Virtual Reality Training for Upper Extremity in Subacute Stroke (VIRTUES) A multicenter RCT

Objective: To compare the effectiveness of upper extremity virtual reality rehabilitation training (VR) to time-matched conventional training (CT) in the subacute phase after stroke. Methods: In this randomized, controlled, single-blind phase III multicenter trial, 120 participants with upper extremity motor impairment within 12 weeks after stroke were consecutively included at 5 rehabilitation institutions. Participants were randomized to either VR or CT as an adjunct to standard rehabilitation and stratified according to mild to moderate or severe hand paresis, defined as >=20 degrees wrist and 10 degrees finger extension or less, respectively. The training comprised a minimum of sixteen 60-minute sessions over 4 weeks. The primary outcome measure was the Action Research Arm Test (ARAT); secondary outcome measures were the Box and Blocks Test and Functional Independence Measure. Patients were assessed at baseline, after intervention, and at the 3-month follow-up. Results: Mean time from stroke onset for the VR group was 35 (SD 21) days and for the CT group was 34 (SD 19) days. There were no between-group differences for any of the outcome measures. Improvement of upper extremity motor function assessed with ARAT was similar at the postintervention (p = 0.714) and follow-up (p = 0.777) assessments. Patients in VR improved 12 (SD 11) points from baseline to the postintervention assessment and 17 (SD 13) points from baseline to follow-up, while patients in CT improved 13 (SD 10) and 17 (SD 13) points, respectively. Improvement was also similar for our subgroup analysis with mild to moderate and severe upper extremity paresis. Conclusions: Additional upper extremity VR training was not superior but equally as effective as additional CT in the subacute phase after stroke. VR may constitute a motivating training alternative as a supplement to standard rehabilitation. ClinicalTrials.gov identifier: NCT02079103. Classification of evidence: This study provides Class I evidence that for patients with upper extremity motor impairment after stroke, compared to conventional training, VR training did not lead to significant differences in upper extremity function improvement.