meta分析的選題、寫作與投稿第三章Meta分析寫作孫敏(猴哥meta)PART05methods：打鐵還需自身硬（中）解讀MOOSE報(bào)告規(guī)范methods的五個(gè)要素1、檢索2、篩選3、提取4、評(píng)價(jià)5、統(tǒng)計(jì)

文獻(xiàn)檢索SearchStrategy納入與排除InclusionandExclusionCriteria數(shù)據(jù)提取DataExtraction質(zhì)量評(píng)價(jià)QualityAssessmentofIncludedStudies統(tǒng)計(jì)方法StatisticalAnalysisJAMA、上市藥品臨床安全性文獻(xiàn)評(píng)價(jià)需要MOOSE聲明在PRISMA規(guī)范推出之前，還有用于觀察性研究的Meta分析規(guī)范MOOSE聲明有雜志（如JAMA）要求對(duì)觀察性研究的Meta分析稿件需參照MOOSE聲明ChinatoinstructMarketingAuthorizationHolder(MAH)forclinicalsafetyliteratureresearchandevaluationonpost-marketdrugs指導(dǎo)藥品上市許可持有人（MAH）開(kāi)展上市藥品臨床安全性文獻(xiàn)評(píng)價(jià)NationalMedicalProductsAdministration(NMPA)releasedthe“GuidelineofClinicalSafetyLiteratureResearchandEvaluationonPost-marketDrugs(Draft)”on18Jun,2019.[1]國(guó)家藥品監(jiān)督管理局于2019年6月18日發(fā)布《上市藥品臨床安全性文獻(xiàn)評(píng)價(jià)指導(dǎo)原則（試行）》JAMA、上市藥品臨床安全性文獻(xiàn)評(píng)價(jià)需要MOOSE聲明在PRISMA規(guī)范推出之前，還有用于觀察性研究的Meta分析規(guī)范MOOSE聲明。有雜志（如JAMA）要求對(duì)觀察性研究的Meta分析稿件需參照MOOSE聲明ThisguidelineprovidesageneralprocessforMAHtoconductpost-marketdrugs’clinicalsafetyliteratureresearchandevaluationwhichincludesdefiningtheevaluationpurposes,preparingandfinalizingresearchprogram,researchingforevidences,screeningliteraturearticles,processinginformation,evaluatingqualityofthearticles’content,analyzinginformation,andproducingreports.ThisguidelinerecommendsMAHtoassigntwoevaluatorstoconducttheresearchtasksandgeneratethefinalevaluationreportasperPreferredReportingItemsforSystematicreviewsandMeta-Analyses(PRISMA)andMeta-analysisofObservationalStudiesinEpidemiology(MOOSE).此指導(dǎo)原則描述藥品上市許可持有人開(kāi)展上市藥品臨床安全性文獻(xiàn)評(píng)價(jià)的一般流程，包括確定評(píng)價(jià)目的、制定和完善研究方案、全面查找證據(jù)、文獻(xiàn)篩選、資料提取、文獻(xiàn)質(zhì)量評(píng)價(jià)、資料分析、形成報(bào)告。鼓勵(lì)由兩名評(píng)價(jià)員獨(dú)立進(jìn)行文獻(xiàn)評(píng)價(jià)，并參考使用系統(tǒng)綜述/Meta分析優(yōu)先報(bào)告條目（PRISMA）和流行病學(xué)觀察性研究的Meta分析（MOOSE）撰寫文獻(xiàn)評(píng)價(jià)報(bào)告MOOSE:觀察性臨床實(shí)驗(yàn)的Meta分析報(bào)告規(guī)范MOOSE(Meta-analysisofObservationalStudiesinEpidemiology),MOOSE工具的主要方面包括:清晰定義研究人群；清晰定義結(jié)局以及結(jié)局評(píng)估；獨(dú)立評(píng)估結(jié)局參數(shù)；足夠的隨訪；隨訪時(shí)無(wú)選擇性失訪；識(shí)別重要混雜因素和預(yù)后因素JAMA.2000;283(15):2008-2012.doi:10.1001/jama.283.15.2008MOOSE:觀察性臨床實(shí)驗(yàn)的Meta分析報(bào)告規(guī)范最近發(fā)表的外科領(lǐng)域系統(tǒng)評(píng)價(jià)中的風(fēng)險(xiǎn)評(píng)估表單對(duì)于不符合預(yù)先設(shè)定的質(zhì)量要求的研究，需要謹(jǐn)慎考慮是否在進(jìn)一步的統(tǒng)計(jì)分析中納入，比如可以通過(guò)敏感性分析識(shí)別異質(zhì)性或通過(guò)累積Meta分析來(lái)識(shí)別時(shí)間對(duì)效應(yīng)量的影響強(qiáng)烈推薦采用這些清單來(lái)嚴(yán)格評(píng)估納入研究的質(zhì)量。完整清單應(yīng)該以文中表格或者補(bǔ)充表格形式出現(xiàn)在系統(tǒng)評(píng)價(jià)中。