seniorityfilezillaftpclient

2014InformHealthSocCare,2014;39(2):67–80!
InformaUKLtd.
ISSN:1753-8157print/1753-8165onlineDOI:10.
3109/17538157.
2013.
812647ManagingmulticentreclinicaltrialswithopensourceyDimitriAristotleRaptis,1*TobiasMettler,2*MichaelAlexanderFischer,3MichaelPatak,3MickaelLesurtel,1DilmurodjonEshmuminov,1OlivierdeRougemont,1RolfGraf,1Pierre-AlainClavien,1andStefanBreitenstein1,41DivisionofVisceralandTransplantSurgery,SwissHPBCenter,UniversityHospitalZurich,Zurich,Switzerland,2InstituteofInformationManagement,UniversityofSt.
Gallen,St.
Gallen,Switzerland3InstituteofDiagnosticandInterventionalRadiology,UniversityHospitalZurich,Zurich,Switzerland,and4DepartmentofSurgery,WinterthurCantonHospital,Winterthur,SwitzerlandBackground:Multicentreclinicaltrialsarechallengedbyhighadministrativeburden,datamanagementpitfallsandcosts.
Thisleadstoareducedenthusiasmandcommitmentofthephysiciansinvolvedandthustoareluctanceinconductingmulticentreclinicaltrials.
Objective:Thepurposeofthisstudywastodevelopaweb-basedopensourceplatformtosupportamulti-centreclinicaltrial.
Methods:WedevelopedonDrupal,anopensourcesoftwaredistributedunderthetermsoftheGeneralPublicLicense,aweb-based,multi-centreclinicaltrialmanagementsystemwiththedesignscienceresearchapproach.
Results:Thissystemwasevaluatedbyuser-testingandwellsupportedseveralcompletedandon-goingclinicaltrialsandisavailableforfreedownload.
Conclusion:Opensourceclinicaltrialmanagementsystemsarecapableinsupportingmulti-centreclinicaltrialsbyenhancingefficiency,qualityofdatamanagementandcollaboration.
KeywordsClinicaltrialsmanagement,computer-supportedcooperativework,contentmanagementsystem,designscienceresearch,opensourceINTRODUCTIONObstaclesinclinicalresearcharecommonlyassociatedwiththeorganisation,supportandperformanceofclinicaltrials.
Theseobstaclescanbeevenmorepronouncedinthecaseofmulticentreclinicaltrials(superiortosinglecentretrials,conductedatseveralclinicalresearchcentresatthesametimewiththeyThispaperwaspresentedatthe15thInternationalSymposiumforHealthInformationManagementResearch(ISHIMR2011),8–9September2011,Zurich,Switzerland.
TheshortversionofthismanuscriptwasincludedintheConferenceProceedingsinprint.
*DARandTMcontributedequallyasfirstauthors.
Correspondence:DrDimitriAristotleRaptis,DivisionofVisceralandTransplantSurgery,SwissHPBCenter,UniversityHospitalZurich,Raemistrasse100,CH-8091,Zurich,Switzerland.
E-mail:Dimitri.
Raptis@usz.
chInformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
aimtoincreasethesamplesize,generalisabilityandhenceimpactoftheresults)wherecommunication,coordination,administration,datacollectionanddataentryareextremelyimportant.
Furthermore,additionalfundingisneededtocovertheextracostsofsuchrequirements.
Thesubstantiallyincreasedadministrativeburdenhassignificantlyreducedtheenthusiasmandcommitmentofthephysiciansinvolved,andinsomecases,totheextentthatthephysiciansdeclinetoparticipateinsuchstudies(1,2).
TheintroductionoflegalrequirementstoconductclinicaltrialsinEurope(3)hasledtoadropinthenumberoftrialsbeingconductedbyindependentacademicgroups(4)partiallyduetothefactthatreportinganddocumentationrequirementsarenowmoredifficultandtheprocesshasbecomeunnecessarilycomplicated(5).
Amajordifficultyincludesthefactthatclinicaltrialgroupsdonothavetheskillsorresourcestoestablishandusesoftwaresystemsrequiredtomanagetrialdata(2).
Thissituationisfurtherworsenedfornon-commercialacademicresearchgroups.
Hence,opensourceclinicaltrialmanagementsystems(CTMS)arevitalinconductingmulticentretrials.
Forthemostcommonlyused,commercialCTMSs(6),suchasOracleClinical(7)andClintrial(8),thecostsinvolvehundredsofthousandsofdollars,makingtheminaccessibletomostacademiccentresanddevelopingcountries.
Severalopen-sourceCTMSsoftwareexist(9–13),howevermostofthemalsoneedcommercialdatabasestorun,suchastheMicrosoftSQLServer(14)ortheOracleDatabaseManagementSystem(15)fordatamanagement(2).
Furthermore,theyneedspecialistITsupport,whichisalsoverycostlyandtimeconsuming.
Inadditiontheyaredifficulttobehandledevenbyexperiencedusersaswellastheyhaveseveraltechnicallimitations,sincetheyarerelativelyrigidandnotfullycustomisable(2).
Basedonthisbackground,theobjectiveofthisstudywastodevelopanopensourceweb-basedmulticentreclinicaltrialmanagementsystemtosupportaprospectivedouble-blindrandomisedclinicaltrial(RCT)(16)investigatingtheinter-observeragreementamongradiologistsindiagnosinglivertumoursbyMagneticResonanceImaging(MRI).
