第2讲-循证医学的证据来源和检索打印版课件.ppt

1、循证医学研究证据的循证医学研究证据的来源与检索来源与检索复习：什么是循证医学？Evidence-based Medicine？2Evidence-based medicine(EBM)Evidence-based medicine(EBM)is the name given to the the conscientious,judicious and explicit use of current best evidence in making decisions about the care of the individual patient(Sackett et al.BMJ 1996).

2、EBM，循证医学（Evidence based Medicine）：是临床医生经过认真尽责地、正确判断地且很明确地做出决定：将最佳证据应用于临床实践诊治病人以便于取得最佳诊疗效果。4EBMThe thinking behind evidence-based medicine is very soundly based on Clinical Epidemiology;and the practice will be familiar to those of you who have experience of problem-based learning.EBMThe philosophy o

3、f evidence-based medicine is not new.What is new is the enormous and growing capacity to access knowledge in an efficient manner.Evidence-based medicine exploits the information explosion that might otherwise overwhelm us.The skills being to ask the right questions,look for answers,and interpret and

4、 apply those answers for the best health outcomes.EBMEvidence-based medicine is a tool with many applications:to complement clinical expertise in providing best care for patients;to make best,rational,use of limited resources;as an aid to continuing education and professional development.EBMThree co

5、m ponents of EBMEvidencePatientClinicianEBM 四大因素的有机结合四大因素的有机结合9临床流行病学方法学基础最佳研究成果素质良好医师患者的参与循证医学STEPS IN EBMEvidence-based Medicine starts with a well-defined clinical problem and then seeks best evidence to solve this problem for the individual patients situation.There are four main steps:STEPS IN E

6、BM1.Framing Clinical Questions.2.Searching for Evidence.3.Appraising the Evidence.4.Applying the Evidence.复习：怎么构架问题？或者都有哪些问题需要来回答的？12Step 1-Framing Clinical Questions询证医学的第一步是问正确的临床问题：这些有关患者保健问题都是医生每天所要面对的。如何解释这些临床结果?哪些诊断试验有助于诊断和管理病人?哪些治疗将会带来最大的“风险与获益”比?预后是怎样的?Step 1-Framing Clinical QuestionsEight

7、central tasks of clinical work that generate clinical questions:1.临床结果（临床结果（Clinical findings）:怎样恰当地采集病史，并根据病史和身体检查来解释临床结果。（how to properly gather and interpret findings from the history and physical examination.）2.病因学（病因学（Etiology）:怎样来识别或鉴定病因（how to identify causes for disease(including its iatroge

8、nic forms，医源性的)）？Eight central tasks of clinical work that generate clinical questions:3.Differential diagnosis（鉴别诊断）（鉴别诊断）:when considering the possible causes of a patients clinical problem,how to rank them by likelihood,seriousness and treatability.4.Diagnostic test（诊断检查）（诊断检查）:how to select and

9、interpret diagnostic tests,in order to confirm or exclude a diagnosis,based on considering their precision,accuracy,acceptability,expense,safety etc.Eight central tasks of clinical work that generate clinical questions:5.Prognosis（预后）（预后）:how to estimate the patients likely clinical course over time

10、 and anticipate likely complications of the disease.6.Therapy（治疗）（治疗）:how to select treatments to offer the patients that do more good than harm and that are worth the efforts and costs of using them.Eight central tasks of clinical work that generate clinical questions:7.Prevention（预防）（预防）:how to re

11、duce the chance of disease by identifying and modifying risk factors and how to diagnose disease early by screening.8.Self-improvement（自我完善）（自我完善）:how to keep up-to-date,improve you clinical skills and run a better,more efficient clinical practice.Step 2-Searching for EvidenceHaving framed your ques

12、tion,the next step is to identify sources of information.These might include:TextbooksJournals and Newsletters.For example:American College of Physicians Journal Club Evidence-based Medicine Scientific reports/Government reports.For example:National Health and Medical Research Council Health Advisor

13、y Committee Step 2-Searching for EvidenceGuidelines.For example:The United States Preventive Services Task Force-Guide to Clinical Preventive Services EMJA Clinical Guidelines Databases.For example:Cochrane Library Best EvidenceBibliographic databases.For example:Medline EMBASEOther internet sources

