ASTM E3106-18基于科学和风险的清洁工艺开发和验证的标准指南

StandardGuidefor

Science-BasedandRisk-BasedCleaningProcessDevelopmentandValidation1基于科学和风险的清洁工艺开发和验证的标准指南

ThisstandardisissuedunderthefixeddesignationE3106;thenumberimmediatelyfollowingthedesignationindicatestheyearoforiginaladoptionor,inthecaseofrevision,theyearoflastrevision.Anumberinparenthesesindicatestheyearoflastreapproval.Asuperscriptepsilon(´)indicatesaneditorialchangesincethelastrevisionorreapproval.1.Scope范围1.1Thisguideappliesthelife-cycleapproachtocleaningprocessvalidation,whichincludesthedevelopment,qualification,andverificationofcleaningprocesses.Itisapplicabletopharmaceuticals(includingactivepharmaceuticalingredients(APIs);dosageforms;andover-the-counter,veterinary,biologics,andclinicalsupplies)andisalsoappli-cabletootherhealth,cosmetics,andconsumerproducts.此指南将生命周期方法应用于清洁工艺验证（包括清洁工艺方法开发、确认、验证）。它将适用于多种类型药物（包括API、剂型、OTC、兽用药、生物药、临床用药），并且适用于其他保健品、化妆品和消费品。1.2Thisguideisfocusedonlyonthecleaningofequipmentproductcontactsurfacesanddoesnotcoverdisinfectionornon-productcontactsurfaces(whicharecoveredunderotherexistingguides:Ref(1),2USP<1072>,GuideE2614,andISO14698).此指南仅关注设备与产品接触面的清洁，不包括设备消毒或不与产品接触的表面（此部分指导见其他指南：USP<1072>,指南E2614,ISO14698）。1.3ThevaluesstatedinSIunitsaretoberegardedasstandard.Nootherunitsofmeasurementareincludedinthisstandard.本指南使用SI单位作为标准单位。除此之外指南内未使用其他计量单位。1.4Thisstandarddoesnotpurporttoaddressallofthesafetyconcerns,ifany,associatedwithitsuse.Itistheresponsibilityoftheuserofthisstandardtoestablishappro-priatesafety,health,andenvironmentalpracticesanddeter-minetheapplicabilityofregulatorylimitationspriortouse.本标准指南的目的并不是解决与使用相关的所有安全问题(如果有的话)。本标准的使用者有责任在使用前建立适当的安全、健康和环境规范，并确定法规的适用性。1.5Thisinternationalstandardwasdevelopedinaccor-dancewithinternationallyrecognizedprinciplesonstandard-izationestablishedintheDecisiononPrinciplesfortheDevelopmentofInternationalStandards,GuidesandRecom-mendationsissuedbytheWorldTradeOrganizationTechnicalBarrierstoTrade(TBT)Committee.2.ReferencedDocuments参考文献2.1ASTMStandards:3ASTM标准E1325TerminologyRelatingtoDesignofExperimentsE2476GuideforRiskAssessmentandRiskControlasitImpactstheDesign,Development,andOperationofPATProcessesforPharmaceuticalManufactureE2614GuideforEvaluationofCleanroomDisinfectants2.2ICHStandards:4

Q8PharmaceuticalDevelopmentQ9QualityRiskManagementQ10PharmaceuticalQualitySystemQ11DevelopmentandManufactureofDrugSubstances2.3ISOStandards:5

ISO9000QualityManagementSystems—FundamentalsandVocabularyISO14698GuideforEvaluationofCleanroomDisinfectants,Parts1–3.2.4FederalStandards:6

21CFR211.67EquipmentCleaningandMaintenance2.5USPStandards:7

USP<1072>DisinfectantsandAntiseptics3.Terminology术语3.1Definitions定义:3.1.1acceptabledailyexposure,ADE,n—representsadosethatisunlikelytocauseanadverseeffectifanindividualisexposed,byanyroute,atorbelowthisdoseeverydayforalifetime.可接受的日暴露量，ADE，n—在个体一生中的每天通过各种途径暴露于低于或等于某毒性药物一个剂量下，不太可能引起个体副作用的剂量。3.1.1.1Discussion—ThisisthetermusedintheISPERisk-MaPPGuide(1)andisequivalenttotheacceptabledailyintake(ADI)butisassociatedwithanyrouteofadministration.讨论—此术语在ISPE基于风险的药品生产指南中也被使用，并且相当于指南中的可接受每日摄入量（ADI）。但是此术语的任何途径包括任何给药方式。3.1.2acceptabledailyintake,ADI,n—measureoftheamountofaspecificsubstance(originallyappliedforafoodadditive,lateralsoforaresidueofaveterinarydrugorpesticide)infoodordrinkingwaterthatcanbeingested(orally)onadailybasisoveralifetimewithoutan

appreciablehealthrisk.

