The high support needs of elderly people with cognitive disability raise questions about the cost-effectiveness of different treatments. Associations between costs and cognitive disability could be influenced by other factors, particularly comorbidities.

To examine the links between costs and cognitive disability in the context of covariates.

Secondary analyses of data from the UK Office of Population Censuses and Surveys disability surveys for over 4500 elderly people living in households were used to examine associations between cost and cognitive disability.

Costs varied considerably, and were associated with severity of disability along a number of dimensions. The cost-raising effects of cognitive disability were smaller when the analyses controlled for levels of disability in other domains.

Cognitive disability is significantly associated with higher costs, but these analyses highlight the need to examine a range of disabilities.

Population ageing and the high costs of care support for elderly people have concentrated attention on economic issues. Is there an association between costs and cognitive disability?

To compare service utilisation and direct costs for elderly people with different degrees of cognitive disability, and between people living in households and in communal establishments.

Secondary analysis of Office of Population Censuses and Surveys (OPCS) Disability Surveys data compared service utilisation and costs for 8736 elderly people with cognitive disability. Cost estimates were constructed for all health and social care services.

A much greater proportion of people at higher levels of cognitive disability lived in communal establishments, where their (direct) costs were much higher than when supported in households. Service utilisation patterns and costs varied with cognitive disability.

It is important to look at the full range of living arrangements and support services when examining costs. The potential cost implications of pharmacotherapies, other treatments or new care arrangements cannot be appreciated without such a broad perspective.

Long-stay patients with learning disabilities (n=214) were assessed in hospital and 12 and 24 months after discharge in order to examine the effects of relocation.

Each resident acted as his/her own control in a prospective repeated-measures design. Skills and behavioural problems were assessed by keyworkers. Self-perceived quality of life was obtained during interviews with researchers who also completed an environmental checklist of the residents' accommodation.

There was little or no change in peoples low pre-discharge skill levels. Certain aspects of problem behaviour improved after 12 months, although socially unacceptable behaviour increased slightly. People were less depressed (P ⩽ 0.01) 12 months after discharge (n=119) and were more satisfied (P ⩽ 0.05) with their new ‘homes’ (n=108). There were few changes in the pattern of activities or the social networks of people 12 months later. Little or no further change in outcomes was reported 24 months after discharge.

The implementation of the deinstitutionalisation policy in Northern Ireland has been limited by the predominance of residential and nursing homes and the lack of ‘ordinary’ accommodation. There is a need for purchasers and providers to give more attention to ways in which the principles of normalisation could be incorporated in the process of contracting and delivering services.