應(yīng)該由至少兩個(gè)評(píng)價(jià)者獨(dú)立進(jìn)行偏倚風(fēng)險(xiǎn)評(píng)估，而且任何差異都應(yīng)該通過(guò)討論達(dá)成共識(shí)選自《傻瓜統(tǒng)計(jì)學(xué)》,特此感謝構(gòu)建MOOSE框架MOOSE框架：英文文獻(xiàn)示例（范文見(jiàn)附件）AlmohmeedYH,AvenellA,AucottL,VickersMA(2013)SystematicReviewandMeta-AnalysisoftheSero-EpidemiologicalAssociationbetweenEpsteinBarrVirusandMultipleSclerosis.PLoSONE8(4):e61110.doi:10.1371/journal.pone.0061110構(gòu)建MOOSE框架構(gòu)建MOOSE框架：第1部分CriteriaBriefdescriptionofhowthecriteriawerehandledinthemeta-analysisReportingofbackgroundshouldincludeProblemdefinitionAroleforEpsteinBarrvirus(EBV)inthedevelopmentofMultipleSclerosis(MS)hasbeensuggestedbyvarioussero-epidemiologicalstudiesthatcomparedtheprevalenceofanti-EBVantibodiesinserumofMScasesandcontrols,withanalmostconsistentfindingofhigherprevalenceinMSpatients.ThelastreviewthatupdatedtheassociationbetweenMSandsero-positivityofdifferentanti-EBVantibodieswasthestudyofSantiago,etal.,(2010).Manysero-epidemiologicalstudieshavebeenpublishedsincethenandthereisaneedforamoreuptodatereviewoftheevidencewithamoreinclusivesearchstrategy.HypothesisstatementMScasesareassociatedwithhigherprevalenceofanti-EBVantibodiesthancontrolsDescriptionofstudyoutcomesPrevalenceofanti-EBVIgGantibodiesinserumofMScasescomparedtocontrolsOddsratio(OR)ofexposuretoanti-EBVantibodiesinMScasescomparedtocontrolsTypeofexposureorinterventionusedIgGantibodiesagainstthefollowingEBVantigens:EpsteinBarrNuclearAntigen1(EBNA1)ViralCapsidAntigen(VCA)EarlyAntigen(EA)TypeofstudydesignsusedCase-controlandnestedcasescontrolamongcohortstudies,thatexaminedtheprevalenceofanti-EBVantibodies(IgGonly)inserumofMScasesandnonMScontrolsStudypopulationNorestrictionapplied構(gòu)建MOOSE框架：第2部分ReportingofsearchstrategyshouldincludeQualificationsofsearchersThecredentialsofthethreeinvestigators(whocontributedtothesearchstrategy)YA,AAandCFareindicatedintheauthorslistandtheAcknowledgmentssection.Searchstrategy,includingtimeperiodMedlinefrom1960–March2012includedinthesynthesisandkeywordsEMBASEfrom1960–March2012SeeMethodologyandAppendix1DatabasesandregistriessearchedMedlineandEMBASESearchsoftwareused,nameandOvidSP,WedetailedtheMeSH/EmtreeHeadingsandtextwordsinversion,includingspecialfeaturesAppendix1UseofhandsearchingBibliographiesoftheretrievedpapers(onlytheincludedstudies)werehandsearchedforadditionalreferences,ListofcitationslocatedandthoseDetailsoftheliteraturesearchprocessareoutlinedinthePRISMAflowexcluded,includingjustificationschart.ThecitationlistofexcludedarticlesisavailableuponrequestMethodofaddressingarticlesWeplacednorestrictionsonlanguage;WewereabletoobtainedallpublishedinlanguagesotherthanarticlespotentiallyeligibleforinclusioninEnglishlanguageEnglishMethodofhandlingabstractsandWecontactedanumberofauthorsforfullreportofrelevantunpublishedstudiesunpublishedstudiesDescriptionofanycontactwithauthorsWecontactedauthorsofrelevantarticlesforfullreportoftheirunpublishedstudies,orforextradataaboutanti-EBVantibodiesprevalenceincasesandcontrols.FordetailsoftheauthorswhoprovidedextraunpublishedinformationseetheAcknowledgmentssection構(gòu)建MOOSE框架：第3部分ReportingofmethodsshouldincludeDescriptionofrelevanceorDetailedinclusionandexclusioncriteriaaredescribedinthepaperappropriatenessofstudiesassembledforassessingthehypothesistobetestedRationalefortheselectionandcodingofAdataextractionsheetwasdeveloped(availableonrequest).Datadataextractedwererelatedtobibliographicdetailsofincludedstudy,methodofidentificationofthestudy,Characteristicsofcases/controls,OutcomesandqualityassessmentAssessmentofconfoundingWeconductedeightsubgroupanalysestocomparetheeffectsofanumberofpotentialconfoundersontheORofexposuretoanti-EBVantibodies.Assessmentofstudyquality,includingWeusedamodifiedversionoftheNewcastleOttawaScale(NOS)toblindingofqualityassessors;assessthequalityofeachstudy.WeconductedasubgroupanalysisstratificationorregressiononpossiblecomparingthestudieswhichscoredabovethemedianintheNOSpredictorsofstudyresultsscaletothosescoringbelowthemedianAssessmentofheterogeneityWeusedtheI2valuetoassessheterogeneityDescriptionofstatisticalmethodsinWementionedtypeofanalysisweused(meta-analysisandsufficientdetailtobereplicatedsubgroupmeta-analysis)andtypeofsoftwareweused(ReviewManager5)ProvisionofappropriatetablesandWeincludedPRISMAflowcharttoshowthemethodofstudiesgraphicsidentification,Table1showingcharacteristicsofincludedstudies,Table2–Table9showingResultsofSubgroupanalysis,Table10showingresultsofQualityassessment,Fourforestplotsofthemaindifferentmeta-analysesconductedandThreeFunnelplotsforpublicationbiasassessment.構(gòu)建MOOSE框架：第4部分ReportingofresultsshouldincludeGraphsummarizingindividualFigure2,Figure4,Figure6andFigure8studyestimatesandoverallestimateTablegivingdescriptiveTable1informationforeachstudyincludedResultsofsensitivitytestingTable2–Table9Indicationofstatistical95%and99%CIintervalswerepresentedforuncertaintyoffindingsallanalysestogetherwithI2valuesforthethreemainmeta-analyses構(gòu)建MOOSE框架：第5部分ReportingofdiscussionshouldincludeQuantitativeassessmentofResultsofsubgroupanalysesarediscussedwithbiasmainpotentialconfoundingfactorsdiscussed.ResultsofFunnelplotandriskofpublicationbiasishighlightedJustificationforexclusionReasonsforexclusionwerereportedmainlyintheresultssection,withthetwomainreasonsbeingeitherthatstudieswereirrelevant(Didnotmeasureoutcomeofinterest)ordidnothaveclearinformationwithnoextradataobtainable.Likelihoodofpublicationbiaswasdiscussed.AssessmentofqualityofImplicationofthesubgroupanalysesrelatedtoincludedstudiesqualityassessmentwashighlighted構(gòu)建MOOSE框架：第6部分ReportingofconclusionsshouldincludeConsiderationofalternativeWementionedinthediscussionthatincludedexplanationsforobservedpatientswithunconfirmedMSdiagnosis(clinicallyresultsisolatedsyndrome)couldhavealteredtheresultsGeneralizationoftheWereportedthefactthatalmostallofthestudiesconclusionswerefromEurope,NorthAmericaandAustralia.Wehighlightedthattherefew/nonefoundfromAsiaandAfricaGuidelinesforfutureFurtherresearchrelatedtoEBVandMScouldfocusresearchoncharacterisingtheriskofMSbasedontitresofanti-EBVantibodies.AlthoughtherehavebeenstudiespublishedwhichexaminedtitrelevelsandMSrisk,ameta-analysiscombiningdatafromdifferentstudiesisyettobeconductedDisclosureoffundingsourceNofundingwasrequiredforconductingthisreview范文解析：風(fēng)險(xiǎn)因素類風(fēng)險(xiǎn)因素類meta分析范文：本Meta分析遵照MOOSE規(guī)范進(jìn)行報(bào)告PassiveSmokingandCervicalCancerRisk:AMeta-analysisBasedon3,230Casesand2,982Controls被動(dòng)吸煙與宮頸癌的風(fēng)險(xiǎn)：基于3230個(gè)病例和2982個(gè)對(duì)照的Meta分析摘要目的：被動(dòng)吸煙已被認(rèn)為是多種癌癥的危險(xiǎn)因素。