Critical-Appraisal-Skills-Programme-(CASP)-秀傳醫(yī)療社團(tuán)法人秀傳-

秀傳紀(jì)念醫(yī)院實(shí)證醫(yī)學(xué)中心

住院醫(yī)師升等實(shí)證醫(yī)學(xué)報(bào)告中醫(yī)部第一年住院醫(yī)師李容妙ForegroundQuestions「耳鳴患者，接受針灸治療與西醫(yī)常規(guī)治療或無(wú)治療相比，是否能使耳鳴癥狀顯著改善？」現(xiàn)病史五十三歲男性患者，因三年前家中發(fā)生變故，開(kāi)

始出現(xiàn)眠差情形，整日昏沉疲倦，兩年前開(kāi)始出現(xiàn)兩耳高頻耳鳴(患者形容為金屬摩擦聲)，左耳較嚴(yán)重，整日皆有，無(wú)特殊加重或減緩因子。近四至五個(gè)月癥狀加重，覺(jué)得耳鳴聲響愈大，難以忍受，經(jīng)西醫(yī)耳鼻喉科檢查無(wú)異樣，並給與常規(guī)治療，但仍無(wú)明顯改善，因此前來(lái)本院針灸科求診。3過(guò)去病史Type2DM

--規(guī)律門(mén)診追蹤及服藥，控制良好無(wú)高血壓、心血管相關(guān)疾病或其他系統(tǒng)性疾患無(wú)開(kāi)刀病史4神志清楚；頭面無(wú)明顯異常，視野無(wú)明顯缺損，眼球活動(dòng)正常；呼吸平順，無(wú)喘鳴音；心跳規(guī)律，無(wú)明顯異常心音；腹部柔軟，無(wú)壓痛，無(wú)可觸及之硬塊，腹部扣診鼓音，腸音無(wú)躁進(jìn)；四肢活動(dòng)無(wú)限制；可觸及周邊動(dòng)脈搏動(dòng)無(wú)發(fā)燒，無(wú)淋巴結(jié)腫大，無(wú)水腫，無(wú)胸悶胸痛，無(wú)小便短少，無(wú)視力模糊，無(wú)感覺(jué)異常，無(wú)肌肉張力或肢體活動(dòng)異常。系統(tǒng)性回顧理學(xué)檢查T(mén)ypeofQuestion

TherapyPICOPatientorProblem耳鳴

Intervention使用針灸治療

Comparison西醫(yī)常規(guī)治療或無(wú)治療

Outcome耳鳴癥狀減輕(包括客觀量表評(píng)估與主觀患者感受)

KeywordTinnitusacupunctureMeSHTinnitusacupunctureSearchStrategy文獻(xiàn)搜尋過(guò)程,keyword、布林邏輯Pubmed鍵入(tinnitus)AND(acupuncture)",搜尋結(jié)果共有95篇在頁(yè)面右上方頁(yè)面點(diǎn)選”systematicreviews”,共篩出13篇在多篇文獻(xiàn)中，考量能提供目前環(huán)境病患的治療方針，選擇最近期之文獻(xiàn)(第一篇)，作為本次可能評(píng)讀之文章CochraneLibrary使用advancedsearch

鍵入"tinnitus"inTitle,AbstractorKeywords

AND"acupuncture"inTitle,AbstractorKeywords

共有二篇文章，因皆屬於Protocol

review與我們想要的主題不符，因而捨棄不選。綜上所述，此次評(píng)讀文章選擇：Acupunctureinthetreatmentoftinnitus:asystematicreviewandmeta-analysis

LiuF1,HanX,LiY,YuS.,

EurArchOtorhinolaryngol.2014Oct25Citations標(biāo)題：Acupunctureinthetreatmentoftinnitus:asystematicreviewandmeta-analysis作者：

FenyeLiu,XiuliHan,YunfengLi,ShudongYu期刊：EurArchOtorhinolaryngol,2014Oct25,DOI10.1007/s00405-014-3341-7EvidenceLevelLevel1aCriticalAppraisalSkillsProgramme

(CASP)：

SystematicReviewScreeningQuestions1.Didthereviewaskaclearly-focusedquestion?

