Evidence for validity of the Swedish self-rated 36-item version of the World Health Organization Disability Assessment Schedule 2.0 (WHODAS 2.0) in patients with mental disorders: a multi-centre cross-sectional study using Rasch analysis.

Background: The World Health Organization Disability Assessment Schedule 2.0 (WHODAS 2.0) is a generic instrument for the assessment of functioning in six domains, resulting in a total health-related disability score. The aim of this study was to investigate the psychometric properties of the Swedish-language version of the self-rated 36-item version in psychiatric outpatients with various common psychiatric diagnoses using Rasch analysis. A secondary aim was to explore the correlation between two methods of calculating overall scores to guide clinical practice: the WHODAS simple (summative) model and the WHODAS complex (weighted) model.

Methods: Cross-sectional data from 780 Swedish patients with various mental disorders were evaluated by Rasch analysis according to the partial credit model. Bivariate Pearson correlations between the two methods of calculating overall scores were explored.

Results: Of the 36 items, 97% (35 items) were within the recommended range of infit mean square; only item D4.5 (Sexual activities) indicated misfit (infit mean square 1.54 logits). Rating scale analysis showed a short distance between severity levels and disordered thresholds. The two methods of calculating overall scores were highly correlated (0.89-0.99).

Conclusions: The self-administered WHODAS 2.0 fulfilled several aspects of validity according to Rasch analysis and has the potential to be a useful tool for the assessment of functioning in psychiatric outpatients. The internal structure of the instrument was satisfactorily valid and reliable at the level of the total score but demonstrated problems at the domain level. We suggest rephrasing the item Sexual activities and revising the rating scale categories. The WHODAS simple model is easier to use in clinical practice and our results indicate that it can differentiate function among patients with moderate psychiatric disability, whereas Rasch scaled scores are psychometrically more precise even at low disability levels. Further investigations of different scoring models are warranted.