Historically this process has relied on the evidence base societal values cultural tensions and political sway but not necessarily in that order We propose a conceptual framework to guide and prioritize this process shifting emphasis toward the prin ID: 71185
Download Pdf The PPT/PDF document "EDITORIAL Open Access Evidencebased deim..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
EDITORIALOpenAccess Evidence-basedde-implementationfor contradicted,unproven,andaspiringhealthcare practices VinayPrasadandJohnPAIoannidis * Abstract Abandoningineffectivemedicalpracticesandmitigatingtherisksofuntestedpracticesareimportantforimproving patienthealthandcontaininghealthcarecosts.Historically,thisprocesshasreliedontheevidencebase, societalvalues,culturaltensions,andpoliticalsway,butnotnecessarilyinthatorder.Weproposeaconceptual frameworktoguideandprioritizethisprocess,shiftingemphasistowardtheprinciplesofevidence-basedmedicine, acknowledgingthatevidencemaystillbemisinterpretedordistortedbyrecalcitrantproponentsofentrenched practicesandotherbiases. Keywords: Evidence-basedmedicine,Reversals,Divestment,De-implementation,Contradiction,Bias Background Divestingfromineffectiveandharmfulmedicalpractices mitigatetheunsustainableriseinhealthcarecosts.Aban- donment(de-implementation)ofmedicalinterventionsmay dependonmultiplefactors.Empiricalevidencefromwell- designedstudiesshouldcount,butotherconsiderations suchasinertia,financialandprofessionalconflicts,cultural andsocietalvalues,knowledgebrokering,andlobbying mayalsobeveryimportanteven tually.Thequestionishow wecanpositionevidencesoastobemoreinformative andinfluentialinthesecomplexprocesses.Herewe provideaframeworktoguidetheevidence-basedde- implementationofinterventions,acknowledginghowon- the-groundrealitiescanentert heseconsiderations.Broadly, wewillconsiderthreecategoriesofhealthcarepractices: thosethatareknownnottowork;thoseforwhichtheevi- dencebaseisuncertain;andthosethatareindevelopment andwherestrategicpreemptiveplacementofevidencemay helptheireventualde-implementation,ifneeded.While medicalpractices,theprinciplesarebroadlyapplicableto allhealthcarefields. Contradictedestablishedmedicalpractices Thenumberofmedicalpracticeswherethebestevidence showsnoefficacyorharmsoutweighingbenefitsis substantial.Onesearchproducedover150potentially ineffectiveorunsafepractices[1],andempiricalreviewsof highimpactmedicaljournalshavegeneratedover140 reversedmedicalpractices[2]. Whenlarge,well-donerandomizedtrialshavecontra- dictedcurrentmedicalpractice,de-implementationmakes sense,butitcanmeetwithfi ercetacticalresistance. Proponentsofcontradictedmedicalpracticescanprocure notonlyeditorials,butalsocounter-evidencethatcuts corners, e.g .focusingonlesserendpoints,highlighting subgroupanalyses,orperformingadditionalstudieswith tailoredeligibilitycriteriaandoutcomeselectiontoshow somebenefit[3],andconflictedexpertguidelinescan followsuit[4]. Takeforexamplethe2007COURAGEstudy,which foundthatamongpatientswithcoronaryarterydiseaseand stableangina,routinepercutaneouscoronaryintervention (PCI)wasnobetterthananinitialstrategyofoptimal medicaltreatment(OMT).Inthemonthfollowing thestudy spublication,PCIandstentingwasdown13% *Correspondence: jioannid@stanford.edu NationalCancerInstitute,Bethesda,MD(VP)andStanfordPrevention