Reliability of the revised functional autonomy measurement system (SMAF) for epidemiological research

The Functional Autonomy Measurement System (SMAF) is an instrument designed to assess disabilities related to 29 functions with a four-point scale (from 0: independent to -3: dependent). For epidemiological studies, a total score and five sub-scores can be obtained. A revised version was developed adding a -0.5 level to many items to indicate an activity accomplished independently but with difficulty. The objective of the study was to verify the test-retest and inter-rater reliability of the total score and sub-scores of the SMAF. Ninety subjects were randomly recruited in nine different residential settings ranging from home to long-term-care hospitals. Half of the subjects were assessed by the same nurse within a 2-week interval (test-retest) and the other half were assessed twice by two different nurses within the same interval (inter-rater). Results show intra-class correlation coefficients (ICC) of 0.95 and 0.96 for the total scores on test-retest and inter-rater reliability, respectively. The ICC were over 0.74 for all sub-scores for both types of reliability. A small systematic bias was present for two SMAF subscores on the inter-rater reliability. The addition of a new level did not modify the reliability of the scale.