三来源捕获-再捕获方法分层三阶段抽样下女性性工作者人群基数估计的统计方法及效度信度评价  

Using Three-list Capture-recapture Method for Estimating the Size of Female Sex Workers in Stratified Three-stage Sampling and Evaluating Validity and Reliability

作　　者：靳宗达 耿国柱[2] 傅颖 杨少春[2] 范玉波 高歌[2] Jin Zongda;Geng Guozhu;Fu Ying(Zhejiang Chinese Medical University(310053),Hangzhou)

机构地区：[1]浙江中医药大学,310053 [2]苏州大学公共卫生学院 [3]苏州市疾病预防控制中心

出　　处：《中国卫生统计》2022年第3期335-339,344,共6页Chinese Journal of Health Statistics

基　　金：国家自然科学基金资助项目(81273188);浙江中医药大学校级科研基金项目(2019SR01)。

摘　　要：目的 应用计算机随机模拟技术,评价三来源捕获-再捕获(C-R)方法下分层三阶段抽样调查方法及统计公式的效度和信度,为估计艾滋病高危人群基数提供准确可靠的复杂抽样调查方法及统计公式,为艾滋病防控策略的制定、防治资源的合理分配提供参考依据。方法 运用数理统计方法推导统计公式;在广西壮族自治区分层三阶段抽样调查女性性工作者(female sex workers, FSW)人群,通过三来源C-R方法估算广西FSW人群基数;以实例调查结果作为模拟总体,应用SAS语言,进行三来源C-R方法复杂抽样下计算机随机模拟调查1000次,评价本研究调查方法及统计公式的效度和信度。结果 推导出总体基数的估计量及其估计方差的统计公式;估算出获取时间段为三个月的广西FSW人群基数及标准误。计算机随机模拟调查估算出的1000个总体基数95%置信区间基本上都包含模拟总体的基数。结论 本文研究的调查方法及统计公式具有良好的效度和信度,可用于艾滋病高危人群、敏感人群或特殊人群的基数估计。FSW高危人群基数提示卫生部门应加强对FSW人群精准监测,对艾滋病精准防控。Objective A Monte Carlo simulation was used to evaluate validity and reliability of three-list capture-recapture(C-R) method with stratified three-stage sampling.This study provided the basic data for control of AIDS and for allocation of limited resources.Methods We derived mathematical formulae.Stratified three-stage sampling design was used to investigate female sex workers(FSW) in Guangxi.Moreover, the three-list C-R method was used to estimate the size of FSW.We simulated investigating FWS in Guangxi 1000 times to assess validity and reliability of methods and formulae in this study.Results The formulae were given for the point estimates and the variance estimates of the population size.During the past six months, the size of the FSW in Guangxi was estimated at 95662,and the standard error was 6922.All the estimated population size of 95% confidence intervals from 1000 different simulations almost contained the simulated population size.Conclusion The three-list C-R method based on stratified three-stage sampling design demonstrated high levels of validity and reliability, as well as provided scientific method for estimating the size of high-risk population.The population size of FSW suggested that the health department should strengthen the monitoring and evaluation aspects of HIV/AIDS high-risk population.

关 键 词：捕获-再捕获方法 基数估计 计算机随机模拟 女性性工作者 效度信度评价 

分 类 号：R195.1[医药卫生—卫生统计学]