ResearchCenter,DepartmentsofMedicineandHealthResearchandPolicy, StanfordUniversitySchoolofMedicine,andMeta-ResearchInnovationCenter atStanford(METRICS),MedicalSchoolOfficeBuilding,RoomX306,Stanford, CA94305,USA Implementation ©2014IoannidisandPrasad;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsofthe CreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse, distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.TheCreativeCommonsPublic DomainDedicationwaiver(http://creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthis article,unlessotherwisestated. PrasadandIoannidis ImplementationScience 2014, 9 :1 http://www.implementationscience.com/content/9/1/1 nationally;however,by2010,thosenumbersreturnedtoprepublicationlevels[5].NotonlyhasCOURAGEandcorroboratingmeta-analyses[6,7]failedtostemtheuseofPCI,butitappearsthattheyhavealsonotimprovedadherencetoOMTattimeofPCI[8].ProponentsofstentingcriticizedCOURAGE,citingselectionbias,cross-over,andpoorstudypower[9].Then,in2012,FAME2showedthatPCIguidedbyfractionalflowreservetestingcoulddecreaseratesofrevascularizationcomparedtoOMT,thoughtherewerenodifferencesincardiacdeath,myocardialinfarction,orstroke[10].YettheresultsofFAME2andCOURAGEarecomparable.Regardingrevascularization,10stentswereplacedtoavert1futurerevascularizationprocedureinFAME2,and12stentsachievedthatgoalinCOURAGE.WhetherPCIwasguidedbyangiographyorfractionalflow,thenetresultissimilar.Moreover,multiplemeta-analysesofPCIco-authoredbyinterventionalcardiologistsand/orsponsoredbytheindustryclaimedbenefitsinPCIbypoolingtogethertrialsofstableanginaandtrialswithresidualischemia,transferringthebenefitsofthelatterpopulationtotheformer.Practiceguidelinesthencouldfollowthesamepath.Anotherexampleistheroutineuseofgownandgloveprecautionsamongpatientscolonizedwithresistantpathogens,whichissupportedonlybyquasi-experimental,before-and-after,studies[11].Yet,todate,twocluster-randomizedtrialshavefailedtosupportthebenefitsofthispractice[12,13].Onestudyshowednoreductioninthetransmissionofmethicillin-resistantStaphylococcusaureusorvancomycin-resistantenterococcus[13],andtheothershowednodifferenceinratesofcolonizationorinfectionwiththesetwopathogens[12].Thesestudieshavefailedtochangethispractice,however,andeditorialambivalencecontinues[14].Somehaveevenclaimedthelikelihoodofharm(morethanclearevidenceofbenefit)shoulddrivethedecisiontoimplementti.e.consideringde-implementationonlyaftercontactprecautionsareproventobeharmful,notmerelyineffective.Suchresistancetoadheretothebestavailableevidenceinflateshealthcarecosts,andmaydistractfromalternativestrategieswithpromisingearlyresultsofefficacy,suchasuniversaldecolonizationprotocols[16].Inbrief,evidencewarscanhinderde-implementa-tion,andpracticeresuscitationmaybesuccessfulatreclaiminglostmarketshares.Ofcourse,itisentirelypossiblethatsomesubgrouptrulybenefitswhenapracticehasnegativeresultsglobally,orthatsomelesserendpointsaremeaningfultopatients.Yet,moreoftensuchcounter-evidenceresurrectionstudiessim-plycreateexcusestonotabandonthecontradictedpractice.Weproposeasimplestandardtocurbsuchpracticeresuscitation:Theevidencetoreviveacontradictedmedicalpractice(whetherinpartorinwhole)shouldinvolveendpointsandcontrolsatleastasrigorousasthecontradictorystudy.Untilsuchevidenceisobtained,payersmayofferdisincentivesbyplacingrestrictionstoreimbursement,andregulatorsmayconsiderrevokingorrestrictingpriorapprovals.UnprovenmedicalpracticesClearlycontradictedpracticesarelesscommonthanunprovenones.Among1,344articlesassessingamedicalpractice,363(27%)testedstandardofcare,with146(11%ofthetotal)contradictingit[2].Manymedicalpracticesarelargelyuntestedorhaveinsufficientevidence.AnempiricalevaluationoftheCochraneDatabaseofSystematicReviewsfoundthattheexistingevidencebasewasunabletosupportorrefute49%ofinterven-tions[17],and48%ofAmericanCollegeofCardiologyrecommendationsweresupportedbyexpertopiniononly[18].Arationalstrategytode-implementmedicalpracticessupportedbylittletonoevidenceistosubjectthemtotestinginsystematicfashion.Ideally,thisassessmentwouldbeperformedundertheauspicesofnon-conflictedbodies,possiblywithinexistinggovernmentalstructures,suchastheAgencyforHealthcareResearchandQuality.Table1highlightspotentialconsiderationstoprioritizeuntestedmedicalpractices.Likelymajorconsiderationsincludetheextantevidencebaseofapracticepreferenceshouldbegiventothosebasedontheleastevidenceandthecostandubiquityofthepracticegiventothosepracticesplacingthelargestburdenonthehealthcaresystem.Additionalconsiderationsincludethepresenceofalternativechoicesinafieldgiventowardreappraisingfieldswithmanyalternativesofvaryingclass,price,andevidencebase;practiceswithclearlydocumentedharms;practiceswherethecosttoobtainthenecessaryevidenceiscontained;andpracticeswheretheresultsoftrialswithunfavorableresultsmayrealisticallychangemindsandpractice.Formalapproaches,suchasvalueofinformationcalculations[19]maybeapplied,investigatingthevalueofspecificproposedrandomizedtrialstowardsde-implementingestablishedunprovenpractices.NovelmedicalpracticesWithmultiplenovelinterventions(therapeutic,diagnostic,prognostic,healthcaresystem,andother)beingintroducedinmedicalcare,akeyconsiderationistotakepreemptivestepsthatwouldallowefficientde-implementationiftheinterventioneventuallyprovesinefficientandharmful.Whilethereisincreasingpressuretoadoptnovelinter-ventionsbeforesubstantialevidencehasbeenobtainedPrasadandIoannidisImplementationSciencePage2of5http://www.implementationscience.com/content/9/1/1 Table1PotentialconsiderationsinprioritizingthetestingofunprovenmedicalpracticesFactortoconsiderGeneralprincipleHowtoimplementthisfactorPriorevidencePriorityshouldbegiventopracticeswherethepresentevidencebaseisweakest.Forinstance,atieredsystemmaybeutilized:Level1(Weak)Randomizedtrialsofinterventionsclaimingsubjectivebenefits,thatareunblindedorfailtousepropercontrols.2(Weaker)Historicallycontrolledstudiesofinterventionsthatpurportsurvivalbenefits,caseseriesdocumentingimprovementsinsubjectiveendpointsandquasi-experimentalstudies.3(Weakest)Practicesbasedonpathophysiologyandexpertopinionalone.Inmanycases,professionalconflictsmayalsoproveproblematic;thus,itmaybereasonabletopursuethistechniqueusingcontent-specificexpertsinstrictlyanadvisorycapacityCost/ubiquityPriorityshouldbegiventointerventionswithsignificantnetfinancialburdenonhealthpayers.Forinstance,orthopedicproceduresforchronicbackandjointpain,includingkneeandhipreplacementsurgeriesarewidelyutilizedintheUnitedStates,incurlargefinancialburdenonpayers,buthavelittleevidenceofsustainedlongtermbenefits.AlternativeoptionsPriorityshouldbegiventopracticesforwhichthereareseveralalternativeoptions,particularlyifalternativesareofcompletelydifferentmechanisms(thusunlikelytoalsobeoverturned),oroflowcostorbolsteredbystrongerevidence.Forinstance,considerthemarketforanti-rheumatologicagents.Maintenancetreatmentofrheumatoidarthritis(RA)withdiseasemodifyingagents(DMARDS)hashistoricallyrelieduponoralanti-immunologicagentssuchasmethotrexate,azathiaprine,cyclosporin,andhydroxychloroquine.Recentyearshavewitnessedaboominnoveldrugs,typicallyexpensivemonoclonalantibodiesagainstcirculatingcytokinesorcellsurfacereceptors.Todate,thismarkethasbeenlimitedbypaucityofheadtoheadtrials,and,oftrialsthathavebeenconducted,themajorityareindustry-sponsoredstudies.Collectively,thereremainsclinicaluncertaintyabouthowbesttousetheseagents[Priorityshouldbegiventotestpracticeswheretheharmsarewelldocumentedandconfersubstantialmorbidity.Forinstance,thereisgrowingawarenessofstrutfracturization,embolism,andmigrationofIVCfilters.Atthesametime,theIVCfilterhasnevershowntoimproveanypatient-centeredoutcomeforanypatientpopulationinaprospectivetrial,andtracesitsapprovalthroughtheFDAs510kmechanism[TestingtheinterventionmakesfinancialsensePriorityshouldbegiventotestpracticeswherethecosttotestisfarlessthanongoingexpendituresofthepractice.Insomerespects,trialistsshouldthinklikeCEOs,weighingthecostsofconductingastudy,whichmayfindapracticeineffectiveversustheongoingexpendituresforthatpractice.Attimes,suchcalculationsmayfavorcostlytrialswheretheexistingevidencebaseisweak,observationalstudiessuggestinefficiencies,andtheongoingcostsarelarge[].Atothertimes,smalltrialsthateliminateboutiquepracticesmaybeemployed[].Whosefinancialbottomlineisbeingaffectedisimportanttoconsider.Forthatreason,nonconflictedbodiesshouldmakethesedeterminations,utilizinginvestigatorswithoutfinancialconflictsofinterest.Proponentsareopen-mindedPriorityshouldbegiventotestpracticeswherenegativeresultsmaytrulygaintraction.Somespecialties(primarycareproviders)maybemorereadytoabandoncontradictedmedicalpractices,anditisreasonabletotestpracticeswhenthereisgenuinebeliefthatcontradictioncangaintraction.Furthermore,somepracticesmaybecumbersome(tightglycemiccontrolintheICU),time-consuming(routinegownandgloveprecautions)orunpleasant,andtheircontradictionmayalsobepalatable.Finally,aspaymentstructuresshiftfromfeeforservicetowardsbundles[],costlycomponentsmaylosefaithfuldisciples.Otherfields,thosewithnumerousandhyperbolicthirdpartyadvocates,havebeennotoriouslyunwillingtotrustresultsthatunderminetheirworldview,nomatterhowrobustthescience.ValueofinformationgainedPriorityshouldbebasedontheexpectedvalueoffundingaspecificstudythatmayinformde-implementation,atthesizeandcostproposed.Valueofinformation(VOI)offersadecision-makingframeworkthattriestocaptureseveraloftheaboveissues,atleasttheonesthatcanbebestquantified[].VOIcanbeusedtoprioritizeandpowerclinicaltrialstakingintoaccountthecostsofincreasingstudysamplesize,thepotentialnumberofpersonsaffectedbychangesinthatpractice,thecostsofthepractice,includingdownstreamcosts,andtheincreasedknowledgeofmarginalchangesinhealthoutcomesthatmayresultfromtestingconvertingalltothefinalcommondenominatorofcostperfavorableoutcomegained.PrasadandIoannidisImplementationSciencePage3of5http://www.implementationscience.com/content/9/1/1 