YesConsiderifthequestionis‘focused’intermsof:thepopulationstudiedasystematicreviewandmeta-analysisofallavailablerandomizedcontrolledtrials(RCTs)usingacupuncture

totreattinnitus.theinterventiongivenorexposure

parallelRCTsoftinnituspatientswhichcomparedsubjectsreceivingacupuncture(oritsotherforms,suchaselectroacupuncture)tosubjectsreceiving

notreatment,shamtreatment,drugsorbasicmedicaltherapy.theoutcomesconsideredVisualAnalogueScale(VAS)subjectivesimplisticevaluations(better,normal,orworse)由上述各細(xì)項(xiàng)檢視，研究母群體、介入方式與療效指標(biāo)均有明確提及，因此從本文描述中有看到一個(gè)明確焦點(diǎn)性問(wèn)題的提出。2.Didthereviewincludetherighttypeofstudy?

YesConsideriftheincludedstudies:addressthereview’squestionOur

reviewremediedthatdeficiencybyincludingallrelevant

randomizedcontrolledtrials(RCTs)publishedinChinese

andconductingasystematicreviewandmeta-analysis,

criticallyevaluatingallofthecurrentlyavailableRCTsthat

usedacupuncturetotreattinnitus.haveanappropriatestudydesignAllthestudiesthatwereviewedincludedparallelRCTsof

acupuncture(oroneofitsotherforms,suchaselectroacupuncture)

comparedwithseveralalternatives,including

notreatment,shamtreatment,placebo,drugsorbasic

medicaltherapyinpatientswithtinnitus.RCTsthatcombined

acupuncturetreatmentwithotherinterventionsand

thencomparedthesewiththeotherinterventionsalone

werealsoincluded.Excludedfromthisstudywereanimalstudies,case–controlstudies,reviews,commentaries,letters,andstudies

thatexaminedotherassociations.Multiplepublications

reportingonthesameRCTsandthesamegroupsofparticipants

wereexcluded,aswereRCTsthatcompared

differentformsofacupunctureordifferentacupuncture

pointsandtrialsforthetreatmentofpointinjection.流程圖(見(jiàn)後頁(yè))由上述各細(xì)項(xiàng)檢視，包含作者提出本研究收集所有(包含中、英文)的針灸治療耳鳴隨機(jī)對(duì)照試驗(yàn)以及明確的納入、排除條件，並繪出清楚的研究設(shè)計(jì)流程圖，，因此本回顧文獻(xiàn)擁有正確的研究型態(tài)。Isitworthcontinuing?由上述兩題的檢視，合乎標(biāo)準(zhǔn)，此篇文獻(xiàn)值得繼續(xù)探討下去DetailedQuestions3.Didthereviewerstrytoidentifyallrelevantstudies?

YesConsider:whichbibliographicdatabaseswereusedThedatabaseswesearchedweretheChinaNetwork

KnowledgeInfrastructure(1979–2013),theChineseScientific

JournalDatabaseVIP(1989–2013),PubMed

(1966–2013),EMBASE(1974–2013)andtheCochrane

Library(Issue3,2013).iftherewasfollow-upfromreferencelists文中未提及iftherewaspersonalcontactwithexperts文中未提及ifthereviewerssearchedforunpublishedstudies文中未提及ifthereviewerssearchedfornon-English-languagestudiesOnlystudiesonhumansandineither

EnglishorChinesewereconsideredforinclusion.由上述各細(xì)項(xiàng)檢視，作者有提及資料庫(kù)的搜尋範(fàn)圍以及語(yǔ)言限定(中英文皆收集)，但未針對(duì)相關(guān)文章更進(jìn)一步的細(xì)節(jié)(含參考文獻(xiàn)的追蹤、專(zhuān)家的接觸面談或未出版文章的蒐集)，總結(jié)本文已有嘗試說(shuō)明蒐集相關(guān)重要文獻(xiàn)的細(xì)部?jī)?nèi)容，但非能有完整呈現(xiàn)。4.Didthereviewersassessthequalityoftheincludedstudies?

YesConsider

ifaclear,pre-determinedstrategywasusedtodeterminewhichstudieswereincluded.Lookfor:ascoringsystem

Datafromthearticleswerevalidatedandextracted

usingapredefineddataextractionform.

Fortheassessmentofstudyquality,thetoolfromthe

CochraneHandbookforSystematicReviewsofInterventions

wasusedtoassessmethodologicalqualityofthe

trials.Foreachquestion,a‘yes’answerindicatedalow

riskofbias(Y),‘unclear’indicatedthattheriskofbiaswas

uncertain(U),and‘no’indicatedahighriskofbias(N).

