5.0 Data Integrity in the Pharmaceutical Microbiology Laboratory 5.0 微生物实验室的数据完整性   5.1 General Considerations and Risks   一般原则及风险   The approaches used to investigate the occurrence of suspected data integrity issues that h recently occurred in a pharmaceutical microbiology laboratory can be challenging and, in some cases, may be very different than those used to evaluate similar occurrences in an analytical chemistry laboratory, Many microbiological methods are performed manually;   subsequently, the recorded results are often based on the visual observations by an individual scientist performing the tests.   制药企业微生物实验室对可疑数据完整性问题的调查方法，越来越成为一个挑战，并且在一些情况下，与同样发生可疑数据的化学分析实验室的调查方法完全不同。很多微生物测试方法都是手动操作，以及所有的测试结果都由微生物测试人员人工检查并记录。   Listed below are a few examples of regulatory observations, Warning Letters, or other institutional accounts that note data integrity problems associated with microbiological laboratory records. These are only examples and are not intended to be an all-inclusive list of concerns:   以下是一些官方检查项，警告信，或其他官方检查关于微生物实验室记录数据完整性问题的情况列举，这些仅是举例并不完全代表所有与微生物实验室数据完整性相关的问题：   ●Company failed to record and report reliable and accurate data for the environmental monitoring (EM) results; for example, contamination in Grade A rooms were recorded and reported as having no viable microorganisms present when, in fact, microbial contamination was present. Specifically, the settle agar plates used in these areas were felsely reported as having “ no colonies present” but were found to contain 16 colony-forming units (CFU), more than could be reasonably overlooked.   ●公司未记录和报告真实准确的环境监测结果数据；例如，A级区实际存在微生物污染但实际记录和报告为零。特别是A级区沉降菌报告为“无菌落生长”，但实际上发现沉降菌结果为16cfu，可能会有更多类似的情况被掩盖。   ●Company failed to implement an adequate quality assurance system as evidenced by: product sterility test failures that occurred and were not reported, investigated, or documented; five batches of viral harvests that were rejected due to contamination，yet no reports were initiated；and a company practice that product sterility test failure(s) were investigated only if more than one test jar per batch of the first or second harvests failed for sterility.   ●公司质量保证体系未有效执行，具体表现为：产品无菌测试失败未报告，调查或记录；5批次产品由于病毒污染而被拒绝放行，但没有任何调查报告被发起，以及某些公司仅在第二次无菌失败或同一批次超过一瓶出现无菌样品失败的情况下，才会发起无菌调查   ●Company used a contract laboratory to perform microbiological testing; however, the company audit checklist used to evaluate the suitability of this laboratory was completed by the contract laboratory, not the company, with no follow-up verification.   ●公司存在委托实验室进行微生物检测的情况，然而，对委托实验室符合性审计的主体是由该委托实验室进行而非公司，且并未进行任何确认。   ●Private testing laboratory claimed to have conducted microbiological testing, yet it did not have the laboratory equipment (i.e., incubators) and/or media necessary to perform the analysis.   ●公司声称内部进行微生物实验测试，但实验室并没有实验相关的设备（如培养箱）和/或测试所必须使用的培养基。   ●Company used an “ unofficial” notebook to record microbial contamination in the plant’s water system; however, there was no official investigation or documentation regarding the water system contamination with a known pathogen (Pseudomonas aeruginosa).[Comment: This type of observation relates to both the microbiology laboratory and operators’ behaviors. Periodically, deceptive individuals will use the same technique to mislead, misrepresent, and/or obscure emerging microbial problems in manufacturing equipment that can impact product quality.]   ●公司使用“非正式”的记录本以记录工厂内水系统的微生物污染情况；然而，当发现水系统中存在已知控制菌（铜绿假单胞菌）污染时，公司并未针对此污染情况发起任何正式的调查或记录。<备注：此类型的污染与微生物实验室及人员操作污染相关。隐瞒人员会定期使用相同的技术来误导、歪曲和/或掩盖生产设备中出现的微生物问题, 从而影响产品质量。>   ●Company recorded results that the growth promotion test on the media used for simulation studies supported growth meeting the standard set by USP Chapter <71> Sterility Tests when, in fact, the microorganisms did not grow (30). In another case, filamentous fungi were seen growing in all five spiked media tubes, indicating contamination, yet laboratory records claimed that all five distinct species of microorganisms were actually growing per the USP standard.   ●公司记录用于培养基模拟灌装的培养基促生长测试结果符合USP<71>无菌检测章节的标准菌种生长要求，然而事实上，XX（30）微生物并未生长。另一个情形中检查人员发现，真菌菌丝在5管液体培养基试管中生长，显示了污染，但实验室记录却显示按照USP标准5管试验管中显示各测试菌种生长良好。   Again, while this list is not exhaustive, it does present actual inspection observations made by several regulatory auditors during their documentation of data integrity anomalies in pharmaceutical microbiology laboratories.   虽然以上显示的清单并不详尽，但却来源于多次官方审计检查中微生物实验室存在的缺陷检查记录清单。   5.1.1 Interviewing Analysts   问询分析人员   One critical element in conducting an audit for data integrity problems in a microbiological laboratory is interviewing the individuals who perform the QA/QC tests, in particular, the laboratory analysts or technicians. When reviewing analytical results reconded on worksheets or data printouts from the LIMS, for example, it is extremely difficult to detect data that should have been recorded but was not. Much of what analysts learn comes from on-the-job training, yet unofficial dialogue with cowokers or supervisors is rarely captured or documented. For instance, when a senior analyst instructs a junior analyst on how to handle the appearance of “unwanted” microorganisms found growing on analytical petri plates (such as“write the numerical count of the suspected colonies on the lid of the petri dish but don’t record it on the official worksheet until the supervisor has a chance to review it.”)