Theresultingartefacthadtobeaseasytouseaspossible,sothatradiologistsintuitivelycouldinteractwiththesystem,aswellastomeettoday'sexpectedperformanceandsecuritystandards.
Thissystemwasfurthertestedandimplementedinseveralotherclinicaltrialswithdifferentneedsandspecifications.
Theoutlineofthispaperisstructuredasfollows:inthenextsectionwewillintroduceourresearchapproach,referredtoasDesignScienceResearch(DSR).
Thiswillformthebasisforthedevelopmentandpre-evaluationofoursolution,whichwillbedescribedinthesectionthatfollows.
Finally,wewillsummarisethemainfindingsandoutlinesubsequentresearchactivitiesinthelastsection.
DESIGNSCIENCERESEARCHASANAPPROACHTOPRACTICALPROBLEMSOLVINGResearchthataimsatdesigningsolutionstocertaincategoriesofpracticaldifficultiesrequiresarelativelydifferentapproachastonaturalsciences,whichratherattemptstoexplainandpredictbehaviouralaspectsofrealitybydevelopingandverifyingtheories(17).
AccordingtoSimon(18),thesciencesoftheartificial''areconcernednotwiththenecessarybutwithD.
A.
Raptisetal.
68InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
thecontingent–notwithhowthingsarebutwithhowtheymightbe–inshort,withdesign''.
Asascience,designsciencethus''dealswiththesystematiccreationofknowledgeabout,andwith,design''(19).
Outofthisconception,differentimplications,meaningsanddefinitionsofwhatconsti-tutesthecentralscientificoutputsofDSRwerediscussedintherecentliterature(20–26).
However,''thereisstillalackofconsensusastothepreciseobjective–andthereforethedesiredoutputs–ofdesignscienceresearch''(27).
Typically,aspecialemphasisisgiventoconstructionofsufficientlynewordecisivelybetter''artefacts'',constitutingthe''possiblysole,orchief,outputoftheresearch''(28).
Theterm''artefact''isusedtodescribesomethingthatisartificial,orconstructedbyhumans,incontrasttosomethingthatoccursnaturally(18).
Awidelyacceptedbreakdownofoutputsistodifferentiatefourdistinctiveartefacttypes(17):(1)Constructs,thatisthelanguagetospecifyproblemsandsolutions,(2)models,thatistherepresentationoftherelationshipsoftheconstructs,(3)methods,thatistheprocedureorgoal-directedplanformanipulatingconstructstosystematicallyelucidatethedesignproblem,andfinally(4)instantiations,thatistheoperationalizationand/orphysicalconversionasproof-of-conceptofthepriorresearchoutputs(e.
g.
software).
InlinewithArcher(29),thelogicalnatureoftheactofdesigningmustnotnecessarilybedependentoftheartefacttobedesigned.
SeveralresearchershavethereforetriedtoclassifyacommonDSRprocessbydecomposingandclassifyingthecentraldesignactivitiesandbyfindingrationalesforeffectiveproblemsolving(30–32).
However,acommonlyagreedproceduremodelstilldoesnotexist.
Inordertobuildandevaluateour''artificial''problemsolution,i.
e.
aneasytouseandcost-efficientsoftwareformulticentreclinicaltrialsmanagement,wefollowedtheproblem-solvingphasessuggestedbyOffermanetal.
(33),whodifferentiatebetweentheproblemidentification,thesolutiondesignandtheevaluationofthedesign.
Followingthisguideline,ourresearchprocessisillustratedinFigure1.
DESIGNINGTHEMULTICENTRECLINICALTRIALSMANAGEMENTSYSTEMAccordingtoPurao(35),thedesignprocessisusuallyinitiatedbya''needandrequireintention''.
However,atthebeginningofthesearchprocesstherequirementsandfeasibilityofasolutionaremostlyambiguous,leadingtoastateofobscurityandpotentiallylackofagreementbetweenneedsandintention(36).
Asoundanalysisoftheproblemor''innovationopportunities''isaccordingtoJa¨rvinen(37),anessentialfirststepofthedesignprocessforclarifyingrelevanceandorientation.
BasedontheoutlineddesignresearchprocessinFigure1andtheresultsofthepracticalproblem-solvingprocedurearepresentedinthissection.
ProblemidentificationTogetadeeperunderstandingofthescopeandimportanceoftheprobleminpractice,weevaluatedtherelevantliteratureaswellasconductedaqualitativestudybymeansofsemi-structuredinterviews(38)askingfivesurgeons,threeradiologists,twoITspecialists,onemethodologist,oneManagingmulticentreclinicaltrials69InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
epidemiologistandonebiostatistician,ofdifferentlevelsofseniorityandwhoweremembersofthe''communityofpractice''.
Table1liststhemaincharacteristicsofthisgroup.
ThelevelsofITliteracywerearbitraryclassified.
''Professional''indicatesITbeingtheirmainprofession,''High''indicatesthattheyhavesomebasicprogrammingskillsandareusersofseveraltypesofmedicalsoftware,while''Low''indicatesnoneoftheabove.
''ITusage''indicatestheexposuretohealth-relatedsoftware(e.
g.
ElectronicHealthRecords(EHR),MedicalImaging)andwasalsoarbitrarilyclassified.
''Heavy''indicatesthatseveralmedicalsoftwareprogrammeshaveanintegralpartoftheirFigure1.