為了檢查它是否也可能對(duì)宮頸癌構(gòu)成風(fēng)險(xiǎn)，我們?cè)谝寻l(fā)表的病例對(duì)照研究基礎(chǔ)上進(jìn)行了薈萃分析。方法：計(jì)算機(jī)檢索PubMed數(shù)據(jù)庫(kù)及截至2012年2月10日的納入研究的參考文獻(xiàn)，對(duì)相關(guān)研究進(jìn)行檢索。在兩位作者獨(dú)立評(píng)估方法學(xué)質(zhì)量和提取數(shù)據(jù)后，使用CMAv2軟件進(jìn)行meta分析。發(fā)表偏倚采用漏斗圖，采用Egger’s和Begg’s檢驗(yàn)。結(jié)果：最終獲得11項(xiàng)合格研究，涉及3230例病例和2982例對(duì)照。結(jié)果顯示，被動(dòng)吸煙的婦女患宮頸癌的風(fēng)險(xiǎn)比不吸煙的婦女高73%(OR=1.7395%CI=1.35-2.21，P<0.001)。亞組和敏感性分析表明，這一結(jié)果是穩(wěn)健的。通過(guò)直觀的漏斗圖、Egger’s和Begg’s試驗(yàn)檢測(cè)到中度發(fā)表偏倚。結(jié)論：根據(jù)現(xiàn)有證據(jù)，這項(xiàng)薈萃分析的結(jié)果表明，被動(dòng)吸煙顯著且獨(dú)立地增加了宮頸癌的風(fēng)險(xiǎn)ZengXT,AsianPacJCancerPrev.2012;13(6):2687-93. /bbs/newweb/pc/post/42932160范文解析：風(fēng)險(xiǎn)因素類LiteraturessearchWeinitiallyidentifiedpublishedandunpublishedstudieswhichtestedtheassociationbetweenpassivesmokingandriskofcervicalcancerbysearchingthePUBMEDdatabasesfromJanuary1st,1988toFebruary10th,2012.Thefollowingsearchtermswereused:(1)“cervicalcancer”or“cervicalcarcinoma”or“uterinecervixcancer”or“CC”or“cervicalneoplasia”;(2)“secondhandsmoking”or“environmentaltobaccosmoke”or“ETS”or“passivesmoking”or“tobaccosmokepollution”;(3)“casecontrol”or“incidence”or“prognosis”or“earlydiagnosis”or“survivalanalysis”or“case-control”.ThesesearchthemeswerecombinedusingtheBooleanoperator“and”inseveralcombinationswithoutrestrictions.Inaddition,wealsoreviewedthereferencelistsofretrievedpapersandrecentreviews.ZengXT,AsianPacJCancerPrev.2012;13(6):2687-93.

文獻(xiàn)檢索檢索了Pubmed數(shù)據(jù)庫(kù)中1988年1月1日~2012年2月10日中發(fā)表的探討有關(guān)被動(dòng)吸煙與CC相關(guān)性的文章。采用以下檢索術(shù)語(yǔ):①“cervicalcancer"or“cervicalcarcinoma”or“uterInecervixcancer"or“CC"or“cervicalneoplasia”;②“secondhandsmoking”or“environmentaltobaccosmoke”or“ETs"or“passivesmoking”or“tobaccosmokepollution";③“casecontrol"or“incidence”or“prognosis"or“earlydiagnosis”or“survivalanalysis"or“case-control”，采用“AND”連接，未行任何檢索限制。此外,還檢索了納入研究及新近綜述的參考文獻(xiàn)/bbs/newweb/pc/post/4293216范文解析：風(fēng)險(xiǎn)因素類StudyselectionWeincludedanystudythatmetallofthefollowingcriteria:1)thestudydesignwasacase-controlstudy;2)investigatedtheassociationbetweenpassivesmokingandriskofcervicalcancer;(3)inclusionofatleast20cases;(4)thediagnosesofcervicalcancerwasconfirmedeitherhistological,pathologicallyorcytological;(5)theoddsratios(OR)andthecorresponding95%confidenceintervals(CIs),orthenumberofeventsthatcancalculatethemwerereported.Twoinvestigatorsindependentlyevaluatedtheeligibilityofallstudiesretrievedfromthedatabaseonthebasisofthepredeterminedselectioncriteria.Studiesnotdesignedascase-control,systematicreviewsandstudieswithmutuallyoverlappingpopulationswereexcludedfromthismeta-analysis.Disagreementswereresolvedbydiscussionorinconsultationwiththethirdone.ZengXT,AsianPacJCancerPrev.2012;13(6):2687-93.

