TMIS实证医学个案讨论会简报范例课件

李文欽醫師實證醫學中心主任2013/11/08實證醫學與個案討論李文欽醫師實證醫學與個案討論簡報大綱l簡述實證醫學五大步驟l如何問問題及形成PICO的內容l資料搜尋的策略、使用資料庫及 搜尋結果l對資料的評讀l如何落實實證醫學於個案討論簡報大綱簡述實證醫學五大步驟過去現在未來當一個醫生該具備的的能力三日不讀書，面目可憎醫病照護教學研究過去現在未來當一個醫生該具備的的能力三日不讀書，TMIS实证医学个案讨论会简报范例课件2000年Sackett將實證醫學定義為“整合最佳研究證據、臨床經驗以及病患價值以達成病患最佳治療的過程”實證醫學的實行是一種終身的、自發的、以問題為導向的學習過程實證醫學的重要性臨床經驗病患價值最佳研究證據實證醫療Sackett et al 20002000年Sackett將實證醫學定義為“整合最佳研究證據實證醫學的五個步驟1.Asking an answerable question (提出可回答的臨床問題)2.Tracking down the best evidence (搜尋最佳實證文獻資料)3.Critical appraisal (嚴謹的文獻評讀)4.Integrating the appraisal with clinical expertise and patients preference (將臨床專業與病人價值觀相結合)5.Evaluation the effectiveness and efficiency in executing steps 1-4 (評估改善)實證醫學的五個步驟1.AskingananswerabAstrokepatientwhohaslongtermgoutyarthritisDoes gout cause a higher risk of stoke?痛風會增加中風的風險嗎？Astrokepatientwhohaslong找資料來回答問題Acquire嚴格評讀文獻Appraisal是否可應用到病人身上Apply將病人的問題寫成PICOAskingEBM的步驟Sir Austin Bradford Hill(1897-1991),EnglishepidemiologistandstatisticianHedescribedtheclinicaltrialasbeingaCAREFULLY,andETHICALLY,describedexperimentwiththeaimofansweringsomePRECISELY FRAMED QUESTIONGreaterprecisionisrequiredintheobjectives.WeneedtohaveAdefinedpopulationDefinedendpointsRelativelyfewquestiontobeanswered.問題要怎麼問？找資料來回答問題嚴格評讀文獻是否可應用到病人身上將病人的問題1.形成問題(ClinicalQuestion)Does gout cause a higher risk of stoke?痛風會增加中風的風險嗎？1.形成問題(ClinicalQuestion)Doe臨床的問題模型(PICO)P:Patient,Population指此病人或族群的年齡層、體質、疾病史、過去史等I:Intervention指治療、診斷、預後、成本效益分析等，為研究的主題C:Comparison指做與不做或不同處置間的比較O:Outcome指預後或與目前問題有關的比較基準臨床的問題模型(PICO)P:Patient,Popul找資料來回答問題Acquire嚴格評讀文獻Appraisal是否可應用到病人身上Apply將病人的問題寫成PICOAskingEBM的步驟PICO 1 Does gout cause a higher risk of stoke?PICO type:TherapyPatient:老年人Intervention:痛風Comparison:無痛風Outcome:中風的風險問題要怎麼問？找資料來回答問題嚴格評讀文獻是否可應用到病人身上將病人的問題2.Tracking down the best evidence直接使用實證醫學資庫(secondary databases)如ACP journal club,Cochrane Library,UpToDate,MD Consult,Practice Guideline找初步篩選的資庫(primary databases)PubMed,OVID,Proquest未經篩選的資庫如Google可搜尋實證醫學的資料庫2.Trackingdownthebestevid 如何尋找與EBM相關的館藏資源？主題切入本院一般綜合性資料庫 1.PudMed：可查出abstract 2.Ovid：可查出abstract、Fulltext 3.Proquest：可查出abstract、Fulltext 4.Clinicalkey：可查出abstract、Fulltext找實證醫學資料時可加上 Systemic Review(系統性評論)、Meta-analysis(整合分析)、Randomized Controlled Trial(隨機對照臨床試驗研究)、Controlled Trials利用資料庫功能鍵如何尋找與EBM相關的館藏資源？