In the Netherlands, there have been three small-scale experiments - all in the city of Amsterdam - involving the prescription of opioids other than oral methadone in the substitution treatment of heroin users. In 1983, the Minister of Health, Welfare and Sports approved a small-scale morphine dispensing program for heroin addicts in Amsterdam. At the same time, the city council of Amsterdam announced its plans for the initiation of a large-scale heroin dispensing program for approximately 300 drug users. This latter plan was not approved by the Dutch government, however, and the program was never implemented (Brussel, 1997a; Derks, 1997).
The morphine dispensing program started at the end of 1983 with a group of 37 very problematic heroin addicts, and was to run for an experimental period of two years. Most of these patients received a combination of intravenous morphine along with a basic dose of oral methadone. The results of the study suggested a beneficial effect (reduced levels of illegal heroine use and criminality) for approximately half of the participants. However, these results could not simply be attributed to the co-prescription of intravenous morphine, because the study lacked a control group. In addition, 11 participants abandoned the treatment program within the first year because they were dissatisfied with morphine and/or were suffering from histamine reactions following its administration (Derks, 1984, 1990). By 1993, ten years after the start of the project - which was supposed to run for only two years - seven of the original 37 patients were still receiving intravenous morphine, 14 patients had returned to oral methadone maintenance treatment, one was known to be drug-free and 15 had died (five committed suicide, six died from AIDS, two had developed lung cancer, and two died of a drug overdose). These data clearly indicate a high mortality rate, although most of these deaths were not considered to be related to the program itself, but instead reflected the severity of the patients' level of pathology at treatment entrance. Nevertheless, the experimental program demonstrated that prescription of injectable opioids is feasible, that very little morphine leaked to the black market and that some of the patients did improve. When the experiment ended, no new patients were recruited because when given a choice, most patients preferred injectable methadone.
In 1990, the Amsterdam Municipal Health Service started a second quasi-experiment, involving the prescription of intravenous methadone to 30 severely addicted AIDS patients, all suffering from a very poor health condition. The treatment had to be terminated for nearly half of the patients, however, because they failed to comply with the treatment regimen. From the remaining group, many patients subsequently died. Among the patients who continued treatment, clear improvements were reported in the therapeutic relationship with the treatment staff. In addition, the heroin use of these patients decreased considerably, although it did not stop (Van Brussel et al., 1996).
In 1995, the Amsterdam Municipal Health Service started a third experimental program, involving the prescription of oral dextromoramide (PalfiumR) in addition to oral methadone to 53 severely addicted non-injecting heroin users with an average addiction career of 21 years, who had not responded well to earlier methadone treatment. The goal of this treatment was to alleviate the patients' suffering and to stabilize their heroin use. Unfortunately, the experimental program was not systematically evaluated. The general impression of the project staff was that the treatment with Palfium resulted in improved social functioning of the patients, and in a better relationship with the treatment services. A clinical evaluation suggested that heroin use had decreased among most patients, and that one patient had actually stopped his heroin use. In addition, urinalysis tests revealed no increase in the use of cocaine. Lastly, the termination of the Palfium prescription to those patients who had not benefited from the treatment did not cause problems (Van Brussel, 1997b).

To summarize, while reflecting an active debate regarding the medical prescription of opioids other than oral methadone to chronic, treatment-resistant heroin addicts, the three experimental treatment programs described above all involved very small groups of heroin addicts with rather specific needs, lacked a control condition, and the results were not adequately documented. Given these limitations regarding the effectiveness of the treatments, the generalizability of their results, and the absence of data in the Netherlands regarding the medical prescription of heroin, the Health Council of the Netherlands (1995) had to turn to scientific evaluations of experiments with heroin in other countries, including the United Kingdom (Hartnoll et al., 1980) and Switzerland (Uchtenhagen, 1996a), when discussing the desirability of medical prescription of heroin to Dutch heroin addicts. These evaluations are summarized in the next paragraph.