Psychometric Properties of a Flanker Task in a Sample of Patients with Dementia: A Pilot Study

BACKGROUND/AIMS: Reliable and valid neuropsychological tests for patients with dementia are scarce. To improve the assessment of attention and inhibitory control in dementia, we determined the feasibility, test-retest reliability, and validity of a Flanker task. METHODS: Participants with all-cause diagnosed dementia (n = 22, mean age 84 years; mean Mini-Mental State Examination [MMSE] score = 19.4) performed a computerized Flanker task twice within 7 days. The Flanker task required participants to indicate the direction of target arrows flanked by congruent or incongruent arrows. Number of completed trials, accuracy, and reaction times (RTs) were recorded, and interference scores were calculated from basic scores. We examined the psychometric properties of the Flanker task and its relationship with the MMSE and Stroop test. RESULTS: The Flanker task was feasible. Test-retest reliability was good for number of correct answers and RTs, and fair to poor for accuracy and the interference scores. The correlation of the Flanker task with Stroop and MMSE performance was fair to poor. CONCLUSION: The Flanker task appears to be feasible, and a reliable and valid measure of selective attention. Although the test-retest reliability for the Flanker RT interference measure was fair, future studies need to confirm its validity to measure inhibitory control in patients with dementia.