可接受的每日摄入量，ADI—指食物或饮用水中某一特定物质（最初仅指食物添加剂，后来范围扩至兽药或杀虫剂残留）的量，该物质（口服）在人的一生中每天都能被消化，不会对健康造成明显的危害。Ref(2)3.1.2.1Discussion—Thistermismorecommonlyassociatedwithfoodandtheoralrouteofadministration.讨论—此术语常指的是食物和口服途径。3.1.3cleaningagent,n—achemicalormixtureofchemicalsfortheremovalofresidualmaterial(forexample,drugsubstance,drugproduct,machiningoil,etc.)fromequipmentsurfacesorothercriticalobjects(suchasamedicaldevice).清洁剂—清除设备表面或其他必须被清洁的物体（例如医学设备）残留的化学物质或化学混合物（例如，原料药，药物，机油等）。3.1.4clean-in-place,CIP,n—methodofcleaningwithoutdismantlingequipment.原地清洗，CIP，n—不拆除设备的清洁方式3.1.5cleanability,n—relativedifficultyforcleaningapieceofequipmentorproduct.洁净度，n—清洁设备或产品的相对困难度3.1.6cleaningcontrolstrategy,n—plannedsetofcontrolsderivedfromtheriskassessmentandcurrentcleaningprocessunderstandingthatensuresreliableandconsistentcleaningprocessperformance.的可靠和性能一致3.1.6.1Discussion—Thecontrolscanincludeparametersandattributesrelatedtomaterialsandtoolsusedforcleaning,cleaningprocedure(s),equipmentoperatingconditions,andtheassociatedsamplingplans,methodsforvalidation,androutinemonitoring.讨论—控制措施可以包括关于清洁、清洁程序、设备操作情况、相关取样计划、验证方法、监控路线的参数和特性。3.1.7cleaningdesignspace,n—multidimensionalcombina-tionandinteractionofcleaninginputvariables(forexample,productcleanability,equipmentdesign,andsoforth)andcleaningprocessparameters(forexample,solvent/cleaningagentconcentration,temperature,time,andsoforth)thathavebeendemonstratedtoprovideassuranceofachievingaccept-ablecleaningoutputs(forexample,activepharmaceuticalingredients(API)residues,cleaningagentresidues).ICHQ8清洁设计空间，n—组合和相互作用的清洁变量（例如，清洁剂浓度，温度，时间，等等）已被证明可以保证达到可接受的清洁结果（例如，API残留、清洁剂残留）。cleaninginputvariables(parameters),n—thosefactorsorsettingswhosevaluesconstitutethecleaningprocessandaffectthecleaningoutputvariables.3.1.7.1Discussion—Theseindependentvariablesincludeproductcleanability,equipmentsize/groups,processresidueload,holdingtimes,cleaningagentconcentration,cleaningagenttype,rinsevolume,pH,time,temperature,velocity,pressure,surfacecoverage,locationandcleaningcycle,andsoforth.3.1.8cleaningoutputattributes,n—theseattributesincludeproductandcleaningagentresiduesremainingontheequip-mentsurfacesaftercleaning.3.1.8.1Discussion—Bioburden/endotoxinlevelsandopera-tionalconsiderationssuchastotalcleaningtime,holdingtimesandcostsmayalsobecleaningoutputattributes.3.1.9cleaningprocess,n—anyprocessdesignedtoremoveprocessresiduesfromproductcontactsurfacesofmanufactur-ingequipmenttolevelsthatensurepatientsafetyandproductquality.3.1.10cleaningprocessparameters,n—cleaningagentconcentration,temperature,time,andsoforth.3.1.11cleaningvalidation,n—collectionandevaluationofdata,fromthecleaningprocessdesignstagethroughcleaningatcommercialscale,whichestablishesscientificevidencethatacleaningprocessiscapableofconsistentlydeliveringcleanequipment.Ref(3)3.1.12cleaningverification,n—confirmation,throughtheprovisionofobjectiveevidence,thatspecifiedcleaningre-quirementshavebeenfulfilled.ISO90000ICHQ10清洁控制措施，n—根据风险评估和当前清洁工艺理解制定的控制计划，确保清洁工艺3.1.13clean-out-of-place(COP)system,n—automatedsystemusuallyusedtocleanlargepiecesofequipmentorpartsofequipmentthataredisassembled,buttoolargetocleanmanually.3.1.13.1Discussion—COPsystemscanrangefromelabo-ratewashingcabinetswithautomaticcontrolsystemstosimpledishwashertypeunits.3.1.14coupon,n—representativesurfacethatistypicallyarectangularpieceofamaterialofconstructioninwhichaknownamountofacompoundisdepositedtosimulateaprocessresidue.3.1.15designspace,n—multidimensionalcombinationandinteractionofinputvariables(forexample,materialattributes)andprocessparametersthathavebeendemonstratedtoprovideassuranceofquality.ICHQ83.1.16exposure,n—processbywhichahumanoranimalcancomeintocontactwithahazard.3.1.16.1Discussion—Exposuremayoccurthroughanyroute(oral,inhalational,dermal,andsoforth).Exposuremaybeshort-term(acuteexposure),ofintermediateduration,orlong-term(chronicexposure).3.1.17groupingstrategy,n—strategyofusinggroupsofproductsorequipmenttosimplifycleaningvalidation.3.1.17.1Discussion—Productsorequipmentorbothareplacedintogroupsandoneormorerepresentativesfromthegrouparechosenforcleaningprocessperformancestudies.Agroupingstrategyshallbescientificallyjustified.3.1.18manualcleaning,v—cleaningofequipment,eitherinplaceoroutofplace,byhandandwiththeaidofbrushes,cloths,detergents,andsoforth.3.1.19marginofsafety,n—differencebetweenthecleaningacceptancelimit(basedonADE)andtheprocessresiduedata.3.1.19.1Discussion—Thisvaluecanbeusedasameasureoftheoverallrisktopatientsafetypresentedbythecleaningprocess.Themarginofsafetycanbemeasuredanumberofwaysincludingtheprocesscapabilityindex(Cpk)andtheprocessperformanceindex(Ppk).3.1.20maximumallowablecarryover,MACorMACO,n—maximumamountofcarryoverfromoneproducttothenext.3.1.20.1Discussion—TheMACiscalculatedasafractionofthelowesttherapeuticdose(usually1/1000)orasafractionofalethaldose(LD50)(usually1/100000or1/1000000).3.1.21maximumsafecarryover,MSC,n—maximumamountofcarryoverofaresidualprocessresidue(API,cleaningagent,degradant,andsoforth)intothenextproductmanufacturedwithoutpresentinganappreciablehealthrisktopatients.Discussion—TheMSCiscalculatedfromtheADEandthetotalnumberofdosesinasubsequentbatch.3.1.22permitteddailyexposure,PDE,n—representsasubstance-specificdosethatisunlikelytocauseanadverseeffectifanindividualisexposedatorbelowthisdoseeverydayforalifetime.3.1.22.1Discussion—ThisisthetermusedbytheEuropeanMedicinesAgency(EMA)andisequivalenttoboththeADEandADI.3.1.23probability,n—likelihoodofoccurrenceofharm.3.1.24cleaningprocessresidue,n—anyresidue,including,butnotlimitedto,APIs,cleaningagents,degradationproducts,intermediates,excipients,andmicrobesremainingafteracleaningprocess.3.1.25qualifiedstatistician,n—individualwithaworkingknowledgeandeducation,training,orbackgroundinstatisticswhocanapplystatisticalanalysistodatafromcleaningandcleaningvalidationstudies.3.1.26qualifiedtoxicologist/pharmacologist,n—individualwithspecificeducationandtrainingintoxicology/pharmacologythatcanapplytheprinciplesoftoxicologytoderivinganADEorPDEvalueforrequiredprocessresidues.3.1.27qualitybydesign,n—systematicapproachtodevel-opmentthatbeginswithpredefinedobjectivesandemphasizesproductandprocessunderstandingandprocesscontrolbasedonsoundscienceandqualityriskmanagement.ICHQ83.1.28representativesurface,n—surrogatesurfacethatmaybeactualprocessingequipmentorhascharacteristicssimilartothatofprocessingequipmentandisusedforspikingstudies.3.1.29visualinspection,n—processofusingthehumaneye,aloneorinconjunctionwithvariousaids,asthesensingmechanismfromwhichjudgmentsmaybemadeabouttheconditionofthesurfacetobeinspected.3.1.30visuallimitofdetection,n—lowestlevelofaprocessresidueonasurface(inµg/cm2orµg/in.2)thatisvisibletoaqualifiedinspectorunderdefinedviewingconditions.3.2DefinitionsofTermsSpecifictoThisStandard:3.2.1CIPsystem,n—inthisstandard,CIPsystemsincludethemanufacturingequipmentitself(mixtanks,transferpiping,andsoforth)aswellastheequipmentusedforcleaning(detergenttanks,rinsetanks,pumps,andsoforth).3.2.2cleaningfailuremodesandeffectsanalysis,FMEA,n—aproceduretoidentifyallpossiblefailuresofacleaningprocessorprocedurethatcouldresultinprocessresiduelevelsthatcouldputapatientatrisk,thetoxicityofthosecleaningprocessfailures,thelikelihoodofthosecleaningprocessfailuresleavingsignificantlevelsofprocessresidue,andtheprobabilitythatthefailureorprocessresidueswillgounde-tected.3.2.2.1Discussion—ThecleaningFMEAcanalsoidentifywaystominimizethefailures,decreasetheirlikelihood,andimprovetheirdetectability.ScaleshavebeendevelopedthatcanbespecificallyusedforcleaningFMEAsandtomeasuretheriskofcleaningfailures(4-8).4.SignificanceandUse意义和用途Applicationoftheapproachdescribedwithinthisguideappliesrisk-basedconceptsandprinciplesintroducedinICHQ9.AsstatedinICHQ9,thelevelofeffort,formalityanddocumentation

forcleaningshouldalsobecommensuratewiththelevelofrisk.