Designresearchprocess.
ModifiedfromOffermanetal.
(33).
ObtainedfromRaptisetal.
(34).
Table1.
Basiccharacteristicsofthemembersofthe''communityofpractice''.
AgeGenderProfessionITliteracyITusageCountry42MaleSurgeonLowLowSwitzerland38MaleSurgeonLowLowSwitzerland35MaleSurgeonLowHeavySwitzerland33MaleSurgeonHighHeavyUK31MaleSurgeonLowLowSwitzerland40MaleRadiologistHighHeavySwitzerland34MaleRadiologistLowHeavySwitzerland33MaleRadiologistHighHeavySwitzerland65MaleITSpecialistProfessionalLowGreece34MaleITSpecialistProfessionalHeavySwitzerland64FemaleMethodologistProfessionalLowGreece59MaleBiostatisticianHighHeavyGreece32FemaleEpidemiologistLowLowSwitzerlandD.
A.
Raptisetal.
70InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
everydayprofession,while''Low''indicatesthatapartfromtheEHRsystemoftheirinstitution,noothermedicalsoftwareispartoftheireverydayprofession.
Topre-evaluatetherelevanceofthespecifiedproblemarea,dataderivedwereusedtoelicitmajorcategoriesoftheparticipants.
Briefly,expectedrequirementstobeimplementedinournewsoftwarehadtoassistwiththeadministration,coordination,communication,collaboration,datasharingandcapturingofdataamongtheusersonapaper-freebasis.
Furthermore,ithadtobesecure,user-friendly,compatiblewithotheropen-sourcesoftwareusedinconjunctionwithit,builtwithnearlynocostsandhaveitasopen-accesstoserveasatemplateforotherclinicaltrialsfromourinstitutionsaswellasotheracademicinstitutionsworldwide.
Belowwestateindetailthedifferenttopics,theproblemsandneedsthatweredepictedduringtheproblemidentificationstage.
Administration,coordinationandcollaborationWiththeuseofanewClinicalTrialManagementSystem(CTMS),aresearchfellow(DAR)hadtoprovidethelocalInstitutionalReviewBoard(IRB)withthedocumentationnecessarytoobtainitspermissiontoconductthestudy(includingtheCTMS).
OurCTMShadtoassistuswiththestudystart-up,identifyeligiblepatientsfromallinstitutionsinvolvedandrecruitparticipat-ingresearchersfromdifferentcentresworldwide.
Allusershadtobetrainedonlineandsupportingmaterials(manuals,instructions)hadtobeavailableintheplatform.
TheCTMSfurtherhadtoassistwiththecoordinationofthewholestudyadministration,datasharing,datacollectionandMRIsharing.
OurCTMSalsohadtohelpusmonitorthesitesforcompliancewiththeclinicalprotocolandensurethattheseniorauthor/primaryinvestigatorreceived''clean''datafromeverysite.
Thedatahadtoberetrievedautomat-icallyandreadyforstatisticalanalysingwithoutanymodifications.
DatasecurityandanonymityAnonymisedpatientdata,includingdemographics,clinicalfindings,medicalhistory,laboratorydata,andMRIs,hadtobestoredinasecuredsitewithaccessgivenonlytotherelevantusers.
Furthermore,ourCTMShadtobesecuredwithaHypertextTransferProtocolSecure(HTTPS)incombinationwithaSSL/TLSprotocol.
Multi-leveluserpermissionsTheadministratorofCTMSneededtohaveaccesstoalldata,includingthepatientinformation,thelaboratoryresults,theMRIsandthehistopathologicalreports.
Forexample,thesurgeonshadtohaverestrictedaccessonlytotheirownpatientinformationandMRIswhiletheradiologists'accesstoallanonymisedMRIs.
CompatibilityandintegrationwithotheropensourcesoftwareTheonlineCTMShadtobecompatiblewithallwebbrowsers,includingversionsformobilephones.
AnMRIopenaccessDICOMviewer(MRIimagereader),integratedintoourCTMS,hadtobeprovidedtotheradiologistswithclearinstructionsonhowtouseittoviewthepatients'MRIs.
Managingmulticentreclinicaltrials71InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
OnlinedataentryanddatacaptureOurCTMShadtoprovidetheradiologistswithanonlinecasereportingform,includingavisualanaloguescaletorecordthelevelofconfidenceintheirdiagnosis.
Furthermore,thedatastoredintheMySQLserverhadtobeeasilyretrievedandconvertedtoafiletypecompatiblewithallstatisticalanalysispackages(e.
g.
txt,csv,excel,sas,sav).
Drop-downmenusandcertaindataentryruleshadtobeemployedtoensurenomissingdataaswellas''correct''data.
Upload,downloadandspeedTheMRIsfromdifferentcentresworldwidehadtobeuploadedbythesurgeonstoourCTMSandthensharedbyallowingallradiologiststodownloadtheMRIstotheirlocalcomputerorworkstation.
TheaverageMRIsizewas400MB,thustheuploadinganddownloadingtimehadtobeasfastaspossible.
SpecificCTMScharacteristicsOurnewCTMShadtobeavailableonline,simpletouse,self-explanatory–withouttheneedforspecialtraining.
Italsohadtobeuser-friendlyandwithasteeplearningcurve.
Instructionshadtobeavailableontheplatforminprintableformat,aswellashelpoptionsatallsections.
