Abstract
Context: GRASSP Version 1 (GV1) was developed in 2010, is an upper extremity measure specifically designed to assess recovery after traumatic tetraplegia. A second version was developed to reduce length of the test and refine instructions/standardization. The purpose of this post hoc analysis was to calculate psychometric properties of GRASSP Version 2 (GV2). Design/Setting: A post-hoc analysis of datasets for the GRASSP cross-sectional (n = 72 chronic,) and longitudinal (n = 127 acute) studies was conducted. Reliability, validity and MDD were calculated from the chronic sample and responsiveness was re-calculated from the longitudinal sample. Both studies were observational. Participants: A chronic sample (n = 72) and acute longitudinal sample (n = 127) of individuals with traumatic tetraplegia (AIS A to D, NLI C2 to C8) were studied. Outcome Measures: GV1, the Spinal Cord Independence Measure III (SCIM), International Standards of Neurological Classification of Spinal Cord Injury (ISNCSCI) were administered in both studies at all centers and the Capabilities of the Upper Extremity Questionnaire (CUE-Q) was administered in North American sites only. GRASSP-Palmar Sensation, GRASSP-Prehension Performance subtest items included in GV2 were re-analyzed for reliability; validity, MDD and responsiveness. Results: Inter-rater and test-retest reliability for all subtests ranged between 0.849-0.971 and 0.950-0.971 respectively. Concurrent validity between domains of GV2 were positively and moderately (0.530-0.830, P < 0.0001) correlated to SCIM, SCIM self-care subscore (SS) and CUE-Q. MDD values were 4 and 3 points for sensation and prehension performance (single side). Responsiveness values were .84-.88 for GR-Sens and .93-1.22 for GR-PP respectively. Conclusions: GV2 retains excellent psychometric properties as does GV1.