《捕获再捕获分析ppt课件.ppt》由会员分享，可在线阅读，更多相关《捕获再捕获分析ppt课件.ppt（85页珍藏版）》请在taowenge.com淘文阁网|工程机械CAD图纸|机械工程制图|CAD装配图下载|SolidWorks_CaTia_CAD_UG_PROE_设计图分享下载上搜索。
1、文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。What is it for?为什么?Capture-recapture analysis is used for counting the total number of people in a population using two or more incomplete lists of those people 捕获-再捕获分析用于两组或多组非完整名单来对某个人群进行计数Why should I be interested?为什么我对此感兴趣为什么我对此感兴趣?Evaluating surveillance sys
2、tems 评价监测系统Magnitude of issues 问题的大小文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Overview概要Origin of method方法的来源Application to epidemiology-why is it useful for us?方法的应用-为什么对我们有用?Principles原则Conditions for using capture-recapture methods应用捕获-再捕获的条件Methods方法Two sources两个来源Multiple sources多个来源Limitations不足之处文档
3、仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Origins of capture-recapture analysis捕获-再捕获分析方法的来源Origins in demography来源于人口学1662-used to estimate the population of London 1662年用来估计伦敦市的人口1783 -Laplace used to estimate population of France 1783年laplace用于估计法国的人口1949 -Sekar and Deming used to estimate birth rate and
4、 mortality in India 1949年Sekar和Deming用于印度的出生率和死亡率Subsequently most often for estimating wildlife populations 随后通常用于野生动物数量More recently applied to epidemiology(Wittes 1968)近期被用于流行病学 文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Application of capture-recapture analysis to human epidemiology捕获-再捕获分析在人类流行病学上的应用E
5、valuating completeness of a surveillance source 评价监测资料的完整性评价监测资料的完整性Passive surveillance被动监测Registers登记Refining incidence and prevalence estimates from surveillance systems or population surveys改进监测系统或人群调查改进监测系统或人群调查得到的发病率和患病率估计得到的发病率和患病率估计Used for cancers,stroke,homelessness,mental illness,drug use
6、,congenital disorders,infections运用于癌症、中风、无家运用于癌症、中风、无家可归、精神疾患、吸毒、生天性疾患和感染可归、精神疾患、吸毒、生天性疾患和感染文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Principles原则Two or more sources(lists)of cases a given disease 某一疾病的两组或多组来源病例Sources considered random capture samples in population 认定来源病例为一总体的的随机捕获样本Cases can be matched
7、by unique identifiers 病例之间可以用唯一识别匹配Estimate total number of cases that are not captured by any source from the matched and unmatched估计未被任何病例来源（匹配和未匹配）捕获的病例总数文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正
8、。文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Critical assumptions/conditions重要假设条件1.Population is closed 人群是封闭的methods exist for open populations有用于开放人群的方法2.Individuals captured on both occasions can be matched两次被捕获的个体间可以匹配3.Capture in
9、the second sample is independent of capture in the first两次捕获之间是相互独立的4.Probability of capture is homogeneous across individuals每个个体被捕获的可能性是一致的Homogeneity of individuals个水平上Homogeneity of lists名单水平上文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Application to humans在人群上的应用“Capture”=appearing on a“list”“捕获”出现在“名单
10、”上“re-capture”=linking by identifying individuals appearing on both lists by criteria name,date of birth etc“再捕获”同时出现在两个名单位并符合标准（姓名、生日等）的连接“Trap fascination”陷阱诱惑“if you feed the animal they are more likely to be caught again 如果你喂食，动物被抓获的可能就增大laboratory confirmed cases are more likely to be reported
11、in other systems实验室确证的病例在其他系统被报告可能就更大“Trap avoidance”陷阱逃逸”if you scare the animal they will avoid the trap 如果吓唬动物，动物将不会被逮a person cant appear on community injecting drug user registry if they are in prison 被关押在监狱的人不可能出现在社区IDU登记册上文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。仅在第一次被捕获仅在第二次被捕获未被捕获第一次和第二次均被捕获未被捕获
12、文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Two sources两次捕获样本Source ASource Bx12x11x21x22?1 included in source 在捕获样本中在捕获样本中2 not included in source不在捕获样本中不在捕获样本中文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Estimated conditionally to number of cases observed in other cellsCapture(Source A)and recapture(source B)捕获（样本
