Search   |  Advanced

Research Summary

Heroin to Treat Opioid Dependence: A Randomized Trial

Although methadone maintenance is effective for opioid dependence, many people drop out of treatment or continue to use illicit drugs. In a Canadian study, 251 people with opioid dependence, 2 previous treatment episodes, and long-term opioid use (mostly heroin) were randomized to receive either oral methadone (n=111) or injectable heroin (diacetylmorphine, up to 3 injections daily) (n=115) for 1 year. The remaining 25 subjects were randomized to receive injectable hydromorphone to facilitate validation of self-reported heroin use by urine tests. All subjects received counseling and psychosocial services.

  • Illegal heroin use decreased more in the heroin group than in the methadone group (from 27 days per month at baseline to 5 and 12 days per month, respectively).
  • Retention in treatment was better in the heroin group than in the methadone group (88% versus 54%) as was the proportion of subjects with a 20% or greater reduction in illegal activities or drug use (67% versus 48%). The heroin group also had better social, employment, and psychiatric outcomes.
  • The heroin group experienced 24 adverse events, including 11 overdoses and 7 seizures. The methadone group had none.
  • Hydromorphone recipients could not tell whether they were taking hydrohydromorphone or heroin, and outcomes were similar to those in the heroin group.


Although these results are similar to those from European studies, heroin treatment is unlikely to be widely disseminated soon. As an editorialist1 points out, the limitations in using heroin as a treatment are historical, political, and cultural rather than primarily clinical. Medical concerns exist too. The adverse events attributable to heroin were not insignificant in this study, and the long-term effects on physical and mental health of a frequently injected short-acting opioid may not be as favorable as those achieved with more constant methadone levels. Nonetheless, in this trial, heroin did show greater efficacy than methadone on multiple important outcomes for people with prior treatment failures. From a pure health perspective, should regulations change, heroin clearly could have a role in treating some patients with this devastating chronic illness. Richard Saitz , MD, MPH
1Berridge V. Heroin prescription and history [comment]. N Engl J Med. 2009;361(8):820–821.


Oviedo-Joekes E, Brissette S, Marsh DC, et al. Diacetylmorphine versus methadone for the treatment of opioid addiction. N Engl J Med. 2009;361(8): 777–786.