【文档说明】循证医学EvidenceBasedClinicalPracticeZSUFS课件.pptx，共(63)页，2.445 MB，由小橙橙上传

转载请保留链接：https://www.ichengzhen.cn/view-254967.html

以下为本文档部分文字说明：

Evidence-BasedMedicine循证医学杨学宁XueningYang,MD广东省人民医院广东省肺癌研究所广东省循证医学科研中心循证医学杂志社Evidence-basedPractice(EBP)Models•Traditionalpr

acticeModel(×)•Model1:Usethehighestqualityinformationtoguideclinicaldecisions•Model2:Search,evaluate,an

dmakeavailablespecialtyspecificbestavailableevidence•Model3:CreateoriginalresearchorsystematicreviewsEBMPr

actice-FivestepsPatientHistoryandPE……•QuestionAskingAnswerableClinicalQuestions.•SearchSearchingforthebestcurrentevidence.•AppraiseCritically-apprais

ingtheevidence.•Application:Integratingtheevidencewithourexpertiseandourpatient’suniquebiologyandvalues•AuditEvaluatingourperformanceStep1Ask

ingAnswerableClinicalQuestions•Setting:64residentsat2NewHavenhospitals•Method:Interviewedafter401consulta

tions•Questions–Asked280questions(2per3patients)–Pursuedananswerfor80questions(29%)–Notpursuedbecause•Lackoftime•Forgottheque

stion•Sourcesofanswers–Textbooks(31%),articles(21%),consultants(17%)Green,AmJMed2000Step1AskingAnswerableClinicalQuestions•Mostofourquestions

areneveranswered•Whenanswered,theinformationislikelytobeneitherthebestnorup-to-dateStep1AskingAnswerableClinicalQuesti

ons•EBMstartswithaclearclinicalquestion•Well-builtclinicalquestionsguidethewholeEBMprocess•Typeofquestion?–Causation/etio

logy–Intervention–Diagnosis–Prognosis•TypeofStudy?–Whatwouldbethebeststudydesigninordertoanswerthequestion?–studydesignsshouldlimitbias

Step1AskingAnswerableClinicalQuestions•Questioncomponents:PICO–WhattypesofParticipants?–WhattypesofInterventions?–WhattypesofComparison?–Whatt

ypesofOutcomes?Step1AskingAnswerableClinicalQuestions•Whattypesofparticipants?–Diseaseorconditionofinterest–Potentialco-morbidity–Setting–Demo

graphicfactors–TaskofclinicalproblemStep1AskingAnswerableClinicalQuestions•Whattypesofintervention?–Typesoftreatment

–Typeofdiagnostictest–Typeofcausativeagent–TypeofprognosticfactorStep1AskingAnswerableClinicalQuestions•Whattypesofoutcomes?–Forcausation

,outcomeisdisease–Fordiagnostictest,outcomeindiseaseofinterest–Fortreatmentincludealloutcomesimportanttopeo

plemakingdecisionstodefinesuccessoftherapy–Forprognosis,outcomeisthechosenendpointofthediseaseStep1AskingAnswerableClinical

Questions•Broadornarrowquestions–Broad•Dogefinitibimprovesurvivalinlungadenocarcinoma?–Narrow•Dogefinitibimprovesurvivalinfemallunga

denocarcinomarefractorytofirst-linechemo?Step1AskingAnswerableClinicalQuestions•Causation–People-Exposure-Comparison-Outcomes•Doesheavysmoker

havehigherincidenceoflungcancer?–Prognosis:People-Exposure-Comparison-Outcomes•IslungadenocarcinomawithhighlevelofCEAaremorelikelytohaverec

urrentdisease?Step2Searchingforthebestcurrentevidence•Textbook–Annuallyrevised–Usehighlevelevidences–Withreference•Guidelines•CochraneLibrary•

MedlineNationalGuidelineClearinghouse(http://www.guideline.gov/)Step2Toomuch?•即使是小的问题也有相关指南Step2Step2Coch

raneLibrary•Advantage–highestqualityevidencewe’lleverhaveontheeffectivenessofhealthcare•Disadvantage–notyetmanysystemicreviews–necessaril

yomitsthenewesttreatmentStep2EBPModels•TraditionalpracticeModel(×)•Model1:Usethehighestqualityinformationtoguideclinicaldecisions•Model

