Implementation of approach bias re-training in alcoholism—How many sessions are needed?

Background: Computerized cognitive bias modification (CBM) programs have generated promising results regarding the treatment of alcohol dependence and anxiety disorders. However, there is hardly any research yet on the implementation of alcohol-CBM into clinical practice. This article addresses the question of the optimal number of training sessions for a specific form of CBM: approach bias re-training in alcohol-dependent patients. Methods: Participants were 111 alcohol-dependent patients of an inpatient rehabilitation clinic. In addition to treatment as usual, they took part in a 12-session CBM protocol, aimed at re-training alcohol approach tendencies, with an adapted Alcohol Approach-Avoidance Task. Learning curves and 2-level fitted mean trends were analyzed. Furthermore, parameters of the fitted mean trend and information on the most effective number of training sessions were used to predict 1-year follow-up data. Results: Two-level analyses revealed 6 training sessions to be the mean optimum, yet many patients improve further after that. Prediction of the individual optimum (speed of response) and parameters of the fitted mean trends failed. The linear term of the graph describing the mean trend of training effects, but not the speed of response, predicted 1-year follow-up data. Conclusions: Suggestions for practical implementations of CBM are discussed