提高中草藥隨機(jī)對照試驗(yàn)的質(zhì)量Ⅰ：臨床試驗(yàn)設(shè)計(jì)和方法學(xué)

1、進(jìn)步中草藥隨機(jī)對照試驗(yàn)的質(zhì)量：臨床試驗(yàn)設(shè)計(jì)和方法學(xué)卞兆祥，李幼平，劉良，吳泰相，繆江霞，關(guān)家倫，宋麗【關(guān)鍵詞】,隨機(jī)對照試驗(yàn)摘要目的：通過對中草藥臨床隨機(jī)對照試驗(yàn)的設(shè)計(jì)及方法學(xué)進(jìn)展質(zhì)量評價(jià)，討論如何進(jìn)步中草藥臨床試驗(yàn)的質(zhì)量。方法：文獻(xiàn)檢索2022年7月前發(fā)表于hrane圖書館的中草藥系統(tǒng)評價(jià)共11篇，包含167個中草藥臨床隨機(jī)對照試驗(yàn)。質(zhì)量評價(jià)方法采用修訂版NSRT聲明9項(xiàng)指標(biāo)以及中草藥劑型及質(zhì)量控制標(biāo)準(zhǔn)指標(biāo)。結(jié)果：所有167個臨床試驗(yàn)都含有試驗(yàn)?zāi)康?、方法、?結(jié)局指標(biāo)、統(tǒng)計(jì)學(xué)方法及中藥劑型；其中16397.6%個臨床試驗(yàn)說明了研究對象的納入標(biāo)準(zhǔn)，只有2615.6%個臨床試驗(yàn)說明了研究對象的排

2、除標(biāo)準(zhǔn)；只有148.4%個臨床試驗(yàn)詳細(xì)說明了隨機(jī)序列的產(chǎn)生方法；42.4%個臨床試驗(yàn)提及了隨機(jī)分配隱藏；絕大局部的臨床試驗(yàn)86.8%屬于開放性的，只有13.2%的臨床試驗(yàn)采用了盲法設(shè)計(jì)；只有1個臨床試驗(yàn)在試驗(yàn)前進(jìn)展了樣本含量的計(jì)算；在中草藥劑型方面，45.5%的臨床試驗(yàn)使用的是湯劑或中藥茶包，只有1個臨床試驗(yàn)提及了制劑的質(zhì)量控制。在167個臨床試驗(yàn)中，所有質(zhì)量評價(jià)指標(biāo)的涉及率只有36.0%。結(jié)論：現(xiàn)階段中草藥臨床隨機(jī)對照試驗(yàn)的質(zhì)量還很低。建議：1試驗(yàn)設(shè)計(jì)者及施行者必須承受正規(guī)的臨床試驗(yàn)根底知識的培訓(xùn)；2推薦采用臨床試驗(yàn)設(shè)計(jì)流程圖，逐一解決臨床試驗(yàn)過程中的關(guān)鍵問題；3在方案正式施行前進(jìn)展預(yù)試驗(yàn)，

3、并根據(jù)預(yù)試驗(yàn)的結(jié)果對臨床試驗(yàn)設(shè)計(jì)方案進(jìn)展調(diào)整；4對臨床試驗(yàn)設(shè)計(jì)的最終方案進(jìn)展注冊登記，并預(yù)先發(fā)表最好是網(wǎng)上發(fā)表臨床試驗(yàn)設(shè)計(jì)方案；5廣泛開展國際合作，特別是與對中醫(yī)藥研究感興趣的國際知名學(xué)術(shù)研究機(jī)構(gòu)進(jìn)展合作，以進(jìn)步中草藥臨床研究的質(zhì)量。關(guān)鍵詞隨機(jī)對照試驗(yàn);中草藥;方法學(xué);質(zhì)量評價(jià)Iprvingthequalityfrandizedntrlledtrialsinhineseherbalediine,part:linialtrialdesignandethdlgyABSTRATbjetive:Tdisussthequalityfrandizedntrlledtrials(RTs)inhinesehe

