Establishing the predictive validity and determining cutoff values of the capabilities of upper extremity test for predicting activities of daily living independence in cervical spinal cord injury.

Abstract

Retrospective observational study. The Capabilities of the Upper Extremity Test (CUE-T) is a tool that assesses upper limb function in detail among individuals with cervical spinal cord injury (CSCI). We aimed to evaluate the predictive validity of the CUE-T for independence in activities of daily living (ADL) at discharge. Spinal cord injury rehabilitation center in Chiba, Japan. The cutoff value for independence in ADL (six self-care items of the spinal cord independence measure III [SCIM Ⅲ]) at the time of discharge from a rehabilitation center was calculated based on the results of the CUE-T conducted within 3 months of injury for individuals with CSCI. We defined a SCIM item score of ≤ 1 as dependence and ≥ 2 as independence. Analyses were performed using a receiver operating characteristic curve; the cutoff value was calculated using the Youden index. Bootstrapping was used for internal validation. Data were collected from 44 individuals with CSCI (median age: 59.0 [interquartile range, 50.8-67.3] years; men: 38, women: 6; American Spinal Injury Association Impairment Scale A/B: 7; C/D: 37). The cutoff values for the SCIM III self-care items ranged from 13-61 points. All area under the curves were 0.8 or greater, which was similar to the results obtained using bootstrapping. The CUE-T cutoff values calculated in this study showed good predictive ability and reproducibility and were clinically useful. These results demonstrate the high predictive validity of the CUE-T.