13、A）和再捕获（样本B）文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Estimation估计If sources independent:P(A+if B+)=P(A+if B-)如果是独立样本：文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Capture(Source A)&recapture(Source B)文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Sensitivity of sources:样本来源的敏感性If numbers in cells small,probability that x11=0
14、is not zero 如果数字较小,X11=0的可能性不为0Estimation 文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Conditions条件Same study period and area 研究期间和地区相同Closed population封闭的人群All cases in any source are true cases 任何样本中的案例是真正的案例True matches are identified 可进行真正的匹配Equal catchability被捕获的可能性要相同Sources are independent独立样本文档仅供参考，不
15、能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Same study period and area相同的研究期间和地区Cases occur during the study period and in study area 在研究期间和研究地区发生案例Different period of capture 不同时间的捕获Probability of recapture x11 overestimates N 再捕获的可能性 x11 高估 N文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Closed population封闭人群Nobody enters o
16、r leaves the population during the study period在研究期间该人群没有人进出No immigration,emigration,death没有移入、移出、死亡开放人群Open populationIndividuals captured in first sample cannot be captured in secondProbability of recapture x11 overestimates N文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。True cases真正的案例All cases in any sou
17、rce are true cases 所有来源的案例是真正的案例False positive cases假阳性案例Positive predictive value(PPV)阳性预测值 1(positive dependence):d d underestimates NOR d overestimates N文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。
18、ExampleEstimation of number of IVDU in Bangkok in 1991(Maestro 1994)1991年曼谷IVDU人数估计Two sources used:两个样本来源Methadone(April May 1991)美沙酮Police arrests(June September 1991)警方逮捕Methadone 美沙酮 Need for drugs毒品需求 Probability of being arrested被逮捕的可能性 =negative dependence,over-estimation of N 相关度为负值,高估N文档仅供参
19、考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Evaluation of source dependenceTwo sources两组独立样本Qualitative analysis of the notification process in each source i.e.there is no statistical method to allow for dependence for two sources定性分析每个样本的告知过程,也就是说没有统计学方法来分析两个样本的依赖度Multiple(2)sources多个来源样本Wittes methodLog-linear
20、 modelling文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Behavioral Surveillance Using Respondent Driven Sampling 用同伴推动抽样法进行行为监测文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Presentation Outline提纲Sampling methods for hard to reach populations;是针对难以接近人群的抽样方法Description of RDS;RDS的描述Lessons learned from Vietnam;越南实践中得到的经验
21、教训文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Probability Sampling 概率抽样(Simple单纯随机,Systematic系统,Cluster整群)Gold Standard-Best methods for samplingBut,do not reach hidden populations:金标准-最好的抽样方法,但不能接近隐匿人群No sampling frame没有抽样方案Stigmatized被歧视Would need huge sample sizes in order to capture a hidden population
22、需要很大的样本量才能找到隐匿人群Expensive费钱文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Sampling Methods to Reach Hidden Populations 针对隐匿人群的抽样方法Time-Location(TLS),Venue-Based 在场所的时间-空间抽样法 -Major Bias:Only captures those who are visible 主要偏倚:仅能找到显性人群Snowball滚雪球 -Major Bias:Not representative of the population(tendency for in
23、-group affiliation,volunteerism and masking)主要偏倚：代表性不好（组内从属倾向，自愿性）文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Background on RDSRDS背景Developed by D.Heckathorn and R.Broadhead with IDUs in Connecticut and in Yaroslavl,Russia;由D.Heckathorn和R.Broadhead在 美国Connecticut州和俄罗斯Yaroslavl的IDU中研发而成Sampling vs.Recruitmen
24、t strategy;抽样 vs 招募策略Different from other chain referral methods because it can give us point estimations with standard errors.不同于其他链式推举方法,因为可以给出一个点估计和标准误文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。How RDS Works如何做RDSUse of a dual system of recruitment through the use of incentives.运用双重招募系统,并给予报酬(小的刺激)Use o
25、f recruitment quotas.运用招募限额Use of peers to recruit peers.运用同伴招募同伴Use of links between recruiters and recruits.运用招募人与被招募人之间的联系文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。The Theory Behind RDSRDS的理论要点Uses prinicples of First Order Markov Theory运用Markov链式理论Long referral chains 推举长链Final sample will be independ
