Four RCTs were included in the review (226 children). Allocation concealment and blinding was adequate for all RCTs. Follow-up varied between trials.
The meta-analysis showed that the GMFM change was higher with intensive treatment than with non-intensive treatment (MD 1.32, 95% CI 0.55 to 2.10; Ι²=0).
The authors reported that the effects of intensive therapy tended to be stronger for children under two years old (MD 5), but this was not statistically significant (95% CI -0.45 to 10.45). For children at least two years old, the difference was small, but significant (MD 1.25, 95% CI 0.47 to 2.03).
There was no significant difference between the two groups when the treatment lasted for less than 60 days; whereas there was a significant difference in favour of intensive therapy when it lasted for at least 60 days (MD 1.42, 95% CI 0.55 to 2.30).
There was no evidence of publication bias.