Convergent and criterion validity of a computer adaptive self‐administered word list memory test and the Mayo Test Drive composite: correlations with traditional measures and group difference by PET imaging biomarker status in persons without dementia.

Abstract

AbstractBackgroundMayo Test Drive (MTD): Test Development through Rapid Iteration, Validation and Expansion, is a web‐based platform optimized for remote self‐administered assessment that includes a computer adaptive word list memory test (Stricker Learning Span; SLS) and a measure of processing speed (Symbols Test). Study aims were to determine 1) convergent validity of MTD through correlations with in‐person neuropsychological measures and 2) criterion validity of MTD by qualitatively comparing the ability of Mayo Test Drive and in‐person administered neuropsychological measures to differentiate biomarker‐defined groups, independent of clinical diagnosis.MethodsMayo Clinic Study of Aging participants (N = 282) completed a brief remote cognitive assessment (mean age = 74, SD = 12; 262 CU and 20 MCI per consensus conference diagnosis; see Table 1). A subset of participants had brain amyloid and/or brain tau PET scans available within 3 years; overlapping groups were formed for 1) those on the Alzheimer’s disease (AD) continuum (A+, n = 31) or not (A‐, n = 57) and for 2) those with biological AD (A+T+, n = 19) or without any AD biomarkers (A‐T‐, n = 47). Primary outcome variables were SLS sum of trials, AVLT sum of trials, SYM Correct Trials response time, Digit Symbol Coding, MTD composite and Mayo‐PACC (AVLT sum of trials, Trails B, animals). Bivariate correlations were performed across all participants. Linear model ANOVAs were used to investigate biomarker subgroup differences; Hedge’s g effect sizes are reported.ResultsConvergent validity was supported through significant correlation (p’s < .001) of SLS and AVLT (r = .63), SYM and Coding (r = ‐.64), and MTD composite with Mayo‐PACC (r = .67); see Table 2. MTD subtests showed expected relationships with age and education (p’s < .05; Table 3). All MTD performances were significantly lower for A+ and A+T+ participants relative to A‐ or A‐T‐ (p’s<.001; Table 4). All MTD outcome measures showed a similar or greater group difference effect size relative to traditional in‐person administered neuropsychological measures for separating A+ vs. A‐ and A+T+ vs. A‐T‐ groups. See Figure 1 and Table 4.ConclusionsMTD subtests show evidence of convergent validity through strong correlations with traditional in‐person cognitive measures. Preliminary analyses suggest MTD has promise for differentiating PET biomarker groups.