26、ent of those selected as“seeds”最后的样本独立于刚开始的”种子”Final sample will be similar to the population of the network from which you are recruiting 最后的样本将与你所招募捐的人群相近文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Wave 1 Wave 2 Wave 3 Wave 4 Wave 5文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Wave 1 Wave 2 Wave 3 Wave 4 Wave 5文档仅供
27、参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Wave 1 Wave 2 Wave 3 Wave 4 Wave 5文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Wave 1 Wave 2 Wave 3 Wave 4 Wave 5文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Wave 1 Wave 2 Wave 3 Wave 4 Wave 5文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Wave 1 Wave 2 Wave 3 Wave 4 Wave 5文档仅供参考，不能作为科学依据，请勿模仿；如有不
28、当之处，请联系本人改正。A Long Referral Chain:Jazz Musicians in New York City 文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Selection of Seeds文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Example in Hai Phong Vietnam越南海防的例子Final Sample size:420 IDUs in Hai Phong and Saigon;418 CSWs in Saigon and 220 in Hai Phong 最后的样本量:海防和西贡的420名I
29、DU,西贡的418名CSW,和海防的220名CSWRecruitment process招募过程20 seeds selected by peer educators 同伴教育员选择20个种子Three coupons to each participant 每个参加对象发三张卡片Participants asked to recruit their peers 要求参加对象招募其同伴Time:March June,2004时间:2004年3-6月Three sites(Hai Phong);Four sites(Saigon)海防的三个地点;西贡的四个地点文档仅供参考，不能作为科学依据，请勿
30、模仿；如有不当之处，请联系本人改正。Eligibility Criteria入选标准CSWs:-Women,18 years or more,living or working in Hai Phong or Saigon;女性,18岁及以上,在海防或西贡生活或工作-Has sold sex for money in the last 30 days;在过去的30天内以性换钱-Has a green coupon(except seeds);有一张绿卡片(除种子外)-Has provided consent.知情同意IDUs-Women(Saigon only)or Men,18 years
31、or more,living in Hai Phong or Saigon;男性或女性,18岁及以上,住在海防或西贡,西贡仅做女性-Has injected drugs during the last 30 days;在过去30天内注射过毒品-Has a yellow coupon(except seeds);有一张黄卡片(除种子外)-Has provided consent.知情同意文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Coupon:Front Side卡片正面LIFE-GAP projectFor Your Health and SafetyReferra
32、l couponAddress_Telephone:_(You can call to make appointment in advance)ID number:Opening hours:Monday-SaturdayFrom 9:00 AM to 5:00 PMLIFE-GAP project:For Your Health and SafetyPayment coupon Address_Telephone:_(You can call to make an appointment in advance)You will receive 15,000 VND for each pers
33、on who you recruit and enrolls into the study(you may recruit up to 3 persons)ID number:Please call us in advance.You must present this coupon for payment文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Coupon:Back Side卡片背面LIFE-GAP project:For Your Health and SafetyBring this coupon to see if you are eligible to
34、 participate in this study.If eligible,you may receive:-Up to 45,000 VND-Free,voluntary/confidential HIV counseling and testing This coupon will not be accepted under the following conditions:-The project has reached the target sample size of 400-This coupon is torn up,tampered with and/or unreadabl
35、e-Coupon holder has already been interviewed with same color coupon Please come and participate in our study LIFE-GAP project:For Your Health and Safety This coupon will not be accepted for the following reasons:-The project has reached the target sample size of 400-This coupon is torn up,tampered w
36、ith and/or unreadable You will receive your incentive if each person you recruit:-is eligible for the study-has not already participated in this study with same color coupon-has completed the behavioral questionnairePlease come and participate in our study 文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Network
37、s of CSWs in Hai Phong文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。A network in Hai PhongSeed文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Initial Lessons from Vietnam越南的初步经验Seeds should have high degree-initial focus group may be important;种子应该具有高学历,最初的中心小组很重要No slow down mechanism to end RDS;没有减速机制来停止RDSNeed for sec
