The Key Features Inventory: Early diagnosis of dementia to improve quality of life

James E. Galvin, M.D.

Dementing illnesses such as Alzheimer's disease are significant health problems in the aging population. More than 4.5 million Americans have dementia but few are correctly diagnosed at the earliest stages of the disease. This is due, in part, to the lack of easy-to-administer sensitive clinical tools to measure early cognitive decline.

In a pilot study of 250 individuals participating in a study of aging and memory, we identified clinical variables that were felt to reliably distinguish between individuals with very mild dementia and those without dementia. These "Key Features" were then used to develop a brief informant-based diagnostic tool, the Key Features Inventory (KFI).

In this application, we propose to test the validity and reliability of the KFI in a consecutive series of patients presenting to the Memory Diagnostic Center at Washington University for an evaluation of their memory and thinking abilities. The Memory Diagnostic Center is a faculty practice of five experienced dementia neurologist and four geriatric nurse clinicians. We will test whether the KFI can reliably distinguish between nondemented and demented individuals. We also hope to demonstrate that this brief tool gives reproducible results when administered to the same individual at two points in time.
We will determine whether the KFI can be combined with commonly used performance measures such as the Mini-Mental State Exam to increase the diagnostic accuracy of the physician as compared to our "gold standard", the Clinical Dementia Rating Scale. A sensitive tool such as the KFI that can be administered in less than 2 minutes and can reliably differentiate between nondemented and demented individuals could improve diagnostic accuracy for general practitioners. The KFI would also be applicable to clinical trials, epidemiological studies and community settings.

Final Report:
Dementia is a significant public health problem that is under-recognized and underdiagnosed. Reliance on objective performance in office-based cognitive tests such as the MiniMental Status Exam is hampered by insensitivity to the early stages of dementia and contains biases based on gender, race and educational attainment. Informant-based assessments on the other hand, are unbiased and more sensitive to early cognitive change but are time-consuming ... Read the full Final Report