结论：分型定错，漏电流限值和隔离要求都会跟着错
医用电气设备的电气安全要求中，「应用部分」的识别和分型是一个基础判断。 它决定了多项要求的严格程度——特别是漏电流限值和隔离要求。
分型判错的后果是实实在在的返工： 按较宽松的类型设计和测试，实际应当按较严的类型，那么漏电流限值不满足、隔离距离不够、甚至电路结构要改。这些问题发现在测试阶段还好，发现在注册审查阶段就更被动。
所以这个判断应当在设计之初就做对，而不是等送检时由实验室来提。
应用部分的识别
先要识别哪些是应用部分。基本思路是：正常使用中，必然要与患者发生物理接触，才能实现设备功能的部分。
识别时容易出问题的地方：
看起来不接触但实际接触的。 设备外壳的某些部位在正常使用中会被患者触及。
通过介质接触的。 不直接接触但通过液体、气体与患者连通。
可触及但非功能必需的。 这部分是否算应用部分，要按具体情况判断。
附件与连接件。 电缆、传感器、夹具，往往是应用部分的一部分。
识别要覆盖正常使用中的所有接触情形，包括患者可能触及的部位。
分型的依据
| 类型 | 大致适用情形 |
|---|---|
| 较基本的类型 | 一般接触，不用于直接心脏应用 |
| 中间类型 | 需要较高防护等级的接触 |
| 较严格类型 | 用于直接心脏应用的场合 |
分型的依据是使用场景，不是设备的复杂程度。 一个结构简单的设备，如果用于直接心脏应用，就要按较严的类型。
判定的核心问题是：这个应用部分与患者的接触方式和位置是什么？ 用途说明中的预期使用场景是判定的主要依据。
判定的步骤
第一步，写清楚预期用途。 设备用于什么、用在哪里、患者是什么状态。
第二步，列出所有与患者接触的部分。 包括附件。
第三步，对每个接触部分判断接触性质。 接触哪个部位、是否涉及心脏。
第四步，确定类型。 按接触性质对应。
第五步，核对设计是否满足该类型的要求。 漏电流、隔离、标识。
第六步，在资料中说明判定依据。
第一步是基础。 预期用途写得含糊，分型就没有明确依据。很多分型争议的根源在于预期用途描述不清。
判错的具体后果
漏电流限值不满足。 不同类型的限值不同，按宽松的设计可能超出严格类型的限值。这往往需要改电源设计或增加隔离。
隔离要求不够。 爬电距离、电气间隙、绝缘配合，都与类型相关。改这些通常涉及结构改动。
防除颤要求。 某些情形下需要防除颤能力，未考虑就要重新设计。
标识错误。 设备上的符号要与类型对应，标错需要重新制版。
说明书内容。 相关的说明和警告要匹配。
测试项目遗漏。 某些类型要求的额外试验没做。
改电路结构是成本最大的后果，因为它可能涉及重新设计、重新打样、重新做全套测试。
送检前的自查
建议在送检前自查以下内容：
预期用途的表述是否明确。 能否据此判断接触场景。
应用部分清单是否完整。 包括所有附件。
每个应用部分的分型及依据是否写明。
设备标识是否与分型一致。 符号、说明书中的描述。
漏电流是否按对应类型的限值自测过。 自测数据不必精确，但能发现明显超标。
隔离设计是否核对过。 关键位置的距离和绝缘配置。
多个应用部分的类型是否可能不同。 一台设备的不同应用部分可以是不同类型。
最后一项容易被忽略。 企业往往给整机定一个类型，但实际上不同的应用部分可能需要不同的类型。
与其他判断的关联
分型不是孤立的判断，它与几项内容相关：
与设备分类的关联。 防电击类型（与供电方式相关）是另一个维度，两者要分开判断但共同影响要求。
与使用环境的关联。 使用场所影响某些要求。
与风险管理的关联。 分型判断的依据应当与风险文件中的使用场景描述一致。
与说明书的关联。 说明书中的使用限制应当与分型的假设一致。
这几处的一致性建议在送检前核对一遍。 不一致会在审查中被指出，而修改说明书比修改设计容易，前提是设计本身按较严的假设做了。
附件与选配件的处理
应用部分的分型不只涉及主机，还涉及附件：
附件是否构成应用部分。 与患者接触的附件通常是。
不同附件的类型可能不同。 一台主机配多种附件时，各附件的接触方式可能不同。
第三方附件。 如果允许使用第三方附件，要说明兼容要求和限制。
附件的标识。 附件本身也可能需要相应的标识。
送检时附件要一起送。 只测主机不测附件，覆盖不完整。
「第三方附件」是实务中的风险点。 如果设备接口通用，用户可能接入未经验证的附件，而附件的特性会影响整体的安全性能。资料中应当明确说明允许使用的附件范围。
设计阶段的核对清单
建议在设计冻结前核对：
预期用途是否写清楚了接触场景；应用部分清单是否完整（含附件）；每个应用部分的类型是否确定并有依据；对应类型的漏电流限值是否在设计中体现；隔离设计是否满足对应要求；标识方案是否与类型一致；说明书的相关内容是否匹配；以及风险文件中的使用场景描述是否与分型依据一致。
这份清单在设计阶段过一遍，比在测试阶段发现问题要省事得多。 设计阶段改的是图纸，测试阶段改的是实物。
我们的做法
承接医用电气设备的委托时，我们会在测试开始前先确认应用部分的识别和分型。如果我们的理解与委托方不一致，会先讨论清楚再开始。
这个确认动作看起来多余，但它避免的是最坏的情况——按一个类型测完，审查时被认定应当按另一个类型，全套重来。
如果你在准备送检，不确定应用部分怎么识别和分型，可以把预期用途和结构说明发过来一起判断，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，更多内容见知识库。
English version
Conclusion. Within the electrical safety requirements for medical electrical equipment, identifying and classifying the applied part is a foundational judgement. It determines the stringency of several requirements, particularly leakage current limits and isolation requirements. Getting it wrong means real rework: design and test to a less demanding type when a more demanding one applies, and leakage limits are not met, isolation distances are inadequate, and circuit architecture may need changing. Discovering this during testing is bad enough; discovering it during registration review is worse. The judgement therefore belongs at the start of design rather than being raised by the laboratory at submission.
Identifying applied parts. First establish which parts they are. The basic idea is a part that must necessarily come into physical contact with the patient during normal use for the equipment to perform its function. Difficulties arise with parts that appear not to contact but do, where areas of the enclosure are touched in normal use; parts contacting through a medium, not touching directly but connected to the patient through liquid or gas; accessible parts not necessary to function, whose status depends on the circumstances; and accessories and connecting parts, since cables, sensors and clamps often form part of the applied part. Identification must cover every contact situation in normal use, including parts the patient may touch.
