Estimating Tobacco-Attributable Disease Burden

Through the above data sources and methods, we can estimate the DALYs and death for health outcomes in China. GBD 2017 estimated the tobacco-attributable DALYs as the total DALYs multiplied by the population-attributable fraction (PAF) for the tobacco-related risk–outcome pair by age, sex, and year. The method for estimating the tobacco-attributable deaths is similar to those of the DALYs. The PAF for each individual risk–outcome pair was estimated independently. The estimates of PAF are based on Chinese tobacco exposure data.

The original data of tobacco exposure were obtained from diversified sources. These sources are cross-sectional nationally representative household surveys; household composition module, including the Demographic Health Surveys, Multiple Indicator Cluster Surveys, and Living Standards Measurement Surveys, self-reported tobacco exposure from cross-sectional surveys, including the Global Adult Tobacco Surveys and WHO STEPS Surveys; national and subnational censuses; and published studies. Spatiotemporal Gaussian process regression was used to estimate PAF.

Two points must be declared. Firstly, in the estimate of the smoking-attributable disease burden, we used the 5-year-lagged smoking prevalence as the exposure. Secondly, for GBD 2017, the current definition of chewing tobacco is the definition of smokeless tobacco in the previous version of the GBD study, including snuff and chewing tobacco and e-cigarettes. Detailed methods for the estimation of the tobacco-attributable burden are presented in published studies (4).