Usefulness of CogEvo, a computerized cognitive assessment and training tool, for distinguishing patients with mild Alzheimer's disease and mild cognitive impairment from cognitively normal older people

Aim This study aimed to assess whether CogEvo, a computerized cognitive assessment and training tool, could distinguish patients with mild Alzheimer's disease and mild cognitive impairment from cognitively normal older people. Methods This cross‐sectional study enrolled 166 participants with Alzheimer's disease, mild cognitive impairment and cognitively normal older people. In CogEvo, five types of cognitive tasks were carried out, and the z‐scores were used as a composite score. Logistic regression and receiver operating characteristics analyses were then carried out to evaluate the usefulness of CogEvo in distinguishing between the three groups. Results CogEvo and Mini‐Mental State Examination scores showed excellent correlation, and could significantly differentiate between the Alzheimer's disease, mild cognitive impairment and cognitively normal older people groups (Mini‐Mental State Examination 20.4 ± 3.5, 25.5 ± 1.6 and 27.6 ± 2.0, respectively; CogEvo: −1.9 ± 0.9, −0.8 ± 0.8 and 0.0 ± 1.0, respectively; both P < 0.001 by analysis of variance). Logistic regression analysis adjusted for age, sex and years of education significantly differentiated the mild cognitive dysfunction group (mild cognitive impairment plus mild Alzheimer's disease; n = 78) from the cognitively normal group ( n = 88) ( P < 0.001), whereas receiver operating characteristics analysis showed moderate accuracy (area under the receiver operating characteristic curve 0.830). Conclusions These results suggest that CogEvo, a computerized cognitive assessment tool, is useful for evaluating early‐stage cognitive impairment. Further studies are required to assess its effectiveness as a combination assessment and training tool. Geriatr Gerontol Int 2021; 21: 192–196 .