|
4.0 Qualification Qualificationis a part of cleaning validation involving the traditional activities ofequipment qualification and process qualification. For cleaning validationpurposes, equipment qualification focuses on qualifying (or verifying) theequipment used as part of the cleaning process, such as a CIP skid andautomated parts washer. For fully manual cleaning operations, such as abrushing or scrubbing, there may be no equipment qualification activities.Design qualification has also been considered as another qualificationactivity, which is addressed in the design and development stage. 确认是清洁验证的一部分，它包括常见的设备确认和工艺确认。在清洁验证中，为确保验证达到预期目的，设备确认主要是对清洁工艺中使用的设备进行确认（或确证），例如CIP 模块和自动清洗机。对于刷洗、擦洗等全手工清洗操作，则不需要进行设备确认。此外，设计和开发阶段的设计确认是另一种的确认活动。
Theemphasis for this section is on process qualification activities. Process qualificationinvolves the runs performed under a protocol designed to demonstrate theconsistency of the cleaning process. The traditional approach for cleaningvalidation has been to focus on the qualification protocols to demonstrateeffectiveness and consistency. The lifecycle approach that the industry hasbeen moving toward involves a different approach with a more comprehensiveview, with qualification runs being only one of the stages of validation. Thelifecycle approach also includes design/development activities and validationmaintenance (ongoing controls). 本节重点是讨论工艺确认活动，工艺确认指按照设计方案完成的用以证明清洁工艺一致性的所有活动。传统的清洁验证注重证明清洁工艺的有效性和一致性。但是在制药工业领域最引入生命周期方法提出了更加全面的观点，它将确认活动视作整个验证中的一个阶段。生命周期的验证方法还包括前期设计、开发活动和验证的持续维护（持续控制）。
Thissection covers protocol elements and specific important issues for cleaningvalidation protocols, including the number of validation runs required, mocksoiling for validation runs, worst-case process conditions, and the dispositionof equipment/product during validation runs. It also covers grouping approachesfor products and equipment as well as important considerations in clean holdtime studies. It ends with a discussion of documentation for “cleaningverification”. 本节内容包括清洁验证方案要素和验证方案中关键点，包括验证次数、验证中的模拟污物、最差工艺条件和验证期间设备/产品处置。此外，还包括产品和生产设备的分组方法、清洁保持时间研究要点以及对“清洁效果确认”文件的讨论。
4.1ProtocolElements 验证方案要素 Cleaningvalidation protocols have many of the same elements as process validationprotocols. For reasons of clarity, the format of a cleaning validation protocolusually follows the same approach (as appropriate) as used for processvalidation protocols for a given company. Common elements include (but are notlimited to) purpose, validation design/strategy, scope, responsibilities,applicable product(s) and equipment, cleaning procedure and associateddocumentation, acceptance criteria, training, and a requirement for a finalreport. Key elements for cleaning validation protocols include residue limits(see Section 5.0), sampling procedures (see Section 6.0) and analytical methods(see Section 7.0). 清洁验证方案要素与工艺验证方案相似。为便于表述，特定公司的清洁验证方案通常采用与工艺验证方案相同的模板（适当时），一般包括（但不限于）验证目的、验证设计/策略、验证范围、职责、适用产品和设备、清洁程序和相关文件、可接受标准、人员培训和最终验证报告的要求。清洁验证方案中关键要素包括允许残留限值（见5.0节）、取样程序（见6.0节）和分析方法（见7.0节）。
Twoapproaches are used for documentation of elements. One general approach is toreference other documents for details regarding that element. For example, specificationof swab sampling sites can be in the protocol while the rationale for selectionof those sites can be in another document that is referenced in the protocol.The advantage of referencing other documents is that only the detailedinformation required for executing the protocol is included in the protocol;supporting information is only referenced thus allowing for more “streamlined”protocols. Another approach is to include all relevant details for a givenelement in the protocol. The advantage of having more details in the protocolis that greater clarity is provided to those executing the protocol. Theapproach used should consider the knowledge management systems within a givenfirm. 有两种方法可用于组织和编写验证文件，常见的方法是引用其他文件相关要素的详细内容。例如，将擦拭取样布点写在验证方案中，而取样点的选择依据在另一个文件中，则可以在验证方案中引用该文件的相应内容。由于方案中只需要详细说明项目需完成的操作，相关支持信息引用即可，这种编写方法可以使方案更加流畅。另一种方法是将方案特定要素的所有相关信息全部描述清晰，验证方案执行者可直接获取明确的信息。企业应当根据自身的知识管理系统选择适当的方法。
