美國生產(chǎn)過程工藝驗(yàn)證總則指南中英文版

{生產(chǎn)管理知識(shí)}美國生產(chǎn)過程工藝驗(yàn)證總則指南中英文版GUIDELINEONGENERALPRINCIPLESOFPROCESSVALIDATIONMay,1987Preparedby:CenterforDrugsandBiologicsandCenterforDevicesandRadiologicalHealthFoodandDrugAdministrationMaintainedby:DivisionofManufacturingandProductQuality(HFN-320)OfficeofComplianceCenterforDrugsandBiologicsFoodandDrugAdministration5600FishersLaneRockville,Maryland20857GeneralPrinciplesofProcessValidationMay1987GENERALPRINCIPLESOFPROCESSVALIDATIONI.PURPOSEThisguidelineoutlinesgeneralprinciplesthatFDAconsiderstobeacceptableelementsofprocessvalidationforthepreparationofhumanandanimaldrugproductsandmedicaldevices.II.SCOPEThisguidelineisissuedunderSection10.90(21CFR10.90)andisapplicabletothemanufactureofpharmaceuticalsandmedicaldevices.ItstatesprinciplesandpracticesofgeneralapplicabilitythatarenotlegalrequirementsbutareacceptabletotheFDA.ApersonmayrelyuponthisguidelinewiththeassuranceofitsacceptabilitytoFDA,ormayfollowdifferentprocedures.Whendifferentproceduresareused,apersonmay,butisnotrequiredto,discussthematterinadvancewithFDAtopreventtheexpenditureofmoneyandeffortonactivitiesthatmaylaterbedeterminedtobeunacceptable.Inshort,thisguidelinelistsprinciplesandpracticeswhichareacceptabletotheFDAfortheprocessvalidationofdrugproductsandmedicaldevices;itdoesnotlisttheprinciplesandpracticesthatmust,inallinstances,beusedtoplywithlaw.-1-Thisguidelinemaybeamendedfromtimetotime.Interestedpersonsareinvitedtosubmitmentsonthisdocumentandanysubsequentrevisions.WrittenmentsshouldbesubmittedtotheDocketsManagementBranch(HFA-305),FoodandDrugAdministration,Room4-62,5600FishersLane,Rockville,Maryland20857.Receivedmentsmaybeseeninthatofficebetween9\a.m.and4\p.m.,MondaythroughFriday.III.INTRODUCTIONProcessvalidationisarequirementoftheCurrentGoodManufacturingPracticesRegulationsforFinishedPharmaceuticals,21CFRParts210and211,andoftheGoodManufacturingPracticeRegulationsforMedicalDevices,21CFRPart820,andtherefore,isapplicabletothemanufactureofpharamaceuticalsandmedicaldevices.SeveralfirmshaveaskedFDAforspecificguidanceonwhatFDAexpectsfirmstodotoassurepliancewiththerequirementsforprocessvalidation.ThisguidelinediscussesprocessvalidationelementsandconceptsthatareconsideredbyFDAasacceptablepartsofavalidationprogram.Theconstituentsofvalidationpresentedinthisdocumentarenotintendedtobeall-inclusive.FDArecognizesthat,becauseofthegreatvarietyofmedicalproducts(drugproductsandmedicaldevices),processesand-2-manufacturingfacilities,itisnotpossibletostateinonedocumentallofthespecificvalidationelementsthatareapplicable.Severalbroadconcepts,however,havegeneralapplicabilitywhichmanufacturerscanusesuccessfullyasaguideinvalidatingamanufacturingprocess.Althoughtheparticularrequirementsofprocessvalidationwillvaryaccordingtosuchfactorsasthenatureofthemedicalproduct(e.g.,sterilevsnon-sterile)andtheplexityoftheprocess,thebroadconceptsstatedinthisdocumenthavegeneralapplicabilityandprovideanacceptableframeworkforbuildingaprehensiveapproachtoprocessvalidation.DefinitionsInstallationqualification-Establishingconfidencethatprocessequipmentandancillarysystemsarecapableofconsistentlyoperatingwithinestablishedlimitsandtolerances.Processperformancequalification-Establishingconfidencethattheprocessiseffectiveandreproducible.Productperformancequalification-Establishingconfidencethroughappropriatetestingthatthefinishedproductproducedbyaspecifiedprocessmeetsallreleaserequirementsforfunctionalityandsafety.