主題切入TMIS实证医学个案讨论会简报范例课件TMIS实证医学个案讨论会简报范例课件TMIS实证医学个案讨论会简报范例课件TMIS实证医学个案讨论会简报范例课件MeSHtoidentifyeverytermforexampleMeSHtoidentifyeverytermfo找資料來回答問題Acquire嚴格評讀文獻Appraisal是否可應用到病人身上Apply將病人的問題寫成PICOAskingSearchingStrategyFindingoutTheCorrectKeywords原始關鍵字Primary Term 或MeSH Term 同義字1同義字2P(oror)andI(Goutor Hyperuricemia or)andC(oror)andO(strokeoror)找資料來回答問題嚴格評讀文獻是否可應用到病人身上將病人的問題找資料來回答問題Acquire嚴格評讀文獻Appraisal是否可應用到病人身上Apply將病人的問題寫成PICOAskingSearchingStrategyFindingoutTheCorrectKeywordsKeywordsfromPICOitem:(“gout”OR“hyperuricemia”)AND“stroke”MeSHtoidentifyeverytermforexample找資料來回答問題嚴格評讀文獻是否可應用到病人身上將病人的問題TMIS实证医学个案讨论会简报范例课件找資料來回答問題Acquire嚴格評讀文獻Appraisal是否可應用到病人身上Apply將病人的問題寫成PICOAsking善用搜尋Systematicreview關鍵字:(“gout”OR“hyperuricemia”)AND(“stroke”OR“cardiovasculardisease”)找資料來回答問題嚴格評讀文獻是否可應用到病人身上將病人的問題TMIS实证医学个案讨论会简报范例课件找資料來回答問題Acquire嚴格評讀文獻Appraisal是否可應用到病人身上Apply將病人的問題寫成PICOAsking找資料來回答問題嚴格評讀文獻是否可應用到病人身上將病人的問題TheEvidencePyramidV.Animal researchV.In vitro(test tube)researchIV.Case series/ReportsV.Ideas,Editorials,OpinionsIII.Case Control StudiesII.Cohort studiesI.Randomized Controlled Studies(RCT)I.Randomized Controlled Double Blind Studies Meta-analysis統計方法統計方法統計方法統計方法Meta-analysisMeta-analysis 圖示結果Forest plot*Hierarchy of evidence:arranges study designs by their susceptibility to bias.(Level IV)From:Oxford Center for EBM研究設計與證據強度TheEvidencePyramidV.AnimalLevel Therapy/Prevention,Aetiology/HarmPrognosisDiagnosisDifferential diagnosis/symptom prevalence studyEconomic and decision analyses1aSR(withhomogeneity*)ofRCTsSR(withhomogeneity*)ofinceptioncohortstudies;CDRvalidatedindifferentpopulationsSR(withhomogeneity*)ofLevel1diagnosticstudies;CDRwith1bstudiesfromdifferentclinicalcentresSR(withhomogeneity*)ofprospectivecohortstudiesSR(withhomogeneity*)ofLevel1economicstudies1bIndividualRCT(withnarrowConfidenceInterval)Individualinceptioncohortstudywith80%follow-up;CDRvalidatedinasinglepopulationValidating*cohortstudywithgoodreferencestandards;orCDRtestedwithinoneclinicalcentreProspectivecohortstudywithgoodfollow-up*Analysisbasedonclinicallysensiblecostsoralternatives;systematicreview(s)oftheevidence;andincludingmulti-waysensitivityanalyses1cAllornoneAllornonecase-seriesAbsoluteSpPinsandSnNoutsAllornonecase-seriesAbsolutebetter-valueorworse-valueanalyses2aSR(withhomogeneity*)ofcohortstudiesSR(withhomogeneity*)ofeitherretrospectivecohortstudiesoruntreatedcontrolgroupsinRCTsSR(withhomogeneity*)ofLevel2diagnosticstudiesSR(withhomogeneity*)of2bandbetterstudiesSR(withhomogeneity*)ofLevel2economicstudies2bIndividualcohortstudy(includinglowqualityRCT;e