产品分类界定不清时，检测清单怎么先行

结论：分类影响的是证据形式，不是产品本身的性能

分类界定没出结果，检测就完全停着，这是很多项目浪费时间的地方。实际上分类决定的是证据由谁出、要不要第三方、审查有多严，而不是产品的物理性能会变。

产品的强度就是那么大，电气安全就是那个状态，这些数据无论最终按哪一类管理都成立。所以有相当一部分工作可以先做。

哪些可以先做

前五行可以立即启动。 这些数据在分类确定后，如果需要正式报告，可以用同样的样品和方法重新出具，或者部分沿用，时间上已经跑在前面了。

分类为什么会不清楚

几种典型情形：

产品形态较新，现有分类目录里没有明确对应的条目。

功能跨界，产品同时具备两类产品的特征，比如既有治疗功能又有监测功能。

预期用途表述影响分类，同一个硬件，用途写法不同可能落到不同类别。

与已上市产品的归类不一致，市面上类似产品有的按这一类、有的按那一类。

最后一种最让人困惑，但它往往说明这些产品的预期用途或技术特征确实存在差异，值得逐个去看它们的注册信息而不是只看产品外观。

并行推进的具体做法

建议分三条线同时走：

第一条线：走分类界定程序。 通过正式渠道申请，拿明确答复。这条线周期最长，所以要最先启动。

第二条线：按较严口径准备检测。 假设是较高类别，按那个口径列项目清单。如果最终界定为较低类别，多做的部分是冗余而非浪费——资料只会更充分。

第三条线：完善预期用途与技术要求。 因为用途表述会影响分类，这份文件定稿本身也是界定申请的输入。

三条线里，第二条的投入要控制。可以先做那些无论如何都要做的基础项目，把分类敏感的项目（比如需要特定资质机构出具的）放到界定结果出来之后。

样品怎么安排

并行推进时样品安排有个技巧：先用试制件做摸底，正式检验留到分类确定后用定型产品做。

这样处理有两个好处：一是摸底阶段发现的设计问题还来得及改；二是正式检验用的是定型产品，代表性没有争议。

需要注意的是样品数量要提前规划。摸底消耗一批，正式检验还要一批，两批的需求要一起算，避免临时补件。

预算怎么分配

分类不确定意味着预算存在不确定性。建议这样处理：

把检测项目分成三组——无论哪一类都要做的、只有高类别才要做的、只有特定情形才要做的。第一组的预算是确定支出，先锁定；第二组按较严口径预留；第三组等结果明确再定。

这样预算的不确定性被限制在第二三组，而不是整体待定。 内部审批时也更容易说清楚。

界定结果出来之后

拿到明确答复后，要做三件事：

核对已完成的检测数据在新的口径下是否可用；补齐分类要求而尚未做的项目；调整技术要求和申报资料的结构。

如果界定结果比预期严格，前面按较严口径准备的工作就派上用场了；如果比预期宽松，可以砍掉一部分计划中的项目，省下的预算用在别处。

常见的判断误区

误区一：看别人怎么分类就跟着分。 别人的产品预期用途未必和你的一样，跟着分可能跟错。

误区二：按分类的低标准准备，赌界定结果宽松。 赌输了代价很大，而赌赢省下的成本有限，风险回报不划算。

误区三：等界定结果完全出来再动。 时间成本往往比多做几个项目的成本高得多。

与注册顾问的分工

分类界定的主导方通常是注册顾问或企业的法规人员，实验室在这件事上不是决策者。但实验室能提供两类有用的输入：

一是同类产品的检测口径。 做过相近产品的实验室知道那些产品实际按什么标准测、报告做成什么形式，这可以反推它们的管理类别。

二是标准适用性的判断。 某些标准的适用范围本身就与产品分类相关，把标准适用性理清楚，分类的线索也会清楚一些。

所以分类讨论时把实验室拉进来是有价值的，虽然最终判断不由实验室做。

一个容易被忽略的成本

分类不确定期间，除了检测，还有几项工作会受影响：包装和标签的设计（不同类别的标识要求不同）、说明书的编写、质量体系文件的覆盖范围。

这些工作如果都等分类确定再启动，会集中压在后面。建议同样按较严口径先做框架，细节等结果出来再调整。

界定申请材料里的检测信息

提交界定申请时，通常需要说明产品的结构、原理和预期用途。这里可以适当引用已有的检测数据来支撑描述，比如输出参数的实测值、材料的成分信息。

有实测数据支撑的描述比纯文字更可信，也减少了因描述含糊导致界定结论偏离预期的风险。这也是前面建议先做基础检测的另一个理由。

我们的做法

遇到分类未定的委托，我们会把项目清单标注成三组：与分类无关的、按高类别需要的、以及可能需要特定资质机构出具的。委托方据此决定先做哪些。

这个标注不收费，但它能让委托方在不确定的情况下把钱花在确定的地方。 实际操作中，第一组通常已经覆盖了大部分工作量，先做完这部分，界定结果出来后补的往往不多。

如果你的产品分类还在界定中，可以把产品资料和预期用途描述发过来，我们帮你把清单分组，或者直接联系：132 4819 8029。能力范围见服务介绍，流程见检测流程，其他问题见常见问题。

English version

Conclusion. Waiting for a classification determination before starting any testing is a common waste of time. Classification governs the form the evidence must take, who may issue it and how strictly it will be reviewed. It does not change the product's physical performance. Strength is what it is, electrical safety is what it is, and those data hold regardless of the eventual class. A substantial part of the work can therefore start immediately.

What can start now. Performance characterisation, electrical safety pre-verification, EMC pre-verification, basic material and biological data, and structural strength and durability are all independent of classification, because what applies is determined by the product's own attributes. What does depend on classification is formal registration testing, where the issuing body's accreditation requirements differ, and the form and holder of the report.