-3-Prospectivevalidation-Validationconductedpriortothedistributionofeitheranewproduct,orproductmadeunderarevisedmanufacturingprocess,wheretherevisionsmayaffecttheproduct'scharacteristics.Retrospectivevalidation-Validationofaprocessforaproductalreadyindistributionbaseduponaccumulatedproduction,testingandcontroldata.Validation-Establishingdocumentedevidencewhichprovidesahighdegreeofassurancethataspecificprocesswillconsistentlyproduceaproductmeetingitspre-determinedspecificationsandqualityattributes.Validationprotocol-Awrittenplanstatinghowvalidationwillbeconducted,includingtestparameters,productcharacteristics,productionequipment,anddecisionpointsonwhatconstitutesacceptabletestresults.Worstcase-Asetofconditionsenpassingupperandlowerprocessinglimitsandcircumstances,includingthosewithinstandardoperatingprocedures,whichposethegreatestchanceofprocessorproductfailurewhenparedtoidealconditions.Suchconditionsdonotnecessarilyinduceproductorprocessfailure.-4-IV.GENERALCONCEPTSAssuranceofproductqualityisderivedfromcarefulattentiontoanumberoffactorsincludingselectionofqualitypartsandmaterials,adequateproductandprocessdesign,controloftheprocess,andin-processandend-producttesting.Duetotheplexityoftoday'smedicalproducts,routineend-producttestingaloneoftenisnotsufficienttoassureproductqualityforseveralreasons.Someend-producttestshavelimitedsensitivity.1Insomecases,destructivetestingwouldberequiredtoshowthatthemanufacturingprocesswasadequate,andinothersituationsend-producttestingdoesnotrevealallvariationsthatmayoccurintheproductthatmayimpactonsafetyandeffectiveness.2Thebasicprinciplesofqualityassurancehaveastheirgoaltheproductionofarticlesthatarefitfortheirintendeduse.These1Forexample,USPXXIstates:"Nosamplingplanforapplyingsterilityteststoaspecifiedproportionofdiscreteunitsselectedfromasterilizationloadiscapableofdemonstratingwithpleteassurancethatalloftheuntestedunitsareinfactsterile."2Asanexample,inoneinstanceavisualinspectionfailedtodetectadefectivestructuralweldwhichresultedinthefailureofaninfantwarmer.Thedefectcouldonlyhavebeendetectedbyusingdestructivetestingorexpensivetestequipment.-5-principlesmaybestatedasfollows:(1)quality,safety,andeffectivenessmustbedesignedandbuiltintotheproduct;(2)qualitycannotbeinspectedortestedintothefinishedproduct;and(3)eachstepofthemanufacturingprocessmustbecontrolledtomaximizetheprobabilitythatthefinishedproductmeetsallqualityanddesignspecifications.Processvalidationisakeyelementinassuringthatthesequalityassurancegoalsaremet.Itisthroughcarefuldesignandvalidationofboththeprocessandprocesscontrolsthatamanufacturercanestablishahighdegreeofconfidencethatallmanufacturedunitsfromsuccessivelotswillbeacceptable.Successfullyvalidatingaprocessmayreducethedependenceuponintensivein-processandfinishedproducttesting.Itshouldbenotedthatinmostallcases,end-producttestingplaysamajorroleinassuringthatqualityassurancegoalsaremet;i.e.,validationandend-producttestingarenotmutuallyexclusive.TheFDAdefinesprocessvalidationasfollows:Processvalidationisestablishingdocumentedevidencewhichprovidesahighdegreeofassurancethataspecificprocesswillconsistentlyproduceaproductmeetingitspre-determinedspecificationsandqualitycharacteristics.