Uncovering Common Rater Errors In Cognitive Assessments For Alzheimer's Clinical Trials

By Marcela Roy, MA; Sayaka Machizawa, PsyD.; Marta Pereira, PhD.; & David S. Miller, MD, MA

Cognitive assessments are foundational to Alzheimer’s disease (AD) clinical trials, with the Mini-Mental State Examination (MMSE) playing a central role in evaluating cognitive function across key domains. Despite its widespread use, the MMSE is highly susceptible to rater variability, leading to significant administration and scoring inconsistencies that can compromise data integrity. A review of 10,203 MMSE assessments from two large Phase 3 trials revealed error rates exceeding 26%, with frequent issues in orientation and attention subtests. These discrepancies, including inconsistent prompts and unauthorized cues, can introduce substantial noise into clinical data.

To address these challenges, centralized data surveillance and targeted rater training are critical. Signant Health’s integrated approach—combining real-world error data with rigorous training and ongoing remediation—supports accurate, standardized administration across global sites. By enhancing reliability in cognitive outcome measures like MMSE and ADAS-Cog, these strategies are vital to improving trial validity and advancing effective AD therapies.