The original Cochrane review on this topic was published in 2003; this update (published June 2011) integrates results from European and Canadian clinical trials.1 The objective was ‘to compare heroin maintenance to methadone or other substitution treatments for opioid dependence regarding: efficacy and acceptability, retaining patients in treatment, reducing the use of illicit substances, and improving health and social functioning’.

In all, 3,346 article titles and abstracts were screened, of which 22 were eligible for the review but only eight met the inclusion criteria.2 The methodology used followed the accepted Cochrane Review guidelines on search strategies, selection of studies, data extraction and management.3 

While most of the studies compared heroin (plus flexible doses of methadone) against oral methadone only, there were some variations.2  One study compared injectable heroin to a control group on a drug treatment waiting list, and one study compared supervised inhaled heroin and another injectable methadone to oral methadone. Depending on the outcome of interest, the relevant studies were included in the various pooled analyses carried out to investigate the effect of heroin maintenance treatment.

1.                   Retention in treatment  

Analysis of the relevant studies (1,388 patients) found that clients on supervised injectable heroin plus flexible dose of methadone were retained in treatment longer compared to patients in oral methadone maintenance (RR  1.44 [95% CI 1.19, 1.75]).4 Compared to any other drug treatment, heroin (supervised or not) (1,535 patients) improved retention in treatment (RR 1.44 [95% CI 1.16, 1.79]).

 

2.                   Relapse to illicit drug use

This was measured as self-reported illicit drug use (including illicit heroin). The authors decided that meta-analysis was not suitable to carry out on the pooled data for this outcome and therefore reported the results of each individual study. They found a statistically significant reduced use of illicit drug use in all the relevant studies in the groups on heroin maintenance compared to the groups on methadone maintenance. The different types of other illicit substances used were not identified in almost all of the studies. 

3.                   Mortality

In four relevant studies (1,477 patients) six deaths were recorded in the group on supervised injectable heroin maintenance, compared to 10 in the group on oral methadone maintenance. There was no statistically significant difference in mortality found between the injecting heroin group and the groups on oral methadone (RR 0.65 [95% CI 0.25, 1.69]) or on any other drug treatment (RR 0.78 [95% CI 0.32, 1.89]).

4.                   Adverse medical events related to the study medication

For this outcome measurement (type of adverse event not specified), only data from four studies could be pooled for meta-analysis, with the results from the other studies presented separately. The results from that analysis were that those in the supervised injectable group had a higher risk of adverse medical events than those in the oral methadone maintenance group (RR 14.42 [95% CI 2.74, 75.97]).

 

5.                   Secondary outcomes

Although the results could not be combined, the results of the appropriate studies indicated a reduction in criminal offences. The only other secondary outcome where prescribed heroin appeared to have a protect effect was for imprisonment, compared to other treatments. 

 

In relation to overall completeness and applicability of the evidence, the authors concluded that the studies included were able to answer the main review questions. The risk of bias was reduced because of the acceptable sample sizes of and the consistency of results across the included studies. One area that was lacking was a clearer description of the characteristics of clients who would most benefit from this intervention.

The authors concluded that there were statistically significant results which showed supervised injectable heroin plus flexible doses of methadone improved retention in treatment, and led to a reduction in rates of illicit drug use, criminal offending and imprisonment. This intervention would be of benefit to clients who have failed to remain in methadone treatment.

Some limitations to this mode of treatment were identified by the authors. The review showed the increased risk of adverse medical events, indicating that prescribed heroin should only be given in treatment centres that were equipped to deal with emergencies. Clients’ attempts to re-integrate into society could be restricted by having to attend a treatment centre two or three times a day for injections. There are many factors related to non-compliance with drug treatment, including poverty, lack of family support and psychiatric co-morbidity. The authors considered that in the current climate of limited and diminishing resources, the appropriateness of providing a more expensive medical treatment rather than attempting to address the known social predicators of non-compliance and relapse. Other studies are anticipated on this topic. 

·         Prescribed heroin provides added value to methadone treatment;

·         Risk-benefit of prescription heroin must be evaluated fully before implementation in clinical practice due to the higher rate of serious adverse events;

·         Prescribed heroin should be considered for clients who are have failed to remain in methadone substitution treatment programmes;

·         It should only be provided in treatment centres if proper follow-up is available;

·         The capacity of treatment services and the cost of the programme must be assessed before provided prescribed heroin.

1. Ferri M, Davoli M and Perucci CA (2011) Heroin maintenance for chronic heroin-dependent individuals. Cochrane Database of Systematic Reviews 2010, Issue 8. Art. No.: CD003410. http://onlinelibrary.wiley.com

2. The studies included were: RIOTT (2010), NAOMI (2009), Haasen (2007), PEPSA (2006), CCBH (A) (2002), CCBH (B) (2002), Perneger (1998) and Hartnoll (1980). Once pooled, a total of 2007 participants were available for analysis. Further details and references of each study included can be found on pp. 25–41 of the review.

3. Higgins JPT and Green S (eds) (2008) Cochrane handbook for systematic reviews of interventions. Version 5.1.0 [updated September 2009]. The Cochrane Collaboration. www.cochrane-handbook.org

4. Risk Ratio (RR) and 95% Confidence Interval (95% CI)