Clinical outcomes for people after stroke are enhanced when management is delivered according to clinical guidelines. However, a large evidence-practice gap exists which is being addressed by knowledge translation research. The effectiveness of these activities in changing clinical practice or improving patient outcomes is not known.
A systematic review of randomised controlled trials determined the effectiveness of knowledge translation activities in stroke rehabilitation. Studies with the primary aim to translate research evidence into practice were included. Study bias was assessed.
From 1046 citations identified, 76 full text articles were reviewed and 14 trials included.   Twelve trials were cluster randomised trials. Interventions were either directed at a single profession (n=5) or multi-disciplinary teams/multiple professions (n=9). Nine studies described resource intensive multi-faceted knowledge translation interventions; seven used a site facilitator or had access to a mentor. Four described education interventions, and one used electronic reminders. Risk of bias was generally low. The type and number of outcomes varied considerably (range 1-21, mean 8) with four studies reporting patient outcomes only, five reporting clinician practice outcomes only and five providing data on both. Studies with six or less outcomes produced significant results in at least one primary outcome; the effect size of these was modest. 
Translating research evidence into practice using large clustered randomised controlled trials with multiple professions and multiple outcomes produces modest improvements in some but not all of the clinician practice and patient outcomes.  Trials with site facilitators or which focus on fewer outcomes appear to be more effective.

Systematic Review of Knowledge Translation Trials in Stroke Rehabilitation