RESUMO
OBJECTIVE: We developed new subroutines and scoring procedures of the CDT to detect markers of early cognitive impairment. SUBJECTS AND METHODS: We compared the performance of 41 mild Alzheimer's patients (AD) (MMSE 22.7+/-3.2); and 18 mild cognitive impairment patients (MCI) (MMSE 28.3+/-1.4), with 33 age and education matched normal controls (NC) regarding their ability in drawing a clock on command, copying a printed clock, setting hands and reading the time. We search for differences in the placement of the hands using a colloquial hour code ("a quarter to ten") and a formal numeric code ("10:45"). We focused our attention on placement and clock hands size. The drawing of hands pointing at 2:50 hs and 8:40 hs were of particular interest. RESULTS: Planning strategies, placement of clock hands using a formal numeric code, clock hands size and hands drawing pointing at 2:50 hs. and 8:40 hs. were the variables that best discriminated AD, MCI and NC ( p<0.01). A stepwise regression analysis using as dependent variables (AD - NC ) and (MCI - NC) showed in both cases that the best predictor model was that formed by: planning strategies, clock hands pointing to 8:40, clock hands size, reading hours and set up clock hands following a formal numeric code (p<.001). CONCLUSIONS: Our proposed modifications in this test convert it in an appropriate tool for cognitive impairment screening because they do not only detect visuoconstructional abilities but also frontal planning, contents and processing of the hour memory code and transcodification of semantic memory processes.