ABSTRACT

Objective: This study is aimed at finding the effects of task-oriented robot-assisted therapy for stroke patients in terms of the motor function of the upper extremities and ADL performance based on a systematic review and meta-analysis. Methods: RCTs were identified through a database search of MBASE, Ovid-Medline, PubMed, CINAHL, and Cochran CENTRAL from 1980 to December 2015, and through additional manual searches. Meta-analyses were separately conducted in terms of the motor function of the subjects' upper extremities and ADL performance. A quality assessment was conducted using SIGN, the level of evidence was reviewed based on GRADE, and the risk of bias was assessed using the Cochrane collaboration tool. The statistical heterogeneities were also investigated, and RevMan 5.3 was used for a Meta analysis covering the averages and standard deviations. Results: Twelve studies were selected and reviewed. The overall task-oriented robot-assisted therapy group was made up of 341 participants (235 male, 106 female) with an average age of 60.50±5.51. The control subjects totaled 301 (207 male, 94 female, with an average age of 58.55±17.1). The improvement in the entire motor function of their upper extremities showed a statistical significance (mean difference, 0.34; 95% CI, -0.30 to 4.38; p<.05; I2=65.0%) with an FMA of 2.98 (mean difference, 2.10; 95% CI, 0.72 to 3.48; p<.05; I2=52.0%), BB of 2.25 (mean difference, 2.34; 95% CI, 0,30 to 4.38; p<.05; I2=28.0%), and FIM of 3.04 (mean difference, 4.14; 95% CI, -1.47 to 6.80; p<.05; I2=73.0%). However, their WMFT, overall ADL performance, ABILHAN, and MAL did not reach a level of statistical significance. Conclusion: Task-oriented robot-assisted therapy was shown to have a positive possibility for the recovery of stroke survivors, particularly in the motor function of their upper extremities and ADL performance. Further research will be needed to ground the evidence and support the clinical use of task-oriented robot-assisted therapy.