The medications prescribed in the study were heroin (in injectable hydrochloride and inhalable base form) and oral methadone (hydrochloride). As described in paragraph 2.2.3, methadone was selected as medication in the control condition of the trials, because methadone treatment constitutes the most applied standard reference treatment in the Netherlands. In addition, methadone was used in combination with heroin in the experimental condition of the trials to prevent withdrawal symptoms during periods that the prescribed heroin is not available to the patient. Hence, subjects in the experimental condition received co-prescribed heroin and methadone, and subjects in the control condition methadone alone. Methadone prescription and dispensing took place in existing treatment locations with an existing treatment-staff, whereas the combined prescription of methadone and heroin took place in newly established locations, with specially recruited staff-members.

Methadone
Similar to the procedures followed in most methadone programs in the Netherlands, methadone was prescribed in the study in oral form once a day (standard methadone treatment). The methadone dose prescribed to an individual patient was determined by the treating physician during the qualification period of the study, using clinical titration procedures. In phase IIa and IIb of the study, an effort was made to maintain the subjects on this initial methadone dosage. A minimum daily methadone dose level of 30-50 mg was recommended. The maximum daily dosage of methadone permitted in the study was 150 mg.

Heroin
Heroin was prescribed in intravenously injectable form (heroin hydrochloride) and inhalable form (heroin base). Given the shorter pharmacological action of heroin compared to that of methadone, the heroin was dispensed three times a day, divided over the day. To avoid the risk of the prescribed heroin being sold "on the street", the heroin had to be taken under supervision at the treatment site. After randomization to the experimental treatment condition, the treating physician - in consultation with the patient - established the initial dosage of prescribed heroin, taking into account the patient's physical health condition, his methadone dosage, his concurrent use of illicit substances, as well as the possible use of additional medications that may affect the pharmacokinetics of the prescribed opioids. This initial dose level was subsequently increased, using a clinical titration procedure. At the end of the titration procedure, the resulting dose level had to be sufficiently high to prevent the subject from seeking additional illicit heroin, while taking into consideration possible adverse effects, such as respiratory depression or overdose, in case the patient would not tolerate the established dosage. The maximum daily dosage of co-prescribed heroin permitted in the study was 1000 mg, and the maximum single dosage 400 mg. Data from the methadone registration system and from several studies have indicated that only a small minority of the Dutch heroin users consume more than 1000 mg of "street heroin" a day (Toet, 1990; Barendregt et al., 1995; Blanken et al., 1996b), while the current purity of the Dutch street heroin is estimated to generally range from 30% to 50%. In Switzerland, the median daily dosage after an initial adjustment phase amounted to approximately 500 mg for injectable heroin (Rehm et al., 2001), and to 1000-1850 mg for smokeable heroin (Uchtenhagen et al., 1997). With regard to these high dose levels of inhalable heroin, it should be noted that the bioavailability of heroin in the applied heroin cigarettes was very low (10-15%), whereas considerable individual variations occurred in both application forms (Uchtenhagen et al., 1997). Based on a pilot study (see below), the bioavailability of the inhaled heroin in the present Dutch trial is estimated to range between 35-45% (Hendriks et al., 2001).

In the course of the qualification period, the treating physician determined whether the subject should receive injectable heroin or inhalable heroin, taking into consideration the patient's physical health condition and his history and current pattern of heroin self-administration. The patients who participated in the injectable heroin trial of the study received the prescribed heroin in the usual (injectable) form. The patient administered the prescribed heroin himself, provided that this was done in a safe manner. For safety reasons, and to standardize the method of self-injection as much as possible, the patients were instructed by the treatment staff about the method of self-injection at the start of the study. If necessary, the patient could be assisted with the self-injection by a nurse. As described earlier in paragraph 2.2.2, subjects who were assigned to the injectable heroin trial at the start of the study had the possibility to switch to inhalable heroin under certain conditions, but they remained in the injectable heroin trial, and were analyzed as participants in the injectable heroin trial.
The procedures for the patients who were assigned to the inhalable heroin trial of the study were very similar to those for the patients in the injectable heroin trial. For example, the participants in both trials were admitted to the heroin (injection or inhalation) administration rooms in small groups at the same time, for reasons of efficiency and comparability to the social setting in which heroin users often use their heroin. However, given the increased health risks associated with injecting heroin use, no injectable form of heroin was provided to patients who had been assigned to the inhalable heroin trial at the start of the study.

In light of the predominance of heroin inhalation in the Netherlands, the CCBH took the initiative to conduct a pilot study on the development and testing of an inhalable heroin compound, which would yield a sufficient and reproducible amount of heroin smoke after heating and volatilization, did not produce toxic pyrolysis products, would not have toxic effects on the central nervous system, and would be sufficiently stable and easy to manufacture in large quantities (Bronner, 1997). The resulting heroin compound consisted of a mixture of heroin base and caffeine. Caffeine was added to the heroin base, because caffeine has been demonstrated to lower the temperature of volatilization of heroin base, to slightly enhance the recovery of heroin in smoke, and to reduce its pyrolytic decomposition (Huizer, 1987; Cook and Jeffcoat, 1990). In addiction, the combination of heroin base and caffeine has been used in illicit street heroin for decades (Grund, 1993; Grund and Blanken, 1993), and has never been observed to produce toxicity.
The developed heroin compound was subsequently investigated in another pilot study (Hendriks et al., 2001) in a sample of 10 heroin addicts, with the objectives to (1) determine the acceptance by the heroin users of the heroin compound and different methods of heroin inhalation (i.e. chasing the dragon from aluminum foil versus inhalation from a heating device), (2) determine the bioavailability of heroin after inhalation, (3) investigate the physiological, behavioral and subjective effects during and following heroin inhalation at various dose levels, and (4) determine the reproducibility of the bioavailability and pharmacodynamic effects of the inhaled heroin. The study findings showed, among others, that the participating heroin addicts strongly preferred the inhalation by means of chasing the dragon, and that an average of 38-47% of the inhaled heroin could be recovered as total morphine in the urine samples of the subjects (Hendriks et al., 2001). Based on the results of this study, a mixture of heroin base and caffeine was chosen as the compound for inhalable heroin trial of the national study, and the - for Dutch heroin addicts usual - method of chasing the dragon was chosen as the route of inhalable heroin self-administration.