Test-retest and inter-method reliability of the self-rating Barthel Index

Objective: To develop a self-rating Barthel Index questionnaire (SE) for assessing stroke patients living at home, the test-retest reliability of SE versions 1 and 2 and the inter-method reliability of SE version 3 were examined. Design: Case series. Setting: Clinics of the Department of Rehabilitation Medicine at the University Hospital and two affiliated hospitals. Patients: Thirty-one, 140 and 65 chronic stroke patients living at home, being seen for regular follow-up, and having no severe aphasia for SE3 versions 1, 2 and 3, respectively. Main outcome measure: Kappa coefficients of items in SE versions 1 and 2 for test-retest reliability; intra-class correlation coefficients of the self-care index, mobility index and total score, kappa coefficients of items and Friedman's two-way ANOVAs of ranking in SE version 3 among the self-ratings, family ratings and occupational therapist (OT) ratings; a regression analysis of age, gender, duration, time:or filling in the SE, OT ratings, Mini-Mental State score and aphasia. Results: Kappa coefficients in SE version 2 were higher than those in version 1. Intra-class correlation coefficients in SB version 3 among the three ratings were good, and kappa coefficients in each item were-also good or excellent. By Friedman's two-way ANOVA, all the items except walking up/down stairs had no significant difference in ranking. Regression analysis revealed that OT ratings and time for filling in the SE affected the absolute difference of scores between the self-ratings and OT ratings. Conclusion: SE version 3 is sufficiently reliable for practical use.