Search

A 10% sample of private households on the electoral register of the London borough of Lambeth was screened for disable persons aged 16 and over, using a postal questionnaire. After three mailings and individual follow-up of non-responders, 87% of the sample households returned questionnaires. Disability was defined in the screening questionnaire as functional limitations or activity restrictions consequent upon disease or impairment. The overall point prevalence of disability was estimated at 15.4% and the most frequently reported impairments were those of the sense organs, bones, central nervous, circulatory, and respiratory systems. Hearing difficulties were the single most frequently reported functional limitation. A log-linear modelling procedure identified age, marital status, and working status as the factors most strongly associated with disability for both men and women. In addition, men aged 50-64 and not working, and men in manual occupations and living alone, were more likely to report disability. These findings indicate that some population groups are disable by functional limitations and activity restrictions not included in office criteria of identification and assessment. These criteria might be broadened, and serves planned for those population groups with higher rates of reported disability.

An estimate of the prevalence of physical disability in the community based upon a sample survey may be influenced by the sample design and the response to the method of data collection employed. In this paper we describe a postal survey of a sample of households in the London borough of Lambeth and the procedures used for calculating the influence of these factors on the estimate produced. These procedures can be used to adjust the estimate to take account of the relative chance of households falling into the sample and to correct for non-response bias.

The 30-item General Health Questionnaire (GHQ) has been validated against the criterion of a standardized psychiatric interview in the community. Used for screening, the GHQ misclassified 25 per cent of 105 respondents, being worse for men. Specificity and sensitivity were above .70. The alternative use of the GHQ as an estimator and as an indicator of morbidity in epidemiological surveys is described and discussed.