Development of Chinese Card Sorting Test and Its Validity Assessment

Qihao Guo,Qianhua Zhao,Xinyi Cao,Yan Zhou,Zhen Hong

DOI: https://doi.org/10.1016/j.jalz.2010.05.060

2010-01-01

Abstract:Objective To develop Chinese Card Sorting Test (CCST) as an efficient, brief test for evaluating executive function in Chinese elderly. Methods Thirty ⁃ five patients with mild Alzheimer's disease (AD), 69 cases with mild cognitive impairment (MCI) [amnesic MCI⁃single domain (aMCI⁃s), 25; amnesic MCI⁃multiple domain (aMCI⁃m), 24; non⁃amnesic MCI (naMCI), 20] and 124 cognitively normal controls (NC) were administered a battery of neuropsychological tests including CCST. Three sets (including figures, words and numbers) of cards were adopted. Each set had 6 cards. Subjects were asked to sort cards according to shape, number, color and meaning. Full mark was 4 for each set and 12 for total 3 sets. Results 1) Comparison among NC, MCI and AD: according to test score, in NC group, Set Ⅱ was most difficult while Set Ⅰ was the easiest. However, in MCI and AD group, patients' scores were similar among 3 sets, which indicated that the discriminative efficacy was same among 3 sets. Set Ⅰ and Ⅲ sorting started with "color" and followed with "meaning". Set Ⅱ sorting started with "meaning". The accuracy score for Set Ⅱ was the lowest which may due to the inhibition effect of "meaning" against other sorting strategy (similar as "the Stroop effect"). "Number" was the most difficult and least adopted sorting dimension. There were few misclassification which indicated that the test design produced few misunderstanding. When the total score of CCST ≤ 6, the sensitivity and specificity for differentiating AD was 94.32% and 74.24%, respectively. When the total score of CCST ≤ 5, the sensitivity and specificity for differentiating AD was 74.35% and 89.52%, respectively. For subjects with low education, optimal cut⁃ off score should be selected as ≤ 5. For subjects with high education, optimal cut⁃off score should be set as ≤ 6. 2) Comparison among 3 MCI subgroups: the score in naMCI subgroup was the lowest (4.25 ± 2.07), while in aMCI ⁃ s subgroup was the highest (6.68 ± 1.65). The score in aMCI ⁃ m was between the 2 above subgroups (5.16 ± 2.26). There were significant differences among the 3 subgroups (P < 0.01, for all). CCST can be used to differentiate amnesic and non⁃amnesic MCI. There were significant differences between aMCI ⁃ s and naMCI for all 3 sets of cards. Significant differences only existed for Set Ⅱ between aMCI⁃m and naMCI. There were significant differences among 3 subgroups in "color" dimension. There were no differences among 3 subgroups in "number" and "meaning" dimension. There were significant differences between aMCI and naMCI in "color" and "number" dimension. There were no differences between aMCI and naMCI in "shape" and "meaning" dimension. Conclusion Chinese Card Sorting Test is an effective, rapid and usable test for differentiating amnesic and non⁃amnesic MCI. DOI：10.3969/j.issn.1672-6731.2010.02.014