-6-Itisimportantthatthemanufacturerprepareawrittenvalidationprotocolwhichspecifiestheprocedures(andtests)tobeconductedandthedatatobecollected.Thepurposeforwhichdataarecollectedmustbeclear,thedatamustreflectfactsandbecollectedcarefullyandaccurately.Theprotocolshouldspecifyasufficientnumberofreplicateprocessrunstodemonstratereproducibilityandprovideanaccuratemeasureofvariabilityamongsuccessiveruns.Thetestconditionsfortheserunsshouldenpassupperandlowerprocessinglimitsandcircumstances,includingthosewithinstandardoperatingprocedures,whichposethegreatestchanceofprocessorproductfailureparedtoidealconditions;suchconditionshavebeewidelyknownas"worstcase"conditions.(Theyaresometimescalled"mostappropriatechallenge"conditions.)Validationdocumentationshouldincludeevidenceofthesuitabilityofmaterialsandtheperformanceandreliabilityofequipmentandsystems.Keyprocessvariablesshouldbemonitoredanddocumented.Analysisofthedatacollectedfrommonitoringwillestablishthevariabilityofprocessparametersforindividualrunsandwillestablishwhetherornottheequipmentandprocesscontrolsareadequatetoassurethatproductspecificationsaremet.-7-Finishedproductandin-processtestdatacanbeofvalueinprocessvalidation,particularlyinthosesituationswherequalityattributesandvariabilitiescanbereadilymeasured.Wherefinished(orin-process)testingcannotadequatelymeasurecertainattributes,processvalidationshouldbederivedprimarilyfromqualificationofeachsystemusedinproductionandfromconsiderationoftheinteractionofthevarioussystems.V.CGMPREGULATIONSFORFINISHEDPHARMACEUTICALSProcessvalidationisrequired,inbothgeneralandspecificterms,bytheCurrentGoodManufacturingPracticeRegulationsforFinishedPharmaceuticals,21CFRParts210and211.Examplesofsuchrequirementsarelistedbelowforinformationalpurposes,andarenotall-inclusive.Arequirementforprocessvalidationissetforthingeneraltermsinsection\211.100--Writtenprocedures;deviations--whichstates,inpart:"Thereshallbewrittenproceduresforproductionandprocesscontroldesignedtoassurethatthedrugproductshavetheidentity,strength,quality,andpuritytheypurportorarerepresentedtopossess."-8-SeveralsectionsoftheCGMPregulationsstatevalidationrequirementsinmorespecificterms.Excerptsfromsomeofthesesectionsare:Section211.110,Samplingandtestingofin-processmaterialsanddrugproducts.(a)"....controlproceduresshallbeestablishedtomonitortheoutputandVALIDATEtheperformanceofthosemanufacturingprocessesthatmayberesponsibleforcausingvariabilityinthecharacteristicsofin-processmaterialandthedrugproduct."(emphasisadded)Section211.113,ControlofMicrobiologicalContamination.(b)"Appropriatewrittenprocedures,designedtopreventmicrobiologicalcontaminationofdrugproductspurportingtobesterile,shallbeestablishedandfollowed.SuchproceduresshallincludeVALIDATIONofanysterilizationprocess."(emphasisadded)VI.GMPREGULATIONFORMEDICALDEVICESProcessvalidationisrequiredbythemedicaldeviceGMPRegulations,21CFRPart\820.Section820.5requireseveryfinisheddevicemanufacturerto:"...prepareandimplementaqualityassuranceprogramthatisappropriatetothespecificdevicemanufactured..."-9-Section820.3(n)definesqualityassuranceas:"...allactivitiesnecessarytoverifyconfidenceinthequalityoftheprocessusedtomanufactureafinisheddevice."Whenapplicabletoaspecificprocess,processvalidationisanessentialelementinestablishingconfidencethataprocesswillconsistentlyproduceaproductmeetingthedesignedqualitycharacteristics.Agenerallystatedrequirementforprocessvalidationiscontainedinsection\820.100:"Writtenmanufacturingspecificationsandprocessingproceduresshallbeestablished,implemented,andcontrolledtoassurethatthedeviceconformstoitsoriginaldesignoranyapprovedchangesinthatdesign."Validationisanessentialelementintheestablishmentandimplementationofaprocessprocedure,aswellasindeterminingwhatprocesscontrolsarerequiredinordertoassureconformancetospecifications.Section820.100(a)(1)states:"...controlmeasuresshallbeestablishedtoassurethatthedesignbasisforthedevice,ponentsandpackagingiscorrectlytranslatedintoapprovedspecifications."