TheCTMShadtobedevelopedrelativelyfastsothatitdoesnotaffecttheoverallstudyconductiontimeandmanuscriptsubmission.
CostThedirectcostsforthedevelopmentofthesoftwarehadtobekeptataminimumlevelastherewasnofundingavailable.
Theanticipatedcostshadtoberelatedonlytotheregistrationofdomainnamesfor3years,theserverrental,andtheHTTPS/SSL/TLSprotocol.
Furthermore,ournewCTMShadtobebuiltandstoredinanopen-accessdatabase,availableasanopen-accesssoftwaresothatitcanactasatemplateforotherclinicaltrials.
Figure2illustratestheprocess-basedrequirementsthatthenewCTMShadtosupportforourclinicaltrialinvestigatingtheinter-ratedagreementamongradiologistsindiagnosingprimarylivertumoursbyMRI.
PatientshadtoberecruitedfrompotentialfivelivercentresinEurope,UnitedStatesandOceanPacific,andtheclinicalinformationaswellastheirMRIshadtobeuploadedtoasecurecentralserver.
Ourteamalsohadtoassureauser-friendlyinterfaceavailablefordownloadandevaluation,allowingtheMRIstobeviewedsimultaneouslybyatleasteightindependentradiologistsworldwideusinganopen-sourceDICOMviewercompatibletobothPCsandMacs.
Radiologiststakingpartinthisstudyhadtoprovideuswiththeirprimary,secondarydiagnosis,levelofcertainty,aswellasindicatethefeaturesinfluencingtheirdiagnosisinfreetext.
Thesedatawouldthenhavetobeanalysedbystatisticalsoftwareatinterimanalysisandattheendofthestudy.
Theanalyseddatawouldserveasbasisforpreparingthefinalreport,allowingaccesstoseveralphysicianstosimultaneouslyworkandmodifyit.
SolutiondesignSoftwarearchitectureTheCTMS(accessibleat(39))wasbuiltonDrupalversion6.
20(TheDrupalAssociation,Portland,OR),afreeandopensourcecontentmanagementD.
A.
Raptisetal.
72InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
system(CMS)writteninPHPanddistributedundertheGNUGeneralPublicLicense(40).
ThestandardreleaseofDrupal6.
20,knownasDrupalcore,containsbasicfeaturescommontomostCMSs.
Theseincludetheabilitytoregisterandmaintainindividualuseraccountswithinaflexibleandrichpermissionsystem,createandmanagemenus,customizepagelayouts,performlogging,andadministerthesystem.
DrupalwasalsodesignedtoallownewfeaturesandcustombehaviourtobeaddedtoextendDrupal'scorecapabilitiesviainstallationofplug-inmodulescreatedandcontributedtotheprojectbyopensourcecommunitymembers.
Furthermore,weusedanApacheserver(41)andMySQLasthedatabase(42).
UserinterfacesAfterthephysiciansfromthefivecentresuploadedall108patientcaserecordsandMRIstooursecuredserverusingtheopen-sourceFTPFileZillaClient3.
4.
0(FreeSoftwareFoundation,Inc.
,Boston,MA)(43)whichwasintegratedintoourCTMS,wepreparedtheuserinterface(Figures3–5,videodemoaccessibleat(44)),whichincludedasimpletouseviewingandreportingplatform.
Theeightradiologistswerethenabletologinwiththeirpersonalusernameandpassword,viewthepatientcaserecords(Figure3),viewtheMRIs(Figure4),andsubmittheirdiagnosisusingoursimpleonlineform(Figure5).
ViewingtheMRIswaspossiblewiththeintegratedopensourceDICOMviewerOsiriXv3.
7.
1(45)forMacandDicomWorksv1.
3.
5b(46)orK-PACSv1.
6.
0(47)forPC.
TheradiologistshadalsotheoptiontoviewtheMRIsattheworkstationoftheirowninstitution.
Avideodemonstrationoftheuserinterfaceisaccessibleat(44).
CLINICALTRIALDESIGNMULTICENTERPATIENTRECRUITMENTMRIuploadPatientinformationretrievalPROVIDECASESTORADIOLOGISTSMRIdownload/viewingCaserecordsavailableOBTAINRADIOLOGISTSREPORTSInterim&finaldataretrievalOnlinedatacaptureDATAANALYSIS&RESULTSPREPARATIONOFMANUSCRIPTSubmitforpublicationSharemanuscriptwithauthorsCompatibility-statisticalsoftwareResultssharingFigure2.
Flowchartfortheclinicaltrialdesign.
Managingmulticentreclinicaltrials73InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
Pre-evaluationOurCTMSwasreadyforuseafterapproximatelyonemonthofdevelopment.
Our''communityofpractice'',thelocalteamconsistingofsurgeons,radiolo-gists,ICTspecialists,methodologist,epidemiologistandbiostatisticiandeliveredourCTMSontime.
Itwasdesigned,empiricallyevaluatedandre-designedbasedonourmulti-disciplinaryteammeetingsthatoccurredtwiceweekly(Figure5).
Figure4.
ExampleofapatientMRIusingtheOsiriXopen-sourceimageviewer(45)integratedintoourCTMS.
Radiologistswereabletomanipulatetheimagesandmaketheirdiagnosisinthesamewaytheywouldatworkstationsintheirinstitution.
Figure3.
Exampleofapatientcaserecordwhereclinical,laboratorydataandMRIswhereavailableforviewinganddownloading.
