Optimizing mobile cognitive assessment reduces administration time while maintaining screening accuracy in older adults

Growing dementia prevalence underscores the need for efficient screening methods, but lengthy digital assessments often cause fatigue among older adults. To address this, we developed and validated the Digital Assessment of Cognitive Impairment (DACI), a brief mobile application designed to accurately identify cognitive impairment (CI). Initially, 304 older adults (272 healthy, 32 cognitively impaired) completed both a pencil-and-paper Cognitive Impairment Screening Test (CIST) and a full-length DACI. The best-performing CatBoost model achieved an area under the curve (AUC) of 0.813, with sensitivity of 0.903, requiring an average completion time of 321 s. Subsequent feature selection identified two essential subtests for a compact DACI. This compact version was then validated with an additional 297 participants (227 healthy, 70 cognitively impaired), achieving improved diagnostic performance (AUC = 0.871) in only 91 s. DACI effectively distinguishes cognitively impaired older adults with minimal test duration, potentially reducing fatigue and improving adherence. By providing rapid, accurate remote assessment without additional equipment, DACI may significantly enhance accessibility of cognitive screening among older adults. Future studies are warranted to validate DACI's feasibility and performance in unsupervised home settings.