Assessment and Evaluation of a Crosswalk Between the Functional Independence Measure and the Continuity Assessment...

To evaluate crosswalks between the Continuity Assessment Record and Evaluation (CARE) measure and the FIM motor subscale scores in a sample of persons with traumatic brain injury given the CARE measure's development by the Centers for Medicare and Medicaid Services in 2005 to assess motor function in all postacute rehabilitation settings and its intended continuity with previous measures.

Crosswalks were created using 3 methodologies (expert opinion, equipercentile, and Rasch). The dataset was split into training and validation datasets. Each crosswalk was evaluated using a reduction in uncertainty, the percent of each crosswalked score falling within ½ SD of the reference measure, population invariance, comparison of statistical moments, and effect size.

Record abstraction.

A total of 982 persons (N=982) who were hospitalized and received inpatient rehabilitation after traumatic brain injury and had both FIM and CARE collected at inpatient rehabilitation admission and discharge.

Not applicable.

Crosswalk between FIM and CARE.

In the training dataset, the expert opinion crosswalk met all criteria except the direction of population invariance within the race category. The equipercentile methodology satisfied all criteria. The Rasch model met all criteria except for a difference in directionality in the skewness of the distributions and 80% of scores not falling within ½ SD of the reference assessment. Results from the validation sample differed from the population invariance criteria, in which the age categories were in opposite directions and had observed differences between standardized mean differences for age that exceeded the threshold of 0.08 for both the equipercentile and Rasch crosswalks.

All 3 crosswalk methods produced acceptable criteria for use, indicating that motor/physical functional outcomes can be compared between cohorts assessed using CARE and FIM measures. Researchers can compare cohorts that have been assessed using these instruments on any of these crosswalks. However, we do not recommend using these crosswalks for longitudinal analyses.