Report on Chinese Residents’ Chronic Diseases and Nutrition 2015中国居民营养与慢性病状况报告(2015)
上QQ阅读APP看书,第一时间看更新

1.2 Main Analysis Methods

Results in this report were representative of the whole nation and in terms of urban area,rural area, sex and groups of different ages.
Each monitoring or survey had been based on the process of data clearing, data validation and standard database establishment. Analysis and report had all applied scientific method,definition or standard, and every indicators or content results had been verified by relevant experts repeatedly, so the results were credible. According to different survey purposes, different methods had been adopted for analysis of different items.
1.2.1 Complex sampling and weighting of means and rates
The Chinese Nutrition and Health Surveillance 2010-2013, Alcohol consumption, Global Adult Tobacco Survey 2010 and Chinese Youth Tobacco Survey 2014 all adopted the population data released by the NBS, and carried out processing of complex sampling and weighting of means and rates, to gain f nal weight through computing sampling weight (basic weight) and post stratif cation weight and adjustment of responses, for indicator computing.
(1) Sampling weight (basic weight): computed according to sampling design;
(2) Post stratification weight: stratified by area, age and sex, etc., for adjustment of deviation of some important indicators in sample and general distribution due to sampling;
(3) Final weight: product of sampling weight and post stratif cation weight.
1.2.2 When computing main chronic disease mortality variation trend and geographical distribution, the research method in Global Burden of Disease Study 2013 was used.
1.2.3 Cancer registration data were stratified by area (urban/rural), sex, age, cancer incidence rate and mortality; then combining national population data 2010, to estimate incidence of national total malignant cancers and main cancers.
1.2.4 COPD prevalence analysis was made with literature meta-analysis method.Literature retrieval method and literature inclusion and exclusion criteria were standardized. US AHRG-recommended cross-sectional research quality evaluation standard was used to evaluate research quality, and scientif c statistical analysis method was adopted for results analysis.
The analysis methods for different data source are detailed in Appendix III.