The goal of this project is to improve the quality of
cause of death information for public health research and planning.
The study provides a more scientific
basis for using cause of death data needed for estimating global and regional
mortality patterns. A related broad
goal studies the transition in non-communicable disease mortality rates to
shed light on a key controversy in public health: whether age-specific death
rates for non-communicable diseases rise or decline with economic development.
The first component of the project focuses on
developing analytical procedures to correct for miscertification of deaths from
ischaemic heart disease, cancers and other major chronic diseases. Multiple regression methods with independent
variables having a predictable relationship to disease levels (e.g. cigarette
consumption) are being applied to national cause of death statistics to develop
disease-specific correction algorithms to apply to registered cause of death
data. In order to make better use of
cause of death data collected in sentinel surveillance systems by verbal
autopsies of relatives of the deceased, validation of cause of death data from
sample sites in China and Tanzania are being carried out. Validation methods include
growth-balance demographic techniques to estimate completeness of death
recording by age and sex. Verification
of the reliability of cause of death information are being carried out using
repeat verbal autopsies of family members (China) and validation against
hospital clinical diagnosis (Tanzania).
A critical component of the project is to
use corrected cause of death data to analyze long-term trends in age-specific
non-communicable disease mortality rates to examine the extent to which
income-elastic risk factors for major non-communicable diseases affect mortality
trends. We firstly describe mortality trends back to 1900 for selected
industrialized and middle income countries using national sources. We then apply the correction algorithms
developed in the project to re-estimate the apparent trend in age-sex-specific
NCD mortality rates over the last 100 years in these countries to test the
research hypothesis.
