Thirty RCTs were included in the review. The quality of the trials was moderate, with a mean quality score of 5.5 (range three to eight). Two trials were rated as high quality; 17 trials were acceptable quality; and 11 trials trials were poor quality.
For 60 to 72 hours for two weeks: Significant benefits with constraint-induced therapy were observed for faster hand and arm use in two high-quality trials (WMD -0.46, 95% CI -0.59 to -0.33; Ι²=0), and it increased the amount of motor activity (WMD 0.85, 95% CI 0.68 to 1.01; Ι²=0) and quality of motor activity (WMD 0.69, 95% CI 0.45 to 0.93; Ι²=13%). There were no differences observed in self-care between the treatment group and a control group in one trial.
For 20 to 56 hours over two weeks: There were significant improvements with treatment, compared with control, in the amount of motor activity (WMD 1.11, 95% CI 0.89 to 1.34; Ι²=14%; two high-quality trials) and quality of motor activity (WMD 0.78, 95% CI 0.20 to 1.36; one high-quality trial). There were no differences between groups in functional independence measures.
For 30 hours over three weeks: Statistically significant benefits were found with the intervention in the amount of motor activity (WMD 0.63, 0.17 to 1.10; Ι²=55%; four trials, with five comparisons), quality of motor activity (WMD 0.79, 95% CI 0.26 to 1.32; Ι²=63%; four trials; five comparisons). Significant benefits were found in three trials for self-care with the intervention, but there were differences between the text and figures for this result.
For 15 to 30 hours over 10 weeks: Hand mobility was significantly enhanced with treatment, compared with control (WMD 11.37, 95% CI 4.17 to 18.57; one trial) and compared with no treatment (WMD 15.34, 95% CI 4.05 to 26.63; one trial).