P4‐348: Self‐reported memory fails to substitute for objective memory measures

Abstract

Background: Cognitive assessment of geriatric patients is now widely recognized as necessary to improve safety and medication compliance. Physicians prefer cognitive assessment methods that are fast and inexpensive. Doctors sometimes just ask patients to assess their own cognition. How accurate are self-reports of cognitive abilities when compared to objective measures obtained by computerized testing? Methods: We examined the relationships between objective memory and executive function measures and three subjective patient questions concerning mental clarity, concentration and memory. 7778 subjects were tested in US medical facilities (6/22/2011-2/15/2012) with the CANSMCI, a computer-administered cognitive examination designed for the earliest detection of MCI. The sample included normal, mildly impaired and demented patients: Age=67.9 (SD=15), 59% female. The median test time was 34 minutes, including answers to depression, cognitive ability, pain and head injury questions. In a subset, CANS-MCI measures were also validated against more traditional measures administered within an independent full neuropsychological evaluation. Results: Patients self-reporting problems with mental clarity and concentration actually performed better on immediate, delayed and guided recall (all p 25). Regression found the most powerful predictor of self-reported cognitive symptoms was age, with younger people unexpectedly reporting more symptoms than 65+ patients. Analyses restricted to age > 65 found that those patients who report problems with their mental clarity and concentration perform better on objective memory tests than do patients who do not report those problems. In all age groups, subjective memory symptoms were not significantly correlated with objective memory measures. In the > 65 cohort, the strongest predictors of self-reported cognitive symptoms were depression, pain, and head injury history. In the validation subset (N=169), the CANS-MCI factor scores were as accurate at detecting impairment as full neuropsychological examinations when DRS2, WMS1 and WMS2 are considered the criterion standards. Conclusions: Self-reported cognitive problems do not match objective memory measures and may deceive medical professionals, since age, depression and pain all affect cognitive complaints. While the utility of caregiver/family assessment is well-established, asking patients to self-identify memory problems is not. The failure to obtain positive results in this very large data set suggests that clinicians using subjective patient memory concerns as substitutes for objective testing are often misguided. The CANS-MCI is a valid screening measure to determine longitudinal changes and the need for further assessment. *Research Director, Screen, Inc. **Director of Clinical Services, Screen, Inc. ***Founder and CEO, Screen, Inc.