Blindingtotheallocatedinterventionswasassessedseparately

forparticipantsandoutcomeassessors.Table3presentsasummaryoftheriskofbiaswithinthe18studiesincludedinthisreview.Biaswithinmoststudieswasofhighorofunclearrisktotheprimaryoutcome.Thiswasgenerallyduetopoorreportingofsequencegeneration,allocationconcealmentorblinding.文中有設(shè)立表格，說(shuō)明偏差風(fēng)險(xiǎn)，分?jǐn)?shù)的評(píng)量特別以VAS為outcome的研究，無(wú)法做到隱匿和盲性可能會(huì)影響結(jié)果之判讀，也會(huì)影響證據(jù)的等級(jí)。本研究中，也詳加描述各研究的品質(zhì)程度。MorethanoneassessorThetwoauthors(SDYuandFYLiu)extracteddata

accordingtopre-definedcriteria.Discrepancieswerenoted

anddiscussedbetweentheauthorsandresolvedbyconsensus.

由上述各細(xì)項(xiàng)檢視，作者有提及檢視納入試驗(yàn)的品質(zhì)評(píng)估(有分是、未確定、否三個(gè)層級(jí)，但未有分?jǐn)?shù)性的評(píng)分機(jī)制)，並經(jīng)由兩位評(píng)估者進(jìn)行檢視，因此回顧者有針對(duì)納入研究進(jìn)行品質(zhì)評(píng)估。5.Iftheresultsofthestudieshavebeencombined,wasitreasonabletodoso?

Yes

Considerwhether:theresultsofeachstudyareclearlydisplayedtheresultsweresimilarfromstudytostudy(lookfortestsofheterogeneity)舉例：The

resultsofthemeta-analysisshowedithadsignificantdifference

betweenacupunctureandshamtreatments.Analysis

oftheacupunctureandshamtreatments(measuredwith

simplyscoring)favoredtheregimen(RR=2.49,95%

CI=1.06–5.84).Thereweresignificantdifferences

betweenthesetwogroups(p=0.04),whilenoheterogeneity

existedbetweenthesetwoRCTs(I2=39.9%).thereasonsforanyvariationsinresultsarediscussedTheacupuncturepointsusedinthese18RCTsweredifferent;

thismaybebecauseofthespecificityofacupuncture

trials.Thenumberoftreatmentsessionsalsovariedfrom

studytostudy,althoughformostoftheRCTsconductedin

China,about30sessionswereheld.Thisvariablelikelyhas

aneffectontreatmenteffectivenessandfurthermorecould

bethereasonforunsatisfactoryresultsinstudiesinvolving

fewertreatmentsessions(suchasthoseEnglishstudies).由上述各細(xì)項(xiàng)檢視，包括各研究詳細(xì)的結(jié)果、研究之間的同質(zhì)性以及可能影響分析結(jié)果的各研究變異性，作者均有提及，因此併合研究分析是合理的。6.Howaretheresultspresentedandwhatisthemainresult?Consider:howtheresultsareexpressed(e.g.oddsratio,relativerisk,etc.)由上表中可知研究結(jié)果以relativerisk作為呈現(xiàn)以此圖表為例，比較各研究針灸與一般西醫(yī)治療的差異性，以RR(relativerisk)為統(tǒng)計(jì)指標(biāo)，並配合95％CI

(信賴(lài)區(qū)間)評(píng)估。各單獨(dú)研究，當(dāng)RR在1的左右任何一端，且CI線(xiàn)不超越1的另一頭，視為有顯著差異。此外，在不同分組群裡，分別做出混合權(quán)重的統(tǒng)合分析，並以p-value評(píng)估顯著差異性，最後再做整體的統(tǒng)合分析，最終顯示RR=1.18,95%CI=1.02–1.37,p=0.03，表示針灸治療組與西醫(yī)一般醫(yī)療組相比是有顯著差異性的。howlargethissizeofresultisandhowmeaningfulitisThe18RCTsincluded1,086patients:580patientsineither

atraditionalacupunctureorelectricalacupuncturegroup

and506patientsinthecontrolgroup.Moststudieshad

relativelysmallsamplesizes.Onlytwohadmorethan100

samples.Meanageamongstudypatientswas

46years(rangingfrom18to72)and49%ofthepatients

weremales.Thecausesofthetinnitusreportedinthese

studieswerelargelyunknown.Thedurationoftinnitus

amongthepatientsvariedwidely,rangingfrom3daysto

30years.本研究完整性蒐集各國(guó)的針灸治療耳鳴臨床試驗(yàn)，是至今最完整的一篇系統(tǒng)性回顧文章。howyouwouldsumupthebottom-lineresultofthereviewinonesentenceTheresultsofthisreviewsuggestthat

