Effects of a virtual reality video game exercise program on upper extremity function and daily living activities in stroke patients.

The trial had only 24 patients, 12 in each group, and the text gives no dropout information, so the results may be unstable. The paper does not clearly name a primary outcome, so it is hard to know which result should carry most weight. The experimental group received video-game exercise in addition to traditional rehabilitation, so improvements within that group cannot be assigned to the games alone. The groups did not receive the same number or length of sessions: control traditional rehabilitation was 5 times per week for 30 minutes, while the video-game program was 3 times per week for 40 minutes in addition to traditional rehabilitation. The daily-living outcome used the Stroke Impact Scale, which depends on participants' subjective opinions. The reported experimental Stroke Impact Scale scores are identical before and after, yet the paper states a significant within-group difference. Only one occupational therapist delivered the intervention and assessments, and the therapist's training and experience may have affected results. Participants were selected for Brunnstrom stage 4 or above in the arm and hand, so the findings may not apply to more severely impaired patients. The paper reports no adverse events, safety, or cost information.

Do not conclude that video-game exercise improved daily living more than usual therapy. The arm/hand tests showed no between-group difference, and the daily-living numbers reported for the video-game group are inconsistent.