Background: Longitudinal cognitive assessments in the community, which can aid in diagnosis of AD, are difficult to achieve due to limited resources of healthcare systems: tests are generally administered by trained staff and are time consuming. Reliable cognitive assessments, self- administered at home, can decrease the burden both within the healthcare system and in clinical trial settings. This study assesses the feasibility of self-administered at home cognitive testing, using a digital device, in healthy controls (HC) and cognitively impaired (CI) patients.
Method: 81 healthy control and 30 cognitively impaired (23 Mild Cognitive Impairment, 7 mild Alzheimer’s dementia) participants, ages 60-76, were tested at home on 9 Cambridge Neuropsychological Test Automated Battery (CANTAB) tasks (4 memory, 2 attention, 2 executive function and 1 emotion recognition) on a provided iPad. Written instructions were distributed prior to first testing session and tasks had integrated computerized instructions, but there was no formal instruction or prior familiarization with tests. Each task was administered two times, across 5 sessions, with each session containing up to 4 tasks; sessions were separated by 2 weeks. Subjects were videoed during task performance by the iPad front camera.
Result: 1. Compliance. Identical proportions of both groups, 67%, had full compliance (each test completed twice); 93% of HCs and 87% of CIs completed each test at least once. 2. Test results. Mann-Whitney tests showed significant differences between groups, with healthy controls performing better than cognitively impaired participants on all but one test. Highest effect sizes, d>1, and significance levels, p<2x10-5, were observed in the memory domain (all results provided in the accompanying table). 3. Test/retest. Highest test-retest values, with an intraclass correlation coefficient (ICC) > 0.71, were observed in 2 memory tests. 4. Subject identity. 33% of videos, randomly chosen, were manually checked; one person used a substitute in one session.
Conclusion: Healthy controls and cognitively impaired patients, ages 60-76, were able to successfully self-administer CANTAB battery tests at home on a personal device, without prior familiarization or oversight by an administrator. CANTAB battery discriminated between healthy controls and cognitively impaired participants with good test-retest reliability.