Objectives: There is currently no appropriate cognitive screening test available to diagnose dementia cross-culturally in South Africa. The aim of this pilot study was to investigate the efficacy of an informant questionnaire in detecting cognitive decline in the elderly.
Design: The Deterioration Cognitive Observee (DECO), an informant questionnaire previously used abroad, was administered to relatives of elderly patients. Relatives were also asked a series of open-ended questions about the patient's cognitive abilities and behaviour. The DECO results were compared with patients' scores on the Mini-Mental State Examination (MMSE), the cognitive measure currently used to assess a patient's level of cognitive decline, as well as with the clinicians' diagnosis.
Setting: The interviews were completed at the Groote Schuur Hospital Geriatric Clinic during the months of May and June 1994.
Subjects: The subjects were patients (N = 20) and their relatives (N = 20) attending the Geriatric Clinic.
Results: DECO scores correctly predicted normal functioning in 7 patients and dementia. In 8. The DECO scores correlated with the MMSE scores (r = 0.625; P < 0.01) and MMSE scores correlated with the clinicians' diagnosis (chi 2 = 0.114; df = 1; P = 0.73). Open-ended questions confirmed the clinicians' diagnosis.
Conclusion: The DECO was found to predict dementia correctly in all but the severely demented patients. As the DECO appears to be a suitable alternative to cognitive testing, it should be considered as a possible screening measure for dementia in elderly people in South Africa.