本指南所述方法的应用适用于ICHQ9中引入的基于风险的概念和原则。如ICHQ9所述，清洁程度、形式和文件应与风险程度相称。

4.1Applicationoftheapproachdescribedwithinthisguideappliesmanyofthescience-based,risk-based,andstatisticalconceptsandprinciplesintroducedintheFDA’sGuidanceforIndustryProcessValidation:GeneralPrinciplesandPractices(3).本指南中描述的方法的应用了许多FDA行业指南：工艺验证：通则及实践中的基于科学、基于风险和统计的概念和原则4.2Thisguidesupports,andisconsistentwith,elementsfromICHQ8,ICHQ9,ICHQ10,andICHQ11.本指南支持并与ICHQ8、ICHQ9、ICHQ10和ICHQ11中的原理一致。4.3KeyConcepts—Thisguideappliesthefollowingkeyconcepts:(1)qualityriskmanagement,(2)science-basedapproach,(3)statistics-basedapproach,(4)processunderstanding,and(5)continuedimprovementasdescribedintheICHQseries.关键概念—本指南应用以下关键概念：（1）质量风险管理；（2）基于科学的方法；（3）基于统计学的方法；（4）工艺的理解；（5）ICHQ系列指南中所述的持续改进5.Science-Based,Risk-Based,andStatistics-BasedCleaningProcessDevelopmentandValidation基于科学、风险、数据的清洁工艺开发和验证