納入文獻(xiàn)標(biāo)準(zhǔn)：納入了符合以下標(biāo)準(zhǔn)的研究:①研究設(shè)計(jì)為病例-對(duì)照研究;②探討被動(dòng)吸煙與CC風(fēng)險(xiǎn)間的相關(guān)性;③包括至少20例患者;④CC經(jīng)組織學(xué)病理學(xué)或細(xì)胞學(xué)確診;⑤直接報(bào)告了OR(oddsratios)及其95%CI(confidenceintervals),或者可以從文章報(bào)告的數(shù)據(jù)進(jìn)行計(jì)算(病例對(duì)照研究主要就是提取這2個(gè)數(shù)據(jù))。由兩名評(píng)價(jià)者獨(dú)立進(jìn)行篩選,不一致通過(guò)討論或咨詢第三位評(píng)價(jià)者解決。/bbs/newweb/pc/post/42932160范文解析：風(fēng)險(xiǎn)因素類DataextractionTworeviewersindependentlyextractedthefollowingdataforeacheligiblestudy:firstauthor’slastname,yearofpublication,siteoforigin,histologicaltypeandstageofthetumor,sourceofcontrols,numberofcasesandcontrols,adjustedestimatesofrisk.Anydisagreementswereresolvedbyconsensus.

數(shù)據(jù)提?。河蓛擅u(píng)價(jià)者獨(dú)立提取符合要求的資料：第一作者姓名,發(fā)表年份,研究地點(diǎn),腫瘤的組織學(xué)類型和臨床分期,對(duì)照組來(lái)源,病例組和對(duì)照組的樣本量,校正因素。通過(guò)討論解決分歧ZengXT,AsianPacJCancerPrev.2012;13(6):2687-93. /bbs/newweb/pc/post/42932160范文解析：風(fēng)險(xiǎn)因素類MethodologicalqualityassessmentTworeviewersindependentlyassessedthemethodologicalqualityoftheincludedstudieswiththeNewcastle–OttawaScale(NOS)(Wellsetal.,2009)forcase-controlstudies,whichconsistsofthreeparametersofquality:selection,comparability,andexposureassessment.TheNOSassignsamaximumscoreof4forselection,2forcomparability,and3forexposure.Hence,ascoreof9isthehighestandreflectsthehighestquality.Discrepancieswereaddressedinconsultationwiththethirdone.ZengXT,AsianPacJCancerPrev.2012;13(6):2687-93.

文獻(xiàn)質(zhì)量評(píng)價(jià)：由兩名評(píng)價(jià)者按照NOS量表獨(dú)立評(píng)估納入研究的方法學(xué)質(zhì)量。文章介紹了NOS量表的內(nèi)容TheNOSevalutiontoolincluded:(1)selectionIsthecasedefinitionadequate?RepresentativenessofthecasesSelectionofControlsDefinitionofControls(2)ComparabilityComparabilityofcasesandcontrolsonthebasisofthedesignoranalysis(3)ExposureAscertainmentofexposureSamemethodofascertainmentforcasesandcontrolsNon-Responserate/bbs/newweb/pc/post/42932160范文解析：風(fēng)險(xiǎn)因素類StatisticalanalysisWecomputedapooledORand95%CIbyusingtheComprehensiveMeta-Analysissoftware,version2.2(Biostat,Englewood,NewJersey)(Borensteinetal.,2005)togenerateforestplots,todeterminewhethertherewasastatisticalassociationbetweencasesandcontrolsandtoassessheterogeneityoftheincludedstudies.HeterogeneitywasquantifiedevaluatedusingthechisquarebasedCochran’sQstatistic(Higginsetal.,2002)andtheI2statistic,thisstatisticyieldsresultsrangedfrom0to100%(I2=0-25%,noheterogeneity;I2=25-50%,moderateheterogeneity;I2=50-75%,largeheterogeneity;andI2=75-100%,extremeheterogeneity)(Higginsetal.,2003).Ifheterogeneityexisted,therandomeffectsmodelwasused,otherwise,thefixedeffectsmodelwasused.

