Treatment engagement can improve without a demonstrated reduction in drug use during the same study. A trial should be credited for the outcome it measured successfully, while keeping the separate drug-use result visible.
What was randomized and analyzed?
Stauffer and colleagues’ 2020 double-blind trial evaluated intranasal oxytocin 40 IU or placebo before six weekly motivational interviewing group sessions for men who have sex with men and severe methamphetamine use disorder. Forty-eight participants attended the first session and entered the primary analysis: 24 per arm, across five therapy cohorts per arm.
The attendance model excluded session one, which everyone in that analysis attended. Oxytocin increased subsequent attendance odds: OR 3.26, 95% CI 1.27–8.41, p=0.014. The investigators did not find a methamphetamine-use benefit or a craving difference. They accounted for repeated observations within participants but not dependence within therapy cohorts. People who never attended session one were excluded.
What does an attendance odds ratio mean?
The ratio concerns the odds of attendance in the fitted session model. It is not a statement that 3.26 times as many people completed treatment, nor that methamphetamine consumption fell by that factor. Both the outcome and its statistical scale should accompany the number.
The confidence interval also belongs with the estimate. Reporting only the point estimate would hide the range of effect sizes compatible with the model and data. A small study’s apparently large estimate should not be presented as a precisely established constant for other programs.
Participants and therapy cohorts are different units
Group treatment creates potential shared influences: the therapist, session atmosphere and interactions between members. Repeated measurements from one person and shared experiences within a cohort represent different sources of dependence. Adjusting for one does not automatically account for the other.
The first-session requirement adds another boundary. The result concerns participants who reached that starting point. It cannot establish that oxytocin brings people into their first visit, because those who never attended did not receive the intervention and were outside the analysis.
What would establish a broader treatment benefit?
A larger trial could test whether an engagement benefit persists and translates into the desired clinical outcome. That translation should be measured directly. Improved attendance is a potential step in a treatment pathway, not proof that every later step occurred.
This historical human trial therefore adds a different kind of evidence from animal oxytocin experiments: a specific attendance result in a defined therapy setting, accompanied by null drug-use findings. It does not establish general effectiveness for methamphetamine abstinence or supply a current instruction to use an intranasal medicine.
Source
Stauffer CS, et al. Oxytocin-enhanced group therapy for methamphetamine use disorder: Randomized controlled trial. J Subst Abuse Treat. 2020;108059. doi:10.1016/j.jsat.2020.108059. Full text.