IAEA培訓(xùn)材料近年放射性事故分析

Module2.10:Accidentupdate

–somenewerevents

(UK,USA,France)IAEATrainingCourseQuestionsDoyouthinktheaccidentshavenothappenedinrecentyears?Doyouthinkwell-developedcentresareimmunetotheseaccidents?2OverviewItshouldbenotedthattheintentiscertainlynottoreflectthequotedcentresinthispresentationinpoorlightInstead,thepurposeistodrawlessonsInmanycases,thecentreshaveaqualitysysteminplaceTheeventsarereconstructedfrominformationinthepublicdomain,andmightdifferfromactualeventsduetogapsinthisinformation.3Overview1stexample: Incorrectmanualparametertransfer (UK)2ndexample: Reversalofimages(USA)3rdexample: Inappropriatemeasuringdevice (France)4thexample: Erroneouscalculationforsoftwedges (France)5thexample: IncorrectIMRTplanning(USA)6thexample: Moreinformationneeded…Newerexamplesofaccidentsinradiotherapyfrom2004to200741stexample:Incorrectmanualparametertransfer(UK)IAEATrainingCourseBackgroundJanuary2006attheBeatsonOncologyCentre(BOC)inGlasgow,ScotlandAtthetime:RadiotherapyphysicsstaffinglevelsinScotlandlessthan60%oftherecommendedlevel“Glasgowhasproblemswithrecruitingphysicists,asshownbytheirhighnumberofvacancies.〞TheBeatsonOncologyCentreinGlasgow6BackgroundTreatmentplanningatBOC:14.5wholetimeequivalent(WTE)staffwereavailableforbetween4500and5000newtreatmentplansperyear.WhenstaffinglevelswerecomparedwithguidelinesfromIPEM,itwasseenthat18WTEstaffwouldbetherecommendedlevel.7BackgroundTreatmentplanningatBOC:PlanningstaffmembersandplanningprocedureswerebothcategorizedAtoCdenotesseniortojuniorstaffAtoEdenotessimpletocomplexplansThemaindutiesperstaffcategoryisoutlinedincolumn4Tablefrom:“ReportofaninvestigationbytheInspectorappointedbytheScottishMinistersforTheIonisingRadiation(MedicalExposures)Regulations2000〞8BackgroundTreatmentplanningatBOC:Practicepriorto2005hadbeentoletthetreatmentplanningsystem(TPS)calculatetheMonitorUnits(MU)for1GyfollowedbymanualmultiplicationwiththeintendeddoseperfractionforthecorrectMU-settingtouse.9BackgroundTreatmentplanningatBOC:InMay2005,theRecordandVerify(RV)systemwasupgradedtobeamoreintegratedplatform.ThecentredecidedtoinputthedoseperfractionalreadyintheTPS,formostbutnotalltreatmenttechniques.10Whathappened?5thJanuary2006,LisaNorris,15yearsold,startedherwholeCNStreatmentatBOCThetreatmentplanwasdividedintohead-fieldsandlowerandupperspine-fieldsThisisconsideredtobeacomplextreatmentplan,performedaboutsixtimesperyearattheBOC.LisaNorris11Whathappened?Thebulkoftheplanningwasdoneby“PlannerX〞inDec’05,ajuniorplanner“PlannerX〞hadnotyetbeenregisteredinternallytobecompetenttoplanwholeCNS,ortotrainonthese“PlannerX〞gotinitialinstructionsandtheopportunitytobesupervisedwhencreatingtheplan12Whathappened?WholeCNSplansstillwentbythe“oldsystem〞,whereTPScalculatesMUfor1GywithsubsequentupscalingfordoseperfxA“medullaplanningform〞wasused,whichispassedtotreatmentradiographersforfinalMUcalculationsTablefrom:“ReportofaninvestigationbytheInspectorappointedbytheScottishMinistersforTheIonisingRadiation(MedicalExposures)Regulations2000〞13Whathappened?HOWEVER–“PlannerX〞lettheTPScalculatetheMUforthefulldoseperfx–notfor1GyasintendedSincethedoseperfxtotheheadwas1.67Gy,theMU’senteredintheformwere67%toohighforeachofthehead-fieldsTablefrom:“ReportofaninvestigationbytheInspectorappointedbytheScottishMinistersforTheIonisingRadiation(MedicalExposures)Regulations2000〞14Whathappened?ThiserrorwasnotfoundbythemoreseniorplannerswhocheckedtheplanTheradiographerontheunitthusmultipliedwiththedoseperfxasecondtime2.92GyperfxtotheheadTablefrom:“ReportofaninvestigationbytheInspectorappointedbytheScottishMinistersforTheIonisingRadiation(MedicalExposures)Regulations2000〞15“PlannerX〞calculatedanotherplanofthesamekindandmadethesamemistakeThistime,theerrorwasdiscoveredbyaseniorchecker(1stofFeb‘’06)Thesameday,theerrorincalculationsforLisaNorriswasalsoidentifiedDiscoveryofaccident16ThetotaldosetoLisaNorrisfromtheRightandLeftLateralheadfieldswas55.5Gy(19x2.92Gy)ShediedninemonthsaftertheaccidentImpactofaccident17LessonstolearnEnsurethatallstaffAreproperlytrainedinsafetycriticalproceduresAreincludedintrainingprogrammesandhassupervisionasnecessary,andthatrecordsoftrainingarekeptup-to-dateUnderstandtheirresponsibilitiesIncludeintheQualityAssuranceProgramFormalproceduresforverifyingtherisksfollowingtheintroductionofnewtechnologiesandproceduresIndependentMUcheckingofALLtreatmentplansReviewstaffinglevelsandcompetencies18ReferencesUnintendedoverexposureofpatientLisaNorrisduringradiotherapytreatmentattheBeatsonOncologyCentre,GlasgowinJanuary2006.ReportofaninvestigationbytheInspectorappointedbytheScottishMinistersforTheIonisingRadiation(MedicalExposures)Regulations2000(2006)CancerinScotland:RadiotherapyActivityPlanningforScotland2021–2021.ReportofTheRadiotherapyActivityPlanningSteeringGroup’TheScottishExecutive.Edinburgh.(2006)TheGlasgowincident–aphysicist’sreflections.W.P.M.Mayles.ClinOncol19:4-7(2007)Radiotherapynearmisses,incidentsanderrors:radiotherapyincidentinGlasgow.M.V.Williams.ClinOncol19:1-3(2007)192ndexample:Reversalofimages(USA)

IAEATrainingCourseWhathappened?October2007attheKarmanosCancerCenter(KCC)inmidtownDetroit,Michigan,USAAttheGammaKnifetreatmentfacility,apatientwassetupforMRIimagingStandardpracticeistopositionthepatient“headfirst〞Thepatientwaspositioned“headfirst〞,but“feetfirst〞scantechniquewaschosenontheunitTheKCCinDetroit21Whathappened?Theaxialimageswerethereforereversedleft-to-rightThephysicistdidnotseethemistakewhenimportingimagesintotheTPSTheerrorresultedinan18mmshiftofisocentreacrossthemidlineofthebrainStereotactictreatment(imagefromKCC)22LessonstolearnIncludeintheQualityAssuranceProgramProceduresforverifyingleftfromrightinsafetycriticalimages,e.g.byusingfiducialmarkersEnsuretherearewrittenprotocolsposted,knownandfollowed,forsafetycriticalprocedures23ReferencesGammaknifetreatmenttowrongsideofbrain.EventNotificationReport43746.UnitedStatesNuclearRegulatoryCommission(2007)243rdexample:Inappropriatemeasuringdevice(France)