4、rbalediine(H)ithrespettdesignandethdlgy,andprvidesuggestinsfrfurtheriprveentinfuturelinialtrials.ethds:AsearhfthehraneLibraryasndutedtidentifyRTsfHnlineinJuly2022.QualityftheRTsasassessedusinga11iteheklistdifiedfrtherevisedNSRTstateent,ith2itesspeifitH(i.e.herbpreparatinfrandqualityntrlfherbs).Resul

5、ts:Thesearhyielded167RTsthatereseletedfrassessent.Alltrialsinludedstateentsabuttheinterventins,bjetives,priaryutedesign,statistialethds,andherbpreparatinfr.Althugh163(97.6%)trialsreprtedinlusinriteria,exlusinriteriaerenlyreprtedin26(15.6%)trials.Feerthan10%ftrialslearlystatedtherandallatinsequenegen

6、eratinethds,andnly2.4%entinedallatinnealent.Thevastajrity(86.8%)ftrialserepenlabel,hilenly13.2%usedblinding.Alsthalf(45.5%)adinisteredtheHinterventinasateardetin.nlynetrial(0.6%)reprtedasaplesizealulatin,andasingletrial(0.6%)disussedqualityntrlftheHinterventin.nlusin:TheverallethdlgiqualityfRTsinHas

7、pr.ItisessentialtiprvethedesignffutureRTsinthislinialarea.Reendatins:(1)InvestigatrndutingRTsshuldhavefraltrainingabutlinialtrialdesign;(2)Aflhartisreendedtensurethatallessentialstepsflinialtrialdesignareinluded.(3)ndutingpiltstudiesprirtRTsayhelpiprvethEirdesign;(4)Registratinflinialtrialsandpublis

8、hingtheirprtlsprirtenrlentayreduepubliatinbiasandsliitpeerreviesftheprpseddesign;(5)llabratinbeteenHinvestigatrsandtraditinalediineaadeiresearhentersinterestedinintegrativeediineayleadtqualityiprveentfRTsfH.KEYRDSrandizedntrlledtrial;hineseherbalediine;ethdlgy;qualityassessent1INTRDUTINhineseherbale

9、diine(H)isbeinginreasinglyppularinindustrializednatinsasnefrf“alternativer“pleentaryediine.Inrdertbringitfullyintthenventinaledialsystesftherldandtherebyutilizeitsnsiderablebenefitsevidenefthesafetyandeffiayfherbsandherbalprdutsareneessary.Thequestinthenis“hereannefindsuhevidene?Bthrandizedntrlledtr

10、ials(RTs)andsysteatireviesarenlythug rvidethestrngestlevelfevideneregardingtreatenteffiayfpetingtherapeutiinterventins.TheredibilityftheevidenetsupprtatreatentapprahsuhasHtherefredependsnthequalityfRTs.Previusreprts1,2haveshedthatsysteatireviesinpleentaryediineinvlvingHlakhighqualityRTstprvidelearev

11、idenefeffiay.Infat,thequalityfRTsithinterventinfHasadalityhasbeenatpifdisussinfrlngtie3,4,thughnevideneasavailabletakeadefinitivejudgentnthetpi.If,asseaysuspet,thequalityfRTsregardingHisntinfatsatisfatry,adisussinustthenfllregardingreendatinsfriprvingthissituatin.ehavetherefreprpsedafurpartseriesfar

12、tileshihfusnthefurbasieleentsfRTsinH:(1)linialtrialdesignandethdlgy,(2)ntrlgrupdesign,(3)qualityntrlfHusedinRTsand(4)reprtingfratfRTsinH.elldesignedRTsereneidelyregnizedasprvidingthestrngestevideneftheeffetivenessfhealthareinterventins.iththedevelpentfsysteatirevieandusagefetaanalytialtehniques,syst