Why classification is sometimes unclear. The product form may be new, with no clear entry in the catalogue. Functions may straddle categories, combining therapeutic and monitoring features. The wording of the intended use may itself drive classification, so the same hardware described differently could fall either way. Or comparable marketed products may be classified inconsistently, which usually indicates genuine differences in their intended use or technical characteristics and is worth investigating product by product rather than by appearance.

Running three tracks in parallel. First, apply for a formal determination; this has the longest lead time so start it first. Second, prepare testing to the stricter reading, since if the outcome is the lower class the extra work is redundancy rather than waste. Third, finalise the intended use statement and technical requirements, since the wording feeds into the determination application itself.

Control the investment on the second track. Run the baseline items that will be needed either way, and hold back the classification-sensitive items, such as those requiring a specifically accredited body, until the determination arrives.

Sample planning. Use pilot units for preliminary testing and reserve definitive product for formal testing once classification is settled. Design problems found in the preliminary phase can still be fixed, and formal testing then uses product whose representativeness is not in question. Plan quantities for both phases together to avoid an urgent reorder.

Budgeting. Split the items into three groups: needed regardless of class, needed only for the higher class, and needed only in specific circumstances. Commit the first group's budget now, reserve for the second at the stricter reading, and leave the third pending. Uncertainty is then confined to the second and third groups rather than the whole programme, which is also easier to explain in internal approval.

After the determination. Check whether completed data remain usable under the confirmed basis, add whatever the class requires that has not been done, and adjust the structure of the technical requirements and dossier accordingly.

Common misjudgements. Following how other companies classify, when their intended use may differ from yours. Preparing to the lower standard and gambling on a favourable outcome, where the downside is large and the saving modest. And waiting for full certainty before moving, where the time cost usually exceeds the cost of a few extra test items.

How we handle it. Where classification is unsettled we mark the item list in three groups: independent of class, required for the higher class, and potentially requiring a specifically accredited body. Clients then decide what to start. The first group usually covers most of the work, so completing it first leaves relatively little to add once the determination arrives.

Send us the product information and intended use statement and we will group the list. Phone or WeChat: +86 132 4819 8029.