EDITOR—Moore et al and Kovacs and Gil del Real criticise the BMA's report on acupuncture. 1 2 Our review of the evidence base of acupuncture rested heavily on the comprehensive work of Ernst and White, which summarised the clinical evidence for and against the effectiveness of acupuncture.3

The conclusion of this work is that acupuncture seems to be more effective than sham acupuncture or other control interventions for some conditions, including nausea and vomiting, back pain, dental pain, and migraine. However, for smoking cessation, weight loss, and a range of other conditions the present evidence is unclear. We discussed the problems introduced in basing conclusions on poor quality studies or reports.

Our survey of general practice throughout the United Kingdom showed that acupuncture is the complementary therapy most used by general practitioners, with most patients being referred for pain relief and musculoskeletal disorders. Acupuncture is now reported to be used routinely ahead of physiotherapy and drug delivery systems in 86% of chronic pain services.4

The thrust of our recommendations seems to have been missed. The BMA calls for substantial research funding, the production of guidelines, and a formal appraisal of acupuncture. Kovacs and Gil del Real should note that our recommendation about availability of acupuncture in the NHS was subject firstly to having policies, guidelines, and mechanisms for making this treatment generally available—hence the need for appraisal by the National Institute for Clinical Excellence (NICE). Improvements in training and regulation of non-medical practitioners are required, and doctors need to know the basics of complementary and alternative medicine so that they are better able to advise patients. Our detailed review of safety and adverse reactions to acupuncture should reassure Moore et al that the treatment is comparatively safe—the more important risk is likely to arise through misdiagnosis and the withholding of orthodox treatment.

There are more than 5500 acupuncturists in the United Kingdom, of whom over 3500 are statutory health professionals, an increase of 51% in two years.5 Acupuncture treatment has flourished despite a lack of widespread knowledge of its efficacy, and without comprehensive guidelines for either general practitioners or patients. Recommendations clarifying whether acupuncture should be used in the NHS are urgently needed.

EDITOR—The letter by Moore et al denouncing the BMA report on acupuncture is couched in strong language, but their account of the report is selective and misleading.1 They ignores its recommendation that acupuncture is effective for nausea and vomiting (particularly postoperative symptoms in adults), for which there is a sound body of evidence.2

Moore et al misrepresent the BMA's position on smoking cessation; in fact, the report states clearly that “at present there is no evidence to support any role for acupuncture in the management of smoking cessation.”

Moore et al state: “There is evidence that it [acupuncture] harms” without reference; in fact, current evidence shows that the incidence of adverse reactions to acupuncture is low.3

The evidence remains equivocal on the use of acupuncture for chronic pain. The most recent systematic review found that acupuncture is better than no treatment (waiting list controls) but that it is premature to draw conclusions about the effectiveness of acupuncture compared with placebo or standard care.4

Performing double blind placebo controlled trials of acupuncture is exceptionally difficult. Pending such gold standard evidence, the BMA accepted the task of dispassionately evaluating the available literature to define an appropriate role for acupuncture in the NHS. Its report is not “quite simply wrong,” and such dogmatism does not serve our patients or enhance the quality of debate on this important subject.

EDITOR—Neither the BMA in its report on acupuncture nor the comments of Moore et al are entirely right or wrong.1 Lack of evidence for efficacy does not equal evidence for lack of efficacy. Obtaining evidence of efficacy for acupuncture has been hampered by methodological problems unique to this kind of manual therapy, particularly that of finding a credible, truly inactive, control procedure. There are now credible sham acupuncture procedures in which skin penetration in the control group is avoided, and the first trial to use such a procedure indicates a specific effect for acupuncture.2

Systematic reviews of acupuncture for back pain include trials that use different techniques and control procedures. These would usually be considered far too heterogeneous to be included in a review. The highest quality trials compare needling of classic acupuncture points with control procedures that entail exactly the same type of needling at other points. The intragroup effects in these trials nearly always indicate a noticeable improvement after needling, but, inevitably, the difference between what is described as real acupuncture and what is described as placebo is rarely significant. As discussed by Moore and McQuay, the controls used in blinded studies of acupuncture for chronic back pain were 50% effective.3 These controls entailed skin penetration, so one form of acupuncture was compared with another. A 50% response rate is typical of effective treatments for acute and chronic pain.4

In their drive for academic rigour, reviewers are distracted from taking a logical overview of the subject. There is no evidence that acupuncture points exist, so subjecting acupuncture points to rigorous testing is unlikely to be rewarding. There is a wealth of evidence, however, that somatic sensory stimulation can modulate pain.5 Needle penetration of tissues is a potent form of sensory stimulation. It is on this basis that the British Medical Acupuncture Society trains doctors in an evidence based approach to dry needling therapy. Safety issues are important, but for general practitioners or pain specialists acupuncture is still probably one of the safest of the physical or pharmacological interventions they use.

There is a dearth of randomised control trials with positive results, but this may be due more to methodological difficulties than a lack of efficacy. The positive results in lower quality trials may not be attributable solely to bias. The pain community would be done a disservice if acupuncture techniques were not tested in both a logical and methodologically sound manner.