Rasch analysis of the Rivermead Mobility Index: A study using mobility measures of first-stroke inpatients

Objective: To evaluate the validity and item unidimensionality of the Rivermead Mobility Index (RMI) by using Rasch analysis. Design: Application of Rasch analysis on the RMI partial data set. Setting: A stroke program at a rehabilitation hospital in Italy. Participants: A total of 308 consecutive patients (155 women, 153 men; avg age, 62.79+/-11.94y) hospitalized between 1990 and 1996. Average interval between stroke onset and admission was 52.48+/-36.22 days. Intervention: Medical inpatient rehabilitation. Main Outcome Measures: Patients' mobility status was assessed using the RMI administered at admission and discharge. Ratings were assigned by 4 staff members working as a team. We performed separate Rasch analyses on the RMI data, gathered from different groups of first stroke inpatients examined before and after rehabilitation treatment. Results: Rasch analysis showed the overall good validity of the RMI, except for item 15, which did not fit the unidimensional continuum estimated through the Rasch rating model. Conclusion: The RMI is a unidimensional scale with a hierarchy of easy-to-hard test questions. Item difficulty level was stable when processed on different groups of patients assessed on different occasions.