Abstract
BACKGROUND AND OBJECTIVES: Rapid and valid cognitive screening tools are a mainstay of clinical practice in the assessment of mild cognitive impairment and dementia. Although there are several available tools, there is concern for reduced sensitivity over time. This study evaluates the Duke Iterative Cognitive Examination (DICE) to establish concurrent validity with the Montreal Cognitive Assessment (MoCA) and to establish the clinical utility of the DICE in differentiating clinical groups. METHODS: Participants were enrolled prospectively and included a convenience sample of 126 healthy older adults, 59 individuals with mild cognitive impairment (MCI), and 66 with dementia, all age 55+. All participants were administered the MoCA and the DICE in counterbalanced order. Spearman rank-order correlation examined the relationship between DICE and MoCA scores. Likelihood ratio (L-R) chi-square tests assessed the predictive value of cognitive scores, age, sex, and education in diagnostic classification. Receiver operating characteristic curve analyses evaluated classification accuracy for both tests. RESULTS: The correlation between the DICE and MoCA was strong, suggesting concurrent validity (ρ = 0.89, p < 0.001). Both measures significantly predicted diagnostic groups and demonstrated strong classification accuracy (MoCA AUC = 0.97 (controls), 0.85 (MCI), 0.95 (dementia); DICE AUC = 0.97, 0.94, 0.83, respectively). DISCUSSION: The DICE performed comparably to the MoCA in the current sample of older healthy adults and clinical groups with MCI and dementia. Findings suggest that the DICE is a reasonable and valid screening measure, and future studies should examine other psychometric aspects of the test, such as test-retest reliability, latent factor structure, and longitudinal clinical prediction.