4.2Key Protocol Issues 验证关键点 Thevalidation protocol is not written and approved until the cleaning process hasbeen designed and developed (see Section 3.0). The execution of the protocolshould not begin until the protocol is approved. However, execution of theprotocol as an engineering or practice run can be helpful in some circumstances(e.g., for activities that are highly complicated or new to those executing theprotocol). Any problems in the execution of the engineering/practicerun can be corrected before actual validation runs. The time spent in suchruns may lead to the higher likelihood of “right first time” protocol executionfor the formal qualification runs. 清洁验证方案应在清洁工艺设计和开发（见3.0）后起草和批准，并在获得批准后实施。在某些情况下（如操作高度复杂或方案执行者首次实施的活动），可以先模拟实施验证方案。在模拟验证中发现的任何问题均可在实际验证实施前纠正。这种模拟有助于提高正式验证的一次成功率。
Keyissues for protocols (aside from limits, analytical methods and samplingprocedures, which are covered elsewhere) are discussed below. 下面对验证中的关键点进行讨论（残留限值、分析方法和取样程序在其他章节讨论）
4.2.1Number of Runs in a Protocol 验证次数 Thetraditional approach for cleaning validation protocols has been to require anevaluation of three consecutive runs of the cleaning processes. “Consecutive”means that no cleaning events of that same process are skipped without anappropriate rationale. For example, if the cleaning validation is for cleaningof Product A, there may be manufacture and cleaning of Product B in betweenmanufacture of lots or batches of Product A. 传统的清洁验证要求评价连续三次清洗活动。“连续”意味着没有充分理由，必须连续完成相同的清洁工艺，不得中断。例如产品A清洗验证实施过程中不可以穿插生产和清洗产品B。
Basedon lifecycle approaches to validation, as well as several regulatory documentsincluding the 2011 U.S. FDA process validation guidance, the newer approach hasbeen to provide a rationale, based on an understanding of the cleaning process,documentation from the design and development phase, and data fromsufficiently similar cleaning processes, for a specific number of validationruns required (9,10). This might result in fewer than three runs or greaterthan three runs. It should be recognized that this new U.S. FDA processvalidation guidance does not formally cover cleaning validation. However, anumber of principles in that document may be applicable to the validation ofcleaning processes. 基于验证生命周期方法以及其他几个法规如2011年美国FDA颁布工艺验证指南，提出一种新的验证方法。基于对清洁工艺的充分理解、设计和开发阶段记录和足够相似清洗工艺的的数据，该方法提供了确定需要完成的验证次数的依据（9,10）。清洁验证次数可少于三次，也有可能多于三次。应该认识到，虽然FDA新的工艺验证指南并不涵盖清洁验证，但是指南中的一些原则可用于清洁工艺的验证。
4.2.2Mock Soiling 模拟污染 Ordinarilya cleaning validation run is performed by cleaning on a commercial-scale batch.An alternative approach is to use what is called “mock soiling” or “artificialsoiling” to simulate the nature and condition of the manufactured product onthe commercial equipment at the time of initiating the cleaning process. Ifmock soiling is used, a rationale must be provided for its use as well as whythe mock soiling simulates a “realistic” manufacturing situation. A commonreason for mock soiling has been to obtain three consecutive cleaningvalidation runs without being forced to make three commercial-scale batches ofthe cleaned product. “Mock soiling” (a process) should be distinguished from a“mock soil” (sometimes called a “surrogate soil”), which is a product whichsimulates the physicochemical properties of the actual soil. 通常情况下，清洁验证需要在产品商业规模生产时进行。还有种方法是在启动清洁工艺时采用“模拟污染”或者“人造污染”模拟所生产产品在商业生产设备上的特性和状态。使用这种方法时，必须提供“模拟污染”的依据，并说明它如何模拟“实际”生产状态。常见的原因是采取模拟污染的方式可以完成三次连续的清洁验证活动，而无需生产三批次商业规模产品。应将“模拟污染”（一个过程）同“模拟污物”（也称作污物替代物）区分开来，模拟污物是一个产品，用来模拟生产中真实污物的理化性质。