-10-Validationisanessentialcontrolforassuringthatthespecificationsforthedeviceandmanufacturingprocessareadequatetoproduceadevicethatwillconformtotheapproveddesigncharacteristics.VII.PRELIMINARYCONSIDERATIONSAmanufacturershouldevaluateallfactorsthataffectproductqualitywhendesigningandundertakingaprocessvalidationstudy.Thesefactorsmayvaryconsiderablyamongdifferentproductsandmanufacturingtechnologiesandcouldinclude,forexample,ponentspecifications,airandwaterhandlingsystems,environmentalcontrols,equipmentfunctions,andprocesscontroloperations.Nosingleapproachtoprocessvalidationwillbeappropriateandpleteinallcases;however,thefollowingqualityactivitiesshouldbeundertakeninmostsituations.Duringtheresearchanddevelopment(R&D)phase,thedesiredproductshouldbecarefullydefinedintermsofitscharacteristics,suchasphysical,chemical,electricaland-11-performancecharacteristics.3Itisimportanttotranslatetheproductcharacteristicsintospecificationsasabasisfordescriptionandcontroloftheproduct.Documentationofchangesmadeduringdevelopmentprovidetraceabilitywhichcanlaterbeusedtopinpointsolutionstofutureproblems.Theproduct'senduseshouldbeadeterminingfactorinthedevelopmentofproduct(andponent)characteristicsandspecifications.Allpertinentaspectsoftheproductwhichimpactonsafetyandeffectivenessshouldbeconsidered.Theseaspects3Forexample,inthecaseofapressedtablet,physicalcharacteristicswouldincludesize,weight,hardness,andfreedomfromdefects,suchascappingandsplitting.Chemicalcharacteristicswouldincludequantitativeformulation/potency;performancecharacteristicsmayincludebioavailability(reflectedbydisintegrationanddissolution).Inthecaseofbloodtubing,physicalattributeswouldincludeinternalandexternaldiameters,lengthandcolor.Chemicalcharacteristicswouldincluderawmaterialformulation.Mechanicalpropertieswouldincludehardnessandtensilestrength;performancecharacteristicswouldincludebiopatibilityanddurability.-12-includeperformance,reliabilityandstability.Acceptablerangesorlimitsshouldbeestablishedforeachcharacteristictosetupallowablevariations.4Theserangesshouldbeexpressedinreadilymeasurableterms.Thevalidityofacceptancespecificationsshouldbeverifiedthroughtestingandchallengeoftheproductonasoundscientificbasisduringtheinitialdevelopmentandproductionphase.Onceaspecificationisdemonstratedasacceptableitisimportantthatanychangestothespecificationbemadeinaccordancewithdocumentedchangecontrolprocedures.VIII.ELEMENTSOFPROCESSVALIDATIONA.ProspectiveValidationProspectivevalidationincludesthoseconsiderationsthatshouldbemadebeforeanentirelynewproductisintroducedbyafirmorwhenthereisachangeinthemanufacturingprocesswhichmayaffecttheproduct'scharacteristics,suchasuniformityandidentity.Thefollowingareconsideredaskeyelementsofprospectivevalidation.4Forexample,inordertoassurethatanoral,ophthalmic,orparenteralsolutionhasanacceptablepH,aspecificationmaybeestablishedbywhichalotisreleasedonlyifithasbeenshowntohaveapHwithinanarrowestablishedrange.Foradevice,aspecificationfortheelectricalresistanceofapacemakerleadwouldbeestablishedsothattheleadwouldbeacceptableonlyiftheresistancewaswithinaspecifiedrange.