Background and Objectives:
The World Health Organisation Disability Assessment Schedule 12-item version (WHODAS-2.0) is a measure of disability representing six International Classification of Disability, Functioning, and Health activity and participation domains, including cognition, mobility, self-care, interpersonal functioning, life activities, and participation.  Preliminary validation analyses of the WHODAS-2.0 as an outcome measure for mild traumatic brain injury (MTBI) using classical test theory methods, were promising. The objective of this study was to further examine the psychometric properties of the WHODAS-2.0 for MTBI by applying Rasch Analysis, specifically considering unidimensionality and responsiveness to injury recovery.

Method:
Adults (n = 148) who were recruited from outpatient clinics in Vancouver following MTBI completed the WHODAS-2.0 at on average 10 weeks and then again 20 weeks following injury.  We used Rasch analysis to determine overall and individual item fit to the Rasch model at each time point, differential item functioning (DIF), local independence, unidimensionality and person-separation reliability of the WHODAS 2.0 using RUMM2030 software. Paired t-tests were used to evaluate change in scores with time.

Results:
Participants were on average 40 years old (SD 12.1), predominantly Caucasian (66%), with slightly more women (66%) than men in the sample. The most common injury mechanism was motor vehicle accident (48%).   Best fit to the unidimensional Rasch model at time one was achieved after locally dependent items were combined into three subtests or super-items (Chi square = 3.63, df 6, p = 0.73). The modified WHODAS-2.0 demonstrated high internal consistency (Person Separation Index (PSI) = 0.91) and there was no DIF across person factors examined. Time 2 analyses although exploratory, due to a small (n=80) sample, were also suggestive of good fit to the Rasch model and high internal consistency (Chi square = 22.21, df 22, p = 0.45; PSI = 0.90).  The results support retaining the original response format and content of the measure.  From these analyses we created ordinal-to-interval conversion tables that will allow more precise assessment of MTBI outcome across individuals on an interval scale.  Finally changes in WHODAS-2.0 total raw scores between times one and two were significant (mean difference = 5.59; 95% confidence interval 4.05, 7.14; p<0.01), and person-item threshold distributions from Rasch analysis reflected reducing disability with time in this sample.

Conclusions:
With minor modifications, the WHODAS-2.0 demonstrated adequate properties as a unidimensional scale and showed responsiveness to improvement with time and MTBI recovery. Rasch analyses supported the WHODAS-2.0 as a psychometrically sound measure of outcome in adults after MTBI.

Psychometric evaluation of the World Health Organization Disability Assessment Schedule 2.0 for Mild Traumatic Brain Injury using Rasch Analysis