蒙特卡罗模拟在批质量保证抽样(LQAS)样本量估计中的应用  

Monte Carlo simulation for sample size estimation of lot quality assurance sampling(LQAS)

作　　者：于石成[1] 冯国双[2,3] 王琦琦 李洋[4] 胡跃华[1] 李俊琦[1] 龙晓娟 向祥龙 石佳欣 Yu Shicheng;Feng Guoshuang;Wang Qiqi;Li Yang;Hu Yuehua;Li Junqi;Long Xiaojuan;Xiang Xianglong;Shi Jiaxin(0ffice of Epidemiology,Chinese Center for Disease Control and Prevention,Beijing 102206;National Center for Children's Health,Beijing Children's Hospital,Capital Medical University,Big Data and Engineering Research Center,Beijing,100045;Beihang University&Capital Medical University,Beijing Advanced Innovation Center for Big DataBased Precision Medicine,Beijing 100083;National Institute of Viral Disease Control and Prevention,ChineseCenter for Disease Control and Prevention,Beijing,102206)

机构地区：[1]中国疾病预防控制中心,北京102206 [2]国家儿童医学中心,首都医科大学附属北京儿童医院,大数据和工程研究中心,北京100045 [3]北京航空航天大学&首都医科大学,北京大数据精准医疗高精尖创新中心,北京100083 [4]中国疾病预防控制中心病毒病预防控制所,北京102206

出　　处：《中国疫苗和免疫》2020年第4期468-473,共6页Chinese Journal of Vaccines and Immunization

基　　金：北航-首医大数据精准医疗高精尖创新中心计划。

摘　　要：本文在回顾批质量保证抽样(LQAS)样本量计算的统计学原理基础上,用超几何分布计算在一定条件下LQAS一次抽样的样本量;应用Excel和SAS软件展示二项分布蒙特卡罗模拟的原理和实现;采用SAS超几何分布随机数函数完成LQAS样本量估计的蒙特卡罗模拟。超几何分布计算的与蒙特卡罗方法模拟的LQAS一次抽样样本量与现有LQAS样本量表的结果一致。蒙特卡罗模拟估计LQAS样本量具有深刻的统计学思想,其计算简单,易于理解;LQAS可广泛应用于公共卫生领域的快速评价。In this article, we review statistical principles of sample size estimation for lot quality assurance sampling(LQAS);compute sample sizes for LQAS under a certain conditions using a hypergeometric distribution;elaborate principles and processes of Monte Carlo simulation for binominal distributions using Microsoft Excel and SAS;and perform Monte Carlo simulation for LQAS sample size estimation using the random function of hypergeometric distribution in SAS. Sample sizes computed using hypergeometric distributions or Monte Carlo simulations for onetime LQAS sampling were the same as sample sizes from LQAS tables. Monte Carlo simulation for LQAS sample size estimation has a profound statistical basis;calculations are simple and understandable. LQAS could be more widely applied for rapid public health assessments.

关 键 词：蒙特卡罗方法 批质量保证抽样 超几何分布 二项分布 模拟 

分 类 号：R195.1[医药卫生—卫生统计学]