13、eatireviesfRTsarenthug rvidethebestlevelfevideneabuttheeffetivenessfaninterventin5.Hever,suhlaisarefundedntheassuptinthatRTsarefsuffiientquality,espeiallyithrespettthelinialdesignandethdlgy,andthatthedetailsarereprtedlearly.Infat,theethdlgyfRTsdeterineshitsresultsanbeinterpretedandtheextentthihitsre

14、sultsanbetrusted.Previusstudieshaveshnthaterrneusnlusinsanbedranbasednisinterpretatinsfastudysdesignandliitatins6,7.Anyinadequateethdlgialapprahes,suhaspatientseletin,saplesizealulatin,randizatinpredure,uteassessent,handlingfdrputsandfllups,illthreatenthevalidityfalinialtrialithptentiallyexaggerated

15、treatenteffets8.InrdertassessthequalityfthelinialdesignandethdlgyinstudiesfH,thegalfthisstudyastreviethequalityfrelevantRTsinH,andtprvidereendatinsfriprvingtheinthefuture.2ATERIALSANDETHDS2.1IdentifiatinfrandizedtrialseidentifiedRTsrelatedtHfrthehraneLibrarydatabasefsysteatireviesinJuly2022.Thesearh

16、strategyusedasasflls.Step1:“hineseherbalediineasusedasasearhter,hihyielded25systeatirevies.Step2:SearhrerdsererevieedfrtheirrelevanetH.Thereere11reviesrelevanttHthatereinluded(Table1),and14reviesthaterentrelatedtHandthusexluded(Table2).Step3:PriarystudiesfRTsfHfrthe11revieseretabulatedandrssreferene

17、dteliinatedupliates.Thisprduedalistf167RTsand2quasiRTs;thelatterereexludedduettheirdesign.Thislistf167RTsasthebasisfrfurtherqualityassessent.2.2Assessentfethdlgyqualityethdlgiqualityasdefinedasnfidenethatthetrialsdesign,ndut,analysis,andpresentatininiizedravidedbiasesinthetrialsfindings9.Inthispaper

18、,eadptedanineiteheklistrelatedttheethdlgyfRTfrtherevisednslidatedstandardsfreprtingtrials(NSRT)stateent10,andaddedtitesespeiallyfrHinvlvingpreparatinfrfherbsandqualityntrlfH.Thus,theheklistaspsedfelevenites(Table3).Table1InludedhraneLibrarysysteatireviesabuthineseherbalediine略Table2ExludedhraneLibra

19、rysysteatireviesabuthineseherbalediine略Table3heklistfRTsithhineseherbalediine略2.3DataextratinandanalysisTbservers(ZhaXiangBIAN,AndreK.L.KAN)assessedthequalityfRTethdlgyardingttheheklistdesribedabve.Alldisagreeentsduetinauratedataextratinerereslvedthrughfurtherverifiatinftheriginalartiles.Inallases,n

20、sensusbeteenthetbserversasahievedprirttheanalysis.ThedatafrtrevieersereenteredintanExelfilefranalysis.3RESULTSeidentified167RTsrelatedttheusefHfr11differentnditins.Table4reprtsthebasiharateristisftheseRTs,inludingtheirindiatin,partiipants(ttalnuber,saplesizerangeandaveragesize),nuberfHinterventinste

21、sted,asellasthejurnalandyearfpubliatin.Attalf18058partiipantserestudiedin167RTs,ithindividualtrialsenrllingbeteen20and374(edian107)fthe.Eighteentrials(10.8%)ntndutedinainlandhinaerepublishedinEnglishjurnals,hile149RTs(89.2%)ndutedinainlandhinaerepublishedinhinesejurnalsntlistedinthe2022editinfSienei

