Subjects over 65 with DSM-111-R dementia and living in the community (with carers) who were General Health Questionnaire (GHQ) cases were recruited. Subjects were assessed using the Mini Mental State Examination and the Geriatric Mental State (GMS). Carers were interviewed using the GMS or the CIS (Clinical Interview Scale). Information was collected on behavioural disturbance in their relatives using the Present Behavioural Examination; the most distressing behaviours were noted. A multidisciplinary team generated an individualized plan aiming to reduce the most distressing behaviours. Couples were then randomized into groups 1 and 2. For group 1, plans were implemented over 16 weeks and they were then reassessed using the same rating scales as before (phase 1). They were followed up a further 16 weeks later to see if the benefits of intervention were sustained (phase 2). Group 2 were assessed after a period as waiting list controls (phase 1). Their plans were then implemented and they were reassessed (phase 2). There was an improvement in carer mental health and in the behavioural difficulties of the person with dementia for group 1; the two measures of outcome were associated. For group 2, there was no improvement in either outcome measure. The findings are discussed in relation to their relevance for clinical practice. KEY woRDs-behaviour; dementia; carers; intervention; randomised controlled trial About 80% of people with dementing illnesses are estimated as having one or more troublesome behaviours (Swearer et al., 1988;Burns et al., 1990), and the burden of coping with these behavioural problems is falling increasingly on relatives and friends.
There is little literature on the management of behavioural difficulties in dementia. However, a recent metanalysis of 17 controlled trials reported that the effects of medication were modest but consistent and reliable (Schneider et al., 1990). Other work found that a period of hospitalization in