4.2.3Worst-Case Process Conditions 最差工艺条件 Thetraditional approach for cleaning validation protocols has been to includeworst-case process conditions in the three protocol runs. Rationales forworst-case conditions should be given in or referenced in the protocol. Forexample, worst-case process conditions may include maximum dirty hold time,maximum batches or elapsed time in a campaign, shortest allowed time for manualcleaning steps, lowest allowed temperature for manual cleaning processes, andworst-case circuits for CIP skid selection. 在设计传统的清洁验证方案时已经考虑到需在最差工艺条件下完成三次验证，评价最差工艺条件的原则应包含或引用在验证方案中。例如，最差工艺条件可包括最长的生产后保持时间、阶段性生产中最大批量或者最长运行时间、最短的手工清洗时间、最低的手工清洗用水温度和最差的CIP模块循环回路。
Parameterssuch as temperature, cleaning agent concentration, flow rates, and process steptimes for automated cleaning processes are generally controlled in a narrowrange such that challenging the cleaning process in the validation runs at thelower or upper end of the specification is not appropriate. If those narrowlycontrolled parameters are to be challenged in the extremes or outside thespecified range, those challenges can be evaluated in development studies todemonstrate the robustness of the cleaning process. 由于自动清洁工艺中温度、清洁剂浓度、流速和工艺步骤运行时间等参数一般控制在一个狭窄的范围内，所以采用参数的上下限进行清洁工艺挑战验证是不合适的。清洁工艺开发研究时，可以通过挑战规定范围内极限值或者超出范围的参数来证明所建立清洁工艺的耐用性。
Theremay be different approaches for addressing worst-case process conditions. Inone approach, a worst-case process condition is addressed in each of therequired validation runs. An alternative approach is to address a specificworst-case condition in the design and development of the cleaning process suchthat the cleaning process is developed to address a worst-case condition. Datafrom such design and development studies may support the use of worst-caseconditions in fewer runs. 有多种方法可用来确定最差清洁条件，一种方法是在每一清洁验证批次均采用最差清洁参数。另一种方法是在清洁工艺设计和开发过程时针对一个特定最差条件进行研究，这样开发的清洁工艺已经涵盖了最差条件。这种设计和开发研究中获取的数据有助于减少采用最差条件的验证批次。
Anotherexample of worst-case conditions is the number of batches in a campaign wherevalidated cleaning is only performed at the end of campaign. In such cases,there may be no cleaning between batches or there may be only “minor cleaning”(such as vacuuming for solids manufacture or a water rinse for liquidsmanufacture). In this case, the maximum number of batches may represent theworst case. Therefore, the validation protocol should consider the effect ofthe maximum number of batches in the campaign. 另一类的最差条件是在阶段性生产后进行清洁验证，批次之间不进行清洁或者仅进行“小清洁”（例如固体生产时真空吸尘或者液体生产时的水冲洗）。在这种情况下，最多连续生产批次即为最差条件，因此验证方案应考虑阶段性生产时最多批次的影响。
Insuch an approach, it may not be feasible to schedule three consecutivecampaigns with the same maximum number of batches. One practical way to addressthis is to manufacture and perform cleaning validation after a specifiednumber of batches that may represent a minimum campaign length. When a campaigninvolves more than the previous number of batches, a validation protocol is executedon that longer campaign. Data from the longer campaign are then compared withdata from the earlier validation runs to determine whether the data areequivalent. The specifics of the results will indicate whether additionalvalidation runs are needed to extend the validated length of the campaign. 采用这一方法，安排具有相同批次的三个连续阶段性生产并不可行。一个解决办法是在生产指定批次（可代表最小阶段性生产批次）后进行清洁验证。如连续生产批次超过该指定数量，则应对更长的连续生产批次进行清洁验证。然后对比分析前后两次验证数据是否一致，是否需要额外的验证来延长阶段性生产批次。