22、tatinIndexJurnalitatinReprts.Table5reprtsthe11iteheklistentinedinthe167RTs.verall,36.0%(903/2505)ftheitesntheheklistfrallstudieserereprtedintheartiles.Ites2,3,10intheheklistserefullyet,hilese(ites1,4,5,6,7,9,11)erepartiallyet,andtherserepletelyitted(ite8).All167trials(100.0%)hadanexplanatinabutinter

23、ventins,generalbjetives,priaryutedesignandpreparatinfrsfherbsused.Althugh163(97.6%)trialsreprtedinlusinriteria,exlusinriteriaerenlyentinedin26(15.6%)trials.nlyfurtrials(2.4%)rialreprtedapririestiatinfsaplesize.nly8.4%(14/167)ftrialslearlystatedsequenegeneratinethds.nly4trials(2.4%)entinedrandizatinn

24、ealent.nly22trials(13.2%)ipleentedblindinginthetrials,hilerethan86%erepentrials.Frthsesingleblindanddubleblindtrials,ntalltrialsgavelearexplanatinsabuttheblindingethd.nlynetrialentinedqualityntrlftheHinterventin,andthistpiillbedisussedfurtherinPartfthisfurpartseriesfartilesnRTsinH.Table4Basiharateri

25、stisfinludedRTs略Table5Breakdnfallitesrelatedithethdlgy略4DISUSSINurreviedisveredthatanyRTsfHarefprethdlgialquality,uhlikeearlyRTsfrtherdisiplinesinintegrativeediine,suhashepathyandaupunture4.Despitethisveralleakness,stRTsdidadequatelyreprtthespeifibjetivesftheirstudy,asellastheinterventins,priaryutes

26、andpreparatinfrsftheherbsused.theraspetsfethdlgy,suhasseletinfpartiipants,saplesizealulatin,randizatin,allatinnealentandblinding,erestlyignred.Adisussinftheiprtaneftheseharateristisisasflls.4.1SeletinfpartiipantsTheseletinfpartiipantsservesthepurpsefensuringthatallpartiipantsareequalintersfdiseases,

27、ensuringthatthefindingsintheRTsauratelyrepresenthatisfrtheinterestftheppulatin,andhelpingedialpratitinersdeidehetherRTresultsanbegeneralizedtthepatientsthatpresentttheirffie.Absentstritinlusin/exlusinriteria,thereaderayntknhetherthetrialresultsanrefletthetrueeffetfinterventinsinthetargetppulatin,and

28、hetherthetrialresultsanbegeneralizedttheirnpratie.ItistherefreessentialthateveryRTshuldbelearlydefinedhsubjetsereseletedfrenrlent.Theinlusinriteriaaitdefinetheainharateristisfthetargetppulatin,hileexlusinriteriaeedutsubsetsfthetargetppulatinthatightunderinethevalidityftheresults.nreasnsfrexlusinarea

29、sflls:ahighlikelihdfbeinglsttfllup,aninabilitytprvidegddata,beingathighriskfsideeffets,rharaterististhatakeitunethialtithhldtreatent.Failuretspeifyexlusinriteriaaythreatentheinternalnsistenyfthestudy.Frexaple,netrial11aiedttestanherbalixturentype2diabetesusedrldHealthrganizatin(H)riteriafrinlusin,bu

30、titdidntspeifyanyexlusinriteria.Patientseetinginlusinriteriabutpresentingithvariusrbiditiesuldtherefrebeenrlled,despitebaselinedifferenesintheirhealthstatusthatayaffetantiipatedrespnsettreatent.nthetherhand,verlyrestritiveexlusinriteriailljepardizeenrlentandliitgeneralizatinfstudyfindingsttherpatien

31、tppulatins.Speifyinginlusin/exlusinriteriaarealsiprtantfrsysteatireviestdeidehethergrupsfrseparatestudiesanbebinedfretaanalysis.4.2Saplesizealulatinfthe167RTsrevieed,nly1(0.6%)ndutedapririsaplesizealulatin.AninreasingnuberfhealthrelatedjurnalsarerequiringRTstdesribethispressinardaneiththeNSRTstateen

