Abstract
Purpose
To compare the test–retest reliability and minimal detectable change (MDC) of the commonly used versions of the Alzheimer's Disease Assessment Scale-Cognitive Subscale (ADAS-Cog) (the ADAS-Cog-11 (11 items), ADAS-Cog-3 (three items), ADAS-Cog-5-Subset (five items), ADAS-Cog-6-Subset (six items), and ADAS-Rasch (11 items)) in people with dementia.
Materials and methods
A repeated-assessments design (2 weeks apart) was used to examine the ADAS-Cog-11, ADAS-Cog-3, ADAS-Cog-5-Subset, ADAS-Cog-6-Subset, and ADAS-Rasch. Participants with dementia were recruited from one hospital, one elder care center, and two day-care centers using convenience sampling.
Results
Fifty-two participants finished the assessments twice in two weeks. All versions showed high intraclass correlation coefficients (ICCs) (0.82–0.96), minimal standardized response means (–0.07 to 0.08) and low to acceptable MDC% (9.2–28.6%). The ADAS-Rasch had the highest ICC (0.96) and the lowest MDC%. The ADAS-Cog-3 had an ICC lower than 0.90 (0.82) and the highest MDC% (28.6%).
Conclusions
The ADAS-Rasch seems to be the most reliable version of the ADAS-Cog for group- and individual-level comparisons. The ADAS-Cog-3 may be a better choice for researchers for group-level comparisons because it requires fewer items to achieve acceptable reliability. The ADAS-Cog-11, ADAS-Cog-5-Subset, ADAS-Cog-6-Subset, and ADAS-Rasch could be considered for clinical usage for individual-level comparisons.
The ADAS-Rasch is the most reliable version of the ADAS-Cog for group- and individual-level comparisons due to its excellent test–retest reliability, lowest random measurement error and absence of a practice effect.
The ADAS-Cog-5-Subset and ADAS-Cog-6-Subset might be good substitutes for the ADAS-Rasch in clinical settings because of their comparable reliability features and superior administration efficiency.
Implications for rehabilitation
Correction Statement
This article has been corrected with minor changes. These changes do not impact the academic content of the article.
Disclosure statement
The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.