Fourteen studies were included in the review and meta-analysis (including 122 children, ranging from two to 31 children per study): three RCTs (57 children); two cohort studies (20 children); seven case series (37 children); and two studies described as single-subject designs (eight children).
For RCTs, total quality scores ranged from 5 to 11 (out of 16). For other study designs, quality scores ranged from 1 to 9 (out of 11).
For the three RCTs, no statistically significant difference in post-intervention upper extremity function was found between virtual reality groups and conventional therapy groups (d=1.97, 95% CI -0.26 to 4.20). In comparisons between virtual reality pre-intervention versus post-intervention group scores for this outcome, effect sizes were greater post-intervention and were either small (d=0.30) or large (d=1,31, d=5.09) (no other details reported).
When the results of all 14 studies were meta-analysed, upper extremity function showed statistically significant greater improvements post-intervention in virtual reality groups (d=1.00, 95% CI 0.45 to 1.56; Ι²=56%) compared with pre-intervention scores. Similar results were shown in studies that measured this outcome through participation (d=1.92, 95% CI 1.19 to 2.66) or body structure and function (d=0.70, 95% CI 0.10 to 1.30).
The strongest effects of virtual reality were shown with younger children, home or laboratory settings, engineer-built virtual reality systems, and among studies with RCT or case series designs.
Further results were reported in the review.