Hierarchical linear modeling of FIM instrument growth curve characteristics after spinal cord injury.

Objective: To examine the recovery of aspects of functional independence as a continuous process using growth curve analysis.

Design: Retrospective database review of functional outcome assessment data from inception cohort.

Setting: Inpatient rehabilitation unit; community.

Patients: A total of 142 subjects (79.6% men; age range, 18-77yr; mean age +/- standard deviation, 36.2 +/- 15.5yr) who were admitted to a rehabilitation unit between March 1986 and November 1994 with a minimum of 4 postinjury FIM assessments. Neurologic subgroups included 63 individuals with paraplegia, 36 with low tetraplegia, 24 with high tetraplegia, and 19 with incomplete injury.

Main Outcome Measure: FIM instrument.

Results: Growth curve analyses with hierarchical linear modeling using a decelerating recovery function yielded a reliable model in which longer rehabilitation length of stay was associated with a more rapid rate of recovery but lower plateau. Neurologic injury category had expected effects on rate and degree of recovery. Level of impairment-specific results included an age effect in which older age was associated with lower level of plateau. In specific neurologic groups there was a significant gender effect, in which men made more rapid recovery than women, and a significant effect of level of education, in which higher education was associated with more rapid rate of recovery. Rate of FIM recovery was reliably modeled in the sample with incomplete injuries, but none of the demographic predictors was significant.

Conclusions: Functional recovery can be modeled as a decelerating rather than simple linear function. The study of predictors of recovery characteristics, including rate of recovery and plateau, offers a valuable way of understanding rehabilitative needs and outcomes. Gender and education effects on the recovery process are intriguing and warrant further investigation.