[For whom is training after stroke most beneficial? Selection method exists and should be used]

In the light of recent Swedish health care cost cuts, this review addresses the need of limiting stroke rehabilitation to those it will benefit most. In many prognostic stroke studies, end-points have been adopted that vary as to time and place--e.g., duration of hospitalisation or discharge to a nursing home. Even modest improvement in ability may significantly enhance quality of life, but is not necessarily associated with beneficial effects on lifestyle or the length of hospital stay. Accordingly, we need to use reliable tests with more realistic end-points. The use of the Functional Independence Measure (FIM) and the Katz ADL scale as prognostic instruments is discussed.