14、.For example:DARE 202223242526循证医学实践的步骤循证医学实践的步骤 检索相关文献 根据提出的临床问题，确定“检索词”利用各种权威的检索系统检索相关文献。从检索结果中找出与问题关系密切的资料，作为分析评价之用。文献检索只是循证医学实践中的一个环节文献检索只是循证医学实践中的一个环节。27 寻找外部证据的途径寻找外部证据的途径 教科书与专著：教科书与专著：从教科书与专著循证是传统的方法，但不是最佳的方法。因为许多临床的问题，难以从教科书中找到令人满意的解答。各个层次的教科书一般来说已经成熟的理论多体现在教科书，专著也多是每数年再版一次，除成熟的理论外，多有作者比较成熟

15、的经验和观点。对于全科医生和低年资的专科医生，教科书和专著往往是循证的重要途径。即使是高年资的专科医生，也常常需要浏览教科书和专著，以较全面地了解本学科的总体发展。对于特殊临床问题的循证，往往需要从后面几个角度去寻找外部的证据。28医学期刊是循证医学最重要的源泉医学期刊是循证医学最重要的源泉国内外际上有数以万计的医学期刊。循证医学，需要学会快速地检索和获取医学期刊文献。通过互联网络检索是最快获取资料的方法之一，有许多检索方法和网址，其中较常用的之一是美国国家生物学技术信息中心的PubMed网页从互联网络上所能获得的文献的原文是很有限的，图书馆中的期刊是循证医学的源泉。在获得所需要的文章以后，还

16、要运用临床流行病学的方法判断该研究的可靠性和有效性。29一次文献：专著，期刊论文，科技报告，学位论文。二次文献：索引，文摘，目录，相应的数据库。三次文献：综述，评述，进展，现状，发展趋势等期刊文献和百科全书，年鉴，手册等参考工具书。零次文献：实验数据，观测记录，调查材料。30313233PubMed：美国国立医学图书馆创建，收集全球70多个国家和地区3900余种杂志，Evidence-Based Medicine是主题词（MeSH Term）http:/ncbi.nlm.nih.gve/PubMed Cochrane Library：每年四期，http:/中文生物医学文献数据库（CBM）：中国

17、医学科学院医学信息研究所研制 33Search for Primary SourcesUse methodological filters to target the right type of study.For instance,PubMED filters for:therapy diagnosis prognosis aetiology 34查找二级资源Guidelines:UK National Library for Health,NICE,SIGN;US National GuidelinesClearinghouse;Canadian Medical Association;N

18、ew Zealand Guidelines Group.CATs:CAT Crawler Evidence-Based Summaries:Bandolier,Clinical Evidence Structured Abstracts:EBM Online,ACP Journal Club Systematic Reviews:Cochrane Library To search several of the databases simultaneously you can use: 35 证据的检索证据的检索1.EBM数据库2.综合性数据库3.EBM期刊4.临床实践指南5.卫生技术评估36

19、Cochrane图书馆图书馆Cochrane图书馆的资料也是循证医学的一个重要的组成部分。Cochrane医生生前极力倡导，各专业应在世界范围内收随机临床试验的研究，进行Meta分析，向世界各国医生提供临床决策的最佳证据。Cochrane图书馆，以出版光碟的形式向全世界传播治疗学方面的，以Meta分析为主的系统评价形式的研究论文。Cochrane图书馆的内容无疑是非常重要的循证医学方面的来源，因为它为治疗学所提供的决策依据具有较高的可靠性。37 1.Cochrane Library 据库据库Cochrane Library 是获取循证医学证据的主要来源，由Cochrane协作网创建。电子出版物

20、，每年四期，可免费获取文摘。http:/主要有三个子库38 Cochrane Library 的子库(1)The Cochrane Database of Systematic Reviews 考科兰协作网系统评价数据库 该库收录由Cochrane协作网50余个系统综述专业组在统一工作手册指导下完成的系统综述，包括系统综述全文(Review)和研究方案(Protocol)，并随着读者的建议和评论以及新的临床试验的出现不断补充和更新。39 Cochrane Library 的子库(2)The Database of Abstracts of Reviews of Effectiveness 效果

