A small randomized approach-bias training trial showed numerically favorable abstinence outcomes without statistically significant group differences. Adverse psychological reactions in both groups also made safety and acceptability important parts of the result.
Manning and colleagues randomized 24 residential-treatment patients to six sessions of personalized approach-bias modification or sham training. Ten received active training and 14 sham. A blinded researcher assessed reported use, craving and dependence severity one and three months after discharge. The active task trained avoidance of personally relevant drug pictures and approach toward goal-related pictures.
Among participants followed, abstinence proportions favored active training numerically, but the differences did not approach significance in this very small sample. Craving and dependence severity also did not significantly differ between groups. Three active and five control participants experienced adverse psychological reactions to the task, with two active and one control participant withdrawing. Positive ratings among training completers therefore cannot describe everyone’s experience.
The follow-up denominators were smaller than the randomized groups, and chance baseline differences could not be adequately controlled in such a small sample. A pilot can identify feasibility issues and uncertainty, but cannot turn a promising direction into established relapse prevention. Benefits previously found for alcohol-related training also cannot simply be transferred to methamphetamine. The study supports cautious further investigation with a stronger safety framework and adequate statistical power. It does not justify claiming proven efficacy, absence of discomfort or reliable individual benefit from the task.
See a different multicomponent digital-therapy intervention.