Evaluating change in mobility in people with multiple sclerosis: relative responsiveness of four clinical measures

Background and objectives: Determining responsiveness of measures across different samples and settings is important for selecting measures of mobility and understanding multiple sclerosis (MS) study results. Currently such information is limited. Methods: This exploratory study examined the relative responsiveness of four mobility measures (walking velocity, 6-minute walk distance, Rivermead Mobility Index and MS Walking Scale) in a community sample (n = 70), after three annual assessments. Distribution based estimates and anchor-based methods (comparison against transition questions) determined responsiveness. A head-to-head comparison was made. Results: While variations in individuals occurred, the group mean change scores for all measures was small, suggesting that there was minimal deterioration in the total sample. Consistent with this, total sample Effect Size (ES) was negligible to small (ES -0.32 to +0.03) for all measures. Differentiation between sub-groups, defined by the participants' perception of change in mobility over the past year (transition questions), showed that some instruments could detect clinically significant changes (small sample sizes limited this interpretation). Correlation analyses between change scores demonstrated that these measures captured related, but different information (r < 0.364). Conclusions: The measures were broadly comparable in detecting mobility changes in this community sample. These correlations highlight that in selection of measures, one should also consider the discrete mobility dimension that the intervention intends to impact.