Basis for classification. More basic types cover general contact not intended for direct cardiac application. Intermediate types cover contact needing a higher degree of protection. More stringent types cover direct cardiac application. Classification rests on the use scenario rather than the equipment's complexity: a structurally simple device intended for direct cardiac application must follow the stringent type. The central question is the manner and location of contact between the applied part and the patient, and the intended use description is the principal basis.
Steps. Write the intended use clearly, covering what the equipment does, where it is used and the patient's condition. List every part contacting the patient, including accessories. Judge the nature of contact for each, covering which site and whether the heart is involved. Determine the type accordingly. Check the design against that type's requirements for leakage, isolation and marking. And state the basis of the classification in the documentation. The first step is foundational: vague intended use leaves classification without a clear basis, and most classification disputes trace back to unclear intended use descriptions.
Consequences of misclassification. Leakage current limits may not be met, since limits differ by type and a design to a lenient limit may exceed a stringent one, often requiring changes to the power supply design or added isolation. Isolation may be inadequate, since creepage, clearance and insulation coordination all depend on type, and changing them usually means structural change. Defibrillation-proof requirements may apply in some circumstances, requiring redesign if not anticipated. Marking may be wrong, since symbols must match the type and correcting them means new artwork. Instruction content must match. And test items may have been omitted where a type requires additional testing. Changing circuit architecture is the costliest consequence, since it may entail redesign, new prototypes and a complete retest.
Pre-submission self-check. Check whether the intended use statement is explicit enough to determine the contact scenario. Check whether the applied part list is complete, including all accessories. Check whether the classification and its basis are stated for each applied part. Check whether equipment marking matches the classification, in symbols and in the instructions. Check whether leakage has been measured in-house against the corresponding limits; the measurements need not be precise but will reveal clear excursions. Check whether isolation design has been reviewed at the critical locations for distance and insulation arrangement. And check whether different applied parts might require different types, since one device's applied parts need not share a classification. The last is easily overlooked, with manufacturers often assigning one type to the whole device when different applied parts may need different ones.
Related judgements. Classification does not stand alone. It relates to the equipment's protection class, which concerns the supply arrangement and is a separate dimension judged independently while jointly determining requirements. It relates to the environment of use, which affects certain requirements. It relates to risk management, since the basis of classification should agree with the use scenarios described in the risk file. And it relates to the instructions, whose stated limitations should agree with the assumptions behind the classification. Check consistency across these before submission: inconsistency is raised during review, and amending instructions is easier than amending a design, provided the design was built to the stricter assumption.
How we handle it. For medical electrical equipment we confirm the identification and classification of applied parts before testing starts, and where our understanding differs from the client's we resolve it first. The step looks redundant but it avoids the worst case, which is completing testing to one type and being told during review that another applies, requiring the whole programme again.
Send us the intended use and construction description and we will work through the classification. Phone or WeChat: +86 132 4819 8029.