IAEATrainingCourseBackgroundReported2007atH?pitaldeRangueilinToulouse,FranceInApril2006,thephysicistinthecliniccommissionedthenewBrainLABNovalisstereotacticunitThisunitcanoperatewithmicroMLC’s(3mmleaf-width)orconicalstandardcollimatorsTheH?pitaldeRangueilinToulouse26BackgroundVerysmallfieldscanbedefinedwiththemicroMLC’sHighdosetoa6x6mmfieldiswithincapabilityTheTPSrequirespercentdepthdoses,beamprofilesandrelativescatterfactorsdowntothisfieldsizeCaremustbetakenwhenmeasuringsmallfields!27Whathappened?DifferentmeasuringdeviceswereusedbythephysicistAmeasuringdevicenotsuitableforcalibratingthesmallestmicrobeamswasused“…anionisationchamberofinappropriatedimensions…〞accordingtoNuclearSafetyAuthority(ASN)inspectors28Whathappened?TheincorrectdatawasenteredintotheTPSAllpatientstreatedwithmicroMLCwereplannedbasedonthisincorrectdataPatientstreatedwithconicalcollimatorwerenotaffected29BrainLABdiscoveredthatthemeasurementfilesdidnotmatchupwiththoseatothercomparablecentres,duringaworldwideintercomparisonstudyItshouldbenotedthatthecompanydoesnotvalidateorholdresponsibilityforlocalmeasurementsorimplementationDiscoveryofaccident30ImpactofaccidentTreatmentbasedontheincorrectdatawentonforayear(Apr′06–Apr′07)AllpatientstreatedwithmicroMLCwereaffected(145of172stereotacticpatients)Thedosimetricimpactwasevaluatedassmallinmostcases,with6patientsidentifiedforwhomover5%ofthevolumeofhealthyorgansmayhavebeenaffectedbydoseexceedinglimits31LessonstolearnEnsurethatstaffUnderstandthepropertiesandlimitationsoftheequipmenttheyareusingIncludeintheQualityAssuranceProgramIntercomparisonwithotherhospitals,i.e.independentcheckofnewequipmentbyindependentgroup(usingindependentequipment)beforeequipmentisclinicallyused32ReferencesReportconcerningtheradiotherapyincidentattheuniversityhospitalcentre(CHU)inToulouse–RangueilHospital.ASN–AutoritédeS?retéNucléaire(2007)334thexample:Erroneouscalculationforsoftwedges(France)

IAEATrainingCourseBackgroundInMay2004atCentreHospitalierJeanMonnetinEpinal,France…itwasdecidedtochangefromstatic(hard)wedgestodynamic(soft)wedgesforprostatecancerpatientsInacountryoffewMedicalPhysicists(MP),thisfacilityhadasingleMPwhowasalsooncallinanotherclinicTheJeanMonnetHospitalinEpinal35BackgroundInpreparationforthechangeintreatmenttechnique,twooperators(treatmentplanners?)weregiventwobriefdemo’sTheoperatorsdidnothaveanyoperatingmanualintheirnativelanguage36BackgroundWhenthesoftwedgeswereintroduced:TheindependentMUcheckinusecouldnotbeusedanymore(unlessmodified)Thediodesusedforindependentdosecheckcouldnotbecorrectlyinterpretedanymore37Whathappened?TreatmentplanningwithsoftwedgesstartedNotallthetreatmentplannersdidunderstandtheinterfacetotheplanningsystem38Whathappened?TreatmentplanningwithsoftwedgesstartedNotallthetreatmentplannersdidunderstandtheinterfacetotheplanningsystemSomeselectedtheplanningformechanicalwedgewhenintendingdynamicwedgev39Whathappened?TreatmentplanningwithsoftwedgesstartedNotallthetreatmentplannersdidunderstandtheinterfacetotheplanningsystemSomeselectedtheplanningformechanicalwedgewhenintendingdynamicwedgeInsteadtheyshouldhaveselectedDynamicWedge…v40Whathappened?TreatmentplanningwithsoftwedgesstartedNotallthetreatmentplannersdidunderstandtheinterfacetotheplanningsystemSomeselectedtheplanningformechanicalwedgewhenintendingdynamicwedgeInsteadtheyshouldhaveselectedDynamicWedge……whichwouldhaveletthecorrectplanningtoolappearv41Whathappened?Whenplanningwasfinishedandtheisodosedistributionapproved…theparametersweremanuallytransferredtothetreatmentunitManuallytransferredMU’swouldhavebeencalculatedformechanicalwedgesandwouldbemuchgreaterthanwhatisneededforgivingthesamedosewithdynamicwedges42Detailsnotclear,BUT:itmighthavebeenwhenMUchecksoftwarewasreplacedandupdatedtobeabletohandleindependentcheckingofdynamicwedges.Discoveryofaccident43ImpactofaccidentTreatmentbasedonincorrectMU’swentonforoverayear(6May2004–1Aug2005)Atleast23patientsreceivedoverdose(20%ormorethanintendeddose)BetweenSeptember2005andSeptember2006,fourpatientsdied.Atleasttenpatientsshowsevereradiationcomplications(symptomssuchasintensepain,dischargesandfistulas)44Informationfollowingaccident15Sep2005,twodoctorsfromtheclinicpassedoninformationthatwenttotheRegionalDept.ofHealthandSocialSecurity(DDASS)5Oct2005ameetingwasheldatDDASS.Decisionswerenotdocumentedoruniformlyinterpreted.Nationalauthoritiesinchargewerenotinformedatthisstage,butonlyafullyearaftertheaccident(July2006)45Informationfollowingaccident7patientswereinformedduringthelastquarterof2005.16otherpatientswere(wrongly)considerednotobeaffected.Ofthese……3wereinformedbyanotherdoctorthantheirradiotherapist…1learntfromathirdpartyperson…1learntfromthepress…1learntbyoverhearingadoctorspeakingtoacolleague…4wereinformedbymanagement2daysbeforepressrelease…1diedbeforebeinginformed46LessonstolearnEnsurethatstaffUnderstandthepropertiesandlimitationsoftheequipmenttheyareusingAreproperlytrainedinsafetycriticalproceduresIncludeintheQualityAssuranceProgramFormalproceduresforverifyingnewtechnologiesandproceduresbeforeimplementationIndependentMUcheckingofALLtreatmentplansInvivodosimetryMakesuretheclinichasasysteminplaceforInvestigationandreportingofaccidentsPatientmanagementandfollowup,includingcommunicationtopatientsInstructionsshouldbeinalanguagethatisunderstood47ReferencesSummaryofASNreportn°2006ENSTR019-IGASn°RM2007-015PontheEpinalradiotherapyaccident.G.Wack,F.Lalande,M.D.Seligman(2007)Accidentderadiothérapieàépinal.P.J.Compte.SociétéFran?aisedePhysiqueMédicale(2006)LessonsfromEpinal.D.Ash.ClinOncol19:614-615(2007)48PostscripttoaccidentinEpinalGoingthroughtherecords,twofurtherepisodeswerereportedsubsequentlyReportedinFeb2007:Inthetimeperiod2001-2006,portalimagingwasusedrepeatedlywithouttakingintoaccounttheaddeddose(estimatedtohavebeen+8%oftotal)for412patientsundermedicalsurveyReportedinJuly2007:Inthetimeperiod1989-2000,useofanin-houseTPSnotupdatedafterchangeintreatmenttechnique,mighthaveledto300patientsreceivingupto7%addeddose.495thexample:IncorrectIMRTplanning(USA)IAEATrainingCourseBackgroundMarch2005,somewhereinthestateofNewYork,USAApatientisduetobetreatedwithIMRTforheadandneckcancer(oropharynx)51Whathappened?March4–7,2005AnIMRTplanisprepared:“1Oropharyn〞.AverificationplaniscreatedintheTPSandmeasurementsbyPortalDosimetry(withEPID)confirmscorrectness.ExampleofanEPID(ElectronicPortalImagingDevice)(Picture:P.Munro)52Whathappened?March8,2005Thepatientbeginstreatmentwiththeplan“1Oropharyn〞.Thistreatmentisdeliveredcorrectly.