A psychometric evaluation of the free and cued selective reminding test: Immediate recall (FCSRT-IR) in a phase III prodromal Alzheimer's disease study

Background: Impaired memory is an early manifestation of Alzheimer's disease (AD), thus reliable and valid assessment of memory is critical. The FCSRT-IR assesses recall under controlled conditions. Performance has been associated with early dementia and the FCSRT is now commonly used as a screening measure for prodromal AD clinical trials. The current analyses provide psychometric data for the FCSRT-IR. Methods: Data from 773 subjects (128 sites, 24 countries) enrolled in an ongoing phase III prodromal AD clinical trial were used in these analyses. Screening and baseline (M duration = 51 days) FCSRT-IR, MMSE and ADAS-Cog data were collected, entered, and reviewed as part of an in-study data monitoring program managed by inVentiv Health Clinical. Analyses of the collected FCSRT-IR scores assessed: descriptive statistics (M, SD) for free, cued, and total recall scores for the screening and baseline visits, changes across the screening and baseline visits for the FCSRT-IR (dependent samples ttest), test-retest reliability (Pearson r), and construct validity with MMSE recall and ADAS-Cog word and delayed word recall (Pearson r). Results: Dependent samples t-tests showed significant differences in FCSRT-IR scores from screening to baseline, although the effect sizes were extremely small indicating no meaningful difference. Test-retest reliability coefficients were strong for the total free (0.76) and total recall (0.83) scores, and were slightly lower for total cued recall (0.69). FCSRT-IR total free recall scores were minimally to moderately correlated with MMSE recall, ADAS-Cog word recall, and ADAS-Cog delayed word recall scores. Conclusions: The results support the test-retest reliability of the FCSRT-IR over a 7 week period, particularly the total free and total recall scores, which are the scores typically used for study inclusion criteria. FCSRT-IR scores were not strongly correlated with MMSE recall and ADAS-Cog word recall/delayed scores, potentially suggesting that the FCSRT-IR measures unique memory processes to identify memory impairment that is not secondary to other cognitive deficits.