Athird approach is to address campaign length during the design and developmentphase. If data or a rationale can be developed to support no change in thedifficulty of cleaning regardless of the campaign length, then the validationruns can be at any campaign length. 第三类方法在设计和开发阶段对阶段性生产批次进行研究。如果有实验数据和原理能够证明阶段性生产批次不影响清洁难易程度，清洁验证过程中就可以任意选择阶段性生产批次进行验证。
4.2.4Disposition of Products and Equipment during Validation 验证期间产品和设备的处置 Acleaning process generally only affects the next product manufactured in thecleaned equipment. Therefore, following protocol execution the “cleaned”product may be released following company procedures for product release. Thatrelease of product is independent of the data obtained for the immediatelyfollowing cleaning process. The data from that cleaning process is used for therelease of the cleaned equipment. 一个清洁工艺通常只影响已清洁设备中生产的下一个产品，因此，按照验证方案执行后，可按照产品放行程序对所生产产品进行放行。产品的放行与生产后的清洁过程数据无关，清洁过程所获得数据可用于已清洁设备的放行。
Thereare several approaches used for disposition of the equipment following thecleaning process. One approach is to not release the equipment until acceptabledata (including meeting all residue criteria) are obtained for that specificvalidation run. At that time, the equipment may be safely released formanufacture of the same or another product. 清洁验证中，清洁后的生产设备有多种处置方法。一种方法在确定本批次清洁验证数据符合要求（包括符合所有残留限度）前该设备不得放行使用，设备放行后方可用于该产品或其他产品的生产。
Analternative approach is to release the equipment following company proceduresat risk for manufacture of the next product. However, that next product cannotbe released until acceptable data (including meeting all residue criteria) areobtained for that specific validation run. If the cleaning validation run failsto meet its acceptance criteria, then the impact on that specific nextmanufactured product should be assessed as part of the investigation into thatnon-conformance. The results of the investigation will determine whether thatnext manufactured product can be released. 另一种方法是在符合公司风险控制程序的前提下放行设备，用于生产下一产品。但该产品需等到该批次清洁验证数据符合要求（包括符合所有残留限度）后放行。如果该批次清洁验证不符合要求，那么作为该“不符合”调查的一部分，应评估其对该产品的影响。调查的结果将决定该产品能否放行。
Ifthere are separate validation protocols for equipment items in a train used tomanufacture the product on which the validation is being performed, eachequipment item can be released based on the protocol data for that validationrun for that equipment item. It is not necessary to wait until validation iscomplete on all equipment items in the train before any item can be releasedfor subsequent manufacture. 如果一条生产线中的不同设备需单独进行清洁验证，那么每一设备都可以基于特定批次清洁验证数据放行使用，无需等待所有设备均验证成功。
4.3Grouping/Family Approach 分组/分类的方法 Groupingis a strategy whereby manufactured products and/or equipment are consideredtogether, and a formal protocol is performed on a representative from thegroup. The representative from the group is usually the worst case amongproducts or equipment in a group. Grouping is also called matrixing, familyapproach or bracketing. The rationale for grouping is to generate optimum valuefrom cleaning validation tasks based on a risk approach. One requirement forgrouping is that product and equipment be cleaned by the same cleaning process.The use of products and equipment grouping may be used to streamline cleaningvalidation programs while ensuring sufficient data are available to supportthe validation of procedures, processes, and equipment associated withcleaning. The grouping program for a given facility or company should bespecified or referenced (e.g., by pointing to a facility cleaning rationale)in a well-designed validation program/validation master plan. 分组法是综合考虑所有产品/设备后对其进行分组，选取组内代表性的产品/设备来替代整组进行验证的策略。用来作为代表的产品或者设备通常是组内最难清洗的。分组法也被称为矩阵法、分类法或交叉法，它是一种运用风险分析的方法在清洗验证中选择最合理的验证目标的方法。分组的原则是划分为一组的产品和设备必须采用同一清洗工艺。运用产品和设备分组的方法可以在简化清洁验证程序的同时又获得足够有效的数据来支持所验证的程序、工艺步骤和设备达到预期。特定工厂或企业的分组方案应当在一个精心设计的验证计划/验证主计划中予以明确或引用 (例如索引至一个关于工厂清洗验证原则的文件)。