2:Search,evaluate,andmakeavailablespecialtyspecificLevel1information•Model3:Createoriginalresearchorsystemati

creviewsStep2PubMed/MEDLINE•http://www.ncbi.nlm.nih.gov/PubmedStep2Medline-PubMedAdvantage–exhaustiveness,flexibilityofsearching–journalc

overage,currency(on-lineversions)–widespreadavailabilityandsupportDisadvantage–havetodoyourownqualityfilteri

ng–puttingtogethergoodsearchesisdifficult–gapsincoverage(medical,geographicalandlinguistic)Step2Step2Step2Step2St

ep2Step2Step2Searchingforthebestcurrentevidence•Forprimaryliterature–Notjustabstract!–TrytogetFull-textStep3Critically-appraisingtheevidence•Va

lid?–Designandmethodology:logical?–Sample–ExperimentorMeasure:carriedoutinasoundwayandConvincible？Follow-

uprate?–Analysis–Conclusion–Ethical?•Importance?–Clinicalsignificance？–Cost-effective？•Applicable?–Willtheresultshelploc

ally?Step3Critically-appraisingtheevidence•Quickscreeningbytitle,abstractandrelatemessage–Authors•Well-known？Pastrecord？Creditable？–In

stitute•Competencetodothisstudy？Well-known？–Specialty•Hasorhadpublishedincorejournals?Yourpractice？Valuable？•GoforFull-textDis

crimination？Rational？EBM？Step3Critically-appraisingtheevidenceTypeofQuestionSuggestedbesttypeofStudyTherapyRCT>cohort

>casecontrol>caseseriesDiagnosisProspective,blindcomparisontoagoldstandardEtiology/HarmRCT>cohort>casecontrol>caseseriesPrognosisCohortstud

y>casecontrol>caseseriesPreventionRCT>cohortstudy>casecontrol>caseseriesClinicalExamProspective,blindcomparisontogoldstandardCostEconomicanal

ysisRightmethodsiscrucialTheEvidencePyramidHierarchyofevidence:arrangesstudydesignsbytheirsusceptibilitytobiasPrimaryResearch

ExperimentalobservationalSecondaryResearch•Forquantitativestudies•ForqualitativestudiesNoconsensusabouttherelativerigourofdifferentmethod

sSystematicreviewStep3Critically-appraisingtheevidenceImprovedsurvivalandsamplesizedControlSurvival（%）5%10%15%101,290394134151,89454

4208202,444676328252,926788374PhaseIIIRCTheadtoheadComparesurvivalAdequatesamplesizeStep3Critically-appraisingtheevidence•

Measure–Objective：Labtest，Radiology–Subjective：pain，symptoms•Endpoint–alternative：relieveofsymptom,shrinkoftumor–GoldStandard：prolongsurvivalStep3

Critically-appraisingtheevidence•Conclusion–Basedonthedata?–New?•Updated?•Comparewithotherstudy•And,Whatnext?Step3Critically-appraisingtheevidence

•Attention!–Don’ttrustitjustbecauseitisfromabigboss,meta-analysis,RCT,orguideline.–Evidenceshouldbeinconsistencewi

thknownmedicalscience(physiology,pathology,pharmacology,etc）UsageofscientificlanguageisnotnecessarilymeanscientificbasedScientificmeth

odscreditNotalltheevidencesarethesameimportanceHollingworth&Jarvik,Radiology2007;244:31-38Flowchartofcost-e

ffectivenessateachlevelofTAhierarchyTechnicalPerformanceDiagnosticPerformanceDiagnosticImpactTherapeuticImpactI

mpactonHealthStep4Application•Integratingtheevidencewithourexpertiseandourpatient’suniquebiologyandvalues–Yourquestion–Yourpatient–

Yourexpertise–Yourhospital–Howto……Step5Evaluatingourperformance•Follow-up•Evaluate–Thepatientsbenefit……?•Next……

•……Oragainanother5steps•CollectdataandwriteanarticleEBPModels•TraditionalpracticeModel(×)•Model1:Usethehighestqualityinformationtoguideclinicaldecis

ions•Model2:Search,evaluate,andmakeavailablespecialtyspecificbestavailableevidence•Model3:Createoriginalresearchorsystematicre

viewsSelectiveEBP•Maynotneedtocarryoutall5steps•1.AskingAnswerableClinicalQuestions.•X2.Searchingforthebestcurrentevidence.•X3.Critically-ap

praisingtheevidence.•4.Application:Integratingtheevidencewithourexpertiseandourpatient’suniquebiologyandvalues•5.Evaluatingou

rperformanceSelectiveEBP•3differentmodesofEBP–Searching&appraising•providesE-Bcare,butisexpensiveintimeandresources–S

earchingonly•much,quicker,andifcarriedoutamongE-Bresources,canprovideE-Bcare–Replicatingthepracticeofexpert

s•quickest,butmaynotdistinguishevidence-basedfromego-basedrecommendationsSelectiveEBP•EvenfullyEBM-trainedclinician

sworkinall3modes–Searching&appraising•fortheproblemsIencounterdaily.–Searchingonly•forproblemsIencounteronceamonth.–Replicating•forproblemsIe

ncounteronceadecade(andcrossingmyfingers!).Case1•Male,ZSUProfessor，62yearsold•Smallcelllungcancer,limiteddisease,chemofor6cycle

swithcompleteresponse–Medicaloncologist：prophylacticcranialirradiation（PCI）notimporvesurvival,andwithAE！–Radio-oncologist：PCIimprove

