The paper by Peter Blanken and colleagues  reports on 4-year outcomes of 147 patients who completed the 12-month Dutch randomized controlled trial (RCT) of injectable or inhalable heroin-assisted treatment (HAT) , and who then subsequently continued in HAT following the initial trial. Patients were able to remain in HAT if they did not transgress programme rules (e.g. attempted diversion of medications), and continued to have positive outcomes with regard to substance use and related health and psychosocial domains. Patients who did not demonstrate positive outcomes were discontinued from HAT and returned to oral methadone treatment. During the 4 years, 66 of 147 subjects were discontinued from HAT for the following reasons: poor outcomes (30.3%), administrative discharge (42.4%), left to their own volition (22.7%, including 10.6% who achieved abstinence) or incarceration (4.6%). Within this context, the findings of this report are to be expected—that those individuals who responded positively (and remained) in HAT over the 4 years had better outcomes than those individuals who were discharged from HAT, even though the majority (85%) of HAT discharges were in methadone treatment at the 4-year follow-up.
The paper again highlights that long-term HAT can be delivered safely and effectively for patients over extended periods of time, replicating German and Swiss reports [3,4]. The vast majority of patients remaining in HAT were using inhalable heroin, which may explain the low rate of serious adverse events compared to earlier reports [3,5]. As with an earlier report of the Dutch heroin trials , there were no baseline patient characteristics that were predictive of treatment outcome over the 4 years.
The paper raises a number of questions that highlight many of the continuing uncertainties regarding this treatment approach. For example, for how long should HAT be continued? The authors suggest that HAT should be continued ‘for as long as possible . . . unless there is a compelling medical or social contra-indication’. However, the vast majority of positive changes in the patients' lives were achieved within the first 1–2 years of HAT, and that thereafter treatment outcomes remained largely unchanged (with the exception of cocaine use, but the reduction in cocaine use could be due to multiple ‘non-treatment’ factors such as patient maturation or changes in price or availability). This raises the question as to whether there should be a greater emphasis on encouraging stable patients to change back to conventional treatment modalities. It could be viewed that many HAT patients are ‘trapped’ within the rigidity of the treatment model—attending a specific clinic on average twice a day, 28 days a month for their heroin dose. Nevertheless, while similar concerns of being ‘trapped’ may be raised for methadone treatment, stable methadone patients progress to regular take-away doses and increased flexibility over time. The rigidity of current HAT models does not allow for this.
This paper highlights the paradox regarding who is suited to this treatment approach. HAT has been proposed largely as a treatment modality for those individuals who fail to benefit from conventional substitution treatment [7,8]. However, the approach used here is that HAT was continued only for those patients who ‘did well’ and achieved some degree of stability—the patients most likely to be able to progress successfully to less restrictive treatment approaches; while HAT was withdrawn from the ‘difficult-to-treat’ patient, who was then returned to methadone treatment.
The question of what to do with those individuals who do not achieve positive outcomes while in HAT remains. We do not know, for example, if the poor outcomes of those patients discharged from HAT in this study would have been any different had they been allowed to continue in HAT. Further research is required examining different approaches for this group—including the consequences of continuing HAT, of returning to optimized methadone (or buprenorphine) treatment or of having additional interventions (e.g. contingency management) introduced. Unfortunately, it may be that treatment systems may be too stretched to be able to respond adequately to the small minority of individuals who remain in the ‘too difficult to treat’ category, and who do not respond to front-line (e.g. methadone, counselling) or second-line (e.g. HAT) treatment modalities.
Alternatively, this ‘too difficult to treat’ group may be considered as individuals who may not want to ‘respond to treatment’, or participate in rigid treatment modalities. Indeed, the very poor uptake of HAT in almost every setting in which it has been offered (both under trial or as routine treatment conditions, as evidenced by slow recruitment rates in almost all studies) suggests that perhaps current approaches to HAT are inherently unable to meet the expectations of those users who desire access to high quality legal heroin available without medical intermediaries. There continues to be a clash of expectations between treatment providers who frame HAT as a strategy for targeting a clinical disorder and individuals who simply want easier access to their drugs without criminal sanctions, and without necessarily wanting to be ‘treated’.