32、t.Thisprvidesreadersiththedetailsusedtperfrthealulatin,andhelpstensurethattrialsillbelargeenughsthattheinvestigatrsanbservethedifferenestheyareinterestedindeteting.Ifasaplesizeistsall,thentheresearhightntdetetthepreseneftruedifferenebeteentdifferentinterventins(i.e.typeerrrrfalsenegative),andthestud

33、yightbeaastefresuresandptentiallyunethial.nthetherhand,asaplesizeustntbeexessivelylargesastasteresures,espeiallytiespentinreruitingandsreeningptentialpartiipants,nrunneessarilyexpsetanypartiipantstanexperientalinterventin.TherearesevenbasieleentsrequiredtndutasaplesizealulatinfraRTinhiheffiayispared

34、nanueriutebeteenntrlandtreatentgrups:(1)designthestudyteetthespeifiais;(2)setaeptableliitsfrtype()(i.e.pssibilityffalselydelaringapsitiveeffetfrthetreatenthenthereisnne,ftensetat0.05)andtype()(i.e.pssibilityffalselydelaringneffetfrthetreatenthenthereisne,ftensetat0.2)errrs;(3)deterinetheiniulinially

35、eaningfuldiffereneandvarianebeteengrupsfrprirliteraturerexpertpinin(e.g.treatentillresultina20%iprveentparedt0%frtreatent);(4)selethetherthetestillbenesided(i.e.isthetreatentbetterthanthentrl)rtsided(i.e.isthetreatentbetterrrsethanthentrl);(5)estiatetheretentinrateatthefinalflluppintandadjustsaplesi

36、zeardingly(e.g.itha50%drputrate,thesaplesizeustbedubled);(6)alulatetherequirednuberfstudypartiipantsineahgrup;(7)revisestudyparaetersasrequired12,13.Slightvariatinsfand,andexpetedeffetandvarianeinbthgrupsillaffetthefinalsaplesizealulatinandtherefretrialsts.Frexaple,reduing,inreasingper,andreduingthe

37、expeteddifferenebeteengrupsillinreasethesaplesize14,15.Theapprpriateeffetsizevariesidelybeteenstudies,sineitshuldrepresentthesallesteffetthatuldberegardedasliniallyeaningfulandiprtant.Thislastpintustbeephasizedsinestudiesthatenrllanypartiipantsayreprtstatistiallysignifiantdifferenesthataretsalltbefi

38、nteresttliniians(e.g.a5%redutininpain).Reprtingthesedifferenesbeteenthetreatentandntrlgrupsaygiveafalseipressinthataninterventinisbenefiial.therfatrstnsiderfrthisstepinludediseaseppulatin,treatentst,fundingfrtrials,et.Therefre,estiatinftheeffetsizefinterestshuldrefletbthlinialauenandtheptentialpubli

39、healtheffet.linialinvestigatrsustpartiipateinthispresssinestatistiiansarentlikelytgenerateusefulsaplealulatinsithuttherequireddata12.ThefeRTsthatreprtedalulatingthesaplesizedidntgivetheethdanddetailsusedtperfrthisalulatin,severelyliitingitsusefulness.Itisneessarythatsaplesizeshuldbealulatedbefrethel

40、inialtrialbegins,andalulatinethdsshuldbereprtedlearly.4.3RandizatinRandizatinisaperfultltassignatargetsaplettreatentgrupsbalanednallpssibleknnandunknnnfunders,thushelpingtavidsysteatibias.ThequalityfresultsfrRTsishighlydependentnthequalityfrandizatin.Theiprtantfatrinthepressfrandizatinistheallatinse

41、quenegeneratin;nversely,sequenegeneratinethdsareusedtjudgethequalityfrandizatin.Unfrtunately,urdatashedthatjust8.4%(14/167)trialsprvidedlearstateentsabutsequenegeneratin,hile91.6%(153/167)erelystatedthatpartiipantsererandizedithutexplanatinasththeallatinsequeneseregenerated.Frthse18trialspublishedin