onthem,onemethodtocurbthespreadofineffectivepracticesistorestricttheirusepriortowidespreaddissemination,asdemonstratedbythecaseofpercu-taneoustransluminalangioplastyandstenting(PTAS)forintracranialstenosis.In2005,theWingspanintracranialarterystentwasgrantedhumanitariandeviceexemptionfromtheUSFoodandDrugAdministration,baseduponprovisionaldatathatitcouldimproveintracranialarterylumendiameterinpatientswithstenosisrefractorytomedicalther-apy[25].However,thesingle,uncontrolledstudythatledtoapprovalwasunabletoinformanypatient-centeredendpoint.In2006,CentersforMedicare&MedicaidServices(CMS)announcedthattheywouldpayfortheprocedureonlywithintheconfinesofarandomizedtrial.Theyadheredtosuchaposition,despitepressurefromthemanufacturerin2008[26].In2011,theonlyrandomizedstudyofthedevice,theSAMMPRIStrial,foundthatPTASamongpatientswitharecenttransientischemicattack(TIA)orstrokeanddocumentedstenosisofamajorintracranialarterynearlytripledthe30-dayriskofstrokeordeathcom-paredtooptimalmedicalmanagement(14.7%versus5.8%)[27].Duringtheyearsitwasapprovedbuttrialdatawaslacking,CMSspolicydramaticallylimitedoff-protocoluseofthedeviceandeffectivelyprotectedthepublic.Altogether,onlyafewhundredpatientsreceivedthedevice(200+treatedonprotocol)intheUS[26]contrastthisagainstthemillionsofpatientswhore-ceivedPCIforstableangina.CMSswise2006deci-sionlikelyavertedacatastrophicoutcomeforthousandsofpatientswhomightotherwisehavebeentreatedwiththedevice.ThelessonofPTASisthathigherupfrontstan-dardshavepotentialtoprotectpatientsfromultim-atelyflawedcare.Unfortunately,regulatoryagenciesappeartomoveincreasinglyintheoppositedirec-tion,notablywiththecreationoftheFDAbreak-designation[28],andemergingguidancetoindustryforexpandedoptionsofacceleratedapproval[29,30]aregulatorymechanismwheredevelopershavehistoricallyshirkedpost-marketingcommitmentsofconductingtrialsexaminingclinicallymeaningfulend-points[31].EmpiricaltestingTheopinionsthatweexpressheremaywellbebiased.Webelieve,however,thatthereisnoreasonthatexperi-mentalstudiescannotbeleveragedtoprovideclarityforhealthpolicieswithbroadsocietalrepercussions.Todate,regulatorypolicieshavebeenbasedontheoryorscantretrospectiveobservationalstudies,butatleastsomepoliciesmaybetestedcreativelywithrandom-izedcontrolledtrials[14].Forinstance,novelagentsmayberandomlyassignedtoacceleratedortraditionalapproval.Thismighthelpinformwhetherprovisionalapproval,widedissemination,andsubsequentconfirma-torytrialsbenefitorharmsocietymorethanrestrict-iveapprovalstrategiessatingrobustendpointspriortodissemination.De-implementingpracticesreflectsarecommitmenttoevidence-basedhealthcare.Thisisimportantformedications,devices,procedures,behavioralorpsy-chologicalinterventions,screeninganddiagnostictests,andanyotherinterventionundertakenbypeopleinthehealthprofessions.Strategiestoelimin-ateineffectiveandharmfulpracticesmayhelpcontainhealthcarespendingandoptimizeoutcomes.Ideally,themajorityofmedicaldecisionsshouldbesupportedbyrobustdata,withambiguousdecisionsmadeonlywithintheconfinesofongoingstudies.However,aswestated,rational,quantitativeevidencemaynotnecessarilybetheonlyorevenmainfactordrivinghealthcaredecisions.Researchtounderstandbettertheother,cognitiveorpoliticalfactorsthatfacilitateorhinderde-implementationisthusalsowarranted(seeBox1). Box1:NotefromtheeditorsTheEditors-in-ChiefofImplementationScienceinvitedthiseditorialfollowingaconsultationwithoureditorialteamandEditorialBoard.Theyidentifiedasanimportanttheme,whichdeservesmoreattentionthanitcurrentlyreceives.Weregardde-implementationbroadlyasstoppingpracticesthatarenotevidence-based.Weencouragefurtherpapersonthisthemeandwillincludetheseinaspecialarticleseriesinthejournaltoenhancetheirvisibility.Allsubmissionswillbereviewedandhandledaccordingtoournormalprocedures.Inaddition,wewelcomeandencouragecommentsinresponsetotheaccompanyingeditorial,usingthecommentfeatureofthejournalsplatform.Thesecomments,whilemoderated,areintendedtostimulatediscussionanddebatewithintheimplementationresearchcommunity.Inbothpapersandcomments,wewelcomearangeofperspectivesandrigorousstudiesonthethemeofde-implementation,including(butnotlimitedto)contributionsthatcoverpsychological,organizationaloreconomicfactors.Weintendtopromoteotherthemesinthefuture.PrasadandIoannidisImplementationSciencePage4of5http://www.implementationscience.com/content/9/1/1 Competinginterests Theauthorsdeclarethattheyhavenocompetinginterests. Authors contributions VPandJPAIdraftedthemanuscript.VPandJPAIeditedthemanuscriptfor intellectualcontent.JPAIprovidedsupervision.Bothauthorsreadand approvedthefinalmanuscript. Received:20November2013Accepted:11December2013 Published:8January2014 References 1.ElshaugAG,WattAM,MundyL,WillisCD: Over150potentiallylow-value healthcarepractices:anAustralianstudy. MedJAust 2012, 197: 556 560. 2.PrasadV,VandrossA,ToomeyC, etal : Adecadeofreversal:ananalysisof 146contradictedmedicalpractices. MayoClinProc 2013, 88: 790 798. 3.StamatakisE,WeilerR,IoannidisJP: Undueindustryinfluencesthatdistort healthcareresearch,strategy,expenditureandpractice:areview. EurJ ClinInvest 2013, 43: 469 475. 4.LenzerJ,HoffmanJR,FurbergCD,IoannidisJP: Ensuringtheintegrityof clinicalpracticeguidelines:atoolforprotectingpatients. BMJ 2013, 347: f5535. 5.WinsteinKJ: Asimplehealth-carefixfizzlesout. TheWallStreetJournal 2010.DateAccessed:December9,2013athttp://online.wsj.com/news/ articles/SB10001424052748703652104574652401818092212. 6.StergiopoulosK,BrownDL: Initialcoronarystentimplantationwith medicaltherapyvsmedicaltherapyaloneforstablecoronaryartery disease:meta-analysisofrandomizedcontrolledtrials. ArchInternMed 2012, 172: 312 319. 7.KatritsisDG,IoannidisJPA: Percutaneouscoronaryinterventionversus conservativetherapyinnonacutecoronaryarterydisease: ameta-analysis. Circulation 2005, 111: 2906 2912. 8.BordenWB,RedbergRF,MushlinAI,DaiD,KaltenbachLA,SpertusJA: PAtternsandintensityofmedicaltherapyinpatientsundergoing percutaneouscoronaryintervention. JAMA 2011, 305: 1882 1889. 9.SiontisGC,TatsioniA,KatritsisDG,IoannidisJP: Persistentreservations againstcontradictedpercutaneouscoronaryinterventionindications: citationcontentanalysis. AmHeartJ 2009, 157: 695 701. 10.DeBruyneB,PijlsNHJ,KalesanB, etal : Fractionalflowreserve guidedPCI versusmedicaltherapyinstablecoronarydisease. NEnglJMed 2012, 367: 991 1001. 11.JainR,KralovicSM,EvansME, etal : Veteransaffairsinitiativetoprevent methicillin-resistantstaphylococcusaureusinfections. NEnglJMed 2011, 364: 1419 1430. 