統(tǒng)計(jì)分析采用CMAv2(ComprehensiveMeta-Analysis)軟件計(jì)算合并OR值及其95%CI，首先采用卡方檢驗(yàn)及I2檢驗(yàn)評(píng)估異質(zhì)性,按照值將異質(zhì)性大小劃分如下:0~25%,無(wú)異質(zhì)性;25%~40%,中度異質(zhì)性;40%~75%;異質(zhì)性很大;75%~100%;異質(zhì)性極大。如異質(zhì)性存在,則采用隨機(jī)效應(yīng)模型;反之;則采用固定效應(yīng)模型ZengXT,AsianPacJCancerPrev.2012;13(6):2687-93. /bbs/newweb/pc/post/42932160范文解析：風(fēng)險(xiǎn)因素類Statisticalanalysis統(tǒng)計(jì)分析Inaddition,weinvestigatedtheinfluenceofasingle此外;作者采用依次逐個(gè)剔除單個(gè)研究的studyontheoverallriskestimatebyremovingeachstudyineachturn,totesttherobustnessofthemain方法探討單個(gè)研究對(duì)整體結(jié)果的影響程results.Ifsignificantheterogeneityisidentified,度,同時(shí),按照組織學(xué)類型和臨床階段、subgroupanalysiswasalsoconductedaccordingtohistologicaltypeandstageofthetumor,sourceof對(duì)照組來(lái)源、研究地點(diǎn)進(jìn)行亞組分析control(population-basedandhospital-basedcase-controlstudies),andcontinentinwhichthestudywasconducted(NorthAmericaandAsia).Ifpossible,若有可能,采用漏斗圖觀察是否存在發(fā)表potentialpublicationbiaswasassessedbyvisual偏倚,同時(shí)采用Begg和Eger's檢驗(yàn)對(duì)發(fā)inspectionofthefunnelplotsoftheprimaryoutcome(Eggeretal.,1997).TheBeggrankcorrelationtestwas表偏倚進(jìn)行定量檢測(cè)usedtoexaminetheasymmetryofthefunnelplot(Beggetal.,1994)andtheEggerweightedlinearregressiontestwasusedtoexaminetheassociationbetweenmeaneffectestimateanditsvariance(Eggeretal.,1997)./bbs/newweb/pc/post/42932160ZengXT,AsianPacJCancerPrev.2012;13(6):2687-93.范文解析：預(yù)后meta分析范文預(yù)后meta分析范文：本Meta分析遵照MOOSE規(guī)范進(jìn)行報(bào)告SurvivalandClinicopathologicalSignificanceofSIRT1ExpressioninCancers:AMeta-AnalysisSIRT1的表達(dá)在腫瘤中生存和臨床病理意義:一項(xiàng)meta分析摘要背景：沉默信息調(diào)節(jié)因子2同系物1(SIRT1)是一種進(jìn)化保守的酶，具有依賴于煙酰胺腺嘌呤二核苷酸(NAD+)的脫乙酰酶活性。SIRT1參與多種細(xì)胞過(guò)程，如基因組穩(wěn)定性、能量代謝、衰老、基因轉(zhuǎn)錄和氧化應(yīng)激等。長(zhǎng)期以來(lái)，SIRT1被認(rèn)為既是腫瘤的促進(jìn)劑又是腫瘤的抑制者。它在癌癥預(yù)后中的作用仍然存在爭(zhēng)議。方法：對(duì)PubMed、EMBASE和CochraneLibrary的63篇文章中的13，138名受試者進(jìn)行薈萃分析，以評(píng)估SIRT1在不同腫瘤中表達(dá)的生存率和臨床病理意義。結(jié)果：Meta分析結(jié)果顯示，SIRT1的高表達(dá)提示癌癥患者的總體生存率(OS)[危險(xiǎn)比(HR)=1.566，95%CI：1.293~1.895，P<0.0001]、無(wú)病生存率(HR=1.631，95%CI：1.250~2.130，P=0.0003)、無(wú)事件生存率(HR=2.534，95%CI：1.602~4.009，P=0.0001)。無(wú)進(jìn)展生存率(HR=3.325，95%CI：2.762~4.003，P<0.0001)。SIRT1水平升高與腫瘤分期[相對(duì)危險(xiǎn)度(RR)=1.299，95%CI：1.114~1.514，P=0.0008]、淋巴結(jié)轉(zhuǎn)移(RR=1.172，95%CI：1.010~1.360，P=0.0363)、遠(yuǎn)處轉(zhuǎn)移(RR=1.562，95%CI：1.022~2.387，P=0.0392)相關(guān)。