The randomization was organized centrally by an independent monitoring organization, and conducted separately for the trials on injectable heroin and inhalable heroin. In order to maximize comparability of the treatment groups within the participating treatment sites and within gender and ethnic groups, the randomization was conducted for each treatment site separately, utilizing gender and ethnicity as stratification variables in randomization blocks of three (inhalable heroin trial) or two (injectable heroin trial) participants. To achieve the predetermined proportions of the treatment group sizes, the subjects within each block were randomized to the treatment conditions with a predetermined ratio of 135 : 115 : 125 for groups A, B, and C respectively in the inhalable heroin trial, and with a ratio of 135 : 115 for groups A and B respectively in the injectable heroin trial (see paragraph 3.1)
Randomization took place at the end of the qualification period, following confirmation of the eligibility of the subject by means of the baseline assessment. To this end, the local study co-ordinator sent a request for randomization of the subject to the central monitoring organization by telefax, in which he confirmed the subject's eligibility. The monitoring organization subsequently assigned a randomization group code (A, B or C) to the subject, and telefaxed the group code to the local treatment site, and a copy to the local study co-ordinator. This procedure ensured that the randomization was conducted fully independently from both the treatment staff, local research teams, and the National Research Board.