论文部分内容阅读
目的用自加权分层3阶段PPS系统整群抽样(自加权分层3阶段抽样)技术,做大规模疾病流行病学调查抽样设计,为我国疾病流行病学调查的抽样设计提供参考思路。方法以结核病患病率现场调查为例,引用2000年我国第4次结核病患病率调查资料,参考二重抽样技术进行随机整群抽样。依此估计的结核病患病率整群调查所需的群数,试作自加权分层3阶段抽样设计。结果二重抽样结果表明,开展结核病患病率整群调查群数约180群。由此,据中国现时人群结核病患病率(预估为2.5‰),以整群抽样为基础,每群调查1 800人,第1阶段选以县为初级抽样单位,将各省按卫生服务类型作2~3级分层后,依各层所辖县人口数用分层等距(PPS系统)抽取60个县。第2阶段亦用分层PPS系统抽样法,在抽中的县分别抽取1个居委会/社区、2个行政村。第3阶段在抽中的村(居)委会/社区整群调查1 800人。结论自加权分层3阶段抽样技术样本代表性好,便于组织调查及其质控实施,节省调查经费,目标值估计简单,调查精确度和效率较高,适用于我国人群大规模疾病流行病学调查及疾病监测抽样设计。
OBJECTIVE: To provide a reference for the sampling design of epidemiological survey of diseases in our country by using the self-weighted stratified 3-stage PPS system cluster sampling (self-weighted stratified 3-stage sampling) technique to do the sampling design of large-scale disease epidemiological investigation. Methods Taking the field survey of prevalence of tuberculosis as an example, we used the data of the 4th TB prevalence survey in China in 2000 and randomized cluster sampling with reference to the double sampling technique. Based on the estimated number of clusters required for the prevalence of tuberculosis in the survey, a self-weighted stratified 3-stage sampling design was tried. Results The results of double sampling showed that the number of cohorts carrying out the prevalence survey of tuberculosis was about 180. Therefore, according to the prevalence of tuberculosis in the Chinese population (estimated at 2.5 ‰), based on cluster sampling, 1,800 people were surveyed in each group. In the first stage, counties were selected as primary sampling units, After 2 or 3 stratification, 60 counties were selected by stratified equidistance (PPS system) according to the population of counties under each jurisdiction. The second stage also uses the stratified PPS system sampling method, and draws one neighborhood / community and two administrative villages respectively in the selected counties. Phase 3 In the extraction of the village (neighborhood) committees / community cluster survey of 1,800 people. Conclusion The self-weighted stratified 3-stage sampling technique has good representativeness, which is convenient for tissue investigation and quality control implementation, saves investigation funds, has simple estimation of target value, high accuracy and efficiency of investigation, and is suitable for large-scale epidemiology of population in China Survey and Disease Surveillance Sampling Design.