survival,andtolerable•Question:Why?ShouldthepatientbegivePCI?Improvesurvival?Howmuch?Sideeffect？When?How?TextbookPublish20061.Cha

nceofSCLCbrainmet：50%-80%for2years2.1995，aMETA-analysis：SCLCwithCR,PCIimprove3yearsurvival15%-21%，butnot

analysisthecomplication3.Reducebrainmetfrom30%→5%☆Conclusion：SCLCwithCR,PCIisrecommendafter2cyclesofche

moComplications?case1Guideline&SR•NCCN–LungcancerandProphylacticcranialirradiation1–2002TheCochraneLibrary,Issue2•Prophylacticcranialirr

adiationsignificantlyimprovessurvivalanddisease-freesurvivalforpatientswithsmallcelllungcancerincompleteremission.case

1例1FromSecondarySource•PCIshouldstartearlyafterchemo•complications：–notRCT–2studyreportnodifferenceinmentalo

rneurologicaldisorder,butonlystudypartofthestudypatients，40%and60%（1995，1997）•N=300•N=314–InoneRCT（N=46）,reportamentaldisorder（1997）case1PUBME

D•"CranialIrradiation"[MESH]ANDlungneoplasms[mh]•("CranialIrradiation/adverseeffects"[MESH]OR"CranialIrradiation

/mortality"[MESH])ANDlungneoplasms[mh]☆25papers，3papersafter2006，onrelatepaper（level5）例1PUBMED•2009,JCOpeer-revi

ew•Complications•Mentalandneurological•Increasebyconcurrentchemoirradiation•Increaseindoselevel≥40Gy，no

morethan30Gy，recommend20Gy–SupportPCIinSCLClimiteddiseasewithchemoCR.例1Conclusion&Decision•PCIreducebrainmet,

improvesurvival•WhenRT<30GY,rarecomplications•Morebenefitthanharm•Recommendation：PCI，rightnowcase1EventsLikeThis……Sanlubabyformulami

lkpowdercontaminatewithmelamine•Jul.16,2008–甘肃省卫生厅接电话报告，称今年该院收治的婴儿患肾结石病例明显增多，近几个月已达十几例，经了解均食用了三鹿牌配方奶粉。•Sept.11,2008–新华社报道：“高度怀疑”三鹿公司生产的三鹿牌婴幼

儿配方奶粉受到三聚氰胺污染。case2Sanlubabyformulamilkpowdercontaminatewithmelamine•Sept.17,2008国务院有关部门负责人和科学家–国家质检总局局长李长江：……共检

验了109家企业的491批次产品，有22家企业的69个批次的产品检出了三聚氰胺，检出不合格产品的企业约为20％……需要指出的是，这次被检出含有三聚氰胺产品的企业，不是所有批次的婴幼儿奶粉都有问题，比如，伊利、蒙牛、雅士利、施恩等企业被抽检的产品有几十个批次，但有问题的只有

几个批次。case2Sanlubabyformulamilkpowdercontaminatewithmelamine–卫生部陈竺部长：三聚氰胺……主要对膀胱和肾脏有影响，引发动物膀胱炎、膀胱结石、肾脏炎症等，但没有发现对人类有致癌作用。婴幼儿食用……出现不

明原因的哭闹、呕吐、发热、尿液混浊、血尿、少尿或无尿等症状，应立即就近到医疗机构筛查就诊。食用含有三聚氰胺的其他品牌婴幼儿奶粉，只有出现上述症状时，才需要立即到医院筛查就诊。–从2008年9月12日至17日8时，各地报告临床诊断患

儿一共有6244例。其中，4917例患儿症状轻微，生命体征稳定，正在进行院外随访治疗或已经治愈。现仍留院观察治疗患儿有1327人。所有临床诊断病历中，有158人发生过急性肾功能衰竭，占总病例数的2.5%。经有效治疗，目前158例重症患儿当中，已经有94人病情平稳或已经治愈出院。此外，回

顾性的调查发现有3例死亡病例，其中甘肃省有2名，浙江有1名，不包括在刚才提到的6244名诊断病例之中。case2TheEvidencePyramidHierarchyofevidence:arrangesstudydesignsbytheirsusceptibilitytobiasPrimaryR

esearchExperimentalobservationalSecondaryResearch•Forquantitativestudies•ForqualitativestudiesNoconsensusabouttherelativerigourofdifferentmeth

odsSystematicreviewSanlubabyformulamilkpowdercontaminatewithmelamine•Celloranimalisnoequaltohuman•Butther

eareanycorelation?•Noevidencedoesnotmeannoharm!•Howshouldweact?Summary•EBMisaLifelonglearningmodel–lifelong,self-directed,proble

m-basedlearning•Know-howinfinding,appraising,andapplyingevidenceFurtherReadingonEvidenceBasedMedicine