EDITOR—The BMA report on acupuncture is regrettable. It suggests, among other things, that acupuncture is effective for back pain, dental pain, and migraine. Three recent systematic reviews show the importance of basing judgments on high quality information.

For back pain, four randomised and blind studies showed no benefit; five open studies showed benefit.1 The BMA's conclusion that acupuncture was effective in back pain was based on all nine studies.

For dental pain, a review of 16 studies concluded that it was effective.2 Many of these were not randomised, were not blind, or had major flaws. Only three small studies were adequate, and these showed no convincing benefit.3

For migraine, trials showing a significant benefit from acupuncture were inadequately randomised or not blind.4 The reviewers themselves were highly circumspect about ascribing any clinical significance to acupuncture.

The BMA report concluded that results for acupuncture are inconclusive in other conditions. These are weasel words. For smoking cessation the 12 month cessation rate with acupuncture was 14% (95% confidence interval 11% to 17%), which was no different from the placebo response with nicotine gum of 12% (11% to 13%).

Trials of acupuncture suffer problems of poor quality, which leads to bias. Reviews with poor quality studies overestimate treatment effects. Original reports may come to the wrong conclusion from their own data, a fact true of two of 13 studies of acupuncture in neck and back pain.5

For those areas where the BMA report thought there was evidence of effectiveness of acupuncture, either there was none or what quality evidence there was indicated lack of effectiveness. For those areas where the BMA thought the results inconclusive, there was either no useful information or acupuncture was shown to be ineffective.

Doctors should beware. There is no useful evidence showing that acupuncture helps; there is evidence that it harms. Perhaps the important point is that we should not deceive ourselves, or people who trust our recommendations. There is no gold standard evidence that acupuncture improves pain or anything else. The BMA report is quite simply wrong.

EDITOR—The BMA has concluded that acupuncture should be made more widely available to British people through the NHS and that general practitioners should receive training in it.1 The association seems to base its conclusion on three things: evidence showing that “acupuncture is more effective than control interventions for back pain, nausea and vomiting, migraine and dental pain”; the fact that 47% of general practitioners have arranged for their patients to receive acupuncture; and the wish of 46% of those professionals to receive training in acupuncture in order to treat their patients.2

The evidence on the effectiveness of acupuncture in the treatment of back pain seems to have been misinterpreted. The Cochrane Collaboration Back Review Group has published a major systematic review of the effectiveness of acupuncture in low back pain.3 This review followed a rigorous methodology and an exhaustive search for information. Its results indicated poor research methods and contradictory results from studies of acupuncture in low back pain. The review was therefore inconclusive and could not serve as a basis for recommending acupuncture. This was consistent with the results of past systematic reviews4 and with a randomised trial that compared the effectiveness of acupuncture with that of massage and self care education (D C Cherkin et al, fourth international forum for primary care research on low back pain, Eilat, Israel, 2000).

Although scientific evidence in this respect has not changed much in several years, public and medical opinion does seem to have changed. The establishment of a double standard for the approval of a treatment technique, bowing to the pressure of public opinion and not taking into account evidence based recommendations, is harmful to the public's health and to the economy of the NHS. In time it could also be harmful to the treatment approved with the lower standard and to the credibility of its practitioners and the institutions that recommend it.

Clinical practice is not always based on scientific evidence and the search for an efficient use of resources. Many years ago patients were convinced of the effectiveness of leeches for the treatment of infectious diseases, doctors prescribed them, and apothecaries sold them. Nevertheless, despite public demand and medical interest, evidence of the efficacy, safety, and cost effectiveness of the treatment was lacking. This lesson from the past should be kept in mind.