Identifying older people with dementia: the effectiveness of a multiservice census.

A census of all relevant services in an area can be used to identify people with mental impairment suggestive of dementia. Two censuses in Tayside, Scotland, were used to test the effectiveness of this method. False positives accounted for 12% of returns. After excluding false positives, by comparison with expected dementia prevalence based on EURODEM, 66% of all sufferers and 50% of those living in the community were identified by the censuses. By pro-rating for non-response, the proportion of sufferers known to services was estimated as 72%. The characteristics of those not known to services are unclear and further research is needed on this. The cost of a census in an area of 250,000 population is under pounds 3000. A multiservice census offers a simple, inexpensive, practicable method of constructing a sample frame for population needs assessment.