.g.,80%follow-up;CDRvalidatedinasinglepopulationValidating*cohortstudywithgoodreferencestandards;orCDRtestedwithinoneclinicalcentreProspectivecohortstudywithgoodfollow-up*Analysisbasedonclinicallysensiblecostsoralternatives;systematicreview(s)oftheevidence;andincludingmulti-waysensitivityanalyses1cAllornoneAllornonecase-seriesAbsoluteSpPinsandSnNoutsAllornonecase-seriesAbsolutebetter-valueorworse-valueanalyses2aSR(withhomogeneity*)ofcohortstudiesSR(withhomogeneity*)ofeitherretrospectivecohortstudiesoruntreatedcontrolgroupsinRCTsSR(withhomogeneity*)ofLevel2diagnosticstudiesSR(withhomogeneity*)of2bandbetterstudiesSR(withhomogeneity*)ofLevel2economicstudies2bIndividualcohortstudy(includinglowqualityRCT;e.g.,80%follow-up)Retrospectivecohortstudyorfollow-upofuntreatedcontrolpatientsinanRCT;DerivationofCDRorvalidatedonsplit-sampleonlyExploratory*cohortstudywithgoodreferencestandards;CDRafterderivation,orvalidatedonlyonsplit-sampleordatabasesRetrospectivecohortstudy,orpoorfollow-upAnalysisbasedonclinicallysensiblecostsoralternatives;limitedreview(s)oftheevidence,orsinglestudies;andincludingmulti-waysensitivityanalyses2cOutcomesResearch;EcologicalstudiesOutcomesResearchEcologicalstudiesAuditoroutcomesresearch3aSR(withhomogeneity*)ofcase-controlstudiesSR(withhomogeneity*)of3bandbetterstudiesSR(withhomogeneity*)of3bandbetterstudiesSR(withhomogeneity*)of3bandbetterstudies3bIndividualCase-ControlStudyNon-consecutivestudy;orwithoutconsistentlyappliedreferencestandardsNon-consecutivecohortstudy,orverylimitedpopulationAnalysisbasedonlimitedalternativesorcosts,poorqualityestimatesofdata,butincludingsensitivityanalysesincorporatingclinicallysensiblevariations.4Case-series(andpoorqualitycohortandcase-controlstudies)Case-series(andpoorqualityprognosticcohortstudies*)Case-controlstudy,poorornon-independentreferencestandardCase-seriesorsupersededreferencestandardsAnalysiswithnosensitivityanalysis5Expertopinionwithoutexplicitcriticalappraisal,orbasedonphysiology,benchresearchorfirstprinciplesExpertopinionwithoutexplicitcriticalappraisal,orbasedonphysiology,benchresearchorfirstprinciplesExpertopinionwithoutexplicitcriticalappraisal,orbasedonphysiology,benchresearchorfirstprinciplesExpertopinionwithoutexplicitcriticalappraisal,orbasedonphysiology,benchresearchorfirstprinciplesExpertopinionwithoutexplicitcriticalappraisal,orbasedoneconomictheoryorfirstprinciplesOxford Centre for Evidence-based Medicine-Levels of Evidence(March 2009)http:/ and risk of stroke:a systematic review and meta-analysis.