Is early speech and language therapy after stroke a waste?

Research into rehabilitation is a relatively young discipline and there are still few centres of academic excellence in stroke rehabilitation. Randomised trials are difficult to conduct in this area. Large variations in patients and disease characteristics make designing trials tricky; blinding may be difficult; and identifying appropriate control interventions and ensuring that interventions are standardised, especially in multi-site studies, is challenging. However, such problems are not unique to rehabilitation research and can be overcome using complex intervention evaluation methods, as has been shown in other areas of stroke care.1 In a linked research paper (doi:10.1136/bmj.e4407), Bowen and colleagues report the findings of the ACT NoW (Assessing Communication Therapy in the North West) study, which is a large scale multicentre randomised controlled trial of speech and language therapy in the rehabilitation of patients after stroke.2 This study is welcome because, as with other treatment interventions, it is essential that rehabilitation is subjected to rigorous scrutiny.

Single case studies, observational studies, and small single centre randomised trials provide weak evidence for guiding clinicians and planners of care, and they cannot answer fundamental questions about how and what services should …