The Clock Drawing Test: performance among elderly with low educational level.

OBJECTIVE To compare the accuracy of four scales, and to determine the inter-rater reliability and the influence of schooling on Clock Drawing Test performance in screening for dementia in elderly with low educational level. METHOD DESIGN Criterion validation, concurrent type. SETTING Outpatient unit of Universidade do Estado do Rio de Janeiro Hospital, Brazil. PARTICIPANTS A total of 306 individuals (> or = 65 years) were assessed by a geriatrician and a neuropsychologist using functional and cognitive instruments. The diagnosis of dementia was established by consensus in a meeting between these professionals. Four different researchers blinded as to the cognitive status of the patients scored the Clock Drawing Test using Shulman's, Sunderland's, Manos', and Wolf-Klein's methods; and a fifth researcher assessed Manos' method for an inter-rater reliability study. RESULTS The data of 211 individuals (< or = 4 years of schooling) were analyzed. The sensitivity and specificity of the four methods were similar (about 65%). The inter-rater reliability of Manos' method was excellent (ICC = 0.944). The areas under the ROC curves were small (0.657 to 0.681), and the differences between them were not statistically significant. CONCLUSIONS The Clock Drawing Test is a reliable but not valid test to screen for dementia in older outpatients with 4 or less years of formal schooling.