Diagnostic accuracy of the clock drawing test: the relevance of "time setting'' in screening for dementia.

Objective: We aimed to evaluate the overall diagnostic accuracy of clock drawing test methods which require a defined time setting. In addition, diagnostic relevance of time setting was analyzed qualitatively.

Methods: Clock drawing test performance of 462 consecutive memory clinic patients were analysed by 5 different clock drawing test methods. The sensitivity, specificity, positive and negative predictive value as well as the overall misclassification rate was calculated against a clinical diagnosis of dementia. Further, qualitative analysis of error types was done when subjects failed in the time setting task.

Results: All clock drawing test methods that require time setting revealed higher sensitivities and superior negative predictive values as well as overall misclassification rates compared to methods which do not. Failure in the time setting task did not exclusively occur in dementia patients. However, error types which reflect a severely reduced capacity in abstract and conceptual thinking occurred exclusively in dementia patients.

Discussion: Clock drawing test methods that require time setting should be favoured above methods which do not. An additional analysis of time setting errors may yield valuable diagnostic information.