The American Journal of Bioethics

Louis C. Charland's (2002) paper arguing that heroin prescription to heroin addicts violates ethical standards lacks basic information that should be weighed in determining whether true consent is possible and, indeed, what are the risks and benefits to the prospective participant. First, the definition of heroin prescription cited in the article is incomplete. Heroin is also prescribed for severe pain in the United Kingdom and other countries, and many physicians feel that it is a superior analgesic to medications available in North America. Thus heroin is not an experimental drug with unknown side effects. Second, the standard treatment for heroin addiction is methadone. While there are numerous clinical trials showing that methadone is effective in saving lives, there are so far no randomized clinical trials showing efficacy for heroin maintenance. It is not correct to say that it may be a "violation of justice" for North American addicts to not have heroin maintenance available because heroin has not been proven effective. There are numerous patently ineffective medications prescribed in other countries that are not permitted in countries with a high standard of science. The Dutch heroin trial does meet scientific standards, but it is not yet complete and published. Little can be concluded from the Swiss experience, because it was uncontrolled. A relevant fact is that heroin could be legally studied in the United States if an appropriate experiment were proposed and approved. At one time the National Institute on Drug Abuse issued a request for research proposals on heroin maintenance, but none of the applications met scientific standards and thus none received a high enough priority score to be funded.

The Swiss were concerned about the number of HIV infections and deaths from heroin overdose among addicts. Since many of the addicts refused to enter standard treatment, a group of clinicians were motivated by humanitarian concerns to induce more addicts into treatment. They offered them the same drug that they were currently using on the street without the attendant risks of impurities, infections, and overdoses. There is no evidence that the subjects taken into the Swiss heroin study were put at any increased risk as a result of the program. They were safer, not exposed to infections and overdose, given medical care and counseling, and given help with an abstinence program if they desired. On the other hand, the Swiss study produced no evidence that heroin maintenance is effective in leading to rehabilitation. This requires a randomized control group. They did demonstrate feasibility. From a pharmacological perspective, methadone is superior to heroin because one dose per day is compatible with normal function. Methadone patients can function in demanding jobs such as teaching school, practicing law or medicine, or driving long-distance trucks. Heroin has a short duration of action, requires injection rather than oral dosing, and thus interferes with normal activities.

While there was no designated control group, a number of applicants for the Geneva section of the Swiss study were put on a waiting list when heroin slots were not available. When they were contacted six months later, most had started on methadone and were no longer interested in the heroin program. It may turn out that the greatest benefit of heroin availability is attracting into the treatment system those heroin users who would not consider methadone. Some of them could later be moved to methadone as a transition and perhaps later to abstinence and naltrexone to block relapse.

Heroin prescription is not at all analogous to the alcohol example used by Charland. There the potential research subjects are abstinent, and the alcohol used experimentally could theoretically increase the risk of relapse. None of the heroin users offered heroin had achieved abstinence, and thus risk of relapse was not an issue.

Note that I have not addressed the issue of whether or not drug craving can influence decisional capacity. Clearly it can, depending on the state of the patient, the time since the last dose, and the presence of other drugs or illnesses affecting mental status. This question, however, is not relevant to the studies of heroin maintenance reviewed in the article where the "choice" is between street heroin of unknown quality with high known risk and medically supervised heroin in a comprehensive treatment program.

In summary, the conclusion that heroin maintenance studies violate ethical standards is simply incorrect. It would be completely different if the research involved offering heroin to former heroin addicts who might then be harmed. In the context of the studies conducted in Europe, I can see nothing but reduction of harm to vulnerable individuals already at great risk and in many cases already infected. My main concern about the studies is that they will divert resources from more definitive and proven treatments that are pharmacologically superior to heroin such as methadone, LAAM, buprenorphine, and naltrexone in combination with appropriate counseling and rehabilitation programs. Since many heroin users refuse these standard treatments, attracting them to the treatment system through heroin availability may save their lives.