4.3.1Product Grouping 产品分组 Productsmay be grouped together if they are manufactured on the same or equivalentequipment, and cleaned by the same cleaning procedure. Products may be assessedfor their relative cleanability by several methods. Relative cleanability maybe affected by the nature of the active ingredients, of the excipients, and/orof degradation products. One example of assessing relative cleanabilityinvolves selecting the product with the least soluble active ingredient in thecleaning solution. This approach may be appropriate for small-molecule APIsynthesis cleaned with a solvent or for finished drug product manufactureinvolving water-soluble formulations. Such an approach may also be possiblefor solid dosage drug products provided that the excipient portion of thedifferent drug products has the same effect on the difficulty of cleaning. 可以将在同一或者等同设备上生产，同时清洁工艺又相同的产品定义为一组。组内产品相对可清洁性有多种方法来评价。活性成分、辅料和/或降解产物的特性都会影响相对可清洁性。例如可选择API在清洁溶液中最难溶解的产品，进行相对可清洁性的评估。在使用溶剂清洗合成的小分子API或者清洗水溶性配方的制剂时，这种方法比较适用。此外，如果固体制剂产品中辅料部分（赋形剂）对清洁难易程度具有相同影响时，也可以采用这一方法。
Anotherapproach involves determining relative difficulty of cleaning using laboratorystudies. For laboratory studies, cleanability is assessed on coupons or smallequipment parts using representative surfaces, with stainless steel being themost common because of its predominance in pharmaceutical equipment. Forcoupons, the roughness of the surface should be the same or rougher (as aworst case) than actual equipment surfaces. From the lab results, the relativecleanability of each product is defined, typically by determining underproposed cleaning parameters which product requires the longest time to clean.Bioactivity and clinical effects may also be considered for the selection of arepresentative product. 另一种方法是通过实验室研究确定相对可清洁性，采用材质试样或设备小部件表面，评估可清洗性，由于不锈钢设备在制药行业广泛使用，一般选择不锈钢材质的部件进行研究。材质试样的表面粗糙度应该与生产设备表面相似或者更加粗糙（作为最差条件）。根据实验室研究结果，确定每一产品的相对可清洁性，通常采用推荐的清洁工艺参数确定哪个产品清洁耗时最长。此外，产品的生物活性和临床药理活性也应该作为代表性产品选择的依据。
Oneoption for product grouping is to use a surrogate worst-case product. In thissituation, the worst-case product is an artificially constructed product (whichmay not be a commercial product) designed to be more difficult to clean thanproducts expected to be routinely manufactured. One rationale for this approachis to maintain continuity of the worst-case product (in cases where acommercial product might be discontinued). Another rationale is to minimizesituations in which new worst-case products are added. 产品分组的另一个方法是引入一种最差条件产品替代物。通常，用作替代品的物质是人工创造出的（可能不是商业产品），它比组内其他产品更难清洁。这种方法的一个优点是可以保证组内最差产品的延续性（如选择上市产品，存在停产的可能)，另一个优点是可以降低引入新的最差产品时重新验证的可能。
Aqualification protocol on the representative (worst-case) product is performed.The acceptance criterion for that worst-case product is generally the moststringent acceptance criterion of all products in the group (that is, thelowest residue limit). Successful cleaning validation of the representative(worst-case) product means the cleaning of the other products in the group isalso validated. Based on risk assessments (addressing both quality risks andbusiness risks), one approach is to perform a single confirmatory validationrun on every other product in the group. Also based on a risk assessment, anotherapproach is to perform qualification protocols on both the most difficult toclean product and the product with the lowest limit. 采用代表性产品（最差条件产品）进行确认。最差条件产品的最低可接受残留限值可作为组内所有产品最严格的可接受标准。代表性产品（最差条件产品）清洁验证的成功同样意味着组内其他产品的清洁工艺已验证合格。但基于风险评估(包括质量风险和商业风险)，一种方法是组内的其他产品都进行一批次清洁确认；另一种方法是同时选择组内最难清洁产品和允许残留限值最低的产品进行验证。