，this “unofficial” practice will not be found in the company’s standard operating procedures（SOPs）. An auditor can best assess the potential for inappropriate practices, first, by verifying the acceptance criteria described in SOPs and，then，by inquiring of the analysts or technicians if they have been instructed to adjust or modify data or divert from the laboratory SOPs in any form. Without conducting such foce-to-face interviews, this kind of microbiological data manipulation would be extremely difficult to detect.   对微生物实验室数据完整性问题的审计一项重要的基本要素是问询执行微生物测试的QA/QC人员，特别是实验室的分析师和技术人员。当审核记录于工作表或从LIMS系统中打印的分析结果时，很难发现本应被记录而未被记录的数据。大部分分析人员的知识来源于在岗培训，因此同事间的或与主管间的非正式谈话就很难被捕捉或记录。例如，当一个高级技术员指导新员工进行培养皿微生物计数，当发现培养皿上“不好的”（或“非预期生长的”）微生物时（例如，在主管审核结果之前将计数结果写在培养皿上而不是记录表格中），这种“非正式”的操作就不会出现在公司SOP描述中。审核员可以评估潜在不当做法的最佳方式是，确认sop中描述的验收标准，然后，通过询问分析员或技术人员，去分析他们是否被指示调整或修改数据或以任何形式从实验室sop转移。如果不进行这种面对面的面谈，这种微生物数据操纵将极其难以察觉。如果没有面对面的这种沟通，这种类型的微生物实验室数据操作将很难被发现。   5.1.2 On-site Laboratory and Sampling Review   公司内实验室和取样环节审核   There are several procedural steps when handling, shipping» and storing microbiological samples that can dramatically and negatively impact final analytical results. Specific examples of managing samples that can diminish recovery of microbiological or endotoxin results and, thus, should be avoided through procedures and adequate employee training, are listed below:   文件中有很多关于微生物样品取样，运输，储存的步骤会对最终微生物测试结果存在显著的、不良的影响。为避免影响微生物及内毒素测试结果应当包括完善的管理文件和足够的人员培训，例子包括：   ●Collecting in-process product or water samples in an inappropriate container that may bind (or remove) a bacterial endotoxin from the sample may later reduce the true level of bacterial endotoxins within that test portion.   使用不合适的容器取中间产品样品或水的样品可能会弱化（或去除）样品中的细菌内毒素水平，随后可能减少被测试样品中真正的内毒素水平。   ● The excessive use of disinfectants to spray the sampling port(s) prior to taking the sample may result in the disinfectant dripping into the sample bottle and reducing the actual bioburden of the product sample. Sampling from the sample ports should closely simulate the common practices performed during routine production operations.   取样前对取样口使用消毒剂过度喷洒可能导致消毒剂滴落在取样瓶中导致降低所取样品中真正的微生物负载水平。从取样口取样应当正常模拟日常生产时使用该取样点的情况。   ● When using in-line sample collection manifolds for collecting intermediate-stage product manufacturing, the manifold needs to cool down after heat sterilization so it does not kill the microbes collected for laboratory testing.   采用在线样品收集管道抽取中间体产品时，收集管道在加热灭菌后需要冷却，以免杀死供实验室检测用样品中收集到的微生物。   ● Product sample collection bags should not be held directly under a UV light within a storage cabinet. Exposure to UV light will kill the microbes present before the sample collection bags are delivered to the laboratory for testing.   取样瓶不应该直接放在有UV紫外灯的传递窗中。样品瓶传递至实验室之前暴露在UV紫外灯下将杀灭取样器中的微生物   ● Microbiological sample bags should not be stored in a freezer or extremely cold refrigerator, which may injure or stress indigenous microbes. Microbial samples should be placed within a controlled environment to preclude any harmful or deleterious impact on the samples; then, analyses should commence as soon as reasonably possible.   微生物取样瓶不应该储存于冰箱或超低温冰箱中以免使样品中微生物受损或阻碍生长。微生物样品应当被放置在受控的环境中排除任何对样品有害的影响；否则，分析人员应当立即进行测试。   Finished products or in-process samples that are undergoing sterility testing in a Class 100 clean room (or isolator) should be disinfected using an antimicrobial solution and a validated contact time. Disinfectant vapors or solution passing through the product packaging or container-closure and into or onto the product prior to sample analysis may prevent or kill microbes in the media-enrichment test environment, which may inhibit recovery of product contaminants. For example, if the product sample was contaminated prior to the package disinfection step, a vulnerable container-closure system prepared with excessive disinfectant would likely result in false negative data. The sample container disinfection process should be included in laboratory qualification studies and the suitability tested to ensure that residual disinfection solution does not alter the integrity of the sample results.   100级房间（或隔离器）中进行成品或中间产品的无菌测试，应当使用消毒剂及经验证的接触时间对样品表面进行消毒。在样品检验前消毒剂喷雾或溶液如果穿透待检品包装或容器密封系统进入或覆盖在包装外则会阻止或杀灭充满培养基的环境，从而阻碍产品中微生物生长。例如，在包装消毒步骤前产品取样遭到污染，那么过度的消毒将容易透过容器密封系统产生假阴性的结果。样品包装容器消毒程序应当包含在实验室确认研究工作以及方法适用性监测中，确保残留的消毒剂不会改变样品微生物结果的完整性。   Because microbiological results are not captured in a chromatograph, as with HPLC or GC analysis, opportunities to directly review any anomalies between the raw and recorded data are few. One way to verify the accuracy of the test data documentation is to set aside all available microbial petri plates, broths，identification strips，etc., in a secure area or refrigerator just after they have been removed from the incubator and analysis is complete. A careful review of the raw data against the analytical worksheets or LIMS data entries should reveal any discrepancies and provide an opportunity to catch any problems with either personnel training or equipment recording, and thus improve, laboratory procedures.   由于微生物测试不像高效液相和气相测试那样通过色谱仪进行分析，因此能够直接审核出原始数据和所记录的数据之间异常情况的几率很小。