Computer-assisted cognitive rehabilitation in persons with multiple sclerosis: Results of a multi-site randomized controlled trial with six month follow-up.

The comparison group was not a no-treatment control: it received usual care plus weekly check-in calls and a referral to MyBrainGames, so improvement cannot be compared with doing nothing. The active intervention bundled weekly group sessions with home computer training, so the paper cannot show which part produced the effects. Both groups improved over time on all measures, and the paper describes the between-group differences as small. Participants were recruited from three Texas cities, were mostly well-educated English speakers, and had mild to moderate MS impairment, so results may not apply to less educated, more impaired, or non-English-speaking patients. Missing data were handled by carrying the last observed value forward, so reported outcomes include imputed values rather than only completed measurements. The paper does not report how many participants dropped out after randomization, only that 183 were included in analysis. Daily-function outcomes were measured with a shortened problem-solving test, and the authors say available measures have limited ecological validity.

The authors declared no conflicts of interest. The active computer component was contracted with Lumos Labs, and the supplied publication types list NIH extramural support.

Do not conclude that computerized brain training alone improved cognition. The active group received 8 weekly group sessions plus home computer training, and the comparison group received usual care plus MyBrainGames and weekly check-in calls. The study cannot separate which component caused the small between-group differences.