4.3.2Equipment Grouping 设备分组 Equipmentmay be grouped together if they are similar and can be cleaned by the samecleaning procedure. Grouping of equipment is an effective method forencompassing equipment from a limited population of systems undergoing cleaningvalidation without redundant testing. The grouping strategy is based on designating equipment as “identical”or “similar,” based on design, mode of operation, and cleanability. Such adetermination usually involves evaluating the equipment qualification, with thestipulation that qualification differences that do not affect the cleaningprocess may allow one to conclude that two equipment items are identical forcleaning purposes. Regulatory documents such as the U.S. FDA SUPAC guidance mayassist in that determination (11). 设备分组要求组内设备相似，并且清洁工艺相同。在清洁验证中，设备分组是一种将需清洁验证的设备合理分组，避免多余测试的有效方法。分组策略是以设备设计、操作模式和可清洁性的“等同”或“相似”为基础的。判定设备是否“等同”或“相似”，需要对设备确认进行评估，如果设备确认的差异不影响清洁工艺，在可判定两个设备在清洁方面是等同的。法规如美国FDA 颁布的SUPAC指南（药品放大和批准后变更指南）有助于“等同性”判定（11）。
Onceequipment has been placed within a designation, the designation defines thecleaning validation requirements. If it involves identical equipment, aprotocol involving any combination of identical equipment items in the group isperformed. Provided an adequate rationale is given for determining theequipment items are identical, there is no need to perform validation runs onevery item in the group. For similar equipment, the representative equipment isthe worst case or may involve bracketing of the equipment. For example, forstorage tanks that are of the same size but different complexity due to thenumber of baffles, the more complex equipment is chosen as the worst case. Forsimilar equipment of different sizes, the largest and smallest (representingthe extremes) may be chosen for the formal validation runs (unless one size canbe determined as the worst case). If there is no worst case or bracketinginvolved, then any equipment items in the group of similar items may be chosenfor validation runs. Confirmatory validation runs (perhaps only one run) are anoption for other equipment (not a worst-case) within the group. 一旦将设备划分到指定组，就可以对这个组的验证要求进行定义。当组内设备均等同时，组内等同设备的任意组合进行验证。假如有足够的证据证明组内设备等同，就没有必要对组内每个设备进行验证。当组内设备相似时，可以选择最难清洁或者通过交叉法选择。例如大小相同，但内部隔板数量较多、结构复杂的罐子就是最难清洁设备；大小不同的设备，最大和最小两个规格（两个极端）均作为最难清洁设备（除非最大和最小中有一个可以作为最差条件）。如没有最差条件设备或者没有采用交叉法确定代表性设备，可以选择组内任一设备进行验证，组内其他设备（非最差条件）完成清洁确认即可（也许只需进行一个批次)。
Aspecific case of equipment grouping involves “minor” equipment, such asutensils, small parts, and smaller equipment. In the case of such minorequipment, it may be appropriate to evaluate a cleaning procedure for thoseparts and to validate the cleaning process using equipment grouping. The groupingof the parts may involve selection of worst-cases based on complexity, size andfunctionality. 容器、小部件以及较小设备等小设备分组时，也可以采用设备分组的方法评价和验证这些部件的清洁工艺。分组时可以通过对比小设备的复杂性、尺寸和功能，来选择最难清洁的部件。
4.3.3Introduction of a New Product or Equipment into a Group 组内引入新产品或设备 Theintroduction of a new product into an already validated group is assessed usingthe same science and risk-based evaluation process (e.g., based on solubilityin the cleaning solvent, a laboratory coupon study, and/or information fromother process cleaning studies) to initially determine the worst-case product.It is recommended that if each new product is tested in a lab evaluation, asuitable control, such as the previous worst-case product, be included.Relative product cleanability is then used to determine validation requirementsfor that new product on equipment used for other products in that group. Therelative cleanability of the product in relation to the preceding worst-caseproduct, as well as any change in the lowest limit for products in the group,will dictate the validation requirements. Based on a documented riskassessment, introduction of an easier-to-clean product may just requirelaboratory and/or scale-up studies to confirm ease of cleaning or may requireone confirmatory run. Introduction of a more difficult-to-clean productrequires validation of that new worst-case product. 