Exploring The Statistical Methods Of The Death In Coronary Heart Disease And Stroke On Cohort Study

The objective of the research is to explore the statistical methods of follow-up data in chronic and noncommunicable diseases' with regard to the characteristics of prospective study data of occurrence and death of coronary heart disease (CHD) and stroke from Sichuan Province, some statistical methods was applied to determine the risk factors of death in CHD and stroke, and disability adjusted life years (DALYs) was utilized to evaluate their disease burdens. The main research works and results obtained are as follows:(1) According to the example, the methods of data entry and logic audit were systematically studied. Seven advices available and easy to use were given, in order to provide scientific bases for logic auditing large data.(2) The death situation of CHD and stroke more than aged 40 in Sichuan Province was quantitatively described. The results show that the accumulative death rate of CHD and stroke is 3.4 and 13.5 per thousand, respectively. And the standardized death rate of CHD in male and female is 13.4 and 11.5 per ten thousand respectively; the one of stroke in male and female is 61.0 and 33.4 per ten thousand respectively. They were under the lower level as compared with other countries.(3) Based on the lower death rate of CHD and stroke, the rare events logistic regression was applied to determine the risk factors of the death in CHD and stroke, meanwhile compared with unconditional logistic regression, Poisson regression and Cox proportional hazards model. The results reveal that the age, the family history of high blood pressure and the 5 levels of blood pressure are the risk factors of CHD;the age, the history of drinking, the history of stroke and the 5 levels of blood pressure are the risk factors of stroke.(4) Based on the association of the outcome, the generalized estimating equation (GEE) was applied to determine the risk factors of CHD and stroke, the results are similar with previous study. The proper method of selecting the working correlation matrices was proposed.(5) As examples of the death in CHD and stroke, The DALYs was utilized to comprehensively evaluate their disease burden. The results show that the disease burden of CHD and stroke in male is 225 and 929 DALYs respectively, accounting for 1.32% and 5.44 of all DALYs lost due to all causes in male; the one in female is 217 and 809 DALYs respectively, accounting for 1.72% and 6.43 of all DALYs lost due to all causes in female. The DALYs lost due to CHD per thousand persons is 63 and 57 respectively in male and female, the DALYs lost due to stroke per thousand persons is 260 and 211 respectively in male and female.