(2009)Source:ArthritisRheum.2009Jul15;61(7):885-92Level:2a找資料來回答問題嚴格評讀文獻是否可應用到病人身上將病人的問題Largeepidemiologicstudieshaveshownthathyperuricemiaisassociatedwithanincreasedincidenceofcoronary heart disease(CHD)andincreased mortality inthosewithandwithoutpreexistingCHD.FangJ,AldermanMH.SerumuricacidandcardiovascularmortalitytheNHANESIepidemiologicfollow-upstudy,1971-1992.NationalHealthandNutritionExaminationSurvey.JAMA2000;283:2404.FreedmanDS,WilliamsonDF,GunterEW,ByersT.Relationofserumuricacidtomortalityandischemicheartdisease.TheNHANESIEpidemiologicFollow-upStudy.AmJEpidemiol1995;141:637.BrandFN,McGeeDL,KannelWB,etal.Hyperuricemiaasariskfactorofcoronaryheartdisease:TheFraminghamStudy.AmJEpidemiol1985;121:11.NiskanenLK,LaaksonenDE,NyyssnenK,etal.Uricacidlevelasariskfactorforcardiovascularandall-causemortalityinmiddle-agedmen:aprospectivecohortstudy.ArchInternMed2004;164:1546.ChoiHK,CurhanG.Independentimpactofgoutonmortalityandriskforcoronaryheartdisease.Circulation2007;116:894.LargeepidemiologicstudieshaProposed mechanisms forsuchanincreaseinriskincludethedevelopment of hypertension andoxidative stress.JohnsonRJ,SegalMS,SrinivasT,etal.Essentialhypertension,progressiverenaldisease,anduricacid:apathogeneticlink?JAmSocNephrol2005;16:1909.BakerJF,KrishnanE,ChenL,SchumacherHR.Serumuricacidandcardiovasculardisease:recentdevelopments,andwheredotheyleaveus?AmJMed2005;118:816.ProposedmechanismsforsuchaItisunclearifhyperuricemiahasacausal effect or,ashasbeenmoreoftensuggested,issimply a marker forotherriskfactors,suchashypertension,dyslipidemia,anddiabetes.CulletonBF,LarsonMG,KannelWB,LevyD.Serumuricacidandriskforcardiovasculardiseaseanddeath:theFraminghamHeartStudy.AnnInternMed1999;131:7.WannametheeSG,ShaperAG,WhincupPH.Serumurateandtheriskofmajorcoronaryheartdiseaseevents.Heart1997;78:147.WheelerJG,JuzwishinKD,EiriksdottirG,etal.Serumuricacidandcoronaryheartdiseasein9,458incidentcasesand155,084controls:prospectivestudyandmeta-analysis.PLoSMed2005;2:e76.FeigDI,KangDH,JohnsonRJ.Uricacidandcardiovascularrisk.NEnglJMed2008;359:1811.Itisunclearifhyperuricemia文獻評讀三部曲文獻評讀最主要的三個步驟，即是VIPV：Validity/Reliability;效度/信度我們能相信這篇文章嗎?I：Importance/Impact;重要性我們相信它，但這個結論重要嗎?P：Practice/Applicability;臨床適用性如果我們相信它，這個結論可以應用在我們所照護的病患嗎?