Meta回歸和亞組分析顯示，種族背景對(duì)SIRT1表達(dá)在預(yù)測(cè)腫瘤生存和臨床病理特征中的作用有影響。在亞洲人群中，尤其是在中國(guó)人群中，SIRT1的過(guò)表達(dá)預(yù)示著OS更差，TNM分期更高，淋巴結(jié)轉(zhuǎn)移更多。結(jié)論：SIRT1高表達(dá)預(yù)示OS、DFS、EFS、PFS較差，但對(duì)無(wú)復(fù)發(fā)生存期(RFS)和腫瘤特異性生存期(CCS)無(wú)明顯影響。SIRT1過(guò)表達(dá)與較高的腫瘤分期、淋巴結(jié)轉(zhuǎn)移和遠(yuǎn)處轉(zhuǎn)移相關(guān)。SIRT1介導(dǎo)的分子事件和生物學(xué)過(guò)程可能是腫瘤轉(zhuǎn)移的潛在機(jī)制，SIRT1是抑制腫瘤轉(zhuǎn)移的治療靶點(diǎn)SunM，F(xiàn)ront.Endocrinol.,13March2019|/10.3389/fendo.2019.00121范文解析：預(yù)后meta分析LiteraturessearchWeretrievedliteraturepublishedinbetween1966andApril1st,2018bysearchingPubMed,EMBASE,andCochraneLibrarywiththekeywords(1)“SIRT1”O(jiān)R“sirtuin1”O(jiān)R“SIR2”O(jiān)R“SIR2L1”O(jiān)R“SIR2alpha”O(jiān)R“silentmatingtypeinformationregulation2homolog-1”AND(2)“tumorORcancerORcarcinomaORneoplasm”andusingthesearchstrategiesasillustratedinSupplementaryTables1A–C.WeselectedandevaluatedallrelevantstudiesandreviewarticlesaboutSIRT1andinquiredtheauthorsforunpublishedrawdata.Thesearchandselectionofarticlesforthestudywereseparatelyconductedbasedonacommonsetofcriteria.Thedivergenceinopinionweresettledthroughdiscussionamongourselves.

文獻(xiàn)檢索應(yīng)用關(guān)鍵詞(1)“SIRT1”、“sirtuin1”、“Sir2”、“SIR2L1”、“SIR2alpha”、“SIR2alpha”和(2)“TumororCancerorCancerorneoplasm”，檢索PubMed、EMBASE和CochraneLibrary在1966-2018年4月1日期間發(fā)表的文獻(xiàn)，并使用補(bǔ)充表1A-C所示的搜索策略進(jìn)行選擇和評(píng)價(jià)。這項(xiàng)研究的搜索和選擇文章是根據(jù)一套共同的標(biāo)準(zhǔn)單獨(dú)進(jìn)行的。意見(jiàn)分歧是通過(guò)我們之間的討論解決的范文解析：預(yù)后meta分析InclusionandExclusionCriteriaThismeta-analysiswasconductedaccordingtoMeta-analysisofObservationalStudiesinEpidemiology(MOOSE)Checklist.Studiesenrolledinthisanalysissatisfiedthefollowingrequirements:(i)patientsmustbediagnosedwithcancerviapathology;(ii)TheexpressionofSIRT1mustbedeterminedbyquantitativereal-timepolymerasechainreaction(q-PCR),immunohistochemistry(IHC),orinsituhybridization(ISH);(iii)ThecorrelationbetweenSIRT1expressionandprognosisorclinicopathologicalfeatureswasinvestigated;(iv)TheHazardRatio(HR)andits95%confidenceinterval(CI)forsurvivalindicatoronthebasisofSIRT1expressionlevelwerereadilyavailableorcouldbecalculatedindirectly;(v)Themostrepresentativeandmostaccuratestudywasadoptedwhenasinglesamplesourcewasusedinmultiplestudiestoavoidunnecessarycohortoverlapping

