Evidence for the effectiveness of serious games (SGs) and their various features is inconsistent in the motor rehabilitation field, which makes evidence based development of SGs a rare practice.
To investigate the effectiveness of SGs in motor rehabilitation for upper limb and movement/balance and to test the potential moderating role of SGs features like feedback, activities, characters and background.
We ran a meta-analysis including 61 studies reporting randomized controlled trials (RCTs), controlled trials (CTs) or case series designs in which at least one intervention for motor rehabilitation included the use of SGs as standalone or in combination.
There was an overall moderate effect of SGs on motor indices, d = 0.59, [95% CI, 0.48, 0.71], p < 0.001. Regarding the game features, only two out of 17 moderators were statistically different in terms of effect sizes: type of activity (combination of group with individual activities had the highest effects), and realism of the scenario (fantasy scenarios had the highest effects).
While we showed that SGs are more effective in improving motor upper limb and movement/balance functions compared to conventional rehabilitation, there were no consistent differences between various game features in their contribution to effects. Further research should systematically investigate SGs features that might have added value in improving effectiveness.