确认测试数据准确性的一种方法是，在待检验工作完成后将用于测试的微生物培养皿、营养肉汤、鉴定条等从培养箱中取出，将它们放在安全区域或冰箱中，然后仔细核对检验工作表或LIMS数据。可以发现任何不符之处，并提供机会发现人员培训或设备记录方面的任何问题，从而改进实验室程序。   Another technique may prove effective in checking data from samples stored in refrigerators and freezers. The ordinal test vessels from completed analyses of all positive samples (in-process or finished product bioburden test results, sterility test results, water testing, EM, personnel, etc.) usually contain microbial isolates. Conducting an inventory of the stored samples can provide a starting point to determine if there are differences between what was actually recovered during sample analysis and the data recorded on worksheets or in the LIMS. While taking inventory of samples stored in freezers and refrigerators，be sure to note the product name, lot numbers, tracking numbers, microbial identification, etc., marked on the exterior of test tubes and petri dishes. This collected information will provide a list that can later be compared to the results recorded on batch records, worksheets, and LIMS. The information gathered from reviewing the stored sample plates and tubes can indicate whether a company’SOPs are effective or if the analysts are following procedures correctly when there are objectionable microorganisms that need to be investigated. If the number and types of actual microbial isolates found from the stored samples do not match those in the company’s quality control records, these data integrity issues maybe due to intentional misreporting of the data.   另一个可能证明有效的手段是检查储存在冰箱或超低温冰箱中的样品数据。保存所有原始的阳性测试结果的测试容器（中间产品或成品的微生物限度测试、无菌检查、水测试，环境测试，人员表面微生物测试等）通常都包含微生物菌落。检查被存储的阳性样品可以从最开始的数据检查检验记录单或LIMS系统与实际数值之间是否存在差异。如果要将阳性样品储存起来，应当确保在试管壁和培养皿外表面注明产品名称，批号，追踪号，微生物鉴定结果等信息。收集这些信息是用于提供足够信息用于与检验记录结果，批记录，表格和LIMS系统中的数据核对。对于这些搜集信息的审核是为了考察公司SOP规定的是否完善以及分析人员是否正确按照规程执行操作，对有任何可疑菌落生长发起必要的调查。如果被储存起来的冻存管，培养皿等观察到的信息与QC报告的数据不一致，则显示有故意为之的嫌疑而造成微生物数据完整性不符合要求。   5.1.3 Worksheet Review   检验记录的审核   Microbiology data often relies on less automated recordkeeping that can be manipulated, often without clear traceability to the original record, when recording forms and worksheets are not well controlled or reconciled. In order to cross-check and confirm data in worksheets, entries in linked documentation should be reviewed (e.g., logbooks of equipment, batch records). Control over the issuance of original paper records, forms or worksheets should provide mechanisms that allow the original records to be distinguished from copies and reconciled. An expert microbiologist recommends reviewing analytical worksheets carefully during an inspection for the following indications of possible falsification: (a) lab reports containing more tests than can reasonably be run per day or per week or by an individual analyst; (b) backlogged sample worksheets reviewed and approved as a batch at the end of the month; (c) worksheets completed by personnel not present in the lab the day of the tests; (d) worksheets reviewed and approved by some other than an authorized reviewer/approver; (e) changes in handwriting signatures; (f) changes in reporting format, e.g” <1 CFU being replaced by zero; (g) results with counts reported as merely meeting specifications; (h) a change in frequency of OOS result reporting; (i) perfectly filled-out worksheets, indicating a risk that they are a second copy; (j) use of media before it passes growth promotion; (k) missing worksheets or worksheets out of numerical order; (l) discrepancies between investigation reports and worksheets; and (m) results from a single batch reported twice (31)   微生物数据记录通常无自动记录装置依靠手工操作，如果记录本和台账未良好受控或管理，这样通常会导致很难可清晰追溯至原始记录数据。为了能够对记录表格中原始数据进行交叉审核，在录入相关文件记录时应当对其录入的数据进行审核（例如：设备使用日志，批记录）。实验室原始记录和表格应当有受控发放机制，以区分原始记录和复印件及受控件的区别。一位微生物学专家建议在审核实验员检验记录时应当关注以下可能出现的数据造假：（a）实验室某个实验员出具的检验记录数量远远超过理论上一个单独实验员一天或一周正常的数量；（b）月底开始集中审核和批准积压的一批次样品检验记录；（c）检测记录上签字的人员在检测当天并未在岗；（d）未授权的记录审核员；（e）书写或签名处有涂改；（f）报告形式变更，例如用0替代＜1cfu；（g）报告的结果几乎接近标准值；（h）OOS结果报告周期的变更；（i）刚好填满一整张记录，可能显示为重新誊抄的记录；（j）在培养基适用性检查之前就使用该培养基；（k）记录表格丢失或表格编号不连续；（l）记录表和调查报告之间存在差异；以及（m）同一批次报告2次结论（31）。   5.1.4 Contract Laboratories   合同实验室   Companies that do not have the laboratory space, specialized equipment, or capacity to conduct highvolume microbiological analyses may outsource this type of testing to contract laboratories. In such cases, the company is responsible for thoroughly investigating the reputation and quality control performance of any new or unfamiliar contract laboratory with which they engage in a quality agreement. Before signing an agreement, it is recommended that a manufacture(a) confirm that the contract laboratory complies with CGMP regulations regarding laboratory qualifications and can perform the required compendial testing for microbial recovery and bacterial endotoxins (30,32— 34); (b) conduct an on-site audit; and (c) submit a sample batch for testing to ensure the quality of testing and data integrity. The company may wish to submit a blinded challenge of test samples that should produce OOS results. Documentation that the laboratory management is qualified to interpret microbiological sample results is required, especially if those individuals will be used as subject matter experts for data review or in the event the lab management needs to conduct a laboratory investigation for questionable sample results.   