Is the Cambridge Cognitive Examination - Revised a good tool for detection of dementia in illiterate Brazilian older adults?

Abstract

Few studies have been published on the use of the Cambridge Cognitive Examination Test - Revised (CAMCOG-R) for cognitive assessment of low educational level older adults. The aim of the present study was to determine the accuracy of the Brazilian version of the CAMCOG-R (Br-CAMCOG-R) within a sample of low educational level and illiterate older adults. The Br-CAMCOG-R was administered to outpatients in a public geriatric clinic. The diagnosis of dementia was based on the Diagnostic and Statistical Manual of Mental Disorders Fourth Edition criteria. The receiving operator characteristic curves were plotted, and the best trade-offs between sensitivities and specificities were calculated. A total of 189 participants were evaluated. The mean age was 77 ± 6.9 years. The mean educational level was 3.1 ± 2.2 years. The mean test score was 66.5 ± 13.1 points; there were 56 (29.6%) participants with dementia. The best cut-off score for illiterate participants was 50/51; sensitivity, specificity and area under the curve (AUC) were 69%, 69% and 0.75, respectively; for participants with a low educational level, the best cut-off point was 60/61; the sensitivity, specificity and AUC were 83%, 85%, and 0.93, respectively; for participants with a middle educational level, the best cut-off point was 69/70; the sensitivity, specificity and AUC were 90%, 76% and 0.91, respectively. The Br-CAMCOG-R was useful for identifying cases of dementia among older adults with middle and low levels of literacy, but inadequate for the illiterate individuals.