21、评价文摘数据库 该库包括非Cochrane协作网成员发表的普通系统评价的摘要和目录，是对Cochrane系统评价的补充，由英国约克大学的国家卫生服务部评价和传播中心提供。DARE的特点是其系统评价的摘要包括了作者对系统评价质量的评估。与CDSR不同的是它只收集了评论性摘要、题目及出处，而没有全文，并且不一定符合Cochrane系统评价的要求。40 Cochrane Library 的子库(3)The Cochrane CENTRAL Register of Controlled Trials 考科兰临床对照试验注册中心 资料来源于协作网各系统评价小组和其它组织的专业临床试验资料库以及在MEDL

22、INE上被检索出的随机对照试验（RCT）和临床对照试验（CCT）。还包括了全世界Cochrane协作网成员从有关医学杂志会议论文集和其他来源中收集到的CCT报告。41 Cochrane Library 的检索42数据库列表数据库列表检索方式检索方式 2.综合性数据库(1)Ovid数据库()由由4个子库组成个子库组成CCRCT即即考科兰临床对照试验注册中心；ACP Journal Club 由两本期刊组成，它按照循证医学文献要求的标准，从100多种生物医学期刊中选择和提炼相关文献，并对该文献临床应用价值进行评论。CDSR即即考科兰协作网系统评价数据库DARE即即效果评价文摘数据库。43 2.综合

23、性数据库(2)PubMed数据库方法一：字段限定 例：iron deficiency anaemia AND systematic sb4445字段限定字段限定方法二：Clinical Queries46方法三：Limits的文献类型中有Meta分析、实践指南和随机对照试验。47 3.EBM期刊(1)Bandolier 月刊 由英国Oxford HS R&D Directorate于1994年创办，网络版始于1995年，可免费获取全文。提供两种检索方式：Subject Index：浏览学科专题目录，逐层点击，直至所需文献。Search：输入检索词，可检索到Bandolier的所有文献，并以星级

24、表示检出文献的相关度。48 3.EBM期刊(2)ACP Journal Club 双月刊 由美国内科医师协会和美国内科协会联合主办 http:/www.acpjc.org49 3.EBM期刊(3)Evidence-based Medicine 双月刊 由英国医学杂志和美国内科医师协会 联合主办。http:/50 3.EBM期刊(4)循证医学 双月刊 由广东省循证医学科研中心、广东省人 民医院和中山大学附属第三医院主办。重点是肿瘤学领域的循证医学实践。http:/51 4.临床实践指南(1)National Guideline Clearing house(NGC)NGC是由Agency for

25、 Healthcare Research&Quality、American Medical Association和American Association of Health Plans主办的循证医学临床实践指南数据库。提供全文。有检索、浏览和比较等功能。http:/guideline.gov/52Search（检索）可进行Basic Search或Detailed Search方式检索，其检索规则与PubMed相似，支持布尔逻辑组配AND、OR、NOT，短语加“”号，截词符用“*”号。Browse（浏览）分为多个栏目，可通过浏览目录，层层点击直至所需指南。主要有Diseaseconditi

26、on、Treatment/Intervention、Measures/Tools、Organization等栏目。53 Compare guidelines（比较）Add to my Collection：将选中的指南加入指南集合中。Compare Selected Guidelines：对选中的指南进行比较。View Guidelines Collection：浏览当前指南集合Remove Selected Guidelines from Collection：选中欲去除指南前的方框，去除选中指南Remove All Guidelines in Collection：去除指南集合中所有指南。

27、54 4.临床实践指南(2)CMA Infobase：Clinical Practice Guidelineshttp:/www.cma.ca/cpgs/加拿大临床实践指南数据库于加拿大临床实践指南数据库于1995年由加拿大年由加拿大国家、州或地区医学卫生组织，专业协会，政府国家、州或地区医学卫生组织，专业协会，政府机构和专家小组共同主办并认可。它提供多种检机构和专家小组共同主办并认可。它提供多种检索途径：索途径：Keyword Search：输入关键词或短语进行检索：输入关键词或短语进行检索，也可选择也可选择Basic Search或或 Advanced Search方式方式检索：检索：55

