Prescription of heroin to treatment resistant heroin addicts

Editor—The study of medical prescription of heroin by van den Brink et al should be interpreted with caution.1 Firstly, the selection of opiate dependent patients with at least four weeks of continuous treatment in the past five years does not define resistance to treatment with methadone treatment, but rather, early treatment intervention. This may explain the substantial treatment response to longer term methadone in the control group, where the only intervention is randomisation into a controlled trial. Secondly, in such a study taking methadone and heroin dosage levels into account is crucial when evaluating clinical outcome; adequate doses of opiate replacement are critical to treatment success.2 Actual doses are reported only in the electronic version of the paper and show that doses of combined heroin and methadone in the heroin treatment groups are about 20% higher in terms of methadone equivalents compared with the control methadone only treatment group. Furthermore, this latter group received a mean methadone dose of about 75 mg/day, which may be suboptimal. The discrepancy in methadone dosage equivalents between the two groups may account for the apparently favourable effects of additional heroin. Lastly, the deterioration after discontinuing heroin at the end of the study may simply reflect a transition from adequate to inadequate amounts of prescribed opiate replacement therapy. Heroin treatment needs robust evaluation as it is intensive and expensive compared with other opiate replacement drugs, including methadone and buprenorphine.3 Clinical studies comparing heroin with other opiate replacement therapies need to be conducted on a level playing field.​field. Figure 1