2.2.1 Randomized controlled trial

The study is a multicenter trial, including a total of 625 patients (1) who are treated in eight treatment units located in six cities in the Netherlands. After an intensive debate, the CCBH decided to develop the study as a traditional randomized controlled trial and not along the lines of a so-called pre-randomization or Zelen design (CCBH, 1997, 1999a).

The study was developed as a randomized controlled trial (see Figure 2). Three consecutive phases are distinguished in the study. During the four to eight weeks qualification period (phase I), the potential participants initially selected from the methadone treatment registration systems were screened on all selection criteria (see paragraph 2.1.2). At the end of the qualification period, and following the baseline assessment, a decision was made as to which trial (injectable or inhalable heroin) the patient would participate in, and all eligible patients were randomized to one of two (trial on injectable heroin: group A or B) or three (trial on inhalable heroin: group A, B or C) treatment groups (see chapter 4).

Figure 2. Design of the trials

Patients randomized to group A received oral methadone during the 12 months experimental phase of the study (phase II), in combination with a standard offer of psychosocial interventions. Patients in group B received a combination of oral methadone and heroin on medical prescription during the 12 months period of phase II, in combination with the same standard offer of psychosocial interventions. In group C, which is only applicable in the trial on inhalable heroin, patients received oral methadone alone during the first six months of the experimental phase (phase IIa), and a combination of oral methadone and co-prescribed heroin during the second six months of the study (phase IIb), both again with the same standard offer of psychosocial interventions. Major outcome assessments were conducted after six (end of phase IIa) and twelve (end of phase IIb) months of treatment, i.e. six and twelve months after randomization. The primary outcome assessment of the study took place at the end of the experimental study period, at the month 12 assessment-point.

Follow-up phase
Following the end of the experimental phase, all subjects entered a naturalistic follow-up period of six months (phase III). Patients who had been assigned to group A were then given the opportunity to receive the experimental treatment with methadone and co-prescribed heroin for a period of six months. The prescription of heroin in groups B and C was terminated at the end of phase IIb and replaced by an offer to return to the methadone maintenance treatment program or to any other form of standard addiction treatment ("most appropriate care"). This was true for patients who had been successfully treated with co-prescribed heroin ("responders"), and for those who had not benefited from the experimental treatment with heroin ("non-responders"). For the group of non-responders in phase II, the termination of the prescription of heroin was final. A similar situation occurred for those responders to the experimental treatment who had not deteriorated two month after the termination of the heroin prescription: no further heroin was prescribed to these patients. Heroin prescription could, however, be reinstated in those responders to the experimental treatment who subsequently demonstrated substantial deterioration (i.e. 20% of the problem severity at the time of the baseline assessment) in their functioning after two months following termination of the experimental treatment. In these cases, the treating physician had the possibility to offer co-prescribed heroin on medical indication in individual cases.
The decision to terminate the treatment with co-prescribed heroin at the end of the experimental phase for all patients was made on both medical-ethical and scientific grounds. If the experimental treatment has not been shown to be effective for a patient, there is no medical justification to continue the experimental treatment. If, on the other hand, the experimental treatment has been shown to be effective for a patient, it can not be excluded that there will be enduring stabilization or improvement after termination of the experimental treatment. In this case, continuation of the experimental treatment in these subjects would be unjustified. Hence, continuation of the experimental treatment only applied to patients who had benefited from the treatment and who showed considerable deterioration after stopping the treatment. To obtain information about these possible scenarios, the treatment must first be terminated, and the consequences of termination must be investigated. In addition, it is important to note, that the effectiveness of treatment with co-prescribed heroin was not (yet) demonstrated on the group level at the end of the experimental treatment period of a patient. In the absence of such evidence, and given that heroin is not yet registered as a medication for the treatment of heroin addiction, it is therefore not justified to continue the treatment with heroin, with the exception of heroin on medical indication and in individual cases as a form of compassionate use.

(1) In the original study protocol, three treatment conditions were distinguished for both injectable and inhalable heroin (CCBH, 1997), but given the low prevalence of intravenous injecting among heroin users in the Netherlands, the third treatment condition (group C) was omitted from the injectable protocol. As a consequence, the intended total sample size of the study decreased from n=750 to n=625 (see chapter 4).