Can a Barthel score be derived from the FIM?

Objective: To establish whether a Barthel score derived by translation from the motor items of the Functional independence Measure (FIM) would equate to the directly scored measure. Design: Conversion criteria for motor item scores on the FIM scale to Barth el scores were first developed. To test these criteria, 40 consecutive pati ents were assessed for Barthel and FIM scores by the multidisciplinary team who were unaware of the conversion criteria. The derived Barthel score was compared with the directly scored Barthel Index. Results; A very high degree of correlation was observed between total score s of the direct and derived Barthel (Spearman's rho = 0.99), which is highl y significant, and no significant differences were seen between scores for any of the individual items (Wilcoxon signed rank test). Item by item analy sis across the study population was undertaken to confirm the conversion cr iteria. Absolute agreement between the two methods ranged from 75 to 100% a nd kappa values from 0.53 to 1.0. Conclusions: This study demonstrates that a Barthel Index can be derived fr om the motor items of the FIM and there is a good agreement with the direct ly assessed Barthel score. Although a larger study may help to delineate th e exact conversion criteria for one item, the current system provides an ac curate and usable translation of the total score and serves as a major step towards achieving a common language in outcome measurement for rehabilitat ion.