Conservative estimates are that 3 million elderly Americans suffer from all forms of dementia. This project will determine variations in the practice of caring for this population and whether these variations affect important outcomes -- nursing home placement, reduction of non-cognitive symptoms of dementia, and quality of life. Practice variations to be examined are: 1) whether persons meeting criteria for dementia have been diagnosed, 2) whether a full clinical assessment has been done, 3) use of pharmacologic treatment for symptom management, 4) use of specialty physician care (psychiatrists, geriatricians, neurologists), and 5) use of community-based services (paid in-home assistance, adult day care, meal or other social services). Most studies of persons with dementia are drawn from clinical populations at major medical centers. This project takes advantage of pre-existing samples of older people at high risk for dementia identified through cognitive screening in four population-based studies in Maryland (13,260 persons screened with approximately 1,050 at risk for dementia based on low MMSE scores). Since these studies selected subjects through random sampling, each provides a cohort representative of at-risk older people in the geographic area sampled. Subjects for this study will be selected from this initial at-risk cohort based on a neuropsychological battery administered to determine dementia (expected yield, 681 cases). In addition, contact will be made with primary physician and, when applicable, diagnosing and prescribing physicians for specific information regarding diagnosis and treatment. Data on practice variations and outcomes will be collected at baseline and in 2 annual follow-up interviews. Medicare claims data will be abstracted to address research questions related to selected service use and costs. The findings will be of use in determining current patterns of care and, more importantly, the effectiveness of current methods of care for people with dementia.
