GOS-E Properties

Reliability:
Based on a sample of 50 neurosurgical patients, using 2 independent raters, inter-rater agreement was 78% using the GOS-E structured interview (Wilson et al., 1998). The weighted kappa coefficient was .85, which is nearly perfect.

Validity:
In a sample of 135 neurosurgical patients assessed at 6 months post-injury, outcome category as assessed by the GOS-E was related to initial injury severity, to outcome as assessed by the Disability Rating Scale, and to scores on self-report measures of health outcome, including the Beck Depression Inventory, the General Health Questionnaire, and the SF-36 (Wilson et al., 2001). Outcome as assessed by the GOS-E was also associated with frequency of symptoms reported on the Neurobehavioral Functioning Inventory (Wilson et al., 2001). In a sample of 87 patients with mild TBI, moderate TBI, or general trauma, Levin and colleagues (2001) reported that the criterion validity of the GOS-E (as determined by association with functional outcome measures and neuropsychological performance) was superior to that of the original GOS. They also found that the GOS-E was more sensitive than the GOS to changes in functioning from the 3-month to the 6-month follow-up.