42、Englishjurnals,nly5trialsstatedlearlyabutthesequenegeneratinethds.Thissituatinissiilarttheredialspeialties16,17.Frexaple,inperidntlgy,althugh91%ftrialseredesribedasrandized,adequateethdsfrrandizatinerenlyreprtedin17%16.Nearly50%frthdntistrialspublishedinAerianJurnalfrthdntisandDentfaialrthpedis,Brit

43、ishJurnalfrthdntisandEurpeanJurnalfrthdntisfr1989t1998didntentintheethdusedtgeneratetherandizatinsequene17.ThisdefiienyistherefrentliitedtRTsreprtingnH.Absenefthisexplanatinillpreventreadersfrjudginghethertheethdsereprper.Thekeyissuefrtherandizatinisthatpartiipantsareallatedtdifferentarsrandly,thati

44、s,byhaneandntbyhie(invlvinginvestigatrshieandpartiipantshie),eitherthrughthesiple(unrestrited)randizatin,restritedrandizatinrstratifiedrandizatin18,19.ThesequenegeneratinethdfrandizatinillhelpdeterinethesientifiaurayandredibilityfRTs.Therefre,stritlyfllingtheneessarystepsfrrandizatin,asellaslearlyre

45、prtingtheseethds,areiprtantfrRTs.4.4AllatinnealentUndubtedly,randizatinisaperfulethdtprduebalanedtreatentgrups,hileallatinnealentistheritialfatrtensurethesuessfrandizatin.Randsequenegeneratinerelyeansthatthegrupassignentsequenefapartiipantasgeneratedrandly,butdesntensurethattheallatinsequeneasnsiste

46、ntlyflledduringipleentatin.Inrdertallatethepartiipantstbalanedtreatentgrups,thesequeneshuldbenealedtinvestigatrs,partiipants,andalltherstudypersnnel.Ifnt,seletinbiasaybeintrdued,herebythetreatentassignentisnlngertrulyrandandanibalaneinprgnstifatrsayurbeteentreatentgrups.Previusstudieshaveshnthatinad

47、equaterunlearallatinnealentanexaggeratelinialeffetsupt40%,espeiallyinprlyndutedtrials6,7,20,anditanalsausegreaterhetergenEityinresults21.Thusprperrandizatinshuldinvlvebthrandsequenegeneratinandpleteipleentatinfthatsequenetiniizebias.urstudyreprtedthatalthugh167trialslaiedtberandized,nly2.4%(4/167)re

48、prtedtheethdsfrallatinnealent,paredith17%trialsreprtingallatinnealentinperidntlgy16.Thereasnhytheratefrandizatinnealent(and/rthereprtingfit)isslisunknn,thughseresearhersbelievethatithblindingandespeiallydubleblindingallatinnealentisntneessary.Butblindingandallatinnealentaredifferent22.Blindingnentra

49、tesnpreventingstudypartiipantsandpersnnelfrdeteriningthegrupthihpartiipantshavebeenassigned(hihleadstasertainentbias);itsafeguardsthesequeneafterallatin.Inntrast,allatinnealentnentratesnpreventingseletinandnfundingbiases;itsafeguardstheassignentsequenebefreanduntilallatin.Seresearhersayntunderstandt

50、hetnepts.hateverthereasn,RTinvestigatrsneedtunderstandthesetpriniplesandtheratinalesfrusingthe.nlyusedethdstnealallatininludeallingaentral,rdinatingffiefreahpatientassignentatthetiethatthepatientpresentsfrstudyinlusin;usingsequentiallynubered,paquesealedenvelpes;andusingnuberedbttlesrntainers23.4.5B

