Makoto Tokunaga, MD, PhD, Katsuhiko Sannomiya. RPT, Yukihiko Nakashima, OTR,
Shinichi Nojiri, RPT, Kaori Tokisato, MD, Kenichi Katsura, MD,
Susumu Watanabe, MD, PhD, Ryoji Nakanishi, MD, PhD, Hiroaki Yamanaga, MD, PhD
Jpn J Compr Rehabil Sci 6: 6-13, 2015
Objective: To predict FIM gain and discharge FIM
score by multiplying the standard value by influence
coefficients for age, cognitive function, and transfer
interval.
Methods: The subjects were 1,118 stroke patients
admitted to a Kaifukuki rehabilitation ward of hospital
A. The median value of motor FIM (mFIM) gain and
discharge mFIM based on mFIM at the time of
admission was used as the standard value. We then
created a formula for predicting mFIM gain and
discharge mFIM by multiplying the standard value by
the influence coefficients for age, cognitive function,
and transfer interval.
Results: The correlation coefficient between the
actual and predicted values was 0.681 in the prediction
of mFIM gain and 0.874 in the prediction of discharge
mFIM. The residual of the subtraction of the predicted
value from the actual value was 1.4±12.5 (median
value: 0) in the prediction of mFIM gain, and 1.3±12.6
(median value: 0) in the prediction of discharge mFIM.
Conclusion: The correlation coefficient is comparable
with those of reports that use multiple regression
analysis. This new method clearly showed the
relationship between factors and mFIM gain/discharge
mFIM.
Key words: Functional Independence Measure, FIM gain, multiple regression analysis, formula, factors