The WHODAS II is a valid, generally reliable and useful instrument for the assessment of stroke patients over the first year after stroke.

Information about resources for support after stroke is unsatisfactory and should be established at a point in the delivery of care chain where all those affected by stroke and their caregivers are accessed.

The WHODAS II (World Health Organization Disability Assessment Schedule II) is a questionnaire derived from the ICF classification system (International Classification of Functioning, Disability and Health) for the evaluation of disabilities and handicaps. Data on its validity and reliability with respect to the consequences of stroke are largely lacking. The present study aimed at measurements of reliability of the WHODAS II in its application to stroke patients and their closest others. Eighty-four patient-relative pairs were assessed six months after stroke with the self- and observer-rating versions. From the patients' and relatives' judgements, Cronbach's alpha was computed as measurement of internal consistency, intra-class correlation coefficients as measurements of the inter-rater reliability of subscales and total scores, and Spearman's rho (rho) for the inter-rater reliability for single items. In addition, the inter-rater correspondence was calculated as the percentage of responses. The internal consistency was found to be good to excellent (alpha=0.81-0.99) both for patients' and relatives' judgements. Inter-rater reliabilities ranged from satisfactory for the subscale Understanding and Communication to excellent for the total score (ICC 0.64-0.94). However, the inter-rater reliability of the items as well as their percental correspondence was hardly satisfactory, with few exceptions in the realms of Life Activities and Self-Care (rho=0.24-0.90 and percental correspondence 29.6%-75.7%). Our results demonstrate that the WHODAS II is a reliable instrument for the assessment of stroke patients, both as a self- and an observer-rating questionnaire. Whereas the correspondence between patient and relative may be disparate with respect to single items - especially those not directly observable -, subscale and total scores seem to allow predictions based on observer judgements. The assessment of stroke patients' disabilities and handicaps through relatives' judgements with the WHODAS II scales is reliable.