Informant Questionnaires in Dedicated Memory Clinics: How Much Do They Contribute?

BACKGROUND/OBJECTIVES The diagnostic process in a university memory clinic is based largely on cognitive testing. However, input from informants, acquired through interview or questionnaires, may significantly impact diagnosis. We sought to evaluate whether informant questionnaires for basic and instrumental activities of daily living, or for identifying progressive cognitive decline would improve diagnostic predictability of neurodegenerative disorders compared with either the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) neuropsychological test battery or the Mini‐Mental State Examination score alone. DESIGN Retrospective data analysis using logit models. SETTING University hospital outpatient memory clinic. PARTICIPANTS A total of 394 patients with dementia, mild cognitive impairment (MCI), depression, or subjective cognitive impairment were assessed. MEASUREMENTS Bristol Activities of Daily Living Scale, Functional Activities Questionnaire, Informant Questionnaire on Cognitive Decline in the Elderly, and the Physical Self‐Maintenance Scale questionnaires were obtained. Analyses through logit models were performed to predict outcome diagnoses, based on cognitive scores alone or in combination with one or more informant questionnaires. RESULTS The four questionnaires were highly correlated (.31–.86). The addition of informant questionnaires improved diagnostic predictability between differential diagnoses of MCI and dementia, or dementia and depression. However, the misprediction rate was reduced by up to 6 percentage points only. Adding more than one questionnaire or all CERAD subtests instead of their sum score never improved prediction in regularized logit models to a clinically relevant extent. CONCLUSION Although questionnaires contribute to a statistically better prediction of the outcome diagnosis, for some sets of differential diagnoses, the benefit may not be clinically pertinent when routine semistructured informant interviews are used by trained personnel. However, standardized assessment, particularly when patients are seen longitudinally, should not be underestimated.