“Modelview〞oftreatmentplan(Picture:VMS)53Whathappened?March9-11,2005Fractions#2,3and4arealsodeliveredcorrectly.VerificationimagesforthekVimagingsystemarecreatedandaddedtotheplan,nowcalled“1AOropharyn〞.“Modelview〞oftreatmentplan(Picture:VMS)54Whathappened?March11,2005Thephysicianreviewsthecaseandwantsamodifieddosedistribution(reducingdosetoteeth)“1AOropharyn〞iscopiedandsavedtotheDBas“1BOropharyn〞.“Modelview〞oftreatmentplan(Picture:VMS)55Whathappened?March14,2005Re-optimizationworkon“1BOropharyn〞startsonworkstation2(WS2).Fractionationischanged.Existingfluencesaredeletedandre-optimized.NewoptimalfluencesaresavedtoDB.Finalcalculationsarestarted,whereMLCmotioncontrolpointsforIMRTaregenerated.Normalcompletion.MultiLeafCollimator(MLC)56Whathappened?March14,2005,11a.m.“Saveall〞isstarted.AllnewandmodifieddatashouldbesavedtotheDB.Inthisprocess,dataissenttoaholdingareaontheserver,andnotsavedpermanentlyuntilALLdataelementshavebeenreceived.Inthiscase,datatobesavedincluded:(1)actualfluencedata,(2)aDRRand(3)theMLCcontrolpointsADigitallyReconstructedRadiograph(DRR)ofthepatient57Whathappened?March14,2005,11a.m.Theactualfluencedataissavednormally.NextinlineistheDRR.The“Saveall〞processcontinueswiththis,butisnotcompleted.SavingofMLCcontrolpointdatawouldbeaftertheDRR,butwillnotstartbecauseoftheabove.ADigitallyReconstructedRadiograph(DRR)ofthepatient58Whathappened?March14,2005,11a.m.Anerrormessageisdisplayed.Theuserpresses“Yes〞,whichbeginsasecond,separate,savetransaction.MLCcontrolpointdataismovedtotheholdingarea.Thetransactionerrormessagedisplayed59Whathappened?March14,2005,11.a.m.TheDRRis,however,stilllockedintothefaultyfirstattempttosave.Thismeansthesecondsavewon’tbeabletocomplete.Thesoftwarewouldhaveappearedtobefrozen.Thefrozenstateofthesecond“SaveAll〞progressindication60Whathappened?March14,2005,11.a.m.TheuserthenterminatedtheTPSsoftwaremanually,probablywithCtrl-Alt-DelorWindowsTaskManagerAtmanualtermination,theDBperformsa“roll-back〞toreturnthedataintheholdingareatoitslastknownvalidstateThetreatmentplannowcontains(1)actualfluencedata;(2)notthefullDRR;(3)noMLCcontrolpointdataCtrl-Alt-Del61Whathappened?March14,2005,11.a.m.Within12s,anotherworkstation,WS1,isusedtoopenthepatientsplan.Theplannerwouldhaveseenthis:

Validfluenceswerealready saved.Calculationofdose distributionisnowdonebythe plannerandsaved.MLCcontrol pointdataisnotrequiredfor calculationofdosedistribution.Sagittalviewofpatient,withfieldsanddosedistribution62Whathappened?March14,2005,11.a.m.Nocontrolpointdataisincludedintheplan.

Thesagittalviewshouldhavelookedliketheonetotheright,withMLCs63Whathappened?March14,2005,11a.m.Noverificationplanisgeneratedorusedforcheckingpurposes,priortotreatment(shouldbedoneaccordingtoclinicsQAprogramme)Theplanissubsequentlypreparedfortreatment(treatmentscheduling,imagescheduling,etc)–afterseveralcomputercrashes.ItisalsoapprovedbyaphysicianAccordingtoQAprogramme,asecondphysicistshouldthenhavereviewedtheplan,includinganoverviewoftheirradiatedarea