Varied effectiveness of outpatient rehabilitation interventions in reducing long-term complications following childhood brain tumours: a systematic review.

Eight studies, with sample sizes from 9 to 60 children. With that many participants, a non-significant result may mean the study was too small to detect an effect rather than that the treatment does not work. Nine different interventions and more than 50 different outcome measures meant the results could not be pooled, and the review could not estimate how large any treatment effect was. Effectiveness was judged by counting how many results had a p-value below 0.05, which gives a 60-child trial and a 9-child trial the same weight. Grey literature, studies published only as conference abstracts and non-English papers were excluded, so some evidence was left out of the review. The included reports did not give enough detail for the reviewers to separate the findings by type of therapy. Participants and therapists could not be blinded to which treatment was given, which may have influenced patient-reported outcomes such as quality of life. One included study had no comparison group at all, and two had differences between the groups before treatment started. Follow-up varied and was often short, so it is unclear whether any gains last.

The authors declared that no financial support was received for the work or for its publication. The review does not report the funding sources or conflicts of interest of the eight included studies in the text provided.

Do not read the improvements in movement as proof that any of these interventions works. Each positive result comes from a single small study, in a set where sample sizes ranged from 9 to 60 children, and one of them (constraint-induced movement therapy) came from an uncontrolled before-and-after study of 9 children with no comparison group. The authors themselves say the evidence does not allow firm conclusions about which interventions are more effective.