Dropout Rate of Participants in Randomized Clinical Trials That Use Virtual Reality to Train Balance and Gait in Parkinson's Disease. A Systematic Review With Meta-analysis and Meta-regression.

Virtual reality is an effective system to train balance and gait in Parkinson's disease, but attrition of this intervention needs to be further examined. This study aims to review and meta-analyze the dropouts of participants in randomized clinical trials that used virtual reality for balance and gait training in people with Parkinson's disease. An electronic search was conducted in PubMed, Web of Science, Scopus and CINAHL. The PEDro scale and Revised Cochrane risk-of-bias tool for randomized trials 2.0 were employed to assess methodological quality. Proportions meta-analysis calculated dropout rate. Odds ratio meta-analysis under 1 indicated lower attrition in experimental participants. Meta-regression identified possible dropouts' moderators. A total of 18 studies were included. The pooled dropout rates were 5.6% (95% CI, 3.3%-9.3%) for all groups, 5.33% (95% CI, 3.03%-9.21%) in virtual reality, and 6.60% (95% CI, 3.84%-26.31%) in comparators. No statistical differences were found in the dropout occurred between the groups (OR 0.83; 95% CI, 0.62-1.12). Number of weeks was the unique moderator (coefficient 0.129, 95% CI 0.018- 0.239; p=0.02). Our overall pooled dropout should be considered in the sample size calculation of future studies. Adequate follow-up of the CONSORT guidelines in the loss report and their reasons could help design suitable retention strategies.