Diagnostic Accuracy of the Swedish Version of the Rowland Universal Dementia Assessment Scale (RUDAS-S) for Multicultural Cognitive Screening in Swedish Memory Clinics

BACKGROUND: The number of people with a migration background and dementia is increasing in Europe. All patients with suspected dementia have the right to an appropriate cognitive assessment and correct diagnosis for optimal treatment and support. Rowland Universal Dementia Assessment Scale (RUDAS) cognitive screening instrument is less affected by language, culture, and educational background, and adapted for use in multicultural populations. OBJECTIVE: To compare the diagnostic accuracy of RUDAS-S to the Swedish version of Mini-Mental State Examination (MMSE-SR) for detecting dementia in a multicultural group of outpatients in Swedish memory clinics. METHODS: We tested 123 outpatients (36 nonnative Swedish), in 4 memory clinics in Southern Sweden with RUDAS-S to supplement the usual cognitive assessment. RESULTS: RUDAS-S had moderate to good diagnostic performance for detecting dementia in a multicultural population in Sweden, with an area under the receiver operating characteristic curve (AUC) of 0.81. At a cutoff score <25 its sensitivity was 0.92, specificity 0.60, and accuracy 76%. The AUC for the MMSE-SR was 0.79. At a cutoff score <23 its sensitivity was 0.65, specificity 0.81, and accuracy 73%. CONCLUSION: RUDAS-S is at least as accurate as MMSE-SR for detecting dementia in memory clinics in Sweden and can be used for all patients undergoing a cognitive assessment, irrespective of their cultural, language, and educational background. However, there is a need for other cross-cultural cognitive tests to complement RUDAS-S to extend cognitive examination.