The intranasal dexmedetomidine plus ketamine for procedural ...

There is evidence that a combination of dexmedetomidine and ketamine (ketodex), administered intranasally (IN), could provide adequate sedation ... Skiptomaincontent Advertisement SearchallBMCarticles Search DownloadPDF Update OpenAccess Published:06January2021 Theintranasaldexmedetomidineplusketamineforproceduralsedationinchildren,adaptiverandomizedcontrollednon-inferioritymulticentertrial(Ketodex):astatisticalanalysisplan AnnaHeath  ORCID:orcid.org/0000-0002-7263-42511,2,3,JuanDavidRios1,EleanorPullenayegum1,4,PetrosPechlivanoglou1,4,MartinOffringa1,4,5,MarynaYaskina6,RickWatts6,ShanaRimmer6,TerryP.Klassen7,8,KamaryCoriolano9&NaveenPoonai10,11onbehalfofthePERC-KIDSCANKetodexStudyGroup Trials volume 22,Article number: 15(2021) Citethisarticle 1298Accesses Metricsdetails AbstractBackgroundProceduralsedationandanalgesia(PSA)isfrequentlyrequiredtoperformclosedreductionsforfracturesanddislocationsinchildren.Intravenous(IV)ketamineisthemostcommonlyusedsedativeagentforclosedreductions.However,aschildrenfindIVinsertionadistressingandpainfulprocedure,thereisneedtoidentifyafeasiblealternativerouteofadministration.Thereisevidencethatacombinationofdexmedetomidineandketamine(ketodex),administeredintranasally(IN),couldprovideadequatesedationforclosedreductionswhileavoidingtheneedforIVinsertion.However,thereisuncertaintyabouttheoptimalcombinationdoseforthetwoagentsandwhetheritcanprovideadequatesedationforclosedreductions.TheIntranasalDexmedetomidinePlusKetamineforProceduralSedation(Ketodex)studyisaBayesianphaseII/III,non-inferioritytrialinchildrenundergoingPSAforclosedreductionsthataimstoaddressboththeseresearchquestions.ThisarticlepresentsindetailthestatisticalanalysisplanfortheKetodextrialandwassubmittedbeforetheoutcomesofthetrialwereavailableforanalysis.Methods/designTheKetodextrialisamulticenter,four-armed,randomized,double-dummycontrolled,Bayesianresponseadaptivedosefinding,non-inferiority,phaseII/IIItrialdesignedtodetermine(i)whetherINketodexisnon-inferiortoIVketamineforadequatesedationinchildrenundergoingaclosedreductionofafractureordislocationinapediatricemergencydepartmentand(ii)thecombinationdoseforINketodexthatprovidesoptimalsedation.AdequatesedationwillbeprimarilymeasuredusingthePediatricSedationStateScale.Assecondaryoutcomes,theKetodextrialwillcomparethelengthofstayintheemergencydepartment,timetowakening,andadverseeventsbetweenstudyarms.DiscussionTheKetodextrialwillprovideevidenceontheoptimaldosefor,andeffectivenessof,INketodexasanalternativetoIVketamineprovidingsedationforpatientsundergoingaclosedreduction.ThedatafromtheKetodextrialwillbeanalyzedfromaBayesianperspectiveaccordingtothisstatisticalanalysisplan.Thiswillreducetheriskofproducingdata-drivenresultsintroducingbiasinourreportedoutcomes.TrialregistrationClinicalTrials.govNCT04195256.RegisteredonDecember11,2019. PeerReviewreports BackgroundOrthopedicinjuriesarecommoninchildrenvisitingtheemergencydepartment(ED)[1,2],sometimesrequiringaclosedreduction[3,4],whichusuallyrequiresproceduralsedationandanalgesia(PSA).Intravenous(IV)ketamineisacommonsedativeagentthatfacilitatesclosedreductions[5].However,IVinsertionisdistressingforchildrenandtheircaregiversandcanbechallengingaschildrenhavesmallveinsandresistpositioning[6].Intranasal(IN)administrationofsedativeagentsisalessinvasivepotentialalternativetoIVinsertion[7]withacombinationofdexmedetomidineandketamine(ketodex)showingpromiseforPSAacrossseveralindications[8].However,itisunclearwhetherINketodexwouldoffersufficientanalgesiaandsedationforclosedreductionsinchildrenwithorthopedicinjuries,comparedtoIVketamine,andthereisnoevidenceonthemosteffectivecombinationofdexmedetomidineandketaminewithintheketodexformulation.Theintranasaldexmedetomidineplusketamineforproceduralsedationinchildren(Ketodex)studyaimstodetermine(i)whetherthesedationprovidedbyINketodexisnon-inferiortothatprovidedbyIVketamineforchildrenundergoingaclosedreductionand(ii)thecombinationdoseofINketodexthatprovidesoptimalsedation.Toachievetheseobjectives,theKetodexstudyusesaninnovativeBayesianresponse-adaptive,comparative-effectivenessdesign[9].Thispaperoutlinesthestatisticalanalysisplan(SAP)fortheKetodextrialasthestudyprotocolispublishedseparately[10].ThisSAPhasbeenpublishedbeforeanalyzinganystudydataandfollowsthereportingguidelinesforSAPs[11].ObjectivesTheprimaryaimoftheKetodexstudyistodetermineifINketodexisnon-inferiortoIVketaminewithrespecttotheproportionofparticipantswhoachieveadequatesedationforthedurationofaclosedreduction.WewillenrollpreviouslyhealthychildrenwhopresenttooneofsixCanadianparticipatingpediatricEDsrequiringareductionforafractureordislocation.ThesecondaryobjectivesofthestudyaretodeterminetheoptimalcombinationdoseforINketodexandtocharacterizethesedationcharacteristicsofINketodexwithrespecttolengthofstayintheED,needforadditionalsedation,nasalirritation,satisfactionwithsedation,andadverseevents.Methods/designDesignandsettingTheKetodextrialisaphaseII/III,double-dummy,controlled,randomized,Bayesianresponseadaptive,dosefinding,non-inferioritytrialbeingconductedinsixCanadiantertiarycarepediatricEDs.Eligibleparticipantsareage4to17 yearsandrequirePSAtoundergoaclosedreductionofafractureordislocation.Followinginformedconsent,patientswillberandomizedina2:3ratiotoreceiveeitherIVketamine,dosedat1.5 mg/kg(maximum100 mg),withanINplacebocombinationtreatment,oroneofthreecombinationdosesofINketodex,i.e.,2,3,or4 mg/kg(maximum400 mg)ofketaminecombinedwith4,3,or2 μg/kg(maximum200 μg)ofdexmedetomidine,respectively,withanIVplacebo.Patientswillthenbefurtherrandomizedtoreceiveoneofthethreedosecombinations,initiallyina1:1:1ratio(see“Randomization”).MoststudydatawillbecollectedintheEDwithin1–2 hofstudyenrollment.Afollow-upphonecallwillbemadebetween24and48 hafterenrollmenttocollectdataonissueswiththesedation.Ifaparticipantislosttotelephonefollow-up,aresearchnursewillattempttoregaincontact3–5times,usingemailoralettertotheparticipant’slastknownaddress,dependingonlocalresearchethicsboard(REB)regulations.StudyprotocoldevelopmentandconductTheKetodexstudywasregisteredonClinicalTrials.govonDecember11,2019,withregistrationnumberNCT04195256.Ateachinstitution,theREBwillgiveethicalapprovalbeforecommencinglocalenrollment.Informedconsent,followinginstitutionalREBguidelines,willbeobtainedfromcaregiversbeforerandomizationanddatacollection.TheKetodexstudywillbesupportedbytheKidsCAN-PERCiPCTnetwork[12],aCanadiantrialsnetworkusingcentralizedinfrastructurefordatamanagementandtrialoversightforfourtrials.Anindependentdataandsafetymonitoringboard(DSMB)ofsixoverseesthestudy.RandomizationParticipantswillberandomizedinaratioof2:3toeitherIVketamineorINketodexusingblockrandomizationwithblocksofsizefive,stratifiedbysite.ParticipantswillundergoafurtherrandomizationtooneofthethreeINketodexcombinationsusingtheREDCapelectronicdatacapturesystem[13].ForparticipantsrandomizedtoIVketamine,thesecondrandomizationwillbetotwoplacebo(saline)solutions.Thesecondrandomizationstepwillbeadaptedsomoreparticipantsreceivethemoreeffectivedosecombinations.Figure 1displaysthistwo-steprandomizationprocedure. Fig.1DepictionoftherandomizationprocedurefortheKetodextrialFullsizeimageTheinitialrandomizationratioforthedosecombinationswillbe1:1:1.Interimanalyseswillupdatetherandomizationratioatapproximately150,200,250,300,and350participants.Asoutlinedinthe“Analysisfortheprimaryendpoint”section,eachinterimanalysiswillcomputetheposteriorprobabilitythateachdoseisthemosteffective.Thenumberofpatientsrandomizedtoeachdosebeforethenextinterimanalysiswillequalthisposteriorprobabilitymultipliedbythenumberofparticipantstobeenrolledbeforethenextinterimanalysis,roundedtothenearestwholenumber.If,atagiveninterimanalysis,theprobabilitythatadoseisoptimalfallsbelow0.05(thresholdchosenbysimulation),wewillnotrandomizeparticipantstothisdose.Ifallthreecombinationshaveaprobabilityofbeingthemosteffectiveoflessthan50%once250participantshavebeenenrolled,thensafetyandtolerabilitywillbeassessedtodeterminethemostpromisingcombination.ThisdosecombinationwillbeusedforallparticipantsrandomizedtoINketodex.TherandomizationlistforthefirststepwillbegeneratedandheldsecurelyattheWomenandChildren’sHealthResearchInstitute’s(WCHRI)DataCoordinatingCentreattheUniversityofAlberta[14].PharmacieswillprepareidenticallyappearingIVketamineplusINsalineorINketodexplusIVsalinekitsforuseatthebedside.Thesecondrandomizationwilltakeplaceatthebedsidefromamasterrandomizationlist,accessedsequentiallyasparticipantsareenrolledacrossallsites.Site-levelstratificationisnotusedforthesecondrandomization.Theresearchnursewillbeblindedtotheinterventionbutnottothedosecombination.Outcomeassessorswillbeblindedtothedosecombinationandtheintervention.DatastorageandcollectionAllparticipantdatawillbestoredinREDCap[13]andheldattheWCHRIDataCoordinatingCentre[14].AlldatawillbeenteredintoREDCapusingaWiFi-enabledencryptediPadorrecordedonpaperintheeventofatechnicalfailure.Priortoanalysis,allpersonalidentifierswillberemovedandparticipantswillbeidentifiedusingauniquestudyidentificationnumber.PrimaryoutcomeTheprimaryoutcomeisadequatesedationforthedurationoftheclosedreduction.Thiswillbeascertainedovertheintervaloftimefromthefirstapplicationoftractionormanipulationoftheinjuredlimbforthepurposeofanatomicalrealignmenttothelastapplicationofarealigningforce.Adequatesedationisdefinedasfulfillmentofallthreeofthefollowingcriteria: i) APediatricSedationStateScale(PSSS)[15]scoreof2or3forthedurationoftheprocedureAND ii) NoadditionalmedicationgivenduringtheclosedreductionforthepurposeofsedationAND iii) Thepatientdidnotactivelyresist,cry,orrequirephysicalrestraintforcompletionoftheclosedreduction. TheprimaryoutcomewillbeassessedbytwoindependentoutcomeassessorsblindedtothestudygroupandobjectiveswhowillusevideorecordingsoftheclosedreductiontocalculatethePSSSscoreevery30 sduringtheprocedure.Videoscoringforeachparticipantwillbeundertakenwithin24–48 hofdatacollection.Thesecondoutcomeassessorwillscore25%oftheparticipantstogenerateaninterrateragreement.SecondaryefficacyoutcomesTheKetodextrialwillinvestigatethreesecondaryoutcomes: 1. Lengthofstay:Time(inminutes)recordedinthemedicalrecordbetweenEDtriageandEDdischarge. 2. Timetowakening:ThisisthetimeintervalbetweenthefirstpairorINspraystothefirstPSSSscore > 3post-closedreduction. 3. Adverseevents(AEs):TheoccurrenceofanyAE,recordedusingthemedicalrecordandqueriesofthehealthcarestaffduringsedationandrecovery.ResearchnurseswillbetrainedontherecognitionanddefinitionofexpectedandunexpectedAEs.AEswillbebasedonHealthCanadareportingstandards. AdditionaloutcomesTheKetodextrialwillrecordsevenadditionaloutcomes: 1) LengthofstayduetoPSA:Time(inminutes)fromthefirstpairofINsprays/IVdosetoEDdischarge. 2) Durationofprocedure:Time(inminutes)ofthefirstpairofINsprays/IVdosetotheendofcastorsplintapplication(closedreduction). 3) LengthofEDstay:timeintervalbetweentriageassessmentanddischarge. 4) Caregiver,participant,bedsidenurseorrespiratorytherapist,andphysiciansatisfaction:Satisfactionfromthecaregiverandparticipant,measuredusinga100-mmvisualanalogscale,obtainedimmediatelypriortodischarge.Satisfactionfromeachhealthcareprovider,measuredusinga100-mmvisualanalogscaleimmediatelyfollowingcast/splintapplication. 5) Nasalirritation:Discomfortassociatedwithnasalsprays(ifrecalled),assessedbytheresearchnurseusingtheFacesPainScale-Revisedatdischarge[16]. 6) Volumeofintervention:thevolumeofINinterventionapatientreceived,comparedtothevolumeofINinterventiontheywerecalculatedtoreceive,recordedattheindexvisit. 7) AdjunctiveIVtherapyandmedications:therequirementofanIVfortherapyunrelatedtosedation. 8) Patientpreferenceforthemethodofsedation:Wewillasktheparticipanttochoosetheirpreferredsedationmethod,iftheyhaveone,attheindexvisit SamplesizecalculationWeusedtheaveragelengthcriterion(ALC)forBayesiansamplesizeestimation[17]byselectingthesmallestsamplesizesuchthatthe95%high-densityposteriorcredibleintervalforthedifferenceintheprobabilityofadequatesedationhadanaveragelengthof0.07,sixtimesshorterthantheprior95%highest-posteriordensitycredibleinterval(“Analysisforprimaryendpoint”).Thisreductionwasselectedbasedonpracticalconstraints[17].WealsousedALCtodeterminetherandomizationratiobetweenIVketamineandINketodex;weconsideredrandomizing20%,30%,40%,and50%ofpatientstoIVketamine.TheALCinitiallyselectedthesmallestsamplesizeforwhichtheaveragelengthofthe95%high-densityposteriorcredibleintervalfellbelow0.07andthenselectedtherandomizationratiothatledtothemostbalancedtrial,providedtheALCremainedbelow0.07.Wesimulated2000samplesfromthepriorpredictivedistributionofthedata,acrossallfourrandomizationregimes,forsamplesizesincreasingfrom350to500inincrementsof10.Foreachsimulation,weestimatedthelengthofthe95%high-densityposteriorcredibleintervalusing2000simulationsfromtheposterior.Basedonthis,thesamplesizefortheKetodextrialis410patientswitha2:3randomizationratiobetweenIVketamineandINketodex.Weexpectminimalmissingdataastheprimaryoutcomeiscollectedduringaprocedurethatmustbecompletedbythephysicianbeforedischarge.Thus,thesamplesizeisnotadjustedforlosstofollow-up.Wewillmonitormissingnessandadjustourrecruitmenttargettoensurethat410patientsrecordtheprimaryoutcome.InterimanalysisandstoppingguidanceTheKetodextrialwillhaveseveninterimanalyses,atincrementsof50enrolledparticipants.TheDSMBwillreviewsafetyoutcomesDSMBateachinterimanalysisbasedondescriptivestatisticsofthesafetyoutcomesbetweentreatmentgroups.Theymayalsoreceiveposteriorcredibleintervalsorpredictiveprobabilities.ThedecisiontostopthetrialforsafetyreasonsisatthediscretionoftheDSMB.Duetotheuneventreatmentallocation,theDSMBwillbeunblindedtotreatmentassignment.Therandomizationratioatthe2ndlevelofrandomizationwillbeupdatedafterrecruitmenthitsapproximately150,200,250,300,and350participants.Wewillnotundertakecomparativeeffectivenessanalysesattheinterimanalysesorstopforefficacyorfutility.StatisticalanalysisplanStatisticalprinciplesTheprimaryanalysisoftheprimaryoutcomewilltakeplaceaftereveryparticipanthascompletedtheprotocolandalldatahavebeencollectedandcleaned.Thestatisticalanalysiswillbeperformedunblindedtoparticipantallocation.TheprimaryanalysiswilldetermineiftheoptimalINketodexcombinationisnon-inferiortoIVketamine.AllotheranalyseswilltestforsuperiorityofINketodex.Wewillundertakeanintention-to-treatanalysis,includingallrandomizedparticipants,andaper-protocolanalysis,concludingnon-inferiorityifboththeseanalysesconfirmnon-inferiority.AllanalyseswilluseaBayesianperspectivewithsignificancedeclaredbasedontheposteriorprobabilitythattheproposedhypothesisistrue.Noadjustmentswillbemadeformultiplicityduetothelikelihoodprinciple[18]withthethresholdsfordeclaringsignificancechosenusingsimulationstocontrolthetypeIerrorofthetrial[9].Specifically,ifthisprobabilityislessthan3.7%,wewilldeclaresufficientevidenceagainstthehypothesis.Iftheprobabilityisgreaterthan60.8%,wewilldeclaresignificantevidenceforthehypothesis.Ifthisprobabilityisbetween3.7%and60.8%,wewilldeclarethatthetrialisinconclusive.Wewillreporttreatmenteffectestimatesusing95%highest-densityposteriorcredibleintervals.ThestatisticalanalysiswillbeundertakeninR[19]asaninterfacetoJAGS[20].HandlingofmissingdataWeanticipateminimalmissingdataasthemajorityofoutcomesarecollectedwithintheED.Iftheproportionofmissingdataisbelow5%,wewillundertakeacompletecaseanalysis.Ifthelevelofmissingnessexceeds5%,wewilluseajointBayesianmodelforthemissingnessandtheoutcome.PatientflowWewillpresentpatientflowwithaCONSORT2010flowdiagramreportingthenumberofparticipantsdeemedeligibleforthetrialatscreeningandthoseexcludedastheymetastudyexclusioncriterionandthenumberofparticipantswhowererandomizedandreceivedtherandomizedallocation.Participantscanwithdrawfromthestudyatanytime,foranyreason.Thenumberofparticipantswhowithdrawandthenumberofparticipantslosttofollow-upwillbesummarizedbytreatmentarm.ProtocoldeviationsandadherenceAprotocoldeviationisanynoncompliancewiththeclinicaltrialprotocol,InternationalConferenceonHarmonizationGoodClinicalPractice,orTrialManualofProceduresrequirements.Anychange,divergence,ordeparturefromthestudydesignorproceduresconstitutesaprotocoldeviation.Thenoncompliancemaybeonthepartofeithertheparticipant,theparticipatingsiteinvestigator,orthestudysitestaff.Theproportionofprotocoldeviationswillbepresentedbythetreatmentgroupalongsidedescriptiveinformationaboutthedeviation.Adherencewillbedefinedasaparticipantwhoreceivedanyofthestudymedicationsandwillbepresentedbythetreatmentgroup.BaselinecharacteristicsWewillcollecttheparticipants’ageandsex,typeoffractureordislocation,locationoffractureordislocation,theidentityofthepersonperformingtheclosedreductionand,ifrequired,theidentityofthepersonperformingthecasting.Baselinedatawillbesummarizedusingfrequenciesandpercentagesforcategoricalvariablesandmeans,medians,standarddeviations,andinterquartilerangesforcontinuousvariables.Analysisfortheprimaryendpoint Interimanalyses Theadaptiverandomizationwillbebasedontheprobabilityofadequatesedationforeachdose,denotedpi,i = 1,2,3.Weassumeabinomiallikelihoodforthedataandabetapriorwithparameters6.25and0.25foreachdose.Thispriorisinformedbypreviousdata[21]andthendown-weightedby4toaneffectivesamplesizeof6.5[22].Usingthesamepriorforallthreedosesassumestheyhavethesameeffectivenessunlessthedataindicateotherwise.Thispreventsearlystoppingofadosecombination.Wewillcomputetheprobabilitythateachdosecombinationhasthehighestproportionofsuccessfullysedatedparticipantsfromtheposteriordistributionsforpi,i = 1,2,3.Thisprocedurehasan83%chanceofrandomizingthehighestnumberofpatientstotheoptimaltreatment[9]. Doseresponsemodeling Participantswhoarenotadequatelysedatedcanbeover-sedated(PSSSscoreof0or1)orunder-sedated(PSSSscoreof4or5)[15].Thefinalanalysiswilluseamonotoniclog-logisticdoseresponsemodelfortheprobabilityofover-andunder-sedation[23]andamultinomialdistributiontoinfertheprobabilityofadequatesedation.Specifically,letXi,fori = 1,2,3,beathree-vectorcontainingthenumberofpatientswhoexperienceunder-,adequate,andover-sedationfromtheNipatientsthatreceivedosei.WemodelXi~Multinomial(Ni, pi),where\({\boldsymbol{p}}_{\boldsymbol{i}}=\left({p}_i^{(1)},{p}_i^{(2)},{p}_i^{(3)}\right)\)and $${p}_i^{(1)}=\mathrm{logit}\left(\{\beta}_1+{\beta}_2\log\left({K}_i\right)+{\beta}_3\log\left({D}_i\right)\right),$$$${p}_i^{(3)}=\mathrm{logit}\left(\{\beta}_a+{\beta}_b\log\left({K}_i\right)+{\beta}_c\log\left({D}_i\right)\right),$$ withKithedoseofINketamineandDithedoseofINdexmedetomidineintheKetodexcombination.Interactionscannotbeestimatedaccuratelyandarenotrequiredindosefindingstudies[24].Theoptimaldosecombinationisthedosewiththehighestexpectedprobabilityofadequatesedation. Priorsforthedoseresponsemodel Wewillusenon-centraltdistributionswithprecision0.001anddegreesoffreedom1,suggestedby[25].Themeanofβ1andβaassumethat5%ofparticipantsareunder-sedatedand2%over-sedated,settingthepriormeansfor\({p}_i^{(2)},i=1,2,3\)to0.93,theeffectivenessfromtheliterature[21].Thepriorsforβ2,β3,βbandβcarecenteredon0soalldosecombinationsinitiallyhavethesameeffectiveness. Effectivenessanalysis TheprimaryeffectivenessanalysiswilldeterminewhetherINketodexisnon-inferiortoIVketamineintermsofprovidingadequatesedation.ThisanalysiscomparestheeffectivenessofIVketamineandtheoptimalINketodexcombination.TheprobabilityofadequatesedationwithIVketamine(pIV)isestimatedusingabinomiallikelihoodandabetapriordistributionwithparameters15.1and0.4.Thesepriorparametersareestimatedfrompublisheddata[26]anddown-weightedtopreventsignificantimpactonthetrialresults. TheprimaryanalysisisbasedontheposteriorprobabilitythatINketodexisnon-inferiortoIVketamine. γ = P(pIV − pIN > η), whereη = 0.178isthenon-inferioritymargin,estimatedfromourteam’ssurveyof204EDphysicians(outlinedinthesupplementarymaterial),andpINistheprobabilityofadequatesedationfortheoptimalINketodexcombination.WewilldeclarethatINketodexisnon-inferiortoIVketamineifγ ≤ 0.037.Othervaluesofγwillinducealternativeconclusions.ThisdecisionrulebasedonγhasatypeIerrorof5%whenpIV = 0.97andpIN = 0.792andatypeIIerrorof7%whenpIN = 0.9[9]. Analysisforsecondaryendpoints Secondaryoutcomes Forlengthofstay,onsetofsedation,anddurationofsedation,wewillusealineardose-responsemodeltoestimatethemeandurationfortheoptimalINketodexcombination.Distributionalassumptionsforthismodelwillbeassessedusingposteriorpredictivechecks[27].Ifanormaldistributionisnotsuitable,alternativemodelfunctions,e.g.,gammaandlog-normal,willbeconsidered.ForIVketamine,wewillestimatethemeanlengthofstay,onsetofsedation,anddurationofsedationbyfittingasuitabledistribution,chosenusingposterior-predictivechecks.ThecomparisonofthemeanswillthendeterminetheprobabilitythatINketodexissuperiortoIVketamineanddeclaresignificanceusingthealgorithmoutlinedabove. ToanalyzetheAEs,wewillusealogisticdose-responsemodelforINketodexandabinomialdistributionforIVketamine.WewillthencomputetheposteriorprobabilitythatoptimaldoseforINketodexhasalowerAEratethanIVketamine,declaringsignificanceusingthealgorithmabove.EachAEwillbecountedonceforagivenparticipant.Wewillalsoreporttheseverity,frequency,andrelationshipofAEstothestudyinterventionbySystemOrganClassandpreferredtermgroupings.ForeachAE,wewillalsoreportthestartdate,stopdate,severity,relationship,expectedness,outcome,andduration.AEsleadingtoprematurediscontinuationfromthestudyinterventionandserioustreatment-emergentAEswillbepresentedineitheratableoralist. Priorsforsecondaryanalyses Priorsforthemodelinterceptswillbeobtainedfromtheliterature,wherepossible,anddown-weightedtoreducepriorinfluenceontheresults.Priorsfortheregressioncoefficientswillbecentraltdistributionswithprecisionof0.001anddegreesoffreedom1[25].Priorsforthestandarddeviationswillusenon-centraltdistributions(truncatedat0)[28].Wewillensurethatthesepriorsarevaguewithrespecttothescaleoftheobserveddataandundertakesensitivityanalysestothepriorspecification. AdditionaloutcomesWewillreportadditionaloutcomeswithdescriptivestatistics,usingfrequenciesandpercentagesforcategoricalvariablesandmeans,medians,standarddeviations,andinterquartilerangesforcontinuousvariables.AdditionalanalysesWewilluselogisticregressiontoinvestigatetheinteractionbetweenbaselinepain,measuredusingtheFacesPainScale-Revised,consideredasacontinuousvariable,andtreatmenteffect.Themodelwillhaveacoefficientfortreatment,baselinepain,andaninteractiontermbetweentreatmentandbaselinepain.Thisanalysiswillonlyusedatafromtheoptimaldosecombination.Wewillconcludeasignificantinteractioneffectifthe95%credibleintervalforthetreatmentinteractiondoesnotincludezero.TrialstatusTheKetodexstudywasregisteredonDecember11,2019,andstartedrecruitmentinMarch2020.RecruitmentisexpectedtobecompletedbyDecember2022.Forthefinalanalysis,thedatabasewillbecleanedandcheckedforcompletenessbeforebeinglockedandthestatisticalanalysiswillthenbeundertakenusingthemethodsinthisSAP. Availabilityofdataandmaterials