do these.

Mike Matthews editor