D.
A.
Raptisetal.
74InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
Duringtheinterimanalysisandattheendofthestudy,wecouldsimplydownloadtheresultsinanexcelfilewhichwasautomaticallyimportedintotheSPSSversion19(48),whichisthesoftwareforperformingstatisticalanalyses.
Themanuscriptwaspreparedonlineallowingtheco-authorstocollaboratewithouthavingtonecessarilymeetoremailthedifferentversionsofthedocument,aspreviouslydescribed(34).
Thefinalversionofthemanuscriptwasfinallysubmittedtothespecifiedjournalbeforethepre-defineddeadline.
Wedidnotencounteranyseriousproblemsduringtheactivestudyphase.
Oneuserexperiencedalowdownloadrate.
AftercontactingusviaourCTMS,weadvisedhimtodownloadtheMRIsathisinstitutioncomputerratherthanfromthehomecomputerandtheproblemwasresolved.
Thisnotificationimmediatelybecameavailabletotherestoftheusers.
Figure5.
ThespecialisedformdesignedfortheRadiologistsinordertoreporttheirfindingsafterhavingfirstreviewedthepatientinformationandtheassociatedMRI.
Theprimarydiagnosisoptionswererandomisedforeachcasesothattheorderdidnotinfluencetheirdecision-making.
Thecertaintyofdiagnosisvisualanaloguescalewasalsoincluded.
Selectionlistswereusedtoensureauniformformatofreporting,aswellasfreetexttoindicatethefeaturestheirdiagnosiswasbasedon.
Allfieldsweremandatorytocompleteandthesystemdidnotallowsubmissionofdataunlessfullycompleted,indicatingthemissingparametersinredcolour.
Managingmulticentreclinicaltrials75InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
TheuserssimplyfollowedtheinstructionsavailableatourCTMSandprovideduswithextremely''clean''dataaswellaswithcomplete(nomissingdata).
Fourofthe''communityofpractice''membersconcludedthat''withouttheuseofournovelCTMS,theywouldnothavebeenabletosoefficientlyconductthestudywithinsuchashortperiodoftime''(28days).
Nearlyallusersfeltthatthesoftwarewas''self-explanatory''andthey''didnothavetoreadtheinstructions(availableattheCTMS)morethanones''.
Allusersalsofeltthat''thissystemwassuperiortoanytypeofpaper-baseddatacollectiontools''asitwasavailableatwork,athomeandonthego(iPad)''aswellasthatiswastime-efficient''.
ThedirectcostsfortheimplementationofthisCTMSwereE128(E9forthedomainnamepurchase,E119fortheserverrental(12months)).
However,theindirectcostsforthedevelopmentoftheCTMScouldnotbeaccuratelycalculatedasDARandTMbuiltduringtheirfreetime.
Atotalof96work-hourswereneededforthedevelopmentofthisCTMS.
Assumingarateof40Europerhour,thecostsfordevelopmentofthisCTMSwouldhavebeen3840Euro.
Furthermore,therewerenocostsinvolvedforthemaintenanceofthisplatformastheDrupalModuleswereupdatedbytheindividualDrupaldevelopersandautomaticallyinstalled.
ThisCTMSwasusedasatemplatetodevelopseveralothersystemswithminormodifications,withoutencounteringanyproblemsandwithinarelativelyshortperiodtofinaliseit(range5–12days).
Forexample,thisCTMSwasusedtosupportadouble-blindrandomisedcontrolledtrialassessingthebodyimageafterconventionalversussingleportlaparoscopiccholecystectomy(49,50).
OurnextprojectwastodevelopaCTMStosupportadiagnosticefficacyprospectivemulticentretrialassessingthevalueofanovelinflam-matorymarker(PancreaticStoneProtein)inpredictingacuteappendicitisinpatientsthatpresenttotheAccidentandEmergencydepartmentwithrightsidedabdominalpain(51,52).
AsDrupal6evolvedtoDrupal7,wedecidedtouseDrupal7todevelopthisCTMSinordertoensurethatthemodulesusedwerecontinuouslymaintainedandupdatedbytheindividualDrupaldevelopers.
WedidnotencounteranyproblemsbyswitchingtoDrupal7andthelearningcurvewasrelativelyshort.
ThisisbecauseDrupal7usedthesamemodulesasforDrupal6withthemaindifferencebeingamoreuser-friendlyadministrationplatform.
TofurtherexpandthedatamanagementpotentialofourCTMS,thesameplatformwasusedtosupporttheconductionof15SystematicReviewsoftheliteratureinvolving102collaboratorsworldwideandpresenttheirresultsatthe2012NET-Liver-MetastasesConsensusConference(PROSPEROProtocolRegistrationID:CRD42012002660,CRD42012002659,CRD42012002658,CRD42012002657,CRD42012002656,CRD42012002655,CRD42012002654,CRD42012002652,CRD42012002651,CRD42012002650,CRD42012002649,CRD42012002648,CRD42012002647,CRD42012002644,CRD42012002643(53,54)).
Finally,ourplatformwasexpandedtosupportinternationalpatientregistries(55)aswellasaninternationalonlineDelphimethodforthedevelopmentofpatientdecisionaids(56).
D.
A.
Raptisetal.
76InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
OurCTMSisdistributedunderGNUlicenceandisavailableforfreedownloadat:https://www.
psptrial.
com/qdownloadDISCUSSIONANDCONCLUSIONSOurnovelCTMSbuiltontheopen-sourcecontentmanagementsystemDrupal6.