51、lindinghyisblindingneessaryfrRTs?Briefly,thereasnisrelatediththeaiflinialtrial:tfindutthebjetiveeffiayfthetargetinterventin.Inrdertaintainthisbjetivity,eshuldakesurethattheresultserentdilutedrisledbythesubjetivepreferenes(bias)frthepartiipants,investigatrs,rassessrs24.Ifnt,alinialtrialanntgenerateau

52、rateresultsandthusannttrulyadvaneurknledgefhealthare.Frthisveryiprtantreasn,blindingisthegldstandardfrlinialtrialdesignandshuldbearriedutevenfrthsetrialsithbjetiveindexesfrassessent.theriseitilldaagethevalidityftheresultstheygenerate.Previusreprtshaveshnthattrialsthaterentdubleblindedyieldedlargeres

53、tiatesftreatenteffetthantrialsinhihauthrsreprteddubleblinding(ddsratisexaggerated,naverage,by17%)6.Therefreitisneessarytexeutetheblindingtest.Infat,theiprtanefblindingasregnizedfralngtie25,andanyresearhersrldideunderstandblindingterinlgy,butappliatinfprperblindingpreduresinRTsustbeiprved.urdatashedt

54、hatnly13.2%(22/167)ftrialseredesribedasblinded,hile86.8%(145/167)erepentrialsinhihthepatientsrdtrs(rbth)ereaareftheassignentttreatentrntrlgrup.fthese22blindedtrials,13laiedtbedubleblindedbutfailedtprvidedetailsabuththisasipleented.Inthe9singleblindedtrials,5prvidedalearexplanatinfblindingethdlgy.Thi

55、sbservatinassiilartthatintrialsfnventinalediine,herenly10.4%f173RTspublishedfr1979t2000intheJurnalfIntensiveareediinedesribedtheblindingpredure26.Inrderthangethissituatin,reattentinsshuldbegiventblindinginlinialtrials.ThisisespeiallytrueftrialsinHassessenthereeffiayisainlysubjetive.AntherhallengeinH

56、trialsisthepreparatinfrftheherbsusedasaninterventin,hihaybeadinisteredasatea,tablet,apsule,rdetin(extratfarudedrugadebybilingrsieringherbsinater,usuallyhasstrngereffetthanatearinfusin).Duetharateristidrs,flavrs,andlrs,theseayeasilybeidentifiedbypartiipantsand/rliniians.Ideally,interventinsfrdifferen

57、ttreatentgrupsshuldbegiveninthesaefriththesaerutefadinistratin,asisninnventinalediinedrugtrials.Adubleduytrialdesignuldbeusedtverethishallenge,herebyeahuldreeivetpreparatinfrs(e.g.teaandapsule),nlynefhihuldntaintheativetreatenthilethetheruldntainplaeb;thistpiillbedisussedfurtherinPartIIfthisfurparts

58、eriesfartilesnRTsinH.Insuary,blindingshuldbestrengthenedinRTsfH.Inthepressfblinding,investigatrsshulddeidehshuldbeblinded,hillbeinhargefthepress,andhatpreparatinfrftheinterventinshuldbetakenthelptheblinding.5LIITATINTheainliitatinfthisstudyisthatresultsareappliablenlytthseRTsidentifiedfrthehraneLibr

59、aryfrqualityassessent.ThughthesestudiesanntrepresentallRTsinH,theyarebelievedtfrarepresentativesaple.6NLUSINIngeneral,thequalityflinialtrialdesignandethdlgyfRTsithHisl.reattentinsshuldbegiventthedesignandethdlgyfthesetrials.Speifially,itessuhasseletinfpartiipants,randizatinsequenegeneratin,allatinne

60、alent,blinding,saplesizealulatin,andplianefpartiipantsandinvestigatrsshuldbeipleentedinallRTstiprvetheirverallqualityandensurethevalidityandusefulnessftheirresults.7REENDATINSFllingurreviefthequalityfRTsinH,ehavedevelpedsereendatinsfriprvingtheirethdlgy.(1)InvestigatrshanttexeutelinialtrialsinHshuld