向已验证过的组内引入新产品，需使用与最初确定最难清洁产品时相同的科学风险评估过程进行评估，例如评估产品在清洁溶液内的溶解性、进行实验室材质试样研究和/或者其它的清洁工艺研究。假如每新增一个产品都进行实验室测试应采用适当的对照，如前最差条件产品进行研究。通过比较新产品和组内产品的相对可清洁性来确定引入新产品的验证需求。而新产品与前最差条件产品的相对可清洁性，以及组内产品最低允许残留限值的变化都决定着验证要求。基于书面的风险评估，组内新增较容易清洁品种一般只需要在实验室和/或中试放大研究时确认其容易清洁或进行一批次清洗确认即可；组内引入更难清洁品种则需要对新最差条件品种进行清洁验证。
Basedon risk considerations, introduction of new identical equipment may simplyinvolve a determination that it is equivalent or may require an additionalconfirmatory run. Introduction of new similar equipment requires an evaluationif that new equipment represents a new worst case or a new bracketing extreme.If not a new worst case or new extreme, special attention should be paid to thefirst commercial cleaning event to confirm effectiveness. If the new equipmentis a new worst case or bracketing extreme, the validation requirements for theprevious worst case or bracketing extreme should be performed for the new worst-caseor bracketing extreme equipment. 基于风险的考虑，组内新增等同的设备时，可确认新旧设备是等效的，或者需要额外进行一个批次清洗确认即可。当组内引入相似的生产设备，需要评价新设备是否形成最差条件或新的交叉法极端条件，如果不是，则应特别注意新设备第一次商业生产后的清洁后效果确认，确认清洗工艺有效；如果形成新的最差条件或交叉法极端条件，新设备需按照原先最差条件或交叉法极端条件设备的验证要求重新完成清洁验证。
4.4“Cleaning Verification” Documentation “清洁效果确认”文件 “Cleaningverification” as used in this Technical Report refers to documentation whichsays that a one-off cleaning event is effective for cleaning equipment so thatthe equipment can be used for subsequent manufacture of a product. There maybe a variety of other terms for this same concept that are used by variouscompanies. Examples of where cleaning verification might be used includecleaning after manufacture of a clinical trial product or cleaning afterproduct manufacture where there is a deviation (e.g., the dirty hold time isexceeded) that affects a validated cleaning processes. 在本技术报告中，“清洁效果确认”是指用来证明生产设备经一次清洁操作后可以用于后续生产的文件。对于“清洁效果确认”，不同企业用语各不相同。清洁效果确认适用于如临床样品生产后的清洁，或者出现偏差（如超出了规定的生产后保持时间）影响了清洁工艺验证状态。
Documentationfor cleaning verification purposes is similar to the documentation for cleaningvalidation, except that the verification data is specific to one cleaningevent. From a compliance perspective, the data applies only to the one cleaningevent (although from a scientific perspective the data may suggest similarperformance if the cleaning event were repeated). Another difference is thatbecause cleaning verification is typically performed on a unique cleaningevent, there may be limited cleaning design and development before execution ofthat event. One approach is to utilize a cleaning SOP and a cleaningverification protocol. Alternatively, companies might use a concept thatdefines explicit requirements for cleaning verification in an SOP and documentsthe specific activities, sample positions and so on in a form which will beapproved. It is generally not appropriate to consider three cleaningverification runs as constituting a “validation” especially if the element ofappropriate design and development is absent. 除了清洁效果确认针对一次性清洁活动，清洁效果确认的记录与清洁验证相似。从合规的角度来看，清洁效果确认数据只用于证明一次的清洁效果符合要求（尽管从科学角度来说数据可暗示重复清洁能够得到相似效果）。另一个区别是由于清洁效果确认通常是针对一次单独的清洁活动，在进行清洁活动前可能只进行了有限的清洁设计和开发。一种方法是建立一个清洁SOP和一个清洁效果确认方案。另一种方法是在一个SOP中明确规定清洁效果确认活动的各项要求，并在批准的表格中记录所进行的活动、取样点位置等。需要指出的是，将三个批次清洁效果确认看做“清洁验证”是不合适的，尤其是在缺少适当设计和开发工作的前提下。 |
|
|
