Abstract

Background: Mobile digital neurocognitive assessment has the potential to facilitate mass screening for MCI and dementia in vulnerable populations by delivering accessible, accurate, standardized measurement of cognition at a vastly reduced administration cost.
Method: 11 subtests of Savonix, a mobile/tablet-based neurocognitive assessment system, were administered to elderly people in hospital and community sites in China in order to assess their applicability in an elderly Chinese population: 29 Mild Cognitive Impairment (MCI) patients, 21 mild Alzheimer’s disease (AD) patients, and 265 neuro-normal elderly people. The diagnostic validity in detecting MCI and AD cases from the neuro-normal cases was assessed with Receiver Operating Characteristics (ROC) curve analyses. The convergent validity was assessed by correlating to traditional paper-and-pen screenings such as Mini- Mental State Examination (MMSE), the Alzheimer Disease Assessment Scale Cognitive subscale (ADASCog), and Clinical Dementia Rating (CDR).
Results: Diagnostic accuracies for all 11 tests were established for AD vs. neuro-normal cases (sensitivity 100%, specificity 95%), MCI vs. neuro-normal cases (sensitivity 85.7%, specificity 93.3%), and AD vs. MCI cases (sensitivity 81%, specificity 85.7%). Fewer numbers of tests also showed comparable accuracies to overall 11 tests. However, the three different total scores demonstrated only weak to moderate correlation to MMSE, ADAS-Cog, and CDRSB for the MCI and AD subjects.
Conclusion: Savonix demonstrated sensitive diagnostic performance in detecting both MCI and AD for Chinese population even when a relatively small number of tests are employed making it suitable for mass screening.