Introduction

Physical therapy has been found to improve sensorimotor function following stroke.–5 Early intervention might be better than late6 but even late poststroke physical therapy may be beneficial.3,7–9 Increased intensity of physical therapy may bring increased benefit,10–12 but this might not always be the case,13 as the severity of motor impairment may influence response to additional treatment.14 Not surprisingly, the content of physical therapy might also influence sensorimotor response,15 but at present, choice of treatment by physiotherapists appears to be mostly determined by the treatment approach which was prevalent during their training.16,17 As there are several different physical therapy approaches (Partridge18 describes nine), the potential for variation is extensive. Variation also arises from location of groups of therapists, e.g. the emphasis on the Movement Science Approach in Australia and the emphasis on the Bobath approach in the UK. There is a strong impression that physiotherapy practice has often been shaped by fashion, underpinned by clinical intuitions, based on uncontrolled observations and rationalized by quasi-scientific theories. An example of this is the Bobath approach, which we shall use as an illustration of the current inadequate scientific basis for commonly-used physical therapies.