Bt. Mast et Pa. Lichtenberg, Assessment of functional abilities among geriatric patients: A MIMIC modelof the functional independence measure, REHAB PSYCH, 45(1), 2000, pp. 49-64
Objective: To investigate the influence of population heterogeneity among g
eriatric patients on the factor structure and differential item functioning
on the Functional Independence Measure (FIM) and thereby evaluate the appr
opriateness of using 2 factor composite scores (FIM motor and FIM cognition
) among geriatric patients. Design: A specific application of structural eq
uation modeling (i.e., multiple indicators, multiple causes [MIMIC] model)
was conducted on FIM admission scores from 718 geriatric rehabilitation pat
ients. Setting: A large urban rehabilitation hospital. Participants: Seven
hundred eighteen inpatients referred for medical rehabilitation (M = 77.5 y
ears; 62% African American; 68% female). Measurements: All participants' fu
nctional abilities were rated on the FIM by trained physical and occupation
al therapists on admission to the rehabilitation hospital. In addition, all
patients completed the Geriatric Depression Scale. Results: Three motor fu
nctioning items and 3 cognitive functioning items differed systematically a
cross groups within this geriatric sample (i.e., young-old vs. old-old, mal
e vs. female, depressed vs, nondepressed) even after controlling for group
differences on levels of overall motor and cognitive functioning, Conclusio
ns: Composite FIM motor and FIM cognition scores are not invariant across s
ubgroups of geriatric patients, and therefore these composite scores may le
ad to inaccurate comparisons of functional abilities between these groups.
Specific recommendations for using the FIM among geriatric populations are
discussed.