OBJECTIVETo evaluate the completeness of the death registration system, so as to understand the death patterns in Xuanwei.

METHODSThe investigation on under-reported deaths was conducted in 30 villages selected with a multi-stage random sampling strategy. Participants were asked about changes of their family members (family members born or dead) during past 3 years with door to door visit. Then, death cases obtained in our investigation were matched with those from routine death registration system and under-reported rate of deaths during 2011-2013 was calculated employing capture-recapture method.

RESULTSTotal under-reported rate of deaths was 31.88%. For people aged between 0-14, 15-39, 40-69 and 60 above, under-reported rates of death were 33.35%, 34.93%, 29.10%, and 32.88%, respectively. And they were 31.72% and 32.02% for males and females, respectively. There was no significant difference shown in under-reported rates among deaths in different age groups (χ² = 7.24, P = 0.065) and genders (χ² = 0.06, P = 0.803). The under-reported rates in high-mortality, medium-mortality and low-mortality regions were 17.48%, 38.01%, and 36.22%, respectively with a significant difference (χ² = 213.25, P < 0.001). Death in local regions with mortality rate higher than 600.00/10(5), between 400.00/10⁵ and 600.00/10⁵ and lower than 400.00/10⁵ were adjusted with under-reported rates in three regions above respectively. The total adjusted morality rate in Xuanwei during 2011-2013 was 776.47/10⁵. For males and females, they were 918.73/10⁵ and 617.46/10⁵, respectively.

CONCLUSIONOverall under-reported rate of death was high in death registration system in Xuanwei. It was necessary to adjust mortality data reported with under-reported rate of death to estimate death patterns in this area.

China ; Data Accuracy ; Death Certificates ; Female ; Humans ; Male ; Mortality ; Registries

Background and objective Since 1970s, Xuanwei in Yunnan province has been one of the towns with highest lung cancer mortality in China. Moreover, the characters of high female lung cancer mortality and sub-regional clustering high lung cancer mortality have not changed. In this study, we further described the exposure situa-tion of risk factors of lung cancer in Xuanwei nowadays, in order to explore the trend of the distribution of lung cancer there.Methods Firstly we divided the 26 towns of Xuanwei city to high-, median- and low- lung cancer areas by the lung cancer mortality in 2010-2012. We chose 2 towns within each area according to topography and orientation, and randomly picked 4 villages in each town to be our study area. We did a questionnaire about lung cancer related risk factors upon the sample population in the study area. We calculated the exposure percentages of each risk factor, in whole sample population and subgroups, for nowadays and for 10 years ago (only living environmental risk factors), and compared them between areas or time points using standardized rates and the statistical test of standardized rate comparison, or chi-square test.Results 65%-80% male in the study area has a history of smoking; 60%-90% non-smoker has been exposed to second hand smoke. These situations are worse in high and median lung cancer areas. 50% male in median lung cancer area have coal mining work experience, which is 2 times of the percentages in the other two areas; while 15%-25% people in high lung cancer area have other occupational exposure history to particulate air pol-lution, which is 3-5 times of the percentages in the other two areas. From ten years ago until nowadays, 80% families in median lung cancer area use 2 tons or more smoky coal per year; more than 90% families burn coal for household heating; more than 60% families suffer from smog in the kitchen during cook; 60% families most frequently use stove in the ground with chimney. Only 20% families in high lung cancer area now use 2 tons or more smoky coal per year. Now 50%-75% families in the study area use 700 kilowatt-hours of electricity per year, much more than ten years ago. 80% residents in low lung cancer area eat fatty or pickled or smoked food at least 3 days per week; while in high and median lung cancer areas the percentages are 50%-60%.Conclusion According to data obtained in this survey, current distribution of smoky coal use has differed from the distribution of high-, median- and low- lung cancer areas. Tobacco use and second hand smoke, the use of smoky coal and occupational exposure to particulate matters could be the main risk factors for lung cancer in Xuanwei now. The relations between lung cancer and stove type, dietary habit and so on deserve further study.