Since the 1980s, the United States has been experiencing growing inter-county inequalities in mortality. Understanding these differentials is increasingly viewed as a key to improving the health and longevity of the population, especially in the current context of a health reform which has had differential impacts by areas. However, a full understanding of the driving forces behind these trends is for now limited by the accessible data and tools. Information on the state/county of residence is not included in the mortality files published by the National Center for Health Statistics for years since 2004, outside of the restricted-access Census Research Data Centers. Futhermore, there are no life table series or complete series of age-specific death rates publicly accessible for all U.S. counties by calendar year despite their value for the analysis of geographic disparities in mortality and changes in the structure of local health systems. The project goal is to construct a publicly-available database of county-level life tables for all-cause mortality in the U.S. by sex and age, for all counties with populations above 10,000 and for county aggregates within each state for all counties below 10,000 for all years 1982-2015 (or later). The approach builds on current gold standard methods designed to resolve issues of small-area demographic estimation and to quantify uncertainty. The model to be implemented has an underlying functional form that captures regularities in age patterns in mortality. It has been tested using simulated data and published in the December issue of the journal Demography. A Bayesian hierarchical framework then builds on this functional form, penalizing departures from the characteristics shapes across age, as well as sharing information across counties and ensuring a relatively smooth trend in mortality over time. The originality of the proposed model is to exploit empirical regularities in mortality rates simultaneously over age, time, and space. Improved information about the areas and the age groups contributing to ongoing disparities in life expectancy and other mortality indicators across the United States will help researchers identify the underlying conditions and processes that have produced these patterns. County-specific data are particularly useful to measure the role of contextual factors in geographic disparities. County-level mortality indicators can also help guiding policy efforts to investigate the historical impact of public health interventions and changes in the structure of local health programs. The proposed database is expected to result in new and valuable research in demography, epidemiology, economics, and related fields by making county-specific mortality data readily available to the broader research community as well as to public health analysts and policy makers.
The purpose of the project is to construct a publicly-accessible collection of age-specific death rates and other life table values by sex and by county for years 1982-2015 using a statistical technique designed to improve information about mortality patterns in small populations. The resulting time series will be presented in a publicly accessible database to guide future policy efforts to improve the health and longevity of the population, to monitor geographic inequalities, and to investigate the historical impact of public health interventions.
