Regarding Byford et al’s paper, ^{Reference Byford, Barrett, Metrebian, Groshkova, Cary and Charles1} the authors present an analysis of the results of the Randomised Injectable Opiate Treatment Trial (RIOTT). ^{Reference Strang, Metrebian, Lintzeris, Potts, Carnwath and Mayet2} Participants of RIOTT were very few in number - fewer than 45 individuals in each of the three arms of the study (injectable heroin, injectable methadone and ‘optimised’ oral methadone). It required 3 full years at 3 sites to screen 301 volunteers, of whom 127 (40%) began the trial and only 89 completed the 26-week treatment protocol.

All of the participants had been receiving ‘conventional’ methadone treatment for more than 6 months and continued ‘to inject “street” heroin regularly’. On average, they had had over four prior treatment episodes. Accordingly, it is reasonable to assume that the overriding motivation of those who volunteered was the hope of receiving injectable opiates, and it is likely that participant bias may have had a substantial impact on outcomes. Indeed, it is revealing that among those assigned to receive optimised oral methadone, 7 (17%) never began the trial and of the remaining 35 only 24 were still enrolled 26 weeks later.

Some of the reported findings seem to underscore the severe limitations that must be kept in mind in drawing even the most tentative conclusions. For example, although the oral methadone group claimed to have committed roughly three times as many crimes as the intravenous methadone group (mean 21 v. 7 crimes), the latter group spent 15 times more nights in prison (mean 6.1 v. 0.4). Surely provision of oral methadone did not somehow make patients more successful in their criminal pursuits.

Perhaps inevitably, the limited ability to extrapolate has been ignored in the wider distribution of the findings. Thus, one report (which refers readers seeking more information to the Press Officer of King’s College London, with which the principal author and five of the seven co-authors are affiliated) had the unqualified headline: ‘Injectable opioid treatment for chronic heroin addiction more cost-effective than oral methadone’, and claimed that ‘total cost savings of providing injectable opiate treatment for this chronic group in England could be between £29 and £59 million per year’. ³

The criticisms noted above must not detract from the bottom-line, common sense, conclusion with regard to injectable opioid treatment: in the interests of addicts as well as the general community, it is essential that those who respond poorly to treatment (any treatment) be provided information on and referral to the broadest possible array of alternative services.