内审发现检测证据不足怎么补

结论：先判断缺的是数据、记录还是关联，三者补法完全不同

内审发现检测证据不足，第一反应常常是「补一份报告」。但缺失的类型不同，补救方式差别很大，补错了等于白做。

缺数据是指该测的没测，真的没有这个数据。缺记录是指测了但过程记录不完整。缺关联是指数据和记录都有，但无法证明它们对应关系。

三者里，缺数据可以补测；缺记录要看缺的是什么，有些补不了；缺关联多数可以通过整理建立。

三类缺失的处理路径

缺数据怎么补

这一类相对直接：确定缺的项目，安排检测。需要判断的是：

用什么样品测。 如果原来的产品批次还有留样，用留样测能覆盖当时的状态；如果没有留样，只能用现在的产品测，那么数据反映的是当前状态，对历史批次的覆盖需要另行论证。

要不要覆盖历史批次。 如果缺失的项目涉及安全性，可能需要评估已上市产品的风险，而不只是补一份当前的数据。

这一点是内审整改中容易回避的。 补一份新报告放进文件夹，看起来整改完成了，但历史产品的符合性问题并没有解决。

缺记录怎么处理

这一类要分情况：

可以补的： 方法依据的说明、设备信息的补充说明、判定规则的明确。这些属于对已有数据的解释性补充，只要能查证就可以补写，但要注明是何时补充的及依据。

不可以补的： 原始测量数据、当时的环境条件记录、操作人员的实时记录。这些是当时才能产生的，事后补写就是编造。

面对不可补的缺失，正确做法是承认并评估影响： 这份数据的可信度受影响到什么程度、是否需要重测、对产品符合性结论有什么影响。然后在整改措施中确保今后不再发生。

事后补填原始记录是严重问题。 它把一个记录不完整的问题，升级成了记录真实性的问题。审核员对补填的痕迹相当敏感——笔迹、墨色、日期逻辑都可能露出破绽。

缺关联怎么建立

这一类的处理是整理工作。做法是：

从现有记录里找出能建立对应关系的线索——报告上的样品编号、委托单上的批号、生产记录上的日期。把这些线索串起来，形成对应关系表。

如果线索不足以建立一一对应，要如实说明关联的不确定性，而不是牵强指定。

整改措施应当是从源头改进：今后在委托时就写明批号，报告上体现样品标识，使关联天然存在，而不是事后整理。

整改的有效性怎么证明

内审整改的有效性验证，需要新产生的证据。常见的做法：

抽查新记录。 整改措施实施后产生的记录，按同样的标准抽查，看问题是否消除。

流程验证。 走一遍完整流程，确认新的控制点确实起作用。

时间跨度。 有效性验证不能只看整改后第一批，建议观察一段时间，因为有些措施短期有效长期松懈。

只提交整改后的文件修订版，不构成有效性证据。 文件改了不代表执行改了。

内审自身的改进

如果内审发现了大量检测证据问题，值得反思内审本身：为什么这些问题积累到现在才被发现。

可能的原因包括：内审的检查方式偏重文件存在性而非内容质量；内审员对检测专业不熟悉，看不出问题；内审频次不足；或者发现了问题但整改跟踪不到位。

改进内审的检查方式，比单纯整改发现的问题更有价值。 具体可以做的是：内审时引入追溯演练的方法、让懂检测的人参与、把重复出现的问题列入重点检查项。

整改优先级的排序

内审通常一次发现多项问题，全部同时整改不现实。排序建议按两个维度：对产品安全有效性的影响，以及问题的系统性程度。

影响安全的优先；涉及系统性缺陷（比如整个流程缺失）的优先于个别记录缺失。按发现顺序逐条整改是常见但低效的做法，因为它不区分轻重。

与外审的衔接

内审发现的问题如果在外审之前没有整改完，要准备好说明：问题已识别、原因已分析、措施在实施中、预计完成时间。

已识别并在处理中的问题，性质远好于未被发现的问题。 所以内审记录本身是有价值的证据，不要因为担心暴露问题而弱化内审记录——外审员看到严格的内审记录，对体系的信心反而会提高。

