Design: For each patient FIM scores were assessed 4 times during hospitalization. Ordinal logistic analyses were performed on group data.
Subjects: Fifty patients with hemiplegia after stroke staying in a long-term rehabilitation facility.
Results: Analyses revealed that FIM-motor scores accounted for much of the variability of independence levels for most of the single FIM™ items, including dressing upper body, and transfers to bed/chair/wheelchair and to toilet. For these items, the independence levels were proportionally associated with FIM-motor scores. For eating, higher FIM-motor scores (>60) were associated with modified independence and lower FIM-motor scores (<40) correlated with attainment of supervision/set-up levels. For dressing lower body, greater independence was apparent when FIM-motor scores were higher (>60).
Conclusion: For single FIM™ items, relative difficulty was comparable with results from previous literatures using Rasch analyses. Moreover, our results revealed that relative difficulty for single items varied greatly between independence levels. With regard to disability task targets, probability of independence evaluated from logistic modelling is an aid to efficient rehabilitation scheduling.