Effect of virtual reality-based upper limb training on activity of daily living and quality of life among stroke survivors: a systematic review and meta-analysis.

Most trials did not blind the participants or the therapists and did not analyse by intention to treat, and the authors name deviation from the intended intervention as the main source of bias, so part of the measured advantage could come from people knowing they got the new treatment. Studies differed in age, sex, stroke type (both ischaemic and haemorrhagic survivors were pooled, and their recovery paths may differ) and in training dose and duration, so the pooled result averages over quite different programmes. Several trials combined VR with robotic exoskeletons or neuromuscular stimulation, so in those trials the effect cannot be attributed to VR alone. Most individual trials were small and follow-up was short; the review cannot say whether gains last. Only 8 trials measured quality of life, and the authors themselves say those results should be interpreted cautiously. Only English-language papers were included, 20 of the 30 trials came from Asian countries and 9 from Europe, and no trial came from Africa, so most settings and health systems are unrepresented.

The only declaration given in the text supplied names a funder, the NASCERE Project. No competing interests are stated.

Do not read the large quality-of-life figure as proof that VR produces a big quality-of-life gain. It comes from only eight trials measured four different ways, the authors warn that these results should be interpreted cautiously, and they give the intervention only a conditional recommendation.