Clock drawing test in screening for Alzheimer's dementia and mild cognitive impairment in clinical practice

Objectives: The clock drawing test (CDT) is a commonly used brief cognitive measure. We evaluated diagnostic accuracy of subjective ratings of CDT by physicians (with/without specialty in cognitive neurology) and neuropsychologists in discriminating amnestic mild cognitive impairment (aMCI), Alzheimer's dementia (AD) and cognitively healthy older adults. We further compared the diagnostic accuracy of subjective categorical ratings with complex scoring of CDT. Methods: Three cognitive neurologists, three neuropsychologists and six neurology residents without experience in cognitive neurology blinded to the diagnosis rated 187 CDTs (50 mild AD, 49 aMCI and 88 cognitively healthy older adults) using a “yes” (abnormal) versus “suspected” versus “no” (normal) classification. The rating suspected was combined with yes or no to obtain two sets of sensitivity estimates. We also used a 17-point CDT rating system. Results: When using the categorical rating, neuropsychologists had highest sensitivity (89%) in differentiating patients with mild AD (yes/suspected versus no), followed by neurologic residents (80%) and cognitive neurologists (79%). When differentiating patients with aMCI (yes/suspected versus no), the sensitivity was 84% for neuropsychologists, 64% for cognitive neurologists and 62% for residents. The sensitivity using the complex scoring system was 92% in patients with mild AD and 69% in patients with aMCI. Conclusions: A categorical rating of CDT shows high sensitivity for mild AD even in non-experienced raters. Neuropsychologists outperformed physicians in differentiating patients with aMCI from cognitively healthy older adults (specificity), which was counterbalanced by the lower specificity of their ratings. The diagnostic accuracy was not substantially improved by using complex scoring system.