Measuring disability, the agreement between self evaluation and observation of performance

In 1981, eight countries and the World Health Organization (WHO) reached consensus about applying a similar disability questionnaire in their health surveys that was related to the ICIDH. In 1992 a revised version, the WHO-disability questionnaire, was recommended. In this paper the concurrent validity of 10 items of the questionnaire is evaluated by measuring the agreement with observed performance of corresponding tasks.

The disability questionnaire was sent to 750 subjects. From the respondents (468) 93 were selected, based on their reported disability, and 10 performance tasks were observed. The agreement was measured for individual items and for all items together in distinguishing between persons with no, or at least one disability.

A low to moderate agreement was found between self reported and observed disability. In general, observation resulted in lower disability ratings. Among 24% of those who reported at least one disability, no disability was observed. Persons among whom at least one disability was observed, only 2% (n = 1) reported no disability in the questionnaire.

Although the concurrent validity is low to moderate between the results of the WHO-disability questionnaire and observation, the questionnaire is highly sensitive for detecting persons with at least one disability. The specificity of the questionnaire must be increased by additional questions.