Four RCTs were included in the review (n=86). Study validity was rated as poor to moderate. Blinding of participants or people who implemented the intervention was not applicable in any of the studies. Blinding of assessors was only properly addressed in two studies.
Three small RCTs assessed differences in FMSA. Only one (n=13) reported a statistically significant difference in effect size in favour of MI, 11 points (95% CI: 1, 21). Of the other two small RCTs assessing differences in ARAT, only one (n=11) reported a statistically significant difference in favour of MI in comparison with control, 15.8 points (95% CI: 0.5, 31.0).