20(40)wassuccessfullydevelopedandreadyforusewithinarelativelyshortperiodoftime.
Itprovidedapaper-freeelectronicplatformwherephysiciansfromfiveinstitutionsandeightradiologistsdistributedworldwide,shareddatainaverysimplemanneraspartofamulti-centreclinicaltrial.
CurrentlyexistingcommercialCTMS(6)areextremelyexpensive,espe-ciallyforacademicinstitutionsorgroupsbasedindevelopingcountries.
Investinginsuchsystemswouldcosthundredsofthousandsofdollars,dependingonthesizeofthetrialandnumberoflicensesneeded.
Furthermore,thereisalackofguidancefromregulatoryagenciesabouthowtoevaluate,aswellaswhattheexactrequirementsare,formulticentreclinicaltrials(2).
Open-sourceCTMSsoftwareexists(9–13),however,themajorityofthemneedcommercialdatabasestorunandtocaptureandmanagedata.
Furthermore,theyneedspecialistICTsupport,whichcouldalsobecostlyandtimeconsuming(2).
Asanopen-sourceCTMS,oursoftwarewastestedasthesystemtosupportanothermulti-centreclinicaltrial,whichprovedtobeeasilyre-designedandimplemented(49).
WeusedDrupalasthebackboneofourCTMS.
Drupalisafreesoftwarepackagethatallowsuserstoeasilyorganize,manageandpublishtheircontent,withanendlessvarietyofcustomization.
Drupalisanopensourcesoftwaremaintainedanddevelopedbyacommunityof630000usersanddevelopers.
ItisdistributedunderthetermsoftheGNUGeneralPublicLicense.
ThisopendevelopmentmodelmeansthatdevelopersworldwideareconstantlyworkingtoensureDrupalisacutting-edgeplatformthatsupportsthelatesttechnologiesthattheWebhastooffer.
TheDrupalproject'sprinciplesencouragemodularity,standards,collaborationandease-of-use(40).
Firstevidencefromourpre-evaluationindicatesthatourCTMSisbeingconsideredsuperiortoothers(2,9–13)sincethedirectcostsinvolvedtodevelopwereverylow,duetotheeaseforcustomisationaccordingtospecificneedsandthefactthatnotrainingwasneededapartfromclearinstructionspresentinthesystem.
However,thehiddencostscouldnotbeaccuratelycalculatedandthisposesalimitationtoourstudy.
Oursystemwasmulti-centricandsharingcomplexdataandfileswasnotaproblem.
Inaddition,communicationbetweenthecentreswasuncomplicated.
However,furtherevaluationactivitiesareneededinordertoprovethesuperiorityofourCTMSascomparedtoothers.
Althoughtheneedforsuchanopen-sourceCTMSisgreatinacademicinstitutionsandmulti-centrestudies,ifdesignedappropriately,itsimpactcanalsobebeneficialtoalltypesofresearch,includingindustry-driven.
The''open-sourceapproach''hasbeensuccessfulinassistingmanyscientificandbiomedicalstudies,suchasthelanguagePERLthatwassaidtohave''savedthehumangenomeproject''(2).
WebelievethatournovelCTMSwouldbehighlycostefficientbyexcludingtheneedtouseexpensivecommercialCTMS,henceenablingthecompletionofmoreclinicaltrials.
Additionally,morephysicianswouldbeencouragedtoconductmulti-centreclinicaltrials,knownManagingmulticentreclinicaltrials77InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
tobesuperiortosingle-centretrialsduetotheincreaseinthegeneralizabilityofthefindings(2).
Thisstudyhassomelimitations.
OurCTMSwasnotyetformallyevaluatedatitsfinalstage.
Evaluationsandre-designwereperformedduringthedevelopmentphaseinanempiricalmanner.
Thenextobviousstepistorequestallend-userstoevaluateitwithavalidatedinstrument.
Inconclusion,ournovelCTMSbuiltonDrupalatrelativelylowdirectcosts,actedasabackboneforourmulti-centreclinicaltrial,bysupportingadministration,collaboration,communicationandinformationsharingneedsamongphysiciansfromseveralcentresworldwide.
Italsohelpedtodeliverhighlyvaluablereportswithinaveryshortperiodoftime.
ThisCTMSisavailableasanopen-sourceandcanactasatemplateforfurtheradjustments,customisationsandimprovements.
Thenextstepistoconductaprospectivetrialformallyevaluatingandcomparingtheusers'experienceanddataentryqualityofthisweb-baseddesktopPCCTMSwiththatofamobileapplication(e.
g.
iPad)(http://clinicaltrials.
gov/ct2/show/NCT01473238).
DECLARATIONOFINTERESTTheauthorshavenoconflictsofinteresttodeclare.
ETHICALCONSIDERATIONSOurworkcomplieswiththeDeclarationofHelsinkiandtheCantonofZurichEthicsCommitteehasapprovedourresearch(KEK-ZH-Nr.
2010-0475).
REFERENCES1.
DelawiD,DhertWJ,OnerFC.
ConductingaEuropeanmulti-centertrial:firstexperienceswiththenewEUClinicalTrialsDirectivefromanacademicperspec-tive.
EurSpineJ2008;17:1113–15.
2.
FeganGW,LangTA.