5.1Science-basedapproachesshouldbeappliedthroughoutthecleaningprocessdevelopmentandvalidationprocess.基于科学的方法应应用于清洁工艺的整个开发和验证过程。5.2Qualityriskmanagementshouldbeappliedthroughoutthecleaningprocessdevelopmentandvalidationprocess.质量风险管理应贯穿于清洁工艺开发和验证过程。5.3Appropriatestatisticalanalysisshouldbeappliedthroughoutthecleaningprocessdevelopmentandvalidationprocess.相应的数据分析应贯穿清洁工艺开发和验证过程。6.RiskAssessment风险评估6.1UnderICHQ9,riskassessmentisbrokenintothreestages:riskidentification,riskanalysis,andriskevaluation.根据ICHQ9，风险评估被分为三个阶段：风险识别，风险分析，风险评估。6.2Riskcanbedefinedas:risk=f(probabilityofoccurrenceofharmandtheseverityofthatharm).风险可被定义为：风险=f（危害发生的可能性和危害的严重性）。6.3Forthepurposesofcleaning,riskcanbefurtherdefinedasafunctionoftheseverityofthehazardsofprocessresidues,likelihoodandlevelofprocessresidues,anddetectabilityofprocessresidues.基于清洁的目的，风险可以被进一步定义为工艺残留危害的严重性，等级，可能性，可检测性等参数间的函数。6.4Forareliableassessmentofrisk,scientificmeans(forexample,riskmanagementtools)shouldbeusedtoidentifythehazardpresentedbyaprocessresidue(forexample,API,degradationproducts,intermediates,cleaningagent,bioburden/endotoxin,andsoforth),theabilityofacleaningprocesstoremoveprocessresiduestolevelsthatareacceptable,andtheabilitytodetectandquantifythepresenceofprocessresiduesaftercleaning.对于一个可靠的风险评估来说，应在识别工艺残留（例如，API，产品降解产物，中间体，清洁剂，生物负载/内毒素等）危害、识别清洁过程将过程残留物降至可接受水平的能力、识别检测和量化清洁后过程残留物的能力等过程中使用科学的手段（例如，风险管理工具）。6.5RiskIdentification—Riskidentificationshouldencompasstheidentificationofprocessresiduehazards,equipmentdesignhazards,andproceduralhazards.风险识别—风险识别应当包括识别过程残留物危害、设备设计危害和程序危害。6.5.1ProcessResidueHazardIdentification识别过程残留物危害:6.5.1.1Thehazardpresentedbyapotentialprocessresiduemaybedeterminedfromatoxicologicalreviewperformedbyaqualifiedtoxicologistorqualifiedpharmacologist.ForanAPI,thisinvolvesathoroughreviewofallrelevanttoxicologicaldataavailablefortheprocessresidueunderstudy(9).WhenpreclinicalandclinicaldataonAPIsareavailabletoreview,anADEcanbedeterminedandusedasameasureoftheseverityofhazardpresentedbyacompound.Forfurtherinformation,seetheISPERisk-MaPPGuide(1)ortheEMAGuidelineonSettingHealthBasedExposureLimitsforUseinRiskIdentificationintheManufactureofDifferentMedicinalProductsinSharedFacilities(9).潜在的工艺残留物的危害可以由一个合格的毒理学家或合格的药理学家进行毒理学检查来确定。对于API来说，这个识别过程包括对工艺残留物相关毒理学数据进行一个全面评估。如果可获得API临床前和临床数据，ADE可以作为一种衡量化合物危害程度的指标来确定和使用。更详细的信息可见ISPERisk-MaPP指南或EMA指南：关于在共用设施中生产不同药品的风险识别中设置基于健康的暴露限值。6.5.1.2WhenanADEisnotavailable,suchasforintermediates,degradationproducts,orcompoundsinearlydevelopment,alternativeapproachessuchasthethresholdoftoxicologicalconcern(TTC)maybejustified(9,10).Althoughcompoundsinearlydevelopmentmaynothavesufficientsafetydatatoperformacompleteanalysis,usefulinformationcanbefoundinthechemicalstructureofacompoundtohelpdetermineaprovisionalADEforthecompound.“Insilico”(computer-assisted)toxicologicalassessmentorastructureactivityrelationshipcanbeusedtodetermineprovisionalADEsforacompound(11,12).Forexample,acompoundinthesamestructuralseriesofaknownAPIfromagiventherapeuticclasscanbetreatedinthesamewayasthatAPI,forexample,acompoundwithapropylaminestructurewouldbeexpectedtosharepropertiesofthisclassofantihistamines(13).Wheredataareavailableoncomparativepotency,thesecanbeusedtoadjusttheestimatedADE.如果ADE值不可获得，可使用例如中间体、产品分解物、或在合成早期的化合物、毒理学关注阈值(TTC)等其他方法。尽管合成早期的化合物可能没有足够的安全性数据来进行一个彻底的分析，但是可以在化合物的化学结构中获取有用的信息，来暂定一个ADE值。“Insilico”(计算机辅助)毒理学评估或结构活性关系可用于确定化合物的临时ADE值。例如，具有与已知疗效的化合物有相似结构的API，可以用相同的方法来计算。例如，一种具有丙胺结构的化合物有望具有这类抗组胺药物的共同特性。如果有关于比较效力的数据，可用于调整估计的ADE。6.5.1.3Thehazardofpossiblebioburdenfromapreviousproductandthepossibilityofmicrobialproliferationafteracleaningprocessandthehazardsthispresents,includingtheneedforsubsequentdisinfection,shouldalsobeconsidered.Forexample,thehazard(s)presentedbyholdingequipmenteitherinadirtystateorincleanstateshouldbeconsideredorthepossibilityofendotoxinandtheneedforsubsequentdepyrogenationshouldbeconsidered.前一种产品可能产生的生物污染的危害以及清洁过程后微生物增殖的可能性，包括接下来消毒工序的要求，也应当进行考虑。例如，应考虑保持设备处于脏状态或清洁状态所带来的危害，应考虑内毒素的可能性和随后脱硫的必要性。6.5.2EquipmentHazardIdentification—Thepotentialhazardspresentedbyequipmentdesignshouldalsobeconsidered,suchasthepossibilityofproductbuildup.Equipmentshouldbedesignedtofacilitatecleaning,inspection,andmonitoring.设备危害识别—：还应考虑设备设计所带来的潜在危害，例如产品积聚可能性。设备应当被设计成便于清洁、检查、监控的。6.5.3ProceduralHazardIdentification—Beforeuse,cleaningproceduresshouldbesubjectedtoriskassessments,forexample,cleaningFMEAorotherriskmanagementtools,tominimizeriskoffailure(forexample,toensurethatproductbuildupisavoided),,improvethecleaningprocedures,andmakethecleaningproceduresmorereliableandrobust.程序危害识别—在使用之前，清洁程序应当进行风险评估，例如清洁程序FMEA或者其他风险管理工具，来使失败风险降至最低（例如，保证避免产品积聚），改善清洁程序，并使清洁程序更可靠和稳健。6.6RiskAnalysis风险分析:6.6.1Afteridentifyingthehazardsposed,therisksassociatedwiththemshouldbeanalyzed.Thisriskanalysisshouldinvolvethecleaningprocessdevelopment,facility/equipmentdesignreview,cleaningprocedurereview,andtheselectionofanalyticalmethods.Theanalysisshouldalsodeterminewhatstepscanbetakentomitigatetheidentifiedrisks.识别出可能引起的危害后，应当分析与之相关的风险。此风险分析应当包括完善清洁工艺，审核设备设计，清洁程序审查，以及选择分析方法。风险分析还应该确定可以采取哪些步骤来减轻确定的风险。6.6.2Theriskanalysisshouldshowhowcleaningmayaffectthepatientsafetyandqualityofthenextproduct.风险分析应当表明清洁如何影响病患安全和下一生产产品的质量。6.6.3Theimpactofthedifferentfactors(processresidue,cleaning/rinsingagents,equipmentengineering,andsoforth)ontheoutcomeofthecleaningprocessshouldbeanalyzed.应当分析清洁工艺中不同因素的影响（工艺残留，清洁剂，设备设计等等）。6.6.4Thecleaningprocessriskanalysiscanhelptodeterminethenecessarycleaningqualificationsandidentifyappropriateriskcontrolmechanisms.