acupuncturetherapymayoffersubjectivebenefittosome

tinnituspatients.Acupuncturepointsandsessionsusedin

Chinesestudiesmaybemoreappropriate,whereasthese

studieshavemanymethodologicalflawsandriskbias,

whichpreventsusmakingadefinitiveconclusion.本系統(tǒng)性回顧說(shuō)明針灸可能提供耳鳴患者有臨床實(shí)質(zhì)性效益，但因過(guò)去臨床試驗(yàn)研究存在許多方法學(xué)上的缺陷，因此未來(lái)若有更完整的臨床試驗(yàn)，將提供更好的實(shí)證來(lái)說(shuō)明針灸的療效性。7.Howprecisearetheseresults?Consider:ifaconfidenceintervalwerereported.Wouldyourdecisionaboutwhetherornottousethisinterventionbethesameattheupperconfidencelimitasatthelowerconfidencelimit?如右圖，表達(dá)針灸治療組與對(duì)照組是否有差異，有使用信賴(lài)區(qū)間，當(dāng)在RR=1的右側(cè)側(cè)時(shí)，我們就會(huì)選擇CI的下限來(lái)看是否越過(guò)RR=1之線(xiàn)；當(dāng)在RR=1的左側(cè)側(cè)時(shí)，我們就會(huì)選擇CI的上限來(lái)看是否越過(guò)RR=1之線(xiàn)，因此我們會(huì)端視情況來(lái)恰當(dāng)使用信賴(lài)區(qū)間。ifap-valueisreportedwhereconfidenceintervalsareunavailable

如上圖所示，當(dāng)無(wú)使用信賴(lài)區(qū)間表達(dá)時(shí)，p-value的適當(dāng)統(tǒng)計(jì)評(píng)估，均有被使用到綜合上述，本研究的統(tǒng)計(jì)精準(zhǔn)度是可被信賴(lài)的。8.Cantheresultsbeappliedtothelocalpopulation?

YesConsiderwhether:thepopulationsamplecoveredbythereviewcouldbedifferentfromyourpopulationinwaysthatwouldproducedifferentresultsIthad

differenceinacupuncturepointsandsessionsbetween

ChinesestudiesandEnglishstudies.yourlocalsettingdiffersmuchfromthatofthereviewWe

limitedoursearchtoreportswritteninChineseandEnglish,

whichmaycauseapotentialpublicationbias.Results

mighthavebeendifferentifwehadincludedreports

writteninKorean,Japaneseorotherlanguages;thiswould

haverequiredadditionallanguageskillsofourteam

members,agoalwhichwecanworktowardsforfuture

studies.youcanprovidethesameinterventioninyoursetting由以上細(xì)節(jié)檢視，本研究包含中文與英文試驗(yàn)，且中文試驗(yàn)多在中國(guó)執(zhí)行，東西方族群的研究差異可供我們參考，這樣完整收納的文章，對(duì)應(yīng)用於臺(tái)灣患者的評(píng)估，是不錯(cuò)的選擇，且裡面所提的研究設(shè)計(jì)也合乎臺(tái)灣患者的治療模式，因此，本研究的結(jié)果，可作為我們解決臨床問(wèn)題的一個(gè)良好參考。9.Wereallimportantoutcomesconsidered?

NoConsideroutcomesfromthepointofviewofthe:individualTheresultsofthisreviewsuggestthatacupuncturetherapy

mayoffersubjectivebenefittosometinnituspatients.

policymakersandprofessionals未提及

family/carers未提及widercommunityResults

mighthavebeendifferentifwehadincludedreports

writteninKorean,Japaneseorotherlanguages;thiswould

haverequiredadditionallanguageskillsofourteam

members,agoalwhichwecanworktowardsforfuture

studies由上述細(xì)節(jié)檢視，本研究結(jié)果，作者有推論於應(yīng)用於個(gè)別患者的效益，並針對(duì)研究限制，推展未來(lái)的研究目標(biāo)，如此可應(yīng)用於更廣闊的族群上。但針對(duì)本研究可提供的醫(yī)療政策制定或?qū)ζ浼彝?、照護(hù)的相關(guān)效益，作者並未表述，因此本研究並無(wú)將各面向之結(jié)果一一呈現(xiàn)。10.Shouldpolicyorpracticechangeasaresultoftheevidencecontainedinthisreview?

No

Consider:whetheranybenefitreportedoutweighsanyharmand/orcost.Ifthisinformationisnotreportedcanitbefilledinfromelsewhere?Datafromthe

articleswerevalidatedandextractedusingapredefined

dataextractionform.NearlyallofChinesestudiesreported

positiveresults,whilemostofEnglishstudiesreported

negativeresults.