28、BrowseBy developers：从创建：从创建guideline的机构名着手进行检索。的机构名着手进行检索。Recent additions：最近新增加的：最近新增加的guideline.NewsCurrent health topic：从最新卫生专题着手检索。：从最新卫生专题着手检索。In the news：新闻。：新闻。ResourcesFor developers：机构的信息。：机构的信息。Other CPG databases：其他国家则临床实践指南数据库。：其他国家则临床实践指南数据库。FAQs：常见问题解答。：常见问题解答。56http:/www.cma.ca/cpgs/5

29、.卫生技术评估(1)INAHTA 国际卫生技术评估网(The International Network of Agencies for Health Technology Assessment)成立于1993年，秘书处在瑞典。http:/www.inahta.org(2)ISTAHC 国际卫生技术评估会 成立于1985年，秘书处设在加拿大。http:/www.istahc.org57 检索实例1一名内科医生在临床实践中提出问题：是否能够将溶栓联合冠状动脉介入治疗急性心肌梗塞，目前有无充分的相关证据？58该医生检索文献的目的并非广泛收集信息，进行科学研究，而是希望能够获取解决临床问题的最佳证据

30、，因此属于循证医学实践范畴。从查全率和查准率的角度看，查准更为重要，因此，在检索时应首先查看是否有相关的高质量的临床实践指南、系统评价和Meta分析。若无，再查看其他等级证据，如单个样本量足够的随机对照实验、设对照组但未用随机方法分组的研究等。59n选词选词急性心肌梗塞（急性心肌梗塞（Acute myocardial infarctionAcute myocardial infarction）血栓溶解疗法（血栓溶解疗法（Thrombolytic therapyThrombolytic therapy）急诊冠状动脉介入治疗（急诊冠状动脉介入治疗（Primary coronary interven

31、tionPrimary coronary intervention）经皮经腔冠状动脉血管成形术（经皮经腔冠状动脉血管成形术（Percutaneous coronary Percutaneous coronary transluminal angioplastytransluminal angioplasty）60数据库的选择数据库的选择National Guideline Clearinghouse National Guideline Clearinghouse http:/www.guideline.gov/http:/www.guideline.gov/PubMedPubMed61Coc

32、hrane review 示例62Selective serotonin reuptake inhibitors(SSRIs)for stroke recovery:a Cochrane reviewClinical64Clinical questionDo selective serotonin reuptake inhibitors(SSRIs)prescribed during the first year after stroke improve recovery at the end of treatment and the end of follow-up?Source:Mead

33、GE,Hsieh CF,Lee R,Kutlubaev MA,Claxton A,Hankey GJ,Hackett ML.Selective serotonin reuptake inhibitors(SSRIs)for stroke recovery.Cochrane Database of Systematic Reviews 2012,Issue 11.Art.No.:CD009286.DOI:10.1002/14651858.CD009286.pub2.65Context Stroke is the major cause of adult disability.Selective

34、serotonin reuptake inhibitors(SSRIs)have been used for many years to manage depression,including in people who have had a stroke.Recently,small trials have demonstrated that SSRIs might improve recovery after stroke,even in people who are not depressed.66MethodsThe registers of the Cochrane Stroke a

35、nd Depression Anxiety and Neurosis Groups were searched,along with the Cochrane Central Register of Controlled Trials,MEDLINE,EMBASE,CINAHL,AMED,PsycINFO and PsychBITE.Trial registries and pharmaceutical websites were examined,reference lists were checked and citation tracking was done of included s

36、tudies.Meta-analyses used the random effects models,with risk ratios(RR)for dichotomous data and standardised mean differences(SMD)for ordinal and continuous data.The Cochrane Risk of Bias tool was applied,with sensitivity analyses conducted that restricted each meta-analysis to trials at low risk o

37、f bias for each quality criterion.67PICO(S)to assess eligible studiesParticipants:People who had had a stroke(not a transient ischaemic attack,TIA)in the previous 12 months.Intervention:Selective serotonin reuptake inhibitor(SSRI)of any type and duration,used for any indication.Comparison:Placebo or

38、 usual care without the SSRI.Outcomes:Primary outcomes:dependence and disability.Secondary outcomes:neurological impairments overall,cognition,motor function,depression,anxiety,quality of life,death,adverse events,premature trial withdrawal,fatigue and healthcare costs.Studies:Randomised trials.68De

