Empirical evidence of cognitive change: Calculating reliable change indices for digital tests of cognition in individuals living with Alzheimer’s disease

Abstract

AbstractBackgroundThe identification of individuals at risk of progressing from normal health to Mild Cognitive Impairment (MCI) and later to Alzheimer’s disease (AD) has focused recently on using evidence of progressive decline. This approach relies upon employing a reliable metric of cognitive change. One possible approach to this issue involves employing a Reliable Change Index (RCI) methodologyMethodA sample of N = 56 participants were assessed on exams from the Neurotrack assessment suite at baseline and 1‐week later. This assessment is composed of well‐known and previously validated measures of episodic memory, working memory, attention and executive function (see Table 1 for details of tests).ResultsUsing the data from our study cohort we calculated a RCI for each measure using the Jacobson‐Truax method. This requires the determination of a threshold that separates those who are impaired from individuals who are not impaired. It also requires a comparison of an individual’s change from assessment time point 1 and 2 to the standard error of measurement (SEM) of the outcome. RCI and 95% confidence interval (CI) was calculated for each of the Neurotrack tests. By applying the 95% CI to a baseline score, a statistically normal range of later performance can be determined. Values falling outside the CI can be considered evidence of real change.ConclusionsIn this study, we provided individual RCI values and associated CI for each test in the Neurotrack battery. We are currently evaluating the Neurotrack assessment in a longitudinal validation study to investigate sensitivity to change over time.