若公司内部没有实验场地，特殊实验设备或缺乏足够的实验能力完成大量的微生物测试，则会寻求将微生物测试外包给合同实验室。这种情况下，寻求合同实验室的公司将负责彻底调查任何新的或不熟悉的合同实验室的名声和质量控制能力，并将审计结论签署在双方质量协议中。在签署协议之前，建议委托方（a）确认合同实验室严格遵守cGMP法规中实验室确认的要求以及能按照药典要求进行微生物测试和细菌内毒素测试（30,32-34）；（b）进行现场审计；（c）以及提供给合同实验室一些样品进行测试判断其测试水平和数据完整性。委托方甚至可以提供一个阳性挑战对照样品以判断其是否能产生OOS结果。实验室管理文件应确认在有需求的时候能够解释样品结果，特别是如果这些人将被用作数据审查的主题专家或实验室管理人员需要对可疑样品结果进行实验室调查的情况下。   Reports written based on samples tested by a contract laboratory should include complete analytical worksheets from the analysis performed (e.g., raw material, in-process product intermediates, EM sampies, water, finished products). AJ1 of the original data (including calculations) and all of the positive (growth promotion results) and negative controls used during analysis should be included in the response report as they are equally important to the final product analysis results. The full investigation report of any OOS must also be provided. A certificate of analysis with a summary of the analytical results does not provide all of the information needed to verify sample test data or to satisfy a regulatory audit.   合同实验室的实验报告应当基于合同实验室测试员执行的全部分析测试表格（例如：原料，中间产品，环境监测结果，水，成品）。所有测试过程中的原始数据（包括计算公式或计算过程）以及所有阳性对照（促生长测试结果）、阴性对照应当包含在最终报告中，这些数据与最终产品的微生物实验结果报告一样重要。如果出现OOS结果，完整的OOS调查报告也应当提供。仅有一份样品的检验报告及分析结果小结并不能完整的提供确认样品测试结论的信息或满足官方审计的要求。   Two laboratory areas that the sponsor or owner company must carefully oversee are: (a) monitoring of source data, including electronic data and metadata, at the contract site, and (b) establishing contractual provisions to ensure the contract site does not delay, deny, refuse, or limit the ability of the sponsor or owner company to inspect their source data. As observed by experienced consultants, some contract sites only allow the sponsor company to review printouts, refusing access to their source electronic data and metadata. Risk-based reviews of critical source data and metadata and reconciliation of source data with reported information is essential for a meaningful data integrity audit.   涉及有两个实验室场所的情况委托方必须特别关注的要点为：（a）检查合同实验室的原始数据，包括电子数据和元数据；（b）制定合同实验室数据管理规程确保合同实验室不会拖延、逃避、拒绝或限制委托方对其原始数据的审计。据一些专家顾问的审计发现，有些合同实验室只允许委托方检查其打印数据，拒绝其进入数据系统查看其原始电子数据或元数据。基于风险的关键源数据和元数据的审查以及源数据与报告信息的核对对于有意义的数据完整性审计至关重要。   If a company obtains services from more than one contract laboratory~~a common practice, particularly when special laboratory equipment or expertise is needed—it should thoroughly investigate the reputation and quality control performance of each laboratory engaged. Though rare, the possibility exists for a contract laboratory to falsify data or adjust reported analytical results for microbiologically compromised samples, threatening the company’s product and reputation as well as patient safety. Another potential risk is that each contract laboratory uses a different microbial identification platform (e.g., genotypic vs. phenotypic) that may generate different species names for the same isolate. This can make it difficult for investigation of any analytical OOS results to find the true potential source of product contamination with manufacturing EM isolates that may have used yet another type of microbial identification system.   如果一家公司同时选择多家合同实验室为其服务---这种情形很常见，特别是当需要特殊的实验室设备或专业技术进行测试时----应当彻底的对每一家合同实验室的知名度及质量控制水平进行评估审计。但也有可能，合同实验室会对存疑微生物测试样品伪造或调整报告的数据，这种情况会威胁委托方产品的声誉，同时危害病人的健康。另一个风险是不同的实验室运用不同的微生物鉴定平台（例如基因测试和表型测试）从而导致对同一个菌落的鉴定产生不同的鉴定种属名称结果。这种情况就会导致一旦发生生产环境测试OOS结果难以找到根本污染原因从而需要寻求另一种微生物鉴定系统。   Lending credibility to the concern about the honesty of private testing laboratories is a court case prosecuted by the U.S. Department of Justice for the FDA. In 2008, the FDA brought criminal charges against a private contract testing laboratory that was caught falsifying data for more than 350 sample worksheets. All 350 sample worksheets showed that the laboratory president performed all analytical steps at “ 1PM” each day. The worksheets indicated that the submitted nutritional supplement product samples were analyzed for yeast and mold. In most cases, the analytical results were made available on the same day the samples were received. The FDA investigators became suspicious of the results on the worksheets because the laboratory had none of the reagents, media, or equipment needed to perform these analyses. Tlie unsuspecting manufacturers that sent their products to this contract laboratory and failed to detect that these sample results were fraudulent (25)   对第三方分析实验室诚信度的关注来源于一次美国FDA司法部门的指控。2008年，FDA指控一家第三方分析实验室，有证据证明该实验室至少有超过350个记录表格造假。所有350份造假数据中均显示所有的测试步骤都由该实验室负责人于每一天的下午13:00中进行。记录显示这些表单是为测试营养补充剂产品的真菌和霉菌。大多数情况下，测试结果都是在接收样品的当天出具。FDA检察官对该实验室存疑是由于该实验室完全没有任何测试用的试剂，培养基，设备。而不知情的委托方根本不知道他们送检的样品检验结论完全是伪造的。   A challenge that some license holders or product owners may have when auditing contracted facilities is limited access to records or information that may reflect a widespread systemic problem at the contracted site and may go undetected. These challenges should be anticipated early in the business and quality agreement discussions to ensure sufficient access to the appropriate records, procedures, and systems during record audits, especially when dealing with OOS or data integrity events. A dual responsibility exists to ensure that all products will be produced and tested according to the regulatory requirements, especially when regulators see contracted facilities as an extension of a manufacturing operation.   