39、scription of eligible studies52 studies(4059 participants)were included in the meta-analyses,with most recruited people having had an ischaemic stroke.The most common SSRI was fluoxetine,which was used in 28 trials.Trials were from USA,Europe,Australia and China.Mean age of participants in each tria

40、l ranged from 55 to 77 years.Most studies excluded people who could not consent for themselves(e.g.due to aphasia or confusion).69Results primary outcomesOne trial(112 participants)provided data on dependency at the end of treatment,RR:0.81(95%CI 0.68 to 0.97).22 trials(1343 participants)could be co

41、mbined in the meta-analysis of disability score at the end of treatment,SMD:0.91(95%CI 0.60 to 1.22);but with high heterogeneity(I2=87%).Only 7 trials provided data on follow-up after the end of treatment.70Results disability at end of Secondary outcomes at end of OutcomeTrials(participants)Effect s

42、ize(95%CI)Heterogeneity(I2)Neurological score29(n=2011)SMD-1.00(-1.26,-0.75)86%Depression(dichotomous)8(n=771)RR 0.43(0.24,0.77)77%Depression(continuous)39(n=2728)SMD-1.91(-2.34,-1.48)95%Anxiety8(n=413)SMD-0.77(-1.52,-0.02)92%Cognition7(n=425)SMD 0.32(-0.23,0.86)86%Motor deficits 2(n=145)SMD-0.33(-1

43、.22,0.56)n/aDeath46(n=3344)RR 0.76(0.34,1.70)0%Leaving trial early49(n=3851)RR 1.02(0.86,1.21)31%Seizures7(n=444)RR 2.67(0.61,11.63)0%GI side effects14(n=902)RR 1.90(0.94,3.85)31%Bleeding 2(n=249)RR 1.63(0.20,13.05)0%Follow-up after the end of treatmentOutcomeTrials(participants)Effect size(95%CI)Di

44、sability2(n=155)SMD 1.78(-1.01,4.57)Neurological deficit 4(n=275)SMD-0.63(-1.30,0.04)Depression(continuous)4(n=275)SMD-1.10(-2.16,0.04)Depression(dichotomous)1(n=99)RR 0.77(0.35,1.76)Cognition1(n=99)SMD 0.02(-0.37,0.42)Death3(n=257)RR 0.08(0.00,1.41)Sensitivity analyses for disability at end of trea

45、tment,restricted to trials at low risk of Quality criterion SMD(95%CI)for disabilityAll trials(i.e.no quality criteria applied)0.91(0.60,1.22)Randomisation0.32(0.06,0.58)Allocation concealed0.70(-0.73,2.13)Patient/personnel blind0.29(0.04,0.53)Outcome assessor blind0.78(0.04,1.52)Incomplete outcome

46、data 0.77(0.22,1.33)Selective reporting0.11(-0.26,0.47)Conclusions for practiceAt the end of treatment,SSRIs reduced dependency,disability,neurological deficit,depression,and anxiety;but with substantial statistical heterogeneity and multiple sources of bias.There are no data on fatigue or costs.The

47、re are insufficient data to draw conclusions about long-term outcome.Nevertheless,SSRIs are a promising intervention to facilitate stroke Future researchFurther,large-scale randomised trials are needed find out whether fluoxetine improves stroke recovery.Three such trials areFOCUS(www.focustrial.org

48、.uk)in the UKAFFINITY(www.affinitytrial.org)in AustraliaEFFECTS(www.effects.se)in SUseful linksCochrane Journal Club discussion pointsSelective serotonin reuptake inhibitors(SSRIs)for stroke recovery荟萃分析（Meta analysis)77Meta-AnalysisMeta-Analysis的基本原理的基本原理 在meta-analysis中的每个研究都被看做是一个体层面，并应用统计方法把所有层结合起来并且获得总的估计效果量。但不要忘记这样估计的可信性和有用性是与每个研究相关的。进行Meta-Analysis的步骤我们应该怎样编码数据和萃取数据？不同研究中对干预、暴露及结果的表述是足够相似的可以合并吗？我们怎样来评价每个研究的质量？识别出作用大小（Identifying the effect size(ES)）合并研究 展示结果 解释异质性