-13-1.EquipmentandProcessTheequipmentandprocess(es)shouldbedesignedand/orselectedsothatproductspecificationsareconsistentlyachieved.Thisshouldbedonewiththeparticipationofallappropriategroupsthatareconcernedwithassuringaqualityproduct,e.g.,engineeringdesign,productionoperations,andqualityassurancepersonnel.a.Equipment:InstallationQualificationInstallationqualificationstudiesestablishconfidencethattheprocessequipmentandancillarysystemsarecapableofconsistentlyoperatingwithinestablishedlimitsandtolerances.Afterprocessequipmentisdesignedorselected,itshouldbeevaluatedandtestedtoverifythatitiscapableofoperatingsatisfactorilywithintheoperatinglimitsrequiredbytheprocess.5Thisphaseofvalidationincludesexaminationofequipmentdesign;determinationofcalibration,maintenance,andadjustmentrequirements;andidentifyingcriticalequipmentfeaturesthatcouldaffecttheprocessandproduct.Informationobtainedfromthesestudiesshouldbeusedtoestablishwrittenprocedurescoveringequipmentcalibration,maintenance,monitoring,andcontrol.5Examplesofequipmentperformancecharacteristicswhichmaybemeasuredincludetemperatureandpressureofinjectionmoldingmachines,uniformityofspeedformixers,temperature,speedandpressureforpackagingmachines,andtemperatureandpressureofsterilizationchambers.-14-Inassessingthesuitabilityofagivenpieceofequipment,itisusuallyinsufficienttorelysolelyupontherepresentationsoftheequipmentsupplier,oruponexperienceinproducingsomeotherproduct.6Soundtheoreticalandpracticalengineeringprinciplesandconsiderationsareafirststepintheassessment.Itisimportantthatequipmentqualificationsimulateactualproductionconditions,includingthosewhichare"worstcase"situations.6TheimportanceofassessingequipmentsuitabilitybaseduponhowitwillbeusedtoattaindesiredproductattributesisillustratedinthecaseofdeionizersusedtoproducePurifiedWater,USP.Inonecase,afirmusedsuchwatertomakeatopicaldrugproductsolutionwhich,inviewofitsintendeduse,shouldhavebeenfreefromobjectionablemicroorganisms.However,theproductwasfoundtobecontaminatedwithapathogenicmicroorganism.Theapparentcauseoftheproblemwasfailuretoassesstheperformanceofthedeionizerfromamicrobiologicalstandpoint.Itisfairlywellrecognizedthatthedeionizersarepronetobuild-upofmicroorganisms--especiallyiftheflowratesarelowandthedeionizersarenotrechargedandsanitizedatsuitableintervals.Therefore,thesefactorsshouldhavebeenconsidered.Inthiscase,however,thefirmreliedupontherepresentationsoftheequipmentitself,namelythe"recharge"(i.e.,conductivity)indicator,tosignalthetimeforregenerationandcleaning.Consideringthedesiredproductcharacteristics,thefirmshouldhavedeterminedtheneedforsuchproceduresbaseduponpre-usetesting,takingintoaccountsuchfactorsasthelengthoftimetheequipmentcouldproducedeionizedwaterofacceptablequality,flowrate,temperature,rawwaterquality,frequencyofuse,andsurfaceareaofdeionizingresins.