Test-retest reliability of the computer-based cognitive assessment tool for community-dwelling older adults in Japan:...

Early detection of dementia is crucial for effective intervention. The computer-based cognitive assessment tool (CompBased-CAT) was designed to assess cognitive function using a tablet computer. While its predictive validity for mild cognitive impairment (MCI) is established, its test-retest reliability remains unclear. This study aimed to evaluate the test-retest reliability of CompBased-CAT among older Japanese adults in a community setting.

This is a methodological study that examined the test-retest reliability of CompBased-CAT. Community-dwelling older adults aged 65 years or older who participated in both the pre-test and post-test were included. Pre-test assessments were conducted during the 2023 Otassha study from 28 September to 8 October 2023, with the post-test administered 62 days (±14 days) later. Subtest scores were normalized to z-scores, and the total CompBased-CAT score was computed by summing these z-scores. Test-retest reliability was assessed using paired t-tests and intraclass correlation coefficient (ICC) two-way random-effects models, with analyses stratified by age, sex, and MCI status.

A total of 80 participants (mean age: 76.8 years; 27.5% male) were analyzed. Paired t-tests showed no significant difference in total scores between the pre-test (mean = -0.75; standard deviation [SD] = 3.73) and the post-test (mean = -0.42; SD = 4.45). The total score exhibited an ICC of 0.64 (95% confidence interval [CI] = 0.49-0.75) for all participants, increasing to 0.74 (95% CI = 0.48-0.88) among those with MCI.

CompBased-CAT demonstrated good test-retest reliability, with an ICC of 0.64 among all participants, which increased to 0.74 among participants with MCI over a 2-month period, indicating its potential for monitoring cognitive function through repeated assessments.