Effects of different rehabilitation training on balance function in stroke patients: a systematic review and network meta-analysis

INTRODUCTION: Stroke is a major noninfectious disease that endangers patients’ health. About 83% of patients have some degree of balance dysfunction. Rehabilitation training is an effective means to improve the balance function of stroke patients. The aim of this meta-analysis was to assess which of the eight rehabilitation training methods was more effective in promoting balance recovery. MATERIAL AND METHODS: Eight studies of rehabilitation training to improve balance function in stroke patients were searched in PubMed, Scopus, Web of Science, Embase, Cochrane, Chinese National Knowledge Infrastructure (CNKI), Wanfang, and VIP databases. Data extraction was carried out by two independent researchers. The improvement in balance function included in the study was measured using the Berg Balance Scale (BBS). A random effects model was used to assess the mean difference in these values between patients who received different rehabilitation training and controls. RESULTS: A total of 46 randomized controlled trials (RCTs) were included, including 3741 patients. Meta-analysis results showed that the clinical effectiveness in promoting balance recovery in stroke patients was ranked as follows: core stability training (odds ratio (OR) = 14.98, 95% confidence interval (CI): 8.18–27.44) > whole-body vibration training (OR = 10.27, 95% CI: 4.36–24.18) > mirror therapy (OR = 5.15, 95% CI: 2.40–11.04). CONCLUSIONS: This results suggested that core stability was more beneficial for improving balance function in stroke patients.