Many marketed antidepressants, dopamine agonists and γ-aminobutyric acid compounds have been trialled and discarded on the long road towards a pharmacotherapy for psychostimulant dependence. Illicit methylamphetamine (MA) can create a particularly debilitating disorder and James Shearer and his colleagues are to be congratulated for their timely trial of modafinil among this population [1]. The headline finding is that modafinil was not associated with adverse events or medication misuse, and that medication compliant participants gave more MA and illicit psychostimulant negative urine tests during treatment.

Modafinil is a novel wake-promoting agent and the front-line treatment for narcolepsy idiopathic hypersomnia. It also improves some aspects of neuropsychological test performance in normal volunteers [2] and, in part, this trial can be seen as a means of ameliorating neurocognitive dysfunction among patients and thereby enhancing their acceptance of clinic recovery services in these critical early weeks of recovery.

Along with several other stimulant medications, there are now encouraging signs that modafinil may be an effective pharmacotherapy for cocaine and synthetic stimulant dependence [3]. This ‘off label’ application makes sense from a neurobiological perspective although much remains to be understood about the precise mechanisms of action and, of course, larger more powerful trials are needed to firm the evidence base. Good implementation features of the present trial included monitoring of self-administration of trial medication, regular medical assessment of side effects, tolerable patient study attrition and good primary outcome measurement in the form of weekly urinalysis for target substance use (specifying for MA, cocaine and MDMA). There were also challenging aspects for the analysis: up to one third of participants completed their treatment, and the modafinil effect appears to have been moderated by concurrent polydrug use and the over-representation of HIV positive patients in the experimental group. More powerful studies are needed to investigate the pattern of outcome and the characteristics of non-response. Also, the assessment of neuropsychological problems and response during treatment may prove to be a useful avenue on the search for mediation of outcome and patient non-response.

Many medication trials in our field rest on a platform of structured psychosocial intervention (four sessions of cognitive behavioural therapy in this instance) as a standard treatment. The Shearer group suggest that modafinil may emerge as a combination therapy alongside a behavioural intervention: this makes good sense and offers real-world generalizability to clinic programmes. It will also be important for future triallists to ensure a high fidelity of their studied behavioural interventions, with their rationale, procedures and monitoring arrangements clearly set out in an accessible protocol or manual. This will facilitate replication of efforts and the translation of results into practice [4]. Making these materials available to the scientific community is now Addiction policy.

Declaration of interest

Dr Marsden holds a research grant from the Medical Research Council to evaluate the feasibility of a multicentre controlled trial of modafinil for cocaine dependence in specialist NHS treatment clinics in England.