納入文獻(xiàn)標(biāo)準(zhǔn)：這項(xiàng)薈萃分析是根據(jù)流行病學(xué)(MOOSE)檢查表中觀察性研究的meta分析進(jìn)行的。本分析研究滿足以下要求：(i)患者必須經(jīng)病理診斷為癌癥；(ii)必須通過(guò)實(shí)時(shí)定量聚合酶鏈反應(yīng)(Q-PCR)、免疫組織化學(xué)(IHC)或原位雜交(ISH)檢測(cè)SIRT1的表達(dá)；(iii)研究SIRT1的表達(dá)與預(yù)后或臨床病理特征的相關(guān)性；(iv)基于SIRT1的表達(dá)水平作為生存指標(biāo)的危險(xiǎn)比(HR)及其95%可信區(qū)間(CI)是現(xiàn)成的。(v)在多項(xiàng)研究中使用同一樣本來(lái)源時(shí)，采用最具代表性和最準(zhǔn)確的研究，以避免不必要的隊(duì)列重疊范文解析：預(yù)后meta分析InclusionandExclusionCriteria納入文獻(xiàn)標(biāo)準(zhǔn)：Studiesthathavesatisfiedtheabovementionedinclusionrequirementswerefurtherruledoutifthey

滿足上述納入要求的研究如果有以下任何缺陷，將進(jìn)一步排除：hadanyofthefollowingflaws:(i)duplicatedarticlesordata;(ii)nothumanstudies;(iii)reviewarticlesor

(i)重復(fù)的文章或數(shù)據(jù)(ii)不是人體研究letters;(iv)lackofsufficientdataorinformationto

(iii)reviewarticlesorlettersgetHR;(v)articlesnotwritteninEnglish.

(iv)缺乏足夠的數(shù)據(jù)或信息來(lái)獲得HR(v)文章不是用英文寫的范文解析：預(yù)后meta分析DataextractionWeextractedthefollowingdatafromthefulltextsofeligiblestudies:(i)thefirstauthor;(ii)publicationyear;(iii)characteristicsofthestudies,whichcomprisedofthepatients'nationality,samplesize,tumortype,andclinicopathologicalcharacteristics;(iv)theassaymethodandcut-offvalueofSIRT1;(v)HRsofSIRT1expressionforOS,disease-freesurvival(DFS),event-freesurvival(EFS),recurrence-freesurvival(RFS),cancer-specificsurvival(CCS),progression-freesurvival(PFS);(vi)iftheHRforOS,DFS,EFS,RFS,CCSandPFSwerecalculatedbybothunivariateandmultivariateanalyses,thelatterwasourfirstchoice,giventhattheseresultswereadjustedforconfoundingfactors.

數(shù)據(jù)提?。何覀儚姆蠗l件的研究全文中提取了以下數(shù)據(jù)：(i)第一作者；(ii)發(fā)表年份；(iii)研究的特征，包括患者的國(guó)籍、樣本量、腫瘤類型和臨床病理特征；(iv)SIRT1的檢測(cè)方法和臨界值；(v)SIRT1表達(dá)對(duì)OS、無(wú)病生存率(DFS)、無(wú)事件生存率(EFS)、無(wú)復(fù)發(fā)生存率(RFS)、癌癥特異性生存率(CCS)的影響。(vi)如果OS、DFS、EFS、RFS、CCS和PFS的HR為通過(guò)單變量和多變量分析計(jì)算，考慮到這些結(jié)果調(diào)整了混雜因素，后者是我們的首選Dataextraction數(shù)據(jù)提取：IfastudydidnotreporttheHR,weestimatedHRand如果一項(xiàng)研究沒(méi)有報(bào)告HR，我們使用theircorresponding95%CIusingthemethoddescribedParmar等人描述的方法估計(jì)HR及其相應(yīng)byParmaretal.(37)andTierneyetal.(38).We的95%CI和Tierney等人用EngaugerecoveredthedataofKaplan-Meiercurvesviathe9.8(http://markummitchell.github.io/EnEngaugeDigitizerversion9.8gauge-DIGITALIZER)恢復(fù)Kaplan-Meier(http://markummitchell.github.io/engauge-digitizer)and曲線數(shù)據(jù)，計(jì)算HR及其95%CI。我們將這calculatedtheHRandits95%CI.Werepeatedthis個(gè)過(guò)程重復(fù)了三次，以減少可變性。在提processthreetimestoreducevariabili