The purpose of this research is to develop methods to forecast life expectancy, active life expectancy and health status changes among the U.S. elderly and oldestold population. To produce such projections it will be necessary to conduct both a.) select substantive studies to examine mortality and morbidity patterns at advanced ages and the potential for modifying those patterns, and b.) methodological studies to integrate stochastic process models of human morbidity, disability and mortality processes into comprehensive forecasting models. In the substantive studies, use will be made of extensive data holdings of: a.) longitudinal studies (e.g., Framingham, Evans County, Albany Civil Servant Study, First and Second Duke Aging Studies, multiple data sets from the WHO ERICA Archiving Project); b.) vital statistics data (e.g., 1962 to 1984 underlying cause of death mortality files, 1968 to 1984 multiple cause mortality files), and c.) national health and disability surveys (e.g., 1982 and 1984 National Long Term Care Surveys; 1969-1977 National Nursing Home Surveys; 1969-1981 Health Interview Surveys). These data sets represent several different types of observational plans and degrees of individual information and will be analyzed with different appropriate models of stochastic processes; i.e., multivariate Gaussian stochastic process model; fuzzy set models of complex discrete state processes, discrete state models of partially observed processes estimated from multiple data sources, and stochastic compartment models. The development of those models will be taken from the stage of theoretical specification where the model is structured to represent the underlying processes of human physiological aging and mortality to the development of practical statistical estimation and forecasting procedures. Projections will be developed for a variety of health and disability endpoints. These projections will be sufficiently detailed to be useful for planning for such programs as the development of geriatric and other ancillary health professional manpower to better meet the increasing volume of health and LTC needs of the elderly and extremely elderly populations. Projections will be made for a series of time horizons and the uncertainty of different endpoints will be assessed.
Showing the most recent 10 out of 15 publications