Nodatasetswereusedtodevelopthisarticleasanalysiswasnotundertaken.Thus,thisconsiderationisnotapplicable. AbbreviationsAE: Adverseevents ALC: Averagelengthcriterion DSMB: Dataandsafetymonitoringboard ED: Emergencydepartment IN: Intranasal IV: Intravenously SAP: Statisticalanalysisplan PSA: Proceduralsedationandanalgesia PSSS: PediatricSedationStateScale REB: ResearchEthicsBoard WCHRI: WomenandChildren’sHealthResearchInstitute References1.ChamberlainJ,PatelK,PollackM,BrayerA,MaciasC,OkadaP,SchunkJ,andothers.Recalibrationofpediatricriskofadmissionscoreusingamulti-institutionalsamples.AnnEmergMed.2004;43(4):461–8.PubMed  Article  GoogleScholar  2.RennieL,Court-BrownC,MokJ,BeattieT.Theepidemiologyoffracturesinchildren.Injury.2007;38(8):913–22.PubMed  Article  GoogleScholar  3.ChengJ,ShenW.Limbfracturepatternindifferentpediatricagegroups:astudyof3,350children.JOrthopTrauma.1993;7(1):15–22.CAS  PubMed  Article  GoogleScholar  4.JonesK,WeinerD.Themanagementofforearmfracturesinchildren:aPleaforconservatism.JPediatrOrthop.1999;19(6):811.CAS  PubMed  GoogleScholar  5.SchofieldS,SchutzJ,BablF.Proceduralsedationandanalgesiaforreductionofdistalforearmfracturesinthepaediatricemergencydepartment:aclinicalsurvey.EmergMedAust.2013;25(3):241–7.Article  GoogleScholar  6.CummingsE,ReidG,FinleyG,McGrathP,RitchieJ.Prevalenceandsourceofpaininpediatricinpatients.Pain.1996;68(1):25–31.CAS  PubMed  Article  GoogleScholar  7.GraudinsA,MeekR,Egerton-WarburtonD,OakleyE,SeithR.ThePICHFORK(PaininChildrenFentanylorKetamine)trial:arandomizedcontrolledtrialcomparingintranasalketamineandfentanylforthereliefofmoderatetoseverepaininchildrenwithlimbinjuries.AnnEmergMed.2015;65(3):248–54.PubMed  Article  GoogleScholar  8.PoonaiN,SpohnJ,VandermeerB,AliS,BhattM,HendrikxS,etal.Intranasaldexmedetomidineforanxiety-provokingproceduresinchildren:asystematicreviewandmeta-analysis.Pediatrics.2020;145(1):e20191623.PubMed  Article  GoogleScholar  9.HeathA,YaskinaM,PechlivanoglouP,RiosJ,OffringaM,KlassenT,PoonaiN,PullenayegumE.ABayesianresponse-adaptivedosefindingandcomparativeeffectivenesstrial.ClinicalTrials.2020.https://doi.org/10.1177/1740774520965173,p.epub.10.PoonaiN,CoriolanoK,KlassenT,HeathA,YaskinaM,BeerD,SawyerS,BhattM,KamA,DoanQ,SabhaneyV,OffringaM,PechlivanoglouP,HickesS,AliS.Studyprotocolforintranasaldexmedetomidineplusketamineforproceduralsedationinchildren:anadaptiverandomizedcontrollednon-inferioritymulticentertrial(theKetodextrial).BMJOpen.2020;10:e041319.11.GambleC,KrishanA,StockenD,LewisS,JuszczakE,DoréC,WilliamsonP,AltmanD,MontgomeryA,LimP,BerlinJ.Guidelinesforthecontentofstatisticalanalysisplansinclinicaltrials.Jama.2017;318(23):2337–43.PubMed  Article  GoogleScholar  12.KellyL,RicherL,AliS,PlintA,PoonaiN,FreedmanS,KnisleyL,ShimminC,HickesS,W’tJongG,PechlivanoglouP.Innovativeapproachestoinvestigator-initiated,multicentrepaediatricclinicaltrialsinCanada.BMJOpen.2019;9(6):e029024.PubMed  PubMedCentral  Article  GoogleScholar  13.HarrisP,TaylorR,ThielkeR,PayneJ,GonzalezN,CondeJ.Researchelectronicdatacapture(REDCap)-ametadata-drivenmethodologyandworkflowprocessforprovidingtranslationalresearchinformaticssupport.JBiomedInform.2009;42(2):377–81.PubMed  PubMedCentral  Article  GoogleScholar  14.WCHRI,“WomenandChildren'sHealthResearchInstitute,(WCHRI).DataCoordinatingCentre.,”https://www.wchri.org/data-coordinating-centre.AccessedAugust7,2018.2018.15.CraveroJ,AskinsN,SriswasdiP,TszeD,ZurakowskiD,SinnottS.ValidationofthePediatricSedationStateScale.Pediatrics.2017;139(5):e20162897.PubMed  PubMedCentral  Article  GoogleScholar  16.HicksC,vonBaeyerC,SpaffordP,vanKorlaarI,GoodenoughB.TheFacesPainScale–Revised:towardacommonmetricinpediatricpainmeasurement.Pain.2001;93(2):173–83.CAS  PubMed  Article  GoogleScholar  17.CaoJ,LeeJ,AlberS.ComparisonofBayesiansamplesizecriteria:ACC,ALC,andWOC.JStatPlannInference.2009;139:4111–22.Article  