关于合同实验室凸显的一个问题是委托方表示在他们对合同实验室进行审计时能够检查能够反映广泛的系统性问题的记录或信息的权限有限，这样的话，很可能问题就不会暴露。因此，这一问题应在双方达成委托检验的早期或签署质量协议讨论时尽早提出，以确保在记录审计期间能够充分的查阅相应的记录，规程和系统，特别是在解决任何OOS或数据完整性事件时。委托检验过程中，委托方和受托方双方都有责任确保所有产品的生产和检验都将符合现行法规要求，特别是检察官提出合同实验室也作为工厂延伸检查一部分。   5.1.5 Equipment and Instrument Review   设备和仪器检查   If not handled correctly, a varietyof equipment and instruments used in a pharmaceutical microbiology laboratory (e.g., incubators, water baths, pH meters, EM devices, temperature monitoring systems, microscopes) can become the source of a data integrity problem; some of these will be discussed in greater detail later in this section. The general concern is that the results derived from the analyses conducted with equipment and instruments not calibrated or not used properly may generate erroneous results. The three examples that follow describe how the review of quality control shortcoming of laboratory or manufacturing equipment can be critical in detecting erroneous results. Lack of quality control could result in overlooking microbiological conditions or contamination that may compromise manufacturing monitoring and valid product testing.   如果没有正确使用，则由于微生物检测设备（例如培养箱，水浴锅，pH计，环境监测设备，温度监控系统，显微镜等）的可变性可能产生数据完整性问题的源头。这些问题会在本节稍后进行详细阐述。通常认为微生物检验的设备未经过计量或未正确使用是产生错误数据的原因。接下来阐述的三个例子便是介绍为什么实验室或生产设备的质量控制缺陷审核对发现错误结果至关重要。缺乏质量控制可能导致忽视微生物条件或污染, 可能会危及生产监测和产品检验的有效性。   √When water activity devices are not calibrated or maintained properly, the validity of the analytical data derived comes into question. For example, if the sensing head of a device is not cleaned properly before a subsequent product sample is placed in the chamber, regardless of the reason for the error, the impact on the analytical value of the final product for release or testing may be significantly altered. Whether or not the post-assay data had been manipulated or unintentionally changed, the analytical results may have been compromised, possibly affecting product disposition (good or bad). A full investigation is required to determine the potential impact.   水系统相关检测设备未经过计量或有效维护,则经过这些检测设备测量出的结果可能会存在问题.例如,如果检测设备探头完成一批次样品检测后未有效清洁即进行下一批次样品检测,暂且不考虑出错的原因,这样的操作会对测试结果有重要的影响.无论这个下一批次含量测试结果数据是否被人为或非人为的修改,这个分析测试结果都会损害,或可能影响产品的最终处置(无论结果是好是坏).需要发起一个全面的调查判定潜在的影响.   √ A laboratory’s automated microbial identification system (e.g., Vitek II", Biolog) generally stores a data file of all the microbial isolates it has identified from all product or environmental microbial isolates tested on it. These identification platforms should have tracking numbers that trace the origin of the microbial isolate back to the product batch record, LIMS, OOS, investigation reports, or similar data. These microbial identification records can be used in an audit to verify the accuracy of the associated paper trail should there be any question of transparency.   实验室自动微生物鉴定系统(例如VitekIT,Biolog)产生的所有来自产品或环境监测的微生物实验的菌落鉴定文件数据将被储存。这些鉴定平台应当有序列编号以方便追溯到被鉴定菌种来源于哪一产品批号，来源于LIMS系统哪个数据，OOS，调查报告或相似的数据。这些微生物鉴定记录能被用于审计以确认相关的记录数据问题的透明性。   √ The dates for when instruments are unavailabledue to needed repair or calibration are usually maintained in a logbook or digital spreadsheet* Comparing the records of nonftmctioning or absent laboratory equipment that might have been used for analysis against the microbiological worksheets may detect potential data integrity problems. The risk for using uncalibrated or nonfunctioning laboratory equipment is greatest when the workload becomes heavy or when tight deadlines are imposed. On the surface, this may appear to be a quality control issue; however, it can become a data integrity concern depending on the level of disrepair or calibration of the instrument used and the ultimate impact on the analytical results.   当设备仪器由于需要维修或计量不可用的情况通常应当记录在设备日志或电子表格中。比较实验室仪器故障时段的数据与微生物实验室纸质记录表格可能会发现一些潜在的数据完整性问题。当工作量增大或检验周期很紧的情况下，实验室设备故障或未被计量而产生的风险将大大增大。表面上这只是一个质量控制的问题，事实上，基于设备故障和未计量设备的使用和对最终分析结果的影响程度，这个问题可能是影响数据完整性的一部分因素。   5.2 Testing for Environmental Monitoring, Sterility, and Bacterial Endotoxins   环境监测、无菌监测和细菌内毒素监测   5.2.1 Environmental Monitoring Equipment   环境监测设备   Some of the equipment used most fiequently in the pharmaceutical microbiology laboratory and manufacturing Facility are environmental monitoring devices used to collect volumes of air or surface contact From specific rooms (e.g., ISO—5 manufacturing areas used for filling sterile products or sterility test suite used to analyze finished drug products labeled as sterile). The neglect or mishandling of EM devices can cause misleading or incorrect analytical results about the microbial presence(bioburden) within the room being monitored.   制药企业的微生物实验室和生产设施中最常用的一类设备是环境监测设备，用于收集来自特定的房间空气或表面接触的环境（例如，用于无菌产品分装的ISO-5级生产区域，或用于无菌制剂无菌检测的设施）。对于环境监测设备装置的疏忽或误操作可能导致对被监测房间内微生物存在情况（生物负载）的误导或给出不正确的分析结果。   