The validity of the Rowland Universal Dementia Assessment Scale (RUDAS) in a multicultural cohort of community-dwelling older persons with early dementia.

The 6-item Rowland Universal Dementia Assessment Scale (RUDAS) is a simple, portable multicultural scale for detecting dementia. Items address executive function, praxis, gnosis, recent memory, and category fluency. It can be directly translated to other languages, without the need to change the structure or the format of any item. The RUDAS was administered to 151 consecutive, consenting, culturally diverse community-dwelling subjects of mean age 77 years, 72% of whom had an informant. Subjects were recruited from various clinics and healthcare programs. All were evaluated for cognitive impairment in a blinded manner by experienced clinicians in geriatric medicine. According to Diagnostic and Statistical Manual of Mental Disorder-IV criteria, 40% of the subjects were normal, 22% had cognitive impairment (not otherwise specified), and 38% had dementia; 84% of whom had questionable or mild dementia. In the primary analysis (normal subjects vs. those with definite dementia), the RUDAS accurately identified dementia, with an area under the receiver operating characteristic curve of 0.94 (95% confidence interval, 0.88-0.97); at the published cut point of less than 23/30, the positive likelihood ratio (LR) for dementia diagnosis was 8.77, and the negative likelihood ratio was 0.14. Additional analyses showed that the RUDAS performed less well when subjects with cognitive impairment (not dementia) were included. In all logistic regression models, the RUDAS was an independent predictor of dementia (odds ratio 0.64, 95% confidence interval, 0.52-0.79, primary analysis model), after adjusting for age, sex, years of education, and cultural diversity, none of which were independent predictors. Further studies are needed across the full spectrum of early dementia syndromes, and in additional ethnic minority groups.