A voice recognition-based digital cognitive screener for dementia detection in the community: Development and validation study

INTRODUCTION: To facilitate community-based dementia screening, we developed a voice recognition-based digital cognitive screener (digital cognitive screener, DCS). This proof-of-concept study aimed to investigate the reliability, validity as well as the feasibility of the DCS among community-dwelling older adults in China. METHODS: Eligible participants completed demographic, clinical, and the DCS. Diagnosis of mild cognitive impairment (MCI) and dementia was made based on the Montreal Cognitive Assessment (MoCA) (MCI: MoCA < 23, dementia: MoCA < 14). Time and venue for test administration were recorded and reported. Internal consistency, test-retest reliability and inter-rater reliability were examined. Receiver operating characteristic (ROC) analyses were conducted to examine the discriminate validity of the DCS in detecting MCI and dementia. RESULTS: A total of 103 participants completed all investigations and were included in the analysis. Administration time of the DCS was between 5.1–7.3 min. No significant difference (p > 0.05) in test scores or administration time was found between 2 assessment settings (polyclinic or community center). The DCS showed good internal consistency (Cronbach’s alpha = 0.73), test-retest reliability (Pearson r = 0.69, p < 0.001) and inter-rater reliability (ICC = 0.84). Area under the curves (AUCs) of the DCS were 0.95 (0.90, 0.99) and 0.77 (0.67, 086) for dementia and MCI detection, respectively. At the optimal cut-off (7/8), the DCS showed excellent sensitivity (100%) and good specificity (80%) for dementia detection. CONCLUSION: The DCS is a feasible, reliable and valid digital dementia screening tool for older adults. The applicability of the DCS in a larger-scale community-based screening stratified by age and education levels warrants further investigation.