LAUSR

OBJECTIVE To link three Spinal Cord Injury - Functional Index (SCI-FI) item banks (Basic Mobility, Fine Motor Function, Self-Care) to the PROMIS® Physical Function (PF) metric. DESIGN Observational study SETTING: 6 SCI Model Systems rehabilitation hospitals in the U.S. PARTICIPANTS 855 adults with SCI and 730 healthy individuals. INTERVENTIONS Not applicable. MAIN OUTCOME MEASURES Three SCI-FI item banks (Basic Mobility, Fine Motor Function, Self-Care), PROMIS PF v1.0 item bank. METHODS AND RESULTS SCI-FI item banks (including 30 items from the PROMIS PF item bank) were administered to 855 adults with SCI as part of the original SCI-FI development study. The data were used to attempt to link three SCI-FI banks to the PROMIS PF metric via two item-response theory methods: fixed-parameter calibration and separate calibration. Sixteen items common to SCI-FI and PROMIS and verified as free of DIF were used as anchor items to implement the methods. Of the three banks, only SCI-FI Basic Mobility could be linked with sufficient precision to PROMIS PF. Comparisons of actual versus linked PROMIS PF scores and test characteristic curves suggested the fixed-parameter method provided slightly more precision than the separate calibration method. CONCLUSION The linkage between PROMIS PF and SCI-FI Basic Mobility was considered satisfactory for group-level usage. Score equivalents computed from SCI-FI Basic Mobility will be useful for researchers comparing functional levels in SCI to those observed in other clinical and non-clinical groups, for example, in comparative effectiveness research.