Couldanopen-sourceclinicaltrialdata-managementsystembewhatwehaveallbeenlookingforPLoSMed2008;5:e6.
3.
InternationalConferenceonHarmonisationofTechnicalRequirementsforRegistrationofPharmaceuticalsforHumanUse[Internet].
InternationalConferenceonHarmonisation'sguidelinesongoodclinicalpractice.
Geneva,Switzerland[cited2012Feb25].
Availablefrom:http://www.
ich.
org/4.
HemminkiA,Kellokumpu-LehtinenPL.
HarmfulimpactofEUclinicaltrialsdirective.
BMJ2006;332:501–2.
5.
GrimesDA,HubacherD,NandaK,etal.
TheGoodClinicalPracticeguideline:abronzestandardforclinicalresearch.
Lancet2005;366:172–4.
6.
ZelenM.
Biostatisticians,biostatisticalscienceandthefuture.
StatMed2006;25:3409–14.
7.
OracleCorporation[Internet].
OracleClinical.
Redwood,UnitedStates[cited2012Feb25].
Availablefrom:http://www.
oracle.
com/us/industries/life-sciences/046720.
html8.
PhaseForward[Internet].
ClinTrial.
Waltham,UnitedStates[cited2010Jan10].
Availablefrom:http://www.
phaseforward.
com/products/clinical/cdm/cis/9.
OpenClinicaLLC[Internet].
OpenClinica3.
1.
2.
Waltham,UnitedStates[cited2012Feb25].
Availablefrom:https://community.
openclinica.
com/D.
A.
Raptisetal.
78InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
10.
CaisisFoundation[Internet].
IntegratedClinicandResearchManagementSystem.
NewYork,UnitedStates[cited2012Feb25].
Availablefrom:http://www.
caisis.
org/.
11.
TolvenInc.
[Internet].
Tolven2.
1.
SanFrancisco,UnitedStates[cited2012Feb25].
Availablefrom:http://www.
tolven.
org/12.
LabKeySoftwareFoundation[Internet].
LabKeyServer11.
3.
SouthLakeUnion,UnitedStates[cited2012Feb25].
Availablefrom:https://www.
labkey.
org/13.
YaleCenterforMedicalInformatics[Internet].
TrialDB.
NewHaven,UnitedStates[cited2012Feb25].
Availablefrom:http://ycmi.
med.
yale.
edu/trialdb/14.
MicrosoftCorporation[Internet].
MicrosoftSQLServer.
Redmond,UnitedStates[cited2012Feb25].
Availablefrom:http://www.
microsoft.
com/sqlserver/en/us/default.
aspx15.
OracleCorporation[Internet].
OracleDatabase11g.
Redwood,UnitedStates[cited2012Feb25].
Availablefrom:http://www.
oracle.
com/us/products/database/index.
html16.
BreitensteinS,RaptisDA,FischerMA,etal.
Pooragreementamongexpertradiologistsindiagnosingprimarylivertumorsbymagneticresonanceimaginginnon-cirrhoticpatients:adoubleblindinternationalmulticenterstudy.
18thAnnualMeetingoftheEuropeanSurgicalAssociation(ESA2011);2011May19–21;Helsinki,Finland.
17.
MarchST,SmithGF.
Designandnaturalscienceresearchoninformationtechnology.
DecisSupportSyst1995;15:251–66.
18.
SimonHA.
Thesciencesoftheartificial.
3rded.
Cambridge(MA):MITPress;1996.
19.
BaskervilleR.
Whatdesignscienceisnot.
EuropeanJInforSys2008;17:441–3.
20.
McKayJ,MarshallP.
Science,design,anddesignscience:seekingclaritytomovedesignscienceresearchforwardininformationsystems.
18thAustralasianConferenceonInformationSystems(ACIS2007);2007Dec5–7;Toowomba,Australia.
21.
WinterR.
DesignscienceresearchinEurope.
EuropeanJInforSys2008;17:470–5.
22.
CarlssonSA.
Designscienceresearchininformationsystems:acriticalrealistperspective.
17thAustralasianConferenceonInformationSystems(ACIS2006);2006Dec6–8;Adelaide,Australia.
23.
IivariJ.
AparadigmaticanalysisofinformationsystemsasaDesignScience.
ScandJInforSys2007;19:39–63.
24.
OrlikowskiWJ,IaconoCS.
Researchcommentary:desperatelyseeking''IT''inITresearch–acalltotheorizingtheITartifact.
InformSysRes2001;12:121–34.
25.
HevnerAR,ChatterjeeS.
Designresearchininformationsystems:theoryandpractice.
NewYork:Springer;2010.
26.
OffermannP,BlomS,Scho¨nherrM,BubU.
Artifacttypesininformationsystemsdesignscience:aliteraturereview.
5thInternationalConferenceonDesignScienceResearchinInformationSystemsandTechnology(DESRIST2010);2010Jun4–5;St.
Gallen,Switzerland.
27.
VaishnaviV,KuechlerB.
Designscienceresearchmethodsandpatterns:innovatinginformationandcommunicationtechnology.
BocaRaton(FL):AuerbachPublications;2008.
28.
GregorS,JonesD.
Theanatomyofadesigntheory.
JAssocInformSys2007;8:312–35.
29.
ArcherLB.
Thestructureofthedesignprocess.
In:BroadbentG,WardA,eds.
Designmethodinarchitecture.
LundHumphries:London;1969:76–102.
30.