清洁工艺分析能够帮助确定清洁确认步骤以及识别适当的风险控制机制。6.6.5ProcessResidueCharacterization工艺残留特性:6.6.5.1Thechemistryofprocessresiduesshouldbeunderstoodtodesignaneffectiveandefficientcleaningcycle,forexample,cleanabilityofprocessresidues(forexample,highlyinsolubleorstronglyadhesiveresidues)andpotentialinteractions(forexample,staining,corrosion)ofprocessresidueswithequipment.需要基于工艺残留物的化学特性来设计一个有效和高效的清洁循环，例如工艺残留洁净度（例如，不溶或强粘附性残留）以及工艺残留物与设备之间可能出现的反应（例如，染色、腐蚀）。6.6.5.2Thechemistryandpotentialinteractionsbetweenprocessresiduesandchemicalsusedaspartofcleaningprocessesshouldalsobeunderstood,forexample,thesolubilityofprocessresiduesincleaningagentsorrinsingagentsshouldbeconsideredtoavoidsituationsinwhichprocessresiduesarenotremovedorwhetherdegradationproductsmaybeformedthatmaybehardertocleanormoretoxicthantheoriginalprocessresidue.还应了解工艺残留物和清洁过程中可能出现的化学和潜在相互作用，例如，应考虑工艺残留物在清洗剂或漂洗剂中的溶解度，以避免出现这种情况：工艺残留物未被去除或形成比原工艺残留更难清洁或毒性更大的降解产物。6.6.6EquipmentDesignforCleanability洁净性设备设计:6.6.6.1Thedesignofequipmenthasanimpactonitscleanability.Equipmentdesignshouldbeconsideredaspartoftheriskanalysistakingintoconsiderationthelikelytypeofcleaningprocessthatwillbeappliedtothatequipment.Thevariablesandattributesrelatedtoequipmentdesignshouldbeidentifiedandlinkedtothecleaningcriticalattributesusingtheappropriateriskassessmenttool(s).Examplesofequipmentdesignconsiderationsmayinclude:materialsofconstruction,presenceofdeadlegsorotherareasinwhichmaterialcouldbecometrapped,ordrainability.设备的设计会影响设备的清洁度。设备设计应当被当做风险评估的一部分，考虑可能应用于该设备的清洗过程的类型。与设备设计相关的变量和属性应加以识别，并使用适当的风险评估工具与关键的清洁属性相关联。设备设计考虑因素的例子可包括:建筑材料、存在死角或其他材料可能被困的区域，或排水能力。6.6.6.2Wheresatisfactorycleaningresultscannotbeachievedbecauseoflimitationsintheequipmentdesign,theequipmentmayneedtobemodified,dedicated,orreplaced.由于设备设计的，无法达到满意的清洗效果时，可能需要对设备进行修改、专用或更换。6.6.7EvaluationofHistoricalCleaningData—Thehistoryofcleanings(alongwithanydeviations,investigations,andcorrectiveactions)shouldbereviewed.Thiscleaningprocessunderstandingandknowledgecanprovideusefulinformationinariskanalysisandmayhelpidentifycleaningprocessparameterstobeusedincleaningprocessdevelopmentstudiesanddeterminethelikelihoodofacleaningfailure(ICHQ10).对历史清洁数据进行评估—应回顾清洁历史。（以及任何偏差、调查和纠正措施）。清洁工艺的信息可以为风险评估提供有用的信息，并可以帮助确定清洁工艺开发研究中使用的清洁工艺参数，并确定清洁失败的可能性6.6.8LevelsofCleaning清洁等级—Manufacturingequipmentmayrequiredifferentlevelsofcleaningandvalidationunderdifferentcircumstances.在不同的情况下，生产设备可能需要不同级别的清洁和验证。Todeterminetheappropriatecleaninglevel,thetypeofproductmanufacturedontheequipment(forexample,intermediates,APIs,finishedproducts)shouldbeconsideredandtheriskstopatientsafetyandproductqualityshouldbeunderstood.Acleaningprocesscanthenbedevelopedtoachievethenecessaryresults.Theremaybeseverallevelsofcleaningbasedonthelevelofrisk,forexample:为了确定相应的清洁等级，应考虑在设备上生产的产品类型(例如，中间体、原料药、成品)，应了解对患者安全和产品质量的风险。然后可以开发一个清洁工艺来实现必要的结果。根据风险级别，可能有几个级别的清洁，例如:6.6.8.1Cleaningbetweendifferentproducts,不同产品之间的清洁6.6.8.2Cleaningbetweensimilarproducts,相似产品清洁6.6.8.3Cleaningduringcampaigning(cleaningbetweenbatchesofthesameproduct),生产过程中清洁（同一产品之间换批的清洁）6.6.8.4Cleaningofdedicatedequipment,专用设备清洁6.6.8.5Cleaningafterequipmentmaintenance,设备维修清洁6.6.8.6Cleaningafterelapseofpermissiblestorage/holdtimeofcleanequipment,清洁设备超过允许储存/保持时间后的清洁6.6.8.7Cleaningaftersampling,and取样后的清洁6.6.8.8Cleaningafternon-routineoperations非常规操作后的清洁CleaningProcessDevelopment—Cleaningprocessesshouldnotbeadoptedrandomlyorchosenbasedonpastuse.Cleaningprocessesshouldbedevelopedtoreduceprocessresidueslevelsaslowaspracticalanddeterminetheappropriatecleaningagentsforthispurpose.Cleaningprocessesthathavebeenoptimizedthroughtheselectionofthemostappropriatecleaningagentsandcleaningparameterscanofferthegreatestabilitytoreduceprocessresiduesintheshortesttimetothelowestlevelofrisk.Theoutputofthecleaningprocessdevelopmentshouldbeusedtocreatethecleaningstandardoperatingprocedure(SOP).清洁工艺研究—不应随意使用清洁工艺或基于历史选择。清洁工艺应当基于将工艺残留物等级尽量降至最低以及决定使用相应清洁剂。经优化的清洁工艺可以保证能最大限度地在最短的时间内将工艺残留物降低到最低的风险水平。清洁工艺研究结果应被用于创建清洁SOP。6.6.8.9Bench-ScaleStudies实验室研究:(1)Laboratoryscaleor“bench-scale”studiescanprovidevaluablesourcesofcleaningprocessknowledgeandcleaningprocessunderstanding(14).Thestudiesmaybeconductedbyspikingtheprocessresidue(s)oncouponsandthensubjectingthecoupons(afterdrying)tovaryingcleaningconditions.Thestudiescouldalsobeconductedinsmall-scaleequipmentdesignedtosimulatetheactualmanufacturingequipment.实验室研究可以为清洁工艺知识和清洁工艺理解提供有用的信息资源。研究可通过将工艺残留物钉在试样上，然后(干燥后)将试样置于不同的清洁条件下进行。研究可以使用设计来模拟实际生产设备的小型仪器。(2)Bench-scalestudiescanbequick,economical,andprovideinformationonhowdifficultaproductistoclean,whichcleaningagentprovidesoptimalcleaning,whichcleaninginputvariablesarecritical,andwhetherdirtyholdtimestudiesmaybenecessary.Cleaningprocessknowledgeandcleaningprocessunderstandinggainedfrombenchscalestudiesmaybedirectlyapplicabletofull-scalecleaningprocessesbutdifferencesbetweenfullscaleandbenchscaleshouldbeconsidered.实验室研究速度快，经济，并能提供相关信息，关于：产品清洁困难程度，使用哪种清洁剂有最佳清洗效果，关键清洁变量，是否有必要对脏滞留时间（DirtyHoldTime）进行研究等。实验室研究中得到的清洁工艺知识和清洁工艺理解可以直接应用于实际清洁，但应考虑实际清洁与实验室模拟情况的差异。6.6.8.10CleaningParameterDetermination—Theeffectsandtheinteractionsofinputvariablesaffectingcleaningshouldbeevaluated.Thevariablestypicallyassociatedwithcleaningare:time,temperature,cleaningagentchemistry,mechanicalaction,productcleanability,andamountofprocessresidue.清洁参数测定—应评估影响清洁的输入变量的影响和相互作用。与清洁工艺相关的变量有：时间、温度、清洁剂化学特性、作用机制、产品清洁度、工艺残留量。6.6.8.11DesignofExperiments(DoE)and“CleaningDesignSpace:”实验设计和清洁设计(1)Toimproveoroptimizecleaningprocesses,experimentscanbedesignedtoexaminetheeffectsofcleaninginputparametersoncleaningoutputvariables.TheseinputscanbeassignedasfactorsinaDoE(TerminologyE1325)andtheeffectsandinteractionsofvaryingthesefactorsontheoutputscanbemeasuredasresponses.