Randomized trial of working memory training as an adjunct to inpatient substance use disorder treatment.

Despite interest in computerized working memory training as a transdiagnostic intervention, little is known about its feasibility and efficacy in addiction treatment settings. This double-blind, randomized controlled trial examined the efficacy of working memory training for improving performance on standardized measures of executive function during inpatient substance use disorder treatment. Secondary outcomes included delay discounting, adherence to cognitive training, and posttreatment relapse. Adults (N = 110) entering inpatient alcohol or drug treatment were randomized to active (adaptive) or control (nonadaptive) training conditions, with up to 25 training sessions spanning inpatient and postdischarge phases. In generalized estimating equation (GEE) analyses, the experimental group showed relatively greater improvements on the primary outcome (digit span performance) based on Time × Treatment interactions in per-protocol (p < .05) and intent-to-treat (p = .07) models. Similar results were observed for one secondary outcome. Lower working memory performance at baseline predicted significantly greater delay discounting; however, no time or treatment effects on discounting were observed. Adherence to cognitive training was generally good during inpatient treatment, but suboptimal overall. These results suggest that implementing cognitive training is feasible in inpatient settings, and that adaptive working memory training can potentially lead to near-transfer effects for select cognitive outcomes. However, further work is necessary to establish training effects on clinical outcomes and to improve adherence in outpatient contexts. (PsycINFO Database Record (c) 2018 APA, all rights reserved)