A systematic literature review of computerized training for attention and executive function in adults who suffered a brain injury (TBI or stroke) has concluded that there is encouraging evidence that such programs can help.
The review found 23 of 28 studies reported significant improvements in attention and executive function, with the remaining five showing promising trends. The studies included 11 that focused on TBI, of which 8 reported significant improvement; 5 that focused on stroke, of which all 5 showed significant improvement; 12 mixed-populations, of which 10 showed significant improvement.
Further studies are needed to confirm these results, as various methodological issues, such as a small number of participants, and inadequate controls, need to be addressed. The 28 studies included 9 that were rates as "class I" (the highest standard), 9 class II, and 7 that were class III (no controls). Almost all (26/28) of the studies involved fewer than 50 participants, with some having as few as 1 to 4. Most studies didn't specify how severe the injuries were, something which makes a big difference to treatment and expectations. Over a third of the studies (11) didn't have any control group, and only a few used the best sort of control - a comparable activity (as opposed to, say, no treatment). Only four studies provided any long-term follow-up.
As you can see, a lot of work is needed yet. Moreover, most programs were unique to the study, so we're still some way off producing recommended protocols. Only one program was used on multiple occasions (5): Cogmed QM (originally called RoboMemo).
Still, notwithstanding all these caveats, the review does support the value of specific training for those suffering brain injury.