为了改善或优化清洁工艺，可以设计实验来测试使用不同变量后的结果效应。这些因子可以在DoE中作为因子，改变这些因素对输出的影响和相互作用可以作为响应来衡量。(2)Typicalcleaninginputparametersincludeproductcleanability,equipmentsize/groups,processresidueconcentration,holdingtimes,cleaningagenttypeandconcentration,wash/rinsetemperature/time/volume,flow,pressure,andsprayballtype/location.典型清洁参数包括产品清洁度、设备大小、工艺残留浓度、滞留时间、清洁剂类型和浓度，清洗/冲洗温度/时间/体积、水流、压力、和喷淋球类型/位置。(3)Thetypicalcleaningoutputvariablesaretheproductandcleaningagentresidues.Bioburdenlevelsandoperationalconsiderationssuchascleaningtimes,holdingtimes,andcostsmayalsobeconsidered.典型的清洁输出变量是产品和清洁剂残留物。生物负载水平，也可以考虑诸如清洗时间、滞留时间和成本等操作方面的考虑。(4)DoEareusedfordetermininga“cleaningdesignspace”8thatprovidesmanybenefitsincludingjustificationofproductorequipmentgroupingandprocesschangecontrolstrategies.Ifchangestothecleaningprocessorequipmentareconsidered,theresultsofariskreviewcanprovideinformationregardingtheimpactoncleaningdesignspaceandtheneedforadditionalstudiesortesting.DoE用于确定“清洁设计空间”，它提供了许多好处，包括产品或设备分组的合理性和过程变更控制策略。如果考虑清洁工艺或设备的变更，风险回顾的结果可以提供关于清洁设计空间和其他研究或测试的信息。6.6.96.6.9.1HoldTimeStudies滞留时间研究:Therearegenerallytwotypesofholdtimestudies:dirtyholdtimesandcleanholdtimes.Themaximumtimeintervalbetweenequipmentuseandcleaningisknownasthedirtyholdtime(DHT).Themaximumtimeintervalbetweenthelaststepofthecleaningproceduretothestartofnextequipmentuseformanufacturingisknownascleanholdtime(CHT).Expiredequipmentholdtime(EEHT)occurswhenthemaximumtimeintervalintheDHTorCHThasbeenexceeded.总共有两种类型的滞留时间：脏滞留时间和清洁保持时间。脏滞留时间（DHT）为清洁后到设备使用的最大时间间隔。从清洗程序的最后一步开始到下一次设备投入生产使用之间的最大时间间隔称为清洁保持时间（CHT）。过期设备保持时间(EEHT)发生在超过DHT或CHT最大时间间隔。6.6.9.2Acleaningriskassessmentcanbeperformedtodeterminethepotentialimpactoreffectofthesetimeperiodsontheabilitytocleantheequipmentormaintainitscleanliness.Thecleaningriskassessmentcanbeusedtojustifythecriticalityleveloftheholdtimeperiodandnumberofcleaningexercisestosupportitsqualification.Iftheconclusionofthecleaningriskassessmentisthattheholdtimeperioddoesnothaveanimpactonthecleaningprocess,thenthisholdtimeperiodshouldnotneedtobequalified.清洁风险评估可用来识别这些时间周期对设备清洁或维持清洁度的潜在影响或效应。清洁风险评估可以用来证明保持周期关键等级和清洁次数。如果清洁风险评估的结论是保持时间周期对清洁工艺没有影响，那么此保持时间周期则不需要被确认。6.6.10CleaningControlStrategy—Acleaningcontrolstrategyshouldbedevelopedbasedontheoveralloutcomeoftheriskassessment,whichincludescleaningprocessknowledgeandunderstandingobtainedduringcleaningprocessdevelopmentandincludestheresultsofcleaningdatacollectedandevaluatedduringcleaningqualificationrunsintheriskevaluationstep.Thecleaningcontrolstrategyshouldberefinedandcontinuallyimprovedasnewknowledgeisgainedfromperiodicassessmentandmonitoringofcleaningprocesses.清洁控制措施—一个清洁控制措施应当基于风险评估后全面结果，包括在清洁工艺建立过程中获得的清洁工艺知识和理解，包括收集的清洁数据结果。在周期性评估以及清洁过程的监控过程出现新信息，那么应根据这些新信息对清洁控制措施进行持续改进。6.6.11OperatorTraining操作培训:6.6.11.1Thesuitabilityofinitialandongoingtrainingshouldbedeterminedbasedontheeducationandexperienceofoperatorsaswellasaperiodicassessmentofeffectivenessofthetrainingprogram.对操作者相应的初始以及持续培训和培训有效性的周期性评估应当基于操作者学历和经验决定。6.6.11.2Thelevelandfrequencyoftrainingshouldbebasedonriskandthetrainingprogramshouldbeappropriateforeachoperatortoensurethattheyperformtasksreliably.Assessmentoftrainingeffectivenessorqualificationshouldbeconsideredtohelpreducevariationintheperformanceofcleaningprocesses.培训的等级和频率应当基于风险和操作者的专门培训计划，来确认操作者能够正确进行操作。对于培训有效性的评估或确认应当考虑是否有助于减少清洁工艺性能的变化。6.6.12Sampling—Instatistics,thepurposeofsamplingistoselectarepresentativesubsetofindividualsfromapopulationthatcanreliablycharacterizethewholepopulation.Incleaning,thismeansselectingrepresentativesamplesfromcleanedequipmentthatcanreliablycharacterizethecleanlinessoftheequipment.取样—在统计学上，取样的目的是抽样的目的是从一个能可靠地描述整个集合特征的集合中选择一个具有代表性的个体子集。对于清洁，这个意味着从已清洁的设备选择具代表性的样本，能可靠地表征设备的洁净度。6.6.12.1SamplingStrategy—Theriskassessmentandpriorexperienceshouldbeusedtodeterminethesamplinglocations,numberofsamples,andsamplingmethodsforeachpieceofequipment.Samplinglocationsshouldconsidertheequipmentdesign,accessibility,andmaterialsofconstruction.Samplinglocationsshouldincludethosedeterminedtobemostdifficulttocleanandwherethereisalikelihoodofcontaminationorcarryovertothenextproduct.Samplingstrategiesshouldhaveastatisticalbasis.取样措施—应使用风险评估和以前的经验来确定每台设备的取样位置、取样数量和取样方法。取样位置应当考虑设备的设计、可达性、以及结构材料。取样位置应包括那些被确定为最难清洗的地方，以及有可能被污染或转移到下一个产品的地方。取样计划应当基于统计学。SamplingTechniques—Thecleaningriskassessmentshouldbeusedtodeterminetheselectionoftheappropriatesamplingtechnique.Theriskassociatedwiththeprocessresiduesandtheequipmentdesignandaccessibilitytotheareatobesampledwilldeterminethetypeofsamplingmethodneeded.Processresidueswithalowriskmaybecandidatesforvisualinspectiononly,whileprocessresidueswithahighriskmayrequireswabsamplingusingspecificanalyticalmethods.取样技术—清洁风险评估应当用来确定恰当的取样技术。与工艺残留物、设备设计和对取样区域的可及性有关的风险将决定所需的取样方法的类型。低风险的工艺残留物可以目视检查，：而高风险的过程残留物可能需要使用特定的分析方法进行擦拭取样。6.6.12.2VisualInspection—Themostbasictoolforinspectingequipmentforcleanlinessasinaccordancewith21CFR211.67(b)(6)hastypicallybeenvisualinspection,anditshouldbethefoundationtechniqueforallcleaningassessmentprograms,includingcleaningprocessdevelopment,andshouldbeapartofthevalidation.Ariskassessmentshouldbeperformedtodeterminethecriticalparametersforvisualinspection.Usingvisualinspectionaloneforvalidationmaybeacceptable(15)onlywhenaRiskAssessmenthasshownthattheriskislowand100%oftheequipmentsurfacecanbeinspectedunderappropriateviewingconditions.