Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Acupuncture for cocaine addiction

A big problem with complementary therapy is that so many of the trials are of such low quality that bias of some degree is likely. When a properly done trial of high quality comes along, it therefore deserves our attention. Such a study demonstrates that acupuncture is ineffective for cocaine addition [1].


This trial compared acupuncture of the ear using criteria of the US National Acupuncture Detoxification Association with needles inserted into non-acupuncture points in the ear and a non-acupuncture relaxation regimen for eight weeks. Six hundred and twenty cocaine dependent adults were randomised. They had to have a recent cocaine-positive urine screen and self-reported cocaine use, but not be dependent on any other drug except nicotine or opiates, or currently receiving other treatment, be actively suicidal or psychotic, or received acupuncture in the previous 30 days.

Patients were told their treatment assignment, because blinding was not possible. Treatments were described in a standardised way, with all the treatment options described as ways to reduce stress and with potential benefits to reduce craving and cocaine use. Weekly (standardised) individual counselling sessions focusing on changing addictive behaviour were offered as adjunctive treatment.

Outcomes were positive urine screens, self-reported cocaine use, and a range of other outcomes.


The three groups were identical in terms of age, sex and ethnic background. Over the eight weeks of the study retention was under 50% and equal in the three groups (Figure 1). The proportion of cocaine-positive urine tests (averaging about three a week) was the same in the three groups (Figure 1). While there was a reduction in cocaine use over the eight weeks, there was no difference between the three treatment groups, either at the end of treatment or at three or six months of follow up (Figure 2).

Figure 1: Percentage of randomised patients who completed eight weeks, and percentage of urines positive for cocaine

Figure 2: Percentage of patients completing the study who were abstinent at the end, and after three and six months


This is a good study that looks at the result both by intention to treat and by completers. It failed to find any difference between acupuncture recommended by the key US association, sham acupuncture, and the non-acupuncture control of relaxation. It dealt with the fact that blinding was impossible, it was randomised, and it was large. The lack of blinding could cause bias, but should have been in the direction of acupuncture doing better if patients or therapists believed that it would. There was no difference.

Acupuncture doesn't work on the basis of these results. What the paper does tell us is how to maximise methodological rigour in a difficult clinical situation, and helps set standards against which we can compare small, low quality studies that more often say that acupuncture does work. It helps explain why small, low quality studies in acupuncture, or anything else, should be ignored.


  1. A Margolin et al. Acupuncture for the treatment of cocaine addiction. JAMA 287: 55-63.
previous or next story in this issue