The Swiss heroin trials. Scientifically sound?

The objective of this article is to critique a study conducted by the Swiss Federal Office of Public Health to evaluate Switzerland's heroin maintenance project. Heroin abusers (N = 1,146) were enrolled in 18 research clinics. Subjects were recruited into three study arms--heroin, morphine, or methadone maintenance, but randomization was unsuccessful, and all received heroin. Medications were self-administered by injection on site. Patients were interviewed at intake and 6-month intervals up to 18 months. A review of the study revealed design weaknesses, including the absence of control groups, lack of corroboration of self-reports, failure to control for the influence of social services on outcome, and the absence of follow-up on those who left the trial prematurely. The program's ability to avert human immunodeficiency virus (HIV) transmission could not be fully evaluated because patients did not consistently submit to HIV testing. The Swiss trials of supervised heroin prescription trials do not withstand scientific scrutiny.