-15-Testsandchallengesshouldberepeatedasufficientnumberoftimestoassurereliableandmeaningfulresults.Allacceptancecriteriamustbemetduringthetestorchallenge.Ifanytestorchallengeshowsthattheequipmentdoesnotperformwithinitsspecifications,anevaluationshouldbeperformedtoidentifythecauseofthefailure.Correctionsshouldbemadeandadditionaltestrunsperformed,asneeded,toverifythattheequipmentperformswithinspecifications.Theobservedvariabilityoftheequipmentbetweenandwithinrunscanbeusedasabasisfordeterminingthetotalnumberoftrialsselectedforthesubsequentperformancequalificationstudiesoftheprocess.7Oncetheequipmentconfigurationandperformancecharacteristicsareestablishedandqualified,theyshouldbedocumented.Theinstallationqualificationshouldincludeareviewofpertinentmaintenanceprocedures,repairpartslists,andcalibrationmethodsforeachpieceofequipment.Theobjectiveistoassurethatallrepairscanbeperformedinsuchawaythatwillnotaffectthe7Forexample,theAAMIGuidelineforIndustrialEthyleneOxideSterilizationofMedicalDevicesapproved2December1981,states:"Theperformancequalificationshouldincludeaminimumof3successful,plannedqualificationruns,inwhichalloftheacceptancecriteriaaremet.....(.).-16-characteristicsofmaterialprocessedaftertherepair.Inaddition,specialpost-repaircleaningandcalibrationrequirementsshouldbedevelopedtopreventinadvertentmanufactureaofnon-conformingproduct.Planningduringthequalificationphasecanpreventconfusionduringemergencyrepairswhichcouldleadtouseofthewrongreplacementpart.b.Process:PerformanceQualificationThepurposeofperformancequalificationistoproviderigoroustestingtodemonstratetheeffectivenessandreproducibilityoftheprocess.Inenteringtheperformancequalificationphaseofvalidation,itisunderstoodthattheprocessspecificationshavebeenestablishedandessentiallyprovenacceptablethroughlaboratoryorothertrialmethodsandthattheequipmenthasbeenjudgedacceptableonthebasisofsuitableinstallationstudies.Eachprocessshouldbedefinedanddescribedwithsufficientspecificitysothatemployeesunderstandwhatisrequired.-17-Partsoftheprocesswhichmayvarysoastoaffectimportantproductqualityshouldbechallenged.8Inchallengingaprocesstoassessitsadequacy,itisimportantthatchallengeconditionssimulatethosethatwillbeencounteredduringactualproduction,including"worstcase"conditions.Thechallengesshouldberepeatedenoughtimestoassurethattheresultsaremeaningfulandconsistent.8Forexample,inelectroplatingthemetalcaseofanimplantablepacemaker,thesignificantprocessstepstodefine,describe,andchallengeincludeestablishmentandcontrolofcurrentdensityandtemperaturevaluesforassuringadequatepositionofelectrolyteandforassuringcleanlinessofthemetaltobeplated.Intheproductionofparenteralsolutionsbyasepticfilling,thesignificantasepticfillingprocessstepstodefineandchallengeshouldincludethesterilizationanddepyrogenationofcontainers/closures,sterilizationofsolutions,fillingequipmentandproductcontactsurfaces,andthefillingandclosingofcontainers.-18-Eachspecificmanufacturingprocessshouldbeappropriatelyqualifiedandvalidated.Thereisaninherentdangerinrelyingonwhatareperceivedtobesimilaritiesbetweenproducts,processes,andequipmentwithoutappropriatechallenge.9c.Product:PerformanceQualificationForpurposesofthisguideline,productperformancequalificationactivitiesapplyonlytomedicaldevices.Thesestepsshouldbeviewedaspre-productionqualityassuranceactivities.9Forexample,intheproductionofapressedtablet,afirmmayswitchfromonetypeofgranulationblendertoanotherwiththeerroneousassumptionthatbothtypeshavesimilarperformancecharacteristics,and,therefore,granulationmixingtimesandproceduresneednotbealtered.However,iftheblendersaresubstantiallydifferent,useofthenewblenderwithproceduresusedforthepreviousblendermayresultinagranulationwithpoorcontentuniformity.This,inturn,mayleadtotabletshavingsignificantlydifferingpotencies.Thissituationmaybeavertedifthequalityassurancesystemdetectstheequipmentchangeinthefirstplace,challengestheblenderperformance,precipitatesarevalidationoftheprocess,andinitiatesappropriatechanges.Inthisexample,revalidationprisesinstallationqualificationofthenewequipmentandperformancequalificationoftheprocessintendedforuseinthenewblender.