Evidence-based Treatment for Stimulant Use Disorder: Time to Hit the Community

We thank Bahji and colleagues for their response ¹ to our article, which adds to the debate on how the available evidence should be interpreted. Bahji and co-authors state that our claims were supported only by the meta-analysis by Tardelli et al., ² while other reviews lacked consistent evidence even though the most recent reviews highlight that prescription psychostimulants (PPs) warrant further investigation for the treatment of cocaine use disorder with the Cochrane review concluding that PPs promote cocaine abstinence. ³

This discrepancy could be because Tardelli and co-authors restricted the analysis to potent PPs: modafinil, amphetamine salts, and methylphenidate. Moreover, it included published results and data obtained directly from authors not included in those earlier systematic reviews, trials demonstrating the effectiveness of PPs given at higher doses of extended-release formulations, and with behavioral measures to ensure retention. More recent systematic reviews, which included results of recent trials with more appropriate methodology also highlighted the potential of PPs, namely prescription amphetamines, for the treatment of psychoStimulant use disorder (PSUD).

The study by Bentzley et al. ⁴ states that behavioural treatments have stronger evidence than pharmacological treatments, however both treatments may have additive or synergistic effects ⁵ and we propose the medication-centred model of care that includes contingency management (CM), psychosocial interventions, and harm reduction measures. Bahji and co-authors mention high dropout rates in PSUD treatment trials which we addressed in our commentary, suggesting that adjunctive CM may help address.

Finally, Bahji and co-authors state “Multiple (…) reviews on PSUD treatment concluded that since treatment dropout was high in many RCTs, high stimulant abstinence (…) could be explained by attrition bias instead,” but this statement is speculative. The imputation techniques by Bentzley and co-authors are timely and methodologically sound. However, if we cannot draw definitive conclusions from the existing evidence from systematic reviews and clinical trials, it is also problematic to draw opposite conclusions from an imputation analysis.

We agree that psychiatric adverse effects are a potential problem of PPs, and therefore propose that patients with increased risk of psychosis are not included in early community-based trials. History of psychosis is a well-recognized contraindication to PPs and we agree with applying this approach to PSUD patients. Furthermore, the population we propose to include has severe PSUD and is therefore frequently exposed to high doses of “street” stimulants. As such, reducing the use of these substances would likely reduce the risk of psychiatric complications. It is noteworthy that the Cochrane systematic review of PSUD pharmacotherapy did not find a significant difference between PPs and placebo for dropouts due to any, cardiovascular, or serious adverse events.

Effectiveness and implementation trials in community-based programmes will be necessary to address social determinants of health for individuals with PSUD, many of whom are excluded from efficacy, academic-based trials due to their comorbidities and adherence difficulties. An agonist-based model could attract those patients to the treatment setting and promote a therapeutic alliance with a multidisciplinary healthcare team. Community-based trials would offer a chance to assess this potentially effective and attractive treatment right where the most vulnerable and difficult-to-reach population is. While other efficacy trials, currently being conducted in many countries, are welcome to add to the existing evidence, community-based trials can also assess additional outcomes and reach different populations.

We recognize the approach we propose remains controversial. Still, we argue that community-based studies with careful patient selection, adequate doses of extended-release preparations, adjunctive behavioral interventions, and the latest methodology should be undertaken. We recognize the proposed intervention has risks, just like agonist therapy for opioid use disorder (OUD). Studies of opioid agonist therapy continued after the OUD treatment with methadone had been accepted and incorporated into clinical practice. Therefore, we continue to advocate that the urgency of the rise in stimulant-related deaths around the world demands community-based trials.