Comparison of the Validity of Screening Tests for Dementia and Mild Cognitive Impairment of the Elderly in a Community：K-MMSE, MMSE-K, MMSE-KC, and K-HDS

Abstract

Objectives: The authors have evaluated the validity of four currently used screening tools of cognitive function in Korea for dementia and mild cognitive impairment in a community elderly population. Methods: We screened 5,085 non-randomly sampled elderly over age 60 in the Gwangju city, Gyeonggi Province. Among those, 927 persons were examined further for evidence of dementia and mild cognitive impairment (MCI). A baseline questionnaire on sociodemographic characteristics and four major screening tests (K-MMSE, MMSE-K, MMSE-KC, and K-HDS) were administered. For the evaluation of validity, we have analyzed the ROC curve and calculated the sensitivity and specificity of each test at its optimal cutoff points. Results: ROC analysis of the 3 MMSEs and K-HDS has shown that area under the curve (AUC) was between 0.897 and 0.923 for dementia. Sensitivity and specificity were 0.768 and 0.870, respectively, when the cutoff line was set at 17/18 in K-MMSE. The other two MMSEs had optimal cutoff lines at 16/17 with sensitivity of 0.768 and specificity more than 0.90. K-HDS had its optimal cutoff line at 18/19 with higher sensitivity (0.874) and lower specificity (0.755). Compared with dementia, the four screening tests showed poor validity on MCI. Sensitivity and specificity was in the range of 0.733-0.756 and 0.840-0.855, respectively. Conclusion: The validity of the screening tools which are currently being used in Korea to screen for MCI was not significantly different among tests. Because of the difference in the population norm and z-score estimation, use of the common norm is advisable.