Abstract
Basic neuroscience research on brain plasticity, motor learning, and recovery has stimulated new concepts in motor rehabilitation. Combined with the development of methodological goal standards in clinical outcome research, these findings have effectuated the introduction of a double-paradigm shift in physical rehabilitation: (a) the move toward evidence-based procedures and disablement models for the assessment of clinical outcome and (b) the introduction of training-based concepts that are theoretically founded in learning theory. A major drive for new interventions has further come from recent findings on the adaptive capacities of neural networks and their linkage to perception, performance, and long-term recovery. In this context, constraint-induced movement therapy, an intervention initially designed for upper-limb hemiparesis, represents the theoretically and empirically most thoroughly founded training concept. Several clinical trials on constraint-induced therapy (CIT) have shown its efficacy in higher functioning patients; however, the transfer of the treatment into standard health care seems slow. Survey research further suggests a rather poor acceptance of CIT among clinical staff and it seems that the implementation of CIT is hindered by barriers constructed of beliefs and assumptions that demand a critical and evidence-based discussion. Within this context, we have conducted a series of experiments on amended CIT protocols and their application in the clinical environment which addressed the following issues: (1) massed practice: are 6 hours of daily training inevitable to achieve clinical benefits? (2) practicality: what is feasible in the standard care setting and what are the clinical benefits achieved by "feasible compromise CIT protocols?" (3) apprehensions: are concerns on increased muscular tone and pathologic movement patterns justified, and (4) learned nonuse: is the assumption of "hidden" residual present paper, the key findings of these studies will be summarized and critically discussed.