欧盟GMP附录8《起始物料与包装材料的取样》中要求对每一个容器中的起始物料都进行取样鉴别：

The identity o f a complete batch o f starting materials can normally only be ensured if individual samples are taken from all the containers and an identity test performed on each sample...
通常 ，只有对每一个包装容器取样，并逐个样品进行鉴别后，才能确保整批起始物料鉴别无误。...

一批物料通常包括很多桶/袋，若逐个容器取样鉴别则工作量巨大，加拿大药监局给出了一种减少检验数量的方案，通过灵活的分组检验实现降本增效。这种方案被ECA（欧盟符合性学会）称赞为世界上最先进的建议。

该方法翻译如下：

Q.2 Does every individual container of a raw material need to be sampled for identification (ID) purposes regardless of the number of containers of the same lot available or are composite samples acceptable provided they are obtained from a maximum of 10 containers?

是否每一桶/袋中的原辅料都需要取样进行鉴别，不管一批物料里面有多少桶/袋？是否接受从不超过10桶/袋中取得的混合样品进行检测？

 

A.2 For human drugs, according to Interpretation 6.1 under C.02.009 Raw Material Testing, each container of a lot of a raw material must be tested for the identity of its contents. Therefore, each container of all raw materials, including excipients and active pharmaceutical ingredients (API), must be opened and sampled. Then, 2 options are available:

回答：对于人用药来说，根据C.02.009原辅料检验6.1的说明，同一批次的每桶/袋原辅料都必须经过检验以确定其内容物的身份。因此，必须打开并对所有原辅料进行取样（包括每桶/袋辅料和活性药物成分（API））。但是，接下来，有两个方案可供选择：

1、To test every sample for ID using a discriminating method (it is not mandatory to perform all ID tests in the specifications, for example United States Pharmacopoeia (USP), but the test must be specific).

对每一个样品的身份进行鉴别（不要求执行标准（如美国药典）中的所有鉴别实验，但必须指定）。

 

2、 If the raw material can be tested for potency, the other option is to mix and pool individual samples taken from each containers in a composite sample but without exceeding 10 individual samples in a composite. A specific ID test is then performed on each composite and, in addition, a potency test is performed to assure the mass balance of the composite. (In such cases, an equal quantity of each individual sample in the composite must be weighed to ensure that the mass balance is representative.)

如果可以测试该原辅料的含量，另一种选择是对每桶/袋原辅料中取得的样品进行分组并混合，每组不超过10个独立样品。先对混合样进行鉴别试验，然后，进行含量检验。（在这种情况下，每个样品必须称量相同重量形成混合样以确保具备代表性。）

As an example, say 72 containers of the same lot of a raw material are received. Each and all containers must be opened and a sample taken from each container. After that, the first option is to test each sample for ID (which implies 72 ID tests). The second option is to combine equal quantities of those individual samples in a way that the number of samples in any composite does not exceed 10 and test those composites for ID and potency. In this case, the easiest way to combine those samples would be 8 composites of 9 individual samples. For a given composite, a potency result of 88.8 % or so would indicate that one of the containers does not contain the right material as each individual sample contributes 1/9 or 11.11% of the total mass of the composite (similarly a result of 77,7 % would indicate 2 containers with the wrong material). In such case, each container selected for this particular composite would have to be tested for ID to pinpoint the one (or more) containers with the wrong material.

举例，接收一批原料有72桶/袋。每桶/袋必须打开取样。然后，按照第一个选项是检验每个样本的身份（这意味着72个鉴定试验）。按照第二个选择则是，将这些样品等量混合，并确保每一个混合样中的样品数量不超过10个，检验这些混合样的鉴别和含量。即，组合这72个样品的最简单方法是，将每9个独立样品混合，形成8个混合样。对于每一个混合样，若检验含量结果为88.8％左右则将表明其中有一桶/袋原料不正确，因为每个独立的样品贡献了混合样总质量的1/9或11.11％（同样地，若结果为77 .7％表示有2桶/袋的原料有问题）。

方法一	方法二
72次鉴别试验	8次鉴别试验+8次含量检验，鉴别试验节省89%

However, the use of a composite sample to establish the ID of a raw material cannot be used when the potency limits are too wide or, similarly, when the precision of the assay method is not sufficient to properly establish the mass balance.

但是，当含量范围太宽时，或者当检验方法的精确度不足以适当地建立质量平衡时，就不能使用混合样来确定原料的ID。