Assessment of potential measures in models of progression in Alzheimer disease
Background: In estimating the potential benefits of treatment, it is often necessary to extrapolate beyond clinical trial results using economic modeling. Previous attempts in Alzheimer disease (AD) were primarily based on the Mini-Mental State Examination (MMSE) due to its widespread use. These models were criticized as not accurately reflecting the total impact of the disease, providing untrustworthy estimates of treatment benefit. We compared 3 alternatives to the MMSE with respect to bridging between clinical outcomes needed for regulatory approval and economic and quality of life (QOL) outcomes important to reimbursement agencies. Methods: The MMSE, Disability Assessment in Dementia (DAD) scale, Clinical Dementia Rating (CDR) scale, and Dependence Scale (DS) were compared in their ability to explain variation in cognitive, functional, and behavioral measures as well as economic and QOL outcomes using univariate (Pearson correlations) and multivariate (linear regression) analyses of data from research sites in the United States and Europe. Results: Subjects with mild to moderate AD (n = 196; mean 75.9 years; 56% female) were evaluated. The DS, DAD, and CDR were moderately correlated with the MMSE (Pearson correlations, range 0.54-0.58) but performed better (higher adjusted R) than the MMSE in explaining variations in subject behavior, QOL, and health status. The DS and DAD performed better in explaining variation in medical costs, caregiver QOL, and caregiver time. Conclusions: Measures of function (DAD) or dependence on others (DS), or global measures (CDR), appear to be better candidates than the MMSE for modeling AD progression.
|Persistent URL||dx.doi.org/10.1212/WNL.0b013e3181f6133d, hdl.handle.net/1765/60307|
McLaughlin, T, Buxton, M, Mittendorf, T, Redekop, W.K, Mucha, L, Darba, J, … Leibman, S. (2010). Assessment of potential measures in models of progression in Alzheimer disease. Neurology, 75(14), 1256–1262. doi:10.1212/WNL.0b013e3181f6133d