Twelve RCTs (21 publications), consisting of 2,813 patients, were included in the review.
Two studies received a PEDro score of 8, three scored 7, three scored 6, and four scored 5.
Functional outcomes.
Seven of the 12 studies investigating functional outcome found in-patient stroke rehabilitation beneficial in comparison with conventional care. Five studies failed to show a beneficial effect.
Mortality.
Three of the 12 studies assessing mortality found in-patient stroke rehabilitation to be beneficial in reducing mortality in comparison with conventional care. Nine studies failed to show a beneficial effect.
Length of hospital stay.
Five of the 8 studies found length of stay to be reduced in patients receiving stroke rehabilitation in comparison with conventional care. Three studies did not show a beneficial effect.
Discharge outcome.
Three of the 11 studies investigating discharge outcome found that, compared with conventional care, patients were more likely to have a good outcome (defined as an increase in the number of patients returning home after hospitalisation, or a decrease in the number of patients admitted to a long-term care facility) having received stroke rehabilitation. Eight studies failed to show a significant difference between the treatment groups.