目视检查—根据《21CFR211.67(b)(6)》对设备清洁度进行检查的最基本手段通常是目视检查，并且它应该是所有清洁评估项目的基础技术，包括清洁过程的开发，并应作为验证的一部分。应进行风险评估以确定目视检查的关键参数。只有在以下情况下才可接受仅使用目视检验进行验证：风险评估表明风险低并且的设备所有表面可以在适当的观察条件下进行检查。6.6.12.3DirectSurfaceSampling:SwabSampling—Swabsamplingoffersdirectphysicalsamplingofresiduesfromsurfaces.Swabsamplingisusuallythetechniqueofchoicewhereequipmentsurfacesarereadilyaccessible.Pointstoconsiderationsmayinclude;swabtype,numberofswabs,swabbingpatternused,swabbingsolventsrequired,andanyspecialrequirementsforspecificareasorequipment.Theareatobeswabbedshouldbedefined.表面擦拭取样—擦拭取样是直接的表面残留物物理取样方法。擦拭取样通常选择在设备表面容易接触的地方。需考虑的点通常有：擦拭类型；擦拭次数；擦拭方法；擦拭溶剂；特殊区域或设备的特殊要求。应确定擦拭区域。6.6.12.4SurfaceScanningSampling—TechnologiessuchasFouriertransforminfrared(FTIR),nearinfrared(NIRS),Raman,fluorescenceandultraviolet(UV)spectroscopythatareusedinvariousin-line/at-lineapplicationsmayhaveapplicationtocleaningverification.Thesetechnologiesshouldbequalifiedfortheirintendeduse.扫描表面取样—FTIR、NIRS、拉曼、荧光和UV光谱等各种应用程序可用于清洁验证。这些技术应当具有符合使用目的的资质。6.6.12.5IndirectSurfaceSampling:RinseSampling—Rinsesamplingallowssamplingofresiduesfromagreatersurfaceareathanswabsampling.Rinsesamplingisusuallythetechniqueofchoicewhereequipmentsurfacesarenotreadilyaccessible,therearephysicallimitations,oritisundesirabletoopenorenteraclosedsystemforprocesssafetyoroccupationalexposureconsiderationsandwhenthedesignofequipmenthasbeencarefullyreviewedtoensurethatlocationsinequipmentthatcannotbecleanedhavebeeneliminatedfromthedesign.Therinsesolutionselectionshouldbebasedprimarilyonproductsolubility.Theuseofrinsesamplingshouldbejustifiedinariskassessment.间接表面取样：冲洗取样—冲洗取样比起擦拭取样来说可以有更大的取样面积。冲洗取样通常是在设备表面不易接触的情况下所选择的技术，可能有物理、或出于为了工艺安全的目的不去打开或进去一个密闭系统、职业暴露考量、当对设备的设计已经进行了仔细的审查，以确保无法清洗的设备位置已从设计中消除。漂洗液的选择应主要基于产品的溶解度。漂洗取样的使用应经由风险评估。6.6.12.6QualificationofSampling—Theabilityofsamplingtechniquestorecoverprocessresiduesofinterestshouldbedeterminedanddocumented.Thisqualificationtypicallyinvolvesarecoverystudy.Thesamplingtechniqueshouldbestandardizedandwelldefinedtoensureconsistencyandreproducibility.Trainingforthesamplingtechniqueisessentialtoobtainreliableresults.Statisticaltechniquescanbeusedtodeterminetheaccuracy,precision,androbustnessofthesamplingtechnique.取样验证—应确定和记录取样技术恢复目的工艺残留物的能力。此类验证应当包括恢复研究。取样技术应标准化和明确定义，以确保一致性和重现性。取样技术的培训对于获取可靠结果来说是关键的。统计工具可以用来确定抽样技术的准确性、精确性和稳健性。6.6.13SelectionofAnalyticalMethods—Thechoiceofmethodsshouldbescienceandriskbased.Thegoalshouldbetousethesimplesttechniquethatisappropriateandcanbejustified.Analyticalmethodsfallintotwogeneralcategories:specificandnon-specific.分析方法的选择—分析方法的选择应该是科学和基于风险的。目标是使用最简单的并且适合的可以被证明的技术。分析方法有两类：特异的和非特异的。SpecificAnalyticalMethods—Specificmethods,suchashighperformanceliquidchromatography(HPLC),arecapableofidentifyingandquantifyingspecificprocessresiduesinthepresenceofpotentialcontaminants.Specificmethodsshouldbeconsideredforhigh-hazardproductsorhigh-risksituations.特异的分析方法—特异方法，HPLC，是否能够识别和量化存在潜在污染物的特定工艺残留物。当是高危害或高风险产品的情况下，应考虑使用特异方法。6.6.13.1Non-SpecificAnalyticalMethods—Non-specificmethods,suchastotalorganiccarbon(TOC),UV,conductivity,pH,andvisualinspectiondetectthepresenceofmultipleingredientswhichcanbeacceptableincertainsituations.非特异的分析方法—非特异方法，例如TOC、UV、导电仪、pH以及目测检查在某些情况下可以接受的多种成分的存在。6.6.14UsingRiskAnalysisforMasterPlanning—Themasterplanforcleaningshouldbedevelopedbasedonthehazardidentificationoftheprocessresiduesandtheriskanalysisoftheequipmentdesignandcleaningproceduresasaminimumandprovidethebasisforthecleaningcontrolstrategy.在主计划中使用风险分析—应根据工艺残留物的危害识别制定清洁主计划，并且设备设计的风险分析和6.7RiskEvaluation风险评估:6.7.1Riskevaluationforcleaningcomparesthecleaningdatacollectedagainsttheacceptancecriteriafortheidentifiedhazards.Riskevaluationsshouldconsidertheprobabilityofpatientexposuretothehazard.清洁风险评估将收集的清洁数据与识别的危害的验收标准进行比较。风险评估应考虑患者暴露于危险的可能性。风险评估应当考虑病人暴露在这一危害下的可能性。6.7.2Riskreductionstepsshouldbeconsiderediftheevaluationofthedataforthecleaningprocessindicatestheriskisaconcern.Thiscanbedeterminedusingoneormoreriskanalysistools.减少风险的步骤应当考虑清洁工艺数据评估是否指示此风险需要进行考量。这可通过一个或多个风险分析工具来决定。6.7.3EvaluationofCleaningData清洁数据评估:6.7.3.1Oneapproachforthisassessmentisthemaximumsafesurfaceresidue(MSSR)basedontheADEforthatprocessresidue(16-18).ThecomparisonofprocessresiduetoMSSRscandemonstratewhetherprocessresiduesonequipmentproductcontactsurfacesposesignificantrisktopatientsandshowwhatthemarginofsafetyisforthatprocessresidue(18).一种评估方法是基于该工艺残留物的ADE的最大安全表面残留物(MSSR)。工艺残留物与MSSRs的比较可以说明设备与产品接触表面的工艺残留物是否将会引起对病人的重大风险，以及显示该工艺残留物的安全边际是多少。6.7.3.2Asimilarprocedurecanbeusedforassessingmicrobialrisk(19,20).类似的程序可以用来评估微生物的风险。6.7.3.3Theresultsofallcleaningrunsshouldbecollectedinadatabaseforthevalidationofthecleaningprocess.所有清洁运行的结果应收集在数据库中，以验证清洁工艺。7.RiskControl风险控制7.1Riskcontrolencompassesactivitiesaroundreducingriskswithperiodicassessmentsandmonitoringtoidentifyproblemsbeforetheyoccur.风险控制包括围绕减少风险的活动，包括定期评估和监测，以便在问题发生之前识别出问题。7.2RiskReduction—Afteridentifyingtherisksthataresignificantenoughthatreductionstepsshouldbetaken,thenextstepistoattemptreducingthelevelofriskpresented.Thiscouldincludestepstoreducetheseverityofthehazard,reducethelikelihoodofoccurrence,andincreasethedetectabilityoftherisk.Aspectsofriskreductioncouldinclude,butarenotlimitedto,theexamplesinTable1.降低风险—在确定了足够重要的风险以采取减少步骤之后，下一步是尝试减少所识别出的风险级别。这可以包括减少危害的严重性、减少发生的可能性和增加风险的可探测性的步骤。减少风险的各个方面可以包括但不限于表1（文末）中的例子。7.3Risk-BasedMonitoring基于风险的监控:7.3.1Risk-basedcontinuedmonitoring(levelsandfrequencyofroutinesampling)ofcleaningprocessparametersandsamplingofcriticalcleaningqualityattributesshouldbebasedonthedatacollectedduringcleaningprocessdevelopmentandqualificationruns.