-19-Beforereachingtheconclusionthataprocesshasbeensuccessfullyvalidated,itisnecessarytodemonstratethatthespecifiedprocesshasnotadverselyaffectedthefinishedproduct.Wherepossible,productperformancequalificationtestingshouldincludeperformancetestingunderconditionsthatsimulateactualuse.Productperformancequalificationtestingshouldbeconductedusingproductmanufacturedfromthesametypeofproductionequipment,methodsandproceduresthatwillbeusedforroutineproduction.Otherwise,thequalifiedproductmaynotberepresentativeofproductionunitsandcannotbeusedasevidencethatthemanufacturingprocesswillproduceaproductthatmeetsthepre-determinedspecificationsandqualityattributes.1010Forexample,amanufacturerofheartvalvesreceivedplaintsthatthevalve-supportstructurewasfracturingunderuse.Investigationbythemanufacturerrevealedthatallmaterialanddimensionalspecificationshadbeenmetbuttheproductionmachiningprocesscreatedmicroscopicscratchesonthevalvesupportingwireform.Thesescratchescausedmetalfatigueandsubsequentfracture.Comprehensivefatiguetestingofproductionunitsundersimulateduseconditionscouldhavedetectedtheprocessdeficiency.Inanotherexample,amanufacturerrecalledinsulinsyringesbecauseofplaintsthattheneedleswereclogged.Investigationrevealedthattheneedleswerecloggedbysiliconeoilwhichwasemployedasalubricantduringmanufacturing.Investigationfurtherrevealedthatthemethodusedtoextractthesiliconeoilwasonlypartiallyeffective.Althoughvisualinspectionofthesyringesseemedtosupportthatthecleaningmethodwaseffective,actualuseprovedotherwise.-20-Afteractualproductionunitshavesucessfullypassedproductperformancequalification,aformaltechnicalreviewshouldbeconductedandshouldinclude:oComparisonoftheapprovedproductspecificationsandtheactualqualifiedproduct.oDeterminationofthevalidityoftestmethodsusedtodeterminepliancewiththeapprovedspecifications.oDeterminationoftheadequacyofthespecificationchangecontrolprogram.2.SystemtoAssureTimelyRevalidationThereshouldbeaqualityassurancesysteminplacewhichrequiresrevalidationwhenevertherearechangesinpackaging,formulation,equipment,orprocesseswhichcouldimpactonproducteffectivenessorproductcharacteristics,andwhenevertherearechangesinproductcharacteristics.Furthermore,whenachangeismadeinrawmaterialsupplier,themanufacturershouldconsidersubtle,potentiallyadversedifferencesintherawmaterialcharacteristics.Adeterminationofadversedifferencesinrawmaterialindicatesaneedtorevalidatetheprocess.-21-Onewayofdetectingthekindofchangesthatshouldinitiaterevalidationistheuseoftestsandmethodsofanalysiswhicharecapableofmeasuringcharacteristicswhichmayvary.Suchtestsandmethodsusuallyyieldspecificresultswhichgobeyondthemerepass/failbasis,therebydetectingvariationswithinproductandprocessspecificationsandallowingdeterminationofwhetheraprocessisslippingoutofcontrol.Thequalityassuranceproceduresshouldestablishthecircumstancesunderwhichrevalidationisrequired.Thesemaybebaseduponequipment,process,andproductperformanceobservedduringtheinitialvalidationchallengestudies.Itisdesirabletodesignateindividualswhohavetheresponsibilitytoreviewproduct,process,equipmentandpersonnelchangestodetermineifandwhenrevalidationiswarranted.Theextentofrevalidationwilldependuponthenatureofthechangesandhowtheyimpactupondifferentaspectsofproductionthathadpreviouslybeenvalidated.Itmaynotbenecessarytorevalidateaprocessfromscratchmerelybecauseagivencircumstancehaschanged.However,itisimportanttocarefullyassessthenatureofthechangetodeterminepotentialrippleeffectsandwhatneedstobeconsideredaspartofrevalidation.