Abstract: Objective To evaluate the effect of virtual reality on motor function in patients with stroke. Methods PubMed, Cochrane Central Register of Controlled Trials, China National Knowledge Infrastructure (CNKI), VIP Database for Chinese Technical Periodicals and Wanfang Data were searched for the randomized controlled trials (RCTs) of virtual reality on motor function in patients with stroke from the date of establishment to January 2013. The included studies were evaluated according to the method recommended by the Cochrane Collaboration. RevMan 5.0 software was used to analyze the extracted data. Results 20 trials were included (13 trials about the upper extremity function and 7 trials about the lower extremity function). Compared with conventional training, the virtual reality training significantly increased the score of Fugl-Meyer assessment (FMA) (WMD=4.27, 95%CI: 2.47~6.06, Z=4.67, P<0.00001) and Box and Block Test (BBT) (WMD=9.29, 95% CI: 5.24~13.34, Z=4.50, P<0.00001). However, the results of Berg balance scale (BBS) (WMD=1.63, 95% CI: -0.83~4.09, Z=1.30, P=0.20) and walking speed (WMD=0.01, 95%CI: -0.14~0.17, Z=0.18, P=0.86) were not statistically significance for the lower extremity function. According to the Jadad Score of included studies, 16 of them were of low quality and only 4 of them were of high quality. Conclusion The virtual reality training could improve the upper extremity function in patients with stroke, but not for the lower extremity function.

Key words: stroke, virtual reality, motor function, randomized controlled trials, systematic review, Meta-analysis