|
|
|
|
The goal of this project is to improve our knowledge of,
and ability to measure, adult mortality in the developing world. Many developing
countries lack conventional data sources permitting the regular measurement of
mortality. Valid substitute methods for measuring mortality in childhood have
been developed, and applied across the majority of developing countries. Success
with development of substitute methods for adult mortality has been more limited,
and no cross-national comparative reviews of age patterns and trends have been
carried out, despite the fact that extensive data exist.
This project compiles data sets relevant to the
measurement of adult mortality for selected countries representing all regions
of the developing world. We evaluate the data sets and carry out adjustments as necessary, explore and
quantify the performance of the methodologies available for evaluating such
data sets and deriving mortality estimates from them, and develop new
methodologies as necessary to improve estimation validity.
Standard methods are being developed to assess
the uncertainty affecting all estimates, and to define plausible ranges around
observed mortality rates within which true rates are expected to lie. The
validated data are being used to compute life tables for the selected
countries for different time points. A
web-accessible data base including both raw data and final life tables is being
created in parallel with the existing Berkeley Mortality Database (BMD) for
developed countries. The validated life
tables are being used to answer three substantive questions: (a) how do patterns
and trends of mortality in LDCs today compare with the historical experience of
now-developed countries as observed in the BMD? (b) what has been the impact of
economic reversals on mortality trends? and (c) how large has been the
mortality impact of the HIV/AIDS epidemic? The mortality information for selected
countries also is being used as a
basis for developing region-specific distributions of deaths by age as a key
input to studying burden of disease.
|
|
|