基于风险的对清洁工艺参数的持续监控(水平和常规取样频率)和关键清洁质量属性的取样应基于清洁工艺开发和验证期间收集的数据。7.3.2Monitoringandsamplingdonothavetocontinueatthelevelusedduringthecleaningprocessdevelopmentandqualificationrunsandcanbemodifiedbasedonthedataandlevelofrisk.Thelevelofmonitoringandsamplingcanbeadjustedtoastatisticallyappropriatelevelandberepresentativeoftheequipment.监测和取样不必在清洁工艺开发以及验证过程中使用过程继续进行，并且可以根据风险的数据和等级进行修改。监测和取样水平可以调整至适当的统计学水平，并且7.3.3StatisticalProcessControl(SPC)Limits—DatathathasbeencollectedintheriskevaluationorriskcontrolstagescanhaveSPClimitscalculatedfromthesedata.TheseSPClimitsshouldthenbeusedformonitoringcleaningprocessesandinplaceoftheMSSRlimitsusedfortheriskanalysisinriskevaluation.统计过程控制限-可以根据在风险评估或风险控制阶段收集的数据计算SPC限。应当在在监测清洁过程中使用SPC限度并且代替在风险评估的风险分析步骤中的MSSR限。7.3.4UseofPATinCleaningValidation在清洁验证中PAT的使用:7.3.4.1PATcanbeusedasasystemfordesigning,analyzing,andcontrollingcleaningprocesses.Thisincludestimelymeasurementsofcriticalcleaninginput/outputvariablesidentifiedanddocumentedduringriskassessmentandduringcleaningprocessdevelopmentwiththegoalofensuringpatientsafetyandproductquality.PAT可以用在设计、分析、控制清洁工艺的系统里。这包括及时测量在风险评估和清洁工艺开发过程中识别并记录的关键清洁参数。7.3.4.2Itisrecommendedthatcleaningprocessesusetechnologiestomeasure,control,andrecordthesevariablestodetectcriticalfailuresandensurereliableandconsistentprocessperformance.推荐在清洁过程中使用测量、控制、记录变量的技术，来监控关键缺陷并且保证持续可靠的工艺表现。7.3.4.3Thedatafromtheseapproachescanbeevaluatedstatistically.Modelscanbedevelopedandsubsequentrunscanbecomparedtothemodeltodeterminewhetherthecleaningisequivalenttothemodelorusedforreal-timereleaseofequipment.这些方法的数据可以进行统计评估。可以设计模型，并将后续运行与模型进行比较，确定清洁效果是否等效于模型或用于设备的实时释放。7.3.4.4On-lineandin-linesamplinghavethemostapplicabilityforcontinuousmonitoringofcriticalcleaninginput/outputvariablessuchastemperature,flowrate,TOC,conductivity,andsoforth.在线和采样最适用于连续监测关键清洗输入/输出变量，如温度、流量、TOC、电导率等。7.3.4.5At-linesamplingmayincludeTOC,conductivity,andvisualinspection.Informationcollectedduringorafteracleaningprocessiscompletecanenableparametricreleaseofequipmentandimprovemanufacturingprocessefficiency(GuideE2476).在线取样可能包括TOC、电导率和目视检查。在清洁过程中或清洁结束后收集的信息可以使设备参数释放，提高制造过程效率率。7.4RiskAcceptance—Afterriskreductionstepsonthecleaningprocesseshavebeencompleted,anevaluationofthesuccessoftheseeffortsshouldbemade.Itisatthispointthattheresidualriskiseitheracceptedorrejectedandscheduledforfurtherreductionefforts.Riskacceptancecanalsooccurafterriskevaluationwhenthereisnoneedtotakefurtherriskreduction.Thesedecisionsshouldbedocumentedinthecorrespondingcleaningriskassessment.风险接受—完成清洁工艺风险降低步骤后，应该对这些效应是否成功进行评估。在这一点上，剩余风险要么被接受，要么被拒绝，并被安排为进一步减少工作。在不需要进一步降低风险的情况下，风险评估后也可以进行风险接受。这些决定应该记录在相应的清洁风险评估中。8.RiskReview风险回顾8.1Theresultsofcleaningandmonitoringshouldbeformallyreviewedonaperiodicbasisbasedonariskassessment.Basedontheriskreview,anewriskcontrolstrategyandanewriskassessmentmayberequired.Thelevelofanymitigationorcorrectiveactionsshouldbecommensuratewiththeriskidentifiedorreviewed.Notethattheresultoftheriskreviewmaytightenorrelaxcontrolsthatwereinstitutedinthecourseoftheoriginalriskassessment.应根据风险评估定期正式审查清洁和监测的结果。在风险回顾的基础上，可能需要新的风险控制策略和新的风险评估。注意，风险回顾的结果可能会加强或放松在最初的风险评估过程收集清洁过程中或结束后的设备参数，改进生产工艺效中建立的控制。8.1.1NewProductIntroduction引入新产品:8.1.1.1Beforeintroducingaproductnewtoasharedmanufacturingfacility,adecisionshouldbemadeanddocumentedabouttheappropriatenessofmanufacturingthisproductinthefacility(seeRef(1)).Thiswouldincludeanevaluationofallrelevanttoxicologicaldata(pre-clinical,clinical,andaftermarket)byaqualifiedtoxicologistorpharmacologisttodetermineanADE(9).在引入新产品共用设施之前，应就在工厂生产该产品的适宜性作出决定并形成书面文件。这包括由合格的毒理学家或药理学家对所有相关毒理学数据(临床前、临床和上市后)进行评估，以确定ADE。Atthispoint,ifthenewproductisacceptable,thecleanabilityoftheproductshouldbedetermined.Ifthenewproductisnotcleanablewiththeexistingcleaningprocess,cleaningprocessdevelopmentwouldbenecessaryincludingqualificationstudiesforthenewcleaningprocess.Otherwise,noadditionalcleaningprocessdevelopmentorqualificationstudieswouldbenecessary.此时，如果新产品是可接受的，则应确定产品的可清洗性。如果新产品不能用现有的清洁工艺进行清洁，则需要开发新清洁工艺，包括新清洁工艺的资质研究。否则，不需要额外的清洁工艺开发或资格研究。9.RiskCommunication风险沟通9.1Riskcommunicationisthesharingofinformationaboutcleaningriskanditsmanagementbetweentherelevantparties.Communicationsmightincludethosewithinacompany,betweencompaniesandcontractmanufacturers,withregulatoryauthorities,andsoforth.Theinformationshouldconsidertheseverity(ADE/PDE),probability(priorcleaningresults),detectability(analyticalmethods),control(cleaningprocedures/cleaningagents),orotheraspectsofriskstopatientsafetyandproductqualitythatmayresultfrominefficientorineffectivecleaningprocedures.风险沟通是有关各方之间关于清洁风险及其管理的信息共享。这种交流的范围可能包括公司内部、公司和合同制造商、与监管当局之间的交流等等。可考虑的所交流信息有严重程度(ADE/PDE)、概率(清洗前结果)、可检测性(分析方法)、控制(清洁程序/清洁剂)、或者其他可能引起病患安全风险的无效的清洁程序。10.Keywords关键词10.1.1.1acceptancelimits;CIP;cleaningdevelopment;cleaningvalidation;PAT;risk-based;science-based.接收限；CIP；清洁程序开发；清洁验证；PAT；基于风险；基于科学。TABLE1ExamplesofRiskReduction表1减少风险的例子RiskReductionElements风险降低因素Reduceseverity

ExamplesofRiskReductionSteps风险降低例子Hazardremoval去除危害

Hazardreplacement(forexample,movetosafercleaningagents)Equipmentdedication

ReducelikelihoodCleaningprocessimprovements

Cleaningoptimization(DoE)Operatortraining

SOPimprovements(forexample,poka-yoke)Equipmentdedication

ModificationsofequipmentdesignImprovementsinequipmentstoragepracticesSelectionofnewequipment

Productcampaigning

IncreasedetectabilityAnalyticalmethodimprovement(lowerDL)

Samplingmethodimprovement(increasedrecovery)CleaningprocessmonitoringStatisticalprocesscontrol

IntroductionofPAT(forexample,at-linerelease)