Abstract
This study sought to determine nationally representative estimates of mortality from or with dementia in the United States, to compare rates derived from various sources of ascertainment, and to estimate the contribution of dementia to morbidity and facility usage in the last year of life. National Mortality Follow back Survey (NMFS) data were analyzed. The NMFS is based on a nationwide probability sample of persons aged 25 and over who died in the United States in 1986. For sampled decedents, information was obtained from death certificates, detailed questionnaires and interviews of family members, and from abstracted records of health facilities that provided care during the last year of life. Rates varied by age, race, and source of ascertainment. Rates based on death certificate diagnoses were generally an order of magnitude smaller than rates based on other sources. Alzheimer's disease was the most frequently reported cause of dementia. In the last year of life, most demented patients required assistance with multiple basic activities of daily living, had severe impairments in cognitive functioning, and received some care in general medical-surgical hospitals. Dementia is an extremely common cause of morbidity and mortality. Rates vary widely depending on the method of ascertainment; in particular, ascertainment based on death certificates grossly underestimates the magnitude of the problem.