5.2.2 Air Sampling Devices   空气采样装置   In general, an air sampling device requires that a petri dish or a specifically designed plastic strip containing nutrient agar be placed into the device; the instrument is then turned on to draw a specific volume of air into contact with the agar surface. At the end of the air sampling dwell time, the petri dish or strip along with any captured microorganisms are removed from the device and placed in an incubator. Subsequently, the number ofgrowing microorganisms on the media is counted. Examples of where data integrity problems can occur with air sampling devices include:   通常，空气采样装置需要将培养皿或含有营养琼脂的特殊设计的塑料条放入装置中; 然后启动仪器以吸取特定体积的空气与琼脂表面接触。 在空气采样结束后，将带有捕获到的微生物的培养皿或条带从装置中取出并置于培养箱中。 随后，对培养基上生长的微生物进行计数。 空气采样设备可能出现数据完整性问题的示例包括：   Visualization Of the Petrl Dish.Signs of agar impingement and/or colonies growing within the agar medium, caused by the directed air being drawn into the device onto the agar surface,should be visible, ensuring that the agar petri dish was placed into the air handling device during the initial room setup or at another time, when appropriate.   培养皿的可视化：由于空气直接吸入设备撞击在琼脂表面引起的琼脂培养皿内的印记和/或菌落的生长迹象应该是可见的，确保在初始房间内操作时或其他合理的时间内，琼脂培养皿已放入空气采集装置中。   Clogged samplers. Over time, the narrow opening of the EM sampler that Funnels the air directly toward the turning petri dish will become clogged, essentially preventing the full volume ofair to impinge on the agar surface. Though a maintenance or operational anomaly; this directly impacts the data integrity of the microbial evaluation of critical worlt areas. Routine cleaning of the device in that narrow opening is essential to ensure data integrity For this equipment.   采样器堵塞。 随着时间的推移，环境取样装置中将空气直接传到培养皿的EM取样器的狭窄开口将变得堵塞，最终不能使得全部的空气撞击琼脂表面。 虽然是维护或操作异常引起，这将直接影响了关键工作区微生物评估的数据完整性。对设备的狭窄的开口定期进行清洁对于确保数据完整性至关重要   Quality control Logbook. As with other laboratory equipment, recording ifor when any ofthe air sampling devices are defectiveoroutofservice in thequalitycontrollogbookisessential.Thedates during which a piece of equipment is listed as “out of service” would not be found in the sample batch records.   质量控制日志:与其他实验室设备一样，所有有缺陷或故障的空气采样设备，必须在质量控制日志中进行记录。故障的设备不应出现在样品批记录中。   Settle Plates.When settle plates are used to passively monitor the manufacturing or laboratory environment, plates located near surfacts exposed to UV light during monitoring may not be suitable.UV light may kill viable airbome microbial contamination, resulting in questionable results. Settle plates exposed too long become desiccate, and unwrapped settle plates exposed within an isolator during vaporised hydrogen peroxide (VHP)or other decontamination steps become unable on support growth. In either case, potentially viable airbome microbial contamination may not be recovered during exposure.   沉降菌检测皿。当用沉降菌监测皿用于被动监测生产或实验室环境时，沉降碟不应放置于在监测过程中会暴露在紫外光下的位置。紫外线可能会杀死空气中的活微生物，导致结果不可信。暴露太久的沉降碟会变得干燥，并且在蒸发的过氧化氢（VHP）或其他去污步骤期间暴露在隔离器内的未包装的沉降碟会变得不适宜微生物生长。 在以上任何一种情况下，暴露期间的微生物污染可能无法真实还原。   Final counts. Any manipulation or adjustment of final microbial counts without sound scientific justification, such as subtracting colony counts near the edge ofthe petri plates or the numberof bacterial colonies appearing at the beginning time period For air sampling plates, is a concernfor data integrity (Figure 5.2.2-1).   微生物最终计数。所有人为的或最终微生物计数时的没有正当理由的调整，比如减去培养皿边缘附近的菌落计数或减去浮游菌采样时在开始时间段出现的细菌菌落数量，是数据完整性的关注点（图5.2.2-1）。 Figure 5.2J-1 Misreading of EM Plates   图5.2.2-1 环境监测计划错误读取   5.2.3 Surface and Personnel Monitoring Equipment   表面和人员监测设备   Disinfectants should not be applied to a work surface or on an operator's gloved hands prior to sampling those surfaces using replicate organism detection and counting (RODAC) plates. Application of a disinfectant to a surface designated to be monitored before the test is performed will change the results ofthe assay, making the surface appear to be free ofmicroorganisms or producing a lower bioburden.   在使用表面接触皿（RODAC）对工作表面或操作员手套进行采样前，不应对其进行消毒。在进行测试之前将把要监测的表面进行消毒将改变测定结果，使表面判定为没有微生物或产生较低的生物负载。   The manipulation of procedural requirements can obscure the reality of a contaminated facility or a person’s poor aseptic technique resulting in the manufacture of a drug product under conditions that adulterate the finished product and impact patient health. Data integrity detection may need to occur at microbiological recovery stages long before an analyst opens the incubator door and transfers the tubes and petri dishes for colonies to be counted, otherwise, the data may be compromised and meaningless. In some cases, the surface monitoring is not done correctly, e.g., too short a contact time or performed improperly (for instance, only the finger tips are sampled but the fingers are not correctly rolled over the agar, or fingers are overlapping the plate).   篡改程序的要求的将掩盖设施已被污染或人员的无菌操作失败的事实，制造出了假药并影响病人的健康。对数据完整性的检查应该在微生物回收阶段（即取样阶段）开始，而不是分析人员打开培养箱门，将培养管和培养皿转移并计数时才进行，否则数据将变得不可信且毫无意义。在某些情况下，表面监测没有正确执行，例如，接触时间太短或操作不当（例如，仅对手指尖进行采样，但手指没有正确地压在琼脂平板上，或者手指重叠。   