Background: Patients 3-5 days post stroke should be evaluated with a toolkit of valid assessment tools to assist in rehabilitation destination planning. A tool to assess stroke severity is a key component. A short form of the Functional Independence Measure (FIM), the AlphaFIM, is widely used. In Quebec, the Functional Autonomy Measurement System (SMAF) is used, however no short version exists.
Objective: To develop a short version of the SMAF, the SF-SMAF, to assist in selecting the optimal rehabilitation destination.
Methods: Patients, on average 5 days post stroke (n = 207) admitted to four academic acute hospitals were assessed with the SMAF. Linear regressions with the SMAF total score as the dependent variable and its 29 items as independent variables identified subgroups of items for inclusion in the SF-SMAF. A panel of experts (n = 10) selected the best subgroup of items, based on clinical experience. Multinomial logistic regression was used to examine the ability of the SF-SMAF to predict discharge destinations (home without rehabilitation, outpatient or inpatient rehabilitation). 
Results: Regression analyses and the expert panel led to retaining 4 of the 29 items of the SMAF: washing, walking inside, judgment, budgeting (R2 = 94%). The SF-SMAF predicts 66% of the rehabilitation discharge destinations. Adding the variables “presence of barriers to returning home” and “communication” to the SF-SMAF raise the prediction to 81%.
Conclusion: The SF-SMAF estimates the SMAF, is quick and easy to administer, and provides a measure of stroke severity to inform the choice of optimal rehabilitation destination.

Stroke, rehabilitation, triage, discharge destinations, stroke severity, Functional Autonomy Measurement System (SMAF)