-22-3.DocumentationItisessentialthatthevalidationprogramisdocumentedandthatthedocumentationisproperlymaintained.Approvalandreleaseoftheprocessforuseinroutinemanufacturingshouldbebaseduponareviewofallthevalidationdocumentation,includingdatafromtheequipmentqualification,processperformancequalification,andproduct/packagetestingtoensurepatibilitywiththeprocess.Forroutineproduction,itisimportanttoadequatelyrecordprocessdetails(e.g.,time,temperature,equipmentused)andtorecordanychangeswhichhaveoccurred.Amaintenancelogcanbeusefulinperformingfailureinvestigationsconcerningaspecificmanufacturinglot.Validationdata(alongwithspecifictestdata)mayalsodetermineexpectedvarianceinproductorequipmentcharacteristics.B.RetrospectiveProcessValidationInsomecasesaproductmayhavebeenonthemarketwithoutsufficientpremarketprocessvalidation.Inthesecases,itmaybepossibletovalidate,insomemeasure,theadequacyoftheprocessbyexaminationofaccumulatedtestdataontheproductandrecordsofthemanufacturingproceduresused.Retrospectivevalidationcanalsobeusefultoaugmentinitialpremarketprospectivevalidationfornewproductsorchangedprocesses.Insuchcases,preliminaryprospectivevalidation-23-shouldhavebeensufficienttowarrantproductmarketing.Asadditionaldataisgatheredonproductionlots,suchdatacanbeusedtobuildconfidenceintheadequacyoftheprocess.Conversely,suchdatamayindicateadecliningconfidenceintheprocessandamensurateneedforcorrectivechanges.Testdatamaybeusefulonlyifthemethodsandresultsareadequatelyspecific.Aswithprospectivevalidation,itmaybeinsufficienttoassesstheprocesssolelyonthebasisoflotbylotconformancetospecificationsiftestresultsaremerelyexpressedintermsofpass/fail.Specificresults,ontheotherhand,canbestatisticallyanalyzedandadeterminationcanbemadeofwhatvarianceindatacanbeexpected.Itisimportanttomaintainrecordswhichdescribetheoperatingcharacteristicsoftheprocess,e.g.,time,temperature,humidity,andequipmentsettings.11Whenevertestdataareusedtodemonstrateconformancetospecifications,itisimportantthatthetestmethodologybequalifiedtoassurethattestresultsareobjectiveandaccurate.11Forexample,sterilizertimeandtemperaturedatacollectedonrecordingequipmentfoundtobeaccurateandprecisecouldestablishthatprocessparametershadbeenreliablydeliveredtopreviouslyprocessedloads.Aretrospectivequalificationoftheequipmentcouldbeperformedtodemonstratethattherecordeddatarepresentedconditionsthatwereuniformthroughoutthechamberandthatproductloadconfigurations,personnelpractices,initialtemperature,andothervariableshadbeenadequatelycontrolledduringtheearlierruns.-24-IX.ACCEPTABILITYOFPRODUCTTESTINGInsomecases,adrugproductormedicaldevicemaybemanufacturedindividuallyoronaone-timebasis.Theconceptofprospectiveorretrospectivevalidationasitrelatestothosesituationsmayhavelimitedapplicability,anddataobtainedduringthemanufacturingandassemblyprocessmaybeusedinconjunctionwithproducttestingtodemonstratethattheinstantrunyieldedafinishedproductmeetingallofitsspecificationsandqualitycharacteristics.Suchevaluationofdataandproducttestingwouldbeexpec