To the Editor: Hall et al1 argue that a hydromorphone trial would break the current deadlock in Australia over prescription heroin research. But would it? If hydromorphone was demonstrated to be an effective intervention for treating refractory heroin users, researchers would then wish to compare both hydromorphone and prescription heroin against the gold standard, oral methadone.

The Prime Minister has stated publicly that proceeding with a prescription heroin trial "would send a wrong message". This claim is still unsupported by evidence. Clinicians and researchers should steadfastly oppose political interference in medical research from however august a level, especially when there is such a strong rationale for the trial and when researchers have so scrupulously followed scientific process. Such interference would not be tolerated in other areas of medical research and should not be tolerated in this field.

The results of a recent large randomised controlled trial2 of prescription heroin in the Netherlands, while yet to be published in a peer-reviewed journal, provide strong additional support for an Australian trial. In contrast, Hall et al1 cite no previous evaluation of the efficacy of hydromorphone in managing heroin dependence.

In Switzerland, prescription heroin is reserved strictly for treating refractory patients and accounts for fewer than 5% of all treatment provided.3 The importance of attracting and retaining this group in treatment is probably far greater than their small numbers might suggest, as there is reason to believe that they contribute disproportionately to the immense social costs of heroin use in the community. (By analogy, the heaviest-drinking 10% in a community account for half the total alcohol consumed.) Those who inject heroin very much more frequently than the community mean are probably responsible for a disproportionate share of crime and enlisting new recruits.

Prescription heroin was selected as the experimental intervention in studies in Switzerland, the Netherlands, Germany and Spain and is now being considered seriously in Canada. The reasons advanced by Hall et al1 are all cogent arguments for conducting a trial of hydromorphone additional to an evaluation of prescription heroin.

Hall et al are concerned about lack of community support for a heroin trial in Australia, but in a recent national opinion poll4 45% of respondents expressed support, while 47% were opposed.

The case for an Australian heroin trial, with or without additional trials, remains compelling.