NunamakerJF,ChenM,PurdinTDM.
Systemsdevelopmentininformationsystemsresearch.
JManageInformSys1990;7:89–106.
31.
CoyneR.
Logicmodelsofdesign.
London:Pitman;1988.
32.
TakedaH,VeerkampP,TomiyamaT,YoshikawaH.
Modelingdesignprocesses.
AIMag1990;11:37–48.
33.
OffermannP,LevinaO,Scho¨nherr,BubU.
Outlineofadesignscienceresearchprocess.
4thInternationalConferenceonDesignScienceResearchinInformationSystemsandTechnology(DESRIST2009);2009May7–8;Philadelphia,UnitedStates.
Managingmulticentreclinicaltrials79InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.
34.
RaptisDA,MettlerT,TzanasK,GrafR.
Anovelopen-sourceweb-basedplatformpromotingcollaborationofhealthcareprofessionalsandbiostatisticians:adesignscienceapproach.
InformHealthSocCare2012;37:22–36.
35.
PuraoS.
Designresearchinthetechnologyofinformationsystems:truthordare.
GSUDepartmentofCISWorkingPaper:Atlanta;2002.
36.
WieringaR.
Designscienceasnestedproblemsolving.
4thInternationalConferenceonDesignScienceResearchinInformationSystemsandTechnology(DESRIST2009);2009May7–8;Philadelphia,UnitedStates.
37.
Ja¨rvinenP.
Onreviewingofresultsindesignresearch.
15thEuropeanConferenceonInformationSystems(ECIS2007);2007June7–9;St.
Gallen,Switzerland.
38.
PattonMQ.
Qualitativeresearch&evaluationmethods.
3rded.
ThousandOaks(CA):SagePublications;2002.
39.
RaptisDA,ClavienPA,BreitensteinS.
MRI-Evaluation[Internet].
Zurich,Switzerland[cited2012Feb25].
Availablefrom:http://www.
mri-evaluation.
com.
40.
BuytaertD.
DrupalOpenSourceCMS[Internet][cited2012Feb25].
Availablefrom:http://drupal.
org/41.
TheApacheSoftwareFoundation[Internet].
ApacheHTTPServerProject[cited2012Feb25].
Availablefrom:http://httpd.
apache.
org/42.
OracleCorporation[Internet].
MySQL.
Redwood,UnitedStates[cited2012February25].
Availablefrom:http://www.
mysql.
com/43.
KosseT[Internet].
FileZillaFTPClient[cited2012Feb25].
Availablefrom:http://filezilla-project.
org/44.
RaptisDA,BreitensteinS.
MRI-Evaluation.
comVideoDemo.
Zurich,Switzerland[Internet][cited2012Feb25].
Availablefrom:http://www.
mri-evaluation.
com/demo45.
RossetA.
OsiriXImagingSoftware[Internet][cited2012Feb25].
Availablefrom:http://www.
osirix-viewer.
com/46.
PuechP.
DicomWorks[Internet][cited2012Feb25].
Availablefrom:http://www.
dicomworks.
com/47.
KnopkeA.
K-PACSDICOMViewingSoftware.
Plauen,Germany[Internet][cited2012Feb25].
Availablefrom:http://www.
k-pacs.
de/48.
IBMCorporation.
SPSSStatisticsv19.
Armonk,UnitedStates[Internet][2012Feb25].
Availablefrom:http://www.
spss.
com/49.
SteinemannDC,RaptisDA,LurjeG,etal.
Cosmesisandbodyimageaftersingle-portlaparoscopicorconventionallaparoscopiccholecystectomy:amulticenterdoubleblindedrandomisedcontrolledtrial(SPOCC-trial).
BMCSurg2011;11:24.
50.
RaptisDA.
Single-portRCTmanagementsystem.
Zurich,Switzerland[Internet][cited2012Feb25].
Availablefrom:www.
Single-Port-RCT.
com51.
TschuorC,RaptisDA,LimaniP,etal.
Thevalueofpancreaticstoneproteininpredictingacuteappendicitisinpatientspresentingattheemergencydepartmentwithabdominalpain.
BMCGastroenterol2012;12:154.
52.
RaptisDA.
PSPClinicalTrialManagementSystem(PSPtrial).
Zurich,Switzerland.
[Internet][cited2012October5].
Availablefrom:www.
PSPtrial.
com53.
InternationalProspectiveRegisterforSystematicReviews(PROPSERO)[Internet][cited2012October5].
Availablefrom:http://www.
crd.
york.
ac.
uk/Prospero/54.
RaptisDA.
Internationalconsensusconferenceonthemanagementofnueroendo-crinelivermetastases(NET-Liver-Metastases),UK[Internet][cited2012October5].
Availablefrom:http://www.
net-liver-mets.
com/qscientific-process55.
SchaddeE,RaptisDA,ClavienPA.
ALPPS.
net.
Zurich,Switzerland.
[Internet][cited2012October5].
Availablefrom:http://www.
alpps.
net/56.
RaptisDA,MelloulM,LesurtelM,ClavienPA.
Livingdonorlivertransplanationinformationanddecisionaid.
Zurich,Switzerland.
[Internet][cited2012October5].
Availablefrom:http://www.
LDLT.
org/D.
A.
Raptisetal.
80InformHealthSocCareDownloadedfrominformahealthcare.
combyRoyalCollegeofGeneralPractionerson02/11/14Forpersonaluseonly.