Constraint-induced movement therapy: some thoughts about theories and evidence.

Constraint-Induced Movement Therapy (CIMT) is a type of treatment for hemiparetic stroke patients in which the patient is strongly encouraged to use the affected arm. One way of doing this is to immobilise the unaffected arm. This treatment is meant to help patients overcome 'learned non-use'. The learned non-use theory is based on deafferentiation experiments in monkeys. In this review four randomised clinical trials are presented systematically. Although the authors of all four studies reported positive results, the effect sizes calculated without covariates yielded no statistically significant differences. In one of the studies a differential effect was found for patients with sensory disorders and hemineglect, leading to the hypothesis that learned non-use may be primarily related to afferent impairments. It is concluded that the learned non-use theory requires further exploration and that the evidence regarding the effectiveness of CIMT is not yet conclusive.