Identifying amnestic mild cognitive impairment in primary care: a feasibility study.

BACKGROUND AND OBJECTIVE: Amnestic mild cognitive impairment (aMCI), characterized by episodic memory impairment in the absence of clinical dementia, often represents a transitional stage between normal aging and Alzheimer's disease (AD). It is not known if non-expert primary-care physicians (PCPs) can differentiate individuals with no cognitive impairment (NCI), aMCI and mild AD in a primary-care practice setting. This study develops an approach to this question, which is necessary for aMCI to become a treatment target.

METHODS: Fourteen experts assessed subjects with memory complaints in terms of their laboratory test results, magnetic resonance imaging findings and scores on the Mini-Mental State Examination, adapted Clinical Dementia Rating Scale and Alzheimer's Disease Assessment Scale-cognitive subscale Delayed Word Recall before designating each subject as having NCI, aMCI or AD. Subjects agreed upon by a consensus committee were assigned to non-expert PCPs who, following brief training, assessed them using the same clinical information and utilizing the same assessment instruments. The chance-corrected inter-rater reliability (expert versus non-expert) measure κ, based on binary outcome (aMCI/not-aMCI), was estimated.

RESULTS: The study recruited 119 evaluable subjects (50 aMCI, 27 mild AD and 42 NCI) and demonstrated fair to moderate agreement (κ = 0.423) between experts and non-experts in designation of aMCI. The percent agreement was 72.3%, sensitivity 62.0% and specificity 79.7%. Overall, non-experts under-rated the level of impairment compared with experts.

CONCLUSION: This study established the feasibility of making the aMCI designation in the community and identified some likely sources of error. The results suggest that when drugs with clear benefit for aMCI patients are developed, community-based PCPs, with additional, more optimized training, will be able to accurately identify those patients who should receive treatment.