Study on smoking-attributed mortality by using all causes of death surveillance system in Tianjin]

To understand the smoking-attributed mortality by inclusion of smoking information into all causes of death surveillance.Since 2010, the information about smoking status, smoking history and the number of cigarettes smoked daily had been added in death surveillance system. The measures of training, supervision, check, sampling survey and telephone verifying were taken to increase death reporting rate and reduce data missing rate and underreporting rate. Multivariate logistic regression analysis was conducted to identify risk factors for smokingattributed mortality.During the study period (2010-2014), the annual death reporting rates ranged from 6.5‰ to 7.0‰. The reporting rates of smoking status, smoking history and the number of cigarettes smoked daily were 95.53%, 98.63% and 98.58%, respectively. Compared with the nonsmokers, the RR of males was 1.38(1.33-1.43) for all causes of death and 3.07 (2.91-3.24) for lung cancer due to smoking, the RR of females was 1.46 (1.39-1.54) for all causes of death and 4.07(3.81-4.35) for lung cancer due to smoking, respectively.The study of smoking attributed mortality can be developed with less investment by using the stable and effective all causes of death surveillance system in Tianjin.