证据补齐的时间压力

内审整改通常有时限，而补测需要时间。如果时限内完不成，正确做法是在整改计划里如实写明分阶段完成的安排，而不是先写「已完成」再慢慢补。

整改记录与实际状态不符，比整改未完成严重得多。 前者是诚信问题，后者只是进度问题。

从内审问题反推流程缺陷

每一个检测证据不足的问题，背后通常有流程上的原因：某个环节没有规定谁负责、某份记录没有对应的表单、某个交接点没有确认机制。

整改时除了补证据，更值得做的是找出这个流程缺陷并补上。否则同类问题会在下一批产品上重复出现，而每次都靠事后补救，成本累积起来远高于一次性把流程理顺。

我们的做法

有委托方因为内审或外审发现证据不足，来补做检测。这类委托我们会多问一句：这次补测是为了填补当前的文件缺口，还是需要覆盖历史产品？

两者的方案不同。填补文件缺口，用当前产品测即可；覆盖历史产品，需要考虑留样、批次代表性等问题，有时还需要做一些论证工作。委托方把目的说清楚，方案才不会做偏。

如果你在处理内审发现的检测证据问题，可以把不符合项描述和现有记录发过来一起看，或者直接联系：132 4819 8029。能力范围见服务介绍，送检要求见送检要求，更多内容见知识库。

English version

Conclusion. When an internal audit finds testing evidence inadequate, the reflex is to commission another report. But the remedy depends on what is actually missing, and the wrong remedy accomplishes nothing. Missing data means the test was never performed. Missing records means it was performed but the process was not documented. Missing linkage means both data and records exist but cannot be shown to correspond. Missing data can be generated by testing. Missing records may or may not be recoverable. Missing linkage can usually be rebuilt through organisation.

Addressing missing data. This is the straightforward case: identify the gap and arrange testing. Two judgements arise. Which samples to use: retained samples from the original batch cover the historical condition, while current product reflects only the present state and its applicability to historical batches needs separate argument. And whether historical batches must be covered at all: if the missing item concerns safety, the risk of marketed product may need assessing rather than simply producing current data. This is the point most often avoided in audit remediation. Filing a new report makes the finding look closed while the conformity question for product already sold remains open.

Addressing missing records. Distinguish what can and cannot be recovered. Explanatory supplements can be added where they can be substantiated: the method basis, equipment details, the acceptance rule. Annotate when they were added and on what basis. What cannot be recovered are raw measurement data, contemporaneous environmental conditions and the operator's real-time record. These could only have been created at the time, and writing them afterwards is fabrication.

Where the gap is unrecoverable, the correct response is to acknowledge it and assess the impact: how far the credibility of that dataset is affected, whether retesting is needed, and what it means for the conformity conclusion. Then ensure the corrective action prevents recurrence. Back-filling raw records is a serious matter, escalating an incomplete-records problem into a records-integrity problem. Auditors are alert to the signs: handwriting, ink, and the internal logic of dates.

Rebuilding linkage. This is organisational work. Find the threads in existing records that establish correspondence, such as sample identifiers on reports, batch numbers on request forms and dates on production records, and assemble them into a correspondence table. Where the threads are insufficient to establish a unique correspondence, state the uncertainty honestly rather than asserting a match. The corrective action should address the source: record batch numbers when commissioning and show sample identification on reports, so the linkage exists naturally rather than being reconstructed.

Demonstrating effectiveness. Effectiveness verification requires newly generated evidence. Sample the records produced after the action and examine them against the same criteria. Walk through the complete process to confirm the new control points operate. And allow time, since some measures work initially and relax later, so observe over a period rather than examining only the first batch. Submitting a revised procedure is not evidence of effectiveness: changing a document does not demonstrate changed practice.

Improving the audit itself. If an internal audit uncovers a large volume of testing evidence problems, it is worth asking why they accumulated undetected. Possible reasons include auditing that checks whether documents exist rather than whether they say anything; auditors without the technical background to see the problems; insufficient frequency; or findings raised but not followed through. Improving how internal audit examines testing is worth more than remediating the individual findings. Introduce traceability rehearsals into the audit, involve someone who understands testing, and add recurring problems to a focused check list.

How we handle it. Clients come to us to fill gaps identified in internal or external audits. We ask one additional question: is the purpose to close a documentation gap now, or to cover product already placed on the market? The approaches differ. Closing a documentation gap can use current product. Covering historical product raises questions of retained samples and batch representativeness, and sometimes requires supporting argument. Stating the purpose keeps the approach on target.

Send us the finding and your existing records and we will work through it. Phone or WeChat: +86 132 4819 8029.