文獻評讀三部曲文獻評讀最主要的三個步驟，即是VIP找資料來回答問題Acquire嚴格評讀文獻Appraisal是否可應用到病人身上Apply將病人的問題寫成PICOAskingAccordingto“SharonE.Strausetal,Evidence-basedmedicine:howtopracticeandteachEBM,Elsevier,2005:33-7.“ValidityImportanceApplicabilityResultsfromsecondarydatabaseResultsfromprimarydatabaseAppraisedbyexpertsNeededtobeappraisedResults(NNT)ResultsConsiderpopulationandfeasibilityConsiderpopulationandfeasibility找資料來回答問題嚴格評讀文獻是否可應用到病人身上將病人的問題效度(Validity)：可以用RAM-bo方式審視效度研究族群是否有隨機分配(Randomized)？隨機分配的方法是否適當(Accounted)？結果的衡量(Measurement)，即是否依照 盲法(blind)及客觀(objective)二要素操作治療意圖分析(Intention-to-treat analysis,ITT)隨機試驗的一種分析方法所有被分配在治療組/對照組的病人無論是否完成該項 治療/安慰劑都應該被放進治療組/對照組(原分派組別)中做分析文獻評讀三部曲效度(Validity)：可以用RAM-bo方式審視效度文獻重要性(Importance/Impact)：研究結果是什麼?研究結果如何被估計?經過多久的時間?文獻結果的評估方式：勝算比(odds ratio)代表生物學上的影響。相對危險(relative risk)、絕對危險降低度(absolute risk reduction)、相對危險降低度(relative risk reduction)、益一需治數(NNT,number needed to treat):為預防一個不良結果或使一個病人達實驗所求之有益結果所需治療的病人數NNT=1/ARR文獻評讀三部曲重要性(Importance/Impact)：文獻評讀三部曲可行性(Practical,Applicability?)：臨床實用性如何？1.應考量病患的生物因素(biological issues):-同樣的治療應用在不同的病患族群是否有不同反應?-我們的病人與研究中的病人是否非常不同，以致無法應用在研究結果？2.其他的社會經濟因素（social and economic issues）的考量:-這個治療適用於我們的診療環境嗎？病患的配合度如何？-醫療提供者的配合度及能力如何？3.流行病學因素(epidemiological issues）的考量:-我們的病人是否有其他共病狀況，可能改變治療的結果？影響有多大？-病人可能從治療中得到什麼好處或壞處？-經由治療而減少的不良後果是否比不治療有明顯的差別？文獻評讀三部曲可行性(Practical,Applicability幫助文獻評讀的工具幫助文獻評讀的工具依所搜尋文獻類別，評讀所問的問題依所搜尋文獻類別，評讀所問的問題4.將臨床專業與病人價值觀相結合您的病人是否與研究中的病人差別很大，以至於無法適用該研究結果？族群/地域/文化差異您期望您的病人從研究結果中獲得多大的好處？還有哪些替代方案？研究結果適用於您的病人嗎？4.將臨床專業與病人價值觀相結合您的病人是否與研究中的病人5.評估執行效果及改善過程您是否開始使用實證手法搜尋最佳證據？您搜尋及評讀證據的速度有多快？您有能力將這些證據應用在適當的病人身上嗎？您是否依循這些新證據來改變您的診療習慣？5.評估執行效果及改善過程您是否開始使用實證手法搜尋最佳證據 如何落實實證醫學於個案討論會?如何落實實證醫學於個案討論會?如何落實實證醫學於個案討論會1.由討論的個案找尋問題(形成PICO)2.進行文獻搜尋(PUB Med,Cochrane,Ovid)3.進行文獻評讀(Evidence level,Recommend grade)會議中報告文章重點與結論4.所搜尋的文獻值得應用到本病人身上嗎?如何落實實證醫學於個案討論會1.由討論的個案找尋問題(形成實證醫學個案討論會順序1.Case report(個案報告)2.提出問題，建立 PICO格式3.文獻搜尋(所使用關鍵字、資料庫)4.所搜尋到文獻的題目、出處、及文獻5.的證據等級 6.5.報告文獻的重點與結論7.6.是否可應用到我們的病人身上 實證醫學個案討論會順序Casereport(個案報告)Clinicalquestion:Does gout cause a higher risk of stoke?PICO type:TherapyPatient:Old patientIntervention:GoutComparison:Without goutOutcome:Risk of strokeClinicalquestion:PICOtypeTitle:Hyperuricemia and risk of stroke:a systematic review and meta-analysis.(2009)Source:ArthritisRheum.2009Jul15;61(7):885-92Level:2aTitle:TMIS实证医学个案讨论会简报范例课件TMIS实证医学个案讨论会简报范例课件科別選定與討論頻率1.科別:內、外、婦、兒、皮膚科 藥劑、護理、(檢驗、放射)2.頻率:每月一次3.會議主題設定:實證醫學個案討論會4.請科部主任定期檢視結果與統計科別選定與討論頻率1.科別:內、外、婦、兒、皮膚科TMIS实证医学个案讨论会简报范例课件Thank You For Your Attention!ThankYouForYourAttention!