5.2.4 Review of Sterilily Test   无菌检测审核   The laboratory methods employed to conduct a sterility test on finished sterile pharmaceutical drug products or medical devices are delineated in international pharmacopeial sterility test chapters such as USP <71> (30). There are two analytical choices: (a) passing the aqueous product through brane filter to collect the potential microorganisms recovered from its contents, followed by rinsing to remove residual antimicrobial preservative(s) contained in the product; or (b) adding the product directly into the enrichment media. The essential step before employing these validated compendialprocedures is to run a product suitability test in the presence of a set of required QC microorganismsto ensure that, under the test conditions,the product ingredients do not interfere with the recoveryand growth of these challenge QC microorganisms. The design of the suitability test establishes the procedures used during routine testing to provide scientific proof that the method will work under routine product-testing conditions.   无菌制剂或无菌医疗器械的实验室方法在国际药典无菌试验章节中有描述，如USP <71>（30），包含两种可选的分析方法：（a）用隔膜过滤器过滤产品溶液，从产品中回收潜在微生物，并随后漂洗以去除产品中所含的残留抗菌防腐剂；或（b）将产品直接添加到培养基中。最关键的步骤是在应用这些药典的方法前需要进行产品适用性实验确保其适应测试条件，产品的成分不会干扰到QC微生物挑战菌的生长。适用性实验的设计为产品方法在日常产品检测条件下使用提供了科学的依据。   When the sterility test is properly performed as described in the compendia, and after determiningthe product suitability with that method, data integrity problems can result if sample handling and product testing is significantly altered or mismanaged from the approved SOPs.The following are risksfor data integrity breaches in sterility testing:   当确认了无菌测试是按照药典要求执行，并且也按照要求通过了产品的适用性测试时，那么数据完整性问题通常体现在样品处理和产品测试方法被明显改变或与SOP要求不一致。以下是无菌测试中数据完整性破坏的风险：   Before product containers (units) are transferred to the sterility testing suite or HEPA-filtered laminar flow hood, the outside of each container is usually subjected to a disinfection procedure:sprayed or wiped with a sporicidal solution to decontaminate the exterior of the container, preventing surface contamination from being transferred into the testing area. Any overexposure to the disinfection solution may penetrate the package or container and kill potential productcontaminants before the sterility test is performed. This critical analytical step for certain producttypes (e.g., medical devices, combination products), if performed incorrectly, can result in a “ falsenegative” sterility test that would allow the release of a potentially contaminated lot of parenteralor implantable products. This data integrity problem may not be obvious on a laboratory work-sheet or detected in LIMS because the destruction of potential microbial results would have curred before the product was tested.   在将产品容器（单元）转移到无菌测试套件或HEPA过滤层流罩之前，每个容器的外部通常需要按照消毒程序进行消毒：用杀孢子剂喷洒或擦拭净化容器外部，防止表面污染物转移到测试区域。在进行无菌测试之前，任何对消毒溶液的过度暴露都可能渗透包装或容器并杀死潜在的产品污染物。对于某些产品类型（例如，医疗器械，组合产品）的这种关键分析步骤，如果执行不正确，可能导致“假阴性”无菌测试，并最终导致受污染的肠外或可植入的批次产品放行。这种数据完整性问题可能不能明显体现在实验室工作记录上，或通过LIMS系统检测到，因为潜在微生物结果的破坏已经在产品测试之前发生。   In a pharmaceutical microbiology laboratory, the preparation of the media and reagents used during sterility testing is an important step, which should be performed carefully to avoid mistakes.For instance, when analyzing products by the “direct inoculation” approach, the accidental use of a lower volume of test medium may create an inhibitory growth condition due to a higher concentration of the product or preservatives in the test medium compared to those levels proven acceptable during the suitability test . The early detection of low volume test tubes can be missed when the tubes are located in the middle of the holding rack.   在制药企业的微生物实验室中，无菌检测过程中使用的培养基和试剂的制备是一个重要的步骤，应该谨慎进行以避免出错。例如，在通过“直接接种”方法分析产品时，意外使用较少体积的测试培养基会抑制微生物的生长，因为与产品适应性测试条相比，该测试培养基中的产品浓度或防腐剂的浓度更高。当试管位于支架中间时，小体积的试管可能会错过早期检测。   One observation that has been documented on several FDA Form 483s when the membrane filter is tranferred to the enrichment broth, or when a powdered product is added directlyto the medium,and the membrane filter adheres to the top of the inner portion of the test tube above the liquid medium (Figure 5.2 . 4 - 1) .This problem has been traced to the four-inch forceps rather than the recommended 10- or 12-inch forceps that may not allow the transfer of the membrane filter into the broth.A similar mishap may occur when hygroscopicpowder products adhere to the neck of the inner tube, preventing the product from entering theenrichment media and allowing potential product contaminants to grow(Figure 5.2 . 4 - 2) .A third example, documented during the inspections of contract laboratories performing sterility testing on sterile medical devices, involves ineffectively cutting up long catheters or complex devices, preventing the product from contacting the enrichment broth. The portions of the productprotruding from the liquid brothmay be the regions of the device thatcontain the indigenousmicroorganisms (Figure 5.2 . 4 - 3) .   在多个FDA483表中记录的一个不符合项是，当膜过滤器转移到富营养培养基中，或者当粉末产品直接添加到培养基中时，膜过滤器粘附在在液体介质上方，测试管内的顶部（见图5.2.4-1）。这个问题原因在于测试过程使用的是四英寸镊子，而不是推荐的10英寸或12英寸镊子，使得膜过滤器不可能转移到肉汤中。当吸湿性粉末产品粘附在内管颈部时，可能会发生类似的问题，产品不能进入富营养培养基使潜在的产品污染物得以生长（图5.2.4-2）。第三个记录在案的例子是，在检查合同实验室对无菌医疗器械进行无菌测试时发现，长导管或复杂装置没有被有效地切割，阻止了产品接触浓缩肉汤。而从液体肉汤中突出的产品部分可能正是装置中含有固有微生物的区域（图5.2.4-3）。   These examples reflect testing conditions that may have gone unrecorded or unobserved by a reviewing supervisor; however,their occurrence may result in false negative data to the final recorded laboratory result. The prevention of problems in the pharmaceutical microbiology laboratory that affect data integrity may need to be fixed at stages before observational or data entry is recorded on a worksheet or in LIMS.   