Computerized Working Memory Training for Children with Moderate to Severe Traumatic Brain Injury: A Double-Blind, Randomized, Placebo-Controlled Trial.

Pediatric traumatic brain injury (TBI) places children at risk for deficits in working memory (WM; comprising a central executive [CE], and two storage systems: phonological loop [PL] and visuospatial sketchpad [VSSP]), which is strongly related to attention and academic skills in childhood. This study aimed to examine whether different components of WM can be improved following adaptive WM training (Cogmed) and whether improvements in WM generalize to other cognitive (attention) and academic skills (reading and mathematics) in children with TBI. Twenty-seven children with moderate to severe TBI were randomized to adaptive (Cogmed; n = 13) or non-adaptive training (active placebo; n = 14) and evaluated at baseline, post-training, and 3-months follow-up. Three children in the adaptive group and one child in the non-adaptive group withdrew from the study before completion of training. Complete case (CC) and intention-to-treat (ITT) analyses were conducted. Children in the adaptive group demonstrated significantly greater gains on select WM tasks (VSSP, but not PL or CE) from pre- to post-training (pre-post) and pre-training to follow-up (pre-follow-up; CC and ITT analyses). No gains were found on tests of attention. Adaptive training resulted in significantly greater gains on select academic skills (reading, but not mathematics): reading comprehension pre-post-training (ITT analyses) and reading accuracy pre-follow-up (CC and ITT analyses). This first, to our knowledge, study to examine the efficacy of adaptive WM training for children with TBI provides preliminary evidence of near and far transfer of training to WM and academic skills, respectively.