RESUMEN
OBJECTIVE: The present paper was to study the validity of screening dementia among Thai elderly by clock drawing test (CDT). MATERIAL AND METHOD: The scoring method selected to apply with CDT was Chula clock-drawing scoring system (CCSS) that was originally developed as clinically-based in Thai elderly patients. The 669 elderly subjects gathered from "Rom Klao" community in Bangkok, Thailand were asked to perform CDT and be examined by a neurologist, using NINCDS-ADRDA diagnosis criteria for probable Alzheimer's disease (AD). CDT was scored by psychiatrists using CCSS. RESULTS: The authors found the demented by clinical diagnosis in 25 cases. Using a CCSS cutoff score of 7, CDT produced positive test results in 191 subjects. Sensitivity was 88%, the specificity was 74% and the area under receiver operation characteristics (ROC) curve was 0.91. The results also showed that comparatively to cutoff point 7, a cutoff point 6 would contribute the higher specificity of 82% and have a similar sensitivity of 88% in this community-based sample. CONCLUSION: The present study provided strong support that CDT scoring by CCSS is efficient to screen dementia in the general community with satisfactory sensitivity and specificity. However modifying the CCSS cutoff score from 7 to 6 increases the specificity and is proposed to be applied in the community.