Abstract
Background: Although the Korean-Mini Mental Status Examination (K-MMSE) has been widely used for screening dementia in clinics and other community settings, there has been little investigation into the validity of the K-MMSE. The present study was conducted to examine the validity of the K-MMSE as a dementia screening tool compared against a comprehensive neuropsychological evaluation. Methods: The subjects were 707 outpatients $(261\;men, 446 women;\;mean\;age=70.3{\pm}6.3\;yr,\;mean\;education=\;6.3{\pm}5.1\;yr)$ who visited the neurology or psychiatry department with complaints of cognitive dysfunction. They were given the K-MMSE and a comprehensive neuropsychological test battery (Seoul Neuropsychological Screening Battery, SNSB). The subjects were classified into "Cognitive Impairment" and "No Cognitive Impairment" groups based on the SNSB as well as the K-MMSE. Results: The agreement between classification based on the K-MMSE and the SNSB was examined. The sensitivity of K-MMSE for detecting cognitive impairments was 80.3% and the specificity was 70.3% (Positive predictive value 90.4%; Negative predictive value 50.7%). The false-positive rate was found to be higher for the subjects with low education. The majority of the subjects in the false-negative group showed prominent frontal/executive dysfunctions as well as memory impairment. Conclusions: These results showed that K-MMSE has high positive- and negative-false rates and a low negative predictive power. They suggest that the K-MMSE should be given with memory and frontalexecutive function tests for improving screening accuracy for dementia.