12.HuskinsWC,HuckabeeCM,O GradyNP, etal : Interventiontoreduce transmissionofresistantbacteriainintensivecare. NEnglJMed 2011, 364: 1407 1418. 13.HarrisAD,PinelesL,BeltonB, etal : Universalgloveandgownuseand acquisitionofantibiotic-resistantbacteriaintheicu:arandomizedtrial. JAMA 2013, 310: 1571 1580. 14.IoannidisJA,PrasadV: EValuatinghealthsystemprocesseswith randomizedcontrolledtrials. JAMAInternMed 2013, 173: 1279 1280. 15.MalaniPN: Preventinginfectionsintheicu:onesizedoesnotfitall. JAMA 2013, 310: 1567 1568. 16.HuangSS,SeptimusE,KleinmanK, etal : Targetedversusuniversal decolonizationtopreventICUinfection. NEnglJMed 2013, 368: 2255 2265. 17.ElDibRP,AtallahAN,AndrioloRB: MappingtheCochraneevidencefor decisionmakinginhealthcare. JEvalClinPract 2007, 13: 689 692. 18.TricociP,AllenJM,KramerJM,CaliffRM,SmithSCJr: Scientificevidence underlyingtheACC/AHAclinicalpracticeguidelines. JAMA 2009, 301: 831 841. 19.ClaxtonK,PosnettJ: Aneconomicapproachtoclinicaltrialdesignand researchpriority-setting. HealthEcon 1996, 5: 513 524. 20.IoannidisJP,KarassaFB,DruytsE,ThorlundK,MillsEJ: Biologicagentsin rheumatology:unmetissuesafter200trialsand$200billionsales. NatRevRheumatol 2013, 9: 665 673. 21.PrasadV,RhoJ,CifuA: Theinferiorvenacavafilter:howcouldamedical devicebesowellacceptedwithoutanyevidenceofefficacy? JAMA InternMed 2013, 1: 3. 22.GotzschePC,NielsenM: Screeningforbreastcancerwithmammography. CochraneDatabaseSystRev 2011(1):CD001877. 23.TerreehorstI,HakE,OostingAJ, etal : Evaluationofimpermeablecovers forbeddinginpatientswithallergicrhinitis. NEnglJMed 2003, 349: 237 246. 24.MechanicR,TompkinsC: Lessonslearnedpreparingformedicare bundledpayments. NEnglJMed 2012, 367: 1873 1875. 25.AccessedatApprovalorderofWingspanstentsystem: FoodandDrug Administrationwebsite ;2012.Availableat:http://www.fda.gov/ohrms/ dockets/dockets/05m0308/05m-0308-aav0001-approval-order-vol1.pdf. AccessedOctober13,2012. 26.GuptaA,DesaiMM,KimN,BulsaraKR,WangY,KrumholzHM: Trendsin intracranialstentingamongmedicarebeneficiariesintheUnitedStates, 2006 2010. JAmHeartAssoc 2013, 2: e000084. 27.ChimowitzMI,LynnMJ,DerdeynCP, etal : Stentingversusaggressive medicaltherapyforintracranialarterialstenosis. NEnglJMed 2011, 365 (11) : 993 1003. 28.KramerDB,KesselheimAS: Userfeesandbeyond TheFDAsafetyand innovationactof2012. NEnglJMed 2012, 367: 1277 1279. 29.KozauerN,KatzR: Regulatoryinnovationanddrugdevelopmentfor early-stagealzheimer sdisease. NEnglJMed 2013, 368: 1169 1171. 30.ProwellTM,PazdurR: Pathologicalcompleteresponseandaccelerated drugapprovalinearlybreastcancer. NEnglJMed 2012, 366: 2438 2441. 31.DhruvaSsRRF: ACceleratedapprovalandpossiblewithdrawalof midodrine. JAMA 2010, 304: 2172 2173. doi:10.1186/1748-5908-9-1 Citethisarticleas: PrasadandIoannidis: Evidence-based de-implementationforcontradicted,unproven,andaspiring healthcarepractices. ImplementationScience 2014 9 :1. Submit your next manuscript to BioMed Central and take full advantage of: Convenient online submission Thorough peer review No space constraints or color gure charges Immediate publication on acceptance Inclusion in PubMed, CAS, Scopus and Google Scholar Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit PrasadandIoannidis ImplementationScience 2014, 9 :1 Page5of5 http://www.implementationscience.com/content/9/1/1