GoogleScholar  18.BergerJ,WolpertR.IMS.Thelikelihoodprinciple.Hayward,California:InstituteofMathematicStatistics;1988. GoogleScholar  19.RCoreTeam.R:alanguageandenvironmentforstatisticalcomputing.Vienna,Austria:RFoundationforStatisticalComputing;2018. GoogleScholar  20.PlummerM.JAGS:aprogramforanalysisofBayesiangraphicalmodelsusingGibbssampling.In:Proceedingsofthe3rdinternationalworkshopondistributedstatisticalcomputing,vol.124,no.125.10;2003.p.1–10. GoogleScholar  21.BhatR,SanthoshM,AnnigeriV,RaoR.Comparisonofintranasaldexmedetomidineanddexmedetomidine-ketamineforpremedicationinpediatricspatients:arandomizeddouble-blindstudy.AnesthEssaysRes.2016;10(2):349.PubMed  PubMedCentral  Article  GoogleScholar  22.MoritaS,ThallP,MüllerP.Determiningtheeffectivesamplesizeofaparametricprior.Biometrics.2008;64(2):595–602.PubMed  Article  GoogleScholar  23.CaiC,YuanY,JiY.ABayesiandosefindingdesignforoncologyclinicaltrialsofcombinationalbiologicalagents.JRStatSoc:SerC:ApplStat.2014;63(1):159–73.Article  GoogleScholar  24.WangK,IvanovaA.Two-dimensionaldosefindingindiscretedosespace.Biometrics.2005;61(1):217–22.PubMed  Article  GoogleScholar  25.GelmanA,JakulinA,PittauM,SuY.Aweaklyinformativedefaultpriordistributionforlogisticandotherregressionmodels.AnnApplStat.2008;2(4):1360–83.Article  GoogleScholar  26.KannikeswaranN,Lieh-LaiM,MalianM,WangB,FarooqiA,RobackM.OptimaldosingofintravenousketamineforproceduralsedationinchildrenintheED—arandomizedcontrolledtrial.AmJEmergMed.2016;34(8):1347–53.PubMed  Article  GoogleScholar  27.GelmanA,ShaliziC.PhilosophyandthepracticeofBayesianstatistics.BrJMathStatPsychol.2013;66(1):8–38.PubMed  Article  GoogleScholar  28.GelmanA.Priordistributionsforvarianceparametersinhierarchicalmodels(commentonarticlebyBrowneandDraper).BayesianAnal.2006;1(3):515–34.Article  GoogleScholar  DownloadreferencesAcknowledgementsTheauthorswouldliketoacknowledgetheiPCTSPORadministrativestaffandourpatientpartners,whoprovidedvaluablesupportandinputonthestudydesignanddocuments.ThefollowingarethemembersoftheKidsCANPERCInnovativePediatricClinicalTrialsKetodexStudyTeam:DarcyBeer,ScottSawyer,MaalaBhatt,AprilKam,QuynhDoan,VikramSabhaney,SerenaHickes,SaminaAli,KarlyStillwell,TannisErickson,ChelseaBowkett,CarolynShimmin,BrendonFoot,ChelseaBowkett,CandaceMcGahern,RedjanaCarciurmaruj,andJeannineSchellenberg.FundingThisworkissupportedbyanInnovativeClinicalTrialsMulti-yearGrantfromtheCanadianInstitutesofHealthResearch(fundingreferencenumberMYG-151207;2017–2020),aspartoftheStrategyforPatient-OrientedResearchandtheChildren’sHospitalResearchInstituteofManitoba(Winnipeg,Manitoba),theCentreHospitalierUniversitaireSainte-Justine(Montreal,Quebec),theDepartmentofPediatrics,UniversityofWesternOntario(London,Ontario),theAlbertaChildren’sHospitalResearchInstitute(Calgary,Alberta),theWomenandChildren’sHealthResearchInstitute(Edmonton,Alberta),theChildren’sHospitalofEasternOntarioResearchInstituteInc.(Ottawa,Ontario),andtheHospitalforSickChildrenResearchInstitute(Toronto,Ontario).ThisstudyissponsoredbyTheGovernorsoftheUniversityofAlberta(Suite400,8215–112Street,Edmonton,Alberta,CanadaT6G2C8).Neitherthestudysponsornorfundershaveanyroleinthecollection,management,analysis,orinterpretationofdata;writingofthereport;orthedecisiontosubmitthereportforpublication.AdditionalsupportwasreceivedfromthePhysiciansServicesIncorporatedFoundation,AcademicMedicalOrganizationofSouthwesternOntario,OntarioMinistryofEconomicDevelopment,JobCreationandTrade,andtheChildren’sHealthFoundationoftheChildren’sHospital,LondonHealthSciencesFoundation.AuthorinformationAffiliationsChildHealthEvaluativeSciences,TheHospitalforSickChildren,Toronto,CanadaAnnaHeath, JuanDavidRios, EleanorPullenayegum, PetrosPechlivanoglou & MartinOffringaDallaLanaSchoolofPublicHealth,DivisionofBiostatistics,UniversityofToronto,Toronto,CanadaAnnaHeathDepartmentofStatisticalScience,UniversityCollegeLondon,London,UKAnnaHeathInstituteofHealthPolicy,ManagementandEvaluation,UniversityofToronto,Toronto,Ontario,CanadaEleanorPullenayegum, PetrosPechlivanoglou & MartinOffringaDivisionofNeonatology,TheHospitalforSickChildren,UniversityofToronto,Toronto,Ontario,CanadaMartinOffringaWomen&Children’sHealthResearchInstitute,UniversityofAlberta,Edmonton,Alberta,CanadaMarynaYaskina, RickWatts & ShanaRimmerUniversityofManitoba,Winnipeg,Manitoba,CanadaTerryP.KlassenChildren’sHospitalResearchInstituteofManitoba,Winnipeg,Manitoba,CanadaTerryP.KlassenLondonHealthSciencesCentre,Children’sHospital,London,Ontario,CanadaKamaryCoriolanoDepartmentsofPaediatricsandEpidemiology&Biostatistics,SchulichSchoolofMedicineandDentistry,London,CanadaNaveenPoonaiChildren’sHealthResearchInstitute,LondonHealthSciencesCentre,London,CanadaNaveenPoonaiAuthorsAnnaHeathViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarJuanDavidRiosViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarEleanorPullenayegumViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarPetrosPechlivanoglouViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarMartinOffringaViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarMarynaYaskinaViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarRickWattsViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarShanaRimmerViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarTerryP.KlassenViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarKamaryCoriolanoViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarNaveenPoonaiViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarConsortiaonbehalfofthePERC-KIDSCANKetodexStudyGroupDarcyBeer, ScottSawyer, MaalaBhatt, AprilKam, QuynhDoan, VikramSabhaney, SerenaHickes, SaminaAli, KarlyStillwell, TannisErickson, ChelseaBowkett, CarolynShimmin, BrendonFoot, ChelseaBowkett, CandaceMcGahern, RedjanaCarciurmaruj & JeannineSchellenbergContributionsAllauthorswereinvolvedintheconceptionanddesignoftheSAP.AHdraftedthemanuscript.DR,EP,PP,MO,MY,RW,SR,TPK,KC,andNPofferedsubstantiverevisions.Allauthorsread,edited,andapprovedthefinalmanuscript.Allindividualsmentionedinthe“Acknowledgements”sectionaremembersoftheKetodexStudyGroup.AHistheseniorresponsiblestatisticianandNPistheclinicallead.CorrespondingauthorCorrespondenceto AnnaHeath.Ethicsdeclarations Ethicsapprovalandconsenttoparticipate EthicsapprovalwasobtainedfromClinicalTrialsOntariofortheleadsite(Children’sHospital,LondonHealthSciencesCentre)andMcMasterChildren’sHospital.Theotherparticipatingsitesreceivedinstitutionalethicsapproval.AllprotocolamendmentswillbesubmittedforapprovaltoHealthCanadabeforebeingcommunicatedtoeachsiteandimplementedonlyafterHealthCanadaandREBapproval.Wewillobtainchildassentandcaregiverconsentfromalltrialparticipantsasappropriate. Consentforpublication Notapplicable. Competinginterests Nonereported. AdditionalinformationPublisher’sNoteSpringerNatureremainsneutralwithregardtojurisdictionalclaimsinpublishedmapsandinstitutionalaffiliations.SupplementaryInformation Additionalfile1. Rightsandpermissions OpenAccessThisarticleislicensedunderaCreativeCommonsAttribution4.0InternationalLicense,whichpermitsuse,sharing,adaptation,distributionandreproductioninanymediumorformat,aslongasyougiveappropriatecredittotheoriginalauthor(s)andthesource,providealinktotheCreativeCommonslicence,andindicateifchangesweremade.Theimagesorotherthirdpartymaterialinthisarticleareincludedinthearticle'sCreativeCommonslicence,unlessindicatedotherwiseinacreditlinetothematerial.Ifmaterialisnotincludedinthearticle'sCreativeCommonslicenceandyourintendeduseisnotpermittedbystatutoryregulationorexceedsthepermitteduse,youwillneedtoobtainpermissiondirectlyfromthecopyrightholder.Toviewacopyofthislicence,visithttp://creativecommons.org/licenses/by/4.0/.TheCreativeCommonsPublicDomainDedicationwaiver(http://creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthisarticle,unlessotherwisestatedinacreditlinetothedata. ReprintsandPermissionsAboutthisarticleCitethisarticleHeath,A.,Rios,J.D.,Pullenayegum,E.etal.Theintranasaldexmedetomidineplusketamineforproceduralsedationinchildren,adaptiverandomizedcontrollednon-inferioritymulticentertrial(Ketodex):astatisticalanalysisplan. Trials22,15(2021).https://doi.org/10.1186/s13063-020-04946-3DownloadcitationReceived:20July2020Accepted:01December2020Published:06January2021DOI:https://doi.org/10.1186/s13063-020-04946-3SharethisarticleAnyoneyousharethefollowinglinkwithwillbeabletoreadthiscontent:GetshareablelinkSorry,ashareablelinkisnotcurrentlyavailableforthisarticle.Copytoclipboard ProvidedbytheSpringerNatureSharedItcontent-sharinginitiative KeywordsProceduralsedationandanalgesiaPediatricclosedreductionIntranasalketodexNon-inferioritytrialBayesianadaptivedesignStatisticalanalysisplan DownloadPDF Advertisement Trials ISSN:1745-6215 Contactus Submissionenquiries:AccesshereandclickContactUs Generalenquiries:[email protected]