这些例子反映了审核主管可能未审核到的，未记录或未观察到的测试条件; 然而，它们的发生可能导致最终假阴性实验室结果。制药微生物实验室中影响数据完整性问题的预防可能需要在观察或数据记录在工作表或LIMS之前的各个阶段进行修复。   5.2 .5 Bacterial Endotoxin Testing   细菌内毒素测试   The bacterial endotoxin test (BET) described in compendial test chapters is an enzymatic assay that employs the reagent limulus amebocyte lysate (LAL) along with a vial of controlled standard endotoxin(CSE).The CSE serves as as artificial, laboratory-derived source of bacterial endotoxin that can be used to perform the suitability test (inhibition/enhancement test); it is also employed as a positive control to ensure that incubation conditions during product testing have not been compromised.The BET can also be conducted using a gel clot procedure or with the use of a spectrophotometer, employing a modified version of the LAL reagent(34).Each of these reagents requires careful handling andstorage, the directions for which are clearly stated in the products package instructions.   在药典测试章节中描述的细菌内毒素测试（BET）是一种酶测定法，使用试剂鲎阿米巴样细胞裂解物（LAL）以及一小瓶受控标准内毒素（CSE）.CSE作为人工从实验室中获取的细菌内毒素，可用于进行适用性试验（抑制/增强试验）; 它也可用作阳性对照，以确保产品测试期间的孵育条件不受影响。BET测试也可以使用改良形式的LAL试剂（34）并采用凝胶凝块程序或使用分光光度计执行。上述每一种需要小心处理和储存，在产品包装说明中有明确说明。   To minimize potential data integrity problems with the final laboratory test results，when performing the BET method for finished product testing of pharmaceutical drugs or medical devices,there are a few critical parameters that need to be understood, satisfied, and documented:   为了最大程度降低最终实验室测试结果的潜在数据完整性问题风险，在对药品或医疗器械的成品测试执行BET检测时，应对一些关键参数进行理解，且这些参数应得到满足并记录：   During product container storage (refrigeratedor at roomtemperature), when vial or ampoulesamples containing aqueous solutions are delivered to the laboratory,the indigenous bacterialendotoxin may form micelles or attach to the glass or rubber stopper surfaces and move out of thesolution. Unless the laboratory SOP for performing the BET on aqueous products includes a mixing step prior to removing the test aliquot, the detectable level of endotoxin determined by the assay will be underestimated and may inaccurately portray the products endotoxin unit value withinspecifications. Consequently, the final analytical results could be impacted, even when using avalidated and properly controlled assay, depending on the outcome of a formal investigation (35) 。   在产品在容器中存储期间（冷藏或在室温下），当将含有水溶液的小瓶或安瓿样品递送到实验室时，固有的细菌内毒素可形成胶束或附着到玻璃/橡胶塞表面，并从溶液中移除。 除非水性产品BET实验室检验SOP中包含了在移除测试等分试样之前的应包含混合步骤，否则测试所得的内毒素水平结果将被低估，最终错误的将产品内毒素单位值描述为符合标准。尽管使用一种已经过验证，且已受用的测试方法，最终仍会影响产品分析结果，具体取决于正式调查的结果（35）   An allowable amount of bacterial endotoxin (as defined in thepharmacopeial product mono graphs) may be present in a drug product based on its dosage, route of administration, and totalamount of endotoxin delivered, e.g., within a one -hour dosing for an injectable product. A well -designed BET method requires that the maximum valid dilution (MVD) be calculated before theassay is performed. If it is not calculated precisely, with the formula variables and the units used inthe MVD formula, an incorrect value may result.The data integrity problem may occur if the cal -cukted MVD value is higher than the true number. That would make it possible to run the assay at a higher product dilution, creating a testing situation where unacceptable levels of endotoxinmaybe present in the product but released without detection.Since endotoxin have dire patient consequences, this is a crucial data integrity matter .   产品中可能存在可允许量的细菌内毒素（如药学产品单图中所定义），基于其剂量，给药途径和存在于药物产品中的内毒素的总量，例如注射剂，一小时的剂量。一个有效的设计较好的BET方法要求在进行检测之前计算最大有效稀释度（MVD）。如果未精确计算，在使用公式变量和MVD公式中使用的单位，可能会导致不正确的值。 如果计算的MVD值高于真实数字，则可能会出现数据完整性问题。 检测可能在较高的产品稀释度下执行，导致产品中实际存在不可接受的内毒素水平，但最终产品未能检测到而被放行。由于内毒素会对病人造成严重的后果，因此这是一个至关重要的数据完整性问题。   In a U.S. Department of Justice criminal prosecution brought by the FDA involving a fraudulentand contaminated hormone product, the drug manufacturer submitted the bacterial endotoxin analytical data as part of the laboratory evidence. The FDA laboratory detected levels of bacterial endotoxin in the vials used for product rehydration above USP specifications.The drug manufactuer employed a private laboratory to test the same suspect lot for bacterial endotoxin andreported the levels of bacterial endotoxin detected as below the compendial specification. Laboratory records showed that the analyst for the private lab manipulated the product dilution to avoiddetecting the actual level of endotoxin in the product.The analytical report had been tailored togive a finished result with an endotoxin level below the USP alowable limit. The FDA laboratory,however, performed the full range of product dilutions toavoid missing the actual endotoxincon-tamination level.The FDA’s test data proved the contract laboratory had manipulated the dilution to produce false results for their client. This example illustrates the need to both understand the analysis procedure and recognize how it can be manipulated.   在美国食品和药物管理局提及的关于欺诈和污染的激素产品的美国司法部刑事诉讼中，药品制造商提交了细菌内毒素分析数据作为实验室证据的一部分。FDA实验室检测到用于产品补液的小瓶中的细菌内毒素水平高于USP要求。该药物制造商雇用了一个私人实验室来测试同一个可疑批次的细菌内毒素，并报告检测到的细菌内毒素水平低于药典要求。实验室记录显示，私人实验室的分析员操纵产品稀释以避免检测产品中的实际内毒素水平。分析报告经过编造，使最终结果的内毒素水平低于USP要求限度。然而，FDA实验室进行了全方位的产品稀释，以避免错判实际的内毒素污染水平。美国食品和药物管理局的测试数据证明，合同实验室已经操纵了稀释液，为客户产生了错误的结果。这个例子说明了理解分析程序和识别它是如何被操纵的重要性。