38、urity-Interviewers have no choice of whom they interview;安全保障-调查员无法选择调查对象Managing multiple sites can be difficult;同时管理几点地点有困难Managing coupon numbers;卡片号码管理No way to control for those who recruit faster.没法控制那些快速招募者文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Initial Lessons from Vietnam(Cont)Difficult to disc
39、ourage recruiters from selling coupons or giving them out in a non random way;没法控制招募人出售卡片或以非随机的方式发出卡片Non response information difficult to obtain (incentives picked up by friends,recruiters do not return for secondary incentive)无应答信息很能获取(报酬被朋友拿走,招募人不回来再取报酬)文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Philoso
40、phical objection?哲学上的异议?Capture-recapture is fun,so it must be epidemiology!捕获-再捕获很有意思,所以客观上一定是流行病学!But,as epidemiologists we are interested in 但是作为流行病学家,我们对三间分布感兴趣Time,place and personCapture-recapture does not capture time-it is a static tool which relies on lists which correspond to prevalence of
41、 a chronic disease(e.g.diabetes)or long time periods for acute diseases(legionella)捕获-再捕获不”捕获”时间,这是一个静态工具,领带于与慢病性(如糖尿病)或长时间急性病(军团病)的名单Can be used for measuring broad trends by repeat analysis(Nardone et al Epidemiol Infect 2003)可以用重复分析来测量总体趋势文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Practical limitations操
42、作上的不足之处Unique identifier has to match in all data sources 必须用唯一识别信息对所有数据来源进行匹配This may contravene confidentiality laws这可能与保密法相抵触Clever statistics cant correct bad data 聪明的统计不能纠正不好的数据 Rubbish in,rubbish out.垃圾进,垃圾出For chronic and expensive diseases(eg diabetes)it may be better to carry out an expensi
43、ve detailed survey than to use quick and dirty methods对慢性和费钱的疾病(如糖尿病),开展一项费钱的详细调查比使用快速但很差的方法更好it may be even more expensive to get it wrong.如果出错,费用更高文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Extrapolation is based on assumptionswe are assuming that the model which describes the observed data also describe
44、s the count of the unobserved individuals.We have no way of checking this assumption.This is analogous to,and has the same dangers as fitting an arbitrary curve to a series of points(x,y),where x0,with the intention of estimating y at x=0.this is analogous to the position of those who automatically
45、assume that the k samples in our problem are independent.我们假设描述观察数据的模型也描述了未观察个体的计数,我们无法检验这个假设.这与随意用一曲线对一组数据(X,Y)进行拟合一样并有同样的危险,当X0时,当X=)估计Y.这与自动假设我们的问题有K个独立样本的某些人的立场一样.Fienberg,Biometrika 1972;59:591-603文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Conclusion小结If conditions are met如果条件符合Potential to use multip
46、le incomplete registers and to estimate population size by capture-recapture有可能利用多组不完整的登记数据,用捕获-再捕获的方法估计人群基数Cheaper than exhaustive registers比彻底完全的登记少费钱 两个来源样本 不可能对相关度进行定量Two sources两个来源样本Impossible to quantify extent of dependenceRequires third sourceMultiple sourcesLog-linear modelling method of c
47、hoiceCan adjust for dependence and variable catchability文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。Caveats警告Use technique but be careful!使用技术但要小心Dont treat this as a black box method 不能当成是一个暗箱方法All prior knowledge should be used to formulate the model就用所有前人的知识来制定模型Know your data!了解你的数据Not the solution to a
48、ll problemsConditions often not met when applied to epidemiology There may still be heterogeneity you dont understandComplementary technique文档仅供参考，不能作为科学依据，请勿模仿；如有不当之处，请联系本人改正。ReferencesWittes JT,Colton T and Sidel VW.Capture-recapture models for assessing the completeness of case ascertainment usin
49、g multiple information sources.J Chronic diseases 1974;27:25-36.Hook EB,Regal RR.Capture-recapture methods in epidemiology.Methods and limitations.Epidemiologic Rev 1995;17(2):243-264 International Working Group for Disease Monitoring and Forecasting.Am J Epidemiol.Capture-recapture and multiple-rec
50、ord systems estimation I:History and theoretical development.1995;142:1047-58International Working Group for Disease Monitoring and Forecasting.Am J Epidemiol.Capture-recapture and multiple-record systems estimation II:Applications in human diseases.1995;142:1059-68LaPorte RE,Dearwater SR,Yue-Fang C